REVOLUTION UNENDING

The CERI series in

Comparative Politics and
International Studies
Series editor CHRISTOPHE JAFFRELOT
This series consists of translations of noteworthy
publications in the social sciences emanating from the
foremost French research centre in international studies,
the Paris-based Centre dEtudes et de Recherches
Internationales (CERI), part of Sciences Po and associated
with the CNRS (Centre National de la Recherche
Scientifique).
The focus of the series is the transformation of politics
and society by international and domestic factorsglobalisation, migration and the post-bipolar balance of power on
the one hand, and societal dynamics, ethnicity and religion
on the other. States are more permeable to external
influence than ever before and this phenomenon is
accelerating processes of social and political change the
world over. In seeking to understand and interpret these
transformations, this series give priority to social trends
from below as much as the interventions of state and
non-state actors.
Founded in 1952, CERI has fifty full-time fellows
drawn from different disciplines conducting research on
comparative political analysis, international relations,
regionalism, transnational flows, political sociology, political
economy and on individual states.

GILLES DORRONSORO

Revolution Unending
Afghanistan: 1979 to the Present
TRANSLATED FROM THE FRENCH BY JOHN KING

HURST & COMPANY, LONDON

in association with the Centre dEtudes et de
Recherches Internationales, Paris

First published in 2000 as La rvolution afghane.

Des communistes aux tlebn by Karthala, Paris
This revised and updated edition first published
in the United Kingdom by C. Hurst & Co. (Publishers) Ltd,
38 King Street, London WC2E 8JZ
2005 by Gilles Dorronsoro
All rights reserved.
Printed in India
The right of Gilles Dorronsoro to be identified
as the author of this publication is asserted by him
in accordance with the Copyright, Designs and
Patents Act 1988.
A Cataloguing-in-Publication data record for this
book is available from the British Library.
ISBNs
1850656835 casebound
1850657033 paperback

Contents
Foreword
Chronology
Glossary

page xi
xv
xviii

Introduction
Reference points
A tribal revolt?
An ethnic war?
A blend of religion and politics?
Issues and hypotheses

1
5
7
9
16
19

Part I. THE ORIGINS OF THE

AFGHAN REVOLUTION
1. The Sociogenesis of the Afghan State
The role of imperialism
Neo-patrimonialism
The legitimacy of the state
Challenges to the state

23
24
25
33
40

2. From Mobilisation to Revolution

The mobilisation of internal resources
International funding
The politicisation of the educated class
The failure of the parliamentary regime
Authoritarian mobilisation and repression under Daud
Utopia and violence: the communists in power

61
61
63
65
76
80
85

Part II. MOBILISATIONS

3. The Commanders
The revolt
Jihad

93
93
105
vii

viii

Contents
Qowm and commanders
The social profile of the commanders
Models of organisation
Finance

4. The Jihadi Parties

The formation of parties
A typology of the parties
The parties and the refugees

108
111
123
129
137
137
149
169

Part III. THE DYNAMICS

OF CONFRONTATION
5. The Kabul Regime
The party under Karmal
The three strategic dimensions
Preparations for the retreat
From the Proximity talks to the Geneva accords
The post-withdrawal survival of the regime

173
173
175
191
198
201

6. The Guerrillas
The professionalisation of the fighters
The evolution of the political map
Inter-party cooperation
The mujahidin after the withdrawal

207
207
212
223
227

Part IV. THE TALIBAN

7. Competition and the Impetus towards Monopoly
(19922001)
The collapse of the state
The regionalisation of the forces (19921995)
Towards monopolisation

235
237
240
250

8. The Ethnicisation of the Conflict

Ethnicisation as a strategy
Ethnicisation as an unintended consequence
The challenge to the ethnic hierarchy

257
258
265
268

9. The Clerical State

The party functionaries

272
273

Contents
State structures
The social backlash: the puritan order and social resistance
The radicalisation of the Taliban

ix
278
284
301

Part V. THE AMERICAN INVASION AND

THE RETURN OF FRAGMENTATION
10. A Splendid Little War?
The strategy of the United States
The failure of the Pakistani solution
The fall of the Taliban

315
317
322
325

11. The Return to Political Fragmentation

The reconstruction of the state
The new guerrilla movement
Ethnicity, social groups and ideologies

329
331
338
342

12. The Policy of Pakistan

Trial of strength with the pro-Taliban movements
Commercial and human movements

346
347
352

Conclusion

354

Select Bibliography
Index

357
363

Maps
Regional context
Physical map
Administrative map
Simplified distribution of macro-ethnicities
The Kabul regime, 1989
Simplified political map, 1992
The regional distribution of political forces, 1996
Territory controlled by Masud, March 2000

xxi
xxii
xxiiii
14
186
247
248
249

Tables
Macro-ethnic groups in Afghanistan (1978)
Political parties before 1978
Principal Sunni parties
Principal Shiite parties

15
72
151
151

Foreword
This book on the Afghan war emerged from a sequence of research
visits undertaken from 1988 onwards. In contrast to an ethnological
approach, I opted to travel from one province to another, seldom
staying more than a few weeks in the same place. A drawback of this
approach was that it restricted my knowledge of any particular region, although the periods I spent in Herat were more substantial,
but it did give me a basis for comparison between different local situations. Perhaps more important, I was able to escape from the stifling routine of daily life in a village and, bit by bit, to discover this
wonderful country.
Travelling in Afghanistan during a war, often alone, was not always easy, particularly with a linguistic ability which was, at the outset, limited; and throughout with restricted financial resources. Each
trip, even though it lasted only a few weeks, became a battle against
physical fatigue and the consequent lowering of spirits. The task of
observing guerrilla warfare necessitates a Spartan existence, with long
periods of inactivity interspersed with infrequent bursts of combat.
Being fired on in an ambush, then being shelled several times, I was
leftwhatever idea I might have had that I was immortalwith a
fear of being wounded possibly even greater than the fear of death.
These things serve to show that the researcher is not untouched
by the landscape in which he works and that much of the effort one
makes goes into the mastery of ones own emotions. The consciously
detached tone of the analyses which follow is no more than the calm
reconstruction of a sometimes far from calm reality.
Desperately clutching his notebooks, whichespecially towards
the end of the triphe is terrified of losing, the researcher himself
becomes the object of sociological interest on the part of those
whom he observes. The same questions are repeatedly put to him
xi

xii

Foreword

about his religion,1 his salary, his food, the existence and status of
Muslim communities in Europeall matters which enable Afghan
interlocutors to make their own estimations of the researcher. The
travellers host derives prestige from his presence, so that he is himself instrumentalised in the politics of local power. In particular he is
appealed to as an exterior or neutral witness regarding the claims
of one party or another, and is asked to make these claims known in
the outside world. A stranger is also viewed as a potential source of
aid for the community, which conditions attitudes towards him. Confusion between journalists, humanitarian aid workers and researchers brings with it an ambiguous situation. This is not the fault of the
Afghans alone: the journalists have often worked with NGOs (nongovernmental organisations), as have the researchers, with the risk
that the ground-rules governing each of these separate activities may
become confused.
In the villages the polite formalities are calculated to minimise the
potential disruption of which the traveller may be the cause. He will
be greeted in a room set aside for guests, he will have no contact
with the women of the household, and he is rarely allowed to move
about alone.2 The researcher will therefore be obliged to take any
excuse to go out, in order to widen his contacts and to maintain a
wider perspective on the situation. In the bazaars the visitor may stay
at the hotel, and thus be more free in his movements. In spite of
these circumstances, the Afghans on the whole showed much patience when I asked silly questions; walked on the table linen (since
one ate sitting on the ground); or made appalling faux pas, such as being understood to ask in a bazaar where there had just been a bloody
settling of scores where the treachery is (the words for treachery
and tailor are very similar). I was particularly startled by the freedom with which Afghans commented on the political situation.
Inaccurate information, on an issue of no direct personal concern,
was often to be taken as no more than a well-meant wish to tell the
1

In Islam, Christians are seen as people of the Book, who as such benefit from
protection, or at the very least enjoy a recognised status.
2 For an analysis of the rules of hospitality, see Julian Pitt-Rivers, Anthropologie de
lhonneur, Paris: Hachette, 1998, and in the Afghan context Encyclopaedia Iranica
IX (1), New York: Bibliotheca Iranica, 1998, pp. 53 ff.

Foreword

xiii

researcher what should be the case rather than what it was in fact;
this did not indicate any intention to deceive him.
Interviews were my primary source concerning local issues, which
were rarely covered by the press or the radio. My preferred informants were often people on the fringes of power.3 Those who had
been through the process of education before the war, often ill at
ease in the rural context, were especially valuable; as also were the
khans4 (notables) who had been pushed aside by political developments; as well as transport professionals, such as truck drivers and
merchants, who were well acquainted with the local political realities. There was a constant falling-short in perceptions of national
politics in comparison with local analyses, which were fascinating in
their precision and insight. Through the opportunities afforded to
him by his position of externality, a foreigner enjoys a privileged
status, especially if he is able to stay aloof from the stereotypes of
various groups. However, certain foreignerswhether journalists,
researchers or humanitarian aid workershave on occasion succumbed to the temptation to side with some particular political
group, and sometimes even to propagate accounts which are virtually mythological.
My first two journeys benefited from the presence of Stphane
Thiollier, who was a superb guide, translator and friend. The present
text owes much to his influence and is dedicated to him. I also offer
my thanks to Hlne Arnaud, Anne-Franoise Basquin and Alain
Besanon, who supervised the thesis on which this book is based; to
Amlie Blom, Philippe Bonhoure, Hamit Bozarslan, Pierre Centlivres, who agreed to read the first draft; and to Micheline CentlivresDemont, Grard Chaliand, Marie-Odile Clerc, Sylvie and Jean Dolbeault, Nathalie Fustier, Marc Gaborieau, Delphine Hery, Mirwais
Jlil, Bertrand Labaux, Cristina LHomme, Rgis Lansade, Chantal
Lobato, Sophie Mousset, Mathilde Pinon, Jean-Jos Puig, Nicolas
Rageau, Rmi Reymann and Alexandre Toumarkine. My research
was funded in part by assistance from Marc Gaborieau and Alexandre
3

It may be somewhat misleading to speak of a process of selection, since it was later

that I became aware of common characteristics among my informants. A degree
of correspondence between the social position of the researcher and that of the
informant was often a source of rapport.
4 Expressions in italics are defined in the glossary on pp. xvi ff.

xiv

Foreword

Popovics Groupe de recherche sur la transmission du savoir dans le

monde musulman priphrique based at EHESS, as well as by a grant
from the Goodbooks foundation and through a contract with DAS
(Dlgation aux affaires stratgiques). Nicolas Rageau kindly drew
the maps on the basis of information I provided. For obvious reasons I do not name those in Pakistan and Afghanistan who agreed to
answer my questions, but they should count themselves the recipients of my sincere thanks.
The Anglophone reader will find here a somewhat different version of the text from that which appeared in French in 2000. This
has been dictated both by the need for factual corrections, and because a reassessment was necessitated by the new phase of the civil
war which began with the American intervention.
Le Couderchet,
March 2004

G. D.

Chronology
1747
18185
18359
183942
184263
186378
187880
1880
1887
1893
1919
191929
192930
19303
1931
193373
195363
1964
1973

Foundation of Durrani Empire by Ahmad Shah. At its maximum, after the battle of Panipat in 1759, this stretched to the
River Indus.
Partition of Afghanistan into three separate principalities
(Kabul, Kandahar, Herat).
Dost Muhammad on the throne.
First Afghan-British war. The British puts Shah Shuja in power.
Dost Muhammad restored to power.
Sher Ali (a son of Dost Muhammad) comes to power after civil
conflict.
Second Afghan-British war. Britain puts on the throne Yaqub
Khan (Sher Alis son).
Beginning of reign of Abdul Rahman Khan (a grandson of
Dost Muhammad). Earliest foundations of the modern state.
Abdul Rahman unifies Afghanistan after bloody campaigns.
Northern frontier of Afghanistan defined.
Eastern frontier defined by the Durand Line.
Assassination of Habibullah. His son Amanullah succeeds. Third
AfghanBritish war leads to Afghan independence.
Reign of Amanullah who attempted to modernise Afghanistan.
First Constitution promulgated in 1923.
Rebellion against Amanullah, who goes into exile. Interlude of
government by self-styled Habibullah II (Bacha-yi Saqao).
Nadir Shah comes to power as king, installing a new dynasty
and a conservative regime.
Afghanistans second constitution comes into effect.
Zahir Shah on throne, after succeeding aged nineteen following
his fathers assassination.
Daud (Sardar Muhammad Daud Khan) Prime Minister.
Afghanistans first liberal constitution announced, with two
elected houses of parliament.
Coup by Daud establishing the Afghan republic. Ex-King Zahir
goes into exile in Rome.

xv

xvi
1978
19789
1979
1979
1986
1988
1989
1992
1994
1995
1996
1998
2001

2002

Chronology
(2728 April) Communist coup brings Taraki and Amin to
power. Daud and his family put to death.
Following the communist coup the countryside rises in rebellion; parties in exile formed in Peshawar.
(March) Rebellion in Herat when defecting army units fought
government troops, Soviet advisers arrive in April to assist the
communist government.
(27 December) Soviet invasion. Babrak Karmal installed in
power by the Soviets.
(4 May) Karmal resigns, replaced by Najibullah under policy of
National Reconciliation.
(April) Geneva Accords.
(January) The Soviet retreat proceeds according to plan.
(20 April) Fall of Najibullah regime. The mujahidin enter Kabul.
War between the victorious parties begins immediately.
(November) Kandahar captured by newly-formed Taliban
movement.
(5 September) Taliban takes Herat.
(27 September) Taliban takes Kabul; Najibullah executed.
Taliban takes Mazar-i Sharif.
(9 September) Masud murdered.
(11 September) Twin Towers of the World Trade Center in
New York destroyed by Al-Qaida-linked guerrillas flying hijacked aircraft.
(20 September) A council of Afghan ulema rules that Osama Bin
Laden should leave the country.
(5 October) The first contingent of US troops arrives in neighbouring Uzbekistan.
(7 October) Operation Enduring Freedom begins; US bombs
Taliban and Al-Qaida positions; US special forces assist Shura-yi
Nazar.
(9 November) Dostums forces and his allies take Mazar-i Sharif.
(12 November) Kabul falls to Shura-yi Nazar.
(25 November) Hundreds of Taliban prisoners killed at Mazar-i
Sharif.
(5 December) Schloss Petersberg agreements provide for setting
up of Afghanistan Interim Authority.
(9 December) Fall of last Taliban stronghold at Kandahar.
(2223 December) Afghanistan Interim Authority established.
Hamid Karzai becomes chairman.
(2122 January) Afghanistan Reconstruction Steering Group
holds international conference in Tokyo for aid donors to
Afghanistan.

Chronology

2003

2004

xvii

(28 January) Joint statement by President George W. Bush and

Hamid Karzai on new partnership between United States and
Afghanistan.
(11 June) Loya Jirga inaugurated by ex-King Zahir Shah.
(24 June) Government formed under presidency of Hamid
Karzai.
(30 June) United States kills forty-two Afghan civilians in bombing raids on villages.
(2223 April) President Karzai visits Pakistan.
(4 June) Major battle between Taliban and government troops
indicates strength of Taliban resurgence.
(11 August) NATO takes over command of ISAF.
(4 January) Loya Jirga concludes with adoption of new constitution, conferring strong presidential powers on Hamid Karzai.

Glossary
asabiyya
agha
akhund
alaqadari
alem
amir
amir al-muminin
arbab
barakat
bey/beg
chador
dari
dawa
daulat
deqhan
faqih
farsi
farsiwan
fatwa
fiqh
gholam bacha
ghund
hadith
hazrat
hajj
haji
hamsaya
hasht nafari
hijra
hukumat
ijaza

solidarity based on belonging to the same group

sir, a term of respect
theologian, master of a school
sub-district administered by an alaqadar
religious scholar or teacher (pl: ulema)
prince/ruler
commander of the faithful
village leader: also has the sense of khan in Hazarajat
blessedness or holiness, which brings personal benefits
major landowner
womans veil
Persian dialect spoken in Afghanistan
religious preaching
state, government
peasant, sharecropper
religious scholar
Persian language
Persian speaker
juridical opinion given by an alem (pl. fatavi)
Islamic jurisprudence
slave in the amirs administration
regiment
reported sayings of the Prophet Muhammad (pl. ahadith)
Excellency/Highness
pilgrimage to Mecca
one who undertakes the hajj
client
system of conscription where one man in eight is called
The Prophet Muhammads migration from Mecca to
Medina
government
certificate granted by an alem to a pupil, authorising him
to teach

xviii

Glossary
imam
imarat
jamiyat
jerib
jihad
jirga
jombesh
kafir
khan
khanaqah
khutba
khwaja
kuchi
kundak
Loya Jirga
madrasa
maktab
malang
malek
mamur
markaz
masjid
mawlawi
maulana
mehman khana
mellat
mir
mirab
mubaraz
mudarris
muhajir
muhallim
mujahid
Muslim
mulgerey
mullah
murid
namaz
namus

xix

one who leads the prayers; more generally the head of an

Islamic community
principality
society, party
one-fifth of a hectare
holy war
tribal council
movement
unbeliever, sinner
man of influence, notable
Sufi center for meeting and meditation
Friday prayers
master, a descendant of the Caliph Abu Bakr
nomad
battalion
Great Council: an institution from the monarchical
period, recently revived
religious school
Quranic school: in Afghanistan a state school
wandering mystic; beggar
proprietor; village chieftain chosen by the elders
government employee
centre; base used by mujahidin
mosque
spiritual master; alem
spiritual master
guest house
nation
equivalent to amir: a chief or lord
person responsible for water supplies
fighter
teacher in a madrasa
religious refugee (pl. muhajjirin): a Quranic expression
elementary teacher (pl. muhallimin) equivalent to Arabic
muallim
fighter in the holy war (pl. mujahidin)
Muslim (pl. Muslimin)
associate of a khan (Pushtun)
religious official of a small community
disciple of a pir
prayer
honor

xx
pushtunwali
pishkhedmat
pir
qanun
qarargah
qowm
qadi
rowhani
sarluchi
sayyed
shabname
shahadat
shah
shahid
shaikh
Shaikh ul-Islam
shariat
shahrwal
shura
silsila
tabligh
takfir
takhie khana
talib
tariqat
tawiz
ulema
uluswali
ushur
wakil
wali
waqf
watan
wilayat
wilayat-i faqih
zakat
zikr
ziyarat
zulm

Glossary
Pushtun tribal law
a particular rank of state employee under Abdul Rahman
Khan
master Sufi, a Sufi saint
non-religious law
military base used by mujahidin
a solidarity group (pl. aqwam)
Islamic judge
man of religion (senior cleric in Iran) (pl. rowhanihun)
literally bare-head: an opponent of Islam
descendant of the Prophet Muhammad (pl. sadat)
tract
Muslim profession of faith
sovereign, ruler
martyr, one who dies in the jihad (pl. shuhada)
man of religion, head of a brotherhood (pl. shuyukh)
title of a major religious figure
Islamic law
mayor
consultative council: an Islamic concept
chain of masters deriving their authority from a given pir
preaching, teaching
excommunication: declaring a person to be an apostate
place of retreat for Sufis; place of worship for Shiites
where the tragic stories of Hassan and Hussein are reenacted
student of religion (pl. tullab Arabic; taliban Persian)
Sufi order or brotherhood
talisman
plural of alem (religious scholars, referred to as a group)
district administered by a local official (uluswal)
an Islamic tax (a tithe)
deputy
provincial governor
religious foundation or endowment (pl. awqaf)
fatherland, nation
province
authority exercise by a faqih
an Islamic tax
Sufi practice consisting of the repetition of the name of
God (literally remembrance)
pilgrimage to the tomb of a saint
oppression

Regional context.

xxi

xxii
Physical map.

xxiii
Administrative map.

Introduction
In the 1970s nothing indicated that Afghanistan was soon to be
plunged into interminable war. The country had enjoyed an unexpected tourist boom, with more than 60,000 visitors in 1969, and
was the obligatory route between Iran and the Indian subcontinent.
Travel books and novels such as Les cavaliers by Joseph Kessel, translated into English and other western languages, had disseminated the
image of a kingdom with pristine lands where proud tribesmen
practised buzkashi and the vendetta.1 Even the coup of 1973 and the
establishment of a republic did not occasion undue concern either
inside or outside the country. The monarchy had been on a constitutional basis since 1964, and its transformation into a republic was
undertaken bloodlessly: King Zahir remained in Italy, from where
he did not return until the spring of 2002. The personality of President Daud, who was a cousin of the king and had already played a
role as Prime Minister between 1953 and 1963, conveyed the impression that the country would be able to summon up the energy it
had lacked.
In April 1978, however, international attention was focused on
the coup by the Hezb-i Demokratik-i Khalq-i Afghanistan (Afghan
Peoples Democratic Party), carried out with the assistance of the
Soviets. For a time there was a debate over whether the government
was truly communist or merely progressive, but the rapid breaking
off of ties with the western countries and the uprising in the countryside made this an increasingly academic question. Under the momentum of the revolution the new regime went on to conduct
bloody purges within its ranks, and in September 1979 Hafizullah
1

Joseph Kessel, Les cavaliers, Paris: Gallimard, 1967. The brilliant photographic
essays of Roland and Sabrina Michaud bring into play, in an almost mystical perspective, the notion of continuity between the past and the present: see Mmoire
de lAfghanistan, Paris: Chne, 1980.

Introduction

Amin rose to the leadership of the country after eliminating his predecessor Nur Muhammad Taraqi, who had been more moderate
and was more highly regarded in Moscow. Meanwhile, as the rebellious countryside became impossible to control, the regime appeared
on the point of collapse, until it was rescued by the arrival of Soviet
troops on 27 December 1979. Commando forces installed in power
a new leader, Babrak Karmal, after killing Hafizullah Amin. Against
all the evidence, the Soviet Union claimed that Amin hadthough
only orallyasked for their intervention. The Soviet contingent
rapidly settled down at about 100,000 men, and the process of integrating the country into the Soviet bloc gathered momentum.
In the months which followed the invasion, historical parallels
multiplied, but also contradicted each other. According to a proverb,
Experience is a light which shines only backwards. Could armed
Afghans overcome the invaders from the north, in the same way as
their ancestors wiped out the British expeditionary force in 1842, or
would Afghanistan end up after a few decades as a dismal Central
Asian colony? Some predicted failure for the Soviets, basing their
arguments on the disastrous outcome of the 19th-century British
campaigns, while others expected the rebels to be defeated, as they
had been in Central Asia in the 1920s when the Basmachis were
overcome by the Red Army.2
Whatever might be the case, western opinion was outraged at
such a blatant violation of the sovereignty of an independent country, and the invasion of Afghanistan signified the end of the period of
dtente. President Jimmy Carter, who said that in the space of a few
days he had learned much about the Soviets, launched a rearmament
programme which Ronald Reagan would in due course continue.3
He also took reprisal measures, such as the boycott of the Moscow
Olympics and halting grain exports, whichcontrary to the general
viewcaused considerably inconvenience to the Soviet Union. The
director of the CIA, William Casey, had the job of coordinating
assistance to freedom fighters, whether Afghans, Angolans or Nicaraguans. These were sometimes put in contact with each other, and
2

Hlne Carrre dEncausse, Les Sovitiques en Afghanistan. Un nouveau Cuba?,

Politique internationale 6, winter 1979/80.
3 The US administration, which had already begun to finance anti-governmental
movements, was in fact warned of the invasion when it was imminent and allowed
the Soviet Union to make the mistake of entering Afghanistan.

Introduction

soon realised that they had little in common except an enemy and a
provider of funds.
After the initial crisis, the Afghan war settled down for the long
haul. Over the years the world began to be familiar with the images
of the fighters, wearing their pakols (Nuristani caps) or turbans, with
a willingness to fight which was hard to explain in view of the trying
conditions of their lives and the disproportionate odds they faced.
Badly armed, deeply divided into competing groups, operating in an
informal style which took observers aback, the Afghan resistance
perhaps even because of all these factorscould not be overcome.
A number of heroic figures emerged, such as Masud, the young and
photogenic commander of Panjshir, a valley to the north of Kabul,
who displayed a strategic mastery in his resistance to the most massive Soviet onslaughts. These attacks, increasingly violent, created
hundreds of thousands of refugees each year: of 12 million Afghans,
3 million would soon have migrated to Pakistan and 2 million to
Iran, making up the largest refugee population of modern times.
In spite of the violence of the fighting, the resistance did not give
way, with the result that the war became too costly for the Soviets,
who in the summer of 1986 decided to withdraw. The Soviet Union, which had been in difficulties since the first oil crisis, had failed
to make economic adjustments, and this had implications for its external policies. For Gorbachev perestroika and glasnost were no substitute for western cooperation, and the settlement of the Afghan issue,
as an earnest of good faith, became a priority. It was also the first step
towards the abandonment of an empire which had finally become
too expensive, although Angolan oil financed the Cubans and Afghan
gas was bought cheaply. In addition, the pointless deaths of young
conscripts were viewed increasingly badly by Soviet public opinion.
Gorbachevs speech of 13 January 1988 was a fundamental break
with the Brezhnev doctrine: the Soviet Union announced its withdrawal from Afghanistan with no conditions as to who would constitute the successor government. This was the end of proletarian
internationalism. The communist bloc began to break up. In the
spring of 1988 the Geneva accords formalised the Soviet withdrawal,
which was completed by 15 January 1989. The accords also envisagedwith some hesitations on the American sidewhat was called
a positive symmetry, where the two superpowers would each remain free to arm their own allies.

Introduction

The United States hoped, in fact, to bring about the downfall of

the pro-Soviet regime within months of the withdrawal. The Americans may, have entertained the fantasy that the humiliation of Vietnam might be avenged by the evacuation of the Soviet ambassador
by helicopter from the rooftops of Kabul. Urgent in their pursuit of
victory, the United States and Pakistan adopted an ill-thought-out
and ill-executed strategy of frontal attack against Jalalabad, a town
whose capture presented particular difficulties. The failure of the resistance fighters, who were unprepared for conventional warfare, led
to US disengagement. In contrast to most forecasts, the Kabul regime, even in its weakened condition, survived the withdrawal and
became the linchpin of the unsuccessful negotiations undertaken by
the United Nations. It was not till the spring of 1992 that the regime
collapsed for lack of external support, some months after the coup
against Gorbachev and the dissolution of the Soviet Union. At this
point a new phase of the war began, as the victors fought among
themselves in rapidly fluctuating coalitions. In 1994, just as Masud
seemed definitively to have taken control of the capital, the appearance of a fundamentalist movement, the Taliban, radically altered the
political balance. The Taliban were supported by a part of the population and received substantial assistance from Pakistan, and they conquered the southern regions of Afghanistan over the space of a few
months. The capture of Herat in 1995 and of Kabul in 1996 marked
the point when the movement broadened out, and with their completion of the conquest of the north during the summer of 1998
they controlled 90 per cent of the Afghan people. Only Masud continued to resist in his mountain fastness in the northeast, with help
from the Russians, from Iran and from the West, especially France.
On 9 September 2001 the death of Masud in a suicide attack
organised by Al-Qaida appeared definitively to mark the success of
the Taliban, but the events of 11 September were to bring about its
downfall. Mullah Omar refused to hand over Bin Laden to the
United States, which launched a military onslaught on 7 October,
largely from the air, leading to the fall of Kabul and the principal
towns in November. The Petersberg accords of 5 December paved
the way for the later formation of a provisional government led by
Hamid Karzai, but dominated by Masuds successors. Several thousand western soldiers of the ISAF (International Security Assistance

Reference points

Force) secured the capital within the framework of a UN operation.

The meeting in June 2002 of a Loya Jirga (traditional assembly) gave
a legal framework to the Afghan government, still under Hamid
Karzais leadership, but it also endorsed the return of the regional
powers, who sometimes became involved in military clashes with each
other, as well as the domination of the Panjshiris in the capital. The
American forces, based mainly in Kandahar, continued to pursue
the Taliban and the Al-Qaida militants. Once the initial shock was
over, the Taliban effectively regrouped in the southeast and southern regions, relying on the help of the tribes of the AfghanistanPakistan frontier to escape from the Americans. In the absence of a
clear victory, mullah Omar and Bin Laden for the time being disappeared from view, leaving the Kabul government without authority
over the country. The American military presence persists, bringing
with it growing resentment in the Pushtun south, the sole theatre of
western military operations.

Reference points
Over the last twenty years the events briefly summarised above have
been the object of changing points of view which have revealed the
assumptions of experts and journalists as much as the internal development of the conflict. In reality the way in which alien societies are
viewed, infiltrated by our own agendas, has a tendency to create
imaginary countries. In another context this has been illustrated by
Lucette Valensis analysis of how from the 17th century onwards the
image of the Turk has been constructed in the light of the entirely
European criterion of despotism, with a resulting distortion of the
Ottomans own concerns.4 In interpretations of the Afghan war two
biases can be singled out, which may be described as the geopolitical
and the essentialist.
During the Soviet occupation from 1980 to 1989 the war was
analysed broadly in terms of the East-West confrontation, and the
local actors and their strategies were interpreted largely in terms of
global issues relevant to American-Soviet rivalry. In this perspective
all actions were decoded as if they had some bearing on the global
struggle. Analyses took little account of random factors, of the auto4

Lucette Valensi, Venise et la Sublime Porte, Paris: Hachette, 1989.

Introduction

nomy of local actors, or of bureaucratic considerations. The Soviet

invasion, in particular, was positioned in the context of an overall
strategy directed at the domination of warm-water seas, in a continuation of the objectives of the pre-1917 Russian Empire. Geopolitics was fashionable, and there was much talk of a continuation of
the Great Game, the period of confrontation between the British
and Russian Empires in the 19th century when the aim of the Russians was to secure a route to the Indian Ocean.5 By the same token
the Taliban are often represented as no more than tools of Pakistans
its anti-Indian strategy and pursuit of influence in Central Asia.6 The
geopolitical interpretation here suggests dubious geographical determinisms and invalid historical analogies which a more precise analysis does not confirm.
After the Soviet withdrawal in 1989 the Afghan crisis, once more
reduced to regional dimensions, invited attempts to discover the
real nature of the conflict. After 1992, when the fall of the regime
in Kabul opened the way to a new war between the victorious parties, there were many such attempts, which followed fashionably on
the heels of the return to realism. Once the Afghan conflict could
no longer be understood in terms of global imperatives, it became
part of what the language of international relations refers to as an example of disorder. Thus the war in Afghanistan was described as a
Hobbesian situation,7 a special case where the war of all against all
fails in reality to provide a foothold for political analysis. Meanwhile
culturalist analyses, which had been put forward since the start of
the conflict, proposed a non-political understanding of the war. This
then became a curiosity, to be interpreted in terms of the exotic and
sub-political aspect of ethnicity, and to be viewed as springing from
factors of a cultural nature, almost in a genetic sense of the term.
The war was seen simply as a return to the natural state of wickedness of the savage; so that that virtually nothing needed to be explained. It only remained to deplore the propensity of the Afghans
to internecine struggle.
5

See Mike Barry, Le royaume de linsolence, Paris: Flammarion, 1984.

Ahmed Rashid, Taliban, Islam, Oil and the New Great Game in Central Asia, London: I. B. Tauris, 2000.
7 See Mohammad Tarzi, Afghanistan in 1992: a Hobbesian state of nature, Asian
Survey 33 (2), 1992.
6

Reference points

In those of its versions which deserve analysis, the culturalist approach reduced actions to unconscious cultural patterns, denying
the capacity of the actors to act independently, and repudiating the
observation of practice in favour of a number of basic texts which,
like all written texts, are capable of an infinite variety of interpretation.8 For instance, the supposed policy of equilibrium between East
and West followed by Muslim statesmen was to be explained in part
by the position of children within the harem: a hypothesis which is
unverifiable to say the least.9 Although they did succeed in expressing
western social fears, such explanations failed to include any consideration of the causes of political violence. Three types of explanationtribalism, ethnicity and Islambelong at least in part to this
deceptive system of observation.
A tribal revolt? Afghan political life has sometimes been represented
as being a recurrent process of fission and fusion, where the integrating tendency of the state is opposed to the centrifugal force of
the tribes: the present war is presumed to be the most recent expression of this tendency.10 Afghanistan actually came into being in the
18th century as the result of the disintegration of the Persian and
Moghul empires, which the Ghilzai and later the Durrani Pushtun
tribal confederations were able to exploit in order to impose their
own will on neighbouring populations. The political domination
of the Pushtun tribes over the other ethnicities did not, however,
obstruct close relations with the modern state which took shape
progressively from the 1880s onwards. In fact, the tribes were organised in a manner which was characterised by the absence of political
institutions. Oppositions at the segmentary level of clan and tribe
normally impeded the emergence of a central power, other than at
moments of crisis. The tribe could therefore be defined as the largest
unit able to unify in the face of an external attack. Tribal identity also
makes reference to a common ancestor,as well as to a code of conduct
8

For a critique of these culturalist theses see Jean-Franois Bayart, Lillusion

identitaire, Paris: Fayard, 1997.
9 Louis Dupre, Afghanistan, Princeton University Press, 1980, p. 193.
10 See Louis Dupre, New York Times, 6 June 1989. This concept describes in theory
the politics of segmentary societies.

Introduction

predicated upon a method of settling conflicts by means of actual or

threatened vendettas, within a system of collective responsibility.11
Tribal revolts against the state were frequent throughout the entire first half of the 20th century, but the most notoriousof which
more will be saidwas that of 1929. In that year the Pushtun tribes
overthrew King Amanullah, who had embarked on a rapid modernisation of society on the Turkish model, though without the
necessary military resources. The government was undermined by
clashes with the ulema, as well as the cost of the programme of modernisation envisaged by Amanullah. An initially small-scale tribal revolt was able to spread because of the weakness of the armed forces
and the support of the ulema, who issued a call to jihad (holy war)
against King Amanullah, the king who was routed and went into exile. Subsequently, to widespread surprise, a Tajik bandit, Habibullah,
known as Bacha-yi Saqao (son of the water-carrier), was able to
take advantage of the interregnum to capture Kabul, where for some
months he ran a fundamentalist regime.12 Only when Nadir, a cousin
of the king, came back was it possible, with difficulty, to restore a coalition of all the Pushtuns and with the discreet help of the British to
establish the Pushtun dynasty which reigned up to 1973.
Can the events of the present day be interpreted in the light of
1929? In 1978 the modernising reforms of the communists precipitated a tribal revolt, followed by the overthrow of the central authority and the occupation of Kabul by Masudwho, like Habibullah,
was a Tajikbefore the reconquest undertaken by the Taliban, who
were mainly Pushtuns. This analogy, attractive as it may be, does not
explain a fundamental novelty: the emergence of political parties,
11

The expression tribe has many definitions. Here we concentrate on those characteristics which are relevant to the Afghan situation, laying particular stress on
segmentarisation. See Pierre Bont, Michel Izard, Dictionnaire dethnologie, Paris:
PUF, 1980, p. 720. This theoretical model does not exclude the occasional exploitation of segmentarisation by the most powerful elements to create a central
authority. See Jean-Pierre Digard, Une contribution quivoque du droit coutoumier Baxtiari la thorie de la segmentalit in Jean-Pierre Digard (ed.), Le
cuisinier et le philosophe, Paris: Maisonneuve et Larose, 1982. In Afghanistan there
sometimes exists a division at the level of tribes, clans and families into two bast
or factions: spin (white) and tor (black). These correspond to voluntary groupings,
or fraternities, a phenomenon encountered in many segmentary societies.
12 See Leon Poullada, Reform and Rebellion in Afghanistan, 19191929: King Amanullahs failure to transfrom a tribal society,Ithaca,NY:Cornell University Press,1973.

Reference points

which were far from being representative of the tribes. Further, it

was not the tribes who took the initiative in the rebellion but the
urban and rural population as a whole. The non-tribal communities,
which are in the majority, were involved in the fighting to the same
extent as the rest. It should finally be pointed out that it was broadly
forbidden to bear arms in Afghanistanwith the exception of a
number of eastern Pashtun tribes attached to the stateand that the
tribes were in decline and had not for some decades represented any
threat to the central authority.
An ethnic war? If an explanation based on tribalism finds too little
foundation in fact, the appeal to ethnicity recapitulates in an apparently more convincing style some aspects of the same argument. Do
ethnic antagonisms, or at least the hierarchy among the principal
ethnicities present in Afghanistan (the Pushtuns, the Tajiks, the Uzbeks
and the Hazaras), contribute to an understanding of the dynamics of
the conflict? Before reaching a conclusion, prudence dictates that
the concept of ethnicity, which has been the subject of a number of
controversies in recent decades, should be more precisely defined.
Up to 1978 a significant part of the literature on Afghanistan was
of an ethnological or anthropological nature, reflecting the interest
of western researchers in a pre-industrial society which appeared to
be a rich repository of cultures and languages. Much discussion concerning the concept of ethnicity was therefore based on the Afghan
situation. Up to the 1950s, in the tradition of colonial ethnographic
literature, an essentialist idea of ethnicity prevailed. At that stage the
concept of ethnogenesis occupied a central position, since ethnic
affiliation appeared to entail characteristics, which were virtually inflexible as well as culturally and biologically defined, as for example
in the case of the warrior races beloved of the British officers.13 A
catalogue of objective criterialanguage, history, community, religion, material interestsconstituted the definition of an ethnicity, in
a theoretical approach which was carried over into Soviet ethnography. Later, by contrast, Afghanistan became one of the prime fields
for a new approach to ethnicity, viewed now in terms of a social rela13

Nirad C. Chaudhuri, The Martial Races of India, Modern Review 68 (4151),

1931, pp. 295307.

10

Introduction

tionship, in a departure from the illusory supposition that communities are natural objects. Henceforth the accent would be on the
construction of communal identity in its relation with other groups,
hence the interest in the concept of an ethnic frontier, implying an
altogether more subjective and relational idea of ethnicity.14
In Afghanistan the identity of each individual is defined by a
series of affiliations, from the most generalto the umma (the Muslim community as a whole)to the narrowest, the close family. A
sense of identity may be based on a shared geographical origin, or on
a common affiliationprofessional, religious, family, ethnic etc. In
the tribal context, the extended family, the clan, the tribe and then the
tribal confederation appear as a series of concentric circles, which
they are not in a non-tribal context. The idea of a qowm specifies
precisely these identities when they are mobilised within a solidarity
network.15 According to context, different identities are mobilised,
exemplifying the jostling and pluralism of collective identities.16
For example a solidarity among Shiites arises out of their confrontation with a largely Sunni society.
Each individual has a place within the networks of solidarity
(qowms) based on obligations which are more or less extended and
defined. The qowm leans towards the nature of a system of exchanges
which expands on the model of the extended family, although its actual basis may be different, e.g. professional, political or confessional.
Affiliation to the same qowm may in particular imply an extended
co-responsibility in the case of a vendetta. This follows the conclusions of Jean Leca and Yves Schemeil relating to the Arab world,
where interpersonal relations are conceived on the pattern of family
relationship.17 The concept of qowm is also close to that of asabiyya,
14

Fredrik Barth (ed.), Ethnic Groups and Boundaries: the Social Organisation of Culture
Difference, Boston, MA: Little, Brown. See also Jean-Pierre Digard (ed.), Le fait
ethnique en Iran et en Afghanistan, Paris: Ed. du CNRS, 1988.
15 In the first instance this Arabic term signified a patrilineal group, but today it is
the word for fatherland used in the Arab world. The Afghans use it in a more
generalised sense for any solidarity group.
16 Le chevauchement et la pluralisme des identits collectives, Pierre Centlivres,
Micheline Centlivres-Demont, Et si on parlait de lAfghanistan, Paris, Neuchtel:
Editions de lInstitut dEthnologie de Neuchtel et la Maison des Sciences de
lHomme, 1988, p. 37.
17 Jean Leca, Yves Schemeil, Clientlisme et no-patrimonialisme dans le monde
musulman, International Political Science Review, December 1983.

Reference points

11

introduced by Ibn Khaldun, which designates the fellow feeling

which results from agnatic solidarity, or something similar to it.18
According to its context, the word qowm may signify solidarities
of a very different kind, within communities of variable size. A solidarity group emerges, for example, through the necessity to maintain
relations with the state via the intermediary of the administrative official at the village level (the malek). Ecological conditions are also
sometimes a factor conducive to cooperation at the village level.
The election of a person to oversee the distribution of water, namely
the mirab, who organises the irrigation system, defines a solidarity
group.19 A phenomenon revealed by the war was the significance of
solidarities of place. This was often defined by attachment to the
same administrative division (province, district or sub-district), a
phenomenon that illustrates the capacity of the state to redefine the
framework within which networks of solidarity operate, rather than
to challenge the structural principle.
A further factor is that a qowm does not necessarily correspond to
a contiguous area. For example, the inhabitants of a valley may be divided into two non-territorial groups due to a religious difference,
with the two qowms so defined enjoying no spatial coherence. Since
the reign of Abdul Rahman Khan (18801901) the majority of the
villages in the south and centre of Afghanistan had lost the homogeneity which dictated that clan solidarities should have a territorial
basis,20 although this continued to be partly the case in the east of the
country. Thus the village did not necessarily represent a qowm and
cooperation is sometimes confined to the upkeep of the mosque.
Classifications by macro-ethnic groups (see table, page 15), which
almost correspond to linguistic groups, therefore constitute only one
level of communal affiliation among others and do not necessarily
18

Olivier Carr, Note critique propos de la sociologie politique dIbn Khaldoun, Revue franaise de sociologie 14 (1), 1973. Ibn Khaldun, Al-Muqaddimah: An
Introduction to History, tr. Franz Rosenthal, 3 vols, London: Routledge and Kegan
Paul, 1958.
19 Barthlmy Amat, Lorganisation paysanne pour la distribution de leau pour
lirrigation dans les villages de la steppe, linstitution du mirab, Afghanistan Journal, 1977. See also R. and M. Poulton, Coopration spontane autour dune
mosque afghane, Revue des tudes coopratives 4, 1979.
20 Louis Dupre, The changing character of South Central Afghanistan villages,
Human Organization 14 (4) 1956.

12

Introduction

imply the existence of operational solidarities. The case of the Tajiks

clarifies this issue. According to some estimates, there were before
the war around 3.6 million Tajiks in a total Afghan population of
12 million, who were therefore numerically the second largest ethnic group after the Pushtuns. In Afghanistan any settled, Sunni,
non-tribalised Persian speaker will be described as a Tajika definition which therefore emerges from a combination of linguistic, religious and sociological criteria, and so lacks both homogeneity and
precision.21 Thus the Aymaqs, who are both tribalised and nomads,
are not designated as Tajiks even though they are also Persian speakers and Sunnis, and do not describe themselves as such;22 the same
applies to the Ismailis, who are distinguishable from the Tajiks only
by their religious affiliation.23 The differences between identity as
claimed and that as attributed by other groups, as well as apparently
anomalous cases such as those of Tajiks who speak Pushtu or Baluchi,
relate to the value attached to classification, a dynamic and conflictual social process in which context influences the choice of declared identity.
Tajik identity relates to and is weakened by the ambient situation,
since affiliation to Persian culture transcends the Afghan context and
is also linked to Iran and Tajikistan. The linguistic criterion does not
by itself permit a simple differentiation of the Tajik population, since
the Persian language is also used by the Hazaras, the Aymaqs and numerous other communities (sedentarised nomads, some of the Arabs,
and so on).
In addition, Tajik identity competes with other stronglyheld
local identities which are linked to more substantial solidarity rela21

la fois dhomognit et de prcision, Pierre Centlivres, Un bazar dAsie Central. Forme et organisation du bazar de Tashqurghan (Afghanistan), Wiesbaden: Ludwig
Riechert Verlag, 1970, p. 155. It should be added that the category of Tajik is
partly the outcome of administrative classification. During the 1970s, in fact, the
State included on its identity cards an indication of linguistic (and hence indirectly ethnic) affiliation of a highly simplistic nature, classifying individuals as
Turkmens, Pushtuns or Tajiks.
22 Alfred Janata, On the origin of the Firuzkuhis in Western Afghanistan, Archiv
fr Vlkerkunde 25, 1971, pp. 5765.
23 Another issue is that Louis Dupre, Afghanistan, op. cit., p. 59, classifies the farsiwan
(Persian-speakers) as an entirely separate ethnic group, defined by their spoken
language (Persian) and their affiliation to the Imamite Shiite sect. For a critique
of the ideas of Louis Dupre, which derive from an objective approach to the issue
of ethnic identity, relying on ethnogenesis, see Jean-Pierre Digard (ed.), op. cit.

Reference points

13

tionships. People refer to themselves, for example, as Herati, Kabuli,

Badakhshi, Panjshiri, Andarabi or Munjani. These identities have
more mobilising power than the more abstract affiliation to a Tajik
group which depends primarily on the linguistic criterion. These
populations do not view themselves subjectively as belonging to a
community: there is no feeling of solidarity with the outside world
nor are there common values, although this statement should probably be qualified in the case of the mountain Tajiks in the north-east,
who are a more homogeneous and self-conscious group.
Finally, the Tajiks are dispersed throughout Afghanistan, and although they are in a majority in the north, the ethnic distribution
here is also especially complex. In contrast to the other ethnicities
the Hazaras, Pushtuns and Uzbeksthe Tajiks are not mainly concentrated in any one region. In this sense there is no Tajik diaspora,
since they have no heartland from which they might be supposed
to have migrated. In fact the Tajiks belong to two distinct groups
according to their geographical locality and their way of life. In Afghanistan the geographical distribution of the communities is to an
extent regulated by the ecological opportunities afforded by economic occupations. The Tajiks live in the towns, or more properly
in the bazaars, and occupy an important position in trade, but the
term Tajik also applies to the long-standing Iranian populations who
farm in the high valleys of the Hindu Kush. Another factor is that a
person is deemed to have become a Tajik when he takes up the urban
way of life and adopts Persian as his everyday language, especially in
such towns as Herat and Kabul. In this case the category Tajik has
least to do with ethnicity24 and can hardly be viewed as a positive
affiliation with the capacity to produce real effects.
Despite the complexity of communal affiliations, however, the
course of the war after the Soviet withdrawal tended towards the
notion of an ethnic conflict as the criterion of analysis, sometimes
as an alternative paradigm to that presented by Islamism.25 The proliferation of texts in which macro-ethnicities are posited as collective
actors has entailed frequent confusion between the two ideas, with
24

reprsente le degr zro de lethnicit, Pierre Centlivres, Micheline CentlivresDemont, op. cit., p. 47.
25 For an instance of this change of paradigm see Olivier Roy, La guerre dAfghanistan. De la guerre idologique la guerre ethnique, LHomme et la Socit,
1993, no. 1078.

14
Introduction

Simplified distribution of macro-ethnicities.

Reference points

15

MACRO-ETHNIC GROUPS IN AFGHANISTAN (1978)26

%
Pushtun
Tajik
Uzbek
Hazara
Aymaq
Turkmen
Baluchi
Nuristani
Pasha
Other
Total

4,800,000
3,600,000
1,200,000
1,000,000
500,000
400,000
200,000
70,000
60,000
170,000
12,000,000

40
30
10
8.3
4.2
3.3
1.7
0.6
0.5
1.4

Pushtun being equated with Taliban. At this point the war was
viewed as a clash between communities, of a type liable to recur.
With the ideological dimension left out of account, the identity of a
political leader is reduced to his communal aspect alone, especially
by the media. In reality the employment of the category of ethnic
warfare to describe the Afghan conflict is far from being neutral,
and is in itself an ideological position. In the case of external observers, the use of such terminology tends to erect this category as in
some way scientific or meaningful, when it is actually part of what is
at stake in the conflict. The antithesis between ethnic war and
ideological war is not appropriate, since nothing could be more
political than the mobilisation of community identities.
However,it was not ethnicitiesthat made war,but political organisations with ideological objectives and particular institutional practices. Furthermore, the initial mobilisation took place in the context
26

Table from A. Janata, Notizen zur Bevlkeringskarte Afghanistan, Afghanistan

Journal 8 (3), 1981, pp. 945, quoted by Pierre Centlivres, Les Nouvelles dAfghanistan 47, April 1990, p. 4. According to other sources, the number of inhabitants
may have been as many as 14 million (it is noteworthy that this is the figure accepted by the World Bank). The percentages of the various macro-ethnicities
are uncertain, due to the lack of reliable sources. Pierre Gentelle estimates the
number of Pushtuns at 3537 per cent, putting them on a par with the Tajiks
(Encyclopaedia Universalis, supplement, Afghanistan, 1980), while Leon B. Poullada
estimates the percentage of Pushtuns to be double that of the Tajiks, see The
Pushtun role in the Afghan political system, Occasional Paper no. 1, the Afghanistan
Council of the Asia Society, 1970.

16

Introduction

of local qowms and not macro-ethnicities. It is also because of this

specificity that, especially at the beginning of the war, it was impossible to discern the disposition of forces on the ground since qowms
were not necessarily territorially based, with different parties often
represented within the same village. Subsequently some parties actually did make use of communal affiliations to mobilise support, since
especially at times of crisisidentities are up to a point susceptible to
manipulation. The appearance of nationalist ideology within certain
communities was a political development which sprang from the strategies of the parties and the history of particular groups. The political
role of the Panjshiris, who exercised a political influence disproportionate to their demographic weight, may perhaps have redefined
them as an ethnic group separate from the Tajiks, thus indicating the
plasticity of identities and their political origin.
After the Soviet withdrawal events moved in the direction of an
alignment of the distribution of ethnicities and political affiliations
on the ground. It should be understood, however, that the ethnicisation of the parties was a consequence of the war. Finally, the focus on
ethnicity excluded the consideration of other dynamics, both social
and ideological, which were equally significant. The new political
equilibrium which had its origin in the American intervention
tended to favour a new interpretation of the war as ethnic, since this
was the only language which the foreign powers understood without difficulty.
A blend of religion and politics? The interpretation of the Afghan war
also hangs on a tendentious question: what is the relationship between politics and religion in Afghanistan?or, to take up a familiar theme, does Islam necessarily identify these two fields? Within
the term Islam a distinction must be drawn between a system of belief
which gives a direction to action, and a theological system, although an
excessive focus on the latter aspect, neglecting social practices, is to be
avoided.For instance it is difficult to argue,as Bernard Lewis does,27 that
in a Muslim society political and religious power are identical, argu27

Especially in Le retour de lislam, Paris: Gallimard, 1985, p. 375. Bernard Lewis is

sympathetic to the theories of Samuel Huntington, for whom the clash of
civilisations is the key to present developments.

Reference points

17

ing from the absence of a word for laicity. In the first place the
Arab-Persian vocabulary distinguishes religious concepts (ulema, jihad and so on) from those of politics (such as hukumat or qanun),
which is some basis for the supposition that the two domains are distinct. Secondly, when the ulema reject the Christian doctrine of the
separation of religion and politics,28 their claim is that it is their prerogative to provide politics with its moral standards, rather than appealing for a somewhat improbable fusion of the two domains.
Above all, throughout the entire Muslim world the separation of
the religious and political functions has historically been the norm:
few ulema have exercised political power other than in situations of
crisis which have been viewed as exceptions to the norm.29 In Afghan
history the amirs (rulers), or those who aspired to power, were always
members of the tribal aristocracy, and were never ulema. This practical and theoretical separation dates from an early era, in the 9th and
10th centuries, with the emergence of the ulema, a body of men of
religion,30 who constituted a relatively autonomous social group,
identifiable principally through their knowledge of the Quran and
its exegesis, acquired by means of recognised processes.
If the fields of politics and religion are not the same, what is their
relation to each other? Such relationships as exist in the West in
modern timesfor example laicity in France and secularism in the
Anglo-Saxon worlddo not exist in the Muslim countries. A further factor is that the frontier between the religious and the political
spheres constantly fluctuate: just as in the West the meaning of the
idea of laicity is not the same at different epochs and in different
28

For example in 1939 al-Maraghi, shaikh al-Azhar, cf. Gustav von Grnebaum,
Lidentit culturelle de lislam, Paris: Gallimard, 1973, p. 56.
29 From the historical standpoint, the proposition that politics and religion are one
is in part vindicated by the nature of the community surrounding the Prophet
Muhammad, where, significantly, the building where the Prophet lived served as
both a place of worship and the seat of government. The practice of the first caliphs showed the same lack of distinction between religious and temporal authority. However, at that time there was no state as such, though a minimal
institutionalisation of power, and the tribal power struggle, were always basic.
Even in this case, the identification is not total, since the politico-religious power
was obliged constantly to compromise with the tribes, whose source of legitimacy lay outside religion. See Maxime Rodinson, Mahomet, Paris: Fayard, 1961.
30 Olivier Carr, Islam laque ou le retour la grande tradition, Paris: Colin, 1993.

18

Introduction

countries, so the language of the religious establishment in Islam tends

to vary in response to historical circumstances.31
In the context of relations between the state and the ulema, the
central question is that of the legitimacy of political power. The state
needs the backing of the ulema to persuade individual citizens of the
religious legitimacy of state domination, since it was scarcely able to
exercise a monopoly over security. The recourse to solidarity in the
private sphere to guarantee the protection of individuals here relates
to anthropological realities preceding Islam, which obstruct or delay
the formation of a central state. Without exception the ulema have
legitimated the political order, while quietism amongst the religious
establishment tends to be prevalent in the Islamic world.
It would be difficult, especially in the case of Afghanistan, to argue
that actors who reject the separation of the political and the religious
spheres embody the essential principles32 of Islamic culture, or that
such a return belongs to the great tradition of Muslim history.33
Other authors, such as Olivier Carr, correctly stress the marginal
and unusual aspect of these tendencies over the long historical perspective.34 In any case, unless one accepts the ideology of certain political groups, Islam does not connote a genetic predisposition to a
unique mode of interaction between religion and politics.
On the other hand, initiatives do emerge from time to time which
are aimed at a religious repositioning of the political order. Shah
Walihullah in India and Ibn Abdul Wahhab in the Arabian Peninsula
are the outstanding examples, both from the 18th century; in addition to the variegated ensemble of contemporary Islamist and funda31

In the case of the Shiite clergy Pierre-Jean Luizard has shown how the political
claims of religious figures have in the past been linked to particular historical situations, and in particular their relationship with the Safavid state (15011786).
In fact the language of the Shiite ulema has been notably variable, and includes a
substantial element of tactical positioning. The imperative to assume the powers
of the Imam, and thus to lead the community, has never met with unanimity,
even among the Shiite clergy themselves. See Pierre-Jean Luizard, La formation
de lIrak contemporain, Paris: Editions du CNRS, 1991, p. 126.
32 les principes essentiels,Bertrand Badie, Les deux Etats,Paris:Fayard,1988,p. 103.
33 la grande tradition de lhistoire musulmane, Jean-Paul Charnay, Sociologie religieuse de lIslam, Paris, Hachette, 1994, p. 39.
34 marginal et dviant,Olivier Carr, Lutopie islamiste dans lOrient arabe, Paris:
Presses de la Fondation nationale des sciences politicques, 1991 p. 10 and p. 31. The
constant danger is of supposing there exists a homo islamicus, an idea discredited
by Maxime Rodinson in La fascination de lIslam,Paris:Presses Pocket,1993, p. 82.

Issues and hypotheses

19

mentalist movements.35 The narrative of the origins of the Muslim

community does duty as a utopia which can be turned to whenever
social conditions are propitious. What distinguishes Islam is therefore not the absence of separation between politics and religion,
whether in the theoretical or the practical sphere, but the availability
of a utopian religious model to contest the political authority. In
practice, however, the sociology of contending groups, ulema or students, provides an explanation of such social mobilisations. From
this viewpoint the Afghan Islamist movements were represented at
the onset of the war as a privileged route towards political modernity.36 However, although this analysis may well illuminate some
aspects of Afghan political movements, it has led to excessive simplification. In the first place, identifying a social groupin this case
young people who have emerged from the educational system
with a political tendency is misleading, since the Islamists never
attracted more than a minority of students. Secondly, the focus on
the Islamists has resulted in an underestimation of the role of the traditional lites, sometimes themselves modernisers, and particularly
of the ulema, who played a central part in the Afghan resistance.

Issues and hypotheses

The present study makes no claim to cover all aspects of the Afghan
crisis. For example, the military, diplomatic and economic dimensions lie largely outside the selected field of study, and are touched
on only to the extent that they explain more strictly political phenomena.37 A number of questions are crucial to an understanding of
developments since the 1980s.
First, how can the origin of this state of political violence be accounted for? In the 1960s a revolutionary situation had developed as
35

On these movements, see particularly Gilles Kepel, Yann Richard (eds), Intellectuels et militants de lislam contemporain, Paris: Seuil, 1990, and Olivier Carr, Lutopie
islamiste dans lOrient arabe, op. cit.
36 Olivier Roy, Afghanistan, Islam et modernit politique, Paris: Seuil-Esprit, 1985.
37 For more complete biographical resources the interested reader should consult
the thesis which I successfully presented at EHESS in December 1995. The section of the thesis devoted to economic issues in the war is partly reproduced
in Afghanistan: from solidarity networks to region in Franois Jean, JeanChristophe Rufin (eds), Economies des guerres civiles, Paris: Hachette, 1996.

20

Introduction

a result of changes in the educational system.38 The emergence of a

group of politicised educated persons in a developing country is
clearly not peculiar to Afghanistan,39 but the parliamentary regime
which was belatedly installed was unable to respond to social demands and gradually forfeited all support. Nevertheless, the communist coup of 1978 was in no way inevitable, and did not bring
about any significant degree of social mobilisationin contrast, for
example, to the Iranian revolution.40 In particular, nothing points to
a profound rejection of the state for cultural or other reasons.
After 1978 how did mobilisation against the state take place? The
origin of the civil war lay much more with the violence of the state
than in the reforms implemented by the new authorities. In contrast
to the situation of most guerrilla movements of the 20th century, the
rural masses were not led by a modernising and revolutionary lite.
With a few exceptions uprisings were spontaneous, and the political
parties were formed after the event or independently of the popular
rebellion. This was initially a counter-revolution, a rejection of atheism and of foreign occupation. The jihad (holy war) was the moral
basis of the individuals engaged in the war, and militated against a
political view of the conflict. Mobilisation took place in practice
within the context of local networks of solidarity, under commanders. The dichotomy between the commanders and the parties in
exile explains why the parties had great difficulty in maintaining a
monopoly over political representation and over the disposition of
resources, and also why the parties disappeared after the fall of the
regime in Kabul.
In what areas did the Afghan war result in revolutionary social
changes? Political ideology and the politics of the ruling class were
38

Charles Tilly draws the distinction between a revolutionary situationthe emergence of a group mobilised in the defence of its interests, and the revolutionary
process itself, cf. Charles Tilly, From Mobilisation to Revolution, Reading, MA:
Addison-Wesley, 1978.
39 See Maurice Martin, Egypte, les modes informels du changement, Etudes,
April 1980, pp. 43552.
40 Which contradicts the postulation of a relationship between massive effects
such as civil war, and profound causes. On these issues, see Michel Dobry,
Sociologie des crises politiques, Paris: Presses de la Fondation nationale des sciences
politiques, 1986, p. 71, and Raymond Boudon, Lart de se persuader des ides
douteuses, fragiles ou fausses, Paris: Seuil, 1990, p. 270.

Issues and hypotheses

21

mostly transformed after the 1970s. As in any crisis, the personal

status of all individuals was re-evaluated, and certain forms of legitimacy became the criterion of access to political or economic resources. An examination of the social background of these new
lites reveals the significance of two groups. One of these, the ulema,
whose role was not especially important before the war, became extremely influential. Almost all the political parties were lead by religious figures. The other group was that of educated individuals,
who emerged from the towns to provide the most dynamic leadership of the resistance. During the war a progressive polarisation
came about between the educated Islamists and the fundamentalist
ulema. With the supremacy of the Taliban a clerical revolution installed a regime unique in the history of the Muslim world.
Finally, what dynamic can be adduced to account for the tendency in the 1990s towards the progressive concentration of power
and the elimination of the weakest actors? Commanders and parties
unable to accumulate resources were in practice driven out, in favour
of complex organisations run by ulema or by the educated. The war
in this instance instigated a modernisation of the forms of organisation. In addition, the international order favoured concentration by
excluding in practice any alteration of the frontiers. Control of the
state and in practice of the capital, Kabul, was the major objectives of
the actors. Since another feature of the situation was that the internal
frontiers could not be stabilised, partition on a regional basis was unable to end contestation. In another context Norbert Elias has described an example of the mechanism of a progressive centralisation
of power.41 Although there were opposing sets of objectives, competition between actors tended to result in one of them achieving a
monopoly. This process operated to the benefit of the Taliban up till
the American intervention, which simultaneously re-established regional authorities and put an end to rivalry for the capital, which was
henceforth placed off limits by the presence of international forces.
This unstable situation, with inadequate international aid, facilitated
the reconstitution of the faade of a state, which however was for the
moment unable to impose its will over the rest of the country, where
the civil war continued.
41

Norbert Elias, La dynamique de lOccident, Paris, Calmann-Lvy, 1975.

22

Introduction

With the intention of testing these hypotheses, the first part of the
book concentrates on the sociogenesis of the Afghan state and developments from 1960 to 1970. The second part examines political
mobilisations in the local and then the national context. The third
part looks at the dynamics of the conflict, from the standpoint of the
government, and that of the guerrillas. The fourth section looks at
developments after the Soviet withdrawal, and especially at the abortive reconstruction of a central political authority, the ethnicisation
of the war, and the aims of the fundamentalist ulema. The fifth and
last section deals with the American invasion and the return of
fragmentation.

Part I. THE ORIGINS OF THE

AFGHAN REVOLUTION
1. The Sociogenesis of the Afghan State
The political history of Afghanistan may be interpreted as the emergence of a state whose development has re-ordered the relationships
between social groups.1 After the death of Aurangzeb (1707), the
vacuum created by the collapse of the Mughal empire and the decline of the Safavid empire allowed Mir Wais, at the head of the
Ghilzai tribal confederation, to drive out the Persians in 1709. In
1747 the election of Ahmad Shah (17471773) as chief of the Durrani tribal confederation marked the inception of an Afghan dynasty
which, after the battle of Panipat against the Indians in 1759, expanded its empire as far as the Indus. Originally elected by the tribes
as an agent of military coordination, the amir attempted afterwards to
keep the proceeds of the conquests for himself. In 1775 Timur moved
the capital from Kandahar to Kabul in order to distance himself from
the Durrani tribes. However, he was unable to escape the effects of
the tribal system, and continued to rule over an empire with fluid
boundaries which progressively shrank to the dimensions of presentday Afghanistan under the pressure of British and Russian imperialism. Up to the end of the 19th century incessant revolts resulted in the
emergence of what were for practical purposes independent provinces in Kabul,Kandahar and Herat.Attempts to establish a centralised
authority inevitably failed due to the weakness of the urban economy and to wars of succession which reduced the country to ruin.2
In any explanation of the sociogenesis of the Afghan state from
the end of the 19th century, account must be taken of the part played
1

On the building of the Afghan state, reference will be made to the studies made
by Hasan Kakar, Government and Society in Afghanistan: the Reign of amir Abdur
Rahman Khan, Austin, TX: University of Texas Press, 1979, and Vartan Gregorian,
The Emergence of Modern Afghanistan, Stanford University Press, 1969.
2 Gregorian, op. cit., p. 69.

23

24

The Sociogenesis of the Afghan State

by the imperial powers, Britain and Russia, which defined the frontiers of Afghanistan and supported the state at critical moments.
Subsequently the development of the institutions of the state took
place side by side with the maintenance of patrimonial practices
within the governing class. Meanwhile the state, whose source of
legitimacy was progressively transformed from Islam to Pushtun
nationalism, faced challenges from the ulema and the tribes.

The role of imperialism

From the close of the 19th century onwards, the internal balance of
forces within Afghan society was diverted into an unnatural course
by Britain and Russia, which laid down the frontiers even before a
state existed which was truly able to impose its authority on the territory. In fact, those two powers on the whole favoured the establishment of an Afghan state, once the impossibility of direct control
was recognised. The frontiers of Afghanistan thus came into being
through a process involving both negotiation and war between the
British, the Russians and the Afghans. In 1891 the British created
the Wakhan corridor, which separated India from Russia and thus
gave Afghanistan a border with China. Two years later Britains Colonel Durand drew the Durand Line which demarcated the frontier
between Afghanistan and the Indian empire.3 Drawn up exclusively
in the light of military considerations, the Durand Line divided the
Pushtun tribes, and was to be the origin of a lingering dispute between Afghanistan and Pakistan. Finally, between 1904 and 1906 the
British demarcated the Afghan-Persian frontier.
In order to forestall disorder on their frontiers, the British were to
make a priority of the stability of Afghanistan, ensuring that conflict
between the amir and the tribes did not result in disadvantage to the
amir. In 1883 Gladstone granted the Afghan sovereign an annual
subvention, financed by a tax levied on the population of India. The
annual revenue of the kingdom of Afghanistan was at that time 12 to
13 million rupees. In relation to this the additional British contribu3

Dr Azmat Hayat Khan, The Durand Line: its Strategic Importance, University of
Peshawar and Hanns Seidel Foundation, 2000. Strictly speaking, the Durand Line
was not a frontier, but rather the limit of the British zone of influence, which was
the justification for Afghan claims on territory populated by ethnic Pushtuns
when Pakistan was established.

The role of imperialism

25

tion was substantial, amounting to 1.8 million rupees per year from
1893, in addition to a supplementary fund available in case of emergency, thus bringing the total to 28.5 million rupees between 1883
and 1901.4 During periods of tension, the British provided money
and arms whenever the central power in Afghanistan appeared to be
in difficulty, especially during the tribal uprisings of 1880, 1882 and
1887. Later, in 1924, during the revolt of the Mangal tribe at Khost,
King Amanullah, with British approval, enlisted Soviet and German
pilots for punitive air raids. After the overthrow of Amanullah in
1929, the British and the Soviets intervened to come to the assistance of the respective parties. When, after some months, British
support enabled a tribal coalition led by the future ruler Nadir to be
established, the Russians went so far as to risk a brief military incursion into the north of Afghanistan in support of Amanullah.5 In the
following period, British assistance in the reorganisation of the army
was equally significant, with the provision of 10,000 rifles, 5 million
cartridges and a sum of 180,000.6

Neo-patrimonialism
The construction of the state, once begun at the end of the 19th
century, was to continue from that date in spite of some reverses.
The concept of neo-patrimonialism7 applies well to a somewhat
ambiguous situation, in which the development of institutions continued but was accompanied by the continued predominance of a
restricted group which was able to perpetuate its hold on power.
The institutionalisation of the state. The project of establishing a state
on the western model came into existence with Abdul Rahman
Khan. He, by refraining from nominating his heirs to positions as
4

Hasan Kakar, op. cit., p. 90.

Georges Agabekov, OGPU: the Russian Secret Terror, New York: Brentanos, 1931;
and Anthony Arnold, Afghanistan: the Soviet Invasion in Perspective, Stanford, CA:
Hoover Institution, 1981, p. 19.
6 Hasan Kakar, op. cit., p. 321.
7 The neo-patrimonial state is defined by the appropriation of the partially institutionalised political centre by a group essentially oriented towards the maintenance
of its own power. On this issue see Jean-Franois Mdard, Etat patrimonialis,
Politique Africaine, September 1990.
5

26

The Sociogenesis of the Afghan State

provincial governors, set a limit to the patrimonialisation of power.8

The very broad autonomy of the provincial governors was suppressed in favour of a more centralised system, which strengthened
the authority of the amir. The size of the provinces was reduced,
which improved administration and offered a more restricted platform to any potential claimant to the throne. Two examples illustrate the transition towards the institutionalisation of power. First,
the royal treasury in 1922 distinguished between the personal expenses of the king and those of the state. The transition towards an
institutional system was thus definitively set in train, all the more so
since a constitution was adopted in 1923. Secondly, the constitution
of 1964 provided for an elected parliament and ruled that members
of the royal family could no longer hold ministerial office.
In addition, the recruitment of administrators reinforced the autonomy of the state. In the 19th century the gholam-i shah (the kings
slaves), in fact orphans or other children under the protection of the
amir, constituted the backbone of the royal administration. The state
thus established for itself a group of loyal supporters, structurally
linked to its interests, and divorced from tribal loyalties. The gholam-i
shah, who often married into influential families, would provide in
due course the personnel of the reformist circles of the first decades
of the 20th century.9 Similarly the pishkhedmat, who were the sons of
notables, kept close to the amir as sureties for the loyalty of the group
whom they represented, performed various duties at the royal court.
This system of recruitment was supplemented by the educational
system, of which one function was to educate administrative lites
for the state. Thus the military college at Kabul, founded by Habibullah in 1904, mainly recruited the sons of the khans of the Durrani
tribal confederation,of whose loyalty the regime wished to be assured.
The college was to provide a significant proportion of the higher
8

The patrimonial state is characterised by leadership of a traditional kind, where

the decision-making power is viewed as the perquisite of the prince and his followers. Before Abdul Rahman Khan, it would be appropriate to speak of
Sultanism, because of the extent to which power appeared to be the personal
property of the amir, divisible between his heirs at successions. Cf. Max Weber,
Economie et Socit, Paris: Plon, 1968, p. 231.
9 See Dr Fazal ur-Rahim Marwat, The Evolution and Growth of Communism in
Afghanistan 19711979, Karachi: Royal Book Company, 1997, p. 106.

Neo-patrimonialism

27

administrators in the first two decades of the 20th century.10 From

the 1950s the university progressively took over the task, but the
administration then found itself unable to absorb the growing cohorts of graduates. The administration remained relatively unexpanded, as the number of officials was estimated at 40,000 in 1969,
in a population of around 10 million people.11
To an even greater extent than the civil administration, the army
was a priority for investment by the Afghan state. The development
of a national army was undertaken in response to the stimulus of the
tribal revolts which punctuated the states drive towards centralisation. Military investment was thus justified by immediate security
considerations, since as late as the 1950s the state found difficulty in
quelling a rebellion solely by the use of its army and was obliged to
raise tribal levies. The clashes of 1924 during the Khost rebellion
were an indication that the balance of forces was not always favourable to the central authority, as the debacle of 1929 a short time later
also showed. In July 1930 the army, unable on its own to put down a
revolt in Kohistan to the north-east of Kabul, was obliged to call on
the Pushtuns. In contrast with these internal threats, the protection
of Afghanistan against neighbouring powers was always a secondary
consideration. The wars against the British in the 19th century, an
essential element in Afghanistans historical memory, were waged by
the tribes. The armys only external engagement, which was in any
case on a limited scale, was the war of independence against Britain
in 1919,12 and even in this case the assistance of the Wazir and
Mahmud eastern Pushtun tribes was necessary. On this criterion the
troubles of 1959 marked a turning-point, since for the first time a
tribal rebellion was easily crushed by the army. The process of imposed modernisation already attempted under Amanullah was able
10

In the first place Habibullah founded a school for the sons of notables (maktab-i
malikzadeh), which later became the royal military college (madrasa-yi harbi-yi
sirajiya).
11 The figure given by the American governmental agency USAID, cited by
Gilbert Etienne, LAfghanistan ou les alas de la coopration, Paris: PUF, 1972, p. 45.
12 Unable to occupy Afghanistan, the British nevertheless imposed their control on
its external relations. After the war of 1919 Afghanistan was accorded full international recognition and welcomed its first foreign embassies. See Ludwig
W. Adamec, Afghanistan 19001923: a Diplomatic History, Berkeley: University of
California Press, 1967.

28

The Sociogenesis of the Afghan State

once more to get under way, and the policies of Daud, from his first
term as Prime Minister onwards (195363), relied on a military institution whose political significance was growing.
At the beginning of the century the army was a crucial item of
expenditure, with 50 per cent of the budget allocated to it.13 Amanullah (191929) was an exception in that at the beginning of the
1920s he cut the military budget, which was to be one of the reasons
for his downfall.14 Subsequently the army became once more a priority. Nadir Shah and his successors devoted more than half the budget to it, and in the 1960s it attracted a third of the current account
expenditure. The large size of the army explains in part the levels of
expenditure. In the first years of his reign Abdul Rahman Khan
set up a paid army of 43,000 men, which had risen to 100,000 by
the time of his death in 1901. After a reduction under Amanullah,
the army rose again from 40,000 to 70,000 men between 1934 and
1941, and with conscription after 1941 reached 90,000,15 in a remarkable contrast to the small complement of civil officials.
As in many countries, the army was a central element in the construction of the state. Owing to the numbers who underwent training, and as a model organisation, it had considerable influence within
the state apparatus. Its evolution may be traced in the light of two
issues, technical modernisation and recruitment.
Rather than real industrialisation, the amirs initially sought technological advance in the field of armaments. Openness to western
ideas in the 19th century was accepted, but was in principle limited
strictly to the technical sphere, and in particular to the manufacture
of arms, which could improve the equilibrium of the amirs power
with that of the tribes. Dost Muhammad (182639 and 184263)
was the first to call on foreign experts, who modernised his army
through the introduction of military uniforms and infantry. Under
Abdul Rahman Khan the closure of the frontiers, in spite of its disastrous effects on the economy and culturally, was no barrier to the
controlled importation of military technologies. The amirs workshops were restricted essentially to the manufacture of ammunition
13

Under the amir Shir Ali in 18778 the budget for the army already represented
43 per cent of the total. Hasan Kakar, op. cit., p. 88.
14 Leon Poullada, op. cit., p. 111.
15 Vartan Gregorian, op. cit., p. 296.

Neo-patrimonialism

29

and guns, and their extent was limited, with a workforce of 1,500 at
the beginning of the century, rising to 5,000 in 1919.16 In the 1940s
the state called on British technicians, but the major turning-point
came in 1955 with the Soviet-Afghan cooperation accords. These
related to the training of officers and the provision of supplies, and
effectively made Afghanistan a client of the Soviet Union.
An analysis of military recruitment displays two phenomena: the
gradual transition to conscription and the training of an officer
corps. In the 19th century several attempts to provide the amir with
forces independent of the tribal levies failed, since the military significance of the Durrani confederations horsemen greatly restricted
the choices open to the government. Nevertheless Abdul Rahman
Khan succeeded in establishing a professional army, paid and hierarchically organised, which was recruited from among the politically
most reliable groups. The process of recruitment followed the changing loyalty of the communities: the Hazara, the Qizilbash and the
Ghilzai, who were dominant up to 1880, gave way after various uprisings to a majority of Durrani. Nonetheless the amir recruited from
all the ethnic groups, stationing his troops outside their province of
origin, a practice continued under all later governments.
Until the 1940s recruitment was based on the supply of a number
of soldiers by the village or the clan, rather than being chosen by the
state. In particular Habibullah set up the system of hasht nafari (eight
men), i.e. the enlistment of one man in eight, chosen by the community. The transition to a conscript army was not achieved without
difficulties. Obligatory military service, decreed for the first time by
Amanullah in the 1920s, was rejected by the Pushtuns, which gave
rise to substantial over-representation in the army of other ethnic
groups. In 1941 the international situation allowed the government
to persuade the Loya Jirga (Great Council)17 definitively to accept universal military service, which led in turn to the abandonment of
16
17

ibid., p. 190.
The Loya Jirga is a council of tribal chiefs, convened for the first time by Mir
Wais at the time of the rebellion against the Safavid Empire. Later the Loya Jirga
became an assembly of tribal chiefs, religious leaders and notables gathered to
endorse a new sovereign or a constitutional change. Cf. Ludwig W. Adamec,
Historical Dictionary of Afghanistan, Metuchen, NJ: the Scarecrow Press, 1991,
p. 150; and Fida Yunas, Afghanistan, Jirgahs and Loya Jirgahs: the Afghan Tradition
(977 AD to 1992 AD), Pakistan, 1997.

30

The Sociogenesis of the Afghan State

community recruitment. However, the men of certain eastern and

Ghilzai Pushtun tribes were exempted from military service because
of their status as guardians of the frontier. In practice this left them
free to indulge in smuggling and participate in the destabilisation of
the Pushtun tribes on the other side of the Durand Line.
In the first half of the century officers were generally recruited
from groups which had links with the authorities: the pishkhedmat,
the gholam-i shah, the sons of notables (khanzadeh), and members of
the royal clan. Later the government of Daud (195363) set up military schools throughout the country. These were free, and became a
means of social advancement for many children from small merchant or peasant families.
The maintenance of patrimonial practices. However, the development of
institutions of state never resulted in a fully coherent system, nor in
the abandonment of patrimonial practices.After the 1950s the Afghan
state stressed its institutional character, but power remained largely
the personal prerogative of the king and the governing class. The
inheritance of power clearly illustrates the low level of institutionalisation. Abdul Rahman Khan laid down rules of inheritance, in
order to avoid the wars which had drawn the numerous claimants to
the throne into mutual conflict in the 19th century. This situation
was complicated by the fact that the amirs frequently had several
dozen children. However, up to the present there has been no
instance of a regular succession to the throne, except that of Abdul
Rahman Khan himself in 1901.18
18

In 1919 Amanullah in fact imposed himself by force against the opposition of his
uncle Nasrullah. Daud, Taraki, Amin, and Najibullah were victims of coups dtat.
Karmal (19806) was ousted by Najibullah on the direct orders of the Soviets,
without what could be called a coup, since the succession procedure, as under
other communist regimes, was not on a firm legal footing. The situation was just
as ambiguous in the case of Zahir who, though the legal heir after the assassination of Nadir, was unable to exercise real power before the 1960s. The coming to
power of Daud as Prime Minister in 1953, without there being a coup, symbolised a new balance of forces and allowed a new generation within the dynasty to
come to power (to the surprise of many, Daud agreed to go at the request of
Zahir in 1963). Even the legality of the presidency of Rabbani (19922001) was
disputed, and mullah Omar (19962001), whose government was not internationally recognised, lost power as the result of his defeat by the United States.

Neo-patrimonialism

31

Up till the 1970s the Afghan governing class comprised several

thousand individuals who, through their wealth or prestige, wielded
a decisive social and political influence. Within this class four groups
were distinguishable: the tribal aristocracy, the intellectuals (university personnel and religious dignitaries), administrators and merchants.19 However, the frequency of marriage between governing
class families, whose members were linked by extended networks,
meant that the division into distinct groups was not absolute, though
without the implication of any automatic convergence of attitudes
or agreement on decisions. Historically the primary element of the
governing class was the royal clan, the Muhammadzai, which had
distanced itself progressively from its tribal origins in order to achieve
a distinctive status within the state. The role of the Muhammadzai
dated from the reign of Abdul Rahman Khan. There was a change
in the attitude of the amir, particularly after the Ghilzai revolt of
18868, after which he systematically raised the prestige of the royal
clan, designating it as the sharik-i daulat (the partner of the state), thus
underlining the privileged position of its members.20 In the contemporary period the Muhammadzai were over-represented in decision-making positions, especially in the armed forces, because of
their supposed loyalty to the dynasty. In the last period of his presidency (19738) Daud, having split with the communists, continued
to seek the support of the royal clan, of which he was a member,
demonstrating the persistence of such loyalties even after the end of
the monarchy.
In the course of the 20th century, the development of the state led
to the gradual enlargement of the governing class, which became an
urban, and even a Kabuli, group. The predominant use of Persian
was an indicator of its increasing distance from the tribes, who were
Pushtu-speakers, although the Durrani aristocracy always maintained their links with Kandahar, their region of origin. With the
19

Abdullah Aziz, Essai sur les catgories dirigeantes de lAfghanistan, 19451963, BerneParis: Peter Land, 1987, p. 55. In addition, Leon Poullada assesses the lites (the
royal family, senior officials, tribal chiefs, wealthy merchants, major landowners)
at 23,000 people: see The Pashtun Role in the Afghan Political System, New York:
Asia Society, 1970 (Afghanistan Council Occasional Paper no. 1).
20 In particular they constituted the royal cavalry (risala-yi-shah-i qandahari). The
amir also employed the Muhammadzai in his personal guard. At moments of crisis, however, he turned more readily to the Safi or the Gardezi.

32

The Sociogenesis of the Afghan State

exception of the religious establishment, this class derived its cohesion from a common aspiration towards western-style modernity.
A foreign university education was commonplace for students from
these families, who frequently went on to find employment in the
higher ranks of the administration. The governing class also laid down
new norms of behaviour, particularly in dress and language. In addition the onset of industrialisation enabled its members to obtain dominant positions within the economy, thanks to their contacts within
the governmental structure. The national bank (Bank-i melli), established in the 1930s, operated in the first instance with private capital
and played a part in the creation of most of the industrial ventures of
significant size, such as the cotton company Spinzar. Not till the
1960s was there any development of smaller-scale industries, which
were never able to obtain sufficient capital.
Beyond the ranks of the national lites, the central authority
could also count on the loyalty of particular communities. From the
time of Timur, the amir employed two strategies, both normal within
imperial structures, to ensure the loyalty of his supporters. In the first
place, an external community, in this case the Qizilbash, who were
Shiites of Iranian origin, became the source of recruitment of court
administrators and of a military lite.21 In addition, the amir favoured
particular Afghan tribes, for example the Safi of Tagab, who was represented in the royal guard, and who provided Abdul Rahman
Khan with his most reliable support. In the contemporary period
the Nuristanis exemplified the attachment of a particular group to
the state. Settled on the Afghan-Pakistan frontier, and converted late
to Islam at the close of the 19th century, the Nuristanis, often recruited into the army, secured numerous material privileges. This
enabled them to develop their valleys while the relations of the
21

The Qizilbash are the descendants of a Shiite community established by Nadir

Shah in the Shindawal quarter of Kabul in the 18th century. They have served
various Afghan amirs in the administration and in the army. In particular, they
constituted the majority of the gholam-i shah. However, after supporting the
British in the war of 183942 they were obliged either to leave the city or to
practise taqiyya (religious concealment). From the reign of Abdul Rahman Khan
the Qizilbash gradually lost their influence, though they held on to a significant
role in the senior ranks of the administration. Many continued to practise a
reduced form of taqiyya (concealment), admitting to being Shiites only in
private.

The Sociogenesis of the Afghan State

33

neighbouring tribes, though Pushtun (the Safi), with the government were poor.

The legitimacy of the state

To ensure its survival the militarily weak state was obliged to avoid
disputes, and accordingly paid particular attention to issues of legitimacy. In this project the language of Islam and that of nationalism
were often employed simultaneously.
Islam. With varying degrees of success all holders of authority have
used Islam as a fountainhead of legitimacy, either primarily or secondarily, with the exception of the communists in 19789. However, in its attempts to lean on religious legitimacy, the Afghan state
has had a less convincing case than, for example, the Moroccan regime with a long tradition behind it.22 In Afghanistan the amir has
claimed Islamic legitimacy since the beginning of the 19th century.
It was laid down at that time that sermons in the mosques should be
preached in his name, and this obligation was reaffirmed in the constitutions of 1923 and 1931.23 When certain mullahs at the time of
the political disturbances in the spring of 1971 no longer preached
the Friday sermon (the khotba) in the amirs name, this constituted a
direct challenge to his legitimacy and as such brought retribution
from the authorities. But although the Afghan state has always
claimed religious legitimacy, its form has often varied greatly. From
Abdul Rahman Khan to Amanullah, legitimacy was seen as coming
directly from God, and the role of the ulema was merely to endorse
this. After 1929, however, clerical recognition became in itself the
source of legitimacy, and the ulema saw themselves as granting legitimacy to the ruler.
Under Abdul Rahman Khan and his successors Habibullah and
Amanullah the ulema were obliged to recognise the divine origin of
22

See especially Mohammed Tozy, Le prince, le clerc et lEtat. La restructuration

du champ religieux au Maroc in Gilles Kepel, Yann Richard (eds), op. cit., and
Rmy Leveau, Le fellah marocain, dfenseur du trne, Paris: Presses de la Fondation
nationale des sciences politiques, 1985, p. 261.
23 The Afghan constitutions of 1923, 1931, 1964, 1977 and 1987 all recognise
Islam as the official religion.

34

The Sociogenesis of the Afghan State

the power of the amir. This was a break with the tradition of their
predecessors,24 for whom legitimacy had been that of a primus inter
pares, the first amir having been elected by a tribal assembly at
Kandahar. In his memoirs25 Abdul Rahman Khan presented himself
as the bearer of a religious mission which he was obliged to fulfil in
the interests of the salvation of the Muslims of Afghanistan. He
strongly rejected tribal justification, favouring by contrast the Islamic legitimation of royalty.26 The practical consequences of this
rhetoric were very real. Afghanistan lost to some extent its character
as a Pushtun empire, and a degree of equality between its Muslim
subjects began to appear. After the suppression of the levy imposed
on non-Pushtun subjects (the sarmardeh), taxes were no longer in
principle variable as between different communities,27 but were equal
for all Muslims. Abdul Rahman Khan also used Islamic terminology
in the introduction of new taxes, which made them more comprehensible and conceivably more acceptable.28 This form of legitimation also allowed royalty to claim a monopoly of the right to declare
jihad29 and to condemn as kafir (heretical) all those who opposed
their power. This measure was also a means of reinforcing his control over the religious establishment. Here may be seen the reflection of the conflicts in which Abdul Rahman Khan found himself in
opposition to the ulema of Kandahar, who had declared a jihad
against him on the grounds that he was supported by the British.
To confirm this religious legitimacy, Habibullah (190119) was to
be solemnly consecrated as amir in a ritual devised by Abdul Rahman
24

This was not, however, a total innovation. See Nikki Keddie, Sayyid Jamal ad-Din
al-Afghani, Berkeley: University of California Press, 1972, p. 55.
25 The Life of Abdul Rahman Khan (1st ed. 1900), Karachi: Oxford University Press,
1980.
26 He also wrote a handbook on the rights and duties of the good Muslim, Zia ul
Mellat va Din (The light of the state and of the faith).
27 The Kandaharis were exempt, the Ghilzai and the Uzbeks paid little, and the
Tajiks were the most heavily taxed.
28 Taxes were justified, on the basis of an interpretation of the Quran, as a religious
duty. On Abdul Rahman Khans economic reforms see McChesnay, The economic reforms of Abdul Rahman Khan, Afghanistan XXI (3), autumn 1968.
29 Abdul Rahman Khan was certainly not the first to launch appeals to jihad. Dost
Muhammad did the same in 1939, and held the title of amir al-muminin (commander of the faithful).

The legitimacy of the state

35

Khan himself. The accession to the throne of Amanullah (191929)

followed the same pattern at the outset, and his declaration of a jihad
in 1919, at the time of the war of independence against the British,
increased further the popularity which he enjoyed in the early years
of his reign. Although he rapidly distanced himself from the ulema,
he maintained his religious legitimacy by posing during the 1920s as
a defender and potential inheritor of the Caliphate.30 In a manner
symptomatic of current thinking, a constitutional amendment of
1923, endorsed by the Loya Jirga (Great Council) in 1924, introduced
yet further discriminatory measures against non-Muslims, including
the obligation to wear distinctive clothes and the imposition of
higher taxes.
However, Amanullahs position was ambiguous because of his simultaneous claim to be espousing a modernising ideology. Thus the
1923 constitution, while confirming the hereditary monarchy, abandoned the principle of the divine legitimacy of royal authority in favour of a constitutional and nationalist model. In addition, the fall of
Amanullah in 1929 led to a substantial modification of the relations
of the ulema with the state, and their link to its legitimation. The
imamist concept of legitimacy, which accorded to the king the right
to govern in the name of God, was replaced by a more contractual
notion of legitimacy, which sprang from the states acknowledgement of the shariat, under the supervision of the ulema. Thus the
king was no longer the servant and protector of the true Islamic
faith as he had been under article 5 of the 1923 constitution.31 Indeed, after the revolt of 1929 the ulema, who had played a key role in
both the fall of Amanullah and the installation of the new ruler,
gained in independence and presented themselves as possessing the
autonomous authority necessary for the rulers legitimation. After
1929 the incorporation of the ulema into the highest ranks of the
state further endorsed the political weight of what was already an influential group. Significantly a Jamaat ul-Ulema (society of the ulema)
was established to certify the conformity of the states laws to the
shariat.32 Its role was significant in the early years,33 but then dimin30

Gail Minault,The Khilafat Movement,New York:Columbia University Press,1983.

Nighat Mehroze Chishti, Constitutional Development in Afghanistan, Karachi:
Royal Book Company, 1998, p. 67.
32 A similar step was taken in 1924 after the revolt of the Mangal.
33 We owe to it the translations of the Quran into Dari and Pushtu.
31

36

The Sociogenesis of the Afghan State

ished progressively with the arrival in power of Daud in 1953 and

the adoption of measures much disputed by the ulema, such as the
abolition of female veiling in 1959. In 1963 the role of the Jamaat
ul-Ulema was devolved to the department of Islamic affairs (Edareyi shoun islami) and then to a parliamentary commission, which
marked its definitive disappearance. By this time, however, the language of nationalism had already become dominant.
Nationalism. The idea of nationalism made its appearance in the first
decades of the century, with the establishment of a constitutionalist
movement (Mashruta Khawan).34 Mahmud Tarzi (18651933) was
one of the key figures of this movement with his periodical Saraj alAkhbar-i Afghaniyah35 (The Illumination of the Afghan News), which
was published from 1911 to 1918. The father-in-law of the future
King Amanullah, Tarzi was an intellectual, open to ideas of modernity, whose influence was to be crucial in the evolution of the intelligentsia at the beginning of the 20th century. The construction of
Afghan nationalism lay at the heart of the concerns of these reformers, who were concerned at the absence of national sentiment and
the countrys lack of development. The difficulty was to bring together nationalism and an acknowledgment of the principles of
Islam, in the context of a situation which somewhat resembled the
end of the Ottoman empire.
In the view of these reformers, national identity should be common to all the ethnic groups, and founded in the last resort on Islamic values. However, an underlying ambiguity remained: should the
direction taken be towards the recognition of the other ethnic groups,
or alternatively towards their assimilation to the dominant group on
the Kemalist model as seen in Turkey? The reformers opted in the
event for the construction of a national identity based on the Pushtun
ethnicity, and in particular the Pushtu language.36 The appeal to
Islam may therefore have justified, or even disguised, a project to
34

See Dr Fazal ur-Rahim Marwat, op. cit., pp. 99 ff.

See V. Gregorian, Mahmud Tarzi and Saraj-ol-Akhbar: Ideology of Nationalism
and Modernisation in Afghanistan, Middle East Journal 21 (3), 1967.
36 Significantly, the obligatory learning of Pushtu was advocated by Tarzi, in the
Saraj al-Akhbar-i Afghaniyah of 14 September 1914, quoted by S. A. Mousavi,
The Hazaras of Afghanistan: An Historical, Cultural, Economic and Political Study,
Richmond: Curzon Press, 1998, p. 158.
35

The legitimacy of the state

37

construct a Pushtun nation-state, and was coupled with a fierce criticism of the conservatism of the ulema, such as one could expect from
the modernists. In spite of its low circulation, Mahmud Tarzis periodical significantly assisted Habibullahs earliest efforts at liberalisation.37 Such modernising ideas were also later to evoke a sympathetic
response from Amanullah. In a sense this movement provided the
mould for the main reformist tendencies up to the coup of 1978.
The communists explicitly acknowledge it, and it is also possible to
identify it more indirectly as a precursor of the Islamists in view of
their modernisation and interpretation of Quranic texts outside the
exegetic tradition of the ulema.
Nevertheless, various constraints held back the states dissemination of a nationalist ideology, and as late as the 1920s the amirs did
not appeal to nationalism as a legitimation of their authority. In fact
it was religious legitimation which allowed them to rally the supportessential for the amirof a faction of non-Pushtuns. Abdul
Rahman Khan and Habibullah even encouraged the use of Persian
rather than Pushtu as an official language, and the language of education was Persian, even in the Pushtun areas. In no sense did Abdul
Rahman Khan owe his power to the Pushtun tribes. The Durrani
opposed him at the time of his accession to power, and most of the
tribal chiefs had been exiled to India.38 Above all, nationalism remained a concept alien to Abdul Rahman Khan, who relied on the
oppositions inherent in a segmented society rather than on broad
Pushtun solidarity, in a manner consistent with a principle of legitimation which remained religious. However, even without consistent
ideological justification, the government was dominated by the Pushtuns, who made use of the state as an instrument of internal imperialism, especially in the colonisation of the north, and with a tax
system which allowed certain privileges to persist. For example,
inequalities remained within the army: the Durrani regiments, and
especially those linked to the royal clan, were better paid than the
others, which gave rise to discontent.39
37

In Central Asia the reformist movement of the Jadids also paid close attention to
this publication. See Leon Poullada, op. cit., p. 42.
38 Moreover, the amir was the last Afghan sovereign to speak Pushtu, Turkish and
Persian. His exile to Russia probably promoted a less narrowly Pushtun view of
Afghanistan.
39 Hasan Kakar, op. cit., p. 113.

38

The Sociogenesis of the Afghan State

Nationalism became the state ideology when Nadir Shah40 was

confirmed as ruler in 1930, after the exile of Amanullah and the fall
of Habibullah II Bacha-yi Saqao (Son of the water carrier), the
Tajik rebel who seized power in January 1929 and held on to it for
nine months. The temporary loss of power by the Pushtuns displayed the potential leverage of the non-Pushtuns, leading the former to close ranks. The role of the tribes, and in particular of those
of the east, was crucial to the capture of Kabul by Nadir. A new terminology emerged from this readjustment of the balance of power:
since the king had been elected by a Loya Jirga dominated by the
tribes, the basis of his legitimacy could not be identical to that of his
predecessors. Nadir was chosen by the people, in recognition of his
service, as clearly laid down by the 1931 constitution, which only
the Loya Jirga had the authority to interpret or modify.
Royalty became once more the political focus of the tribes, and at
the same time there was a slide away from dynastic legitimacy and
towards Pushtun nationalism. The government was unable to control the ulema, and progressively turned to nationalism as a form of
legitimacy alternative to that derived from Islam. As an instance of
this, the legal distinctions between Muslims and non-Muslims were
dispensed with, which tended to promote a nationalist rather than a
religious definition of citizenship. Therefore, the legitimacy of the
state increasingly rested on nationalism, an ideology in which the
feeling of belonging to the national community was presented as
natural, since it was deemed to result from a shared history or even
a shared biological basis.41 In this way legitimation on the basis of
tribal genealogy evolved towards legitimation based on race. The
construction of a historical memory then became a central concern.
From the 1930s Pushtun nationalism, given its impetus by the intelligentsia, made use of history to demonstrate the common origin
and destiny of the populations which had their existence on Afghan
soil. Thus archaeology was pressed into service to establish the Aryan
origin of the peoples living in Afghanistan. The influence of the
Kemalist model was plain to see, as was that of nationalist or even
racist European theories from the inter-war period.42 The debate
40

From 1926 the king bore the title Shah, rather than amir.
Ernst Gellner, Nations and Nationalism, Oxford: Blackwell, 1983.
42 The rise of National Socialism in Germany was favourably received by the
Afghan nationalists, and especially by Daud, who then attempted to impose the
41

The legitimacy of the state

39

which took place within the Loya Jirga leading up to the adoption of
the 1964 constitution displayed clearly the misgivings of the nonPushtuns over these ideas. The non-Pushtun participants demanded
and obtained an amendment of Article 1 to reaffirm that the expression Afghan applied to all citizens and not exclusively to the Pushtuns (in common parlance Afghan was often taken to mean Pushtun).43
The linguistic issue took on a growing significance44 since the
state implemented a consistent policy of Pushtunisation which was
to continue till the 1970s. Although Persian was understood by a
large minority of the population either as a first or second language,
the government was to attempt to impose Pushtu as the national language,45 when hitherto its use had not been obligatory even within
the administration. The Pushtu Academy (Pashto Tulana) founded by
Amanullah expanded, and government publications increased in
number. In 1936 Pushtu was raised to the status of a national language, on an equal footing with the Persian of Afghanistanknown
as Dari to distinguish it, somewhat artificially, from the Persian of
Iran. In the same year the government imposed Pushtu as the language of education, before recognising this as unfeasible, leading to
the adoption of Pushtu/Persian bilingualism in 1946.46 In the 1960s
administrative measures were put in place with the target of publishing 50 per cent of written material in Pushtu.
This policy encountered the hostility of the non-Pushtu-speakers.
The debate in Parliament in 1971 on the requirement that officials
should learn Pushtu exposed once more the dichotomy between the
Pushtu-speaking deputies and the others. A further problem was
that the intelligentsia and the court spoke and wrote Persian, the traditional language of culture. King Zahir himself had only rudimentary Pushtu.
use of Pushtu, at the expense of Persian. See Mir Mohammad Sediq Farhang,
Afghanistan dar Panj Qarn-i Akhir (Afghanistan in the last five centuries), Peshawar: Derarsheh, 1988, p. 632.
43 Nighat Mehroze Chishti, op. cit., p. 96.
44 J. Petrusinska, Afghanistan 1989 in sociological perspective, Central Asian Survey, Incidental Papers Series no. 7.
45 Pushtu, meanwhile, was spoken by less than 10 % of the population as a second
language. See Encyclopaedia Iranica 1 (6), Costa Mesa, CA: Mazda, p. 504.
46 A distinction was drawn between the national language, Pushtu, and the official
language, Persian. See Shafie Rahel, La politique culturelle en Afghanistan, Paris: Les
Presses de lUNESCO, 1975.

40

The Sociogenesis of the Afghan State

In the army the recruitment of officers underwent a major transformation. Daud (195363) wished to use the army as an instrument
for the Pushtunisation of the state, and to that end he gave priory to
the recruitment of Ghilzai and eastern Pushtuns into the military
colleges. In the 1970s the young officers were Pushtuns, or less frequently Tajiks.47 The Hazaras and Uzbeks, at the bottom of the ethnic
ladder, were in practice excluded from the military profession.
This insistence on Pushtun nationalism at the expense of Islam
was a source of tension between communities. The non-Pushtun
populations, who were probably in the majority,48 felt themselves
excluded, while for the rural populations, whether Pushtuns or nonPushtuns, religious legitimacy was the only comprehensible principle. Under the presidency of Daud (19738) Pushtunisation entered
a more aggressive phase in relation to the other communities. For
example, the radio broadcasts in vernacular languages initiated at the
end of Zahirs reign were cancelled. In 19789 the extreme Pushtun
nationalism of the communists, in spite of their theoretical support
of minorities, was to play its part in the outbreak of the insurrection.

Challenges to the state

During the process of its formation the state often faced challenges.
To crush these, Abdul Rahman Khan relied on a police system, with
a network of spies, which deflected all forms of protest short of open
rebellion. For instance, journeys of more than a few kilometres outside Kabul had to be authorised by the police. It was not till the 1964
constitution that Afghans were given the right to free movement inside and outside the country. Breaking with the historiographical
tradition which has represented the amir of Kabul as a positive factor for Afghan unity, the historian Jonathan Lee was one of the first
to draw attention to the detrimental consequences of the hypercentralisation of the state, which was reduced to impotence by the
47

Hasan Kakar, op. cit., p. 211, and Barnett Rubin, The old regime in Afghanistan:
recruitment and training of a state elite, Central Asian Survey 10 (3), 1991.
48 Even if one finds texts where the Pushtuns are said to be in a majority, with the
effect of justifying a posteriori their domination of the political system. Census returns were in fact unreliable, and available data relating to ethnicity has not been
fully explored.

Challenges to the state

41

ill-health of the ruler, and of the campaigns of pacification with their

adverse effects on intercommunal relations.49 The reliance on policing was an indication of the fragility of the state, which was still
threatened by demands for autonomy from unruly communities,
and too weak to negotiate, even when this would have strengthened its legitimacy in the longer term. Habibullah, the successor to
Abdul Rahman Khan, taking advantage of his more certain exercise
of power, was able to abandon some of these police practices, but
conflicts once more became regular occurrences.
The challenge to the state came essentially from the tribes (Pushtuns, Uzbeks and Hazaras) and from the ulema. However, these oppositional thrusts were not necessarily in the same direction. The ulema
protected their privileges, especially in education and the judicial
system, but they did not dispute the necessity of the states existence
since they hoped to utilise itfor example, to impose the shariat in
place of tribal customs.50 On the other hand the tribes denied the
states right to play a part in their affairs, and in the case of the Pushtuns it was understood at best as the representative of the domination exercised by the tribes in Afghan territory. In unsettled times
the behaviour of the tribes tended towards pillage of the towns and
of state property. Thus, after the sack of Kabul in 1929 the eastern
Pushtun tribes were in no particular hurry to re-establish any kind
of central government.
The Tribes. When Abdul Rahman Khan (18801901) set himself the
task of ensuring by military means his domination over Afghan territory, a number of regions still enjoyed de facto autonomy. After the
military campaigns at the end of the 19th century the Ghilzai tribes,
which had been virtually independent till the 1820s, as well as the
Turkmens, the Hazaras and the Durrani of Kandahar, no longer represented a serious threat to the central government. In future these
49

See Jonathan Lee, Abd al-Rahman Khan and the maraz ul-muluk, Journal of
the Royal Asiatic Society 3 (12), 1991, pp. 20842. The amir is said not to have
suffered from gout but probably from cirrhosis of the liver, chronic encephalopathy and porphyria. 100,000 people are said to have been executed by the authorities during his reign. See F. Martin, Under the Absolute Amir, London, 1907.
50 A centralised authority was often favourable to the ulema. These played an increasing role within in the imperial bureaucracies of the Ottoman, Safavid and Mughal
empires.

42

The Sociogenesis of the Afghan State

tribal confederations would not provide the framework for political

mobilisation. Subsequently uprisings mainly concerned the eastern
Pushtun tribes, the only ones to have retained any military strength.51
This semi-autonomy was partly a result of the geographical position
of these tribes, which the state wished to be able to bring into play
against the British empire, and later against Pakistan.52
From the 1950s onwards the state had the military capacity to put
the tribes to flight. Although the Safi revolt of 19479 still proved
difficult to suppress, the troubles of December 1959 showed that the
balance of forces had changed. When the Mangal took arms in protest against the building of a road which threatened their income
and their independence, they were unable to stand their ground
against a mechanised army, and thousands fled into Pakistan from
where they only gradually returned. In the same year the state was
sufficiently strong to demand the payment of taxes by the Kandahar
landowners, who had traditionally refused to comply on the grounds
of an exemption granted to them in the 18th century. The police intervened and then the army, and in the end the state won the day,
even though the eastern Pushtun tribes continued largely to evade
payment of the tax.
However, the state did not always choose to clash head-on with
the power of the tribes. The simple act of imposing administrative
frontiers which did not accord with the boundaries of the tribal territories was in itself a way of undermining the strength of communal
attachments. The result was in some areas the substitution of solidarity based on location for solidarity founded on clan. Another factor
was that the villages were in any case less and less homogenous in
their tribal affiliations, except in the east. Existing communal antagonisms were also exploited, and sometimes encouraged. For example,
the divisions between the Hazaras and the Qizilbash were carefully
51

The most significant of these in recent times have been as follows: in 1933, the
Mohmand tribe; in 1937, the Mohmand, Shinwari and Ghilzai tribes in the east;
in 1945, incidents in the Kurram valley; in 19479 the revolt of the Safi tribe; in
1955 tribal stirrings near Kabul; riots at Kandahar and in the east in 1959; in
1968 the war between the Jaji and Mangal tribes; the agitation among the
Shinwari in 1970.
52 Similar phenomena occur in the case of the Kurds, on the periphery of the
Ottoman Empire. See Van Bruinessen, Evliya elebi and his Seyahatname, in
Evliya elebi in Diyarbekir, Leiden: E. J. Brill, 1988.

Challenges to the state

43

cultivated in order to avoid the emergence of a united Shiite front.

In the same way Amanullah played off the tribes against each other
to curb the Mangal revolt of 1924, although this strategy was to fail
in 1929.
Frequent transfers of population also facilitated control over the
tribes.53 From the reign of Abdul Rahman Khan, there were movementsboth forcible and voluntaryof predominantly Pushtun
groups to the north of the country. The aim of this interior colonisation was to ensure the security of the frontiers and exert more
effective control over tribes prone to rebellion against the central authority. The government granted lands to the colonists, and afterwards traded on Pushtun solidarity in the face of the dispossessed
indigenous populations. The earliest migrations began in 1883, with
a programme of colonisation in Badghis, the objective of which was
to counter the threat of a further Russian annexation. The failure of
this initiative after the loss of Panjdeh in 1885, gave rise to a more
ambitious programme, using Pushtun nomads rather than the Jamshidis, who were prone to be tempted by Russian offers of an accommodation. Colonisation was undertaken thereafter with varying
degrees of success across the whole of northwest Afghanistan. Transplantation was sometimes punishment for a rebellion, as in 1890
when a penal colony of Suleiman Khel was set up at Baghlan. In addition, the populations involved were not always Pushtuns: at the beginning of the 20th century groups of Uzbeks were forcibly installed
in the neighbourhood of Kabul and Jalalabad. The final forced transfers seen to have been those of the Safi after the revolt of 19479.
However, from the 1920s the exploitation of new lands in the north
at Kunduz led to the settlement, though here on a voluntary basis, of
several thousand Pushtun families. Interior migrations organised by
the state continued till the 1960s, making their contribution to the
strength of the central government, which from this point on was in
a position to arbitrate between the different communities.
Under Abdul Rahman Khan the significance of the tribes was
acknowledged by the existence of a darbar-i shahi (council of tribal
chiefs), whose deliberations, were strictly consultative. After the events
of 1929 a strategy of integrating the tribes into the national institu53

Richard Tapper (ed.), The Conflict of Tribe and State in Afghanistan, London:
Croom Helm, 1983.

44

The Sociogenesis of the Afghan State

tions was declared. The 1931 constitution provided that Parliament

should sit from May to October, which necessitated the presence in
Kabul of the principal tribal chiefs. The Shah thus hoped to limit the
risks of rebellion by gathering together the tribal leaders between
the seasons, after the end of the years agricultural labour, at a time
which otherwise lent itself to the mobilisation of the tribes. In addition, in recognition of the role of the tribes, a Loya Jirga was supposed to meet at least once every three years, although this was not
always to be observed, especially as the council duplicated the work
of Parliament. Finally, the institution of a Minister for the frontiers
was evidence of the states vigilance over the turbulent eastern Pushtun tribes.
In one case at least, internal conquest left a trauma whose effects
are still felt today. No campaign was as difficult as that which achieved
the subjugation of the Hazarajat, the central region of Afghanistan
populated by the Hazaras.54 The origin of the Hazaras remains relatively obscure. Contrary to a popular supposition they are not largely
the descendants of the armies of Genghis Khan, as their Mongoloid
physical appearance, which contrasts strongly with the Pushtuns and
Tajiks, might suggest. They are actually an ancient group, with Turkish and Mongol links.55 They speak a dialect of Persian, known as
Hazaragi, in which Turkish and some Mongol words occur. The total
population of Afghanistan is not precisely known, still less its ethnic
composition, but most writers agree that the Hazaras constituted
some 10% of the population, that is to say more than a million people, in the 1970s. They make up 80 per cent of the Shiite population, the remainder of whom are Qizilbash and Tajik.56
The date of their conversion to Shiism is unknown, but it was
probably in the first part of the 16th century, in the early days of the
Safavid dynasty (15011786).57 In any case, not all the Hazaras are
54

In particular the provinces of Bamyan, Ghor, Uruzgan, Wardak, Ghazni and

Jozjan. For a history of the Hazaras see Hasan Poladi, The Hazaras, Stockton, CA:
Mughal Publishing Co., 1989.
55 See S. A. Mousavi, op. cit., pp. 19 ff. The Hazaras do not always have pronounced
physical characteristics, and their social identification is more in terms of cultural
markers. There are also regional variations, perhaps the result of intermarriage
with Pushtuns and Tajiks.
56 Persian-speaking non-Hazara Shiites are found on the frontier with Iran, mainly
in Herat and the province of Nimruz.
57 Hasan Poladi, op. cit., p. 118.

Challenges to the state

45

Shiites. Sunni Hazara communities are to be found to the south of

Mazar-i Sharif, in Dara-i Hazara (a valley adjacent to Panjshir),
Rustaq and Takhar, where Babur reports their presence from the
16th century.58 However the non-Shiite Hazaras tend in general to
repudiate their ethnic allegiance in favour of the claim that they are
Tajiks, on account of the widespread conflation between Shiites
and Hazaras, with their consequent stigmatisation by the Sunnis.
The claim to be an ethnic Hazara implies in practice an affirmation
of Shiite faith. In this case ethnic and religious affiliations reinforce
one another, while on the other hand there are no links of solidarity
between the Sunni and Shiite Hazaras.
Though they recognised their attachment to the kingdom of
Kabul, to which they paid tribute, the allegiance of the Hazaras to
the amirs who succeeded up to the end of the 19th century was
largely nominal. At times the unification of Hazarajat under the authority of a tribal chief even threatened the authority of Kabul.
When Mir Yazdan succeeded in pacifying Hazarajat under his leadership from 1843 to 1863, Dost Muhammad had him imprisoned.
At the end of the 19th century the central authority could no longer
tolerate the de facto independence of Hazarajat and the banditry of
some of its tribes. In consequence the amir of Kabul waged a series of
campaigns, mainly between 1891 and 1893, which resulted in a real
integration of Hazarajat into the Afghan state.59
This war, similar at its outset to the campaigns carried out in
Turkestan or against rebel Pushtun tribes, turned into a bloody repression of which the memory persists even today. Religious mobilisation explains in part the tenacity of the two sides. The declaration
of jihad by the amir of Kabul, and that of the Shiite ulema in response,
was to justify the worst atrocities, and in particular the enslavement
of a segment of the Hazara population; Hazaras were sold in the
markets of the capital as late as the first years of the 20th century. The
amir also sent Sunni mullahs to convert them.60 Already displaced
58

Zahiruddin Muhammad Babur, Babur-Nama (trans. Annette Beveridge), London, 1922; reissue, Lahore: Sang-e Meel, 1987. Sunni Hazaras also exist in significant numbers in Badghis, in the region of Qala-i Naw, where they were settled
by Nadir Shah.
59 See Hasan Kakar, The Pacification of the Hazaras of Afghanistan, New York:
Afghanistan Council, the Asia Society, 1973.
60 Louis Dupre, The Political Use of Religion: Afghanistan in K. H. Silvert (ed.),

46

The Sociogenesis of the Afghan State

from their mountains by the Pushtun expansion of the 18th and

19th centuries, the Hazaras lost a significant part of their cultivable
land. Within a number of years a hierarchical tribal system had given
way to a disorganised assembly of notables of whom the state made
use as local agents. These traumatic events were to give rise to a
strong sentiment of identity which was to have its consequences in
the present conflict.
Following these events, amir Habibullah (190119) proclaimed an
amnesty. However, few Hazara returned from exile and it was not
until the accession to the throne of Amanullah, who repudiated antiShiite discrimination, that some of those who had fled came back.
In the rebellion of 1929 the Hazaras actually supported the king.
Significantly, slavery was abolished by decree in 1921, a measure
confirmed in the 1923 constitution. However, the Hazaras, though
defeated, remained an element unassimilable by Pushtun nationalism,
while their adherence to Shiism placed them in opposition to the
Sunni ulema.61
Alien to the two modes of legitimation of the Afghan state, and
relegated to the bottom of the ethnic hierarchy, the Hazaras suffered
from social and economic discrimination. The nomadism of the Pushtun tribes in the central regions of Afghanistan was a particular source of
oppression against them. For example, during their seasonal displacements a police agent normally accompanied the Pushtun nomads,
and conflicts with the sedentary Hazaras were generally settled in
their favour. Trade with the Hazaras and the lending of money at
usurious rates of interest also enabled the nomads to seize their land.
In the 1970s the nomads went less often to Hazarajat to trade because
of the circulation of trading lorries from the neighbouring towns,
but the competition for land grew more intense because of increasing demographic pressure.
Descriptive and anecdotal observations all point in the same direction.62 In the 1960s and 1970s the Hazaras land was no longer
Expectant People: Nationalism and Modernization, New York: American Universities Field Staff Report, 1967.
61 The Constitutions of 1923, 1931 and 1964 recognised Hanafi Islam as the official rite. The Constitution of 1977 marked a concession to the Shiites by omitting this specification.
62 The cultivation of intensively terraced mountain plots in a good indication of
the pressure for land. See also M. Gaweki, The Hazara farmers of central Afghan-

Challenges to the state

47

sufficient to feed them, which seemed to be the result of two factors. First despite particularly high infant mortality owing to worse
sanitary conditions than elsewhere,63 demographic growth was no
longer containable. And secondly, the Pushtuns developed the raising of flocks and herds, restricting the amount of land usable by
Hazara peasants. From the 1960s poverty led many of the Hazaras to
move to the towns, mainly Kabul and Mazar-i Sharif. In due course
this phenomenon of rural exodus accelerated, displacing the centre
of gravity of the community towards the towns where the Hazaras
became a community generally looked down on, often working as
bakers or as unskilled workers.
These circumstances explain why during the past century uprisings in Hazarajat have been frequent, and popular recollection preserves the memory of the most notorious figures associated with the
resistance to the state.64 During Amanullahs reign, Naim Khan rebelled, together with his two sons, because of the encroachments of
the Pushtun nomads. He was taken captive at Panjao and imprisoned in Kabul. Another well-known chief, Yusuf Beg, a khan from
Shahristan, fought for nineteen years against the government before
finally being arrested. He was taken to Kabul and executed on the
orders of the Prime Minister Hashem Khan (193346). Books about
Yusuf Beg, published in Teheran by the Hazara community, are still
circulated in Afghanistan. However, the most celebrated figure remains Muhammad Ibrahim Khan, a khan from Shahristan known as
Bacha-yi Gaw Sawar (the boy riding the bull). His nickname comes
from a tale told about him in which he is said to have ridden on a
bull which ascended to paradise when its right ear was pulled. During the winter of 19456 Ibrahim Khan headed a rebellion against a
new tax on animal fats. The police post at Shahristan was occupied,
and the central authorities lost control of the district for a whole
winter. In the spring a delegation of Hazara notables led by the govistan: some historical and contemporary problems, Ethnologia Polona 6, Poland
1980.
63 There were only some ten dispensaries for the whole of Hazarajat, and no doctors outside the NGOs.
64 David Busby Edwards, The Evolution of Shii Political Dissent in Afghanistan
in Juan R. I. Cole, N. R. Keddie (eds), Shiism, and Social Protest, Yale University
Press, 1986.

48

The Sociogenesis of the Afghan State

ernor of Kabul came from the capital to offer terms. The government cancelled the tax, but Ibrahim was taken to Kabul and placed
under house arrest. He was accused, together with Ismail Balkhi (see
below) of planning a coup and imprisoned till the 1960s. He died
soon after his release but remained a well-known figure in Hazarajat.
In addition to these revolts led by prominent figures, disturbances
were commonplace, often connected to the presence of the nomads.
The leaders of these uprisings were local khans: apparently there
were no popular uprisings. Uruzgan, the home territory of Ibrahim
Khan, Yusuf Beg and Naim Khan, seemed most susceptible, on account of the Pushtuns settled in the south of the province, and also
because of the minimal presence of the state in this inaccessible region.65 It would be unsafe to claim that these men were the forerunners of Hazara nationalism, especially where documentary records
are so sparse. However, a generation later the first Hazara nationalists,
often the sons of the khans, who by this time were students in Kabul,
held them up as symbols of resistance to the Pushtun state.
The ulema. In contrast to the tribes, who were progressively marginalised by the military power of the state and by social evolution, the
men of religion came to represent after the 1950s a real counterweight to the established authority. The ulema constituted a social
group with its own system of education, both private and governmental, as well as its own transnational networks, and its own material and moral interests, which provided a mobilising impetus. In
contrast to the ulema, the mullahs66 were not a well-defined or homogenous class. There is no priesthood in Islam. A person who is qualified to lead the prayers is known as an imam, and as a mullah if he
fulfils this function regularly and professionally. However, mullahs
educated in the madrasas displayed solidarity with the ulema, especially those whose principal activity was to lead the prayers.
65

The province of Uruzgan long remained under the domination of the khans,
with a strong tendency to conflicts, both internal and with the Pushtuns.
66 In Central Asia and Turkey, as distinct from the Arab countries, the expression
mullah was not restricted to Shiites. The mullahs, less well integrated into the
administration, were sometimes paid by the government, but more often remained in charge of village communities. If there was no school in the village,
they would offer an elementary education to the children. See The Encyclopaedia
of Islam, Leiden: E. J. Brill, 1993, p. 223.

Challenges to the state

49

In the 19th century the ulema were involved in governmental affairs

only in the context of the initial legitimation of a ruler,67 although
they might occasionally play a part in the proclamation of a jihad, or
in opposition to administrative measures deemed to be at variance
with the Quran.68 Through their often conflictual relationships with
the state, the men of religion affirmed their status as a coherent group
with an identity and with collective interests. The state sought to
integrate the ulema, but the influence of the great religious families,
together with the effects of fundamentalist ideological tendencies,
motivated them to retain their autonomy while mobilising against
modernising reforms.
The instrumentalisation of Islam as a source of legitimacy accounts
for the effort made by the state from the close of the 19th century to
integrate the mullahs and the ulema into the administration. Abdul
Rahman Khan assimilated the ulema up to a point into the administrative system by the institution of qualifying examinations, while
mullahs participated in the indoctrination of soldiers in all the regiments of the army. In 1896 Abdul Rahman Khan attempted to
circumscribe the economic autonomy of the ulema and the brotherhoods (tariqat) by taking partial control of the religious endowments
(the waqf ), especially in the towns.69 In certain cases the amir did not
hesitate to punish those reluctant to comply: after the Ghilzai revolt
of 18868 he cut the emoluments of the ulema in the Ghazni region.
The role of the state was later reinforced by Amanullah, and then in
the 1960s by the Etemadi government, which set up a waqf depart67

Because of the absence of strict hierarchy among the ulema, a claimant to the
throne always found a way to ensure his recognition by some group from among
them. At the beginning of the 18th century Mir Wais, in order to justify his revolt against the Safavid Shiites, procured a fatwa from Mecca, which enabled
him win the help of the tribal chiefs of Kandahar.
68 In the 1830s the invasion of a non-Muslim power provided the opportunity for
the ulema, in declaring a jihad, to demonstrate their influence.
69 See Donald Wilber, The Structure of Islam in Afghanistan, Middle East Journal
6 (1), 1952, pp. 418; Ashraf Ghani, Islam and State-Building in Afghanistan,
Modern Asian Studies 12, 1978, pp. 26984. Similar phenomena are found in Iran
(see N. Keddie, Scholars, Saints and Sufis, Berkeley: University of California Press,
1972, p. 7) and in Turkey (see Albert Hourani, Ottoman Reform and the politics of Notables in W. R. Polk, R. Cambers, eds, Beginnings of Modernization in
the Middle East, University of Chicago Press, 1968, p. 58). The clergy continued to
be significant landowners in certain areas, especially, it would seem, in Kandahar.

50

The Sociogenesis of the Afghan State

ment within the Ministry of Justice. In 1970 the state reduced once
more the influence of private individuals by placing certain of the
pilgrimage sites (ziarat) under direct government control.70
However, education was the crucial factor in controlling the ulema.
Before the modern period there was no training centre for them in
Afghanistan. Before the Bolshevik Revolution of 1917 they were
normally educated in Central Asia and till the 1920s at Deoband in
northern India, where Afghans were the second largest group of foreign pupils.71 At the beginning of the 20th century the Afghan amir
went as far as to fund construction at Deoband, and offered an annual
contribution, which the Deobandi school refused in the interests of
preserving its independence. In the 1920s Amanullah attempted to
curb the influence of these educational centres, which were notorious for the their dogmatic rigidity. Education at Deoband, and at
certain Central Asian madrasas was declared no longer to be a qualification for official positions.72 From the 1940s onwards a national
system of religious education was set up by the state, with the aim of
facilitating the integration of the ulema into the state machinery, and
of maintaining control of the content of the teaching provided. The
creation of the School of Shariat in 1944 was complemented in
1951 by the establishment of a faculty of theology, linked to that at
Al-Azhar University in Cairo. The governmental madrasas, opened
one by one in the principal towns of the country, trained ulema who
in due course were attached to the various organs of the state, serving in particular as judges, as academics, or by taking charge of religious education in the secondary schools.
Nevertheless, this governmental enterprise was always limited in
its extent, and by the 1970s the majority of the ulema were still produced by the private madrasas, which had their continued existence
guaranteed by the 1931 constitution. The government madrasas were
in fact to produce the Islamists who were the most antagonistic towards the central authority. This issue is discussed below.
70

See Louis Dupre, Saint Cults in Afghanistan, American Universities Field Staff
Report, South Asia Series XX (1), May 1976. The Charter of the Awqaf Administration of 1969 is given in an appendix.
71 Barbara Daly Metcalf, Islamic Revival in British India: Deoband 18601900,
Princeton University Press, 1982.
72 Amanullah was later obliged to rescind this measure.

Challenges to the state

51

The education of the Afghan ulema was therefore carried out largely
outside the governmental structure, which accounts for the influence of the reformers of the Indian subcontinent. From the beginning of the 19th century, the fundamentalist philosophy inspired by
the Indian reformer Shah Waliullah (17031762) was found in Afghanistan. Although these reformers were stigmatised by the British as
Wahhabites, there is no historical proof of any direct affiliation to
the Arab movement, and a separate origin is more probable. One of
the sources of inspiration common to both Shah Waliullah and the
Salafists, who rejected the designation Wahhabi, lay in the writings
of Ibn Taymiyyah,73 who condemned the excesses of popular religious fervour inspired by Sufism,even though he was himself affiliated
to the Qadiri tariqat. It can be seen with hindsight that he was especially significant for his theory of takfir (apostatisation), although he
failed to define with complete clarity in what conditions a Muslim
could be regarded as an apostate, and did not call for rebellion against
existing Muslim governments. Ibn Taymiyya was nevertheless one of
those rare theologians whose rhetoric did not systematically justify
the status quo. In this he differed from his teacher Ibn Hanbal, who
refused to recognise a right of resistance even to a government which
ordered actions contrary to divine ordinances. Ibn Taymiyya, on the
other hand, recommended disobedience and even revolt. A respected though marginal author, he has since been claimed as an authority both by reformers of the school of Waliullah and by Islamist
writers such as Sayyed Qutb and Maududi, who are not themselves
ulema and find in him a justification for their political theories.
Such reformist ideas were found particularly in the teaching given
at the madrasas of Deoband and Patna in northern India, which educated part of the Afghan ulema. The Deobandis taught strict observance of the Islamic ethical code, and were opposed to the British
presence, as well as to any syncretistic approaches towards Hinduism.
Distinguishing themselves from the school of Deoband, the Ahl-i
Hadith adopted a particularly radical standpoint, and had maintained
relations with the Salafis from the mid-18th century.74 They preached
73

On Ibn Taymiyya see E. Sivan, Ibn Taymiyya: Father of the Islamic Revolution,
Encounter, May 1983.
74 Marc Gaborieau, Nicole Grandin, Le renouveau confrrique (fin XVIIIe
XIXe sicle) in Alexandre Popovic, Gilles Veinstein (eds), Les voies dAllah, Paris:
Fayard, 1996.

52

The Sociogenesis of the Afghan State

the destruction of the tombs of pirs to prevent them becoming places

of worship, but they were not opposed to Sufism as such; indeed
controversy over the links between Sufism and Salafism was often
deceptive, since even among the Salafis there existed a minimal Sufi
current. Consequently, from the time of Sayyed Ahmad Barelwi
(17861831) it would be preferable to speak of reformist Sufism, as
opposed to the popular practices involved in the worship of saints.75
What influence did these movements have over the Afghan
ulema? Deobandi tendencies have long been present in Afghanistan,
where Deobandi texts have been known since the 19th century. For
instance, in 1912 the Pushtuns of Karachi, under the influence of the
Afghan amirs representative, took the side of the Deobandis in a
controversy with those who adhered to the orthodox Hanafi school
over the status of the Prophet Muhammad.76 In the course of the
discussions concerning the 1931 constitution certain ulema disputed
Article 6, which includes a reference to saints in the context of the
Shahs oath of fidelity, thus adopting a typically reformist position.
After the partition of India certain of the Afghan ulema, who
would traditionally have gone to Deoband, were instead educated at
madrasas in the North-West Frontier Province of Pakistan, where the
Deobandi influence was strong. In 195960, 15 per cent (60 out of
397), of the students at the madrasa Dar al-Ulum Haqqaniyah were
Afghans, a proportion which grew to 37 per cent in 1970 (204 students out of 550).77 Movements of the most fundamentalist kind also
exercised regional influence. In particular many ulema in the frontier
provinces of Badakhshan and Kunar were educated in madrasas identified with the Ahl-i Hadith and the Deobandis. In Afghanistan the
Najm ul-Mudaris madrasa near Jalalabad also served as a channel for
fundamentalist tendencies close to this movement.78 Nevertheless, in
75

Frederick de Jong, Bernd Radtke (eds), Islamic Mysticism Contested: Thirteen Centuries of Controversies and Polemics, Leiden: E. J. Brill, 1999. Marc Garborieau, A
Nineteenth Century Indian Wahhabi Tract against the Cult of Muslim Saints:
al-Balagh al-Mubin in Christian W. Troll (ed.) Muslim Shrines in India, Delhi:
Oxford University Press, 1990. Marc Gaborieau and Nicole Grandin, op. cit.,
object to the term neo-Sufism, which implies a break with tradition.
76 Sahrah Ansari, Sufi Saints and State Power: the pirs of Sind, 18431947, Lahore:
Vanguard Books, 1992, p. 80.
77 See Jamal Malik, Colonization of Islam: Dissolution of Traditional Institutions in Pakistan, Lahore: Vanguard Books, 1996, p. 207.
78 Olivier Roy, op. cit., pp. 75 ff.

Challenges to the state

53

pre-war times the majority of Afghan ulema, who followed the

Hanafi rite, were opposed to the Wahhabis, lumping together under that term many of the reformists.79 For example they rejected
the doctrine of the Ahl-i Hadith, since the cult of saints remained
very popular in Afghanistan.80
The dominant faction in the pre-war period professed adherence
to a version of Islam close to the Sufi and Deobandi tradition. The
Mujaddidi family was a typical rallying point for the traditionalist
ulema. This family was also behind the initiation of a modernising
reform in the teaching of the private madrasas. Indeed, in the 1930s
Fazl Omar Mujaddidi founded the Nur ul-Mudaris madrasa, in the
Ghazni province, which was the venue of a genuine reconstruction
of private religious education. In reaction to the setting-up of governmental madrasas his son Ibrahim in the 1970s introduced modern subjects, including science, sport and English, equipping a new
generation of ulema who were able to compete with those from the
government schools.
The autonomy of the ulema in relation to the state was also linked
to the role of a number of influential families, notably the Mujaddidi
and the Gaylani, who dominated the informal Afghan religious hierarchy and played a leading part in all efforts to mobilise them as a
group. From these two families came the most renowned ulema and
the most influential pirs of the two main religious brotherhoods
present in Afghanistanthe Naqshbandiyya and the Qadiriyya.81
The Gaylani and the Mujaddidi were not originally from Afghanistan, but their prestige had allowed them to lead the local ulema.
Qayyum Jan Mujaddidi, who came from India in the early 19th century, set up a madrasa in the Shor Bazaar, from which arose the name
by which he and his descendants were known: Hazrat Sahib-i Shor
79

Here the ulema enjoyed the support of the State, which was apprehensive of
antagonistic positions. As late as the 1970s an alem in Kandahar had Wahhabi
preachers imprisoned (interview, Kandahar, 1992).
80 See Nighat Mehroze Chishti, op. cit. On the persistence of non-Islamic traditions,
see Marc Gaborieau, Typologie des spcialistes religieux chez les Musulmans du
sous-continent indien: les limites de lislamisation, Archives des sciences sociales des
religions 55 (1), 1983, pp. 2251.
81 On the Naqshbandiyya, see Marc Gaborieau, Alexandre Popovic, Thierry Zarcone (eds) Naqshbandis. Cheminements et situation actuelle dun ordre mystique musulman, Istanbul-Paris: Isis, 1990.

54

The Sociogenesis of the Afghan State

Bazaar.82 There were also branches of the family in Herat, Logar and
Ghazni provinces. The second important family, the Gaylani, arrived
in Afghanistan in 1905. Other branches of this family were previously in Afghanistan, but the last to arrive achieved dominance.
Sayyed Hasan Gaylani83 had moved to Afghanistan because of differences with the pir of the Qadiri brotherhood in Iraq. He was well
received by amir Habibullah, who granted him a pension, and settled
near Jalalabad where he became the principal pir of the Qadiri brotherhood in Afghanistan.
These two families, though often opposed to the authorities, belonged to the governing class, and their strategy of forging matrimonial alliances with the royal clan earned them an important position
among the lites. The Mujaddidi, who adhered to the Naqshbandi
tradition, were awarded a role as the counsellors or arbitrators of the
ruler, and in particular took upon themselves the right to confer
legitimacy on new sovereigns and to adjudicate on whether their
actions conformed to Islamic principles.84 The Gaylani, who were
much less confrontational, also had less influence among the ulema.
Their ties with the royal clan were close.
For the Shiites the centres of education have always been outside
Afghanistan, particularly at Qom in Iran and Najaf in Iraq, and the
authorities made no attempt to integrate their religious establishment.
82

Qayyum Jan Mujaddidi was descended from Shaikh Ahmad Sirhindi, who was
born in Kabul in 1564 and buried at Sirhind in India in 1624. Shaikh Ahmad
Sirhindi was recognised as the Mujadid alf-i thani (the bringer of millennial renewal), since according to Muslim tradition a reformer periodically arises to
bring religion back to its principles. See Ludwig W. Adamec, Historical Dictionary
of Afghanistan, London: Scarecrow Press, 1991, p. 167. According to family history Qayyum Jan arrived at the end of the 18th century, see David B. Edwards,
The Political Lives of Afghan Saints: the Case of the Kabul Hazrats in Grace
Martin Smith (ed.), Manifestations of Sainthood in Islam, Istanbul: Isis Press, 1993.
83 A descendant of the founder of the Qadiri brotherhood, pir Baba Abdul Qadir
Gaylani (10771166), was born in Baghdad in 1862. The eldest son of Sayyed
Hasan, Sayyed Ali, who later inherited the position of pir, died in 1964, and his
successor was Sayyed Ahmad Efendi Sahib Gaylani. On the frequent visits by the
Gaylani of Iraq to India at the close of the 19th century, which aroused the suspicions of the British, see Glhan etinsaya, Ottoman Administration of Iraq,
18901900, Ph.D thesis (unpublished), University of Manchester, 1994, p. 55.
84 This Naqshbandiyya strategy of acquiring influence is evident in the significant
number of senior officials who are Naqshbandis, especially from the Barakzai
and Sadozai tribes.

Challenges to the state

55

A genuine cultural renaissance manifested itself in some madrasas

from the 1960s onwards. Ismail Balkhi (192268), an intellectual
known for his abilities as an orator, was one of the leading figures of
this movement.85 Sayyed Beheshti attracted the taliban (students) from
the entire Hazarajat to Waras. At Kandahar, the cultural movement
Sobh-i Danesh (dawn of knowledge), inspired by Asef Muhseni, made
available a modernist education; in addition to strictly religious topics,
the movements monthly publication dealt with social issues, Marxist dialectic and other subjects.86 In the majority of cases training in
Iran or Iraq was part of this renaissance, which gave rise to a growing politicisation of the students emerging from the madrasas. The
politicisation of the Shiite clergy preceded the Iranian revolution.
The autonomy of thought of the ulema and their organisation
accounted for the frequency of clashes with the state. From 1919 the
ulema opposed Amanullah, who wished to speed up the process of
modernisation. In 1924 Amanullah, fearing a plot inspired by the
British, arrested pir Mujaddidi and executed a number of ulema. In
1929 the ulema played a key role both in the spread of the rebellion
and the seizure of power by Nadir. From the 1950s opposition once
more intensified between the ulema and the authorities, who had resumed the modernisation of society. Some fifty mullahs, of whom a
number were members of the Mujaddidi family, were imprisoned as
the result of their opposition to the unveiling of women in 1959, a
measure especially symbolic since Amanullah had already included it
in his programme of reform. The parliamentary regime which took
office in 1964 afforded the opportunity for a substantial number of
men of religion to enter Parliament, around a quarter of the first assembly.87 The ulema and their taliban (students), under the influence
85

Sayyed Muhammad Ismail Balkhi, originally from the province of Jozjan, was a
remarkable intellectual and preacher who studied at Mashad and Qom. His
speeches calling for equality for the Shiites, as well as for the democratisation of
the country, had an impact in student and cultured circles beyond the Shiites
themselves. He was arrested in 1946 for an attempted coup, and remained in detention for fourteen years. His influence on the Shiite movement as a whole in
Afghanistan was critical. On the politicisation of these groups see David Busby
Edwards, The Evolution of Political Dissent, op. cit.
86 See Olivier Roy, op. cit., p. 202.
87 Louis Dupre, Comparative profiles of recent parliaments in Afghanistan,
American Universities Field Staff Report XV (4), July 1971.

56

The Sociogenesis of the Afghan State

of Ibrahim Mujaddidi, who succeeded his father in 1956, were at

that time organised into an informal group known as the Khodam
ul-Forqan (servants of the Quran).
In 1970 a number of ulema were again imprisoned, this time after
clashes with Marxist students. In the spring of the following year, religious figures from Kabul and its neighbourhood organised the
largest demonstration so far seen in the capital, involving tens of
thousands of people. The protest was directed against the appearance
in a communist journal of an article deemed offensive to religion,
and in particular against the use in celebratory coverage of the hundredth anniversary of the birth of Lenin of the term dorud (greetings), an expression reserved for the Prophet Muhammad. Ibrahim
Mujaddidi and Gholam Muhammad Niazi were the instigators of
the movement, which demanded an end to the education of women
and a return to veiling (purdah),88 while the protests of the mullahs,
who roused the tribes in Laghman and Nangrahar, induced the government to negotiate. It will be seen that at this time the political
parties enjoyed nothing like the same ability to mobilise.
As has often been the case in the Muslim world, law and education were the two points of friction with the state. The ulema were to
an extent marginalised in both these fields, above all by the overtly
modernist policies of Daud (195363). In the judicial domain relations between the ulema and the state were initially good. Aiming to
counter tribal and customary law, Abdul Rahman Khan widened
their role, in particular through the nomination of qadis (judges) in
all districts. Ashraf Ghani has shown in a study of the case of the
province of Kunar how the qadis, closely controlled by the state, were
able by employing a jurisprudence antithetical to customary law profoundly to modify the way in which conflicts were resolved.89 Nevertheless, many issues continued to be settled outside the legal system.
88

Leftist movements then mounted a demonstration involving several thousand

women. Shots were fired and a number of the demonstrators were burned with
acid, in Kabul and other towns. See, Linda Clark Richter, The impact on
women of regime change in Afghanistan, Journal of South Asia and Middle Eastern Studies 3 (2), winter 1983, p. 6.
89 With the result that women were empowered to become more autonomous
agents, while they were excluded from customary law. See Ashraf Ghani, Disputes in a court of Sharia, Kunar Valley, Afghanistan, International Journal of Middle Eastern Studies 15, 1983.

Challenges to the state

57

The recent subjection of Hazarajat and Nuristan, which had been

Islamised only at the close of the 19th century, also tended to enhance the power of the ulema, who issued their judgements independently with a right of appeal only to the amir. Amanullah still
devoted one day each week to the dispensation of justice.
In due course conflicts became inevitable between the state, which
introduced a legal code on the European model, and the ulema who
regarded the shariat as the principal legal source and wished to retain
the monopoly of justice. Indeed, the effect of all the legal codifications carried out by the state was to diminish the role of the ulema.
For example, the penal code of 1924 laid down the priority of specialised legal codes in the commercial and military fields, which lay
beyond the jurisdiction of the ulema. Codification also allowed recourse to legal sources in competition with the shariat, such as the
pushtunwali, the customary Pushtun law. After 1929, however, the
secularisation of the law went into reverse, with the civil and criminal law directly inspired by the shariat at the expense of imported
law.90 The movement towards secularisation gradually resumed after
1945, but the shariat and the right of reference to an Islamic judge
continued, so that the coexistence of two alternative principles of
justice led to some confusion. Personal law remained generally under the authority of the Islamic judge in the new civil and penal
code of 1976, but punishments were defined, which deprived the
qadi of a margin of interpretation. Finally, judges increasingly tended
to be drawn from the faculty of law and political science.
In teaching the state attempted to displace the ulema and the mullahs in favour of compulsory education controlled by the administration in which the teaching of non-religious subjects had first
place. Nevertheless, at least in the 1930s and 1940s, the ulema retained a controlling power over education. In consequence the secondary schools for girls were closed till the end of the 1930s, and
religious subjects regained their predominance. After the arrival of
Daud in 1953, secularisation once more became the dominant tendency, not least because of the increasing presence of western experts in this sphere.
90

Nevertheless the constitution was ambiguous on the point. See Night Mehrose
Chishti, op. cit., p. 65. See also M. G. Weinbaum, Legal lites in Afghan society,
International Journal of Middle East Studies 12, 1980, pp. 2957.

58

The Sociogenesis of the Afghan State

The conjunction of the tribal and religious oppositions. Up to the 1950s

the principal danger for the state remained the conjunction of the
two opposition forces, tribal and religious. It proved fatal to Amanullah in 1929 having already dangerously threatened his power in
1924. Though there were major contradictions between the religious code, the shariat, and the tribal law, the pushtunwali, Islam was
nevertheless an essential element of Pushtun identity. In addition to
their performance of the prayers and the rituals, the mullahs interpreted the law during tribal jirgas, or gatherings. Occasionally the
externality of the mullahs to the tribal system enabled them to intervene as intermediaries, similarly to descendants of the Prophet (sadat)
and holy men (pirs). The influence of a particular mullah was in practice variable, and was enhanced if significant waqfs were under his
control. The role of the mullah, as an interpreter of the Quran, implied a constant tension with tribal practices, since he sought to
impose other values, religious rather than customary, as well as an
alternative justice, Quranic as against pushtunwali. In contrast to the
pirs, the mullahs did not necessarily seek to minimise conflicts, but
looked rather for an Islamic solution.91 The mullahs therefore sought
autonomy within the tribal structure.
The conjunction of opposition by the tribes and by the religious
establishment in general occurred through the intervention of charismatic mullahs, who made possible the temporary suppression of
tribal differences and the adoption of a unified stand against an authority denounced as apostate (kafir). This was demonstrated by the
numerous rebellions led by them.92 These mullahs might even be illiterate, since they derived their legitimacy from their charisma and
not, like the ulema, from learning. Even in the absence of a charismatic figure, the ulema were able to sustain a tribal revolt by declaring a jihad, as in 1924 during the Khost rebellion. In that year, for
instance, at the time of the revolt of mullah Abdullah Ahmadzai,
known as mullah-i Lang, against King Amanullah, the Mujaddidi
91

This opposition between mullahs and pirs is found in many tribal societies in the
Muslim world. On the situation in Morocco, see Ernest Gellner, Saints of the Atlas, London: Weidenfeld and Nicolson, 1969.
92 Instances in Pakistan and Afghanistan are numerous: the Akhund of Swat, mullah-i Lang, the Fakir of Ipi, Shami pir or Hadda-yi Sahib; cf. David Busby
Edwards, Charismatic leadership and political process in Afghanistan, Central
Asian Survey 5 (34), 1986.

Challenges to the state

59

supported mullah Abdul Ghani, one of their disciples, who directed

the uprising of the Ghilzai at Ghazni.
However, any accord between the ulema and the tribes would be
provisional, to the extent where one could speak of a mutual instrumentality. In fact, the model of tribal action did not undergo permanent change under the influence of the religious leaders, and alliances
formed under the leadership of charismatic personalities were always
temporary by nature. In reality, tribal revolts were generally occasioned by attempts by the state to place limits on the independence
of the tribes, for example through the suppression of the badraga (the
levy imposed on travellers by the eastern tribes), or the imposition of
compulsory military service, or the introduction of identity cards.
The religious aspect was generally marginal when set against the
preservation of tribal autonomy.
The events of 1929 showed clearly the complementarity between
the strategy of the ulema and that of the tribes. The revolt against Amanullah was launched by the Shinwari tribes for economic reasons,
with no genuine religious dimension. This is why they were able later
to respond to the call to revolt launched in 1938 against Zahir Shah
by a charismatic mullah, Muhammad Saadi al-Keilani, who called for
the reinstatement of Amanullah. The same tribes who engineered his
downfall in 1929 were to call for his return ten year later. However,
in 1929 the ulema opposed to Amanullahs project of modernisation
capitalised on the revolt against him by declaring a jihad against the
ruler, who could not stand up to the tribes because his army was not
strong enough. So far the situation followed the normal pattern,
with the ulema and the tribes in alliance, but the capture of Kabul by
Habibullah Bacha-yi Saqao, a Tajik brigand,93 created an unprecedented situation. Habibullah, crowned ruler by Akhundzada, the pir
of Tagab, was supported by the network of fundamentalist ulema and
by the Naqshbandis of the north;94 he is supposed to have been the
murid (disciple) of the pir Shams ul-Haq Mujaddidi Kohestani.95
93

Habibullah was born in 1890 in Kalakan, the son of a water-carrier (saqao), and
became a bandit after deserting from the army. See Mir Muhammad Sediq
Farhang, Afghanistan dar panj qarn-i akhir (Afghanistan in the last five centuries),
Peshawar: Derarshesh, 1988., p. 561.
94 Olivier Roy, Afghanistan. Islam et modernit politique, Paris: Seuil, 1985, pp. 86 seq.
95 Muhammad Naser Kamal, Afghanistan sarzamin-i aria (Afghanistan, an Aryan
country), Peshawar: Danesh Ketabkhane, 1999, p. 151.

60

The Sociogenesis of the Afghan State

However, the Mujaddidi maintained an ambiguous attitude towards

him, because of their links with the Muhammadzai dynasty and the
Ghilzai tribes.96 After some months power was regained by the
tribes united under the banner of Nadir Khan, when the Mujaddidi
family played a decisive role in rallying the support of the Ghilzai.
The key point was the incompatibility of the various strategies.
Since the ulema at this point were unable alone to sustain the ruler
against the tribes, they rallied to the pretender to the throne. By the
1960s and 1970s such opposition movements seemed to be a thing
of the past, as the state embarked on a modernisation of the country
which led to a different configuration of opposition forces.

96

The ziarat of Shirind was a place of pilgrimage for the Ghilzai nomads, which
was the reason for their connection with the Mujaddidi.

2. From Mobilisation to Revolution

At a time when the state was no longer threatened by opposition from
the tribes and the ulema, developmental processes within Afghanistan produced a revolutionary situation. This was because the educated classes, dissatisfied with their role in society, began resorting to
collective organisation. From the 1970s onwards the largely dysfunctional political system, unable to mobilise support, paved the
way for the communists to carry out a coup, a situation which led
eventually to the civil war and the Soviet invasion.
Internal constraints and the model of development. After the initial efforts
of Abdul Rahman Khan (18801901), the revolt of 1929 against
Amanullah obliged his successor Nadir to exercise great prudence.
Up to the start of the 1950s the state was principally concerned to
acquire the means to maintain its own autonomy, indeed its very existence, since it did not enjoy real military superiority over the tribes.
Its modernising effort was at that time relatively limited, and it suffered some notable defeats. Later, no longer apprehensive of military
defeat, the authorities were able to embark on social change, placing
the emphasis on education, infrastructure and industrialisation.
In Afghanistan the accumulation of capital and the process of modernisation had been the states business. For this reason the Jewish
and Hindu communities, very active up to the beginning of the 20th
century, were progressively excluded from international commerce,
as the state established monopolies over external trade. The difficulty
of mobilising internal resources to embark on the modernisation of
the country led the Afghan state to seek international finance, and
this produced social and political tensions.

The mobilisation of internal resources

Since the destruction of the systems of irrigation in the south and
west of Afghanistan by Genghis Khan and Tamerlane, the countrys
61

62

From Mobilisation to Revolution

territory has seldom afforded the necessary resources to maintain

central authority. This was why the Pushtun empires of the 18th
century were mainly financed from the Punjab.1 Subsequently, the
Afghan state experienced difficulties in raising sufficient revenue to
finance its development. To find the money to maintain the army
and create infrastructures, Abdul Rahman Khan brought in a number of new taxes, but the revenue raised remained small. In the
1920s Amanullahs ambitious modernisation project, as well as the
end of British aid, led him to increase his fiscal pressure substantially
and to raise taxation solely in the form of casha policy made possible by Afghanistans relative prosperity between the two world wars.
Thus the budget of 1928 shows that 30 per cent of the revenues of
the state came from taxation on agricultural produce: a threefold increase over the previous situation. One of the main reasons for the
unpopularity of Amanullah lay in his fiscal exactions and their consequences: corruption within the administration, and banditry. In
1927 there was even a physical attack on the Minister of Finance by
discontented peasants.
Amanullahs eventual failure demonstrated the impossibility of
financing ambitious modernisation projects rapidly without external assistance,2 and this was even more the case after the events of
1929, when the regime of Nadir Kahn came under additional pressure. The new constitution required the consent of the Loya Jirga before the introduction of new property taxes, sounable to modify
the tax regime to any significant extentthe government continued
through the 1930s to depend on customs duties for two-thirds of its
revenue,3 a notably high figure which persisted till the 1970s. The
urban economy was more highly taxed overall, which put a brake on
its development. On the other hand, the great landowners easily
evaded property taxes. Finally, the Second World War brought a substantial fall in the states revenue because of the reduction in the
overseas sales of the fleeces of the karakul (the wavy-haired sheep)
and the flight of capital to the city of Peshawar (now in Pakistan).
1

Hasan Kakar, op. cit., p. 176.

This statement should be qualified, as is shown by the development of Kunduz,
undertaken in the 1930s by the Bank-i Melli without external assistance.
3 Hasan Kakar, op. cit., p. 317.
2

The mobilisation of internal resources

63

International funding
After various hesitations, and in spite of the fact that by accepting a
degree of dependence the Afghan state broke with the isolationism
and wariness institutionalised by Abdul Rahman Khan, the solution
offered by external finance was unavoidable from the close of the
1940s. The principal advantage of this approach was that it avoided
on the one hand argument over the implications of modernisation
and especially its costs, and on the other a clash with the ulema and
the local notables. International circumstances were auspicious.
Rivalry between the United States and the Soviet Union was being
fought out through competition in the field of aid, and at least in the
short term Afghanistan gained the full benefit. From 1950 relations
between the great powers were regulated by an informal agreement
between the Soviets and the Americans which defined their zones
of influence, respectively to the north and south of the Hindu Kush.4
The volume of aid for the period 195077 was $2 billion, or more
than $70 million a year.5 From the 1960s more than 40% of state revenue was derived directly from foreign aid.6 Between 1950 and 1969
this came mainly under bilateral agreements, with more than 80 per
cent of the aid budget divided between the United States, which
provided 31% of the total, and the Soviet Union (54%). In the 1970s
an evolution towards multilateralism began to be evident: the
United Nations and the World Bank provided 20% of the aid, while
the US and Soviet Union shares fell to 21 and 29% respectively.
There was in overall terms a fall in American aid from the close of
the 1960s, while the Soviet presence was strengthened, especially the
number of Soviet specialists. The training of military officers and civilian administrators by the Soviet Union also enabled the KGB and
the GRU to pursue an effective policy of recruitment with effects
that were to make themselves felt in the 1970s.7
4

Ren Cagnat, Michel Jan, Le milieu des Empires ou le destin de lAsie centrale, Paris:
Laffont, 1981, p. 190.
5 Jacky Mathonnat, Une conomie impulse de lextrieure in Pierre Centlivres
et al., La colonisation impossible, Paris: Les Editions du Cerf, 1984, p. 169.
6 Barnet Rubin, Political Elites in Afghanistan: rentier state building, rentier state
wrecking, International Journal of Middle East Studies 24, 1992, p. 97.
7 Marie Broxup, The Soviets in Afghanistan: the anatomy of a takeover, Central
Asian Survey 1 (4), 1983.

64

From Mobilisation to Revolution

By accepting international aid the government progressively lost

control of its development programme, as was clear from its development plans from 1957 onwards, which consisted of efforts to coordinate after aid had been received. The role of foreign experts was
decisive in the selection of aid projects, through which the donors
sought visibility to demonstrate their technological prowess, but
which succeeded only in creating modernised enclaves. Industrial
development, which owed much to the Bank-i Melli, was concentrated predominantly in Kabul where a number of light industries
appeared, although before the revolution the number of workers involved never exceeded 40,000. Similarly, large-scale projects which
needed diverse groups of workers and their families to be assembled
at specified locations, such as the Pul-i Khumri cement works,
enjoyed only limited success.
Instead of facilitating a controlled transition, Afghanistans development strategy only reinforced the distinctions between regions
and social groups. This may be a feature common to all economic development, but the disruption was particularly evident in the Afghan
case since modernisation led to ever greater contrasts between the
ways of life and the aspirations of the urban and rural populations.
Life in the Westernised parts of Kabul, where most foreign experts
worked, was clearly separate from the rest of the country. Two consequences of this strategy were especially detrimental economically:
the agricultural crisis and the dependence on foreign aid.
Apart from certain large-scale irrigation projects, like that in Helmand, agriculture was not at the forefront of the donors priorities,
especially those of the Soviets, which was a reflection of their own
failures. There was not to be a Green Revolution in Afghanistan.8
The end of the civil wars had made possible a steady growth in the
Afghan population from the beginning of the 20th century. From
the 1960s there was pressure on agricultural sector, and the diminishing
improvements in productivity were no longer sufficient to match the
increase in the population. Thereafter peasant income, which had
been stable from 1935 to 1966, fall in real terms.9 The drought of
8

Concerning Afghan agriculture see Gilbert Etienne, LAfghanistan ou les alas de la

coopration, Paris: PUF, 1972.
9 For a critique of the adverse effects of aid see Pierre Gentelle, LAfghanistan et
laide internationale de 1950 1978. Espaces modernisables, socits rtives, Revue du Tiers Monde, October-December 1979.

The politicisation of the educated class

65

19701 and the resulting famine in 1972 revealed the demographic

tensions and the fragility of the agricultural equilibrium. In 1972 some
8,000 people are said to have died of starvation, and some livestock
also perished, which drove the peasants heavily into debt.10 Only the
urban centres were linked at this time by paved roads. The lack of secondary roads accounts for the absence of an integrated marketthe
price of wheat varied from one region to anotherand explains how
localised famines could occur. On the other hand, the following years
showed that in normal times Afghanistan was self-sufficient in food.
More generally Afghanistan experienced the onset of a rural exodus, partly towards Iran and Pakistan. In certain places such as the
Andarab valley north of Kabul, 40% of the male population went to
work in the towns in the winter.11 Kabul underwent spectacular development, from 450,000 people in 1965 to 600,000 on the eve of
the war. A third of those living in the city had not been born there.
Overall the growth of the urban population exceeded that of the rural population, increasing by 3.4% in contrast to 2.5%, although Afghanistan remained an essentially rural country on the eve of the
revolution of 1978, when 86% of Afghans still lived in the villages.
In the 1970s foreign aid declined because of the minimal United
States interest in Afghanistan in a period of dtente. The Soviets, particularly represented through their provision of technical aid, gave
less new credit at the same time as the burden of debt was becoming
more significant, while the dates for repayment of existing Soviet
loans began to fall due. The repayment of loans, which represented
27% of payments in foreign exchange at the start of the 1970s, rose
to 40% in 1975.12 This situation put further pressure on the political
system, which was obliged to find additional sources of finance.

The politicisation of the educated class

However, it was in education rather than the economic sphere that
modernisation had its greatest effects, laying the groundwork for the
10

See Hasan Kakar, The Fall of the Afghan Monarchy in 1973, International Journal of Middle East Studies, 9 May 1978. The production of wheat is said to have
fallen by 20 % and the number of sheep by 40 %. World Bank, op. cit., p. 27.
11 Louis Dupre, Settlement and Migration Patterns in Afghanistan: a tentative
statement, Modern Asian Studies, 9, 1975.
12 Richard S. Newell, The government of Mohammad Musa Shafiq: the last chapter of Afghan liberalism, Central Asian Survey, July 1982, p. 57.

66

From Mobilisation to Revolution

emergence of an educated generation from which almost all political activists were drawn. The earliest generation was educated in the
1920s within the system put in place by the principal western powers in the shape of their patronage of high schools.13 As an outcome
of the liberal episode of 194652, the first trade union and political
movements, illegal but tolerated, saw the light of day in Afghanistan.14 These included the Union of Students, the first student union,
founded in 1946, which was reformist, looked back with nostalgia to
the days of Amanullah, and was anti-imperialist, which meant in
practice anti-American. Another key organisation was the Wikh-i
Zalmayan (Awakening of Youth) formed in 1947.15 This group, first
established in Kandahar, brought together a large segment of the politicised intelligentsia, including notably Daud, the future president;
Nur Muhammad Taraki, the future founder of the communist party;
Shamsuddin Majruh, a senator; Muhammad Musa Shafiq and Muhammad Hashem Maywandwal, both future Prime Ministers; and
Abdul Majid Zabuli, founder of the Bank-i Melli. This groups inclination was nationalist rather than marxist, even though future communists participated in it. It also had an anti-monarchist component,
and the idea of setting up a republic was openly discussedroyalty
never enjoyed great legitimacy in the eyes of the Afghan governing
classes. However, in spite of its moderate demands, governmental
suppression put an end to this first experiment after a demonstration
in 1952 demanding the recognition of political parties.
From the 1960s the states educational policy led to a rapid increase
in the numbers of high school pupils and of students.16 Between
13

These were the French Istiqlal lyce established in 1922, the German Nejat
School in 1924 and the British Ghazi School in 1928.
14 Clandestine movements in fact began to develop early in the century: a Jamiyat-i
Siri-yi Melli (Association of the National Secret) emerged under Habibullah,
though its activities were quickly minimised after it was uncovered and suppressed following a plot against Habibullah in 1909. In addition, the intelligentsia consolidated around Mahmud Tarzi, who supported Amanullahs reforms,
but Amanullahs fall in 1929 and Nadirs Shahs authoritarianism obliged the
modernising intellectuals to observe a certain discretion, at least until the 1940s.
Thus the Halqa-yi Jawan-i Afghanistan (People of the Afghan Youth), a secret
society of supporters of constitutional reform, was broken up under the conservative regime in place from 1929.
15 See Dr Fazal ur-Rahim Marwat, op. cit.
16 There were 8,500 students in 1971 and around 15,000 in 1978. Cf. Etienne

The politicisation of the educated class

67

1957 and 1967 the number of children receiving school education

rose from 126,000 to 540,000.17 Collective action, especially strikes
of students and primary school teachers,18 began to occur, a symptom of increasing politicisation, which was promoted by a number
of factors. Because of the shortage of teachers of Afghan nationality,
foreign assistants, who generally viewed the politicisation of their
students favourably, played an important role within the university.
Two-thirds of the books used by students were in a foreign language (English, French, German or Russian)a significant indicator
of the growing assimilation of foreign culture by the Afghan lites.
The relatively liberal policies of the 1960s, and in particular greater
press freedom, also increased awareness of political ideas.
These mobilisations were in the last resort the result of the students difficulty in integrating themselves into society after leaving
the university. Without a sufficiently developed private sector, the appearance of numerous educated job seekers on the labour market
highlighted the difficulty of providing sufficient employment, with
the public sector remaining the principal outlet.19 Professional success after leaving university depended more on family influence than
on ability or on what qualification had been obtained, which condemned the majority of graduates to the lower ranks of the administration, where they were badly paid and underemployed. Indeed
high school or university education had created a psychological detachment which, especially for those from the provinces who had
been physically cut off from their origins, made return to their families difficult. These former students found themselves in a situation
which they perceived as unjust, given their abilities and the needs of
Gille, Laccession au pouvoir des communistes pro-sovitiques in Pierre Centlivres et al., op. cit., p. 181.
17 World Bank, op. cit., p. 26. In 1974, 30 % of pupils of primary school age were at
school, though there was a very marked under-representation of girls: Centlivres
et al., op. cit., p. 77. By 19734 around 11 % of the population could read. See
Encyclopaedia Iranica, Costa Mesa CA: Mazda Publishers, 1998, p. 239.
18 On the student movements of 19668 see Louis Dupre, op. cit., p. 619. There
were 39 strikes in 1966, 80 in 1967, and 133 in 1968: see Dr Muhammad Anwar,
The third Afghan constitution (part VII): political parties and newspapers, Central Asia 11, Area Study Center, University of Peshawar, winter 1982, p. 5.
Around 15,000 people demonstrated following the death of a student killed by
the police in the June 1969 demonstrators.
19 See Encyclopaedia Iranica, op. cit., p. 239.

68

From Mobilisation to Revolution

society. This frustration was in part due to their material circumstances: the salaries were miserable and fell by half in real value between the 1960s and the end of the 1970s.20 However they were
also psychological, as the personal and collective ambitions nurtured
during their school or student years collided with the reality of a
society which had little tolerance for their ideas. The relative sequestration of the sociable student life, in a country where the population was largely illiterate, reinforced the sense of identity among the
group. The army underwent a similar crisis with its young officers.
In spite of a degree of ethnic and social inclusiveness in the officers
academies, the key positions in the military were reserved for the
Durrani, especially those of the royal clan, who were presumed to
be faithful to the dynasty. In addition, those young officers who had
spent time in the Soviet Union were kept out of positions of responsibility, since the royal authorities feared, sometimes justifiably,
that they might have become communists.
In the 1970s the political parties were not highly structured organisations, but served mainly as discussion forums. The capital was
the centre of political activity, although some militants were to be
found in towns such as Herat and Jalalabad. At the University of
Kabul there were dozens of groups, with only fluid boundaries between them. In the case of the Maoists the various provincial branches were for practical purposes autonomous. These organisations,
which at their outset often wavered between adopting the status of
student unions or that of political parties, generally concentrated on
the publication of a periodical. However, in a largely illiterate country, the public they were able to reach was limited. In the absence of
exact studies, the total number of militants and sympathisers of all
the parties at this time can be estimated at some thousands. Thus the
two branches of the communist movement included at most several
thousand members at the beginning of the 1970s, of whom 2,500
belonged to the Khalq and 1,5002,000 to the Parcham, about the
same as the Islamists and the Maoists.21 The Pushtun nationalist
movement and that of the social democrats, who recruited less exclusively from among the students, had even fewer adherents.
20

World Bank, Afghanistan: the journey to Economic Development, report no. 1777aAF, 1978, p. vii.
21 Louis Dupre,Red Flag over the Indu-Kush,Part V,AUFS report no. 28,1980,p. 3.

The politicisation of the educated class

69

A feature which these political movements had in common was

that they never achieved any penetration in the countryside. Studentsmainly Islamistswho returned to their villages during the
summer to spread political propaganda were only isolated instances.
The parties were never able to mobilise beyond the educated classes
because of their lack of contacts in the rural communities, though
distinctions must be made between the different regions, especially
according to the prevalent type of social organisation. For example
in a highly segmentary tribal context, as with the eastern Pushtuns,
political activism was almost non-existent, while in Badakhshan,
among the Turkmens and in the south of the Hazarajat, where the
politicised sons of local notables exploited their social connections,
the situation was much more open.
In the 1970s movements multiplied and the ideological options
widened.22 A religious tendency solidified around mawlawi Faizani,
an alem from Herat, who established the Khodam ul-Forqan at the
beginning of the 1950s and later moved closer to Ibrahim Mujaddidi.23 Elsewhere ulema inspired by the Egyptian Muslim Brotherhood held their earliest meetings around 1957, led by Gholam Niazi,
dean of the Faculty of Theology and a graduate of Al-Azhar University, later to be executed under Daud.24 Its distinctness from the
student branch, which emerged ten years later more or less independently, was marked from the outset, even though the two movements
were in contact. The Islamist students, who were not ulema, mainly
adhered to the Sazman-i Jawanan-i Mosalman (Organisation of
Young Muslims) and acknowledged as their founder Abdul Rahim
22

For a treatment of the various tendencies see Basir Ahmad Daulat Abadi, Hazab
va Jarayanat-i Siyasi-yi Afghanistan (Parties and political events in Afghanistan),
[Pakistan]: Moallef, 1982.
23 Dr Mohammad Anwar Khan, The Emergence of Religious Parties in Afghanistan in Fazal-ur-Rahim Marwat, S. Waqar Ali Shah (eds), Afghanistan and the
Frontier, Peshawar: Emjay Publishers, 1991, pp. 1 seq.
24 This movement, whose organisation was loose, is supposed to have been known
by the name Tahrik-i Islam (Islamic Movement): cf. Olivier Roy, LIslam, in
Pierre Centlivres et al., op. cit. p. 108. For the origins of the Islamic movement see
also Olivier Roy, The origins of the Islamist movement in Afghanistan, Central
Asian Survey 3 (2) 1984, and Assem Akram, Histoire de la guerre dAfghanistan,
Paris: Balland, 1998, pp. 212 seq. Other groups were active in the 1960s, notably
that of Menhajuddin Gahiz, who ran the magazine Gahiz (Morning) until his
death in 1972.

70

From Mobilisation to Revolution

Niazi, a student, who died after an illness in 1970. Among the Shiites
the dominant personality was Ismail Balkhi, who founded the Qiyami Islam (Revolt of Islam) in the 1950s. In December 1950 Balkhi
was accused of having plotted a coup together with Muhammed
Ibrahim Bacha-yi Gaw Sawar, and was imprisoned until 1964.25
The pro-Soviet communists belonged to the Hezb-i Demokratik-i Khalq-i Afghanistan (Afghan Peoples Democratic party), founded
on 1 January 1965 with Nur Muhammad Taraki as secretary-general
and Babrak Karmal as deputy secretary-general. In the spring of
1967 the party split into two factions, Khalq (The People) under the
leadership of Nur Muhammad Taraki and Parcham (The Flag) under Babrak Karmal.26 A local splinter group, the Setam-i Melli (National Oppression) made its appearance in the early 1970s.27
The Sazman-i Demokrat-i Nawin-i Afghanistan (Modern and
Democratic Organisation of Afghanistan), known as Shola-yi Jawid
(the Eternal Flame), mustered the Maoists within a very loose organisation. This movement was established on 4 April 1968 and led
by Dr Hadi Mahmudi, together with his nephew Abdur Rahman
Mahmudi, as well as Akram and Sadiq Yari, all of whom were Hazaras
from Jaghori. The party recruited mostly among the Hazaras and the
Qizilbash of Kabul, as well as in the Kunar valley in cooperation
with Muhammad Hashem Khan, perhaps because of the presence of
the Chinese in this province in the context of a cooperative agricultural project.
The social democrats set up the Hezb-i Demokrat-i Mottaraki
(Progressive Democratic Party) under the leadership of Hashem
Maywandwal. On 24 August 1964, when Maywandwal was Prime
Minister, this movementnot officially a political partyproclaimed
25

Dr Fazal-ur-Rahim Marwat, op. cit., p. 246.

On the Afghan communist movement, see Anthony Arnold, Afghanistans TwoParty Communism, Parcham and Khalq, Stanford, CA: Hoover Institution Press,
1983.
27 The Setam-i Melli itself in due course split into two factions. In 1974 the founder
of Setam-i Melli, Taher Badakhshi, became leader of the SAZA (Sazman-i Zamatkeshan) while Bahruddin Bahes took the leadership of the SAFZA (Sazman-i
Fedayin-i Zamatkeshan). The origin of the split seems to have lain in the careful
strategy of Taher Badakhshi, who remained in contact with the Parchamis and
was opposed to the more radical vision of mawlawi Bahes, who favoured violent
action.
26

The politicisation of the educated class

71

its allegiance to Maywandwal in a radio transmission.28 In contrast to

most of the other movements, it adopted a reformist position and
refrained from vehement criticism of the monarchy.
The Pushtun nationalists gravitated towards the Jamiyat-i Afghan
Demokrat (Afghan Democratic Party) or the Afghan Mellat (Afghan Nation) set up by the engineer Gholam Muhammad Farhad.29
The latter party,established on 8 March 1966,took an ultra-nationalist
political line, particularly on the question of Pushtunistan and the
Pushtun language. At the same time the Hazaras also joined nationalist organisations, frequently those set up by the diaspora in Pakistan
and Iran, including notably the Hezb-i Moghul (Moghul Party),
which was active among the Hazara community in Iran,30 and the
Tanzim Nesl-i Haw Hazara (Organisation of the Hazara New Generation) in Quetta, Pakistan.31 It appears in addition that Uzbek militants, without setting up a political party, were also in contact with
each other.
28

Dr Fazal ur-Rahim Marwal, op. cit., p. 246.

Gholam Muhammad Farhad, an engineer trained in Germany, had been elected
mayor of Kabul in 1948, before becoming a member of Parliament.
30 The Berberis, the Hazara population in the region of Mashad who had arrived
in Iran from the 18th century onwards and called themselves Khawaris, are said
to have numbered about 500,000.
31 The Hazara population settled in Quetta from the mid-19th century, but Pakistani nationality was not granted to them till 1963. From the mid-1960s the
Hazaras of Quetta developed a strong nationalist tendency, in particular through
the Tanzim Nesl-i Naw-i Hazara Moghul (Organisation of the New Generation
Hazara-Moghul). The Tanzim, established in 1971 and particularly active in the
cultural field among the Hazara community in Quetta, had an ideology based on
an ethnic and racial nationalism, in which the Hazaras portrayed themselves as
the descendants of Genghis Khan. This positive image of conquest and force
contrasted with the humiliating and tragic situation of the Afghan Hazaras. Thus
a tapestry representing the Mongol horde has pride of place in the office of the
director of the Tanzim. Hazara nationalist works dealing with the history of the
Hazarajat do not rely on the religious distinction but rather on race as the criterion for belonging to the community. The Hazarajat, according to most of these
nationalist authors, extends over a remarkably wide area. All people related to the
Hazaras are included, regardless of their avowed identities, such as the Sunni Hazaras of Panjshir and of Qala-i Naw. For example, see Muhammad Gharjestani,
Tarikh-i nowin-i Hazarajat (Modern History of Hazarajat), Peshawar, 1988. In addition, all territories that have belonged to the Hazaras are included, even if today
they are occupied by other populations (this applies especially to areas towards
Kandahar and Ghazni). The Tanzim strives to maintain the Hazara identity of the
29

72

From Mobilisation to Revolution

POLITICAL PARTIES BEFORE 1978

Name

Formation

Membership

Hezb-i
Demokratik-i
Khalq-i
Afghanistan

1965 (split
into Khalq
and Parcham,
1967)

Persian speakers Communist

for the Parcham.
Pushtun speakers for the
Khalq

Ideology

Nur Muhammad
Taraki. Babrak
Karmal

Sazman-i
Jawanan-i
Mosalman

1968

Urban (Kabul)

Abdul Rahim
Niazi

Shola-yi Jawid

1968

Urban (Kabul, Maoist

Herat, Mazar-i
Sharif) Shiites
Qizilbash, Hazaras of Jaghori,
Kunar

Osman Landay,
the Yari brothers,
the Mahmudi
family

Afghan Mellat

1966

Townspeople,
Pushtuns

Pushtun
Nationalist

Gholam Muhammad Farhad

Hezb-i
Demokrat-i
Mottaraki

1966

Urban lites

Social Democrat Hashem

Maywandwal

Islamist

Leaders

It must be asked whether commitment to political parties mobilised feelings of communal solidarity. Political adherences were based
primarily on universal propositions such as Islamism, Maoism and
Communism, since their point of reference is either the umma or
the global proletariat. Only the Pushtun nationalists and the Hazaras used explicitly community-centred language, but in student circles their impact was less than that of the other groups. There was
no especially obvious dominant ethnic element among the Islamists
or the Communists, seen overall. On the other hand, the political
parties, the factions within them and most splinter groups clearly
reproduced communal divisions, whether tribal, religious or regional. Following an initial commitment to an ideological principle, militant activity probably tends to promote solidarity on the
basis of a broadly regional or communal basis. The Afghan political
parties functioned as networks of solidarity (qowm): adherence to a
immigrant community by organising courses and publishing a periodical in Persian and Urdu, Zulficar, but the majority of the Hazaras are today educated in
Urdu. The ability to write Persian is rare among the younger generation.

The politicisation of the educated class

73

political party was developed into a broad solidarity which brought

with it far-reaching implications extending far beyond the avowed
political aims of the movement. These might relate to security, for
example from the consequences of feuds, and to wide-reaching exchanges, as in marriage. Thus Shiite students are over-represented
in the Shola-yi Jawid, led by the Hazaras of Jaghori, although the
Herat group is not principally composed of Shiites. Another example was a splinter group of the Khalq, the Setam-i Melli, which
arose from opposition to the Pushtuns who dominated the movement by Tajiks, broadly from the province of Badakhshan. Among
the Islamists the ulema were most often Persian-speakers, while the
lay student members were mostly Pushtu-speakers.
However, these differences in the field of recruitment are inevitably traceable to social factors, urban and tribal affiliations, which
were more significant than ethnic affiliation as such. Thus factions
within the Hezb-i Demokratik-i Khalq-i Afghanistan at first sight
reproduce the antithesis between Pushtu-speakers and Persianspeakers. However, when Persian speakers such as Gholam Dastagir
Panjshiri, Shah Wali or the Uzbek Sharay Jozjani left Parcham for
the Khalq in 1968, the opposition between provincials and Kabulis
re-emerged. The Parchamwhich, while avowedly Persian-speaking,
was also bourgeois and Kabulicould not absorb these provincial
elements. In this case rural-urban opposition offered a more satisfactory explanation than ethnic affiliation. In addition, Pushtun Islamists in general emerged from mixed Persian- and Pushtu-speaking
areas with weak tribal institutions.32 The case of Sebghatullah Mujaddidi provides further evidence of social contradictions. Mujaddidi,
who had been imprisoned in 1959 for an assassination attempt on
Khrushchev, was close to the Muslim Brotherhood. Nevertheless, his
attempt to become leader of the student branch of the Islamic
movement encountered strong opposition, particularly from Gulbuddin Hekmatyar, because of his family connection with the lites.33
If the interpretation offered here is correct, social trajectoriesmore
than social positionsare a more determining factor than the broad
32

Olivier Roy, Le double code afghan. Marxisme et tribalisme, Revue franaise de

science politique 35 (5), October 1985.
33 David B. Edwards, op. cit., p. 184.

74

From Mobilisation to Revolution

ethnic affiliations of Pushtuns, Tajiks and so on. Within the same

ideological movement individuals gravitate together according to
their social profile. The modalities of political engagement can be
seen as related to two particular social trajectories, namely those of
students with links to the governing class and of provincial students
with no ties with the lites.
Some of the students linked to the governing class are found in
leftist groups such as Parcham, where activists linked to the royal clan
itself are not exceptional. Daud, who worked with the Parchamis after
he was ousted from power in 1963, appealed to them on his return
to politics in 1973, since he believed them to be linked to the existing order and up to a certain point loyal to it. Social status provided
the explanation why, with only one exception, the communists elected
to parliament in 1964 and 1969 were all Parchamis rather than Khalqis.34 The Parchamis, who for the most part were Kabulis, adopted a
more liberal attitude on moral issues than the Khalqis, with female
militants in particular playing an active role in the movement.
The leader of the Parcham, Babrak Karmal,35 was a good representative of his faction. He was born in 1929, the son of Muhammad
Hussein, a general who at one point had been governor of Paktya.
The family was linked to the royal clan and belonged to the wealthy
bourgeoisie of the capital. Karmals Pushtun origin has been questioned, and it was the case that he spoke only Persian. He was a typical
product of the Kabuli bourgeoisie, very distant from the peasant and
tribal world. As a student at the Nejat School, he became a German
translator after studying political science and law. In the 1960s he
attached himself to the Halq-i Jawan-i Afghanistan movement (People of the Young Afghans); afterwards he participated in the foundation of the Hezb-i Demokratik-i Khalq-i Afghanistan and became
leader of the Parcham branch after the movement split in 1967.
Provincial members of the educated class, who had only distant
prospects of cooptation into the establishment, were attracted by re34

Elected in September 1965 to the Wolesi Jirga: Babrak Karmal, Anahita Ratebzad,
Nur Ahmad Nur, Fazl ul-Haq Fezan. The defeated candidates were Muhammad
Taraki, Afizullah Amin (by a small margin), Sultan Ali Keshmand, Abdul Hakim,
Sharay Jozjani. In 1969 only Karmal (in Kabul) and Amin (in Paghman) were
elected. The latter was the only Khalqi ever to have achieved election.
35 See Anthony Arnold, op. cit., pp. 19 seq.

The politicisation of the educated class

75

volutionary movements such as the Khalq or the Sazman-i Jawanan-i

Mosalman. These two movements adopted the same compartmentalised and militaristic style of organisation in a bid to substitute
party solidarity for the natural solidarity of the family, and thus
make up for the inferiority of their members social position. Ideology was therefore an experience both radical and all embracing.
The Islamist students criticised traditional Islamic instruction, and
therefore the ulema themselves. In a similar way Khalq accused the
Parcham of collaboration with the authorities and the bourgeoisie.
The biography of the principal Islamist leader, Gulbuddin Hekmatyar,
exemplified this kind of career.
Hekmatyar was born in the autumn of 1948 near Imam Saheb in
the province of Kunduz. His father was a small peasant farmer from
the Ghilzai Kharuti tribe. After primary education in Imam Saheb
he was sent to the military school in Kabul and in due course passed
the examination for entry to the Faculty of Engineering. By his own
account, Gulbuddins adoption of Islamism goes back to 1966, when
he heard on the radio the news of the hanging of the Egyptian Islamist Sayyed Qutb. Contradicting rumours claiming that Gulbuddin
had been a communist in his youth, witnesses in Imam Saheb are
unanimous in saying that even then he was sympathetic to the Islamists.36 Hekmatyar made himself noticed on the campus at Kabul
through his energy, natural authority and ability as an orator. He
took part in the conference of the Union of Students in 1970 as an
active member, then as an elected official in 19723. At the same time
he became leader of the student branch of the Jawanan-i Mosalman
in the early 1970s. As a fugitive in Peshawar from 1974, he became
leader of the Hezb-i Islami.
Though just as provincial and equally without resources, the students of the government madrasas were in a somewhat different position from the other students. Their politicisation was linked to their
status as functionaries and to their contacts with the Egyptian Muslim Brotherhood, as is shown in the contrary case by the absence of
Islamists among the ulema who came out of the private madrasas. They
shared the same ambitions as the educated group, but their political
positions, though revolutionary, did not imply an insulation from
36

Various interview at Imam Saheb, October 1991.

76

From Mobilisation to Revolution

society as complete as that of the students. In contrast to the rest of

the educated group, they enjoyed a recognised position, and in the
countryside even a certain amount of prestige.
The biography of Rabbani37 is typical of an Islamist alem: provincial, influenced by the Egyptian Muslim Brotherhood, and imbued
with the Sufi tradition. Rabbani was born in 1940 at Fayzabad in
Badakhshan, the son of a small landowner. After primary school he
left for Kabul to study at the Abu Hanifa school (Dar al-Ulum-i
Shariat), and then till 1963 at the Faculty of Shariat at Kabul. He was
a teacher at the Faculty in Kabul from 1963 to 1966, and then went
to Al-Azhar, where he took his degree in 1968. Rabbani, himself a
member of the Naqshbandi Sufi order, then prepared his doctorate
on Jami, the mystical poet of Herat. He also undertook the translation of the book In the Shadow of the Quran by the Egyptian Muslim
Brother Sayyed Qutb. He became professor of philosophy in the
faculty of Shariat, and he edited the periodical Majalle-yi Shariat
(Shariat Review). As a member of the Islamist faction from its origins, he organised student groups under the direction of Professor
Niazi in the 1960s. After going into exile in Pakistan after Dauds
coup, he became in 1974 the leader of the Jamiyat-i Islami.

The failure of the parliamentary regime

The dysfunctionality of the constitutional monarchy had a number
of causes: parliaments ineffectiveness, its inability to respond to economic crisis,and its failure to incorporate the nascent political parties.
The ineffectiveness of parliament was due in the first instance to
the absence of legal political parties, which was a barrier to the construction of coherent political platforms.38 The elections scheduled
for 1974 did not in fact take place, but the king had never intended
to authorise the participation in them of the political parties, which
prevented the Prime Minister Mohammad Musa Shafiq from estab37
38

See Ludwig W. Adamec, op. cit., p. 201.

Within Parliament, several factions were nevertheless distinguishable: the conservatives centred on the Mujaddidi family and the leftists around Babrak
Karmal. In addition, informal groups centred around figures such as Khalilulah
Khalili, an intellectual close to the king, and his movement Wahdat-i Melli (National Unity); Abdul Majid Zabuli, a businessman, and Mir Muhamnmad Sadiq
Farhang, an intellectual and a senior official.

The failure of the parliamentary regime

77

lishing his own movement. In addition, the authorities limited the

access of the candidates to Parliament. In 1969 some Parcham candidates had pressure put on them by the government or were imprisoned. Members of the intelligentsia, disadvantaged by lack of money
and their lack of access to rural locations, were unable to play a part
the institutional process. Further, the absence of political parties
meant that candidates were obliged to defray personally the costs of
their election campaigns, which amounted to considerable sums.
Generally those elected were local notables who were able to mobilise a considerable clientele on the strength of their names. As shown
by the very low levels of participation, of the order of 10%, parliamentary elections offered the occasion neither for political mobilisation nor for the legitimation of the constitutional monarchy.39
The role of Member of Parliament was acted out as a version of
that of the khan, and the extent to which the concerns of the electors
were conveyed to the administration was governed by the conventions
of patronage. Corruption scandals, for example in the organisation of
the pilgrimage to Mecca, were attributable to the desire of the elected representatives to replenish their funds. Due to the legal constraints no professional political class was to emerge capable of giving
expression to social demands and of enforcing party discipline.
Because of the low representation of the educated classes among
its members, Parliament and the government were made up of social
groups whose interests were divergent, which became to some
extent a further obstacle to the system. While the majority of the
members of Parliament were rural notables, ulema or landowners of
low educational attainments, of whom about one-third were illiterate,40 cabinet members and high officials belonged to the governing
class, which was favourable to the rapid westernisation of the country. A particular issue was that Parliament contained a majority of
landowners, who in due course were to take steps towards lowering
the property tax in order to satisfy both their local clientele and their
own interests. This tax had scarcely been readjusted since Amanullah,
39

For a number of examples of the levels of participation see Louis Dupre, op. cit.,
p. 590. The constitution of 1964 envisaged an Upper House (Meshrano Jirga)
representing the provinces, and a Lower House (Wolesi Jirga), respectively elected for five and four years.
40 Nighat Mehroze Chishti, op. cit., p. 121.

78

From Mobilisation to Revolution

except very modestly in 1965. As a result it represented the equivalent of only 0.5 per cent of agricultural income at the start of the
1970s, and in some provinces such as Badakhshan, it was in any case
imposed only very sporadically. By 1978 the ratio of fiscal receipts
to GNP was 7%, a figure which had doubled since 1973 but was still
one of the lowest in the world.41 The state was therefore unable to
deploy significant resources. As a result the salaries of government
officials were derisory, which created a tendency towards corruption and fraud, over customs duties in particular, and in time resulted in a loss of revenue.
In day-to-day terms the representative system installed by the
constitution of 1964 was broadly ineffectual. When a quorum was
lacking, legislation could not be adopted by Parliament, which substantially impeded the process of administration. On occasion laws
approved by Parliament were not promulgated by the king, including for example legislation relating to political parties and to municipal and provincial councils. Thus Zahir Shah forfeited the support
of the business community when he failed to ratify laws to protect
the Afghan market, exposed as it was to competition from Pakistan
and Iran. In the 1970s the parliamentary regime seemed unable to
cope with the mounting economic problems, a failure which indirectly called into question the kings credibility. On the other hand
the Assembly did play some part in the expression of popular dissatisfaction. For example, the inability of the corrupt and ineffectual
administration to distribute international aid to the provinces most
affected by the famine of 1972 was condemned by members of Parliament, who strongly attacked the Prime Minister, Abdul Zahir.42
The parties had no mobilisation strategy which reached beyond
their own militants, and social crises did not lead to popular mobilisation. In Afghanistan there were no de-ruralised peasants43 avail41

Gilbert Etienne, op. cit., p. 64, and World Bank, op. cit., p. 14. To put the Afghan
case into perspective, it is well known that fiscal exactions in Iran have always
been lower than in European states: see Sad Amir Arjomand, The Turban for the
Crown, Oxford University Press, 1988.
42 In the same way, Musa Shafiq was challenged over the treaty signed in March
1973 with Iran on the sharing of the waters of the Helmand river (the treaty was
later repudiated by the Khalqis following 1978).
43 Farhad Khosrokhavar, Hasan K., paysan dpaysann, parle de la rvolution
iranienne, Peuples mediterrannens, 14, 1980.

The failure of the parliamentary regime

79

able for an Iranian-style mobilisation, since the exodus from the

countryside had not yet led to the appearance of shanty-towns or
deprived suburbs. The demonstration mounted by the Hezb-i
Demokratik-i Khalq-i Afghanistan before the coup of April 1978
brought out only 1015,000 people in Kabul, but was later counted
as a success. No popular movement favourable to Zahir Shah was in
evidence during the coup of 1973, any more than there would be at
the time when Daud fell in 1978. In the latter case the combatants,
in spite of the hundreds of victims, did not arouse any following
among the population of Kabul; only the presidents immediate
supporters fought for him, courageously it must be admitted, like
Amins guards who put up a stiff resistance to the Soviet commandos in December 1979.
Hindsight must be avoided in interpreting the events of 1979 in a
way that would lead to the supposition that the communist takeover was inevitable. The salient characteristic of this period was the
opportunism of various protagonists in a situation which none succeeded in resolving to their own advantage. Daud succeeded in
taking power on 17 July 1973, preventing the former Prime Minister Maywandwal from launching a long-planned coup, while the
current Prime Minister Musa Shafiq favoured a take-over by the
kings son-in-law, Abdul Wali. Dauds tenure of the presidency from
1973 to 1978 failed to stabilise the situation, as was shown by the
failed attempts by the Islamists in the summer of 1975, and then in
September 1975 and December 1976 by the military, before that of
the communists in 1978.44
On the other hand, the outcome of the various coups was not entirely a matter of chance, but reflected the ability of various actors to
organise small groups of militants and to infiltrate institutions. Given
the lack of widespread mobilisation of the population, everything
depended in the event on the ability of the actors to control the
countrys institutions, and in particular the army from within. From
the 1950s onwards, the military establishment became the primary
objective for any group which wished to take control of the state.
Up to this point the normal route to power had been to seek external support, tribal or foreign, and to intrigue within the royal clan.
Henceforth, however, conflicts would be settled within the state
44

See Louis Dupre, Afghanistan, op. cit., p. 762.

80

From Mobilisation to Revolution

apparatus by way of coups. Dauds chances of success, or those of

the communists, were directly related to the level of their support
within the army. The coup of 1973 was carried out on the basis of
the support Daud had retained among the officers from his first period in power, in 195363, when he had played a key role in the
armys modernisation. Dauds lack of vigilance subsequently allowed
the Khalqis to infiltrate the police and the army, where they became
the most influential group, thus laying the basis for the coup of
1978. In addition, most of the officers had been trained in the Soviet
Union, which paved the way for the involvement of the Soviet
forces. However, the coup of 1978 demonstrated a lack of cohesion
within the officer corps, which did not rally in numbers around the
communists. On the other hand the Islamists, lacking leverage
within the machinery of the state, had no chance.

Authoritarian mobilisation and repression under Daud

The coup of 17 July 1973 was carried out without difficulty because
Daud had the backing of the officers. Zahir Shah, who had few supporters to call on, chose to prolong his sojourn in Rome into a permanent exile. In spite of the proclamation of the republic, Dauds
rise to power did not take the form of a fundamental split within the
governing class, but seemed rather to be a change of allegiance within
the interior of the royal clan. In the event, Daud set up an authoritarian regime which made the government isolated and fragile, in
spite of an undeniable improvement in the economic situation. Benefiting from an international situation favourable to the Afghan economy, the government dramatically increased taxes at the rate of 18%
per year in 19737.45
In the political arena Daud moved rapidly to undermine all the
representative institutions, and in particular Parliament. The notables
forfeited part of their influence with the administration, since the
state saw itself henceforth as dealing with individuals and not communities. This is why it was ordered that personal names with a tribal
or communal connotation should be altered, in order to counter
communal solidarity, at least symbolically. Daud also planned to enhance his support with the creation in 1976 of a unique political
45

World Bank, op. cit., p. 30.

Authoritarian mobilisation and repression under Daud

81

party, the Hezb-i Enqelab-i Melli (Party of the National Revolution). In the constitution of 1977 a key role was assigned to this institution, when all the deputies of the Melli Jirga (National Assembly)
automatically became members of it and it was given the role of
putting forward the candidature of the president of the republic. After
the constitutions ratification of 1977, Afghanistan officially became
a one-party state, with the president responsible only to an extraordinary session of the Loya Jirga.
However, the party launched by Daud did not put down strong
enough roots to serve as a counterweight to the opposition, and
especially to the communist opposition. The failure illustrated Dauds
inability to create political structures sufficiently attractive to mobilise the officials, the students and the modernist notables within the
framework of the Hezb-i Enqelab-i Melli. The communists were
similarly to fail to establish a numerically significant party after their
seizure of power.
To avoid any challenge Daud systematically suppressed the opposition, both legal and illegal. Following the coup the former Prime
Minister and leader of the social democratic Hezb-i Demokrat-i
Mottaraki, Hashem Maywandwal, who had been in power in 19657,
was arrested in September 1973 and executedperhaps on the orders of Daud, or possibly on the initiative of Parchamis who had infiltrated into the police intending to radicalise and isolate Dauds
government. Daud then attempted to get rid of the Islamists with
the aid of the communists, who were well represented in the police.46
In 1974 some Islamist militants were arrested, while others fled and
took to the hills near Keshem in the province of Badakhshan, where
they were led by Dr Omar.47 The most visible militants, including
Masud and Hekmatyar, went to Pakistan, where they were welcomed by the regime of Zulfiqar Ali Bhutto, who was currently on
bad terms with Daud.
At the beginning of 1975 a movement bringing together Islamist
students and ulema was established in Pakistan.48 Its structures com46

Tahir Amin estimates at 600 the number of Islamists who fell victim to purges
under Daud. See Afghan Resistance: Past, Present and Future, Asian Survey
XXIV (4), April 1984.
47 Afghanews 5 (2), 1989.
48 According to the militants of one group or the other the new structure was
called Jamiyat-i Islami or Hezb-i Islami. Which came first is a factor in the estab-

82

From Mobilisation to Revolution

prised a Shura-yi Marqazi (central council) open to all members,

which elected a Shura-yi Ejray (executive council) of ten members,
including Hekmatyar and Masud. Since the ulemaNiazi, Rabbani,
Tawana et al.were mostly still in Afghanistan, the students took
charge. During the summer of 1974 Rabbani, the designated successor of Gholam Niazi, who was already in prison, arrived in Pakistan and set up in opposition to Hekmatyar, the former leader of the
Islamist students in Kabul and responsible at that time for military
operations and relations with the Inter Service Intelligence (ISI: the
Pakistani secret services): a key post in the circumstances of the
time. Elections took place and an alem, Qazi Amin, was elected as a
compromise candidate with the backing of both Rabbani and
Hekmatyar, but gravitated rapidly towards the latter. Rabbani then
embarked on a pilgrimage to Mecca, which was a way of keeping
his distance.
The Islamist movement began at this point to plot a coup, with
the assistance of the Jamaat-i Islami and the Pakistani secret services,
who trained the militants and provided money for the purchase of
300 rifles. The coup attempt took place in July 1975. The Islamists
pursued a strategy of dispersing small groups of fighters throughout
the provinces. They infiltrated themselves particularly in Badakhshan
under Dr Omar; in Panjshir under Masud; in Herat where they were
led by Nasratyar; in Laghman with mawlawi Habib ur-Rahman;
in Paktya under the directions of Hekmatyar, who was to remain in
Pakistan; and in Nangrahar, where the leader was Adam Khan. In all,
there were no more than 2300 combatants who, without backing
in the army, had no chance of overturning the regime.
In Panjshir, for example, Masud divided his group of about forty
individuals into three sections, to take control of the whole valley.49
Masud himself, accompanied by the engineer Ishaq, took charge of
lishment of legitimacy. It may be that the appellation Jamiyat was used in a
more or less official manner, during the period of the formation of the Islamist
movement. On this subject, see also the serie of articles by Tawana, long established as Rabbanis deputy, Afghanews 5 (7), 1989.
49 An account of the events, together with the list of 37 participants, is given by
Abdul Hafiz Mansur, Panjshir Dar Doran-i Jihad (Panjshir during the Period of
Jihad) Peshawar, 1991 (1369 AH), pp. 48 seq. Reliance is also placed here on
various accounts gathered in the north in the autumn of 1991.

Authoritarian mobilisation and repression under Daud

83

the central sector of the valley (the alaqaderi of Rukha) during the
night of 21/22 July and wounded the alaqadar (deputy governor). In
the morning the inhabitants woke to find banners on the walls of
the government building proclaiming Death to Daud and Death
to the Russians. Supposing the culprits were bandits, they alerted
the army which sent in helicopters and tanks. Isolated and hunted
by both the army and the population, the Islamists fled in disorder,
and most succeeding in returning to Pakistan.
Other groups were even less successful. In all, ninety-three individuals were arrested and tried, of whom three were condemned to
death. In addition to the revolutionaries lack of organisation, the
principal lessonas significant as the results of elections would have
beenwas the total lack of support from the population, who neither knew nor understood the Islamists. This attempted coup also
gives some indication of the movements rootsit was in fact weak
and geographically restricted. Nothing took place in the south, in the
region of Kandahar, or in the west. Only the northeast and the frontier with Pakistan were affected. The coup had no chance of success,
as the Pakistani military men who helped to lay the groundwork for
it must have been aware. It might therefore be supposed that the
Islamists had been manipulated by the Inter Service Intelligence,
whose intention was to deliver a warning to Daud; according to some
sources, the ISI may even have tipped Daud off in advance. In spite
of the acknowledged lack of preparation of the young militants, who
had placed a little too much emphasis on Guevarist foco theory,
the Islamist movement had nevertheless embarked on an adventure
which cost it many lives, particularly as the result of the backlash
subsequently unleashed in Afghanistan. The fiasco also marked the
definitive split in the Islamist movement between the Jamiyat-i
Islami, led by Rabbani, and the Hezb-i Islami led by Hekmatyar.
After the operations against the Islamists, Daud decided to exclude
the communists from the new cabinet formed in January 1977, but
many senior officials of openly Parchami sympathies were nevertheless assigned positions of responsibility, especially in the Ministry
of the Interior. Dauds immediate entourage was infiltrated by the
communists, including for example Yaqubi, the future head of the
secret services under Najibullah. Meanwhile the Khalqis continued
their clandestine infiltration of the army, an operation facilitated by

84

From Mobilisation to Revolution

Dauds purges within the military establishment and the rise of a

new generation of officers.
The position of the communists was also made difficult by the
new foreign policy of Daud, who sought to re-adjust his hitherto
pro-Soviet position. His rise to power in 1973 had actually been
welcomed by the Soviets because of his links with the Parchamis and
the good relations he had maintained with Moscow at the time of
his previous period in power in 195363. In particular, Daud set in
train military cooperation with the Soviets following his visit to
Khrushchev and Bulganin in September 1956. From that time on,
about 100 Afghans went each year for training in Soviet and Czech
military schools, with Tajik translators. In fact the earliest requests for
military assistance had been made to the United States, since mistrust of the Soviet Union remained strong in the Afghan governing
classes of the 1930s and 40s, but in 1955 the United Statesbecause
of poor Afghan-Pakistani relations, refused to finance the modernisation of Afghanistans military machine. Daud because of its claim
to Pakistan-controlled Pushtunistan, turned to the Soviet Union.
Cooperation with the Soviets later took on an economic aspect
when, following the closure of the Pakistani port of Karachi to
Afghan products, Afghanistan arranged to send its exports through
Soviet territory.
Dauds return in 1973 gave a new impetus to Afghan-Soviet cooperation. In the 1970s more than 100 civilian experts arrived each
year, and from 1973 onwards every battalion of the Afghan army had
its Soviet adviser. On the diplomatic front the treaty of neutrality
and non-aggression of 24 June 1931, which replaced the treaty of
friendship of 28 February 1921, was re-adopted on 10 December
1975 to run till 1980 with an automatic renewal every five years.
Lastly, on 30 July 1977 a thirty-year commercial pact was signed.
However, Daud also attempted from this stage onwards to make
regional agreements, with the aim of accelerating Afghanistans development and bringing it out of its seclusion. In 1976 the country
took a step towards the recognition of the Durand Line as the international frontier. This was a move towards abandoning its irredentist
ambitions, tending towards the resolution of the old quarrel with
Pakistan. An agreement would have represented a decisive change in
Dauds strategy, from which he hoped to obtain economic benefits

Utopia and violence: the communists in power

85

and a more propitious regional environment. Daud and Zulfiqar Ali

Bhutto exchanged visits in 1976. Changes in the situation in Pakistanwhere the elections of 1977 led to Zia ul-Haqs successful
coup as well as changes within Afghanistanwould in the end prevent the finalisation of an agreement. During a visit to Iran the same
year, Daud obtained a promise of aid for the construction of a railway in western Afghanistan. The following year, 1977, Daud visited
Egypt, Pakistan and Saudi Arabia, confirming the re-direction of his
foreign policy towards the conservative states of the region, and his
decision to keep his distance from the socialist camp. It was in this
spirit that he attacked Cubas adherence to the Soviet line during
the preparations for the non-aligned summit of May 1978, and the
Cuban delegation which arrived in Kabul a few days before the
coup had a conspicuously cold reception. These internal and external developments in Dauds foreign policy were destined to precipitate his break with his communist allies.

Utopia and violence: the communists in power

The retrenchment of the Americans on the international scene,
together with Dauds policies after 1975, prompted the Soviets to
consider instigating a coup and consequently the reunification of
the Khalq and the Parcham. This took place on 3 July 1977 after several months of negotiations under Soviet auspices. The coup, initially
set for August 1978, was brought forward due to circumstances. In
the event, 17 April 1978 saw the assassination of Mir Akbar Khayber,
a leading Parcham ideologue and a confidant of the Soviets. Though
there is no definite evidence, it seems likely that Amin, the Khalqs
number two, had wished both to rid himself of a rival and to precipitate a political crisis. Mir Akbar Khayber, who controlled the military faction of Parchams membership, was potentially a rival to Amin
by virtue of his age and position. Whatever the truth of the matter,
the funeral of Khayber on 19 April turned into an anti-government
demonstration, since Daud was generally thought to have been responsible for the assassination, and 1015,000 demonstrators shouted
slogans against the government, the United States and SAVAK, the
Iranian secret police. In response, the police on 26 April arrested the
leaders of the Hezb-i Demokratik-i Khalq-i Afghanistan, including
Taraki and Karmal. On the other hand Amin, who had supporters

86

From Mobilisation to Revolution

within the police, was only placed under house arrest and was therefore able to issue orders to launch the coup according to the plan
already agreed. Thus began the Revolution of Saur (April/May).
On 27 April at 9 a.m. the garrison of Pul-i Charkithe Fourth
Armoured Brigade, commanded by Captain Aslam Watanjaradvanced on Kabul with fifty tanks, while Colonel Abdul Qader led a
mutiny at the air base at Bagram, north of Kabul. Captain Watanjar
attacked the Ministry of Defence while another group took control
of Kabul airport, and then at mid-day placed himself with nine tanks
outside the presidential palace, where Daud had assembled his cabinet to decide the fate of the communists arrested the previous evening. Daud resisted, at the head of the 1,300 men of the Republican
Guard, and fighting went on through the night. Seven or eight MiG
21 combat aircraft from Bagram bombarded the palace and the general staff headquarters, while a counter-attack by two fighters, which
had come from Shindand on Dauds orders, failed. Daud was finally
killed, still fighting, at around 5 a.m. on Friday 28 April, together
with eighteen members of his family. A thousand people, mostly soldiers, are also said to have lost their lives in the fighting. The tanks of
the Fourth Armoured Brigade then left for Jalalabad, where some
officers had refused to join the coup, and the commander of the garrison there was also killed. At 7 p.m that evening the coup was proclaimed on Radio Kabul, and the newly-formed Shura-yi Enqelabi
(Revolutionary Council) undertook to embark on socialist reforms
while respecting Islam. On 30 April the Revolutionary Council announced its firman (decree) no. 1:50 Afghanistan would henceforth
be the Democratic Republic of Afghanistan under the authority of
Taraki,51 who became president of the Revolutionary Council and
Prime Minister.
50

Significantly, the text was read first in Pushtu by Taraki and then in Persian by
Karmal.
51 Taraki, who was born in 1917 at Moqur in the Province of Ghazni, was a Ghilzai Pushtun, from a modest though not poor family. He studied at the high school
at Ghazni and was then employed in India by the Pushtun Trading Company.
He met Khan Abdul Ghaffar Khan, the leader of the Pakistani Pushtun nationalist
movement (the Red Shirts), who was a fervent admirer of Lenin. Working as
an employee of the Ministry of Economy, he won a reputation as a novelist, poet
and translator. After going to Washington in 1953 as press attach, he asked for
political asylum in the United States, then spent three years in Pakistan. In 1956

Utopia and violence: the communists in power

87

From then onward, the Soviet Union worked towards the swift absorption of Afghanistan into the socialist bloc, and at the end of May
1978 it became a member of the socialist community. On 5 December 1978 the two countries signed a treaty of friendship, cooperation
and good neighbourliness, which specifically provided for enhanced
military cooperation. In addition, the Afghan police, hitherto trained
by West Germany, began to receive experts from East Germany.
The role of the Soviet Union became all the more crucial since
relations between Afghanistan and the United States were undergoing a period of serious crisis. On 14 February 1979 the American
ambassador Adolph Dubs was kidnapped by militants belonging to
the Setam-i Melli, a splinter group of the communist movement,
which wished to exchange him for their imprisoned leaders, and
especially for Taher Badakhshi, Badruddin Bahes (who was possibly
already dead), and Wasef Bakhtari. The police, with their Soviet officers, immediately attacked, against the advice of the Americans, and
Ambassador Dubs was killed in the exchange of fire. This provoked
an immediate reaction from the United States: the planned level of
civilian aid, which had been set at $15 million, was cut by half, and
military aid of $250,000 was cancelled. Meanwhile the regime had
already initiated a package of reforms aimed at radical social transformation, at a time when the tensions between Khalq and the Parcham
were worsening: the situation which would in the end lead to the
Soviet invasion.
Once in power, the communists emphasised the centralist and authoritarian aspects of Dauds regime. On 9 June the Revolutionary
Council announced its initial thirty-point programme of economic
and social reform.52 Firman no. 6 of 12 July 1978 forbade usurious
loans and cancelled mortgages on land-holdings of less than 2 hectares if they dated from before 1974, cancelling a proportion of mortgages taken out in subsequent years. Because of the agricultural
he was employed as a translator by the American embassy in Kabul, and was
probably working for the KGB at this period. He was a founder-member and
the first secretary-general of the Hezb-i Demokratik-i Khalq-i Afghanistan in
1965, and then became leader of the Khalq after its break with Parcham two
years later. See Anthony Arnold, op. cit., pp. 17 seq.
52 For an English translation of the statutes see Louis Dupre, Red Flag over the InduKush, part 3, American Universities Field Staff Report, South Asia Series, 1980.

88

From Mobilisation to Revolution

crisis, peasant farmers had been obliged to take out loans bearing
rates of interest typically of the order of 50 per cent. The burden of
debt was in consequence bitterly resented in the countryside, but
this firman, which targeted the relationship of patronage between
proprietor and farmer, would in the event be only very partially applied. Firman no. 7 of October codified the practice of mahr (dowry).
The ceiling was fixed at 300 afghanisa nominal sumand women
under the age of sixteen and men under eighteen were forbidden to
marry. This decision was the only one to be justified on the basis of
the shariat. However these measures, well intentioned as they were,
succeeded only in undermining the legal status of women in marriage since traditionally the wife recovered her dowry in the case of
a divorce, which gave her real protection against repudiation. Finally,
firman no. 8 of 2 December 1978 set out the framework of an agrarian reform (islakat-i arzi) which was intended to bring into being a
class of peasant farmers supportive of the regime. A single family was
not supposed to own more than 6 hectares of land of the highest category (the statute defined seven categories of land). Redistribution
would be carried out primarily in favour of day labourers working
the land, then of landless peasants in villages. The reform also provided for the setting-up of agricultural cooperatives for those who
owned less than 5 hectares. In practice it was not possible to implement the reform since it failed to make available to new owners the
water and seeds needed to exploit the land. In addition, in those regions (the majority) where the link between peasant farmers and
proprietors remained strong, it generally met with resistance. These
reforms, as well as the policing practices of the regime, gave rise to
popular uprisings, which exacerbated internal conflicts within the
regime itself. These are examined in detail in the next chapter.
Tensions between the two factions, the Khalq and the Parcham,
were not eased after the acquisition of power. In theory an equilibrium existed between the two factions. Taraki, as secretary-general
of the party, president of the Revolutionary Council and Prime
Minister, had Karmal as his deputy in each of these positions. However the apparent cooperation between the two factions disguised an
advantage for the Khalqis, who were better represented among the
military. Also there was a nucleus of militants, such as Qader, a Parchami and Watanjar (a Khalqi who later went over to the Parcham)

Utopia and violence: the communists in power

89

whose fidelity was primarily to the Soviet Union. These militants

gave their obedience directly to Moscow, which would in due course
extend its protection to them in the context of the internal party
struggles.
From the end of May the Parchamis disappeared from the reports
of official activities in the press, which signalled their fall from grace.
In early July their principal leaders were appointed as ambassadors,
which kept them away from real power.53 The Parchami ministers
were not replaced by others, except for the Minister of the Interior,
who was replaced by Watanjar. The great ease with which the Parcham was eliminated showed that the institutional balancing mechanisms within the party had not been enough to counterbalance the
influence of the Khalqis inside the army. On 27 November 1978 the
majority of the Parchamis were excluded from positions of power, so
that the Political Bureau fell under the control of the Khalqis.
The Parcham leaders, now scattered among the embassies, in due
course planned a coup with Soviet assistance which was to have taken
place in September 1978. However, on 17 August General Abdul
Qader the Minister of Defence, Colonel Rafi the Minister for Public Works, Sultan Ali Keshmand the Minister for Planning and
Shahpur Ahmadzai the chief of staff of the army were arrested and
charged with plotting a coup. Their confessions were published by
the Kabul Times of 23 September 1978. Ahmadzai and Akbar were
executed, but Keshman, Qader and Rafi, who were also condemned
to death, had their sentences commuted, probably at the behest of
the Soviets. A further coup was attempted in mid-September by a
number of soldiers under Watanjars command, but competition was
now mostly between members of the winning side. Indeed the
elimination of the Parchamis at once opened the door for overt expression of the internal tensions within the Khalqi faction. In particular the conflict between Taraki and Amin was aggravated.54 Amin,
53

On 27 June Baryalay, the half-brother of Karmal, and Najibullah were appointed

ambassadors to Pakistan and Iran, followed on 5 July by Babrak Karmal himself,
who was sent to Czechoslovakia. Nur Muhammad Nur, a former Minister of
the Interior, went to the United States, Abdul Wakil to Britain, and Anahita
Ratebzad to Belgrade.
54 Amin was born in 1929 at Paghman. He came from a Ghilzai family of minor
officials, and became the principal of the Avesina High School after training as a

90

From Mobilisation to Revolution

the Minister of Defence, became Prime Minister on 27 March 1979

and increasingly strove to replace Taraki as head of the party.
On Monday 10 September 1979 Taraki, returning from the nonaligned summit in Havana, stopped over in Moscow and in complicity with the Soviet leaders decided to have Amin killed at Kabul airport, with the connivance of Watanjar, Gulabzoy and Asadullah
Sarwari, chief of the AGSA (the Afghan Secret Services). However,
Aminapprised of the plot by Nawab, Sarwaris deputy at the
AGSAchanged the guard at the airport and then arranged for the
dismissal of Gulabzoy, Watanjar and Sarwari, who all took refuge in
the Soviet embassy. Over the following days Taraki invited Amin
several times to the House of the People, the former presidential palace, in order to have him killed, but Amin, aware of a plot, refused to
come. On 14 September he finally accepted the invitation and came
unarmed to the office where he was awaited by Taraki and the Soviet ambassador Puzanov.55 Sayyed Daud Tarun, one of Amins supporters, went in first and was killed by Tarakis men. Amin then fled
to the Ministry of Defence, gathered some troops and had Taraki
arrested, without a fight. Taraki was strangled a week later in his
cell. His death, attributed to illness, was announced in a brief newspaper report. Amin had already had himself proclaimed president, on
16 September, and immediately received a message of congratulation from Brezhnev.
Why did the Soviets fail to keep their Afghan allies under control
at a time when many of them were agents for the KGB or the GRU
(Soviet military intelligence)? This probably had to do with the
revolutionary process, which favoured extremism, as well as to the
resilience of the networks of solidarity within the party, which forestalled purges. At first more favourable to the Parchamis, whose
coup, planned for September, they had supported, the Soviets then
helped Taraki against Amin, whose brutality had alarmed them. The
consistent line taken by Moscow had been to promote a gradual
transition tending towards the long-term integration of Afghanistan
schoolteacher and obtaining an MA at Columbia. He was the only Khalqi ever
to be elected, in 1969, after being trounced in 1965. His active and organised
contribution was crucial to the success of the Hezb-i demokratik-i Khlaq-i Afghanistan in the organisation of the April 1978 coup.
55 Fikriyat Tabeev became ambassador in Puzanovs place in November 1979.

Utopia and violence: the communists in power

91

into the Soviet bloc. In the event, from the takeover of power by
the Hezb-i Demokratik-i Khalq-i Afghanistan, the Soviets sought to
calm the revolutionary passions of their protgs, as they were conscious of the risks of social upheaval which reforms could bring. At
their request the Afghan government held over certain measures
such as the agrarian reform of 1979. The Soviet advisers also suggested, though unsuccessfully, a more cautious line in education.
To win popular favour there was even a major celebration on the
Prophet Muhammads birthday.
The worsening of the situation subsequently induced the Soviets
to increase their military aid and from time to time to intervene directly, as during the Herat rebellion (see the next chapter). The
Afghan communists, especially Taraki, repeatedly sought an increase
in military aid. However, in the spring of 1979 the prospect of a massive direct intervention continued to be rejected by all the Soviet officials responsible for Afghan affairs, highly aware as they were of the
diplomatic costs and the risk of becoming bogged down.56 The deterioration of the military situation during the summer of 1979,
with the capture of the town of Gardez by the rebels, impelled the
Soviets to boost their aid, and in July in particular to station 400 men
at Bagram airport north of Kabul. The signing of the Afghan-Soviet
treaty of friendship and cooperation on 5 December 1979 enabled
the Soviet Union to send in 10,000 men, who would lay the groundwork for the invasion. The assassination of Taraki, who had personal
links with Brezhnev, as well as Amins seizure of power finally put
the issue of direct intervention on the agenda, but the aim was no
longer so much to help a friendly regime as to restore order in the
internal affairs of the party through the elimination of an element
which was out of control.
What is now known about the decision-making process serves as
a good illustration of the workings of the Soviet system as it entered
its decline.57 The decision to intervene was taken following a meet56

In the face of the worsening situation a working committee on Afghanistan was

formed, which included Ustinov (Minister of Defence) Andropov (KGB) and
Ponomarev (International Department of the Party). On Soviet attitudes to the
Afghan communists, see Odd Arne Westad, The Soviets and the Afghan Communists, International History Review XVI (1), February 1994.
57 See Jean-Christophe Romer, Les mcanismes de prise de dcision en URSS. Le
cas afghan 19781979, Relations Internationales 85, spring 1996; also the docu-

92

From Mobilisation to Revolution

ing of a number of the members of the Politburo on 12 Decemberthe day when NATO made public its position on the socalled Euromissiles. Since the KGB, the diplomatic service and the
army were opposed to intervention, it appears that the responsibility
for the decision rested on the alcoholic and depressive Brezhnev and
on ideologues such as Suslov and Ponomarev, who are said to have
desired to safeguard the dogma of the so-called irreversibility of the
construction of socialism. Rather than from an overall strategy, the
decision sprang from a synergy between bureaucratic authority and
the mechanical rehearsal of a repertoire of responses which had brought
results in the past, for example in Hungary and Czechoslovakia.
On 22 December 1979 Amin left the House of the People to set
up his headquarters in the south of Kabul, at Darulaman, where he
thought he would be safer. On 23 and 24 December the Soviets,
who were in control of Kabul airport, sent in reinforcements, and in
the evening of 26 December 300 Soviet commandos took up positions 3 kilometres from Darulaman. On 27 December at 7 p.m the
palace was surrounded and the commandos attacked Amins residence. At the same moment the telecommunications centre of Kabul
was destroyed by an explosion, and the Radio and Television building was stormed. At 8.30 p.m. a message from Karmal, still in Central
Asia, was broadcast on the radio. Amin and his men resisted courageously and the fighting continued until 1 a.m. Amin, who was alive,
was taken to the Soviet headquarters and executed. Engagements
with his supporters continued till the evening of 28 December, but
from then onwards Soviet control was unchallenged.

ments of the Politburo (in German and Russian) in Pierre Allan et al. (eds),
Sowjetische Geheimdokumente zum Afghanistankreig (19781991), Hochschulverlag
der ATH, Zrich, 1995.

Part II. MOBILISATIONS

3. The Commanders
Edicts are the Sultans; the mountains are ours.Dadal Oghlu

In response to the violence of the state, and afterwards to invasion by

a foreign power, the population rose up in revolt and embarked on a
Holy War, a jihad. The mobilisation of the countryside took place in
the context of local solidarity networks, organised around commanders. The social backgrounds of these leaders provide an explanation of the different kinds of organisation which were set up.

The revolt
Why was there a breach between the communists and the population? The reception given to the new regime in its first weeks was
by no means consistently hostile. Indifference was the principal reaction, but certain urban groups, as well as landless peasants in such
areas as Laghman, supported the reforms. Measures taken by the regime, particularly in agrarian reform and the reduction in womens
dowries, did not amount to radical change. Land redistribution had
begun to be tried out under Daud.1 The question of limitation of
dowries was subject to widespread debate,with some Islamists favouring it in the interest of facilitating marriage for the poorest people.
Because of the high value placed on the written word in Afghan society, such projects as the promotion of literacy never attracted fundamental objections from the rural population if they were carried
out in a spirit of respect for local customs, and in particular the separation of men and women. In practice, however, teaching adults to
read without prior dialogue, together with forced labour, visits by
1

Dauds reform in 1976 limited land holdings to 30 jerib for each individual in order to permit the donation of 600,000 hectares to 676,000 families. In practice
there was no genuine redistribution, particularly on account of the manipulation
of the land register by large landowners. Hermann-J. Wald, Asis Nadjibi, Land
Reform in Afghanistan, Internationales Asienforum 8 (12), 1977, pp. 11023.

93

94

The Commanders

the census authorities and agrarian reform, created considerable tension. The failure of the communists was due mainly to their administrative inadequacy, as well as to a proclivity towards regimentation
which did not compensate for the absence of legitimacy. Above all
it was due to their resort to violence.
In the first instance the state did not have sufficient administrators
to put its reforms into practice.2 It did not have any representation
below the level of the alaqadari (the sub-district), and the officials
were mostly remote from the villagers. Because of their location on
the periphery of villages and their contemporary architecture, government buildings symbolised the externality of the state. The typical official, the mamur, dressed in the European style and wore the
karakol cap rather than a turban, or even went bare-headed (sarluchi).
The officials were conduits for the importation of urban manners
and professed modern values alien to the villages. To cap it all, nonPushtuns found themselves confronted largely by Pushtun officials
who were frequently suspected of privileging members of their own
community. Language was a further barrier between the administration and the rural population.
The level of integration was certainly not the same in the case of
the great nomadic tribes, who kept their distance from the government, as it was for the peasants of the oases, who were subject to the
activity of the administration on an everyday basis. Outside the
towns, however, the legitimacy of the state was unrelated to its ability to administer the population, which was undemanding other
than in the provision of schoolteachers, who were often requested
by the villages. In the case of rural communities, the state was held at
arms length by a profusion of micro-strategies, in particular by
corruption, which became a means of limiting the effective influence of the administration on daily life. In some instances officials
were quite simply paid not to intervene. However, this externality of
the state did not present a barrier to its becoming a source of profit,
particularly for the Pushtuns. The antagonism between the world of
the administration and that of the tribes, in theory virulent, did not
prevent nepotism and the exercise of patronage.
2

Thomas Barfield, Weak Links on a Rusty Chain: Structural Weaknesses in Afghanistans Provincial Government in R. Canfield, M. N. Shahrani (eds), Revolutions
and Rebellions in Afghanistan, Berkeley: University of California Press, 1984.

The Revolt

95

Links between the two worlds were necessary, but the abolition of
the arbab, who served as intermediaries between the administration
and the population, had recently removed this linkage. Attempted
under Daud, this measure was rapidly abandoned, as it would also be
after the Soviet invasion. Secondly, the efforts of the authorities to
mobilise the population were in due course to prove counter-productive. Political life in Kabul and even the coups aroused only limited
interest beyond the educated class. Some nomadic communities in
the northwest still remained unaware of the overthrow of the king
and the installation of the republic some years after they had happened. The regime therefore made efforts to overcome popular
indifference. Radio stations throughout the country spread the new
regimes propaganda, but the unfamiliar Marxist-Leninist language
fell harshly on the peoples ears. To the south of Kabul, in Logar, the
villagers were obliged to repaint the doors of their houses red as a
sign of support for the revolution and, bizarrely, to demonstrate against
the Chinese invasion in Vietnam. Officials also compelled the villagers to applaud, which was considered a Russian custom and therefore kafir.3 At the Istiqlal lyce in Kabul, where the doors were also
painted red, the Khalqis held compulsory half-hour political meetings every morning before school. Amin later abolished them.
This strategy of authoritarian mobilisation did not make up for
the illegitimacy of the communists, and the aggressive atheism of
their politics rapidly lost them all credibility with both the rural and
urban populations. This rejection showed that religious legitimacy
continued to be a determining condition for acceptance of the state.
To make up for the abandonment of religious justification, the government played on Pushtun solidarity as well as on the multiple
antagonisms present in a fragmented society, but the systematic nomination of Pushtuns to all key positions, and in particular the posts of
provincial governors, only served to exacerbate opposition by the
non-Pushtun population at a time when the language of nationalism
was failing to attract even the Pushtuns themselves.4 For example, in
3
4

Interviews in Logar, autumn 1992 and February 1997.

This policy was implemented in spite of official pronouncements and a number
of other measures such as the abolition of obligatory Pushtu courses and the
occasional use of Uzbek and Baluchi in the official media.

96

The Commanders

the province of Wardak the government armed the Pushtuns so that

they could attack the Hazaras, while in Nuristan it played on the
conflict between the Pushtuns and the Nuristanis. In both cases the
governments policy failed.
However, it was finally the policy of repression undertaken by the
authorities which alienated the governing class, the non-communist
intelligentsia and, before long, the entire population. Going beyond
anecdotal explanations and immediate causes of the uprising, it was
the violence of the state rather than its reforms that lay at the root of
the crisis. As often with revolutionary movements, a degree of paranoia on the part of those in control unleashed a cycle of increasingly
violent repression. When the doctrinaire vision of the Khalqis, which
in any case ran contrary to their own system of patronage,5 failed to
arouse any social response, they were obliged to impose their reforms
through a widespread resort to violence. Repression first affected
political opponents identifiable as such. Certain generals who had
been close to Daud and two former Prime Ministers, Musa Shafiq
and Nur Ahmad Etemadi, were immediately executed by the Khalqis. The Islamists already imprisoned under Daud were executed en
masse on 4 June 1979 in Pul-i Charkhi prison. Similarly, 18 January
1979 saw the assassination of ninety-six men from the Mujaddidi clan,
who were potential opposition leaders, while forty-two women and
children were arrested.
In proportion to the growth of opposition, the Khalqis took a
hostile attitude to entire classes of the population. Indeed, although
the communists were not the inventors of political violence, it had
hitherto been targeted against declared enemies of the authorities.
The communists broke new ground, launching campaigns of liquidation against the khans and the mullahs where there was no infraction or open hostility. In early November 1979 a list of 12,000 victims
of the Khalqi period was published by Yaqubi, the Minister for
Security, which gave an idea of the scope of the massacres carried
out in the prisons.6 Following the fall of the regime, mass graves
5

For example, from the spring of 1978 Afizullah Amin armed the members of his
tribe (the Kharuti) more on the basis of tribal patronage than on political
affiliation.
6 No precise count exists, but the policy of suppression during the Khalqi period
was probably the cause of tens of thousands of deaths. For the Pul-i Charkhi

The Revolt

97

were discovered, notably at Bamyan and Herat, which provided

evidence of the severity of the repression. Since the numerous notables carried off and executed without trial, including mullahs and
landowners, had been leading members of their community, their
disappearance brought in its wake expressions of solidarity.
During the spring of 1978 revolts multiplied.7 They did not have
an openly political dimension and mostly took the form of the capture of the nearest government outpost, the disarming of troops and
sometimes the execution of officers, mostly without any attempt by
the rebels to broaden their movement. Nevertheless, after some months
the countryside was out of control.
The geography of the rebellion illustrated two important phenomena. First, the role of the tribes was not a determining factor.
The myth which represents the uprisings as a new phase of the confrontation between them and the state must be set aside. The eastern
Pushtun tribes, although since 1929 they had originated the principal rebellions, maintained their calm till the spring of 1979, and in
some places even later. Their wait-and-see policy can perhaps be explained by the attitude of the Khalqi officials, who were able to contain them due to the existence of tribal solidarities. In Kandahar the
revolt did not occur till very late, just before the arrival of the Soviets. The rebellion of 1978 was radically different from previous ones
since its mobilisation took place within all the communities, both
rural and urban. The rejection of the regime was universal, a phenomenon yet more evident after the Soviet invasion.
Secondly, the uprisings were not organised: there was no overall
plan and there were no networks which would enable a revolt originating in one community to spread to the national level. Nevertheless, in the province of Badakhshan a number of Islamists, and in
prison alone Mike Barry suggests a figure of 27,000 executed during this period,
op. cit., p. 264.
7 The first revolt took place in Nuristan in May. See Vincent Schneiter, La guerre
de libration au Nuristan, Les temps modernes, special issue JulyAug. 1980. For a
list and a chronology of the uprisings see amir Etemad Daneshiyar, Jang-i Afghanistan wa shuravi amel forupashi jehani komunism (The war between Afghanistan and
the Soviets, a Factor in the Fall of the Communist World), Peshawar: Markaz-i
taqiqiyat-i enqelab-i islami Afghanistan (Centre for the Study of the Islamic Revolution in Afghanistan), 1992, p. 83.

98

The Commanders

particular Basir Khan, provided the impetus, while to the south of

Kabul, from Logar to Ghazni, the networks of mullahs linked to the
Khodam ul-Forqan coordinated the uprisings, though at the provincial level at most. Three examples will show the diversity of the
different situations: the rebellion in Herat, that of the Shiites in
Hazarajat and Kabul, and the student movements.
The Herat rebellion. The uprising in Herat in March 1979 was unique
in the Khalqi period. Although other provincial towns experienced
uprisings, especially the mutinies of the garrisons in Mazar-i Sharif
and Kabul, nowhere did power fall entirely into the hands of the insurgents. The government here suffered a severe psychological blow
and became aware that it faced opposition even within the towns. In
addition the suppression of the uprising led to the intervention of
Soviet air power directly from the Soviet Union, in an anticipation
of the invasion which was to take place in December. The Herat rebellion showed the complexity of the mechanisms of mobilisation
and in the end demonstrated their broad spontaneity.8
Agrarian reform was not a determining factor in the Herat rebellion. This had in essence been completed before the uprising began,
with little opposition due to the absence of ties of solidarity between
the large-scale landowners, who were often townsmen, and the farmers.9 On the other hand, the persecution of the religious lites and
the notables by the communist authorities seems to have been a decisive factor. The pirs and the ulema, who represented the religious
tradition in its most retrograde aspects, were the prime targets of
the Khalqis. Many were killed or imprisoned, while others went into
exile in Iran or Pakistan.
8

Essentially, these events have been reconstructed on the basis of interviews undertaken during stays in 1988, 1989 and 1993.
9 R. Grnhaug, Scale as a Variable in the Analysis: Reflections Based on Field
Materials from Herat in F. Barth (ed.), Scale and Social Organisation, New York:
Wenner-Gren, 1973; and Olivier Roy, Afghanistan, Islam and modernit politique, op.
cit., p. 143. According to the government press, 50,000 hectares were distributed
to 25,000 families up to the summer of 1979 (Kabul New Time, 6 August 1979); in
that year fifty-one cooperatives were formed, of which thirty-four remained in
1984 (Kabul New Time, 10 March 1984). The later disappearance of most of the
cooperatives is explained by the action of the mujahidin and the lack of government support.

The Revolt

99

Shortly before the uprising in the city, revolts took place in a

number of villages in the province, including Salimi, Ghoryan and
Zindajan. However, these were not organised and differed in their
immediate causes: a census in Salimi and forced labour on a road in
Gharyan. On the morning of 15 March the peasants of the surrounding area gathered around the mosques together with the townsmen
and, encouraged by the mullahs, converged on Herat, attacking all the
symbols of the state and of communism indiscriminately. The revolt
of the 17th division, which began immediately after that in the
town, brought the army on to the side of the rebels. This reinforcement enabled them, in the course of the day, to gain control of all the
government buildings. The success of the revolt in Herat was probably due to the confluence of two separate phenomena. First, the city
formed part of an urban agglomeration which ran from Zindajan in
the west to the Pushtun Zargun in the east. There was no clear geographical or ethnic separation between the town and the neighbouring villages. The rebellion therefore had at its disposal a wider
arena than in most instances, with the government unable to prevent
villagers from the surrounding regions entering the city. At Ghazni,
by contrast, the clear separation between the Pushtun countryside
and the Tajik bazaar hampered effective coordination and therefore
the capture of the town. The second factor related to the participation of the troops. The rebels, who did not have arms, would probably have been scattered by a military contingent that had remained
loyal to the government.
The city remained in the hands of the rebels for a week, and according to witnesses these were days of total anarchy. The bazaar was
repeatedly looted, contradictory rumours circulated, and no organisation appeared able to control events. A number of Soviet advisers
were killed although, contrary to what is sometimes said, their bodies
were neither mutilated nor dragged through the streets.10 Although,
according to one eyewitness, this explosion of violence was a revolution without a leader, a number of individuals nevertheless stand
out. Among the soldiers of the 17th division Sardar Khan, a Maoist
artilleryman, and Gholam Rasul Khan, an officer, played a significant
role in such coordination as existed. The future leaders of the prov10

It is likely that between 150 and 200 Soviet advisers were killed during the revolt. Other foreigners in the city were spared.

100

The Commanders

ince, Captain Ismail Khan and Allauddin Khan, had a lower profile.
On the civilian side an impression prevails that there was great confusion. Gul Muhammad, a khan from Gozargah, a Barakzai Pushtun,
came to the city with a number of rifles and headed one of the earliest armed groups, while two former convicts, Kamar-i Dozd and
Shir Agha Shongar, set up their own band.11 A committee was also
set up during the week of the towns liberation, and in spite of sometimes contradictory reports it appears that Shir Agha Shongar and
Kamar-i Dozd were its dominant personalities. Outside the city all
the uluswali (districts) were captured, except for the command posts
at Obeh and Pushtun Zargun. The surrounding provinces also rose
in revolt. The uprising of the Badghis followed that of Herat several
days later, and because news of the events had come rapidly to Qala-i
Naw, the provincial capital, was directly connected to it. In the south,
in Farah, most of the uluswali were captured several days before the
events in Herat, but there seems to have been no link between these
two uprisings.
After the onset of the rebellion, the Kabul government accused
the Iranians of having encouraged it because of a speech made by
Shariat Madari and an attack on the Afghan consulate in Mashad.
Around half the inhabitants of Herat are Shiites, and the government suspected them of being influenced by the Iranian revolution.
The return of Afghan labourers from Iran also prompted a propaganda move by Kabul, which claimed that 4,000 Iranians disguised
as Afghans had entered the country. Having denounced the foreign
role in the uprisings, the government embarked on a particularly brutal programme of repression. The commander of the base at Kandahar, Major-General Sayyed Mukharam, organised a force of some
thirty tanks and around 300 men to re-take the town. The soldiers
arrived at Herat on 20 March, waving Qurans and green flags, and
the rebels, convinced that the revolt must have spread to the whole
of Afghanistan, allowed the contingent to enter the town, while the
suburbs were bombed by aircraft which came both from the Soviet
Union and from the neighbouring air base at Shindand. It is still difficult today to know how many people were killed in 19789, but
11

Kamar-i Dozd was a thief, as is indicated by his nickname Dozd Shir Agha, on
the other hand, was from a rich family, but had squandered his inheritance and
spent several periods in prison. He was known for his extravagant tastes and predilection for prostitutes, male and female.

The Revolt

101

the discovery in 1992 of a mass grave containing 2,000 bodies northeast of the town enables a guess to be made. A figure of 25,000 victims was the estimate of government officials.12
This spectacular rebellion has been the subject of various analyses.
Giorgio Vercellin suggests an ethnic interpretation, stressing the secular opposition between the Pushtuns and the other communities.13
He refers to the observations of a 19th-century Hungarian Jewish
traveller, Armnius Vambry, according to whom there were sharp
antagonisms at that time between the newly-arrived Afghans, i.e.
the Pushtuns, and the local populations.14 However, the same evidence also acknowledges the assimilation of Pushtuns previously
resident in the country, and it may be supposed that this process had
continued, since none of the eyewitnesses questioned spontaneously
suggested that there had been an ethnic clash. In addition, the personalities of some of the leading figures, such as Gul Muhammad, a
Pushtun Barakzai, appears in itself to rule out such motivations, while
the first uprising took place in Salimi, a Pushtun village. Finally, nothing in the state of inter-ethnic relations in the province would allow
the supposition that the Pushtuns of Herat, often well assimilated
into the urban environment, might have been the object of resentment from the Persian-speaking majority. However, though research
does not permit the hypothesis of a rebellion directed against the
local Pushtuns, there remains the fact that the Khalqi period was
characterised by a takeover of the state by the Ghilzai and by eastern
Pushtun tribes. In this sense the rebellion of Herat, a Persian-speaking
town, did have anti-Pushtun implications. In a similar interpretation,
Richard Newell argues that the rebellion was Shiite and anti-Pushtun,15 but this runs up against the same obstacles, with the additional
difficulty of explaining the uprisings outside the city, where the
Shiites are very much in the minority.
12

Interview with a Parchami expatriate in Germany, 1989. Olivier Roy gives a

wide margin, estimating there were between 5,000 and 25,000 victims: Afghanistan, Islam et modernit politique, op. cit., p. 146.
13 Giorgio Vercellin, Afghanistan 19731978. Dalla repubblica presidenziale alla
repubblica democratica, Venice: Universit degli studi di Venezia, 1979, pp. 614.
14 Cf. Armnius Vambry, Voyages dun faux derviche en Asie centrale, Paris: You Feng,
1987, pp. 2401.
15 Richard Newell, The Government of Muhammad Mussa Shafiq: the last chapter of Afghan Liberalism, Central Asian Survey, 1982, p. 92.

102

The Commanders

Was the Herat rebellion at least partly the result, as Olivier Roy
believes, of the work of the Islamists?16 Although there were contacts
between the militants of the Jamiyat-i Islami and the officers of the
17th division, it is incorrect to suggest an infiltration of the army by
the Islamists. The Jamiyat-i Islami office in Mashad, whose activities
were favoured by the Iranian revolution, had begun to make contacts among the more senior officers of the 17th division in Herat,
including Ismail Khan, Allauddin Khan and Abdul Ahad, some
weeks before the rebellion. After the rebellions spontaneous outbreak the officers joined the rebels, with whomand this is the crucial pointthere had been no prior dialogue, for the good reason
that there had been no pre-existing organisation on the civilian side.
Furthermore, some of the leaders from the army side were Maoists,
such as Sardar Khan, which ruled out a mutiny organised by a group
of officers belonging to the Jamiyat-i Islami. The ulema also did not
seem to have been connected in any particular way with this group,
since the majority of them afterwards joined the Harakat-i Enqelab.
The disparity in the personalities of the leaderskhans, fugitives
from justice and officersallows the hypothesis of a rising organised
by a political movement to be excluded. The evolution of the rebellion was rather an indication of the weakness of political and other
organisations in Herat.17 In this perspective the reticence of the
ulema, who took only a minimal part in the rebellions active phase,
seems to have been especially revealing. In the end those who took
partpeasants, ulema and studentsare convinced of the spontaneity of the movement. The eyewitnesses of the episode, who were
also participants, stress the indiscriminate nature of the mob, the disorder, the air of acute crisis sometimes degenerating into arbitrary
violence. The one cry which echoed through the city was Allahu
Akbar (God is the greatest), while many of the inhabitants huddled
in their houses in fear of being denounced as pro-government or
Maoists, or falling victim to the hunt for those who were sarluchi
with uncovered heads, indicating lack of piety.
16
17

Olivier Roy, Afghanistan, Islam et modernit politique, op. cit., p. 145.

S. C. Stack notes the relative absence in the pre-war period of trade unions or
associations in local life: Herat, a Political and Social Study, unpubl. Ph.D. thesis,
University of California, Los Angeles, 1975.

The Revolt

103

The Shiite uprising. In Hazarajat a number of notables took an optimistic view of the communist coup: it at least delivered them from
Daud, whose aggressively nationalistic policy had led to the widespread installation of Pushtun officials.18 However, troubles were
rapidly unleashed by the announcement of the reforms brought in
by the regime, locally represented by young officials who were both
inexperienced and arrogant. These reforms were never to be applied
in reality, but the uprising was prompted by the provocative attitude
of the young Khalqis and by the victimisation of the khans and the
Shiite clergy in both Hazarajat and Kabul. Notables of the former
regime were particularly targeted, as was shown by the imprisonment of many former parliamentarians. In response government
buildings and especially schools, which were seen as centres of communist indoctrination, were destroyed by the rebels. In Hazarajat the
rebellion took on a character of unequivocal opposition to the state
and to the Pushtuns. It began prematurely in October 1978 immediately after that in Nuristan. Following the winter which severs all
communication in Hazarajat, local uprisings directed against the
administrative centres began in March and April. By June the entire
region was free of government control, which from then on was restricted to the fringes of Hazarajat, in the bazaars of Jaghori and of
Bamyan which is only partly Hazara. This success resulted particularly from the minimal administration present in Hazarajat, as well as
from the mountainous terrain and the lack of means of communication. The rebellion was in general both spontaneous and unanimous:
the entire population would participate in attacks on governmental
administrative posts. In some places khans led the rebels, in the tradition of Bacha-yi Gaw Sawar, but a hitherto unobserved phenomenon was the role played by the ulema, often the heads of private
madrasas, who thereafter formed the backbone of the new political
class of Hazarajat.
In the spring of 1979 the first signs of rebellion also appeared
among the Shiite community of Kabul. In this period of crisis the
indiscriminate repression practised by the government reinforced
18

To qualify this statement somewhat, it may be seen that certain khans still look
back with some affection to the era of Daud, which marked the onset of the
modernisation of Hazarajat, in contrast with the characteristic of Zahir Shahs
administration.

104

The Commanders

the solidarity between the two Shiite communities of the capital,

the Hazara and the Qizilbash. The Shiite religious lites were decimated by the Khalqis in 19789, especially in the towns. Religious
authorities such as the ayatollahs Waez and Akay Alem were incarcerated in the regimes prisons, as were young intellectuals such as
Binesh, Muhsenis deputy. In reprisal for this persecution, the Hazaras, who controlled this market, prevented the supply of wood to
the capital. At the end of May 1979 Amin met the leaders of the
Hazara community, but evidently no agreement was reached since
riots took place in Shendawal, the Shiite quarter of Kabul. An
attempt by four Hazaras to capture a police post caused the death of
a Soviet officer and unleashed extremely brutal measures of repression. The army used light armour, and several dozen died. Three
hundred Hazaras, picked up at random, were taken to the Pul-i
Charki prison and executed the same day. Helicopters dropped leaflets alleging Iranian involvement. It is true that on 21 June Iranian
radio broadcast a message from ayatollah haji Sayyed Hasan Tabatabay, urging the Afghans to continue their resistance. The roster of
the 12,000 victims who died under Taraqi, published by Amin on
10 October 1979, shows that 7,000 Hazaras were shot at Pul-i
Sharki in a few months. The fiercest suppression of the Shiites was
the result of fear of Iranian influence, widely overestimated, and of
traditional contempt for the Hazaras.
The student movements. In the towns, and particularly in Kabul, the
presence of secondary school and university students, sometimes
Maoist or Islamist militants, gave a particular character to the protests.19 In the capital the student demonstrations were incited by
leaflets (shabname) and appeals to jihad proclaimed at night from the
roofs of the houses. In particular the Maoists planned an action on
21 April, the Day of the Flag: the anniversary of the replacement of
the old red, black and green flag by the red flag of the revolution.
Leaflets were circulated on 19 and 20 April, and on 21 April the first
demonstration took place.
19

The focus here, for analytical reasons, is on the students. However, the general
atmosphere of the city was one of resistance even before the arrival of the Soviet
forces. Kabul would later be mobilised in its entirety against the Soviet occupiers
between 21 and 24 February.

Jihad

105

By 26 April the strike at the university was near-total. On 28 April

the police took severe action against a student march, leaving many
dead and wounded. The girls of the Surya and Rabe Balkhi Schools
then took the initiative of beginning strikes at school and encouraged the boys of the Ghazi School to follow suit. On 29 April a demonstration set off towards the university buildings. Six students
died under Parchami fire and some thirty were wounded. A girl
from the Rabe Balkhi School, shot dead by the Parchamis, became
the symbol of the student revolt. In another development the garrison at Bala Hisar, south of the town, rose up on 5 August 1979 at the
instigation of Maoist groups which had infiltrated it. The government then used its MiGs to bombard the fortress and rapidly retook
the base. In the following months police controls would increasingly
forbid open opposition.
These three cases demonstrate that despite a lack of organisation
there was unanimity among the population during this first phase.
The educated class and the town-dwellers in particular sided largely
with the opposition. A further factor was that the authorities failed
in their attempts to exploit communal antagonisms. This unanimity
was due in particular to the religious significance which had been
attached to the rebellion.

Jihad
At the outset the uprisings were fuelled by the multiple demands of
the population, and their moral rejection of political authority; later
their various justifications coalesced under the banner of jihad, which
served to legitimise the rebellion. Jihad is a religious concept which
veiled the political perception of the struggle, since it is an exchange
between man and God, and not between two adversaries. In 19789
the people did not mobilise against the communist government in the
name of an ideology: militants were rare among the rural population,
and their rhetoricincluding that of the Islamistswas incomprehensible to a population whose literacy was as undeveloped as its
politicisation. The political parties had not made themselves known,
or indeed were not established at all, until after the uprising, which
they exploited but did not initiate.
The religious interpretation was imposed on events thanks to a
conjunction of factors. Historical memory enabled Islam to appear

106

The Commanders

as the legitimation of popular rebellions against the invading kafir, as

had been the case with the British in the 19th century, and against
the state, as in 1929. With the Soviet invasion resistance was also
seen as patriotic, and the parallels between Babrak Karmal and Shah
Shoja20 became common currency in the pamphlets calling for jihad
against the occupier. In addition, Islam remained an essential component of individual identity. Religious practice was general in the
pre-war period, while atheism was practically unthinkable, at least
in the countryside, though the Uzbeks and the Hazaras were often
less observant than the Pushtuns. For the Pushtuns Islam was tightly
linked to tribal values, even when these did not have a religious origin, for example in the case of the concept of honour (namus).
Finally, the establishment of an Islamic republic in Iran, which had
not resonated only with the Shiites, presented a contemporary model
of religious revolution. In this initial phase there was no contradiction between these multiple references.
This religious interpretation of the rebellion was promoted by the
ulema and the mullahs, a group strongly united in their struggle against
the communist authorities, who had proclaimed a jihad against the
regime. Represented everywhere in the country, they constituted an
informal but efficient network for the transmission of information,
as the rebellion of 1929 had already shown. In instances where the
uprising was coordinated, for example in Logar or in Ghazni, the
ulema played the leading role. In most insurrections the sermons of
the mullahs were crucial: the people often assembled at the mosque
before marching on the government command post. In the mosques,
the habitual scene for discussion among the villagers, the mullah
would use his influence to put forward a religious exegesis of resistance to authority, and his intervention often served to convince the
hesitant by removing their doubts as to the illegitimacy of the authorities. Later the mujahidin gathered in military centres, but in villages far from the combat zones the mosque played a central role up
to the 1990s in bringing combatants together. Lastly the religious
leaders often enabled communal antagonisms to be overcome, but
this unanimity was only temporary, and the former splits, often reinterpreted in political language, later came to the fore again.
20

From the name of the amir put on the throne by the British in 1839, who became a symbol of collaboration with foreign imperialism.

Jihad

107

With the proclamation of jihad by the ulema the rebellion took on

a universal nature, aiming thereafter at the overthrow of the illegitimate authorities. The justification in terms of Islam gave a direction
to individual efforts: jihad is experienced as both a personal obligation and a religious duty.21 The concepts of mujahid and of shahid lie
at the heart of the conceptions of individuals in relation to the nature of their commitment.22 The declaration of jihad in particular
justifies emigration (hijrat) and the special treatment of mujahidin,
whose status is defined by the religious texts as a balance of duties
and privileges. The rules relating to purification, especially in the
funeral ceremony, are modified. In contrast to the normal practice,
the corpse is not washed, since the body of one who dies as a shahid
is pure, and paradise has already been promised to him. In another
illustration, according to a poster outside a mosque in Badakhshan,
mujahidin in combat are permitted to enter a mosque wearing shoes,
in accordance with an ancient legal principle dating from the era of
the Muslim conquests.
Following a long-standing tradition great fighters who have died
in battle become the object of popular veneration comparable to
that given to the pirs: In pre-war Afghanistan, martyrs were not distinguished from other holy men whose tombs were the object of
visitation.23 Thus at certain mosques in Herat prayers are still said for
the martyrs of 19th-century wars against the British. Frequently a
Front would retain the name of its first commander and in this way
place itself under his patronage. The exploits of the great mujahidin
who have died in combat are recounted, and placards about them are
displayed throughout the country. The prestige attached to the destiny
of a shahid reconciles those left alive to their loss. This may explain
the willingness of the mujahidin to sacrifice themselves, and the high
morale of the population up to the Soviet withdrawali.e. the exact moment when the concept of jihad became less relevant.
21

Jihad (etymologically effort) is the act of participating in the defence of the

Muslim community or in its expansion. See The Encyclopaedia of Islam, Leiden:
E. J. Brill, 1960, p. 551.
22 In the tribalised Pushtun regions the ideal of the warrior, who obeys the
Pushtun tribal code, replicates that of the mujahid, which nevertheless offers
stronger legitimation.
23 Dans lAfghanistan davant guerre, les martyres ne se distinguaient pas des autres
saints hommes dont on visitait les tombeaux, Pierre Centlivres, Micheline Centlivres-Demont, op. cit., p. 58.

108

The Commanders

Qowm and commanders

Conflicts, provided they are short, do not necessarily bring with
them any profound change in social organisation, and may even fulfil certain functions. Confrontations which occur at regular occasions in the year, outside the periods reserved for agricultural labour,
and which rarely occasion more than a few fatalities, are sometimes
seen as demographic controls. This functionalist approach, though
distinctly underestimating both real disturbances of the social balance and the regulatory role of the state, does offer an explanation of
the repetitive aspect of such confrontations.24 However, from 1978
onwards violence had unprecedented effects. The state withdrew
from the countryside, and the consequences of minimal administration and of military constraints meant that the mujahidin became
progressively less identified with the general population. A new social
role appeared: the commander.25 This was an individual, attached to
a party, who fulfilled at the local level the functions of military and
political leadership. In the 1980s there were several thousand commanders in Afghanistan, whose influence varied substantially, ranging from Masud, the organiser of an army of more than 10,000 men,
down to a village notable at the head of a handful of mujahidin. However, the commanders were aware of belonging to groups which
fought or collaborated to achieve the same goals.
What was the relationship between belonging to a qowm and the
process of mobilisation? Before the war elections displayed the mobilisation and coalition of ethnic groups for the election of deputies
24

On this, see Louis Dupre, Tribal Warfare in Afghanistan and Pakistan: a reflection of the segmentary lineage system in A. Ahmed, D. Hart (eds), Islam in Tribal
Societies, London: Routledge and Kegan Paul, 1984. Nevertheless, historical studies demonstrate the frequency of serious disruption in Afghan history. In Afghan
Turkestan the population accepted the arrival of the Kabul government in the
19th century, since the continual conflicts between khans were giving rise to
serious disorder and were perceived as unacceptable. In addition, a functionalist
account brings up all the criticisms usually aroused by an approach of this type,
and in particular the fact that war is only one possible means of demographic
control (in competition with such means as birth control and emigration), so
that why it occurs historically in this particular society remains to be explained.
25 It is also interesting that this word, Commandant, was itself new, and of foreign
origin. Amir is sometimes used of important commanders, but has not gained its
place in the common language to designate of this new social role. An analysis of
the local and regional organisations is given in Barnett R. Rubin, The Fragmentation of Afghanistan, New Haven and London, Yale University Press, chapter 10.

Qowm and commanders

109

(wakil) on the basis of local issues: for example, the Hazaras of Qala-i
Naw got their khan elected ahead of the Pushtun candidate, an
Uzbek might be elected by non-Pushtuns against a Pushtun candidate, and so on. In other cases concrete interests or a common way
of life, independent of ethnicity, provided a basis for electoral mobilisation.26 On the contrary, social classes were not groups with solidarity, nor could they be mobilised: there was no peasantry, and still
less a working class conscious of itself as such.27 During the war it
was possible to distinguish two situations: one where mobilisation
was within the framework of an existing qowm, and another where
the commander had mobilised several qowms or parts of qowms.
If mobilisation took place within the framework of an existing
qowm, the network of solidarity collectively sustained a party and a
commander. At what level is a qowm capable of itself becoming the
framework of mobilisation? Over-broad sources of solidarity, such as
macro-ethnic affiliation (i.e. Pushtun, Tajik and so on), do not constitute a framework for mobilisation, and nor do tribal confederations such as the Durrani or the Ghilzai. In the latter case Ghilzai
identity had become a historical relic, as Anderson remarks.28 In
fact the Ghilzai confederation was broken up by the appearance of
the Durrani confederation, and the political context no longer accommodated it. During the war this identity was not an active basis
for mobilisation, even though the Ghilzai were over-represented in
the Harakat-i Enqelab. The absence of extended qowms capable of
political mobilisation is accounted for in part by the action of the
state, which had destroyed the tribal structures.
A network of solidarity on a smaller scale, not necessarily of a territorial nature, generally served as the primary framework for mobilisation. Among the eastern Pushtuns tribes or the Ghilzai, the mobilised
qowm was generally the clan, several thousand individuals at most.
In the mountains the field of action of the commanders was often
26

See particularly Micheline Centlivres-Demont, Types doccupation et relations

inter-ethniques dans le nord-est de lAfghanistan, Studia iranica (Paris), 5 (2),
pp. 26977.
27 The analyses offered by Richard Tapper seem here to assume rather than to
demonstrate a transition to a class-based structure, Ethnicity and Class: Dimensions of Inter-Group Conflicts in North-Central Afghanistan in R. Canfield,
M. Shahrani (eds), op. cit., pp. 23047.
28 Tribe and community among the Ghilzai Pashtuns, Anthropos, 70, 1975, p. 596.

110

The Commanders

dictated by geography: from one valley to another, or at different

altitudes in the same valley, the solidarity networks differed. Similarly a fraternity of artisans practising the same trade could serve as
the basis for the establishment of a group.
To take a particular example, the fraternity of the delinquents of
Kandahar, the payluch (the shoeless ones), functioned as the framework of mobilisation under the leadership of Abdul Latif, an innkeeper before the war but now leader of the payluch and as such
responsible for relations with the local authorities. Little is known of
this brotherhood, which is similar to that of the koka in Kabul, the
luti (from lut, hooligan) in Iran, or the shabab in Tripoli (Lebanon).29
The brotherhood brought together the young men of the bazaar,
often the sons of butchers, innkeepers and such likeall professions
lacking prestige. Their distinguishing signs of membership were a
large knife, a yellow robe and distinctive slippers. The payluch were
obliged to observe a code of honour and solidarity between members, and therefore constituted a qowm. They organised dogfights
near a ziarat close to the town and smoked quantities of hashish.30
Given the responsibility for the security of certain streets, probably as
some kind of racket, they sometimes also acted as contract killers.
Though on the margin of legality, the payluch formed part of the social life of the bazaar in Kandahar and today former members are not
stigmatised. The son of the chief of the payluch, Gul Agha, who had
been educated up to the top class at the secondary school in Kandahar, had been a government official before the war. At the moment
of the Soviet invasion the payluch were massively mobilised behind
their leader to take part in the fighting against the occupiers. They
seemed afterwards to have disappeared as an organised group, since
many were killed in the early years of the war.
29

Interview with a former payluch (Kandahar, 1992). The organisation of young

men from popular urban origins into futtuwa was an invention of the caliph
Nasr, of which the payluch are perhaps a survival. Cf. E. Ashtor, A Social and Economic History of the Near Middle East in the Middle Ages, Berkeley: University of
California Press, 1976, p. 234.
30 There is a link stronger than that of mere geography between these three activities: the visit to the ziarat, the use of hashish and the dogfights. Pressing the point
somewhat, it could be said that the urban style of Kandahar is here perfectly expressedto which should be added music, which was ubiquitous in the town
before the arrival of the Taliban.

The Commanders

111

Although political mobilisation brought solidarity networks into

play, it was not a mechanical transposition of various qowms.31 Mobilisation might sometimes not be ascribable to a qowm, since a commander may benefit from the support of a number of networks or
partial networks. In some instances the qowm may support more than
one commander, the better to protect itself from political events. Political mobilisation in existing networks is autonomous, but whether
mobilisation takes place within an existing qowm or whether it brings
into existence a new structure, its mode of operation continues to be
that of a network of solidarity.

The social profile of the commanders

The social origin of the commanders mainly corresponded to four
ideal types: the holy man (the pir or the sayyed), the alem, the khan or
the educated man. These were not primarily political or military figures, but could become so through a capacity to mobilise men and
resources. These different types of status were related to resources
of different kinds: respectively religious charisma,32 religious knowledge, wealth, and the modern education sanctioned by the state. In
addition, the khan, the pir and the alem were all important figures in
their local communities; the influence of the educated class was less
definite. However, the prestige attached in the villages to learning
was undeniable, and social contacts made at school or at university
opened up access to political networks. For those lacking initial resources social mobility was limited: ordinary fighters could assert
themselves at the local level through their courage or their organisational abilities, but very few went on to run larger groups.
These four ideal types did not exist in the real world, where for
the most part the situation was more complex. A single individual
could bring to bear a diversity of resources, such as his status as an
31

For example, A. Shalinski studied a group of muhajirin of Fergana and noted that
they belonged simultaneously to the Hezb-i Islami and to the Jamiyat-i Islami:
Ethnic reaction to the current regime in Afghanistan, Central Asian Survey
3 (4), 1985.
32 It should be recalled, following Max Weber, that charisma is a social and not a
psychological phenomenon. It is the recognition by a group of disciples of the
exceptional quality of an individual, in this case a religious quality. Economie et
societ, vol. I, Paris: Presses Pocket, p. 320.

112

The Commanders

alem, his territorial inheritance which might earn him a role as a

khan, and his membership of a family of pirs. So the principal commander in Kunduz, Aref Khan, an educated man who was the son
of the former mayor of Kunduz and a major landowner, had at his
disposal a number of different resources. Men of the educated class
who became commanders often belonged to notable families, and it
is therefore difficult to distinguish between what they owned to
those connections and to their scholarly achievements. Where an
educated man did not belong to the dominant branch of an influential family, such as Amin Wardak or Fazlullah, his educational status
might allow him to be preferred to other candidates for the position
of commander. In other cases pirs were often also significant landowners and acted as khans, to whom they might up to a point be
assimilated. Pirs and sadat might also be ulema, at least in the regions
such as Herat and Ghazni that were dominated by orthodox Sufism.
In Hazarajat khans often gave their daughters in marriage to sadat.33
On the other hand, ulema and khans were normally distinct categories in pre-war Afghanistan, especially among tribal populations.34
The emergence of the new lites allows these types of resources
to be re-evaluated in the context of social upheaval. As a result of the
jihad, religious personalitiesulema, pirs and sadatfound their legitimacy enhanced. In addition, local social structures governed the
probability that an individual of a particular social background might
become a commander. In fact, the political map showed regional patterns in the social origins of the commanders, to an extent which
rules out a random distribution: for example the ulema were dominant in the province of Ghazni and in Hazarajat, and the khans
maintained their hold among the Baluchis of Helmand. In some
sense the value of religious prestige or scholarly achievement, and
their capacity to be transformed into a political resource, varied from
region to region. Afghanistan was not socially homogeneous, and
this was especially true after the disappearance of the state.
33

R. Canfield has described the relations between pirs and mullahs in the Bamyan
region, Ethnic Regional and Sectarian Alignments in Rural Afghanistan, in Ali
Banuazizi, Myron Weiner (eds), Religion, the State and Ethnic Politics, Syracuse
University Press, 1986.
34 However, Jamal Malik, op. cit., p. 244, on the basis of a study of the social origins
of students in the madrasas of the NWFP (some of whom were Afghans), points
to a not insignificant number of sons of khans in the recent period (11.3%).

The social profile of the commanders

113

What tendencies emerged? First the sweeping scale of the changes

must be stressed. The national lites were virtually eliminated. The
presence of the ulema was the most novel factor, together with the
rise of the educated classes. Regional differences were a factor: in
different areas pride of place was taken either by the educated or the
ulema. The changes also had a generational aspect: most of the commanders were born in the 1950s, since the material conditions and
the high mortality rate favoured younger fighters, who were physically capable of guerrilla warfare. Among the khans the new generation took the form of a transfer of power to heirs who had received
an education. These points will become clearer through individual
analysis of each of the four categories defined above.
The ulema. In a number of places ulema or mullahs became commanders, and this direct and long-term exercise of power was something
new in Afghan history. Two factors tend to explain this development: their legitimacy in the eyes of the rural populations and their
capacity for organisation. The retreat of the ulema to the countryside
in the two or three decades before the war restricted their influence
to rural areas. The example of the Mujaddidi family, which transferred its principal madrasa from Kabul to Ghazni, typified a longterm sociological change: the tendency for the ulema to become displaced from urban centres, especially if they had not come from the
government madrasas. Sometimes financially dependent on their congregations, the ulema were only rarely major landowners in the prewar period, and at least in the rural context they remained strongly
rooted in their community, the majority of them exercising their functions in the region of their birth.35 This close link explains why men
of religion have often acted as spokesmen for the population, e.g.
in Badakhshan where they occupied the position of an alternative
authority.36 In addition, when the state withdrew from the countryside after 1978, the ulema, in their capacity as religious scholars, had
at their disposal a legitimate and recognised field of knowledge which
35

Some tribal communities, however, made a point of choosing their mullah from
outside their own group.
36 M. N. Shahrani, Causes and Context of Differential Reactions in Badakhshan
to the Saur Revolution in M. N. Sharhani, R. Canfield (eds), op. cit., p. 152.

114

The Commanders

allowed them to establish legal relationships between individuals.

The war therefore had the effect of giving the ulema a monopoly in
the field of justice, which reinforced their influence.
The second factor was related to the existence of structures and
networks. The ulema and those mullahs who had followed a course of
religious teaching formed part of an institution held together by the
strongest personal ties. An alem would remain faithful to the master
who had given him his ijaza (licence to teach), and to his fellow pupils. If he was the director (mudares) of a madrasa he could also rely on
his students (taliban) and their families. Further, outside the traditional madrasas, where the teaching was rigid and repetitive, the government madrasas and a number of private madrasas educated ulema
who were of a modern tendency and more aware of political issues,
who would become effective administrators.
The ulema-commanders, who existed throughout Afghanistan,
were dominant in Ghazni, in the north of Helmand, in Badghis, to
the south of the town of Mazar-i Sharif, and in Hazarajat.37 What
were the social structures which facilitated their success? First of all,
the weakness of the khans favoured the emergence of the ulema on
the political scene, as is seen in Ghazni. In this province there were
numerous khans of the Andar tribe, but they had little influence as
individuals. The weakening of the ties of patronage therefore had its
effect, and the mullahs had not faced strong competition. In the deprived regions of Helmand, which were very different from the central regions where irrigation projects had brought about a profound
social transformation, the local khans were unable to mobilise a significant clientele. On the other hand, deep-seated segmentation was
a brake on the appearance of the ulema as political leaders, so that the
khans were dominant among the eastern Pushtun tribes, despite the
exception represented by mawlawi Haqqani of the Jadran tribe. In
addition, the rural nature of the communities concerned made the
establishment of the ulema easier since there were no alternative urban
lites. In the second place, the presence of a prestigious madrasa, such
as the Nur ul-Mudares at Ghazni, was a factor common to most of
the regions dominated by ulema-commanders, such as Waras and
Nili in the Hazarajat. These madrasas were the centres of networks of
37

In Logar, the ulema were in competition with the educated class.

The social profile of the commanders

115

ulema who were able to coordinate their activities over what were
sometimes far-flung regions.
The pirs and the sadat. In Afghanistan, as in the rest of the Muslim
world, particular individuals or objects may enjoy the reputation of
possessing a beneficial spiritual quality, barakat, which reflects on to
those who come to view them. The pirs and the sadat are the two
categories of individuals who enjoy this spiritual quality. These posessors of religious charisma have substantial social effects, since their
disciples (murid) are potentially available for mobilisation in a political cause.38
The sadat, descendants of Muhammad through his daughter Fatima
and his son-in-law Ali, enjoy no particular privileges in Islam. They
are not automatically recognised as the bearers of barakat, but nevertheless command a certain prestige39 and may enjoy a privileged economic position, though there also exist poor sadat. Villagers may ask
sadat to live among them in exchange for gifts, in order to benefit
from their barakat.40 Sometimes specialising in pious rituals, they
serve as intermediaries and as arbiters, for example in the case of a
conflict within a tribe. Their status varies: in the east they have little
prestige, while in Hazarajat they are influential.
The pirs are spiritual masters, normally within the context of a
Sufi brotherhood.41 Their relationship with their disciples may be
ritualised through participation once or twice a week, often during
the night of Thursday or Friday, in a zikra mystical exercise consisting of a recitation of the names of God under the pirs leadership.
Professional solidarity may also be expressed through the cult of a
deceased pir, who may be adopted as the patron of a group of artisans
practising the same trade. Popular religion is also expressed, though
38

Louis Dupre, Saint Cults in Afghanistan, American Universities Field Staff Report, South Asia Series XX (i), May 1976.
39 The khwaja, descendants of the first Caliph Abu Bakr, have a lesser degree of religious legitimacy but often play a comparable social role.
40 Micheline Centlivres-Demont, Types doccupation et relations inter-ethniques
dans le Nord-Est de lAfghanistan, Studia Iranica 5 (2) 1976, pp. 26977.
41 The Qadiriyya and the Naqshbandiyya are the most prevalent Sufi orders,
although the Chestiyya and Suhravardiyya are also encountered. There are also
pirs who do not belong to a brotherhood, and malang, vagrant preachers who
have no organised clientele.

116

The Commanders

less consistently, through visits to the pir and through pilgrimages to

relics or to the tomb (ziarat) of a pir, which is often under the guardianship of the descendants of his family.42 These practices are widespread. The ziarat of Mazar-i Sharif, where a local Afghan tradition
maintains that Ali, the Prophets son-in-law, is buried, attracts tens
of thousands of pilgrims every year.
In other places the pir presides over a khanaqah (a place where
Sufis gather) which functions both as a mosque and as a madrasa,
although this function had declined in the pre-war period. A murid
would normally be assured of a welcome from his master, who
would when appropriate exploit the relationship to his own benefit.
The attachment of a murid to his pir would take the form of a gift in
cash or in kind made once or twice a year on the occasion of a visit
by the murid or of a tour by the pir. This source of finance covers the
upkeep of the khanaqah but the surplus of the gifts of the disciples
over the expenses of the pir might be considerable, since the murids
often come from several provinces. The pattern of pir-murid relations
is structurally that of patron and client, although the goods involved
are of a different kind. The hypothesis may be advanced of an extension into the sacred domain of more general social models.43
In the context of Sufism there are two ways of approaching the
exchanges between the pir and the murid.44 Orthodox mystical practices form an integral part of the teaching of the ulema. On the contrary, practices connected with the cult of saints are viewed by the
ulema as magical rituals. Richard Tapper suggests on the basis of a
case-study that individuals who take part in Sufism of the saintly
variety are often of inferior social status and experience feelings of
inadequacy in relation to the values of their group.45 Whatever the
truth of this, the distinction between the two types of Sufism is not a
matter of difference between one brotherhood and another, but between different pirs and the communities from which they recruit
their murids. Among the tribal populations the Qadiri pirs, but also
some naqshbandis, are generally attached to a saintly cult.
42

The guardians (mutawali) usually sell amulets (tawiz) to visitors.

On Christian saints, see Peter Brown, Le culte des saints, son essor et sa fonction dans
la chrtient latine, Paris: Editions du Cerf, 1984.
44 See Olivier Roy, Sufism in the Afghan Resistance, Central Asian Survey 2 (4),
December 1983.
45 Richard Tapper, op. cit.
43

The social profile of the commanders

117

In the pre-war period the flow of both manuscripts and oral transmissions seems to have dried up, although the scarcity of historical
sources must cast some doubt on the reality of the supposed decline
of Sufism, a theme which recurs in literature on the subject. It remains true that the brotherhoods did little recruiting within the
young educated class or in the towns, other than in regions such as
Herat. Bo Utas relates this decline to the hereditary status of the pirs,
so that their disciples were not motivated to aspire to excellence in
order to become their successors.46 This probably accounts for a certain ossification in the style of teaching.However,the principle of succession within families is long-standing, so that the decline, a more
recent phenomenon, must be explained by other factors, in particular
the competition offered by modern schools from the 1950s onwards.
To what extent were such charismatic figures able to make use of
their legitimacy to achieve the status of commanders? In the pre-war
period certain pirs or sadat were personalities of local influence, who
were sometimes even elected as members of Parliament. The economic and political effects of religious charisma did not therefore
appear only with the onset of the war. Because of the long tradition
of struggle of the Naqshbandis against the Russian colonisers in the
Balkans, Central Asia and the Caucasus, it might have been thought
that this brotherhood would become a motivating factor in the jihad.47 In fact the brotherhoods have not been at the heart of the resistance, and have not been the framework for the parties, as has been
seen in Turkish and Iraqi Kurdistan at various periods.48 With some
exceptions, mainly in Herat and Maymana, the pirs did not become
commanders and there was no collective loyalty among murids to
any particular group.
Locally the sadat occupied significant positions, for example in the
northeast, but did not make up greatly extended networks. The only
exception concerns Hazarajat, where the sadat were dominant almost
46

Bo Utas, Notes on Afghan Sufi Orders and Khanaqahs, Afghanistan Journal 7,

1980, p. 2. For the theory concerning the decline, see also David B. Edwards, op.
cit., p. 187.
47 It should be noted that the brotherhoods have not played a determining role in
the current war in Chechnya.
48 See Martin van Bruinessen, Agha, Shaikh and State, London: Zed Books, 1992,
mainly chapters 5 and 6.

118

The Commanders

up to 1982, though in alliance with the ulema.49 To be a sayyed helped

notables who were also major landowners to preserve their local clienteles. In the last resort the war did not produce charismatic figures, mad mullahs such as appeared in the 19th and 20th centuries
in the tribal revolts against the British on the Afghanistan-Pakistan
frontier.50
How might the relative eclipse of the brotherhoods be explained?
The pirs were the victims of the Khalqi repression in the 19789 and
in consequence many left for Iran or Pakistan. Further more the
brotherhoods, with their recruitment among the urban and educated groups in decline, did not have men at their disposal who were
able to become guerrilla leaders, and suffered from the effects of competition from other lites, the educated class and the ulema, who had
taken control of various parties and tended towards action of a more
organised kind. Nevertheless it will emerge in the course of the analysis of the biographies of the leaders of the political parties that membership of a brotherhood was still a considerable advantage. Thus the
influence of the Mujaddidi family depended partly on its position
49

In Hazarajat the influence of the sadat was related to that of the khans, but with
the eclipse of the khans the networks were broken up, since the sadat were not
spread throughout Hazarajat. Though numerous in Yakoalang and at Behsud,
they were virtually absent from the south of Hazarajat. This factor partly explains the initial alliance between the sadat and the ulema who were more evenly
distributed. In addition, most pirs are sadat, which explains the significant clientele they enjoyed and also their competitive relationship with the Shiite clergy.
Traditionally the role of the sadat was to provide links between social groups, especially at times of crisis. Their activity during the war may be seen as an attempt
to perpetuate this role, though on a scale and in ways which were different. The
failure of the sadat was due to the impetus of the Iranian revolution, which
favoured the rise of clergy trained in Iran who were generally hostile to popular
forms of religion. The conflict between the ulema and the sadat was therefore
also a competition for the monopoly of religious legitimacy. Nevertheless, the
sadat were not eliminated in contrast with the khans, and their prestige among
the lowest classes of the population allowed them to resist from their strong
points, for example Waras, Behsud and Naur. A major landowner such as Sayyed
Hasan Jaglan was able to mobilise hundreds of men at Naur. A sayyed such as
Beheshti, on the other hand, was able to use his status as a mudares to speak on
equal terms with the ulema and to mobilise his taliban.
50 See Akbar S. Ahmed, Millennium and Charisma among Pathans: a critical essay in social anthropology, London: Routledge and Kegan Paul, 1976.

The social profile of the commanders

119

within the Naqshbandi brotherhood. The same observation holds

for the relationship of the Gaylani with the Qadiri brotherhood.
The khans. A khan is a local notable who exercises an influence related to his status as a landowner or as a tribal leader. He maintains a
clientele whose size is measured by those who are obliged to him,
from among the relationships in which he is dominant. He feeds
others and obtains services for them: he cannot operate without an
open table, gifts to the local mullah and so on. He therefore invests his
capital and his time in a network of relationships while personifying
the values of his group, in particular that of generosity. He also plays a
role as intermediary between the community and the outside world,
whether government officials or a passing stranger, and he thereby
obtains numerous advantages, particularly through his dealings with
the administration. Nevertheless, his behaviour is directed towards
the acquisition of prestige, rather than material benefits. Thus in his
relationship with a mulgerey (companion) the khan feeds or helps his
client, but this clearly differs from the payment of wages. In effect
the khan makes use of his companion, who remains a free man, as an
emblem of his status as a person of influence.51
The power of the khan consists primarily in his capacity to exert
influence. He has no means to impose his will other than persuasion,
debate and gifts. His power has no legal basis, although in certain regions such as Hazarajat and Badakhshan his domination over the
peasants can be very forceful. This is not a matter of traditional power,
since there is no handing down of immutable and more or less sacrosanct rules, but rather of a system of influence which is relatively
open to newcomers. Between khans there is a degree of competition
for dominance, and personal qualities are obligatory to attract continued fidelity.52 The son of a khan does not automatically inherit the
position of his father, although the power of acquired status tends to
favour family continuity.
Two types of khan are distinguished by the more or less hereditary
nature of their situation. In the eastern Pushtun tribes or among the
51

See Whitney Azoy, Buzkashi: so that his name shall rise, Philadelphia, PA: University of Pennsylvania Press, 1982.
52 Fredrik Barth, Political Leadership among Swat Pathans, London: Athlone Press,
1959.

120

The Commanders

Ghilzai the competition between khans results in a degree of instability among individual positions, while in other cases, there is what
might almost be described as an aristocratic form of authority. The
great families of Kandahar in particular, but also those of Hazarajat,53
of Badakhshan and among the Baluch, enjoy a legitimacy which is
founded not only on their economic resources but also on their
genealogy, which is in itself a source of legitimacy.
After the 1960s economic transformations acted to the detriment
of the clientele relationship between the khans and the peasants. Relations lost their mystique, becoming more overtly economic. Social
relations altered, and the very idea of a clientele began to erode. In
Ghazni, for example, the introduction of tractors broke the cycle of
reciprocity, and the position of khan tended to disappear.54 In Hazarajat the khans lost their erstwhile social supremacy. Their decline did
not date only from the war, but had already been in evidence much
earlier, probably from the end of the 19th century when the Hazara
tribes had been conquered by Abdul Rahman Khan. However, social changes from the 1950s onwards further accelerated a decline
which was to the benefit of the merchants and craftsmen. Everywhere sedentarisation, economic liberalisation and the alleviation of
insecurity tended in the direction of increasing the influence of the
state and the weakening of the khans power.
How did the khans react to the new conditions imposed by the
war? Those whose clientele was at the village level did not have the
same level of resources to dispose of or employ the same strategies as
those who wielded their influence on a provincial or national scale.
The disappearance of the great notables was a major consequence of
the war. In Hazarajat they were overthrown in the very regions where
their power had been most absolute, such as the Uruzgan province
and Jaghori, while many smaller khans had survived in regions such
as Behsud. These were notables who had sometimes possessed several dozen villages. In the Kandahar region the great Durrani fami53

According to the particular place in Hazarajat, there exists a quasi-aristocratic

system with much importance ascribed to genealogies (in the province of Uruzgan), or a system of competition between notables, similar to that of the eastern
Pushtun tribes (this applies in Behsud).
54 Jon Anderson, There are no Khans any more: economic development and social
change in tribal Afghanistan, op. cit.

The social profile of the commanders

121

lies went into exile, sometimes to Quetta, and no commanders of

any significance were drawn from their ranks.
The power of the khans suffered as a consequence of the fall from
grace of the state partly because of their role as its representatives.
However, this effect should not be exaggerated: the khans lost power
mainly because of the link between military force and hostile policies. Assimilated to the old feudal order, they were exposed to the
hostility of the Islamists and, in a less obvious way, of the ulema, not
to speak of the government, for whom they represented an easily
identifiable target. In some cases pressure from the Islamists, which
went as far as assassination,prompted their withdrawal from the political scene, especially in Laghman. In Hazarajat the ulema sometimes
bitterly opposed the landowners, whom they accused of injustice
against the peasants. At Yakaolang, where most of the landowners
were sadat, open conflict was to last for several years until the ulema,
supported by the peasants of the neighbouring mountains, eliminated
the great landowners of the valley. The khans, preoccupied with their
personal rivalries, showed themselves unable to band together to
make common cause against either the ulema or the Islamists. More
significantly, they did not even succeed in holding on to their local
power by calling into play their relationships with their erstwhile clients, which speaks volumes about the violence of the social changes.
In addition, the richest and the most influential disposed of the
economic and cultural resources to be able to establish themselves
easily in Pakistan or in the West. Those khans who had connections
with the governing class therefore went into exile and played only
an indirect, though sometimes significant, political role, such as those
from Kandahar who established themselves in the Pakistani city of
Quetta. On the other hand, at the village level the khans were able to
set themselves up as commanders, since here they did not encounter
the competition of the educated class or the ulema, whose numbers
were insufficient to administer the population in all the villages.
Regional variations were explicable in terms of the intensity of
fighting and the power of the tribal system. In regions little touched
by the war, the khans retained their dominance, since the war was a
less significant agent of change. Furthermore, the lessening of economic activity had the effect of preserving established positions, at

122

The Commanders

least for the land owners. The Baluchis in the south of Helmand and
the Aymaq provided instances of this persistence in the social structures. In addition, in some tribalised areas such institutions as the
jirga became once more the real locus of power, since the commanders were often the heads of clans. This phenomenon of retribalisation was to be seen in all the Pakistan frontier regions, where the
solidarity of communal affiliation and the tribal code did not permit
autonomous political activity.
The educated class. Though restricted in numbers, the return of the
educated group to the countryside was important since it provided
leaders for the resistance and particularly for Islamist parties.55 As the
fighting became more generalised, many educated young men left
the towns or returned from abroad, making their way back to their
villages of origin where their families were known. Masud, a native
of Panjshir, had lived mainly in Kabul, and from 1974 in exile in
Pakistan, before settling in his native valley in 1979. In the same way
Zabihullah returned to Marmul, at a time when he was teaching in
Mazar-i-Sharif. Dr Fazlullah and Amin Wardak, who were students in
Kabul, returned respectively to Baraki Barak in Logar and to Jeghatu
in Wardak.
The lack of leadership was in fact the crucial problem for the
Afghan guerrillas. Organisation, especially in the military field, necessitates a minimal level of technical competence, and in particular the
ability to read and write. Members of the educated class therefore
became commanders, especially since their status also gave them the
ability to negotiate directly with the groups in Peshawar.
However, the elevation of a member of the educated class to the
role of commander generally also required either affiliation to a party
or the support of a family. For example, in the first phase of the war,
militants returned to Afghanistan with a few followers and imposed
themselves locally as commanders, without consultation with the
55

In the case of Badakhshan the principal commanders, with the exception of

some khans, had emerged from the educational system, which was well developed in this region. On the other hand it would be hard to argue, even in
Badakhshan, that all the leaders were educated men, as Shahrani does (op. cit.,
p. 164), in view of the contrary examples of Sayyed Wakil and of Kheyradmandwho were respectively a khan and an alem.

Models of organisation

123

local population, as Masud had done in Panjshir. In certain places in

Hazarajat such as Lal o Sarjangal, or Deh Kundi (Khedir), power was
overtly seized by force, without prior political effort. Several dozen
young Hazaras, who had gone to study or work in Iran, came back
armed and took control of their regions of origin.
Nevertheless, the adjustment of the young townsmen to life as
guerrillas was far from easy. Many stayed in Peshawar, unable to find
a niche in Afghanistan. The rejection of westernisation, which was
identified with the state, placed the educated class in an awkward
position. Anecdotally, but significantly, table mannersthe return to
the traditional style of eating with the hand from a common dish
caused the younger men some distress. Some were never able to
adapt to these new living conditions and preferred to undertake political work in Peshawar or in the West.
The presence of educated commanders was particularly frequent
in the northern regions of Afghanistan, such as Herat in the northeast. In the south, where tribal organisation dominated, the educated
class had difficulty in imposing itself in competition with the khans.
In Hazarajat or in Ghazni the educated group was pushed aside by
the ulema.

Models of organisation
The commanders? They have taken the place of the old khans, but they are
more powerful! (Ahmad Shah Masud in an interview, autumn 1991)

Analysis of the social origins of the commanders allows the way they
exercised power to be scrutinised, with two preliminary remarks. First,
the correspondence between a commanders social background and
a particular way of exercising power is only on the level of probability.
For example, an educated man might exercise power in the manner
of a khan, in a very traditional way. Secondly, one should emphasise
the independence of the commanders from the parties, which explains
why party affiliation is not brought in at this stage of the analysis. In
fact there was little control over the activity of the commanders by
the parties based in Pakistan or Iran. In the absence of national coordination, the strategy of the commanders remained largely autonomous, especially in the early years of the war. Local political alliances
were made on the initiative of the commanders, and within a single

124

The Commanders

party there might be contradictory strategies. Also precise lines were

not laid down by the parties in relation to the organisation of mujahidin groups. Discipline, such training as might be offered, and the
way of life differed little from one party to another, but depended
rather on the social background of the commander and the community within which he operated. On the basis of actual cases, two
models of how power was exercised might be constructed: the patrimonial, which was the more normal at the start of the war, and the
institutional in the case of the complex organisations which progressively took control of widespread areas.
The patrimonial model. In the patrimonial model there was no clear
separation between the public and private domains. The commander made no distinction between his personal revenues and those of
the party, or between his personal actions and those in which he represented a political movement. His status was dependent on his personal reputation or his family rather than an institution. His field of
action was defined by the extent of the networks of solidarity which
supported him, and in particular of his extended family, occupying
all posts of importance. In addition, the system of succession to power
was symptomatic of the low level of institutionalisation and of the
weight of family connections. Since the status of commander was
part of the inheritance of the individual, it could be handed on, in
most cases to brothers, or sometimes to a son. The party to which the
commander belonged would restrict itself to the ratification of the
choice of the new commander.
In a patrimonial system the commander exercised personal power,
without delegation or administrative structure. No attempt would
be made to establish an administration, or to set up schools or courts.
On the contrary, the people were obliged to pay a mullah to give primary instruction to children, and to turn to an alem on judicial questions. Nor did the commander embark on the establishment of a real
military organisation since, even when he headed an armed force, he
behaved above all as a private person. No hierarchical structure was
instituted: rather, the commander maintained relations with his clientele through favours and gifts. If he was sufficiently wealthy, he
lodged his mujahidin in his own house and fed them, which reinforced
his personal authority.

Models of organisation

125

The commander partly took upon himself the traditional functions of the khan in the relationships of the community with the exterior. He would exploit his position to enrich himself and to
become an owner of land. His position as an intermediary effectively
enabled him to appropriate a proportion of humanitarian aid, as well
as of any assistance provided by political or religious factions, and of
local resources, whether by means of taxation or of seizure. He
might also marry into influential families desirous of making an alliance with the local authority. The guesthouse, where travellers of
standing would normally be entertained, was a perquisite of the commander, and he kept open table in order to maintain his reputation
for generosity. The distinction between a khan and a commander lay
in the absence of relations with the state in the case to the latter,
although he might be surrounded by the wider network of partisan
solidarity. While the power of the khans was entirely informal, the
commander could impose his will by force and was not obliged to
rely on appeal to the consensus or on the use of indirect pressure.
The institutional model. As distinct from the patrimonial model, the
institutional model constituted a bid to set up an alternative state,
with regulations and a civil and military administration. Many educated commanders established organisations resembling this model,
including notably Masud in Panjshir, Ismail Khan in Herat, Zabihullah in Mazar-i-Sharif, and Najmuddin in Badakhshan. However,
numerous traces of patrimonialism survived: for example bodyguards
tended to come from the home village of the commander, while personal profiteering by leaders continued to be a frequent occurrence.
While in a patrimonial system the commander controlled networks of solidarity, in an institutional system he exercised authority
over a population occupying a defined territory, imposing consistent
standards concerning the treatment of individuals. Solidarity networks
ceased to be the only paradigm for the recruitment of officials and
mujahidin. There was a meaningful distinction between public and
private goods, and a system of taxation could be established. Officials
ran a rudimentary administrative system, expressing the aims of their
organisation in ideological language. Succession to power, though not
necessarily following precise rules, takes place outside the commanders family. Military organisation was professional, and the mujahidin

126

The Commanders

were sometimes paid, recognising hierarchical authority and occasionally wearing uniforms. Objectives were conceived on a broader
scale, with strategies developed over longer periods. On the basis of
this general model, three ideal types can be constructed: a clerical
model, set up by the ulema; a state model, set up by the educated
class; and a partisan variety, established by the Islamists.
The clerical model sprang directly from the involvement of ulema,
which did not imply that all ulema organised a system of this type:
some remained with the patrimonial system of organisation.Administration was conducted according to the principles of Islamic law.
Official positions were given in principle to ulema or to their taliban,
who were normally pupils of the same madrasa. Two examples of this
model may be cited. In the province of Helmand in the 1980s, the
commander ras Abdul Wahid set up a typically clerical system at
Baghran, in the province of Helmand. All the officials were taliban
from his madrasa. The mujahidin, who were closely controlled and
professional, wore uniform and observed strict discipline, while authorisation from the commander was required for population movements. In the province of Ghazni the administration of the provincial
shura (council) was entirely in the hands of the ulema. However, no
strict hierarchy existed among the ulema, as at Baghran: in contrast a
more consensual system prevailed, although Qari Baba continued to
be the dominant personality up to 1994. These characteristics were
in essence those which would later be found among the Taliban,
who appeared to implement the same system on a national scale.
In the state model a bureaucratic organisation took upon itself the
functions of government, such as education, taxation and justice. No
attempt was made to impose a precise ideological model, beyond the
consensual principle of the Islamic state. This type of organisation
was therefore linked not directly to an ideology but rather to a social
class. The recruitment of officials was carried out among the educated
class, including soldiers and state functionaries,56 but not necessarily
among the Islamic militants as the example of Herat showed. In this
province in 1980 Ismail Khan set up his own regional organisation,
the Emarat, open to commanders belonging to all parties. He was
56

The commanders often recruited their officials in the universities where they
themselves had been students: engineers in the case of Masud, teachers for
Zabihullah, and soldiers in the case of Ismail Khan.

Models of organisation

127

not himself an Islamist and his relations with the local Islamist network, which was run by the Afzali family57 and Nurullah Emat, were
difficult. Ismail Khan was a former officer, and he attempted to reproduce within his organisation the functions of the state. From the
beginning of the war he was unique in recruiting his senior aides
with no regard for communal or political affiliation. Of the four leading figures in the Emarat, three were from outside the town proper,
namely Ismail Khan and Allauddin Khan, who were both from
Shindand,58 and Muhammad Aref from Shamali. These three outsiders had in common the fact that they were all former soldiers.
They were far from being the only military men in the Emarat, and
it was certainly this link with the military which gave the organisation its coherence at the top.
The partisan model was a variant of the state model, in which the
underlying structure was more the party than the state. The structures were of a different type than those which have been described,
since the movement had revolutionary principles. There was some
similarity with Maoist guerrilla movements, although the ideological structure was different. The population was led by militants with
ideological commitment, in addition to, or sometimes instead of,
technical abilities. The objective was to politicise the population.
Masuds organisation was the most developed example of this kind
of structure in Afghanistan.59 He had been a militant Islamist since
adolescence, and his involvement in the movement drove him into
exile in Pakistan in 1974, whence he returned in 1975 to take part in
57

Hafizullah Afzali was a student in the university at Kabul in the 1970s, and took
part in the coup in Panjshir in 1975, where he was killed. See Abdul Hafiz
Mansur, op. cit., p. 52, and the anonymous Biography of Commandant in Chief
Safiullah Afzali and the Resolution of Afghan Mujaheddin and Refugees, Liestal:
Bibliotheca Afghanica, n.d. His brother led a Front in Ghorian, in the province
of Herat, and never acquiesced in the leadership of Ismail Khan, in particular because of his lack of legitimacy in the eyes of the Islamists.
58 Whether the uluswali of Shindand was regarded as belonging or not to the province varied from one period to another.
59 There are many written accounts of Masud and films about him, but few are of
any quality. The demands of the media, which tend to focus attention on
Masuds personality, obstruct understanding of his political organisation. JeanJos Puig gives a relevant analysis of his career in Le commandant Massoud in
Grard Chaliand (ed.) Stratgies de la gurilla, Paris: Payot, 1994. See also Abdul
Hafiz Mansur, op. cit.

128

The Commanders

the Islamists abortive coup.60 After a period in Nuristan he took up

residence in Panjshir, where the rebellion broke out in July 1979,
and eliminated rival groups, in particular the Maoists. For this professional revolutionary ideological issues were central, in spite of the
priorities of combat. In 1980, through the agency of his brother, he
embarked on the establishment of political teaching for the peasants
in Panjshir: in the event the classes failed, and the plan was abandoned. Officials used very political language,61 and their horizons
were not limited to Afghanistan: their movement fell within an international revolutionary context. Another factor was that Masuds
expansionism provoked a long series of clashes with other commanders, which brought about, for example, his military occupation of
the neighbouring valley of Andarab. Masud was also the only commander to organise on a substantial scale training courses for Afghan
guerrilla leaders as well as for foreign Islamists, who included Kashmiris, Turks and others. Internationalism was a characteristic trait of
Masuds model, at least up to the fall of Kabul in 1992.
In the face of attempts by lay members of the educated class to
build political organisations, the rural population kept its distance. In
reality there was a fundamental ambiguity in the relationship between the population and the educated class: where the latter wished
to establish an impersonal political relationship, the rural population
interpreted this new type of authority differently to avoid recognising the regulation being imposed on them, and to continue relying
on personal relationships, shading sometimes into clientelism, which
60

Masud was born in 1956, the third of six sons, into a wealthy family in the
Panjshir valley; his father was an officer. He spent his childhood in Kabul, and attended classes at the French Istiqlal lyce (although he rarely spoke French, he
retained his competence in the language). He studied at the Polytechnic high
school, and joined the Islamist movement. After the failed coup in the spring of
1974, in which he did not take part, he fled to Pakistan. He was trained by Pakistani officers and led the group which carried out the coup of 1975 in Panjshir.
In 1976 Masud, who was close to Rabbani, found himself accused by Hekmatyar of having betrayed the movement, on the basis of statements obtained under
torture from a friend of Masud, Jan Mohammad, who was later assassinated by
Hekmatyar. Though arrested, Masud escaped death and returned to Panjshir in
the spring of 1979. See Anthony Davies, A brotherly vendetta, Asiaweek, 6 December 1996.
61 Quoting for example Mackinder, one of the founders of geopolitics. Interviews
with officials of the Shura-yi Nazar, autumn 1991 and spring 1992.

Finance

129

they found easier to manipulate. The issue was not the acquisition
of power but its justification, and how it worked in practice. Therefore, when Masud travelled in the country, and as the number of
personal petitions increased, there was a short-circuit of the regular
procedures in favour of an appeal to a charismatic personality who
makes just decisions in the light of a moral code. In the short term
the commanders power appeared to be enhanced, but in fact his
political programme was jeopardised. The recognition of charisma
functioned in this instance as a transaction, in effect a compromise,
between the institutional motivations of the educated class and the
apolitical goals of the peasants.
In theory, the cult of personality would be a response to this ambiguity, combining the recognition of a leader with the process of
organisation undertaken by the educated class. The educated class
were the driving force behind the transition to a cult of personality,
since they were able to use it as a means of imposing ideological or
administrative constraints in the name of the charismatic personality.
This phenomenon was observed, up to a point, in the regions controlled by Masud: it was in these regions that tensions were strong
since the model was the most ideological.

Finance
Economics lay behind the choice of a war of attrition, with offensives over a limited time involving only a few men. The commanders did not administer the economy, which could otherwise have
been a source of income, since they had neither the organisation nor
the legitimacy to do so. Even movements between the mujahidin and
the governments regions were only occasionally controlled, generally with the aim of obstructing some move by Kabul. The government did undertake massive purchases of wheat from time to time
with the aim of blocking the mujahidins supply chain, using the nomads as intermediaries; this caused a rise in prices. However, the
economic dimension was not absent from strategy, examples being
the creation of bazaars and the diversion of trade routes as a result
of the clashes and rivalries between commanders, particularly in
Badakhshan and Hazarajat. There was also interaction between military logistics and economics, since the commanders strove to control

130

The Commanders

the means of communication. The resources of the commanders could

be seen as deriving essentially from three sources: taxation, external
assistance62 and the traffic in drugs.
Taxation. The issue of taxation was important in two ways. First,
taxation was an important source of local finance, which enabled the
commanders to be independent of both the parties and external aid.
It was therefore a good indicator of the state of relations between
commanders and the population. The levies raised by the mujahidin
were in theory Islamic taxes.63 Zakat is raised on cash and on goods,
such as flocks, and mineral extraction, to the level of 2.5% of the
value held in a year. Ushur, raised on harvests at the rate of 10%, was
of considerable importance, since the commanders controlled few
bazaars, at least at the start of the war. The proceeds of these two
taxes are normally given to the poor, to needy travellers and to the
mujahidin, which served as a justification for their appropriation by
the commanders. Still, there was an evident dilution of the Quranic
principle that the recipient of this obligatory gift should be freely
chosen.64 As an exception to the rule laid down above, there were
also non-Islamic taxes, in particular tolls on the roads. The commander here enjoyed a resource which was all the more advantageous in
that it did not risk disaffecting the local population. Taxation on
shops, which does not seem to have an Islamic basis, was in general
low. Finally, a widespread form of participation was to feed the
mujahidin in the nearest markaz. This was the first form of popular
contribution to the resistance. For all these taxes self-assessment of
contributions was the rule rather than the exception, owing to the
lack of specialised officials, and also to the refusal of the peasants to
allow the commander to interfere in their affairs. Since many regions
were politically fragmented, the payment of tax was the symbol of
affiliation to a particular commander, from which sprang the strategy
of seeking protection by paying taxes to several commanders.
62

In the absence of reliable information, no attempt will be made to analyse in detail the distribution of military aid.
63 Joseph Schacht, An Introduction to Islamic Law, Oxford University Press, 1982.
64 The zakat should be distinguished from the sadaqat, which is a voluntary act of
charity.

Finance

131

In the early years of the war contributions from the population

were widespread and spontaneous. The turning-point came as the
mujahidin became increasingly professional in 19856, with the growth
in foreign aid, especially from the United States. Popular contributions were sidelined, at least in the case of the major commanders,
who relied on external humanitarian and military aid. In many regions the population found themselves deprived of a bargaining
counter and unable to exert pressure on the commanders.
Humanitarian aid. It may appear controversial to put humanitarian
aid among the resources available to the commanders. However,
although the avowed objective of the NGOs was to give aid to the
population,65 the commanders profited from it directlyand indirectly through its influence on their prestige or legitimacy. The distinction between humanitarian aid and aid not regarded as such is in
any case open to criticism. Assistance given by Islamist networks did
not admit a distinction of this type. The building of a hospital or the
gift of arms served the same objective, namely to ensure the victory
of the jihad. However, the two routes were distinct, both from the
point of view of the donors and in the conditions under which aid
was given. Thus the distinction should be maintained.
In the early years of the war humanitarian aid had two characteristics. First it was crisis aidessentially gifts of cash, together with a
medical presence and the setting-up of dispensaries. The aid came
mainly from private organisations run by well-wishers, rather than
from institutional organisations as became the case still later. By their
nature the medical organisations were more professional, but with
almost that sole exception aid operations at that time had a distinctly
amateur and sporadic character: the amount of aid provided was
economically negligible, its importance resting above all on the reports of the situation it brought to the attention of western public
opinion.66 A significant evolution took place from 1986 onwards
65

Some 256 NGOs have been involved in aid to the Afghans, of which about fifty
operated in Afghanistan. Helga Baitenmann, NGOs and the Afghan war: the
politicisation of humanitarian aid, Third world Quarterly 12 (1), January 1990,
pp. 6285; and Pierre Centlivres and Micheline Centlivres-Demont, Etat, islam
et tribus face aux organisations internationales. Le cas de lAfghanistan, 1978
1998, Annales 54 (4), JulyAugust 1999.
66 For example the testimony given in the United States by Juliette Fournot, the
head of the MSF (Mdecins Sans Frontires) programme in Afghanistan, to the

132

The Commanders

with the arrival of American money.67 Professional sources of finance

provided an impetus towards the professionalisation of the staff. The
NGOs, which were highly dependent on institutions (e.g. agencies
of the United Nations and of the European Community and
USAID), were no longer autonomous in their strategy and became
the agents of their backers. The borderline between NGOs and
international organisations therefore became less well defined.
Humanitarian aid was a significant factor for a commander, as the
number of commanders or their representatives who hovered in the
anterooms of the NGO offices in Peshawar made evident. Besides,
the humanitarian aid agencies were often the sole source of information on the conflict, and they naturally spoke about the regions
they knew. In consequence most journalists followed in the tracks of
the humanitarian workers because of their logistical facilities, and
the existence of a source of information and a viable story. The political map therefore became distorted to the benefit of those commanders who enjoyed humanitarian aid. The reverse of the coin was
that the prestige of a commander might suffer if he could not guarantee the security of the humanitarian aid workers.
In addition, a commander did not usually have the means to undertake administrative action, in particular because of the lack of
doctors, teachers and engineers. Humanitarian aid enabled a response to be made to the more urgent demands of the population,
particularly in medical care. For example, a dispensary brought a
commander little income, but he would be credited for his astuteness in having persuaded an NGO to work in his territory. Aid was
also a resource for the commander in that he was able to find positions for his clientele and above all his family within the NGO project. In addition, since the commander ensured the security of the
NGO, he would arrange, normally at inflated rates, for it to pay a
number of mujahidin. The security argument also allowed the comCongressional Task Force on Afghanistan, Hearing on Medical Operations in Afghanistan, 4 March 1985.
67 US $400 million for the period 19804 (combining humanitarian and military
aid) and $250 million for the year 1985 alone: John H. Lorentz, Afghan aid: the
role of private voluntary organisations, Journal of South and Middle Eastern Studies
XI (12), autumn-winter 1987, p. 103. $250 million were attributed to international agencies between 1985 and 1989: Helga Baitenmann, op. cit., p. 75.

Finance

133

mander to exercise a strong influence over the situation in which

the project operated, which was the guarantee that he could keep it
under his control. In numerous cases personal profit to the commanders was an acknowledged fact, particularly where land was reclaimed. Sometimes such projects related to zones in which the
commander had a personal interest, but in other cases profit came
from a tithe on the aid provided. This was not necessarily seen as illegitimate by the population, all of whom depended on the size of
the amounts held back.
In large organisations, which were emerging from the patrimonial
model, the NGOs played an important role in the functioning of the
administrative apparatus, as with Masud and Ismail Khan, though at a
lower level in the case of the latter. Masud, who had set up a relatively complex administration, was entirely dependent on the aid of
the NGOs in all civil matters, such as health, reconstruction and, to a
great extent, education.
In addition to the western countries, Islamic networks such as the
Muslim Brotherhood or the Salafists (the Ahl-i-Hadith) were active
from their bases in Peshawar. The Muslim World League, whose director in Peshawar in the 1980s was Abdul Hasan, deployed the largest
budget for the maintenance of the mujahidin, amounting to several
million dollars annually. Saudi Arabia, the most important donor, did
not appear to control the use of its funds closely, and local employees
were generally identified with the Muslim Brotherhood. Additionally, the Saudi Red Crescent was funded directly by the Saudi
government, but here too the personnel often belonged to the Muslim Brotherhood tendency. Among the other groups, the Maktab-i
Khidmat-i Mujahidin (Mujahidin assistance office) was centred at
the beginning of the war around Abdullah Azzam and his assistant
Osama Bin Laden. In particular this group published the magazines
Jihad and Bunian-i Makhus. Its finance came principally from the
Gulf and from British Muslims. The distribution of aid to the
mujahidin was not centrally planned, and this fragmentation was a
reason for the multiplicity of sources of finance and the groups present in Peshawar. An additional issue was that sources of finance such
as Syria, Iraq, Saudi Arabia and Libya were often antagonistic to each
other politically. A further factor was the cooperation between the

134

The Commanders

Muslim Brothers and the Saudi financiers, in spite of their ideological separation. The situation of Abdullah Azzam is an indication of
this distribution of responsibilities.68
Drug trafficking. The third source of finance was the opium trade.69
After 1979 circumstances came together to cause a real explosion of
production in Afghanistan. The revolution in Iran had the effect of
bringing to a halt all production in a country which was a traditional
consumer, with around 2 million users, and also an entrepot on the
smuggling route to Europe and the United States. In addition, Ira68

Abdullah Azzam is a figure who throws much light on the mobilised networks.
Born in Palestine in 1941, he was of Jordanian nationality. After reading Muslim
law at Damascus University from 1967 to 1973, he took a doctorate in that
subject at Al-Azhar and taught Islamic Studies at the University of Jordan from
1973 to 1980. In 1981 he became a professor at King Abdul Aziz University in
Jeddah, but lived in Peshawar where he took part in the jihad. He died in November 1989 as the result of a car bomb in circumstances that remain obscure.
69 Though illegal in Afghanistan, the consumption of hashish was tolerated, and it
was not unusual to see smokers in parks or buses, as well as in saqikhana (places
where opium and hashish were smoked). Opium was rarely used, and only
among the poorest classes. Part of the production was exported, mainly to Iran,
via Herat to Mashad and Chakansur, sometimes after being refined in Kabul.
Smuggling to Europe and to the United States was routed through Kabul airport. The profit derived from opium was significant, more so than from hashish,
which explains why the zones of consumption and production did not necessarily coincide. For example, the Jalalabad and Balkh regions were producers, but
without significant consumption, since the opium was destined for export or for
the Turkmens. Opium was frequently grown for sale because of the price it
commanded, while hashish was consumed locally. The cultivation of the poppy,
which is planted in the autumn, forms part of a cycle of cultivation including
wheat and vegetables. In January 1972 the police confiscated pure heroin and refined hashish prepared for export, which marked the discovery of an illegal trade
no longer carried out on a local level but forming part of the international market. The 1970s also saw an increase in consumption. While hashish was frequently
used, opium continued to be confined to the most deprived populations. It is
likely that the opening of Afghanistan to tourism was an important factor, both
in the increase of production for export, and in consumption, through imitation.
In early 1972 the Afghan government set up two committees for the suppression
of the opium traffic, which indicated the onset of an awareness of the problem,
and perhaps reflected an apprehension of appearing to be subject to American
pressure like Turkey. The reduced production of opium in Vietnam after 1973
would also have had the effect of increasing production in Afghanistan and Pakistan.

Finance

135

nian smugglers fled to Pakistan to avoid the authorities and continue

their activities. By coincidence, the harvest was bad that year in Asia,
which increased international prices. At the same time (in February
1979) Pakistan decided, because of Zia ul-Haqs policy of Islamisation, to ban the production and use of opium. Hitherto a low level
of production had been permitted, and use was tolerated in official
government venues. Because the crop in 1979 was exceptional, and
the price increased because of the ban, the Pakistani producers,
mostly in the North-West Frontier Province, converted their opium
into heroin. Thus all the circumstances came together to make Afghanistan a major producing region: the increased international prices, the
existence of refining laboratories and international smugglers in Pakistan, and the absence of control over the Afghan countryside.
Three regions produced 90 per cent of the Afghan opium exported: Badakhshan, Helmand and Nangrahar. The north of Helmand70 was one of the worlds leading areas of production Opium
might be described as the dominant crop, since it occupies the largest land area and yields the most revenue. Two factors favoured the
producers. First the Baluchi smugglers, of Iranian and Afghan nationality, switched from smuggling electrical goods to trafficking
opium. Secondly, the principal commanders of north Helmand, beginning with the most important among them, mullah Nasim, gave
their blessings to the crop by providing transportation and security.
Although mullah Nasim did not himself cultivate opium poppies, he
levied a tax on exported opium which guaranteed him useful revenue.
The situation was very different in Badakhshan. The principal
producing areas were Fayzabad, Keshem, Jurm and Baharak. How70

The province of Helmand may be divided into three areas clearly distinguishable
from each other, ethnically, economically and politically. Baluchi country begins
in the south, from Mirabad, in the alaqaderi of Deh Shu. Durrani Pushtun tribes,
the Alizai, the Ishaqzai and the Alikozai, occupy the northern districts of the
province: Musa Qala, Nausad, Baghran and Sangin. The central part of the provinceMarja, Lashkargah and Girishkis ethnically very mixed because of the
irrigation projects. There are also Pushtuns, sometimes from Farah or Wardak, as
well as Hazaras and Uzbeks. In each of these three regions the economy is different, with a predominance of dry crops in the north. The peasants of north
Helmand began to produce opium in pre-war times for export to neighbouring
Iran via the Baluchis. Production was carried out on a local basis and alternated
with other crops such as wheat.

136

The Commanders

ever, it appears that in Khash 60 per cent of the land was given over
to the production of opium, which made it the largest crop. Similarly, in Peshkan the land was not irrigated and therefore did not
lend itself to other forms of cultivation, so that the peasants grew
only poppies. Traditionally the population used opium, especially the
Ismailis, and poppies were grown on small individual landholdings.
Badakhshan was a province always on the edge of famine. Its situation greatly deteriorated in the 1970s, because of population growth.
In this context the cultivation of poppies was of the highest importance for peasants who were thus able both to obtain money and to
ensure their own personal access to the drug. Although the commanders did not earn large returns from the growth of poppies, it
was politically difficult for them to forbid it because of the overall
level of poverty.
Production in Nangrahar, primarily destined for export, began
long before the war. Opium, which in pre-war times had been a secondary crop, now became the principal source of revenue for many
families, especially in Nangrahars poorest districts. In prosperous regions the proportion of land given over to the cultivation of poppies
was some 10 per cent, and this figure increased for poorer lands. In
contrast to Badakhshan the local commanders were directly involved
in cultivation, as landowners, and in its transportation to Pakistan.
The presence of the same tribes in both Pakistan and Afghanistan
facilitated the movement of the drug and its delivery to the North
West Frontier Provinces laboratories.

4. The Jihadi Parties

In parallel with the drive towards mobilisation inside Afghanistan,
political organisations made their appearance in Pakistan and Iran.
They adopted a stance opposed to the regime in Kabul, whose legitimacy they wholly rejected, and established duplicate authorities, in
a situation characteristic of civil war. In the analysis of the political
situation we first examine the formation of these parties, followed by
the exclusivity of their representation of the mujahidin. Finally a typology is presented, relating to the various types of ideology, organisation and recruitment.1

The formation of the parties

In 19789 the situation was so fluid that dozens of organisations
attempted to turn popular mobilisation to their advantage. This was
a chaotic period characterised by ephemeral alliances, when attempts
to achieve unity were as rapidly followed by schisms. Parties canvassed known personalities to preside over their coalitions, although
in the event all that was on offer was the opportunity to create new
organisations. In opposition to the initial movement towards unity,
which was justified in terms of the military and moral exigences of
the jihad,ideological divergences and competition for resources served
as a contrary impetus towards fragmentation. The pattern of schisms
which led to the emergence of the Sunni and Shiite parties is examined below in detail.
In 1978 Hezb-i Islami and Jamiyat-i Islami, which were already in
existence in Peshawar, came together as the Harakat-i Enqelab-i
Islami (Movement for the Islamic Revolution), under the leadership
of mawlawi Muhammad Nabi, a former member of Parliament and
an associate of the Mujaddidi family. Nabi laid claim to the political
inheritance of the pre-war Khodam ul-Forqan, since the assassina1

See Barnett R, Rubin, op.cit., pp. 196 ff.

137

138

The Jihadi Parties

tion of Ibrahim Mujaddidi and his son Ismail had left the Mujaddidi
family without a leader.2 In fact the most politicised harakati commanders recognised the continuity between the two movements.
However, the strains between Hezb-i Islami and Jamiyat-i Islami
continued, and each soon took its own way, with Harakat-i Enqelab
remaining under the leadership of mawlawi Muhammad Nabi.
Mawlawi Muhammad Nabi subsequently surrendered the leadership to Sayyed Gaylani, a pir of the Qadiri brotherhood and a member of the pre-war governing class, in a new movement, the Mahaz-i
Melli (National Movement). However, mawlawi Mohammed Nabi
rapidly recovered control of the Harakat-i Enqelab, leaving Gaylani
at the head of Mahaz-i Melli. In addition the Jebhe-yi Nejat-i Melli
(National Salvation Front) was established as a new coalition under
the presidency of mawlawi Sebghatullah Mujaddidi, who had been in
Peshawar since October 1978.3 However, most of the partiesincluding Hezb-i Islami, Mahaz-i Melli and Harakat-i Enqelab
refused to participate in this Front. The Jamiyat-i Islami became for a
time Mujaddidis sole partner, but later it regained its independence,
leaving Mujaddidi as the leader of Jebhe-yi Nejat-i Melli.
At the same time Hezb-i Islami split at the end of 1979. Mawlawi
Khales emerged as the leader of a party that retained the same name
it will be referred to as Hezb-i Islami (Khales). His separation from
Hekmatyar may be explained in terms of personalities: mawlawi
Khales probably felt that the difference in their ages and his own
position as an alem gave him precedence over the young Islamist.
Khales style, with little concern for organisational niceties, also put
him at odds with the Hezb-i Islami line.
In January 1980,when the Islamic conference took place in Lahore,
the six partiesHezb-i Islami, Hezb-i Islami (Khales), Harakat-i
Enqelab, Jamiyat-i Islami, Jebhe-yi Nejat-i Melli, Mahaz-i Melli
embarked, at the urging of Jamiyat-i Islami, on a process of unification which resulted on 19 March 1980 in the formation of an alliance, the Ettehad-i Islami Baray Azadi-yi Afghanistan (Islamic Union
2

Another consideration was that the only remaining heir, Muhammad Amin
Mujaddidi, was too young: he spent the war years in Islamabad.
3 Sebghatullah Mujaddidi, the founder of the Jebhe-yi Nejat-i Melli, was unable to
lay claim to the inheritance of his cousin Ibrahim. In the event, Sebghatullah was
not fated to become the pir of the Mujaddidi family, and in any case before the
war he took a controversial position close to that of the Muslim Brotherhood.

The formation of the parties

139

for the Liberty of Afghanistan). The secretary-general Sayyaf, an

Islamist alem recently freed by the communist regime, then went
into an alliance with Hekmatyar against the other movements, but
afterwards broke with Hekmatyar while retaining control of the
Ettehad. At this point the seven Sunni parties came into existence.
Nevertheless in 1980 the number of more or less autonomous
groups was still substantial, and the monopoly enjoyed by the seven
parties was imposed only piecemeal. Besides the Maoist groups (see
Chapter 6), there were regional parties, mainly in the province of
Kunar.4 Kunar was the first region to rebel in 1978, and due to its
position on the Pakistani frontier it rapidly escaped government control. Owing to the violence of the fighting, only 20 to 30 per cent of
the population remained there out of 330,000 inhabitants before the
war. Independent fronts made their appearance: thus Muhammad
Anwar Amin, a Nuristani khan, became leader of the Jihad-i Islam-i
Nuristani (Islamic Jihad of Nuristan), a party which found itself
in competition with Da Islami Jahad Da Para Da Kunar Da Qowmuno Ettehad (Alliance of the aqwam of Kunar for the Jihad), formed
in 1980. In the early months, the parties in Peshawar provided arms
and sent men to fight alongside the Nuristanis, except for Hezb-i
Islami, which tried to take military control of the region. Muhammad Anwar became the principal Nuristani leader, though due to
lack of support his movement was progressively marginalised.
In the case of the Shiites the politicisation of the confessional
parties, which had already begun before the war, became the rule
after 1978.5 Religious sentiment, mobilised through the appeal to jihad, precluded membership of the Sunni parties6 and favoured the
leadership of the ulema, who were seen as the coherent force within
the community. The Shiite political scene was characterised by opposition between the predominantly Hazara parties, including Shura,
Nasr and Sepah, and the parties with predominantly non-Hazara
4

David J. Katz, Responses to Central Authority in Nuristan: the case of Vygal

Valley Kalasha in M. N. Shahrani, R. Canfield (eds), op. cit., pp. 94119.
5 See Rolf Bindemann. Ethnizitt und Gesellschaft. Religion und Politik bei den
schiitischen Hazara in Afghanistan, Iran und Pakistan, Berlin: Das Arabische Buch,
1987, Occasional Papers no. 7.
6 There were several instances of Shiites joining Hezb-i Islami. The Mustazaffin, a
small group whose membership was largely Shiite, also accepted Sunnis.

140

The Jihadi Parties

leadership, such as Harakat-i Islami and the Mustazaffin, which

were present either outside Hazarajat or in its border areas.
During the war a number of parties either coexisted or confronted each other in Hazarajat. However, in the earlier period the
Shura-yi Enqelabi-yi Ettefaq-i Islami-yi Afghanistan (Revolutionary
Council of the Islamic Union of Afghanistan) held sway as a consultative organisation. The Shuras earliest manifestation was a gathering of several hundred independent delegates and representatives of
parties, whose aim was to establish an autonomous authority in
Hazarajat.7 The assembly was held at Waras, in the province of Bamyan, in August and September 1979, and appointed Sayyed Beheshti
as its leader. This Shura was an initiative unique in Afghanistan in
terms of the number of delegates who came together to agree on effective cooperation over such an extensive territory. Even before the
Soviet invasion, the convening of the Shura signalled Hazarajats political renaissance. Meanwhile the central authorities, careful to avoid
gathering together the Hazaras into a single administrative entity, had
divided them up into several provinces. The assembly at Waras demonstrated the power of Hazarajats identity, at a time when among
the Sunnis the shuras of the mujahidin were systematically replicating
the existing administrative divisions. The only similar case among
the Sunnis was the Nuristani party mentioned above.
One of the principal parties which took part in the Shura was the
Sazman-i Nasr (Party of Victory). Nasr was founded by a group of
ulema trained in Iran who had been active in the pre-war years both
in Hazarajat and among the Hazara community in Kabul, Ghazni
and Mazar-i Sharif. Some of them had travelled in Syria, Lebanon and
Palestine, and had fought against Israel in the 1970s. However, the
movement devoted itself primarily to political propaganda: the operation of libraries and publication of leaflets (shabname). Nasr was
officially established in the summer of 1979 in Mashad by three ulema,
Mir Hussein Saddeqi, Azizullah Shafaq and Abdul Ali Mazari.
The founders of the third significant party in Hazarajat, the Sepahyi Pasdaran-i Islam (Army of the Guardians of Islam), were intellectuals who were active in Iran at the moment of the revolution. In the
7

At Yakaolang, after the capture of a military post, a committee was formed to participate in the constitutive assembly of the Shura-yi Ettefaq: in a good example of
the unanimous spirit of those times this included members of Nasr, khans of the
Harakat-i Islami and the principal of the local school,who had no party affiliation.

The formation of the parties

141

liberal atmosphere of the early days of the Iranian revolution, under

the government of Bazargan, various Shiite parties8 came together
to set up the Jebhe-yi Azadibakhsh-i Afghanistan (Front for the Liberation of Afghanistan) whose name was derived from an organisation linked to the Iranian Pasdaran. Two or three months later these
parties separated, after setbacks which resulted from ideological differences and from the antagonism between ulema and lay members.
The young intellectuals who had taken a key role in this Front did
not immediately set up another party, but linked up with the Iranian
Pasdaran to fight in Afghanistan. At the time the leader of this group
was Ibrahim Qazimi, an engineer in his thirties from Behsud, educated at the university in Kabul. The decision to give official recognition to the Pasdaran as an Afghan movement was taken by the
Iranians in 1983, following a two-month trip to Hazarajat by the
Iranian Jaffar Zade, who visited Ghazni and Behsud. The party then
adopted the title of Sepah-yi Pasdaran-i Islam (Army of the Guardians of Islam), choosing as its leader a former member of the Shura,
Akbari, an alem trained in the Iraqi city of Najaf. Ibrahim Qazimi,
who had ceded his influence to the religious establishment, migrated to Canada in early 1984 and disappeared from the political
scene. The creation of the Sepah by Iran was the result of a desire to
take back control of the Shiite movement: Nasr was regarded as too
independent, in particular as the result of its history before the Iranian revolution. Sepah also represented the victory of ulema who
were close to the Iranian clergy over young intellectuals who were
too independent for Teherans taste.
Outside Hazarajat the Khalqi governments assassination of shaikh
Waez, who was the representative of the Ayatollah Khoei in Afghanistan, left shaikh Mohseni as his heir and one of the leaders of the
non-Hazara Shiite community, although he had mainly been known
in Kandahar. After a narrow escape from police who had come to
arrest him in the spring of 1978, Mohseni removed himself to the
Iranian city of Qom where he established the Harakat-i Islami.
The Sazman-i Mujahidin-i Mustazaffin (Organisation of the Poor
Mujahidin) was a small party of intellectuals based at Shashpul in the
province of Bamyan, and was the only organisation in Afghanistan to
8

Nasr,Niru,Rad,Nehzat,Mujahidin-i Khalq (Mustazaffin),Jombesh-i Mustazaffin.

142

The Jihadi Parties

accept both Shiites and Sunnis, both in practice and in terms of

doctrine. The movement seems to have originated in the 1960s in
the time of Zahir Shah, with the Pasdaran-i Enqelab-i Islami (Guardians of the Islamic Revolution), which in 19778 became the party
of the Mujahidin-i Khalq-i Afghanistan (Mujahidin of the Afghan
People). In 1979 the party adopted its present name. The other
Shiite parties were sparsely represented on the ground, sometimes
mustering only a few dozen men. Their origin often lay in the ties of
patronage between Hazara ulema and Iranian ayatollahs: this was the
case with the Nehzat-i Islami-yi Afghanistan (Islamic Movement of
Afghanistan), the Hezb-i Dawat-i Islami (Party of Islamic Preaching),
and the Hezbollah.9 The Nehzat-i Ruhaniyat wa Jawan-i Afghanistan (Reja) (Movement of the Afghan Clergy and Youth), the
Niru-yi Islami-i Afghanistan (Islamic Force of Afghanistan) and Rad
(Thunder) had leaderships drawn more from the laity and attracted
less support from Iran.
Monopoly of representation. Over time the exiled parties gradually
established their exclusive right to speak for the mujahidin, especially
through their ability to define the parameters of legitimate ideology,
as well as through their control of resourcessubject to the control
exercised by the countries in which they operatedand through
the membership of the commanders.
The parameters of legitimate ideology. The concept of jihad, which in
the eyes of the great majority of the population was the sole legitimate principle, entirely governed the doctrine of the political parties.
The plainest indication of the centrality of religion was the position
occupied by the ulema. In the event, the absence of a professional political class in the pre-war era and the significance invested in reli9

Hezbollah was not in reality a structured party but rather an ensemble of groups
financed and armed by Iran. The two important Hezbollah groups were at Herat
and at Kandahar. In Kandahar Hezbollah, led by haji Mukhtar Sarwari, played an
active part in the struggle against the Soviets. Sarwari opposed shaikh Mohseni,
who was accused of accepting American aid. The two groups in Kandahar and
Herat maintained relations, but there do not seem to have been mechanisms for
liaison, even though shaikh Ali Wusuqi was described in Kandahar as the leader of
the Afghan Hezbollah.

The formation of the parties

143

gious legitimation enabled the ulema to dominate the political field:

they led all the parties, with the exception of only one of any significance, Hezb-i Islami. The lay parties, which did not satisfy this requirement, were swiftly eliminated (see Chapter 6). In contrast with
the pre-war period, when the range of ideological systemscommunist, Islamist and nationalistwas unrestricted, the parties in exile
were distinctly more homogeneous. By this token the Islamists lost
their monopoly over legitimation on the basis of Islam, even though
their ideology remained distinctive and opposed to that of the traditional ulema and the members of the former governing class.
The control of resources. The ability to control the political arena and
to maintain a monopoly of representation required the control of
external aid, and in the absence of resources of their own the parties
subsisted by appropriating a portion of the aid intended for the mujahidin and the refugees. Relations with host countries were crucial for
the parties. No movement could survive without the assistance of
Iran or Pakistan, and Harakat-i Islami, for example, had alternately
enjoyed support from both.
For the Sunnis the principal host country for refugees and exiled
combatants was Pakistan, whose cooperation was also essential in the
transfer to the mujahidin of the arms and financial support provided
by Saudi Arabia and the United States. Since Pakistan largely controlled the distribution of aid, its attitude was crucial for the survival
of the groups which came into existence in 19789. In December
1979 the Soviet invasion of Afghanistan was a threat to Pakistan, but
it was also useful to the regime of Zia ul-Haq whose military dictatorship had been ostracised by the international community, especially after the hanging of the former Prime Minister Zulfiqar Ali
Bhutto. After some hesitation Pakistan accepted full American assistance, in the shape of $3.2 billion of direct military aid and total aid
of $7.2 billion during the 1980s.10 In addition, Pakistan was given an
IMF credit of $1.6 billion, the largest hitherto granted to a developing country. In exchange for western support, Pakistan became the
10

P. Dikshit, 1993: Afghanistan Policy, Strategic Analysis, Nov., vol. XVI no. 8,
p. 1073. For a historical approach to Pakistani foreign policy, see S. M. Burke,
L. Ziring, Pakistans Foreign Policy: an historical Analysis, Oxford University Press,
1990.

144

The Jihadi Parties

conduit for western aid to the Afghan resistance and a sanctuary for
the anti-Soviet guerrilla movements. The several billion dollars of
aid available for the resistance was therefore in practice distributed
by Pakistan.11
On the other hand, far from acting simply an instrument of the
West, Pakistan made use of its position as an intermediary to control
the parties and develop its own policy. In institutional terms Pakistani supervision was exercised in three ways: political and military
affairs were supervised by the ISI (Inter-Services Intelligence Directorate), international contacts and negotiations were carried on by
the Ministry of Foreign Affairs, and aid to the refugees was administered by the Commissioner for Afghan Refugees. The role of Pakistan was crucial from the first in the formation of the exiled Afghan
political parties which appeared in 19789. The Pakistani government therefore decided to halt the formation of new groups and to
stabilise the number of parties, in the case of the Sunnis, at seven.
Similarly, for the entire duration of the war a Pakistani general took
part in the meetings of the exiled parties, while strategy on the
ground was broadly laid down by the Pakistan military. The Pakistan
administration also played a decisive part in the allocation of aid
to 3.2 million Afghans within its borders, of which a significant part
seems never to have reached the refugees. Finally, Pakistan acted in
the field of diplomacy as the representative of mujahidin parties who
were never invited to participate directly in negotiations.12
Pakistans Afghan policy, from the Soviet invasion up to the events
of 11 September, displayed great stability and did not depend on the
party in power. For this reason support for Hezb-i Islami continued
after the death of Zia ul-Haq and only ceased because of the movements failure on the ground. Similarly, the alternation between
Nawaz Sharif and Benazir Bhutto had no consequences for Afghan
policy, with Benazir continuing to support the Taliban. The reason
for this continuity lay in the influence of the military establishment
and also in a broadly consensual view of Pakistans national interest.
11

The single American donation amounted to around $2 billion and the Arab
countries provided a similar sum. See Charles G. Cogan, Partners in Time: the
CIA and Afghanistan since 1979, World Policy Journal, 10 (2), 1993, pp. 7382;
William Maley, The Afghanistan Wars, London: Palgrave, 2002, pp. 76 ff.
12 D. Cordoves and S. S. Harrison, Out of Afghanistan: the inside story of Soviet withdrawal, Oxford University Press, 1995.

The formation of the parties

145

The strategic plan, often expounded by the Pakistani military, was

to endow Pakistan with strategic depth in relation to India, through
the installation of a pro-Pakistan government in Kabul, and furthermore to create a Muslim region capable of standing up against India
economically, demographically and perhaps even militarily. The balance of power with India and the issue of Kashmir therefore entirely
determined Pakistans Afghan policy. In particular, a pro-Pakistan
government in Afghanistan should be able to block the return of an
alliance of Afghanistan and India against Pakistan.13
In line with such arguments, the Pakistan authorities backed those
Afghan religious movements which appeared to share an a priori
ideological solidarity with Pakistan, which itself had embarked on a
process of Islamisation. In the 1980s Gulbuddin Hekmatyar had proposed a confederation between Pakistan and Afghanistan, aiming to
please his Pakistani protectors. On the other hand, Pakistan had
played a key role in the elimination or marginalisation of the Maoist
or nationalist movements within the Afghan resistance, because they
represented a potential threat to Pakistani domination.
However, the situation in practice was more complex than might
be supposed from what has so far been said. Pakistans Afghan policy
had actually been set in train by officers who held an orientalist
view of Afghanistan. The model which influenced the ISI officers
who were in charge of contacts with the Afghan commanders was
implicitly that of the Pakistani tribal region, where central authority
manipulated tribal divisions on the British divide-and-rule model.
This is probably why the ISI assisted the development of the Khales
faction of the Hezb-i Islami which opposed Hekmatyars Hezb-i
Islami in the early 1980s, even at a moment when it supported the
latter. In general the Pakistani intelligence services systematically
divided the Afghan parties by setting the commanders against each
other. This strategy did not only affect the military: up until 11 September at least, Afghanistan was seen by the Pakistani political class
as an extension of Pakistans frontier zone, and therefore the Ministry of the Interior rather than that of Foreign Affairs took the leading role, even in diplomatic contacts with the Afghan movements.
13

See Ahmad Iqbal, A mirage misnamed strategic depth, Al-Ahram Weekly, Cairo,
no. 392, 27 August2 September 1998. Also Major Abdul Rahman Bilal, Islamic
Military Resurgence, Karachi: Ferozsons, 1991.

146

The Jihadi Parties

In the implementation of its strategy Pakistan made use of a network of Afghan clients, of which Hezb-i Islami, followed by the
Taliban, were the most important. As has been observed, the relationship between the Islamists and the ISI went back to the 1970s,
especially in the organisation of the coup of 1975. Hekmatyar, the
future leader of Hezb-i Islami, liaised at that time with the Pakistan
intelligence services and had made himself the privileged ally of
Islamabad. Subsequently Pakistan distributed a large proportion of
the aid to Hezb-i Islami: probably around 40%. Hezb-i Islami also
took advantage of its good relations with the Pakistan administration
to establish itself in the refugee camps and to assassinate its political
adversaries, particularly the leftists, nationalists and royalists.14
With the Shiites Iran played a role symmetrical to that of Pakistan, although it did not exercise such close control over the establishment of parties since some were based in Afghanistan and therefore
less dependent on external aid. Iranian policy went through a number of stages. Its complexity and lack of consistency at certain moments arose from there being multiple centres of decision-making
which were sometimes violently opposed to each other. Irans policy
of fragmentation through the creation of parties was probably not so
much a Machiavellian encouragement of discord, even though in
theory Iranian control would thus have been enhanced, but rather
the result of internal struggles within Iranian politico-religious circles.
In Afghanistan, Iran sought above all to preserve its relationship
with the Shiites, while at the same time not providing them with
any significant military assistance in order not to damage its own relations with the Soviet Union.15 The Persian-speaking Sunnis kept
their distance from Iran, which offered them no significant help. The
reality was that the regime in Iran prioritised its struggle against Iraq,
while offering largely verbal support to the mujahidin, although it did
provide some logistical facilities on its territory. Consequently its relations with the parties which were actually fighting the Soviets
were difficult, even when these were Shiites, as with the Harakat-i
14

In the 1980s Hezb-i Islami appeared also to maintain private prisons in Pakistan
where it was able to detain its opponents.
15 Well before the Islamic revolution, the Afghan Shiites felt themselves to be close
to Iran. Portraits of the Shah were to be seen in some Shiite houses, and Hazara
migrants travelled for preference to Iran rather than to Pakistan.

The formation of the parties

147

Islami. On the other hand the Iranians did assist Shiite movements
which took their inspiration from the Iranian revolutionary model,
such as Nasr and Sepah.
The affiliation of the commanders. The parties monopoly over the representation of the mujahidin was yet further extended in Afghanistan
through the recruitment of the commanders, hence the fundamentally asymmetrical relationship between these two types of actors.16
The parties depended on the bellicosity and organisational capacity
of their commanders to attract further foreign donors and enhance
their influence, although media manipulation might in the short
term affect how they were perceived and therefore influence the
flow of aid. On the other hand, what the commanders were able to
raise from the population was used locally, with nothing returning to
the centre of the organisation.
When a commander joined a party, it was because he thereby derived immediate and concrete benefits. In practice commanders
were only able to maintain their influence if they succeeded in obtaining arms and money. Here the parallel drawn by Mike Barry between the activities of the khans and those of leaders of parties is
illuminating.17 The primary activity of the parties was the distribution of arms, while most movements also ran military training courses: for example, teaching the use of anti-aircraft weapons. The parties
also gave support to the commanders or to their mujahidin when
they spent periods in Pakistan or Iran.
At their origin the relationships between the commanders and
the parties did not depend principally on ideological identification,
but on access to resources provided from abroad. However, this statement must be qualified, especially for those commanders who had a
political affiliation which pre-dated the war. Non-material issues
generally played a considerable part in the relationship between a
commander and a party. Party affiliation provided protection, and also
legitimacy in relation to other groups. In the event a commander
not attached to a party could be suspectwas he, for instance, a
16

The leaders of the main Hazara Shiite parties were not in exile, but remained in
Hazarajat. Only Mohseni, who was not a Hazara and recruited principally in the
towns, lived in Pakistan. There did not exist therefore the same kind of relationship between the parties and the commanders as for the Sunnis.
17 Mike Barry, op. cit., p. 15.

148

The Jihadi Parties

Maoist?and other local groups might take this as a pretext to bring

him down, especially since they did not fear reprisals. Similarly, affiliation to a party almost inevitably brought with it benefits of solidarity, especially while travelling.
The relationship between a party and a commander also depended on the commanders importance. Many petty commanders lingered for months in Peshawar, waiting for a party to provide them
with arms and thus enable them to go home.18 However, from the
moment when a commander became important enough to be able
to pay for a representative in Pakistan, his autonomy was enhanced,
since he would henceforth be able to make his own contacts with
the Pakistan intelligence services and with western embassies. For
instance, Abdul Haq, Masud and Amin Wardak had the benefit of
direct aid from France or from Britain. Such direct links weakened
the parties, who saw the exclusivity of their role as intermediaries
threatened.
While membership of a party was the general rule, this did not
prevent a commander from having a large measure of autonomy in
his choice of which particular organisation to join. This would be, in
the majority of cases, a choice for the commander and not the mujahidin or the population. Affiliation was undertaken at the initiative of
the commander, as the leader of an already existing group which
would make contact with a party in order to obtain weapons.19 In
addition, competition between the parties meant that they were
obliged constantly to strive to maintain their existing clientele, which
hampered the establishment of party discipline. In the event this
competition augmented the freedom of action of the commanders
who followed the patrimonial model, and changed their allegiance
without difficulty from one party to another, although the commanders of the institutional type maintained more stable affiliations.
During changes of party the mujahidin remained faithful to their
commander rather than to the party. Various reasons might lie behind
18

Conflict could erupt between the leadership and the commanders if their demands ceased to be met: in some cases violent incidents took place in Peshawar
itself.
19 However, the commanders room for manoeuvre was not unlimited, and an unpopular party affiliation could diminish his power. In addition early affiliation,
often crucial, tended to be made when the mujahidin, and therefore the commanders, were not well differentiated from the population.

A typology of the parties

149

the commanders changes of party, though the prospect of obtaining

more arms, and therefore of consolidating their power, was often
enough. Another common situation was the existence of a conflict
between commanders of the same party. For instance, in Badakhshan Jamiyat-i Islami was largely dominant at the beginning of the
war, except in Keshem which was under the control of Hezb-i
Islami. However, internal conflicts within Jamiyat-i Islami in the
Argu region prompted the expulsion of the commander Jamaluddin
and his replacement by mawlawi Kheyradmand of the Hezb-i Islami.
As the war continued, the parties never succeeded in leaving behind this disunity between the leaderships and the commanders.
With only rare exceptions, occurring mainly when the leadership
was situated in Afghanistan, the parties would continue to be coalitions of commanders.

A typology of the parties

The classification of the Afghan parties generally adopted has proved
inadequate. The customary distinctions drawn between moderate
Islamists, radical Islamists, moderate fundamentalists, traditionalists and so forth are confused and hard to justify, since they depend
on non-empirical criteria which arise from the subjective involvement of an observer, often a westerner, with the movements. In
addition the relevant areas for analysisideology, organisation and
recruitmenthave not been distinguished.
Three models of party may be hypothesised, derived from the
types of commander: the Islamist, the clerical and the patrimonial.
This typology is based on a statistical correspondence between social
position, ideology and method of political action. The ideology of
the educated class tended to be Islamist, while that of the pre-war
lite was conservative, and the ulema of the private madrasas were often
fundamentalists. Similarly, a bureaucratic style of organisation was
linked to the involvement of the educated group or of the ulema,
while the pre-war lites adopted a more informal structure, based on
patronage. Finally, the recruitment of commanders was socially close
to that of the leadership: ulema-commanders were found in clerical
parties, educated commanders in Islamist parties and khan-commanders in patrimonial parties.
The characteristics of an Islamist party were therefore an Islamist
ideology and leadership, of lay or religious origin; with a bureau-

150

The Jihadi Parties

cratic organisational style and recruitment from the educated class.

A clerical party was characterised by a fundamentalist ideology, combined with a clerical bureaucracy, and a leadership consisting of
ulema-commanders. Finally a patrimonial party brought together a
conservative or reactionary ideology, a leadership drawn from the
pre-war lites, a patrimonial organisation, and recruitment from the
khans. These categories applied fairly well to the majority of the parties in terms of their leadership, organisation and ideology, but recruitment was by its nature more complex since, as is shown later,
they involved local considerations.
This typology also enables the identification of mismatches and
irregularities, since the links between the various phenomenaideology, leadership, organisation and the recruitment of commanders
were not in practice absolutely consistent. In practice there might be
a contradiction between recruitment and leadership, or between recruitment and avowed ideology. In all such cases local issues and incidental motivations should be examined for the understanding of
particular situations. The personal histories of the leaders often provided some explanation. In addition, a party might evolve moving
closer to an alternative ideal type. Ideological frontiers between Islamists and fundamentalists had a tendency to become blurred, which
was not unique to Afghanistan. However, recruitment would generally follow a set pattern.
To return to the selected categories, Hezb-i Islami, Jamiyat-i Islami
and, among the Shiites, the Mustazaffin corresponded closely to the
category of Islamist parties, although the situation of Jamiyat-i Islami
was more complex owing to the mixture of ulema-Islamists and lay
Islamists in its leadership. Jebhe-yi Nejat-i Melli and Mahaz-i Melli
were instances of the patrimonial model. Hezb-i Islami (Khales) was
similar to this model in its organisation and recruitment, but its
leader, an alem who did not belong to the pre-war ruling class, expressed himself in fundamentalist terms. Harakat-i Enqelab was the
sole example of a Sunni clerical party, to which may be added the
majority of the Shiite parties, especially Nasr and Sepah. Harakat-i
Islami, headed by an alem but with a membership of notables and
an Islamist ideology, was hard to fit into a category, as was also the
Ettehad, with an Islamist leader, a fundamentalist doctrine and a
frankly opportunist recruitment policy. The proposed classification

A typology of the parties

151

will be more satisfactorily demonstrated by way of a detailed examination of four issues: leadership, organisation, ideology and recruitment, always distinguishing between Sunnis and Shiites.
THE SUNNI PARTIES
Name of party
Hezb-i Islami

Leadership
Orgnisation
Bureaucratic
Islamist
(Gulbuddin
Hekmatyar)
Jamiyat-i Islami Islamist, clerical Bureaucratic
(mawlani Rabbani)
Harakat-i
Clerical
Clerical
Enqelab
(mawlani Nabi)
Jebhe-yi Nejat Patrimonial
Patrimonial
(pir Mujaddidi)
Mahaz-i Melli Patrimonial
Patrimonial
(pir Gaylani)
Hezb-i Islami Patrimonial
Patrimonial
Ettehad
Clerical
Patrimonial
(mawlani Sayyaf)

Ideology
Islamist

Recruitment
Educated
class

Islamist
Fundamentalist

Educated
class
ulema

Conservative

khan

Conservative

khan

Fundamentalist
Fundamentalist

khan
Opportunist

THE PRINCIPAL SHIITE PARTIES

Name of party
Shura

Leadership
Clerical
(sayyed Beheshti)
Nasr
Clerical
(shaikh Mazari)
Sepah
Clerical
(skaikh Akbari)
Harakat-i Islami Clerical
(shaikh Mohseni)
Mustazaffin
Islamist
(engineer Hashemi)

Organisation Ideology
Clerical
Conservative
Clerical

Islamist

Clerical

Islamist

Clerical

Conservative

Bureaucratic Islamist

Recruitment
Hazara,
sadat
Hazara,
ulema
Hazara,
ulema
Shiite khan
Educated
class

The leaderships. With one exception onlyGulbuddin Hekmatyar

the leaders were ulema, but this category was not homogeneous. Some,
such as Burhanuddin Rabbani, pir Sayyed Ahmad Gaylani,20 Abdul
20

Born in 1932, Gaylani was educated at the Abu Hanifa College in Kabul, then at
the Faculty of Theology at the University of Kabul where he took his degree in
1960. The Gaylani family soon linked itself to the royal clan through marriage.

152

The Jihadi Parties

Rasul Sayyaf,21 and Sebghatullah Mujaddidi,22 had emerged from the

government madrasas, while others, such as mawlawi Yunus Khales23
and Muhammad Nabi Muhammadi, came out of the private madrasas.
In 1952 Sayyed Ahmad married Adela, the granddaughter of amir Habibullah, at
a moment when the wife of King Amanullah was also of Gaylani descent. Before the war the pir was mainly occupied in secular pursuits: among other activities he was the Peugeot concessionaire, and his activities as a brotherhood leader
were modest. He placed the preservation of his network of clients above his
pursuit of spiritual practices.
21 Sayyaf was born in 1946 at Paghman, near Kabul. He was a pupil at the Abu
Hanifa theological school where he graduated in 1963, then at the Islamic College at Kabul University where in 1967 he took his degree with distinction, then
becoming a teacher at the Faculty of Shariat where he taught hadith. In 1969 he
went to Al-Azhar University, and graduated with distinction. He returned to
Afghanistan in 1972, where he taught and helped to publish Shariat. His involvement with the Islamist movement was already notorious, which resulted in his
arrest in 1974 as he boarded a plane to the United States to undertake legal studies. He was freed by Babrak Karmal in 1980 and went to Pakistan, where he
achieved election as president of the Ettehad-i Islami (Islamic Alliance).
22 Born in Kabul in 1925, Sebghatullah Mujaddidi was the son of Muhammad
Masum Mian Jan Mujaddidi. He was educated at the Habibia School, and afterward spent six years at the Faculty of Law and Jurisprudence at Al-Azhar University, where he graduated in 1953. Until 1959 he taught at various schools, as
well as teaching Quranic studies and the shariat at the Teachers College in
Kabul. He had links with the Egyptian Muslim Brotherhood, and enjoyed good
relations with Rabbani, with whom he had in common membership of the
Naqshbandi tariqat. In 1959 he was accused of participation in a conspiracy
against Khrushchev during the latters visit to Afghanistan. He was imprisoned
until 1964, then lived in the United States from 1968 to 1970. He founded the
Jamiyat-i Islami-yi Ulema-yi Mohammadi in 1972. Once more exiled from
1974 onwards, he headed the Islamic Centre in Copenhagen, returning to
Peshawar in 1978, where he set up the Jebhe.
23 Born into a poor family around 1919 at Nuqur Khel, Khales was a Pushtun of
the Khagiani tribe, and studied in private madrasas before going to complete his
training at the Haqqaniyyah madrasa in Pakistan. In the 1960s he worked for
Kabul Radio, where he presented religious programmes. He seems to have enjoyed a certain popularity, owing to the direct and often humorous style of his
contributions, and he also collaborated on a monthly magazine, Gahis. Around
1960 Khales published a translation of the book by Sayyed Qutb, Islam wa Adalat
Ijtimai (Islam and Social Justice). His involvement with the radio station and the
magazine ended with the rise to power in 1973 of Daud, who swiftly showed his
antagonism to the Islamists: under Daud one of his sons was killed. Khaless
membership of the Islamist movement appears to date from this time. In 1975 he
went into exile in Pakistan, taking up residence in Islamabad. From 1975 to 1978
he worked within the Hezb-i Islami.

A typology of the parties

153

The life histories of the Islamists, who were often exiled during the
1970s because of their opposition to the authorities, were plainly different from the members of the governing class such as Muhammad
Nabi, a former member of parliament, or Gaylani, a former member
of the Loya Jirga. The Islamist leaders, who had been relatively unknown before the war, had a network of militants at their disposal,
while the other leaders were well known and had support among
the ruling class and, in particular in the case of Gaylani, within the
Sufi brotherhoods. Some leaders held several positions, others did
not. In contrast to Hekmatyar, who belonged only to one category
that of IslamistGaylani belonged to three: he was a member of the
ruling class, a pir and an alem. Rabbani was an Islamist and an alem.
Ethnically Rabbani was a Tajik, Gaylani and Mujaddidi were sadat,
and the others were Pushtuns.
Among the Shiites none of the important leaders belonged to
the ruling class, and education was not a point of difference, since all
were ulema. Their community affiliation was a more influential factor. Three types of leader mobilised different networks: the alemsayyed such as Beheshti; the Hazara alem such as Mazari,24 and Akbari;
and the non-Hazara alem such as Mohseni. The engineer Hashemi,
the leader of the Mustazaffin, was the only lay Islamist among the
Shiites, occupying a position comparable to that of Hekmatyar
among the Sunnis, but leading only a group of very small size.
Models of organisation. The party structures in general duplicated the
administrative pattern, with titles such as amir-i welayati (provincial
governor) and so on. Titles were derived from Islamic terminology,
such as amir, and in general distinguished between military officials
(amir-i nezami) and political ones (amir-i siasi). Beyond these common characteristics, the parties might be organised on patrimonial,
bureaucratic or clerical lines.
The bureaucratic style of organisation necessitated a body of rules
concerning activities, such as meetings and the assignment of tasks,
on an administrative model. Recruitment of students and former
government employees accounted for the propensity to bureaucratic
24

Born in 1946, Mazari came from Nanway, a village near Charkent in the province of Mazar-i Sharif. After studying in Qom and Najaf he returned to Afghanistan in 1978.

154

The Jihadi Parties

organisation. The Hezb-i Islami was the closest to this model, while
Jamiyat-i Islami, which also recruited among the educated class, did
not have the means to maintain an extensive bureaucracy. Hezb-i
Islamis organisation was a blend of centralisation and military discipline on the model of the Pakistani Jamaat-i Islami.25 Its obsession
with rules contrasted with flexibility observed elsewhere. Hezb-i
Islami set up an alternative society as well as an alternative state, and
demanded that its members should end their communal affiliations.
The party was intended to take precedence over family or tribal loyalties, a policy which came up against various kinds of resistance.
The organisations outstanding characteristic was centralisation.
Hekmatyar was in a position to make crucial decisions more or less
on his own, though this seems to have been modified after successive
setbacks and the marginalisation of the movement. While in most
parties the commanders were allowed wide autonomy on the ground,
the leadership of Hezb-i Islami retained as far as possible strict control over local initiatives. In the 1980s the major commanders had in
principle been supposed to stay in daily touch by radio. Internal
purges enhanced respect for the discipline of the party, which had the
use of prisons in Pakistan where several hundred members of other
parties, as well as communists, were detained, tortured and executed.
Among the Shiites the Mustazaffin was the best example of a bureaucratic organisation. In total this group amounted to only several
hundred men, whose time was divided between military and civilian
activities. A sharp distinction was drawn between the officials, who
were all members of the educated class, and the ordinary members.
The officials were drawn from the original pre-war nucleus, and
though the engineer Hashemi was the spokesman and leader of the
group, the leadership was collective. The overall level of organisation
was the best found anywhere in Afghanistan.
The clerical parties adopted a type of organisation in which the
personnel were religious figures. Among the Sunnis this model was
found in the Harakat-i Enqelab, where the leadership consisted in
25

See Marc Gaborieau, Le no-fondamentalisme au Pakistan. Maududi et la

Jamaat-i-islami in Olivier Carr (ed.), Radicalismes islamiques, vol. 2, Paris: LHarmattan, 1986, pp. 3376; also Kalim Bahadur, The Jamaat-i-islami of Pakistan,
Lahore: Progressive Books, 1978.

A typology of the parties

155

principle of ulema and their taliban, and was the dominant one among
the Shiite parties in Hazarajat, especially during the Shura of the
early years, which replicated the pattern of administration of the
state, though with the posts of responsibility occupied entirely by
ulema.26 The system set up by the ulema of the Nasr and the Sepah
was little different from that of the Shura, but was less complex.
Mahaz, Jebhe and Hezb-i Islami (Khales) functioned according to
a patrimonial model where no distinction was made between the resources of the leader and those of the organisation. For example, the
trucks belonging to Hezb-i Islami (Khales) were the property of the
leader, who was thus in a position to embark on commercial ventures
in his private capacity. In these parties the commanders employed
complex strategies to increase their resources and their clientele. The
commanders often directly provided their own finance, and therefore enjoyed wide autonomy. Within Hezb-i Islami (Khales) a number of competing networks could be distinguished.27 The Pakistani
intelligence services controlled the activities of all of them, supporting the various commanders according to their priorities of the moment. In Mahaz the group which surrounded pir Sayyed Ahmad
Gaylani consisted largely of members of the former governing class,
as well as the pirs family members. This was the only party to have
been family-based to this extent, with the possible addition of that of
26

Hazarajat, and particularly Yakaolang, was incidentally the only place where the
population was systematically disarmed. Hazarajat was divided into nine wilayat
(provinces). Small settlements had a shahrwal (mayor). The administration of the
Shura, which took over the government buildings, was top-heavy and inefficient
to a degree. The representatives of the thirty-four liberated uluswali were represented by two or three delegates making up a permanent shura which met once
or twice a week. There were committees for war, economy, culture and law.
Taxes were high, reaching 10% of incomes, much more than among the Sunnis.
Militarily the Shura ran four fronts, in Bamyan, Behsud, Naur and Jaghori, coordinated by Sayyed Hassan jaglan. There was conscription, requiring one year of
military service at the age of twenty-two, with the possibility of sending a paid
replacement.
27 On the one hand, haji Din Muhammad and his brothers haji Qadir, Abdul Haq
and Daud controlled an important part of the partys resources. On the other
hand mawlawi Khales maintained good relations with Engineer Kabir, his
brother, and Engineer Mahmud, which counterbalanced the influence of the
other network. Mawlawi Haqqani and Amin Wardak were virtually independent
powers, each closely linked to the Pakistani intelligence services.

156

The Jihadi Parties

Mujaddidi; in both these cases nepotism reached a level which obstructed their efficiency.28
Ideology. Scripture is its own interpretation. (Martin Luther)29
Three ideological positions define the range of the Sunni political
field. These are Islamist, fundamentalist and reactionary-conservative,
each of which relies in different ways on religion for its validation.
The Islamists, in Hezb-i Islami and Jamiyat-i Islami, had a distinct
predilection for ideological issues. This concernsystematically to
validate the actions of the party on the basis of abstract principles
was a point of distinction from the practice of other movements.
Hezb-i Islami therefore attached great importance to the training of
its members and to the spread of its ideology.30 Its programmes and
those of Jamiyat-i Islami were important texts since they enumerated
in detail the parties principles on religion and education, among other
issues. The Islamic revolution, with its founding myth of the original
community of the Prophet, played a key role in political doctrine
and in the self-conceptualisation of these parties. The principal accent
28

Sayyed Hasan Gaylani, Fatima Gaylani, Naser Zia, and Suleiman Gaylani were
among the most active of the Gaylani family within the Mahaz apparatus. Among
the former members of the senior administration were General Salam, General
Katawasi, Asef Muhammad Ikram and Dr Gholam Faruq Azzam. The leadership
of Jebhe was drawn from lite figures from the former regime: for example Muhammad Gulab Nangrahari, Shahid Zemaray the secretary of Mujaddidi, and
family members such as Abdul Shakur Turyalay Osman. The son of S. Mujaddidi, Zabihullah, was the treasurer of the party, while his brother played a part in
the early years before taking up residence in California. The details of these networks may be followed in Ludwig W. Adamecs A Biographical Dictionary of Afghanistan, Graz: Akademische Druck- und Verlaganstalt, 1987.
29 Oeuvres, vol. I p. xxiv, Paris: Gallimard, 1999.
30 For analysis of the programmes, see David B. Edwards, Summoning Muslims:
Print, Ideology and Religious Ideology in Afghanistan, Journal of Asia Studies 52,
no. 3 (August 1993), pp. 60928. Hezb-i Islami propagated its ideology essentially
through the partys publications in Peshawar. The daily Shahadat and the monthly
magazine Mujahidin Monthly were among the partys most widely distributed
periodicals. In addition, meetings were regularly organised in the refugee camps
to allow the party leaders, and especially Hekmatyar, to put their points of view, a
thing the other movements did less often and on a less organised basis. Hezb-i
Islami attempted to take control of the refugees, for example through setting up
schools for orphan children in the camps. Great attention was typically paid to
education in the desire not to leave it to the mullahs and to teach non-religious
subjects.

A typology of the parties

157

was placed not on re-Islamisation in day-to-day terms but rather on a

radical and violent transformation of political society, and putting
social relations on a new footing. The Islamist project did not give a
special place to the ulema, but legitimised institutions on the basis of
universal suffragemasculine and femininewhile nonetheless rejecting western democracy and especially the concept of multiple
political parties.
Rationality, science and modernity were also essential elements of
the ideology of the Afghan Islamist parties. Hezb-i Islamis model of
modernisation, essentially urban and industrial, favoured investment
in heavy industry and the nationalisation of major companiesas
too did that of Hezb-i Demokratik-i Khalq-i Afghanistanwhile
explicitly excepting small companies. On the other hand, the peasantry did not loom large in the partys thinking. Hezb-i Islami was
the inheritor of the Muslim Brotherhoods ideology, in practice closely allied to dirigiste planning, whereby development could be programmed in foreordained steps. The Islamists were influenced by the
Third World left in more than one aspect of their programme, all the
more since anti-imperialism was a concomitant of the Islamic revolution. Verbal antagonism towards the United States, whatever the
complexities of the actual relationship, was a constant in the language of Hezb-i Islami and, to a lesser degree, of Jamiyat-i Islami.
Hezb-i Islami took a systematically hostile attitude towards westerners, and incidentssome fatalwere frequent.
However, there were differences between the two movements. The
history of the Islamist movement was re-written because of Hekmatyar, who was the object of a real personality cult organised by the
party functionaries, while Rabbani set himself up as an arbiter between the various commanders of the Jamiyat-i Islamic. In addition
the programme of Jamiyat-i Islami was an appeal to all Afghans, and
laid less stress than Hezb-i Islami on the party leader. Although both
movements referred to Islamist authors such as Maududi and Sayyed
Qutb as the source of their doctrine, Jamiyat-i Islami was inspired by
the dominant tendency of the Egyptian Muslim Brotherhood, while
Hezb-i Islami identified itself rather with the Pakistani Jamaat-i
Islami, which was very hierarchical.31 Even though Jamiyat-i Islami
31

Maududi was, however, more conservative. For example, he was radically opposed to all agrarian reform: see Seyyed Vali Reza Nasr, Maudidi and the Making

158

The Jihadi Parties

aimed at the liberation of Soviet Central Asia,32 it put first the patriotic aspect of the struggle against the occupier. Hezb-i Islami rejected
all nationalist ideology and proposed a confederation with Pakistan.
However, this suggestion was made with the aim of enhancing relations with Islamabad, and thus in reality primarily tactical.
Hezb-i Islami and Jamiyat-i Islami also took opposing positions
on the issue of takfir: could a Muslim be declared an apostate by virtue of his recognition of a non-Islamic state, when he continued to
perform his religious duties, prayer, fasting and so on? Sayyed Qutb,33
an important ideologue for the Egyptian Muslim Brotherhood, reintroduced this idea, which was found in the writings of Ibn Taymiyya (12631328). Hezb-i Islami practised individual takfir, that is
to say it took upon itself the right to declare a practising Muslim an
apostate, which could imply the death penalty, but stopped short of
declaring the whole of Afghan society to be irreligious, as certain
groups following Sayyed Qutb had done in the case of Egypt. With
this exception Hezb-i Islami, in denying that law had any autonomous existence, operated according to political imperatives, untrammelled by moral or juridical perspectives. Political assassination,
justified by takfir, was a vital part of the political culture of Hezb-i
Islami. This was a divergence from the juridical viewpoint of most of
the parties, who turned in cases of difficulty to an interpretation in
the spirit of jurisprudence, although the ulema were not always the
of Islamic Revivalism, Oxford University Press, 1996, p. 74. The closeness of
Hezb-i Islami to Maududi may have consisted at a more profound level in its interpretation of texts outside the religious institutions; though Maududi, in contrast to Hekmatyar, had a classical religious education and came from a line of
pirs of the Chishti Sufi brotherhood.
32 As witness Ismail Khan, who declared that he wished to die a martyr at Bukhara
(interview: Herat, autumn 1988). Jamiyat-i Islamis maps showed Soviet Central
Asia as an occupied portion of the umma.
33 On the thought of Sayyed Qutb, who was largely the Islamists inspiration, see
Olivier Carr, Mystique et politique. Une lecture rvolutionnaire du Coran par Sayyid
Qutb, frre musulman radical, Paris: Presses de la fondation nationale des sciences
politiques, 1984. On takfir, see p. 15 ff. For Qutb, takfir was a generalisable concept, implying a break with society rather than individual excommunication.
On this point Sayyed Qutbs view was atypical within the Egyptian Muslim
Brotherhood and only a few groups of extremists, such as Tahrir, Takfir and
Jihad, made takfir the basis of their strategy, in the process carrying Qutbs view
to an extreme. In particular, this was the justification for the assassination of President Sadat of Egypt.

A typology of the parties

159

judges of last resort. For Hezb-i Islami nothing could be just in itself,
and truth had no value outside the party, which retained a de facto
monopoly over the interpretation of the Quran. Jamiyat-i Islami, on
the contrary, employed a more limited concept of takfir, possibly
explained by the presence of numbers of ulema within the party.
While the Islamist movements were linked to the phenomenon
of mass education, it has been observed that the fundamentalist
tendencies, often established by pirs or ulema, originated with the
19th-century reformists of the Indian sub-continent. In contrast to
the Islamists, the fundamentalists preached a return to good Islamic
customs, in relation for example to the status of women, to religious
practice and to artistic activities. However, this dichotomy should
not be stretched too far, since there was also an authentic fundamentalist political project, although it implied the rejection of the modern forms of the state and even of politics as an independent sphere.
Within the fundamentalist movement various tendencies were represented in the Afghan parties, whether inspired by Salafism or more
in continuity with the pre-war tradition.
The war facilitated a breakthrough on the part of fundamentalist
movements whose influence had previously been limited. The ideology of the Ettehad and of certain Hezb-i Islami (Khales) commanders was inspired by such fundamentalist movements. In particular,
these parties rejected parliamentary democracy and elections, to
which they preferred an Islamic system, in which the Ahl-i hal wa
akd (pious Muslims, persons respected within the community) played
a determining role in the legislative system. The frequent denunciations by Khales and Sayyaf of the Shiites, to whom the former even
denied the right to vote, reflected in the case of Khales the traditional tensions between the Shiite and Sunni Pushtun tribes of the
frontier, but it also arose from the scale of Saudi financial support.
However, it seemed that neither the commanders of the Ettehad nor
Sayyaf himself became Salafists, since they continued to observe the
Hanafi Sunni rites. Finally, at the beginning of the war, Harakat-i
Enqelab was closer to Mahaz or to Jebhe. Nevertheless, a growing
strictness on moral issues indicated a rapprochement to the tendencies mentioned above.
The reactionary or conservative parties, Mahaz and Jebhe, were
nostalgic for the old order and the domination of the pre-war lites.
Mahaz had no detailed programme, but its ideology may be deduced

160

The Jihadi Parties

from the publications of the WUFA (Writers Union of Free Afghanistan). Mahaz was a royalist party, which arose from the personal links of Gaylani with the royal family and the pre-war
establishment. Mahaz demanded a constitutional system inspired by
the regime of the West, with separation of powers and a parliament.
It favoured the market economy and opposed the social measures
proposed by the Islamist parties. Jebhe took a similar line, although
it appealed more directly to Islam as the mode of social organisation
and did not insist on support for the king as an absolute principle.
Among the Shiites the distinction between Islamism and fundamentalism was less relevant. The ideological structure was polarised
by attitudes to Khomeini, and placed conservatives in opposition to
revolutionaries. In the revolutionary parties such as Nasr and Sepah
the ulema insisted unanimously on the predominance of the religious leaders in the political field. Their ideology was nevertheless
put into practise in various ways by different leaders in different
places. Nasr was inclined generally to be populist and puritan, but
some officials took a less sectarian and less strictly clerical line. The
demand for social justice was also expressed by many of its leaders,
and attempts were made to distribute land to the peasants. In addition Hazara nationalism, though seldom appealed to as such, was an
element of the ideology of Nasr, Sepah and Nehzat.
By contrast, the Harakat-i Islami of Mohseni, a disciple of ayatollah Khoei, did not seek to set up the Iranian revolution as a model. A
factor was the opposition between Khomeini and Khoei, who had
always rejected the involvement of the ulema in politics, and in addition had failed to return to Iran after the revolution. Although shaikh
Mohseni preached that the social order should conform to Islamist
values, he did not proclaim a clerical model of society and remained
socially a conservative. The Mustazaffin were once more the exception. The principal sources of the partys ideology were Iranian intellectuals, especially Ali Shariati,as well as,more tangentially,Muhammad
Iqbal, a Pakistani Sunni, and Ismail Balkhi. The view taken of Khomeini was somewhat critical, since the party was opposed to the
domination of the ulema in politics. Utopianism was the most unusual
feature of this system of ideas, which sought to recreate a perfect
community modelled on that of the Prophet.34
34

The parallel with the Iranian Mujahidin-i Khalq was striking, and the two parties in fact had contacts. This ideology was organised around two central ideas:

A typology of the parties

161

Distribution on the ground. Since the beginning of the war most

Afghans identified with a party and many held membership cards. It
was not unusual to encounter an individual with two or three cards
from different parties. Multiple adherences in general included a main
membership, genuinely entered into, along with others acquired for
convenience, often to facilitate travel. In the towns membership was
more often to a single party. In practice an individuals choice of a
party depended on the opportunity to develop political awareness in
the context of environments other than the immediate solidarity
group of the extended family, i.e. in school, at the university or in
professional life. Among townsmen who had been through the educational system, the various members of a single family might have
experienced different social contexts, and therefore may have had access to a variety of choices, while in rural areas it was customary to follow family or clan affiliations. However, whether membership was on
an individual or a collective basis, it was only with a commander as
intermediary that the Afghan population joined a party. Only the distribution of the party memberships of the commanders was significant.
The political map of Afghanistan was complex, and one cannot
necessarily discover the individual reasons for the affiliation of each
commander. Of more significance are the imperatives which explain
them in more general terms. Two types of affiliation can be distinguished, one based on the imperative of proximity and the other on
the imperative of differentiation. The first type results from the search
for a leader who is close to a commander in the context of a qowm,
and the second type from the need for differentiation at a local level,
which obliged the commanders to distinguish themselves from each
other by their choice of party.35
first, a unity between Shiites and Sunnis, who ought to overcome their differences, and, second, social justice, to be achieved by way of the collective ownership of the means of production, or of a part of them. The influence of
Marxism, or at least of socialist ideas, was palpable. The very name Mustazaffin
(the poor) was significant. In addition, their positions on the status of women
were very liberal in the Afghan context. Uniquely, as far as is known, the
Mustazaffin gave women military training.
35 In addition some affiliations were purely practical. The Ettehad, in particular,
was set up on the basis of generous contributions from the Gulf, hence the initial
reluctance of observers to regard it as a real party. Opportunist reasons for membership were in this case the rule rather than the exception.

162

The Jihadi Parties

The imperative of proximity. Affiliations were explicable in part on the

basis of social proximity: commanders adhered to the party whose
leadership was socially compatible in terms of geography, ethnicity
or social status. In the end the degree to which a party was able to
extend itself on the ground would depend largely on the capacity of
the leader to bring various networks under his control. In fact, the
real criterion for the choice made by a commander was not in any
direct sense the partys ideological stance, but rather the personal
characteristics of the leader, in communal, geographical, social and
religious terms. The leader of the party was crucial, since the desire
for social proximity, on the model of the qowm, was operative in the
choice of membership. Party leaders brought into play their various
qualities to attract various networks: ulema, Islamists, notables, members of the ruling class, murids. There follows an examination of
communal proximities, shared membership of networks, whether
political or linked to the Sufi brotherhoods, and finally shared membership of a social category.
In all the parties there was an ethnicity which was dominant at
the level of leadership and, though in general less obviously, among
the commanders. For example two-thirds of the leadership of Hezbi Islami was made up of non-Durrani Pushtuns, clearly differentiating it from Jamiyat-i Islami, which was three-quarters Tajik.36 Mawlawi Rabbani was the only party chief who was a Persian-speaker, a
Tajik and a Sunni. Mujaddidi and Gaylani were Persian-speakers
but also sadat. This exclusivity over the representation of the nonPushtuns certainly played a part in the expansion of the Jamiyat-i
Islami. Similarly the leadership of Hezb-i Islami (Khales) was eastern
Pushtun, while that of Harakat-i Enqelab was in essence Ghilzai
Pushtun, while that of Mahaz was Durrani Pushtun. Among the
Shiites the leadership and the commanders of the Harakat-i Islami
were mainly non-Hazara Shiites: Qizilbash and sadat.37 By contrast,
Nasr and Sepah were led entirely by Hazaras.
36
37

Barnett Rubin, op. cit., p. 87.

The most important areas for Harakat-i Islami were: Kandahar (commander Ali
Anwar), Ghazni (Dr Shah Jan, originally from Kakrak), the neighbourhood of
Kabul (commander Anwari), Charkent (commander Din Muhammad), Bamyan
(commanders Sayyed Adi and Mobarez), and the Unay pass (commander Rezavi).
In Herat, Harakat-i Islami benefited in 1989 from the membership of commander Azimi, hitherto the responsible military official in Rad. There was there-

A typology of the parties

163

Affiliations therefore could be established as the result of a common macro-ethnic attachment, but also through local proximity.
Thus the areas of effectiveness of a party were often those where the
leader, in one way or another, had roots. Rabbani, originally from
Badakhshan, Khales from Nangrahar, and Sayyaf from Paghman
were influential in the provinces where they were born. The geographical extent of Harakat-i Enqelab, also resulted in a strong and
durable presence among the Ghilzai and in the province of Ghazni.
Muhammad Nabi was himself born into a Ghilzai tribe, the Astonekzai. In addition Ibrahim Mujaddidi had set up the Nur al-Modares
school in Ghazni, and, since the recruitment of taliban was mainly
local, his pupils were principally Ghilzai. The building of this madrasa
reinforced what was already a long-standing link between the
Mujaddidi and the Ghilzai tribes. Historically there had therefore
been a Mujaddidi presence among the Ghilzai, which Muhammad
Nabi inherited as leader of the Harakat-i Enqelab.
In this Hezb-i Islami was an exception; it was one of those rare
cases where the leader of a party had no local or tribal roots. The regions of Hezb-i Islamis strongest influence, for example the neighbourhood of Kabul, were the result of a party structure in which
town-dwellers predominated. In fact the Pushtuns of Imam Saheb,
Hekmatyars birthplace, were affiliated not to Hezb-i Islami but to
Gaylanis Mahaz. Hekmatyars membership of the Kharuti tribe, part
of the Ghilzai confederation, led to no sizeable recruitment, since
Hekmatyar suffered from two disadvantages: first, he was born in the
province of Kunduz, far from the territorial base of his tribe, and
secondly his family was not especially influential. He was therefore
not seen as a representative of the Kharuti tribe, a fact demonstrated
by Hezb-i Islamis recruitment among the Suleiman Khel, who were
the Kharutis traditional enemies. The Mustazaffin were also a particular case. The members of this group tended to be townsmen,
often from Kabul or the north of Afghanistan, so that basing itself at
Shashpul, near Bamyan, did not constitute a return to the leaders
fore both an urban distribution and a presence at the periphery of Hazarajat.
This geographical position was partly what conditioned the strategy of the party,
which was always in a minority in majority Sunni areas. In addition, though the
commanders were mainly Qizilbash or sadat, the mujahidin were partly Hazaras,
especially in Ghazni and around Kabul.

164

The Jihadi Parties

qowm of origin as in the case of Masud or Zabihullah, but was only

the result of strategic considerations.
The leaders of the parties also mobilised commanders who had
belonged to a particular political or Sufi brotherhood network. Because of the personalities of their leaders, Mahaz and Harakat-i
Enqelab were the two main groups able to mobilise the Sufi brotherhood networks. Harakat-i Enqelabs links with Sufism were close
because the Mujaddidi stood for the orthodox Naqshbandi tradition, and the two networks of pirs and ulema were difficult to distinguish, as in Herat.
Sayyed Ahmad Gaylani does not have links with all the Qadiris of
Afghanistan, and most of those who recognised him as a pir did not
follow the ritual of orthodox Sufism. His murids belonged mainly to
Mahaz,especially in Ghazni and Jalalabad.38 The presence of Gaylanis
murids was substantial in Ghazni, for example among the kuchis (nomads), who joined Mahaz as a group. The largely Qadiri affiliation of
the Ghilzai nomads had already been observed in the 1930s by Captain
J. A. Robinson.39 Among the various eastern Pushtun tribes the
Ahmadzai mainly joined Mahaz at the outset of the war, in particular
under commander Shamali who was active in the region of
Jalalabad. The decision of the eastern tribes to throw in their lot
with Gaylani was related to the presence of murids, as well as to the
recruitment of families of sadat, such as that of Majruh.40
A commander who was a former Islamist militant rejoined his
former party. This accounted for the presence of a Hezb-i Islami
commander near Keshem in the heart of Badakhshan, a region
which belonged mainly to the Jamiyat. This commander, Abdul
Wadud, was the brother of Dr Omar, an Islamist executed by Daud
who had been close to Hekmatyar. There were also some ulema who
were former pupils of Rabbani, such as mawlawi Shirin in Wardak.
38

The pir of Esfandeh, pir Ali Mahmad Khalifa, a Tajik, and pir mullah Sayyed
Palawan had links with Gaylani, who often visited them in pre-war times. In
addition, the amir-i welayati of Mahaz, Sayyed Nazar Jan, was a Qadiri pir also
linked to Gaylani.
39 Captain J. A. Robinson, Notes on Nomad Tribes of Eastern Afghanistan (1st edition
1934), Lahore, 1980.
40 When the Gaylani family arrived in 1905, Habibullah gave him a property at
Chaharbagh, where pir Sayyed Hasan Gaylani was later buried. His tomb became
a ziarat and many of the Gaylanis murids are found today in this area.

A typology of the parties

165

Finally commanders joined various parties as the result of the

similarity of their social originswhether they were educated, ulema,
or khansto those of the leadership of the party. The educated class
mainly joined the Jamiyat-i Islami and the Hezb-i Islami, which was
therefore principally found within those groups. The individual affiliations of students, who were often from the technical faculties as
with most Islamist movements, as well as those of former government employees, were of greater importance here than with most of
the parties. The case of the province of Helmand showed that Hezbi Islamis recruitment took place within groups affected by modernisationschoolteachers, government employees and such likewhile
its recruitment in the rural and tribalised areas was negligible. Because of the lack of religious legitimacy of its leader, a negative
characteristic of Hezb-i Islamis recruitment was the small number
of ulema who adhered to it and their restricted influence. The Islamist ulema in general gravitated towards Jamiyat-i Islami at the time of
the schism in 1975, with none of the Islamists who had been at the
Abu Hanifa government madrasa joining Hezb-i Islami. Those rare
ulema within the Hezb-i Islami were mostly judges excluded from
senior posts. It was not by chance that the two principal splits within
Hezb-i Islami were carried out by ulema.41
Jamiyat-Islamis membership had the same element of educated
laymen, but Rabbanis personality also attracted Islamist ulema trained
at the government madrasas. These two groups were found among
both the leadership and the commanders. As for the commanders,
the party recruited more among the educated class than the ulema,
the great majority of whom joined the Harakat-i Enqelabeven
those who were Persian-speaking, as in Herat, since Rabbanis prestige could not be compared to that of the Mujaddidi family. The
ulema-commanders of the Jamiyat were generally the products of
government madrasas, such as that of Abu Hanifa in the case of mawlawi Alam at Mazar-i Sharif, or of the faculty of theology at Kabul in
the case of Muhammad Ismail Tariq of Laghman.
41

The second of these, which followed that of mawlawi Khales, was that carried out
by Hekmatyars deputy Qazi Amin, an Islamist alem from a government madrasa.
From 1980 Qazi Amin, who did not relish his subservience to Hekmatyar,
attempted to create a network of support for himself within the party. In 1982
he officially split from Hezb-i Islami, probably with financial aid from Sayyaf,
who had an interest in weakening a competitor.

166

The Jihadi Parties

Those ulema who were not from the government madrasas mainly
joined Harakat-i Enqelab because of the personality of its leader
Muhammad Nabi, who had been close to Ibrahim Mujaddidi before
the war. Mawlawi Khaless fundamentalist approach had also attracted
ulema from the private madrasas, both in the west, from Kandahar to
Farah, and in the east as in the case of Jalaluddin Haqqani and Shir
mullah Khel. Among the Shiites, Hazara ulema made up the officials
of Nasr42 and Sepah.
The khans mostly joined Harakat-i Enqelab, Jebhe, Hezb-i Islami
(Khales) and Mahaz. Harakat-i Enqelab was the party which attracted
the most khans at the start of the war, since Muhammad Nabi represented a degree of continuity with the parliamentary regime, in contrast to the Islamist movements.
Within Jebhe, Sebghatullah Mujaddidi was an alem who because
of his early exile had lost touch with the Islamist movements. In the
absence of militant supporters, he relied on his membership of the
Mujaddidi family rather than on his Islamist commitment in order
to set up his party. Jebhes recruitment came largely from two sources. The lite schools connected to the government of Zahir Shah,
which in Dauds time, were often in opposition, provided the senior
staff of the party. Merchants and artisans, often with links to the Sufi
brotherhoods, made up the majority of the Jebhes commanders and
mujahidin.
Mahaz was the party of those linked to the old orderthat of
Zahir rather than of Daud. These were people who had suffered
from modernisation: the bazaaris (shopkeepers and artisans), the governing class and the major notables. Inevitably, therefore, there was a
generational difference between the officials of Mahaz and those of
the other parties. The bazaaris, often Persian-speaking, also made up
a significant portion of the membership of Mahaz because of their
nostalgia for the days of royalty and the economic structures of that
time, which were more favourable to the artisans.
42

The presence of Nasr was particularly strong in the heart of Hazarajat, especially
in Yakaolang, Lal o Sarjangal, Deh Kundi (Khedir). The Nasr officials were all
ulema: ustaz Abdul Husein Sadeqi, of Turkmen, a disciple of shaikh Akay Alem
(the representative of Khomeini in Afghanistan, who was assassinated under
Taraqi); the ayatollah Abdul Ali Mazari of Charkent, who spent most of the war
in Iran, where he was in close contact with the Iranian clergy; Nadeqi (of Deh
Kundi); Azizullah Shafaq of Behsud; ayatollah Eftekhari (of Dara-yi Suf); shaikh
Qurban Urfani; and hojatoleslam Zaedi (of Yaokalang).

A typology of the parties

167

In the tribalised regions of the east, the Pushtun khans, who were
often the leaders of sub-groups of their tribe, followed mawlawi
Khales. His ethnic origin, his reputation due to his pre-war radio
broadcasts, and his status as an alem enabled him to recruit in a tribal
area. His role was therefore not dissimilar to that of the charismatic
mullahs who crop up in the history of the Afghan frontier.
The Shiite Harakat-i Islami commanders were generally not ulema,
which clearly differentiated them from the pro-Khomeini revolutionary parties. Harakat-i Islami recruited notables: merchants, landowners such as Din Muhammad at Charkent, and some from the
educated class such as Dr Shah Jan at Ghazni.
The imperative of differentiation. Membership of a party also followed a
system of differentiation: two adjacent but competing groups would
affiliate to different and even opposing parties. Local rivalries were
translated into political conflicts, and perpetuated as such. There
were many examples: the affiliation of the Andar tribe to Harakat-i
Islami increased the tendency for the Suleiman Khel tribe to join
another partyHezb-i Islamibecause of pre-existing tensions.
With two such groups in the same party there would be the risk for
the minority group of the power of the majority group being reinforced because it would be better represented in the leadership of
the party. In addition, the coexistence of two groups within the same
party would prove unmanageable in the case of an armed clash.
Through its re-introduction of the autonomy of individual actors,
this imperative based on local configurations militated against what
could have become the automatic operation of affiliation founded
on proximity.
Sometimes the process of affiliation by differentiation reproduced
the pattern produced by proximity; however two groups might be
similar in macro-ethnic termsboth Pushtun, for exampleand
still join different parties. This provides an explanation for the presence, at first sight surprising, of some parties in areas which were
alien to them. The adherence of mullah Naqibullah to Jamiyat-i Islami
seems a case in point. Naqibullah was an Alikozai Pushtun, of the
Durrani confederation, from the Arghandab valley north of Kandahar. None of his social characteristicsa Pushtun small landowner
who had not been politically active before the warappeared to predispose him to join Jamiyat-i Islami. However, in this region Mahaz

168

The Jihadi Parties

membership was largely Barakzai, Harakat-i Enqelab was dominated by Ghilzai from the neighbouring provinces of Ghazni and
Logar, and Hezb-i Islami was locally Ghilzai. Therefore it was probably more logical for mullah Naqibullah to join Jamiyat-i Islami,
which was locally neutral in tribal terms and which also possessed
substantial resources. In addition, the fact of being ethnically marginal within Jamiyat-i Islami may offer advantages, since the party
was impelled to give priority to a Pushtun commander who could
endow it with legitimacy nationally. The same explanation holds
good for the adherence of ras Abdul Wahid of Baghran, in the province of Helmand, to Jamiyat-i Islami in 1988, after a long conflict
with Nasim Akhundzada, his neighbour, when both at first belonged to the same party, Harakat-i Enqelab. In this case strategic considerations may have predominated: Ismail Khan, himself a member
of Jamiyat-i Islami, had made an agreement with ras Abdul Wahid
concerning control over the road between Herat and Quetta.
In a situation of marked fragmentation, the likelihood was that all
the parties would achieve representation, for reasons other than the
imperative of proximity, since the groups present were at least as numerous as the parties. Only the choice of the dominant group, if
there was one, followed the imperative of proximity or representation, while the remainder could be regarded as random choices. The
sole exception to this rule was that two strongly opposed groups would
often belong to the most openly antagonistic partiese.g. Hezb-i
Islami and Jamiyat-i Islami. Such motiveless partisan affiliations were
found especially in Kandahar, at least up to 19889, after which the
pattern tended to become simpler. Many groups adopted contradictory and sometimes very fluid affiliations; adherence here was entirely subservient to an imperative of pure differentiation. There was
also a good example of fragmentation around the town of Kunduz,
where the political map was especially complicated. Within the town
there were populations from all over Afghanistan who had come to
cultivate the land in this northern region. The distribution of the
parties seemed largely random, except for Aref Khan, a Pushtun and a
son of the former mayor of the town, whose pre-war Islamist sympathies had led him to Jamiyat-i Islami.
Hezb-i Islamis recruitment among the Pushtuns was a recurrent
theme in these choices of affiliation, which as a result came to resemble an imperative of proximity. Thus the Pushtuns were often

The parties and the refugees

169

members of Hezb-i Islami when they were locally in the minority,

as in the province of Baghlan or at Mazar-i Sharif. Similarly the
Ghilzai of Kandahar, who were mainly of the Hottak tribe, were
largely Hezb-i Islami in a Durrani environment. However in majority Ghilzai regions, the Ghilzai did not join Hezb-i Islami, but
preferred Harakat-i Enqelab. Groups in minority situations joined
Hezb-i Islami since this enabled them aggressively to defend their
specificity, which was not as much encouraged by the other parties.
This provided a good illustration of the effects of the leaders situation: Hekmatyar, himself a minority Pushtun, recruited among
populations who saw themselves in a similar light.43
Affiliations on the basis of differentiation had the general implication that the parties were more broadly instrumentalised, but they
also affected relationships between groups, in general by exacerbating tensions, since party conflicts on a national scale were reproduced
at the local level. Affiliations based on context were more numerous
among populations organised on a segmentary basis, such as in the
eastern provinces and Kandahar, and in multi-ethnic zones, where
the need for differentiation was most obvious.

The parties and the refugees

From 1979 refugees arrived in Pakistan at the rate of tens of thousands
a month.44 They were estimated in October 1979 at some 200,000,
and would reach a million at the end of 1980 and 2 million in 1981.
The largest number of registered refugees was seen in 1990, with 3.2
million, to which around half a million should be added who were
not registered.45 After this, return to Afghanistan became an increasing
phenomenon. Figures provided by the UN High Commission for
43

To take a further example, Hezb-i Islami was very largely Pushtun in the province of Baghlan. The Pushtuns were relatively recently settled in the region, between 1930 and 1950, and their relations with the Tajiks and the Ismailis were
sometimes tense as a result of competition for land. The coming to power of the
Parchamis did not improve the situation of the Pushtuns, since Babrak Karmal
made a preferential alliance with the Ismailis, whose support was necessary for
the control of the crossroads of Pul-i Khumri. The majority adherence of the
Pushtuns of Baghlan to Hezb-i Islami was therefore likely.
44 On the number of refugees in Iran and Pakistan and their situation, see Encyclopaedia Iranica VII (4), New York: Bibliotheca Press, 1995, pp. 383 ff.
45 Ibid., p. 383.

170

The Jihadi Parties

Refugees have been criticised, since several hundreds of thousands

were not counted, either through delay or to escape supervision.
In Pakistan, to simplify the situation, two waves of refugees may
be distinguished. Until approximately 1984 the refugees, who were
up to 80 per cent Pushtuns, came from the south and the southwest.
In some cases their exodus was a protesta hijrabut more often
the population simply fled the fighting. Generally they arrived in an
organised group, a clan or a village, under the leadership of a khan
or mullah. There was no automatic linkage between the number of
departures and the intensity of the fighting, as it is showed by the
province of Kandahar, although the exodus from Kunar could be directly linked to the fighting. After 1984 there were more from the
north, but they were always a minority among new arrivals. Their
arrival was related to the fighting; sometimes to conflicts between
groups of mujahidin, as well as to bombing and to the deliberate destruction of the harvest. Their exodus, often involving only individuals or families, was undertaken clandestinely, since the main roads
had been closed by the government.
The majority of the refugees in Pakistan, around 2.3 million, lived
in the North-West Frontier Province. In Karachi there were also
200,000 refugees who were not in general gathered into camps. In
1982 the Pakistani administration had also set up a camp in the
Punjab, which by 1987 had a population of 172,000.46 Because of
the scale of the aid, the material conditions of life were not bad. The
rate of infant mortality was lower than in neighbouring Pakistani
villages. Solidarity groupsclan, village or tribewere reproduced
within the camps. The difference in the way of life was above all due
to the lack of land, which had consequences in daily activity and
social status. In addition, the size of the groups of people was larger,
in Pakistan averaging several tens of thousands.
The refugee population was a considerable prize for the Afghan
political parties.47 For this reason, out of aid totalling $1 billion intended for the refugee camps in Pakistan throughout the entire conflict, it may be estimated that only half reached its destination, the
remainder being divided between the Pakistani administration and
46

Brigitte Picard, Les Damns du Penjab, Les Nouvelles dAfghanistan, special issue
Les rfugis afghans, December 1987.
47 Among the refugees in Iran the role of political organisations was more limited
since there were no camps and only the Shiite organisations were active.

The parties and the refugees

171

the Afghan parties, which were active in the camps where they competed with the traditional notables as intermediaries with the administration. Even more than in Afghanistan, the possession of a party
card became universalfor the Pakistan administration this document also served as an identity card. The Islamist parties, principally
Hezb-i Islami and Jamiyat-i Islami, were the most active in organising the population, with schools, security committees and so on.
The refugees dependence on aid enabled the parties to present
themselves as intermediaries and therefore to acquire real power;48
they were able to control the refugees and negotiate with the Pakistani state. A second factor, of some importance, was the absence of
arms in the camps, which meant a lower risk of vendettas and the
consequent fragmentation of parties. Finally, the militants did not
need to justify their local strength in numbers. The displacement of
populations gave rise to new relationships and facilitated interventions by outsiders.
One may nevertheless question how real were the affiliations to
Islamist parties. In some camps such as those in Baluchistan49 the low
level of influence of these parties was attributable to the persistence
of strong tribal structures among the refugees. Frequent pro-royalist
demonstrations and the hostile reception given to Hekmatyar during his visits were evidence of this. In 1987 the Afghan Information
Center, directed by Sayed Bahodine Majrouh,50 carried out a survey
of political opinion among the refugees in the camps. The most
48

See the analyses of Pierre Centlivres and Micheline Centlivres-Demont, Hommes dinfluence et hommes de partis. Lorganisation politique dans les villages de
rfugis afghans au Pakistan in Erwin Grtzbach (ed.), Neue Beitrge zur Afghanistanforschung, Liestal: Bibliotheca Afghanica, 1988, and Micheline CentlivresDemont, Les rfugis afghans au Pakistan: gestion, enjeux, perspectives in Rocardo
Bocco, Mohammad-Reza Djalili (eds), Moyen-Orient. Migration, dmocratisation,
mdiations,Genve:Institut Universitaire des Hautes Etudes Internationales, 1994.
49 On the refugees in Baluchistan, see A. S. Ahmed, The impact of the Afghan refugees on ethnicity and politics in Baluchistan, Central Asian Survey 9 (3), 1990,
pp. 4356.
50 B. Majrouh, the son of Sayyed Shamsuddin Majruh (a former Minister and senator), who was close to Mahaz, ran the Afghan Information Center until his assassination in February 1988. He also left a significant body of poetry, part of which
has been translated into French: see particularly Le suicide et le chant. Posie
populaire des femmes pachtounes, Paris: Les Cahiers des Brisants, 1988 (translation
and adaptation by Andr Velter).

172

The Jihadi Parties

striking finding was the popularity of the former king Zahir Shah
and the desire for a negotiated peace.51 The selection of the sample
was perhaps questionable, but it was also clear that there was no
large-scale ideological affiliation to the Islamist parties.

51

62% wanted a negotiated solution and 73% wanted unity between the parties.
The survey was conducted among a sample of 1,787 educated Afghans, a restriction which raised questions of methodology.

Part III. THE DYNAMICS

OF CONFRONTATION
5. The Kabul Regime
Amadand o sukhtand o raftand (They came, they burned, they left). (Hafez)

In the history of the Kabul regime three separate periods can be distinguished. From 1980 to 1986 the Soviets applied a policy whose
long-term aim was to turn Afghanistan into a new Central Asian
Republic. Between 1986 and 1989, the failure of their project drove
them to prepare the ground for withdrawal, while at the same time
attempting to leave their Afghan ally stronger through a policy of
National Reconciliation. From 1989 to 1992 the regime survived
with Russian assistance, but it became progressively weaker. This
chapter examines the internal balance of forces within the administration, as well as the Soviet-Afghan strategy and the preparations for
retreat, together with the survival of the regime until 1992.

The Party under Karmal

Under Babrak Karmal there was neither institutional stability nor
well-defined legal order. His ten-point document published in 1980,
served as a provisional constitution till 1987, when the Constitution
was finally adopted by a Loya Jirga. Legitimation by way of a popular
vote was never a priority, even at the symbolic level. In theory the
authority of the state rested with the twenty-seven members of the
Revolutionary Council who elected the Praesidium, but these institutions were in reality subordinate to the Party. This was structured
in the classic manner: the Central Committee, with seventy-three
members, elected the eight members of the Politburo and the eight
members of the Central Committees secretariat. Some of the members of the Politburoincluding Najibullah, Nur Ahmad Nur, Saleh
Muhammad Zeary and Babrak Karmalalso sat on the secretariat
of the Central Committee, which further reinforced the concentra173

174

The Kabul Regime

tion of power. There were also the primary organisations of the Party
and its mass organisations, which, outside Kabul, were largely theoretical. In any case the Party never achieved functional harmony because of constant factional clashes. The government lacked unity,
since the most powerful ministries were divided between the Khalqis, who held the ministries of the interior and defence, and the Parchamis, who held the Prime Ministers office and controlled the State
Information Service, the KHAD (Khedamat-i Ittalaat-i Daulati).
A further problem was that the leadership was unable to impose
its will without Soviet support. Indeed, the direct implication of
Babrak Karmal in the invasion had undermined his legitimacy even
within Party circles. He was never able to extend his authority over
the whole of the Hezb-i Demokratik-i Khalq-i Afghanistan; still less
was he able to achieve wider credibility. Babrak Karmal never tried
to establish a cult of personality as did Taraqi or Amin, probably on
the instructions of the Soviets whose desire was to avoid the personalisation of power. In fact, he governed with the help of a small
family group, together with a number of figures supported by Moscow. In addition to Karmal himself this nucleus included Anahita
Ratebzad, who was his mistress; Mahmud Baryalay, his half-brother,
who was married to one of Anahitas daughters; Nur Ahmad Nur,
also married to a daughter of Anahita; Shah Muhammad Dost, whose
daughter was married to one of Karmals brothers; and Generals
Yasin Sadiqi and Abdul Wakil, who were Karmals cousins. Besides,
some appointments were made directly by the Soviets, including
Watanjar, Gulabzoy, Mazduryar, Tanai, Mohmand, Kawal, Panjshiri,
Nazar Mohammed and Gul Dad.
Attempts to widen Karmals support within the Party were often
prompted by the Soviets, but ended in failure. Some Parchamis opposed him, notably Wakil, Mazdak, Kawiani, Suleiman Laeq, Mangal
and Khazemjo; while the few Khalqis recruited by him were Soviet
trusties formerly close to Taraqi. At the insistence of this group
Asadullah Sarwari, a former head of the secret services, and in that
capacity responsible for the anti-Parchami purge, even came briefly
to hold the post of Prime Minister under Karmal, before being
named ambassador to Mongolia in July 1980: in effect a sentence of
exile. The Parchamis did not have the resources for a purge which
would have decapitated the army and weakened the Party, with the
result that only those close to Amin were removed, leaving Karmal

The three strategic dimensions

175

to govern with a very limited Party base. In the event the Soviet invasion and the return of the Parchamis, who had been brusquely
displaced by Taraqi, constituted a major shock to the Khalqis, whose
sense of nationalism was unsettled by the presence of foreign troops.
Karmal was therefore obliged to govern with a Party whose membership, 75 per cent Khalqi, rejected him. The Khalqis always retained their majority within it, and the bid by the Parchamis to
gerrymander the national Party conference in March 1982 by annulling the results of elections in the provinces favourable to the Khalqis ended in failure.
In spite of their original intentions, the Soviets never succeeded in
reunifying the Party, since political allegiances worked on the model
of the qowms. Purges, far from eliminating opposition, only succeeded in deepening splits and exacerbating vendettas.1 The tensions between the Khalq and the Parcham were never resolved and
only the Soviet presence, and pressure from the mujahidin, prevented
them on the whole from turning into armed conflicts.

The three strategic dimensions

With the limited resources at their disposal the Soviets adopted three
complementary strategies: Sovietisation, pacification and open warfare. First the authorities attempted to sovietise the Afghan towns
on the model of the Central Asian republics. Secondly, pacification
was initiated, employing various tactics such as the setting-up of
government militias, recruitment of notables, a nationalities policy
and the institution of a government-backed clergy. The objective in
this case was not to sovietise: the government aimed only to maintain the neutrality of the population in order to establish buffer zones
against the regions held by the guerrillas. The intention, therefore, in
contrast with the Khalqi ideal, was gradually to pacify particular regions or segments of the population, while at the same time reestablishing the legitimacy of the state. In the end open warfare broke
out in the regions outside government control, where the principal
objective of the Soviet-Afghan army had been to expel the population and thus weaken the guerrillas.
1

See Olivier Roy, Le double code afghan. Marxisme et tribalisme, Revue franaise
de science politique, 35 (5), October 1985.

176

The Kabul Regime

Sovietisation. The inauguration of the first bridge over the Amu

Darya between Termez, then in the Soviet Union, and the Afghan
town of Ayratan symbolised the incorporation of Afghanistan within
the Soviet sphere. The process of integration depended on direct
control by Moscow, and at the same time the promotion of Afghan
relations with the Central Asian republics. In the event the reaction of
the Central Asian communists to the invasion of Afghanistan was positive, since Afghanistans inclusion in the socialist camp tended to reinforce their influence in relation to Moscow. In particular the Uzbek
leader Rashidov made a point of welcoming Afghan students.2 In
the early years of the occupation, however, the influence of the Central Asian republics remained low. Only after the Soviet retreat were
their relations with Afghanistan closer, especially with the latters
northern provinces.
Integration had an institutional dimension.In addition to the occupying troops, some 3,000 Soviet advisers, of whom half were soldiers, oversaw the totality of Afghanistans institutions, each of which
was directly linked to the corresponding Soviet organ. Russian was
the customary working language in these bilateral contacts. The Soviets took charge of the general direction of the Kabul regime, and
even of day-to-day administration. The most influential Russian figures did not hold official positions. The ambassador3 had little real
influence, but the senior KGB official played a key role. Most Afghans
who were in contact with the occupiers stress the superiority complex of the Soviets in their dealings with the Afghans.
For the people Sovietisation meant ideological training in the
schools and the indoctrination of workers and officials within the
unions, militias, womens organisations etc. Publications in Russian
became more numerous, and television propaganda was insistent.
Thousands of children were sent to study in Central Asia with the
aim of preparing a new Sovietised generation to take the country
over by the year 2000. The transformation of urban society continued its progress in various fields, in particular in relation to women,
who shed their veils and adopted western dress in Kabul and more
2

Marie Broxup, The Soviets in Afghanistan: the anatomy of a takeover, Central

Asian Survey, 1 (4), 1983.
3 The post was occupied in the early 1980s by Fikret Ahmedzhanovich Tabiev, an
Uzbek-speaking Tatar from Bukhara.

The three strategic dimensions

177

timidly in Mazar-i Sharif. Integration also had its economic dimension: Afghanistans gas reserves were exploited by the Soviet Union,
which did not buy its gas at the international market price, while
production fell to half its pre-1978 level. On the other hand the Soviets never attempted any collectivisation of the Afghan economy,
especially of trade.
The counter-insurgency. Sovietisation required mastery of the military
situation. The Soviets therefore tried to crush the resistance by indirect strategies, a technique which had served them well in the 1920s
against the Basmachis.4 In the Afghan case the KGB was the principal instigator, and the Afghan secret service, the KHAD,5 was the key
instrument of the counter-insurgency policy.
During the war the secret services became a complex organisation,
probably the best financed and most efficient institution in Afghanistan. Since accounts of it are fragmentary, the synthesis here suggested does not claim to be more than tentative, being based on a
number of interviews with former KHAD officials. From the moment he took power Karmal was to make the KHAD the principal
instrument of his policy, while ensuring that its recruitment was primarily Parchami, with the particular aim of countering the Khalqi
dominance in the armed forces. The KGB official who oversaw the
KHAD therefore took a detailed interest in appointments to the
Afghan army, where final decisions were taken in consultation with
Soviet officers. Relations between the KHAD and the Afghan army
4

Even before the invasion, the Afghan communists had turned to the Central
Asian model. On 8 October 1978 the Khalqi regime set up the first Committees
for the Defence of the Revolution in the towns, on the pattern of those set up in
Central Asia in the 1920s to counter the Basmachis. See also Marie Broxup, The
Basmachis, Central Asian Survey 2 (1), 1983, and Joseph Castagn, Les Basmatchis,
Paris: Leroux, 1925.
5 The employment of spies was a tradition of the Afghan authorities since Abdul
Rahman Khan, but a modern intelligence service was only set up in the 1930s.
After 1978 the new services were called AGSA (Da Afghanistan da Gatay da
Satanay Edara; Afghanistan Security Service), then under Amin KAM (Da Kargarano Istikhbarati Muasasa; Workers Security Institution). Karmal then set up
the KHAD (Khedamat-i Ettelaat-i Daulat; State Information Services) which
was to become under Najibullah the WAD (Wezarat-i Amniat-i Daulati; Ministry for State Security), before disappearing along with the regime in 1992.
A Riasat-i Amniat-i Melli (National Security Services) was later to operate during the presidency of Rabbani.

178

The Kabul Regime

were therefore consistently bad. The KHAD was directly linked

with the KGB: several hundred Soviet advisers, of Russian, Uzbek
and Tajik origin, ran the service, down as far as the regional level. In
this sense integration was more real than in the Afghan army, where
officers were often in conflict with their Soviet colleagues.
Between 1980 and 1986 the KHAD was run by Najibullah, who
had been personally selected by the Soviets, and in particular by Yuri
Andropov. At the head of the organisation the director was assisted
by between two and four deputies, responsible for some twelve departments, whose number varied from time to time.6 Najibullah appointed to the KHADs key positions people with whom he had
personal ties. On the evidence of known cases, his family and professional networks, including the Faculty of Medicine, seem to have
been crucial.7 The KHAD employed between 30,000 and 40,000
official agents, who were excused military service, and possibly around
the same number of clandestine or unofficial agents. The organisation had its own military unit, including a national guard. The regional organisations were modelled on the central organisation, with
a staff of 1,000 in regions where the KHAD functioned fully, for example in Kandahar, Jalalabad, Herat, Kunduz and Mazar-i Sharif. An
idea of the importance given to the KHAD can also be gleaned from
its finances. At the outset its budget was not fixed, credits being made
available according to need. In 1991 it is said to have been granted
$300 million.8
The primary objective of the counter-insurgency was the reestablishment of authority and of the basic legitimacy of the state in
the countryside. To this end the regime set aside its most controversial policies; thus agrarian reform was discontinued by Karmal
in March 1981, only to be resumed in February 1982, although it
now took account of the traditions and religious sentiments of the
6

For a list of the various offices and Najibullahs deputies in the KHAD, see
Mohammad Nasir Kemal, op. cit., pp. 4423.
7 For example, Najibullahs deputy, General Baqi, a Pushtun engineer from Paktya,
was one of Najibullahs distant relatives. Tareq, a Tajik from Parwan, who was the
official in charge of the KHADs internal security, was Najibulahs childhood
friend, while his successor, Dr Shir, a Pushtun from Nangrahar, knew Najibullah
at the Faculty of Medicine and became his confidant. It was Dr Shir who transferred Najibullahs money to India when the regime fell.
8 This figure was given by a person close to Najibullah (interview, Kabul, 1997).

The three strategic dimensions

179

people, a position which concealed nothing less than the discontinuation of the redistribution of land. In addition, this policy had
four concomitant aspects: the establishment of an official clergy, the
policy of nationalities, the recruitment of notables and the establishment of militias.
In relation to religious policy the era of the Soviets represented
a break from the Khalqi period.9 Atheist propaganda campaigns, a
frequent occurrence under Taraqi, came to an end, and the regime
presented itself as the protector of Islam. Religious freedom was
guaranteed in the provisional Constitution of 1980, particularly for
the Shiites. In the same way, in cases where the legal situation was
fluid, judges were allowed to turn to the shariat as a source of law.
The regime began once more to organise the hajj, and multiplied its
contacts with Islamic institutions in the Soviet Union and elsewhere.
Mosque construction and the restoration of ziarats were also utilised
for propaganda purposes. Also on occasions the regime resumed the
use of Islamic terminology. Those who died fighting for the government were referred to as shahids, and the struggle against the guerrillas became a jihad. The opposition was stigmatised, in a Quranic
expression, as munafiqun, or hypocrites. They were also accused of
destroying mosques and assassinating mullahs, which was sometimes
true, since the mujahidin often targeted religious figures who had
pledged their loyalty to the regime. In addition the mullahs were supposed to aid the regime in the dissemination of the revolution,
which was presented as a continuation of the reforming policies of
Amanullah. The KHADs Bureau 7 was given the task of secretly
disseminating instructions for the mullahs sermons, and of maintaining surveillance over the urban mosques. The Kabul government
also made use of pirs as agents, for example in the province of Logar,
in the uluswali of Baraki Barak.
The indoctrination of an Afghan clergy on the Soviet model was
an important objective of the counter-insurgency. Karmal gave a
new impetus to the transformation of the ulema into state officials, a
process which had already been under way for a century. The Jamiyat
ul-Ulema (Society of Ulema), dormant under Taraqi, was reactivated.
In 1982 the authorities set up an Edare-i Shuun-i Islami (Department of Islamic Affairs) answering directly to the Prime Minister.
9

See Asta Alesen, Islam and Politics in Afghanistan, London: Curzon Press, 1995.

180

The Kabul Regime

The Shura-yi Ali-i Ulema wa Ruhanyun (Council of the Ulema

and the Religious) attached to this department had a consultative
role, with ulema designated by the Prime Minister; from 1983 the
department was attached directly to the Paderwatan (see below) and
in March 1985 it became the Wezarat-i Islami wa Awqaf (Ministry
of Islamic Affairs and Waqfs) headed by mawlawi Abdul Wali Hujat.10
In spite of his efforts, the government did not succeed in attracting
well-known ulema: the majority of the religious officials were illiterate mullahs, educated in the private madrasas.
The second plank of the pacification programme, the policy of
nationalities, also failed. In the Soviet Union the communist theory
of nationalities laid down by Stalin was unquestioned before Gorbachev, and for the Afghan communists it was the unchallenged intellectual framework. A nationality was defined, according to this realist
approach, by a language, a territory, a culturewhich might include
a religionand by common economic interests.In reality the practice
of the Afghan communists put into play a definition of nationalities
which served the political objectives of the authorities. The government distinguished six major ethnic communities: Baluchis, Pushtuns, Tajiks, Uzbeks, Turkmens and Hazaras, together with two minor
ones, the Pashas and the Nuristanis. Following the Soviet model the
communist authorities permitted these groups a large measure of cultural autonomy11 in exchange for a strict ideological control which
relegated cultures to the level of folklore. The objective was to reinforce ethnic segmentation in order to enhance state control. In the
context of Central Asia this policy, applied over the long term, led to
10

A Tajik from the province of Takhar, Abdul Wali Hujat was a former official of
the Ministry of Justice who was imprisoned for four years under Zahir for his
unorthodox views. He was close to leftist movements such as Setam-i Melli, and
became the public prosecutor of Herat under Taraqi, then president of the Department of Islamic Affairs in 1983, finally becoming a Minister in 1985. For an
analysis and an excellent biography, see Chantal Lobato, Un islam conservateur au
service du communisme: Kaboul 19801985, Paris: EHSS, 1986, p. 75 and Kaboul
19801986: un Islam officiel pour lgitimer le pouvoir communiste, Central
Asian Survey, vol. 7 no. 2/3, pp. 838, 1988.
11 The foundations of this policy were in fact laid from the inception of the regime,
with decree no. 4, which allowed the use of Turkmen, Baluchi, Uzbek and
Nuristani as national languages. The authorities established radio broadcasting
in these languages. Two newspapers, Gors and Yulduz (in Uzbek and Turkmen
respectively), were published in Kabul.

The three strategic dimensions

181

the creation of real nationalismsUzbek, Turkmen and more tenuously Tajikwhich had not existed some decades earlier.12
On the institutional level the Ministry for Tribes and Nationalities replaced the Ministry for Frontiers, which had traditionally been
responsible for the management of the eastern Pushtun tribes. In the
context of the policy of nationalities the government published periodicals such as Melliatah-yi Baradar (Brother Peoples) in the principal
languages of Afghanistan. In April 1987 individual systems of transliteration were developed for Baluchi, Pasha and Nuristani, which
at least in theory boosted local particularism at the expense of the
jihad. The policy of nationalities was also a means of exerting pressure on neighbouring states where the inhabitants belonged to the
same ethnic communities. The Baluchis, who were also represented
in Iran and Pakistan, were awarded a national day by the communist
government, which at the time gave refuge to Baluchis of the Pakistani opposition.13 On 19 October 1980 Karmal announced on Radio
Moscow his support for the struggle of the peoples of Baluchistan
and Pushtunistan. There was constant contact between the Pathan
nationalist movements and the Kabul government. In 1988 Khan
Abdul Ghaffur Khan, the historic leader of the Pathan nationalists,
was buried in Jalalabad in accordance with his wishes.
This model was to a great extent inappropriate to Afghan circumstances and therefore impracticable. In reality, as long as the Soviets
were there, belief in the jihad stood out in opposition to all ethnic
and national demands. In addition, the ethnic communities did not
enjoy homogeneous territories owing to the displacement of the
Pushtun tribes, the territorial dispersal of the Tajiks, and so on, while
in general there was an arbitrary quality to the criteria defining a
nationality. For example, the Nuristani dialect adopted as a standard
was not understood in all the valleys of Nuristan. Finally, the means
employed had only a restricted effect: publications had an impact
only on an urban and literate public, i.e. a small minority, although
radio broadcasts had a larger potential audience.
Only one community, that of the Hazaras, approximated at all
closely to the government model. In this territory, which was in any
12
13

Olivier Roy, La nouvelle Asie centrale ou la fabrique des nations, Paris: Seuil, 1998.
Witnesses were able to visit these camps in 1978 (personal information from
Stphane Thiollier).

182

The Kabul Regime

case strategically unimportant, the Soviets refrained from military

operations in favour of an indirect approach. The war was thus the
cause of an alteration in the relationship between the Hazaras and
the state, especially under Karmal. The regimes recruitment policy
and the massive migration of Hazaras to Kabul during the war led in
due course to a marked increase in their numbers within the administration. However, attempts to open a dialogue between the government and the Hazaras made no progress, since Hazarajat was the
scene of a civil war between pro-Khomeini revolutionary groups
and conservatives. The main issue was the refusal of the Kabul government to recognise a defined Hazara territory, probably because
of Khalqi objections, since it would have opened the door to a form
of federalism. During this phase of the war the policy of nationalities
finally collapsed, since the macro-ethnic groupsPushtuns, Baluchis, Uzbeks and Tajikswere unable to provide the framework
for a political mobilisation. To take effect allegiances needed to be
pledged at local level.
After the disaster of the Khalqi period, the Soviets aimed at mastering the situation by enlisting the notables, hoping thus to pacify
the countryside and isolate the mujahidin. The governments aim
was to play its part once again in the traditional pattern of relations
between the state and the local notables, while blocking the rise of
new lites, especially the Islamists. On 15 June 1981 the Jebhe-yi
Melli-yi Paderwatan (National Front for the Fatherland) was set up,
with the purported aim of providing an apolitical structure for the
notables, in the context of which they could support the government.
The Paderwatan brought together peasant organisations, religious and
other groups which were intended to represent the Afghan people,
and strove to continue the modernising and nationalist tradition
established by Amanullah.
In the governments scheme the Paderwatan was intended primarily for the Durrani Pushtuns, in general with a leader from one
of the great families of Kandahar. In spite of this the establishment of
the apparatus of the Paderwatan was more successful in the predominantly Uzbek northwest provinces. In these regions the mujahidin
found keeping up their resistance harder than elsewhere because of
their distance from the sanctuary of Pakistan, their proximity to the
Soviet Union and the difficult terrain, much of it desert. The Uzbeks

The three strategic dimensions

183

were also more receptive to the Soviet approach, perhaps because of

the neighbouring presence of Uzbekistan which presented a model
of modernisation at a time when Uzbeks were in a minority in
Afghanistan and had traditionally been excluded from power.
Although there were some successes, the achievements of the
Paderwatan were disappointing, especially in the view of the amount
of effort invested. By placing their reliance on the traditional notables who were in decline, the Kabul authorities were able to enhance
their power only marginally. In addition, notables who swore allegiance were obliged to flee to the towns, thus losing their rural clientele. But in spite of these shortcomings the Paderwatan continued
to provide the political framework for recruitment between 1982
and 1987.
Beyond the recruitment of the notables, the Paderwatans objective was to facilitate the formation of militias, the distribution of
which was dictated by geographical considerations. These arose
from the need to establish buffer zones between the towns, which
the government sought to Sovietise, and the countryside, where the
mujahidin were too well established for the army to achieve anything
beyond combing operations which always needed to be repeated.14
Gradually militias began to be deployed along the roads, as between
Kabul and Mazar-i Sharif, and between Kandahar and Spin Boldak,
as well as on the outskirts of the towns where they provided an essential reinforcement for the security perimeters set up between 1983
and 1987 to limit the incursions of the mujahidin. It was in these
zones that the efforts of the government began to take effect, and the
level of military pressure frequently obliged villagers to choose between exile and compromise with the authorities. As a quid pro quo
for pledges of allegiance the government provided material help, for
example with electricity supplies and tractors for the harvest, without demanding ideological commitment in return. The governments
classification of the militias was complex. To simplify, a distinction
was drawn between units formed by the government and pre-existing groups which had rallied to the governments side.
The militias formed by the government enlisted their members
on an individual basis, sometimes because of ideological consider14

This section follows the lines of an article written by the author in collaboration
with Chantal Lobato, The militia in Afghanistan, Central Asian Survey 8 (4),
1989, pp. 95108.

184

The Kabul Regime

ations, but more often because for financial reasons. Generally they
were deployed in areas where some measure of industrialisation had
occurred, and where the population was not too strongly in the grip
of traditional communal solidarities. For example in Lashkargah,
in the province of Helmand, Khanos militia recruited men of all
ethnic communities as individuals, though nepotism and family connections, here as elsewhere, played their part.15 In this militia the
ideological content was strong, with a violent rejection of the mullahs and of traditional society. The officers continued to employ
outrightly atheistic language even after the fall of Kabul to the mujahidin. In the province of Jozjan, Rashid Dostams16 militia, one of
the largest in the country, also proceeded on the basis of individual
recruitment. Rashid Dostam, a former gas employee, was typical of
a class of Afghans at odds with the traditional rural values. Sometimes landless peasants who had benefited from the agrarian reforms
organised themselves to defend their gains, for instance in Kandahar
in the Arghandab valley.
The principal purpose of the urban militias was to preserve urban
security and to occupy territory in the wake of Soviet-Afghan offensives, in order to assist the peasants and enlighten them about the
nature of the Revolution. For example, after the Soviet offensive of
July 1984 in Panjshir, 4600 members of the Sepah-yi Enqelab
(Army of the Revolution) were sent for two months into the valley,
in an operation extensively covered by Kabul television. The operation was repeated over a further two weeks in 1985, with equally little result. That same year the dispatch of 200 members of this militia
to Kandahar was also a failure, particularly because of the resignations it prompted among the militiamen, who were made to stand
guard over government buildings and whose sojourn was compulsorily prolonged. On the other hand, the governments urban militias functioned relatively smoothly in the towns of the north-west
such as Maymana and Shibergan, which had never been threatened
by the mujahidin.
15
16

Interview with Khano, autumn 1992.

Dostam was born into a poor family at Khwaja Doki, near Shiberghan. His
father was a farmer. Dostam was employed in the gasworks at Shibergan, where
he was recruited by the Parchamis. He became a militiaman and did six months
military training in Kazakhstan in 1980. Cf. Ludwig W. Adamec, op. cit., p. 66.

The three strategic dimensions

185

However, in the formation of such militias gaining the allegiance

of pre-existing groups was a much more effective procedure than individual recruitment. Such groups, if sufficiently large and aggressive,
formed rural militias (Ghund-i Qawmi or Kandak); others were confined to maintaining the security villages under the patronage of the
Paderwatan. Militia groups came from various ideological backgrounds: leftists, religious minorities and former mujahidin.
The majority of the leftist groups had signed up to the government of Babrak Karmal. In the Taraqi-Amin period government repression constrained the Setamis to retire to the countryside where
the militants, most of whom were townsmen, had few roots.17 In
particular their progressive positions were not enough to earn them
the support of the peasants when they were attacked by the mujahidin, which placed them in an untenable situation. After the Soviet
intervention, which brought about the Parchams return to power,
the Setamis progressively rallied to Babrak Karmal, who was a personal friend of Taher Badakhshi, the historic leader of the Setami
movement. SAFZA rejoined the government in 1981, although it
did not officially take the plunge till 1986 in the framework of the
policy of national reconciliation. The Setamis were the source of important militias in the north-east, where they were well represented
before the war. Masuds defeat at Shahr-i Bozorg in Badakhshan in
1990 was largely due to the intervention of these militias, though
they were helped by Hezb-i Islami.
The Ismaili community, founded upon religious rather than ethnic solidarity, was also courted by the government. The Ismailis lived
in two strategic locations: one at Pul-i Khumri, a key position on the
road linking Kabul with Mazar-i Sharif, and the other close to the
Pakistani frontier in the province of Badakhshan. Although the
community based in Badakhshan had too little demographic weight
to play an independent role, that in Pul-i Khumri found itself at the
heart of the conflict. After a phase in opposition during the Khalqi
period, the Ismailis cooperated with the government, since they were
repelled by the Islamic position; the Sunnis did not regard them as
Muslims. The government was therefore able to recruit them whole17

From the moment of the Parchamis displacement by the Khalqis, the Setamis
were subject to severe repression. Taher Badakhshi and mawlawi Bahes, who had
been imprisoned since the summer of 1978, were executed.

186
The Kabul Regime

The Kabul regime, 1989.

The three strategic dimensions

187

sale from 1980 onwards, in exchange for material benefits and autonomy for their community.
Finally, the vigorous rivalry between commanders led to recurrent clashes between mujahidin. The government was therefore able
to play the role of a third party in these conflicts, tipping the scales in
favour of one side or the other. In general the recruitment of mujahidin was carried out through the Paderwatan, which appointed an
elder (rishsafid, a white beard) to make contact with the target
group. If the negotiations were successful the group officially surrendered its arms, only to take them up again at once, but this time in
the service of the government. Occasionally these encounters turned
out badly: in 1983, seventy government officials were massacred
during an allegiance ceremony near Jalalabad.
The recruited group in no way altered its behaviour, neither submitting to military discipline nor renouncing their religious practices. Recruitment was therefore not synonymous with the return of
the state. The ambiguous position of those mujahidin who rallied to
the government side was a recurrent factor, for example in the region of Herat. Among the most spectacular pledges of allegiance was
that made by Ismat Muslim, an Achekzai Pushtun, whose forces were
deployed along the road between Kandahar and the Pakistan border.
In 1984, following clashes with rival groups, he crossed over to the
government side together with several hundred men of his tribe, a
move he seems likely to have made after payment of $200,000. Ismat
Muslim, the son of a general and educated in the Soviet Union, was
a personal friend of Babrak Karmal and had little sympathy with the
puritan position of the mujahidin. All in all, the success of the policy
of recruitment also depended on the ability of the army to put pressure on the mujahidin.
Military operations. By restricting the size of its occupying forces,
Moscow attempted to escape entanglement of the kind which had led
Americans to commit more than half a million men in Vietnam.18
On the other hand the Soviet contingent, which never numbered
much more than 100,000 men, was unable to compensate for the
weakness of the Afghan army, especially since the Soviet troops, who
18

For an overall view of military operations, see Mark Urban, War in Afghanistan,
London: Macmillan Press, 1990.

188

The Kabul Regime

had little taste for combat as well as being ill-equipped and underfed,
were restricted to useful Afghanistan: the cities and the main axes
of communication. It seems that no reliable information is available
on the make-up of the Soviet forces, not even relating to the number of troops who passed through Afghanistan, though the figure
officially quoted is half a million. The first arrivals were volunteers,
who believed they were defending Afghanistan against the activities
of the CIA, but the truth gradually dawned concerning the hostility
of the population and the serious aggression by their own comrades
faced by the young conscripts, including beatings and rape. From
this point onwards, in the absence of further volunteers, those conscripted were seldom told their true destination.
The prime objective of the Soviet high command was to increase
the strength of the Afghan army, in view of the difficulty of keeping
control over an area larger than France, with terrain very advantageous to the guerrillas. In 1978 the Afghan army numbered 110,000
men but, eroded by desertion, it fell to 25,000 men by 1980. The
number under arms, which was to reach 150,000 in 1988, at the
moment of the Soviet withdrawal, remained insufficient in spite of
programmes of forced conscription, the lowering of the age of conscription, and the extension of military service to three years, with
two years of mobilisation for reservists under thirty-nine. These
measures from time to time occasioned protests: in November 1982
the eastern Pushtun tribes demonstrated against conscription, which
violated their traditional exemption. The Soviets also wanted to reform the structure of the Afghan army, and to revise its strategy to fit
in more closely with their own military system. In 1980 Afghanistan
was divided into seven regions, each with a Soviet commander and
Afghan officers as regional commanders under his authority. However, the Afghan army continued to be badly organised and was divided into the regular army, the gendarmerie (Sarandoy) and the frontier
guard. And up to the end armaments were in short supply.
Relations between the Soviet and Afghan armies were not particularly good, and the Soviets always took the Afghans to task in particular for their untrustworthiness. For this reason Masud was always
aware of operations in the Panjshir well before they took place, even
if he was sometimes surprised by their magnitude. In 1985 a number
of Afghan generals were even arrested for passing information to the
mujahidin. Faced with the multiplication of leaks, the Soviets quickly

The three strategic dimensions

189

adopted the practice of informing the Afghan officers only at the last
minute before joint operations. The Afghans on the other hand
complained about the inertia of the Soviet troops, who declined to
become involved when Afghan troops were attacked. The major
battles, and therefore the largest losses, involved the Afghan forces,
which fought in the front line and defended isolated outposts.
The objectives of the Soviet command were basically to preserve
the security of the towns and the lines of communication, and to
curtail infiltration by the guerrillas from their sanctuary in Pakistan;
frontier control was necessary to deny or obstruct the reinforcement
of the guerrillas in men and supplies. However, the level of Soviet
involvement and tactics changed over time, as was seen by the varying level of their losses.
There were three distinct periods. Between 1980 and 1983 the
Soviets adopted a defensive position, avoiding direct involvement
and relying largely on the Afghan forces. The priority accorded to
static defence of strategic areastowns, bases, roadsaccounts for
the low level of casualties: 1,500 in 1980 and 2000 in 1981. In 1982
Andropov sought to limit the losses and attempted to reach a diplomatic solution. In this period there were few attacks, and government operations were aimed mainly at dislodging the guerrillas from
the regions surrounding the towns, where they were infiltrating at
night and attacking guard posts. From January to March 1982 the
Soviets attempted to improve the defences of the major towns, including Kandahar, Herat and Mazar-i Sharif. They surrounded and
destroyed infiltrated quarters, a tactic which resulted in substantial
civilian casualties, but in spite of these operations security left much
to be desired. At Mazar-i Sharif commander Zabihullah even succeeded in seizing the civil airport in a commando operation, and at
Kandahar there were battles in the town every night.
The absence of real results in due course obliged the Soviets to
become directly involved, so that in 1984 and 1985 operations were
at their height, with a strategy directed at winning the military victory by wearing down the resistance. In 1985 there were six offensives involving more than 5,000 Soviet troops, particularly in Panjshir
and in the east, and in the Paktiya region 30,000 men took part in
one operation. Soviet losses reached a peak, with 2,300 deaths in 1984
and 1,868 in 1985. During these two years the government also strove

190

The Kabul Regime

to protect the towns with guard posts every 100 or 200 metres, the
space between these positions being blocked by anti-personnel
mines. To gain entry to the towns three or four security cordons had
to be crossed. In Mazar-i Sharif, where the operation was especially
successful, the mujahidin were unable after 1984 to get into the town,
and military pressure obliged them to fall back towards the south.
Elsewhere major cleansing operations were regularly launched in
areas held by the mujahidin to drive the population out of regions
the government was unable to control. Finally, after 1986 the Soviets
were getting ready to pull out and launched fewer offensives. Losses
fell significantly at this time, with 1,300 killed in 1986, 1,215 in 1987
and 759 in 1988.19
Inadequate logistics and lack of training for guerrilla warfare were
two factors which accounted for the poor performance of the Soviet
military. The war uncovered the decline of the Soviet system and indeed marginally accelerated it. The lack of experience of the Soviet
troops in mountain warfare, the lack of initiative of the officers, and
insufficient knowledge of the terrain were all factors which led to
tactical errors, and in particular to excessive use of tanks. The Soviets
were insufficiently mobile, surrounding and indiscriminately bombing terrain in reaction to guerrilla offensives, but failing thereafter to
occupy it because of the lack of efficient infantry. The war was waged
too inflexibly, with aerial bombings between set hours and no night
attacks, at least in the early years. Although the ambushes mounted
by the mujahidin were often clumsy, the Soviets ability to respond remained low. In the end the logistical problems reached a point where
attacks had to be drawn to a close after a week, just at the moment
when the mujahidin were running short of ammunition. Most of the
Soviet prisoners were captured while searching for food.
Although logistics remained hopelessly inadequate up to the end,
the patent unsuitability of the Soviets military structure led to changes in tactics. The army was reorganised into smaller units, and commandos were used to seal off frontiers and mount nocturnal raids.
The most significant change related to the increasing role of air
power in the conduct of the war. For the Afghan and Soviet forces
19

Even if these figures are not accurate in absolute terms, the year-on-year variation appears to be authentic (see below).

Preparations for the retreat

191

aviation became the decisive factor in the anti-guerrilla struggle,

since it was more effective than traditional large-scale operations
and allowed casualties to be kept down. In this way bombing raids
with MiG 24 assault aircraft became more frequent, with support
from MiG 8 observation planes. From 1985 onwards the Soviet
forces adopted the technique of covering the flanks of their attacks
with helicopter-borne units, especially during mountain warfare.
However, because the mujahidin became better armed, air support
was less effective from 1986 and 1987, which entailed difficulties for
large-scale operations and in the supply, and therefore the survival,
of isolated outposts.
At the moment when the Soviets withdrew they had succeeded
in establishing satisfactory security in the towns in the north and to
some extent in Herat, although Kandahar remained largely beyond
their control. As for the security of frontiers and lines of communication, their other priority objective, the failure of the Soviets was
obvious. Regular attacks launched to wipe out mujahidin groups
near the roads or in the frontier zone with Pakistan were unsuccessful, largely owing to the nature of the terrain and to the scale of
Pakistani support. The multiplication of commando operations and
the millions of mines laid on the Afghanistan-Pakistan frontier were
similarly ineffective, since the geography of the region meant that no
real closure of the frontier could ever be achieved.20

Preparations for the retreat

The decision to withdraw from Afghanistan was taken in the spring
of 1986. From his accession to power a year earlier Gorbachev had
decided to resolve the Afghan crisis, which was impeding his policy
of reconciliation with the West. The social mobilisation effected by
glasnost and perestroika was in practice insufficient to counter the
shortfall in production in the Soviet Union, which was becoming
increasingly dependent on the West in technology. The Afghan war
became a test of Soviet good will and of the renunciation of its expansionist policy. The withdrawal was all the more necessary due to
20

Mines were much used: according to the US government, between 10 and

16 million were laid. The Soviets are reported to have installed 2,131 minefields,
of which only 1,518 had been cleared at the time of their departure. Pakistan
Times, 16 August 1988, and AFP 25 June 1988.

192

The Kabul Regime

the mounting diplomatic and internal costs of the occupation; the

war had left the Soviet Union diplomatically isolated. The invasion
of 27 December had been carried out in defiance of international
law, and the only justification offered by the Soviet government relied on a verbal appeal made by the Afghan leader Afizullah Amin
before his execution by Soviet special forces. Outside the immediate circle of Soviet client states this was accepted by no-one, and on
20 November 1980 the General Assembly of the United Nations
called for the withdrawal of foreign forces from Afghanistan, and
the resolution was thereafter renewed year by year. At the moment
of each vote the UN became the scene of a diplomatic confrontation between the Soviet Union and its allies on one side, and a
growing majority of UN member states on the other.21 The repeated censure of the Soviet presence in Afghanistan represented a
major setback for Soviet diplomacy in relation to the Third World
and particularly Muslim states.
The war was also a source of tensions within the Soviet Union.
Though hard to calculate, the cost of the war was probably $48 billion a year in the 1980s.22 The official Soviet casualty figures
13,369 dead, 311 missing and 35,578 woundedare an understatement: other estimates tell of 4050,000 deaths. Unofficial estimates
made by the Soviet people at large were considerably higher, while
the lack of any credible rationale for the war made it all the more
unacceptable.23 Drug use, which had not previously been significant,
also became widespread. From 1983 popular opposition to the war
began to be voiced.24 The relative freedom of expression under
Gorbachev later permitted a number of demonstrations to take place,
which made clear the popular view of the posting of conscripts. For
21

The first resolution condemning the Soviet invasion met with the approval of
104 countries against 18, with 18 abstentions. The figures for that of 15 November 1984 were 119 for, 20 against and 17 abstentions; that of 13 November 1985:
122 for, 19 against, 12 abstentions. On 10 November 1987, 123 votes were cast
against the Soviet Union, an indication of its mounting isolation.
22 See Anthony Arnold, The Fateful Pebble: Afghanistans Role in the Fall of the Soviet
Empire, California, Presidio, 1993, p. 185 ff.
23 Official figures given in May 1988 by General Lizichev. For an assessment of
these figures, which seemed considerably to underestimate the losses, see the
article by Valeri Konovalov, Legacy of the Afghan War, some statistics, Radio
Liberty, 1 (4), 1989, and Anthony Arnold, op. cit., p. 190.
24 Anthony Arnold, op. cit., p. 194.

Preparations for the retreat

193

example, on 20 May 1985 there was a demonstration in Erevan of

200 relatives of conscripts who were about to set off for Afghanistan.
A fact understood by few observers is that Najibullahs accession
to power in 1986 signalled a major change in Soviet policy. To safeguard the survival of his regime after the withdrawal, he was to play
a double game: within the party he consolidated his authority and
set in train structural changes, while in his external relations he
attempted to bring the mujahidin into a process of political negotiation, and simultaneously undermined their position by going through
the motions of a liberalisation designed to attract the population into
his camp.
The restructuring of the Party. The advent of Najibullah had been orchestrated by the Soviets against the wishes of a large majority of all
factions of the Hezb-i Demokratik-i Khalq-i Afghanistan. In the
Soviet view his personal qualities and social background were appropriate to their plans.25 Andropov had in any event sounded him
out as a future successor to Karmal as early as 1982. He was young
aged thirty-nine in 1986as well as being energetic and something
of an orator, in contrast to Karmal. He was also well placed as the
former head of the KHAD to develop a counter-insurrectionary
policy and to block any coup attempt. In addition his background
conferred on him greater legitimacy than that enjoyed by Karmal,
both inside and outside the Party. Najibullah was a Pushtun speaker,
born in Paktiya, and married to a Muhammadzai (the royal clan),
which increased his acceptability to the Pushtuns. Even after the fall
25

Najibullah, the son of Akhta Muhammad, was an Ahmadzai Pushtun from

Paktya, born in Kabul in 1947. His father was at one time consul and commercial attach in Peshawar, where Najibullah learned Urdu and English. After
attending the Habibia School, he studied medicine and joined Karmals group.
He took part in student demonstrations and was even briefly imprisoned on two
occasions. He married Fatima, a niece of Dr Anas (formerly Zahir Shahs Minister of Culture) and a member of the royal clan. It is likely that he became a KGB
agent at the close of the 1960s, and he joined the Central Committee at the time
of reunification in 1977. In 1978 he was put in charge of the Kabul Committee,
and thena victim of the purge on the Parchamishe was named as ambassador to Teheran, and afterwards to Yugoslavia. He later became head of the KHAD,
and after appointment as a member of the Politburo in 1981 became its secretary-general at the end of 1985, handing over the running of the KHAD to his
deputy Yaqubi.

194

The Kabul Regime

of the regime he retained the support of certain eastern Pushtun

tribes including his own tribe the Ahmadzai.
Najibullahs ascent was obvious by the end of 1985. On 11 January 1986 the KHAD was elevated to the status of a ministry, as a sign
of the growing importance accorded to an apparatus of which Najibullah was in practice to keep control. After a month-long private
stay in the Soviet Union, Babrak Karmal resigned his post as secretary-general of the party on 4 May 1986, to be succeeded by Najibullah. On 30 September 1987 Najibullah was appointed to the
presidency of the Revolutionary Council. On 2829 November
1987 the Loya Jirga approved the new constitution and the appointment of Najibullah as head of state. He would in due course combine the position of head of state with that of president of the
Council of Ministers and of the Higher Defence Council. This accumulation of power was a symbol not so much of the concentration of power in his hands as of his isolation. Nevertheless he took
care to place his supporters in key posts. Gholam Faruq Yaqubi, his
deputy at the KHAD, went on to the Central Committee, and
Suleiman Laeq, who had worked a great deal with KHAD in his position as Minister for Tribes and Nationalities, became an alternate
member of the Central Committee. Najibullah was also able to rely
on the support of Parchamis opposed to KarmalKawiani and
Wakilas well as several Khalqis. In addition he mobilised his network of provincial officials and notables from the old regime who
had been appointed under the policy of National Reconciliation.
For instance, in 1990 Khaleq Yar, Dauds former Minister of Finance
and governor of Herat, was appointed Prime Minister in succession
to Sultan Ali Keshmand.
The purge of Karmals supporters was spectacular. Anahita Ratebzad was stripped of her position as president of the Womens Democratic Organisation on 6 August 1986,and on 21 November Panjshiri,
Danesh and Qader left the Politburo. To strengthen his grip on the
Central Committee Najibullah increased its membership. Some of
Karmals supporters reacted violently to their eclipse. When Najibullah was appointed, pro-Karmal demonstrations took place, especially in Kabul and Kandahar.On 8 May 1987,a few days after Karmals
departure to exile in Moscow, his supporters organised a demonstration outside the Ministry of the Interior. There were also three
bomb explosions at Mikrorayon, also perhaps their responsibility.

Preparations for the retreat

195

The announcement of the Soviet departure inaugurated a period

of unanimity within the Party, which mobilised to ensure its own
survival, especially during the mujahidin attack at Jalalabad in the
spring of 1989. However, tensions continued to run high and, taking
advantage of the progressive withdrawal of the Soviets, the Khalqis
several times tried to recover control, making use of their positions
in the Defence and Interior Ministries. Two or three attempts were
made during the course of 1988. 5 March 1988 saw the start of the
trial of 128 conspirators, among whom were officers and Party members. On 14 May 1988 twenty-three of the twenty-nine provincial
governors were replaced in order to forestall a coup. In consequence
Najibullah sacked the Khalqis most involved. On 19 October 1988
Saleh Muhammad Zeary, leader of the Khalq, left the Politburo. In
November Gulabzoy, who was suspected of organising a coup, was
replaced as Minister of the Interior by Watanjar. In compensation
Najibullah attempted to gratify the Khalq by appointing Tanai, the
Khalqi party chairman, as Minister of Defence in August 1988. In
the same spirit four Khalqis out of nineteen new members were
appointed to the Central Committee in October 1989.
Najibullah, who had become party chief and president entirely at
the wish of the Soviets, never consolidated enough support within
the party: his supporters were always fewer in number than those of
Karmal, while the Khalqis never accepted him, even though he was a
Pushtun. This initial weakness obliged him to rely to a great degree
on the KHAD, which he continued to manage through his deputy
Yaqubi, while in the interests of maintaining his authority he also
played on the antagonism between the Parchamis and the Khalqis.
These machinations led to a lessening of the Khalq-Parcham split, as
hostile groups sprang up.
The policy of National Reconciliation. The expression national reconciliation (translated into Persian as ashti-yi melli) was employed by
Gorbachev in his Vladivostok speech of July 1986 in relation to crisis
situations in the Third World in general and in particular in Afghanistan. From the end of 1985 Karmal had made use of the expression
naosazi to denote a variety of perestroka. The appointment of Najibullah gave a new impetus to reform and implied renunciation of
the military solution. From now on the accent would be placed on
an indirect strategy.

196

The Kabul Regime

In the event the major military operations conducted by the

Soviet-Afghan command had, more than anything else, given a political boost to the guerrillas. As was often the case, the apparent disparity between the forces led the government to neglect the more
purely political aspects of the struggle, a point proved by the example
of Masud. Had offensives not been regularly launched against Panjshir, whose strategic importance was sometimes overestimated,Masud
could never have achieved the level of recognition he enjoyed. Similarly, attempts to break the position of Ismail Khan in Herat by military means reinforced his legitimacy, solidifying around him the
unity of the mujahidin, who were mostly divided. Military pressure,
which tended to prompt mobilisation, in fact worked against the
Kabul government to the extent that the complete annihilation of a
group of combatants was always extremely difficult.
At first the significance of the policy of National Reconciliation
was misunderstood by observers and by the commanders, who saw
in it only a propaganda exercise. Their incredulity, which was real,
marred the effect of the declaration of the policy, with which the
government had counted on dropping a political bombshell. Nevertheless, the following developments were spectacular: the government
set about nothing less than the deconstruction of the communist
regime. The issue concerned not so much the sincerity of these
changes as their political effects, which would in the end be substantial since they expressed new power relationships. The effects of the
policy of National Reconciliation would only truly be felt after the
withdrawal.
After several last thrusts in 1986, mainly to ensure the security of
the towns, the Kabul government proclaimed the policy of National
Reconciliation in January 1987. Following the plenary meeting of
the party and the declaration of a unilateral cease-fire from 15 January, the government asked the leaders of the armed opposition to
negotiate. In addition a general amnesty was promulgated on 24 January for Afghans engaged in anti-government activities. Finally, on
8 March 1988 Najibullah commissioned the Paderwatan to establish
contacts with opposition groups, which were acknowledged as future
partners.
The lifting of restrictions on commerce26 and the new attitude
in public, at leastof respect for religion brought great satisfaction
26

In addition, a decree of June 1987 raised to 20 hectares per person the initial

Preparations for the retreat

197

to the merchants and officials, who were alarmed by the dissension

among the mujahidin which seemed to increase in proportion to their
closeness to victory. The language of the authorities, as employed
for example by provincial governors at ceremonies of allegiance,
centred around two themes: first, the affirmation of the Islamic character of the government, which respected and protected the faith;
and second, an appeal to patriotic sentiment against Pakistani interference, which some Afghans actually accepted but with reluctance.
The policy of National Reconciliation also had an ethnic dimension. Najibullah, who was able to take advantage of his Pushtun
legitimacy, attempted to oust the Tajiks from the Party leadership.
However, at the same time his link with the Hazaras was maintained.
In 1986 a Hazara Nationality Council (Shura-yi Melli) was established under the supervision of the KHADs 12th Bureau. In addition a Hazara jirga, the Jirga-yi Sarasari-yi Milliat-i Hazara (Central
council for the Hazara nationality), was held in Kabul in September
198727 while Sultan Ali Keshmand (himself a Hazara) was Prime
Minister. Sultan Ali Keshmand spoke in terms with markedly nationalist overtones on the occasion of the inauguration of the Markaz-i
Insijam Omur-i Milliat-i Hazara (Centre for the Coordination of
the Affairs of the Hazara Nationality) in the summer of 1989.28 He
also set up a Hazara regiment, the 520th infantry, in December 1986.
In addition, the new constitution of 1987 recognised the equality of
the tribes and nationalities.
In the event, on 30 November 1987, a Loya Jirga approved the
new constitution, with a general outline that remained close to that
of 1964. A Loya Jirga made up of tribal chiefs, intellectuals, notables et
al. was to designate the president of the republic for a term of seven
years. The president had a right of veto over parliamentary decisions,
which could only be overruled by a two thirds majority. Parliament
was elected for six years, while the senate was half appointed by the
President and half elected. Finally, there were significant symbolic
concessions: Islam once again became the religion of the state, and
Afghanistan adopted a flag in which the colour green predominated.
ceiling of 6 hectares, which was tantamount to recognition of the failure of
agrarian reform.
27 Kabul New Times, 24 September 1987.
28 S. A. Mousavi, op. cit., p. 177.

198

The Kabul Regime

From the Proximity talks to the Geneva accords

Diplomacy was not of major importance in bringing about the Soviet withdrawal. In the first instance, the negotiations were a means
for Moscow to put a diplomatic face on its military defeat, since it
had failed to produce a political formula to end the civil war. The
talks were therefore only marginally useful, but they did allow the
UN secretary-general to strengthen his position. In the event the
settlement of the international aspects of the Afghan conflict marked
the end of the Cold War and the emergence of new international
relationships. The role of the UN at this time changed radically, it
appears, and the Afghan issue was one of the first where the organisation, no longer paralysed by the East-West confrontation, made a
bid to embark on autonomous action.29
From the earliest months after the Soviet invasion, recriminations
and proposals for negotiations proliferated. Demands for unconditional withdrawal were issued in January 1980 by a special session of
the UN, then several weeks later by the Organisation of the Islamic
Conference (OIC). At the end of February 1980 Pakistan proposed
a troika, to include the Foreign Ministers of Pakistan and Iran and
the secretary-general of the OIC, to serve as an intermediary between the mujahidin and the Kabul government, which however refused to entertain even indirect contacts with the parties in Peshawar.
The Iranians refused to negotiate in the absence of the mujahidin, a
factor which, when added to their bad relations with the United
States, largely excluded them from subsequent negotiations.
In its turn Kabul took the initiative and on 14 May 1980 put forward a preliminary proposal to open bilateral talks between Afghanistan and its neighbours Iran and Pakistan relating to a mutual nonintervention pact, to be guaranteed by the United States. In Kabuls
eyes the Soviet presence was non-negotiable, since it related to Afghanistans sovereignty, and negotiations should not be seen as a way to
internationalise the crisis.
Faced with this impasse, the UN adopted a policy of indirect rapprochement. In February 1981 the secretary-general Kurt Waldheim
appointed Xavier Perez de Cuellar as his personal representative for
29

For the text of the agreements, see Diego Cordovez, Selig H. Harrison, Out of
Afghanistan: the inside story of the Soviet withdrawal, New York: Oxford University
Press, 1995.

From the Proximity talks to the Geneva accords

199

Afghanistan. Perez de Cuellar shuttled between Islamabad and Kabul,

but their positions appeared irreconcilable: the Kabul regimes demand was for direct negotiations with the Pakistan government,
while Pakistan insisted on self-determination for the Afghan people. After Perez de Cuellars election as UN secretary-general, Diego
Cordovez became the new special representative and succeeded in
initiating indirect negotiation in Geneva, beginning in June 1982.
With the mediation of UN negotiators these brought together Pakistan and the Kabul government of Afghanistan, in the presence of
Soviet and American observers. The talks were organised around four
issues: a bilateral accord between Kabul and Islamabad on reciprocal
non-intervention, international guarantees, a bilateral agreement on
the return of refugees and a plan for the withdrawal of Soviet troops.
By mutual agreement between Pakistan and the Soviets, the mujahidin were not represented. This first series of negotiations ended in
failure, since the Pakistanis refused to recognise the Kabul government, which in its turn did not wish to discuss a withdrawal.
Andropovs rise to power between 1982 and 1984 was marked by
a relative freedom for diplomatic activity, but without concrete results, due to his rapid departure. At the negotiations of April 1983 in
Geneva, Kabul asked for the closure of the camps in Pakistan and the
end of support for the resistance; while Islamabad wanted a firm
commitment over the duration of the withdrawal, which the Soviets
rejected. Nevertheless Diego Cordovez showed optimism and believed that agreement was close, although the Soviet position later
hardened with the demand that the security of Kabul should be fully
guaranteed before any withdrawal. It appeared in fact that Andropovs
influence had begun to wane because of his ill-health. The Soviets
returned to their initial position: there could be no discussion of
Soviet troops in the course of Afghan-Pakistani negotiations, and the
only role for the Americans would be as guarantors of non-intervention. 1984 was a year without results, although the fourth round
of the negotiations in August was the first to be held on the basis of
proximitytalks were no longer indirect. Henceforth meetings
were held at a venue which accommodated both Pakistani and
Afghan delegations.
In the spring of 1985 the advent of Gorbachev re-launched the
process of negotiation, and in 1986 agreement was reached on three

200

The Kabul Regime

points: the refugees, non-intervention and guarantees. Only the

schedule for the withdrawal of Soviet troops remained as an important point of disagreement. The Pakistanis wanted a rapid withdrawal, over three or four months, while the Soviets and, even
more, the Kabul regime wanted a longer period. Still, the positions
were moving closer. At the round of talks in March 1987 Pakistan
asked for withdrawal over seven months as against eighteen months
called for by the Kabul regimeby the end of 1987 the demands
had become eight months and sixteen months respectively. The issue
of a transitional government was also put on the table, since the end
of the negotiations seemed to be approaching.
Gorbachev, in his Vladivostok speech of 28 July 1987, proposed
the formation of transitional government including the mujahidin.
By this time Diego Cordovez had already tried to negotiate the formation of a provisional government including the king, Zahir Shah,
but he ran into opposition from the Islamist parties, Pakistan and the
United States, which inclined towards a military solution. In a BBC
interview on 11 May 1987, Zahir Shah said he was ready to take up
his responsibilities if the majority asked him to. On 10 June Najibullah said he was also willing to negotiate with the king, but on 16
June Zahir Shah gave his final refusal as a consequence of the negative reactions from the mujahidin.
The Soviets finally unblocked the negotiations by renouncing the
objective of a political solution as a condition of withdrawal, but
maintained their demand for negative symmetry, by which they
meant a ban on arming the adversaries within the country. On 8 February 1988 Gorbachev made a speech removing the last obstacle
with the proposal of a withdrawal within ten months, while setting a
deadline of 15 March for an agreement to be reached. In February
1988 Zia ul-Haq, taken by surprise, hastily set up an interim mujahidin government. On 2 March 1988 the final round of negotiations
began in Geneva. Negative symmetry was finally abandoned after
American hesitations, and each side continued to arm its allies.
On 14 April 1988 the accords between Pakistan and Afghanistan
were signed in Geneva, with an American-Soviet guarantee.30 They
provided for a Soviet withdrawal in early 1989, but were less than
explicit about the political future of the country. On 12 May the
30

Diego Cordovez, Selig S. Harrison, op. cit., pp. 389 ff.

The post-withdrawal survival of the regime

201

withdrawal began at Jalalabad and Kandahar, then on 26 July at

Shindand. The first withdrawal from Kabul took place in August,
by which time, according to Gromov the commander in chief of
the armed forces, no more than half the Soviet forces remained,
concentrated into the six provinces of Kabul, Baghlan, Samangan,
Kapisa, Farah and Herat. The Soviets abided by their commitments
and completed their withdrawal on 15 February 1989. On the whole
the mujahidin did not attack the Soviet forces as they pulled back,
choosing rather to occupy the positions they left behind.
In the end responsibility for the application of the Geneva Accords
was assigned to Diego Cordovez, who was replaced at the end of his
mandate as the secretary-generals representative by a Cypriot, Benon
Sevan. However, many believed that the fall of Kabul was now only a
matter of months away, while the principal problem had become the
reconstruction of the country, entrusted to a UN special agency
UNOCA (United Nations Office for the Coordination of Humanitarian and Economic Assistance to Afghanistan). On 12 May Sadruddin Aga Khan, the former UN High Commissioner for Refugees
and an influential member of the Ismaili community, was appointed
UN coordinator for economic and humanitarian programmes in
Afghanistan, at the head of UNOCA.

The post-withdrawal survival of the regime

Up to the coup of 1992 the situation appeared deadlocked: the regime did not fall, and the mujahidin proclaimed themselves unable
either to push through a military victory or to formulate a credible
political alternative. In the event the ongoing Soviet withdrawal had
had a number of consequences. The spirit of jihad began to flag, while
the Kabul regime was no longer so visibly controlled by Moscow. In
consequence the language of National Reconciliation came to seem
more credible. In addition the regimes military redeployment, departing from a significant number of outposts and small towns, was
not without advantages: a more concentrated security system was
stronger. The regime also had the benefit of Soviet military aid, while
the West had scaled down its assistance to the mujahidin. In the end
the failure of the attack on Jalalabad in the spring of 1989 brought
about a change of view: a mujahidin victory began to seem improbable in the short run.

202

The Kabul Regime

Consequently, the policy of National Reconciliation came fully

into play. Wherever where the government refrained from launching
attacks, de facto cease-fires and more or less peaceful coexistence began to take hold. After the reverses of 19889 military deployment
began to take a more permanent form, while the morale of the mujahidin was damaged by their internal dissent. In the post-Soviet period, the government did not so much attempt to rally the mujahidin
as to strike agreements with them in the aim of keeping clashes
down to an acceptable level. UNOCA tended in the same direction
with the specification of cease-fire zones.
One example will serve to illustrate the success of this process. In
Kandahar, the major Pushtun city of the south, some particularly
aggressive guerrilla groups brought the fighting into the heart of the
town. Here violence against the Soviet occupiers had always been
spontaneous and immediate. In 1980 four Soviet soldiers, on duty in
the heart of the city, were beaten to death by the population in an
unplanned attack. The soldiers did not exaggerate when they used
to say that a posting to Kandahar was a punishment. The governments attempts to dislodge the guerrillas from the outskirts of the
city never succeeded, to the extent that several observers predicted
its imminent fall after the Soviet withdrawal. However, under the
leadership of the governor, Nur ul-Haq Ulumi, a Parchami from an
important local family, the policy of National Reconciliation began
to bear fruit, meanwhile aggravating the divisions among the mujahidin. In exchange for financial and other incentivese.g. permission for the muijahidin to come into the town without their weapons
to visit their familiesthe parties scaled down their attacks, a particular result of which was that any coordination between muijahidin
groups for a general attack on the town became impossible. Pakistani
initiatives to unite the mujahidin were to fail, although some parties,
including Hekmatyars and Khaless Hezb-i Islami, refused to compromise with the government, in order to maintain their access to
Pakistani funding.31
In a bid to gain wider support, the government organised a not
altogether honest reorganisation of political life. Najibullah wanted
31

This split also corresponded to tribal considerations: the Ghilzai, who were
dominant in the Khales and Hekmatyar factions of the Hezb-i Islami, rejected
the agreement, which was accepted by the Alizai and Barakzai Durrani who belonged to the Jamiyat-i Islami and the Mahaz.

The post-withdrawal survival of the regime

203

to set up a new party in order to enhance his credibility as a participant in a putative coalition government with the mujahidin. On
25 Feburary 1990 he called on the party, the Hezb-i Demokratik-i
Khalq-i Afghanistan, to step down from its controlling position. A
new movemement, the Hezb-i Watan (Party of the Nation) was
officially created on 27 June 1990. As a forewarning of this development, the government press had maintained a discreet silence concerning the Hezb-i Demokratik-i Khalq-i Afghanistan, from the
time of the Soviet withdrawal up to the victory at Jalalabad. In fact
the Hezb-i Watan inaugurated no profound change since Najibullah was able neither to broaden his political base nor to appoint
new officials. The executive committee of the Hezb-i Watan was
made up of thirteen members, of whom eleven had belonged to the
Central Committee of the Hezb-i Demokratik-i Khalq-i Afghanistan in 1980; 80% were Parchamis, with a balance between proKarmal and pro-Najibullah factions.
In addition multi-party pluralism was largely an illusion, although
the Hezb-i Watan ceased to play a leading role, it remained the guarantor of the policy of National Reconciliation. In the event Najibullah authorised political parties under the decree of 11 September
1990, but it was Department 7 of the WAD (successor to the KHAD)
which clandestinely organised such opposition forces as the Hezb-i
Deqhanan (Peasants Party) and the Hezb-i Islami (the Party of
Islam: a Hazara nationalist splinter group based in Jaghori), as well as
the independent newspapers such as Akhbar-i Hafta (The Weekly
News). In the Senate twenty-five out of sixty-two seats were assigned
to the opposition.32
In spite of these developments the dangers which would prove
fatal for the regime were contained already in embryo in the policy
followed by the government from 1987. Although the deployment
of the military was generally unchallenged, declarations of loyalty remained sparse, with most of the commanders preferring to remain
autonomous, outside ad hoc agreements with the government. The
policy of National Reconciliation was extremely fragile, as the regime managed its internal divisions badly and progressively lost control of the periphery.
With the period of the withdrawal over, tensions within the government worsened. In March 1990 the Khalqis attempted one final
32

Asta Olesen, op. cit., p. 267.

204

The Kabul Regime

coup, with the help of Hezb-i Islami. The earliest contacts between
the Khalqi tendency and Hezb-i Islami probably took place in early
1990, when Hekmatyar met Khan Abdul Wali Khan, the leader of
the Pathan nationalist movement, the Awami National Party. For
the Khalqis, who opposed Najibullah and his strategy of National
Reconciliation, the coup represented the possibility of acquiring
power while at the same time achieving Islamic legitimacy through
the alliance with Hezb-i Islami, which had already opened its ranks,
for reasons of communal solidarity, to a number of Khalqi deserters.
On 6 March Tanai, the Minister of Defence, launched the coup
from the air base at Bagram, to the north of Kabul. The government,
forewarned of the attempt, kept control of the situation, thanks especially to the KHAD troops who guarded the capital. Najibullah, who
took refuge in the bunker at the Presidency, contacted the provincial
governors, who remained loyal to him, one by one. On the following day, 7 March, Tanai acknowledged his failure and joined the
Hezb-i Islami in Pakistan. On 8 March Bagram air base was recovered from the rebels. On 11 March an anti-Khalqi purge began with
the dismissal of five members of the Politburo and seven members
of the Defence Council. According to the regime, 623 arrests were
made, but the ousting of the Khalqis did not of itself enable Najibullah to reassert his power. He was no longer able to exploit the Khalqis as a counterbalance to Karmals supporters, who would play a
decisive role in his fall in the spring of 1992.
In addition, despite the scale of Soviet aid, which amounted to
$14.2 billion in the year of the withdrawal,33 the financial and political burden of the policy of National Reconciliation progressively
undermined the infrastructure of the Afghan state. To pay the militias the government adopted an inflationary policy which resulted in
a spectacular increase in the money supply, from 112.5 to 222.7 billions of Afghanis between 1988 and 1990.34 However, even more than
the financial burden, the loss of control over the provinces threatened the regime.
33

See Anthony Arnold, op. cit., p. 185. Thirty MiG 27s were delivered in October
1988, as well as SCUD missiles, used against the outskirts of the capital. The Politburo archives show that up to the end the Soviet Union believed it could keep
Najibullahs regime in place (Allan, op. cit.).
34 See Barnett R. Rubin, The Fragmentation of Afghanistan, New Haven and London: Yale University Press, 2002, p. 163.

The post-withdrawal survival of the regime

205

The authorities increasingly relied on the militias and established

with them a relationship of clientelism comparable to that which
existed between the parties and the commanders. The army banked
on the militias to ensure the security of the towns, but also increasingly to carry out major offensives. The militiamen of Abdul Rashid
Dostam, originally from the north-west, fought throughout Afghanistan, which put them in a position of competition with the regular
army. These men were systematically involved in the toughest fighting, since they were volunteers, sometimes paid by results, and were
not liable to desert to the mujahidin, from whom they expected no
mercy. This militia, officially the 53rd division, comprised 2040,000
men and answered directly to the president, bypassing the Ministry
of Defence.
The militias effectiveness presupposed local roots, which at first
the government encouraged, but it would in due course have negative consequences. In some provinces the chronic problems of the
regular army left the militiamen as the predominant local military
force. Their bargaining power grew substantially in relation to the
state, which was obliged to pay out in order to maintain the loyalty
of groups which were nothing less than private armies. This was true
for the former mujahidin who had preserved their autonomy, but also
for militias set up by the government itself, such as that of Rashid
Dostam, who were quite capable either of refusing to fight, or of
pursuing local policies sometimes running counter to those of the
central government. Disobedience was rife, and in many places the
authority of the government did not run; it could no longer keep
order in certain suburbs of Herat, where the militia held travellers to
ransom and settled their differences with the Kalashnikov. In the autumn of 1991 clashes between the militias and the regular army were
an indication of the decline of the situation in the province of Herat.
The same happened in the north where Dostams authority took
precedence over Kabul officials. The government thus lost the credit
it had gained through its policy of moderation and its discontinuation of military operations. Random violence became prevalent, and
that of the militias was as arbitrary as that of some of the mujahidin.
The central government had few means at its disposal to exert pressure: the secret servicesthe KHAD and then WADwere themselves powerless to control the militias which had a real esprit de corps,

206

The Kabul Regime

often based on communal solidarity. In the Uzbek provinces of the

northwest the most powerful militia leaders had taken control of the
Party apparatus and the army. In the face of the militias growing
independence, the central state became unable to control its own
security apparatus.
The weakness of the centre relative to the periphery called into
question the functionality and even the existence of the state. Traditionally power in Afghanistan was exercised from the centre. The
appointment of officials who were generally strangers in provinces
where they served for limited periods safeguarded the control of the
central state, and precluded to a certain extent the establishment of
excessively powerful local authorities. However, for practical reasons
the state had fostered the autonomy of the militias, bringing into being a balance of power increasingly unfavourable to the centre. The
achievements of pacification were in the end too costly for an exhausted state, which could not survive the disappearance of its Soviet
protector. The policy of the government thus led to a gradual loss of
control of the periphery, which facilitated the coup of 1992.

6. The Guerrillas
An account of events from the point of view of the mujahidin presents more of a problem than their examination from the standpoint
of the government. The guerrillas existed in a wide variety of local
circumstances, and the Shiite and Sunni groups developed in different ways, while the priorities of the commanders on the ground
diverged from those of the leaderships in exile. A common thread
nevertheless emerges: the tendency towards professionalisation among
the mujahidin, which emerged in parallel with the prevalence of
institutional styles of leadership over the patrimonial model. The
changing political dispositions on the ground, in addition to interparty relationships and developments subsequent to the Soviet withdrawal, will be analysed below.

The professionalisation of the fighters

At the outset the Afghan guerrillas lacked mobility, while in many places the tribal model of warfare persisted. The mujahidin were strongly
identified with particular territories, which accounts for the problems they encountered in operating outside their home regions. The
guerrillas did not put a high value on secrecy or mobility, since the
government was well aware of their positions on the ground, but
placed their faith in the difficulty of access to their mountains and
high valleys, as well as in the governments inertia. Ambushes involving more than a few dozen men were rarely mounted, and these were
often carried out in the same places,with little attempt at surprise.With
the exception of a few commanders such as Masud, attacks against
government positions generally aimed not at capturing themat least
not before 1988but at keeping up pressure on the government
while providing an opportunity for individual fighters to display
their valour. Overall, the support of the population, in addition to
207

208

The Guerrillas

the topography, compensated for any inadequacies, which in any

case the Soviet-Afghan forces did not fully exploit.
However, over the years the circumstances of the resistance changed, as they became better armed and trained. In the early months of
the rebellion the mujahidin fought with old English Lee-Enfield rifles,
then came the arrival from Pakistan of Kalashnikovs, which remained
their standard weapon throughout the war. At the outset the guerrillas were entirely without heavy arms, but they gradually acquired
suitable anti-tank and anti-aircraft weapons. Plastic mines were effective, since the Soviets had nothing with which to detect them, but
the antiquated RPG7 anti-tank weapons were responsible for most
of the Soviet-Afghan losses. From early 1986 onwards the mujahidin
obtained around 200 Blowpipe anti-aircraft weapons, and then about
1,000 Stingers between 1986 and 1990. The efficiency of these weapons has been exaggerated, however, and they certainly did not alter
the course of the war; Mark Urban has estimated Soviet-Afghan
losses due to these missiles at ninety helicopters and planes, of which
1520 were combat aircrafti.e. less than 20% of the total losses of
some 500 units up to the Soviet withdrawal.1 Communications were
also a weak point for the mujahidin: during the 1982 offensive against
Panjshir, Masud used handwritten messages to communicate with
his troops. Hence the relatively sophisticated radios supplied by
western agencies greatly assisted coordination between groups.
The acquisition of better weapons was directly linked to the provision of aid by western and Arab countries. Between 1980 and early
1992, when military aid officially ceased, the United States gave
$23 billion to the mujahidin, to which can be added a similar sum
from the Arab countries. However, President Reagan had been highly
cautious during his first term, in spite of his anti-communist rhetoric.
For 1983 the CIA asked for a budget of $30 million for the mujahidin
(compared to $24 million for the Contras in Honduras). Charles Wilson, a Democratic senator, obtained an additional $40 million and
some Swiss anti-tank guns.In 1985 a distinct turning-point was reached,
with a budget of $200 million, which was to reach $280 million by
1989. The other western countries, especially France and Britain,
did not publicise the level of their military aid to the mujahidin, but
1

Mark Urban, op. cit., p. 296. It is possible that these estimates are too high.

The professionalisation of the fighters

209

fighters were trained in Europe and Pakistan, while equipment such

as radios, missiles and anti-tank weapons was distributed to the parties. Around 80,000 mujahidin were reported to have been trained in
Pakistan, mainly by Pakistani officers, as well asmore unusually
by Americans.2
However, the improvements in armaments only achieved their effect through the introduction of more elaborate organisation, which
brought with it an increasing level of integration and coordination.
This phenomenon took place with greater or lesser rapidity in different regions. From the earliest years Masud laid the foundations of
a permanent army, while in some distant valleys the mujahidin continued as they had started, as the inhabitants of a village under arms.
These developments were in part related to the patterns of rivalry
between the commanders.
Relationships between commanders were generally characterised
by competition arising out of personal rivalries, logistical issues, control over strategic positions, and humanitarian aid among other considerations. In addition, mobilisations had often been conducted on
the basis of qowms, so that commanders sometimes inherited longstanding communal hostilities. For instance, Pushtuns who had recently settled in the north were sometimes subjected to attacks by
local populations wanting to recover their land. Other cases had a
more social explanation, as in Laghman where the conflict between
Hezb-i Islami and Jamiyat-i Islami could be interpreted as a confrontation between great landowners and tenant farmers.
The reality was that no regulatory mechanism existed which could
enforce an agreed standard of behaviour and underwrite transactions. The parties, who were themselves in a competitive situation,
were hardly in a position to impose discipline on the commanders.
Rivalry frequently spilled over into armed clashes. The jirgas of some
tribal regions, and for a while the ulema of Kandahar, were able to
2

For an assessment of the levels of aid, see James Rupert, Afghanistans slide towards
civil war, World Policy Journal, vol. XI, no. 4, fall 1989; Kirsten Lundberg, Politics of
a Covert Action: the US, the Mujahedin and the Stinger missile, Harvard Intelligence
and Policy Project, Kennedy School of Government, 1999; Alan Kuperman, The
Stinger Missile and U.S. intervention in Afghanistan, Political Science Quarterly,
vol. 114, no. 2 summer 1999, pp. 21963. On the Arab aid see Muhammad
Yousaf and Mark Adkin, The Bear Trap: Afghanistans Untold Story, London: Leo
Cooper, 1992, p. 77.

210

The Guerrillas

play a regulatory role, but such instances frequently came after the
outbreak of conflicts, with the aim of reconciliation. A considerable
number of commanders were pressured into switching their allegiance, driven into exile or assassinated. Mistrust was the most striking characteristic in the day-to-day attitudes of the commanders.
Cooperation between groups was restricted to ad hoc operations,
and the primary objective of each commander was to secure his
own autonomy. Shura councils were rarely effective, and only more
closely-knit systems, where the commanders were more narrowly
constrained by some form of organisation, were able to implement
cooperation. Such were Masuds Shura-yi Nazar (Council of Oversight) and in the west the Emarat of Ismail Khan.
Conflicts between commanders were far from rare even before
1992, and sometimes occurred even among those of the same party,
especially one whose dominance in a particular region concealed
the customary local differences. In 1988, in the west, clashes between
the two commanders of the Harakat-i Enqelab in Helmand, mullah
Nasim and ras Abdul Wahid, left hundreds dead and intermittently
interrupted the traffic of convoys between Pakistan and the Herat
region. Similarly, in Badakhshan battles between Najmuddin and
Basir Khan divided the province politically and to some extent economically over a period of years.
Tashqurghan, on the road between Kabul and Mazar-i Sharif,
provides a good example of such situations. In the 1980s clashes between commanders in this small town were endemic. The bazaar
was split between different commanders, and no mujahid was able to
enter the territory of a different party. In an overt manifestation of
this hostility, the commanders built fortified towers with firing points
and walls two metres thick, sufficient to withstand rocket-launchers.
Movement in the bazaar at night was dangerous because of the possibility of ambush, while political assassinations were ordinary occurrences. Violent confrontations emerged even within parties. In
contrast to other regions, there were no social or historical factors
here which offered a sufficient explanation of these antagonisms.3
The social origin of the principal commanders was not a divisive
factor, since in essence the rural notables, the bays, of Tashqurghan
3

For a portrayal of the bazaar in the pre-war period see Pierre Centlivres, Un bazar
dAsie centrale, Wiesbaden: Ludwig Riechert Verlag, 1972.

The professionalisation of the fighters

211

had retained their power. There was no confrontation with the new
lites, although the rise of educated individuals provided some explanation for the tensions. The ulema had risen in influence, here as
elsewhere, but they were still excluded from the political process.
Nor did the existence of several communities explain the clashes.
The town was predominantly Tajik, with significant minorities of
Uzbeks and Pushtuns, but the commanders were not noticeably
identified with any ethnic group. The case of Tashqurghan shows
that clashes between commanders cannot always be explained by
communal tensions arising from social structures; here rivalry between commanders was the decisive factor. Politics was being conducted for its own sake, and the struggle for power wasat least
sometimesits own justification.
Institutional systems were by nature more likely to prevail in the
competitive struggle between commanders because of their superior organisation and greater ability to attract and make use of foreign aid. Masud and Ismail Khan progressively expanded their spheres
of influence, eliminating or absorbing the smaller groups. The emergence of organisations on a provincial scale was the predominant
tendency in the 1980s. This process of amalgamation brought about
a transformation of structures founded on networks into territorial
structures. At the outset of the war, commanders had generally based
their power on individuals linked by solidarity related to a qowm.
Later, however, the influential commanders were those who controlled a territory. The appearance of territories in due course implied the demarcation of frontiers; real political divisions on the
ground were beginning to emerge.
The territorialisation of politics and the concentration of power
were the result of the predominance of institutional systemseither
clerical or state systems according to their location. In the north
members of the educated class set up the most effective organisations, but the ulema held sway over Hazarajat and the tribal areas of
the south, and the educated class made no inroads into the tribal regions. For example Amin Wardak, himself an educated man, was not
able to extend his territory since his base was in Wardak, a Pushtun
tribal zone, where he faced strong opposition as he attempted to expand beyond his initial network. Up to 1994, therefore, the Pushtun
south remained much more fragmented politically than the north,

212

The Guerrillas

so that the ulema achieved real predominance only in the province

of Ghazni and in Hazarajat.
These developments had their impact on the relationship between the mujahidin and the population, in which a watershed was
reached in the mid-1980s. External aid permitted the mujahidin
increasing independence from the population, especially from the
point of view of finance. The mujahidin tended more to gather in
their own marakiz (bases) which gave concrete expression to the distinction between military and civil territory. In the early days the
mujahidin generally spent several months at these bases, returning
afterwards to cultivate their lands or going to work abroad. However,
the destruction of the irrigation systems and the fact that battles had
often been fought on the best landsnear villages or on the plains
left many individuals unemployed and available to become full-time
fighters. Being a mujahid therefore became a career. The mujahidin
henceforward stood apart from the population,from whom they were
no longer indistinguishable as they had been when the war began.

The evolution of the political map

The fighting, the allocation of aid and rivalries between commanders were all factors which affected the geographical distribution of
the parties. Our analysis of political changes on the ground includes
the elimination of the lay parties, who were the heirs of the pre-war
nationalist and leftist movements,4 and an examination of the Sunni
and Shiite parties.
The elimination of the lay parties. The leftist parties in general had no
social base in the countryside. They had recruited in the urban environment and been unable to adapt themselves to the new political
4

The nationalist movement Afghan Mellat had no presence on the ground. After
1979 its development took place mainly outside Afghan territory. For example,
Dr Mohammad Amin Wakman was resident in the United States at the moment
of his election to the presidency of the party at a meeting in Peshawar on 89
March 1990. Afghan Mellats claim to a greater Pushtunistan caused it to lose all
aid from Pakistan, and doomed its attempts to establish bases in Afghanistan,
though it later moderated its irredentist claims and tried to widen its ethnic base.
Its main enemy was Hezb-i Islami, which it blamed for the assassination of several
of its members in Peshawar, in particular Dr Shewigal on 27 March 1990 and Taj
Muhammad Khan in September 1991.

The evolution of the political map

213

realities. Khalqi policies were discredited, together with progressive ideas such as agrarian reform, the spread of literacy and the liberation of women, all of which had been favoured by a faction
drawn from the urban and educated population. Although leftist
militants, often the sons of notables, were represented in certain
areas, clashes with other parties and a lack of support from Pakistan
led to their rapid marginalisation. In addition the government had
been able to repress or bring over to its side the largely urban movements, as was shown by the arrest of several hundred leftists in Kabul
in early October 1981.
The pre-war Maoist movement, Shola-yi Jawid, had disappeared
in the political turmoil of the 1970s, when its principal activists were
killed or fled.5 New organisations appeared in 1979, of which the
leading example was the SAMA, Sazman-i Azadibakhsh-i Mardomi Afghanistan (Organisation for the Liberation of the Afghan People). Other groups of former Maoists, under different titles, kept up a
limited level of activity in the interior of the country. Their activities
fell into two categories: on the one hand there were autonomous
Maoist commanders, and on the other they made attempts to infiltrate other parties.
At the beginning of the war the leader of SAMA, Majid Kalakani,6
led a Front in the Shamali plain, close to Kabul. Kalakani had been a
Maoist militant from the 1960s and was one of the founding members of SAMA. On 27 February 1980 he was arrested after fighting
the regime for several months from a base in his village. He was imprisoned and in June 1980 executed. Under heavy pressure from
5

For instance Dr Abdul Rahim Mahmudi, the brother of the founder of Shola-yi
Jawid, was able to escape with his brother Adi when some of his relatives were
murdered.
6 Majid Kalakan was born in 1939 and came from the village of Kalakan in Kohdaman, in the Shamali plain. This was also the birthplace of Habibullah Kalakani,
bacha-yi saqao, raising the question whether there was a tradition there of opposition to the authorities. His father and grandfather are reported to have been
hanged in 1945 during Hashems regency. The family was exiled to Kandahar
until 1953, when Kalakani returned to Kabul. He afterwards went into hiding between 1958 and 1963. During the liberalisation of the 1960s, he joined the Maoist movement Shola-yi Jawid. In 1968 he once more went into hiding and a price
was put on his head. A biography of Kalakaniin part the source of this informationwas published in French in Brussels in December 1982 in the periodical
Shah Nama.

214

The Guerrillas

other movements, his men disbanded and his movement disappeared.7 Up to the fall of Kabul in 1992 the Maoists were just as active in the west of the country. In Farah the Jebhe-yi Moallimin
(Teachers Front) brought together several hundred fighters, at Bala
Bolak and near the Iranian frontier, whose affiliation, although they
did not openly advocate Maoism, was known to all in the region.
The unstable political circumstances in Farah afforded them the opportunity to exercise influence on the local level, through alliances
forged with local commanders. In Nimruz a Front based at Zaranj
was led for several years by two Maoist leaders: Parwiz, a Shiite,
and Gul Mahmad, a Pushtun sympathetic to SAMA. The persistence of Maoist Fronts in the west was explained by the relative
weakness of the other parties, particularly Hezb-i Islami. A significant
role had been retained in Nimruz by the notables, which shielded
the Maoists both directly, since they were often linked to the notables by family ties, and indirectly, since the Islamists were less influential in this area. Iranian support, which was overt in Nimruz in
Parwizs case, tended in the same direction. It is likely that the Iranian government wanted a clientele independent of Pakistan in its
frontier region. The establishment of Hezbollah groups in Herat had
a similar purpose.
Attempts to infiltrate are by nature clandestine and tend to be uncovered only if they go wrong. In addition, a distinction between the
individual presence of former Maoists and genuine infiltration was
not always easy to draw. Such attempts brought two issues into focus.
First, they targeted Harakat-i Enqelab, at least in those cases where
some record of events exists, for example in Herat and Balkh at the
start of the war. The reason for this may have been the partys loose
organisation, as well as the family ties between the Maoists on the
one hand and the khans and mawlawis on the other. Secondly, in the
absence of popular support, the Maoists were obliged to act in secret
and their initiatives mostly failed, at least where their existence
aroused the antagonism of the other parties.
Elsewhere the Maoists kept a foothold in Peshawar, despite the
regular assassinations of which they had been the victims since the
7

SAMA was afterwards led by his brother Abdul Qayum Rahbar, who was unable
to make his base in Afghanistan. He was assassinated in Peshawar on 27 January
1990.

The evolution of the political map

215

start of the war, the responsibility for which was generally borne by
Hezb-i Islami.8 There was also a secretive but well established Maoist community in Quetta.
The Sunni parties. During the 1980s Jamiyat-i Islami became the
dominant party in the north, while Hezb-i Islami (Khales) and the
Ettehad also expanded their influence. The losersMahaz-i Melli,
Harakat-i Enqelab, and Jebhe-yi Islamibegan to lose their influence, a tendency which accelerated after the fall of Kabul.
All observers agree on the importance of Harakat-i Enqelab during the first years of the war as the only party which could claim at
that time to exist on a national scale. This initial success was related
to the support of a majority of the ulema and of the khans. However,
from 19823 onward Harakat-i Enqelab declined markedly in the
north, to the benefit of Jamiyat-i Islami, particularly in the provinces
of Balkh, Herat and Takhar, while in the west its decline benefited
Hezb-i Islami (Khales). The inroads of competing parties also threatened Harakats position in Logar, though less seriously. Areas which
stayed solidly under Harakat-i Enqelabs control were the north of
Helmand, up to the point of ras Abdul Wahids defection to Jamiyat-i
Islami in 1989, and above all Ghazni where the young and warlike
ulema monopolised power in spite of the split initiated by mawlawi
Mansur in 1981.9
8

The kidnap and murder of Dr Samad Durrani provides an example of the difficulties of the Maoists in establishing themselves politically in Pakistan. Durrani,
from the province of Paktya, was born in 1947 and studied medicine at the University of Jalalabad. He fled to Pakistan at the end of 1979 and set up the Jebhe-yi
Mobarezin (Combatants Front), a movement based on Maoist principles, as well
as playing an active role in the Afghan Doctors Association. In August 1981 he
took part in the trial of the Soviet invasion organised by the Peoples Tribunal in
Stockholm. In late 1980 he also set up a school for bare-foot doctors. On
25 May 1982 some members of Hezb-i Islami kidnapped Durrani and executed
him shortly afterwards, with the acquiescence if not the complicity of the Pakistan police. A womens organisation, RAWA (Revolutionary Afghan Womens
Association), also maintains a militant stance.
9 This secession, the earliest and the most significant, was carried out by mawlawi
Mansur, a former pupil at the Nur al-Modares madrasa in Ghazni, who had retained his links with the Mujaddidi family as a murid and a former student of the
pir. It is not accidental that mawlawi Mansur gave the name of Khodam ul-Forqan
to his splinter group of Harakat-i Enqelab, echoing the title of the pre-war movement. He relied largely on Harakati commanders from Ghazni who had come

216

The Guerrillas

Weaknesses in Harakat-i Enqelabs leadership account for its decline, but this was probably not decisive: after all, mawlawi Khales was
no better organised. The nature of Harakat-i Enqelabs recruitment
process was itself at issue: many of the commanders were khans or
mawlawis advanced in years who were unable to bring useful cadres
to the movement. The commanders conservatism was a barrier to
innovations in both the military and civilian areas. Harakat-i Enqelabs
recruits guaranteed it a high level of social legitimacy, but not necessarily any particular aptitude for warfare. A final factor was that Pakistan gave no priority to the support of Harakat-i Enqelab. In 1983
many commanders went over to Jamiyat-i Islami because Harakat-i
Enqelab was not arming them sufficiently.
Jamiyat-i Islami gradually achieved predominance in the north,
from Badakhshan to Herat. Its success was related to the conjunction
of a number of factors: the presence of educated individuals among
its members, strong links between its commanders and the rural
population, and a high level of external aid. In addition the decentralised structure of the party, resulting largely from the failure of the
leadership in Peshawar to impose its will, had at least the advantage
of favouring the emergence of powerful local organisations without
too much opposition from Rabbani. Jamiyat-i Islami also gained a
foothold in the south, especially under mullah Naqibullah in Kandahar, though it did not achieve a predominant position anywhere in
the Pushtun region. In the mid-1980s Jamiyat-i Islami became the
leading party in Afghanistan, in spite of low membership in the east
of the country.
The success of Ettehad was essentially due to the financial resources
of its Arab patrons. It recruited on an ad hoc basis among commanders who wished to change their party,a consequence of which was the
random nature of its geographical distribution. Meanwhile Hezb-i
Islami (Khales) recruited mainly new commanders in the west of the
country, thus spreading beyond its regional base but not leaving behind its sociological roots, since it continued to be largely Pushtun
and tribal.
from the Nur al-Modares, and who almost all rallied to him. It may be deduced
that Muhammad Nabi must have had difficulty imposing his will on the former
pupils of the pir Mujaddidi, since he was not an alumnus of the Ghazni madrasa
and had received a more traditional education.

The evolution of the political map

217

The erosion of Mahaz-i Melli and Jebhe-yi Islami continued. The

first retained a regional presence in Kandahar, while the second was
totally marginalised. Lacking Pakistani support and any credible organisation, these two parties rapidly abandoned all hope of influencing military developments.
Finally, the decline of Hezb-i Islami was striking. The issue was
why, after its initial success, this movement progressively lost its influence, in spite of its numerous cadres and its disproportionate share of
Pakistani aid. Its failure is probably connected to rejection by a population unwilling to tolerate the authoritarianism of its officials. In
the end Hezb-i Islami made the same mistakes as the Khalq in its
behaviour towards the rural people. Furthermore it became marginalised as a consequence of its strategy of confrontation with other parties. Its clashes with Harakat-i Enqelab and Jamiyat-i Islami unleashed
what were virtually regional wars, especially in the north, the west
and the province of Laghman. The most serious clashes were with
Jamiyat-i Islami, especially on Masuds supply routes. At the worst
moment of the Soviet offensive of September 1981, Hezb-i Islami
attacked Masud through Koh-i Safi, putting him in the perilous
position of having to withstand an attack on two fronts. In a further
instance on 9 July 1989 Sayyed Jamal, a Hezb-i Islami commander,
ambushed some of Masuds men, after which twenty-five mujahidin
were taken prisoner and executed. A major row ensued and Hekmatyar, who had to explain his commanders actions, spoke of a local settling of accounts. After a long search Sayyed Jamal and three other
Hezb-i Islami mujahidin were arrested and on 23 December 1989
they were publicly executed. This aggressive strategy did not save
Hezb-i Islami, hitherto well established, from being progressively reduced by Masud to isolated Pushtun remnants.
The Shiite parties. From the Shiite standpoint the war was fought in
two distinct theatres. First, inside Hazarajat the conflict was mostly
an internal affair, while outside it the war was waged in coordination
with the Sunnis and presented no distinctive features. After the war
within Hazarajat, there followed changes in the relationship between
Sunnis and Shiites.
Because of Hazarajats isolation and the absence of Soviet-Afghan
troops, the situation there was an interesting one, which revealed the

218

The Guerrillas

political dynamics in a distinct region which had no substantial interaction with the state. Tensions suppressed elsewhere because of
the exigencies of the jihad were fully displayed. Some tendencies such
as the growing importance of the ulema, the decline of the khans and
the emergence of ethnic politics were harbingers of developments
that took place at the national level in the 1990s.
Soviet military operations in Hazarajat were kept on a very limited scale. Certainly Bamyan rose in revolt in July 1979, but the city
was retaken in 1980 by the government, which held it till 1988.
Armed columns which occupied Behsud and Panjao during the
summer of 1980 stayed only a few days. No operation of any size took
place subsequently, other than occasional bombardments of Waras
and Behsud. The reason for the Soviets lack of interest in the interior
of Afghanistan lay in their priorities: control over the economically
exploitable sector of the country and the major communication
routes. This, as has been seen, was no barrier to an awareness by
Karmals government and then that of Najibullah of the opportunities for political leverage presented by Hazara nationalism. In this region the primary dynamic consisted of internal conflicts: first the
elimination of lay forces, then conflicts between revolutionaries and
conservatives.
At the beginning of the war, almost the whole of Hazarajat was
controlled by Shura-yi Ettefaq, led by Sayyed Ali Beheshti, and in
the first two years the khans and the Maoists were eliminated, as well
as the nationalists, who were prey to the hostility of the ulema who
ran the party.
Within the Tanzim Nesl-i Naw-i Hazara-Moghol10 movement
there emerged a Hazara nationalist party, the Ettehadia, whose original feature was that it consisted of both Pakistani and Afghan Hazaras.
10

This nationalist organisation, mentioned in Chapter 2, had been exploited in the

past by the Pakistani government. In 1973 Dauds return to power and his irredentist Pushtun propaganda led the government of Zulfikar Ali Bhutto, using
the Hazara radio station based in Quetta, to give its support to the Tanzim in the
interests of its promotion of a greater Hazarajat in the centre of Afghanistan.
Links with the Afghan community were mediated through the Hazaras, who
traded with Quetta. When the war came the Hazara population of Quetta grew,
bringing an increase in nationalist activities. For example, Ghajestani, a former
journalist and author of a number of books on the history of Hazarajat, lived in
Quetta where he ran a Shura-yi Farangi (cultural association) before being assassinated in Afghanistan after the fall of Kabul.

The evolution of the political map

219

Relations between the Tanzim and the Ettehadia were very close,
with the former giving logistical support to the latter and providing
its intellectual framework. In May 1979 haji Rasul, shaikh Mohaqeq
Turkmeni, Fasei and Akbari (from Jaghori) set up the Ettehadia-yi
Mojahidin-i Afghanistan (Union of Mujahidin of Afghanistan).11
The aim was to provide support for those Hazaras fighting in the interests of a Hazara nationalist renaissance. The movement had its
main bases in Jaghori, where the commander was haji Barakat. This
was the Hazara district nearest to Quetta, the place from which a
substantial part of Quettas Hazara community originated. From
there too came the majority of its militants. The first arms convoy
set off for Quetta towards the close of 1979, led by haji Rasul. At that
point Ettehadia was not yet receiving military aid from the Pakistan
government, but the Harakat-i Enqelab and Jebhe-yi Islami provided some arms. In the autumn a press conference in Quetta attracted
the attention of the Pakistan government and from then on Ettehadia was substantially assisted by the ISI. From Pakistans point of
view, the advantage presented by the Hazara nationalists was that
they acted as a counterbalance to the Iranian revolution. Ettehadia
at that time played the role of an arms supplier for the other Shiite
partiesShura, Nasr and Harakat-i Islami. However, the movement was unable to expand further, due to internal crises and rivalry
with the other parties. Following multiple splits12 Ettehadia ceased
effectively to exist, while it also met resistance inside Afghanistan. In
1981 it was militarily vanquished by the other parties after several
days of fighting, and thereafter its role ceased. Contrary to allegations which have been made, this party was not simply a channel for
11

Saddeqi, a future leader of Sepah who was also in Quetta, refused to become
involved and went directly to Iran, where he already had contacts. Fasei was appointed president, haji Rasul vice-president, and Aedri secretary-general. Aedri,
later president of the Ettehadia, was a Pakistani citizen. Mohammad Akram
Gisabi, a Hazara from Kabul, joined the movement in 1980 and later became
vice-president.
12 During 1980 the first signs of dissent appeared within Ettehadia: Fasei, the president, joined Nehzat and a split ensued, led by haji Rasul, who was opposed to
Abdul Husein Maqsudi, the new president. haji Rasul, who had links with Nasr
and the pro-Iranian parties, became leader of a new Ettehadia. For his part
Akbari, who had set off with an Ettehadia convoy, joined the Shura, for which
he became a spokesman, which further embittered relations between these two
parties.

220

The Guerrillas

the ISI to supply arms to Hazarajat. It failed first of all as the result of
the overwhelming capacity of the Iranian revolution to attract support, as well as its inability in any real way to break into the religious
networks. In reality, though some disciples of ayatollah Khoei from
Najaf were sympathetic to Ettehadia, it was unequivocally lay and
even leftist.
The only bid by the khans to achieve political expression was by
way of the Shura-yi Arbabah (Council of the Arbab),13 whose leading figures, Gharibdad from Behsud and haji Nader of Turkmen,
were assassinated by Nasr in 1983 and 1984 respectively. In any case
this shura was never more than an informal group, rather than a structured party. In fact a significant number of khans joined Harakat-i
Islami, though it was run by a Shiite Qizilbash, because of its conservative ideology and its opposition to the pro-Iranian revolutionary parties. However, the lack of any previous Harakat-i Islami
presence and the intrinsic weakness of the khans precluded its success. In 1985 the Nasr partys assassination of Tawala, the commander of Harakat-i Islami, who was a son of the principal khan of
Yakaolang, initiated the definitive decline of Harakat-i Islami as a
political force within Hazarajat.
Two factors account for the collapse of the khans. First, there was
the absence of any political connection with Pakistan or Iran which
might have given them arms and lent political support. Support
from Iran was out of the question, so there only remained Pakistan,
which however chose to support first Ettehadia and then Harakat-i
Islami. Second, political organisation was lacking, a factor attributable to the traditional rivalry between the khans, as well as to their
age, which meant that they had little appetite for involvement in political and military activism. The generation which had the ability to
organise itself on the political level had already become politically
committed before the war, often in the cause of the left. This seems
likely to be the most convincing reason for the khans eclipse. They
were thus doubly the losers on the political scene, since they were
not only swept from power themselves, but their offspring, who
were often Maoists or nationalists, were also politically eliminated.
The khans were the most modernising and educated group among
the population of Hazarajat, and their abrupt displacement was to
have serious consequences for the population as a whole.
13

In Hazarajat the expression arbab is equivalent to khan.

The evolution of the political map

221

There were many Maoists among the Hazaras from Hazarajats

border regions. In these more developed areas some of the notables
sons attended the university, where most joined Maoist groups. They
were active in the province of Ghazni in particular, and especially in
Jaghori: the Mahmudi brothers were one example. At the beginning
of the war there was apparently a Maoist Front of which five members were hanged in Kabul in 1980. No longer able openly to acknowledge their Maoist allegiance in face of the hostility of the
other parties and a section of the population, the Hazara Maoists of
the SAMA turned to entryism, targeting in particular nationalist
movements such as Ettehadia.Later they mostly withdrew to Quetta.
After the elimination of the lay parties, tensions sharpened between the Shura and the revolutionary pro-Iranian parties, especially
Nasr, which soon allied itself with Sepah. In early 1982 there was
a Nasr offensive against the Shura, with the complicity of some of
the latters commanders. Husein Saddeqi, the Shura representative in
Iran, who had returned to Nili in 1981, went over to Sepah. He led
the principal attack on Waras, Sayyed Beheshtis village, which was
the Shura headquarters, but could not advance further than Panjao
because of a counter-offensive mounted by Sayyed Hasan jaglan,
the Shuras military chief. At the close of 1982, after this first offensive, the new political position on the ground showed an overall loss
for the Shura, which no longer had real control beyond the regions
of Waras, Panjao and Naur. The revolutionary parties, Nasr and
Sepah, were dug in at Deh Kundi, Lal o Sarjangal and Yakaolang, as
well as Dara-i Suf and Turkmen. Shura thus lost control of the centre
of Hazarajat, which contained the most mountainous and inaccessible regions.
After sporadic fighting, the second major offensive by the revolutionary parties took place in 1984. On 6 May 1984 Sayyed Beheshti
was dislodged from Waras and obliged to retreat to Naur, Sayyed
Hasan jaglans fief. Backing from Harakat-i Islami enabled him to retake Waras in September. The positions on the ground stabilised and
thereafter no further large-scale attack was launched, though sporadic fighting continued until 1989. This second offensive in particular revealed the polarisation between the moderates, Harakat-i Islami
and Shura, against the revolutionaries of Nasr and of the Pasdaran.
The fierce but highly localised struggles in which the parties were

222

The Guerrillas

later involved are difficult to detail. There was clearly a tendency

towards fragmentation: clashes took place between all the parties,
for example between Nasr and Sepah at Deh Kundi, and sometimes
even within them.
Political autonomy in Hazarajat led to a resurgence of clashes with
the Pushtuns. The politically-motivated exclusion of the nomads
from Hazarajat became generalised, with the exception of a few areas
such as Naur, a traditional crossing point.14 The collapse of the central government removed the protection enjoyed by the Pushtun
nomads, who were also robbed and involved in clashes which closed
off their traditional migration routes.
In addition to the clashes with the nomads there was also conflict
throughout the war between Hazaras and Pushtuns along Hazarajats
frontiers, especially at Jaghori but also at Anguri in Ghazni province,
near Kajran in Ghor, and at Meydan Shahr in the province of Wardak.
On the other hand, in the northern Ghazni and in Wardak other
than in Meydan Shahr the clashes were limited in scale. In the case of
Wardak the reason may have been that the long-standing Pushtun
presence in the valley of Wardak, dating from their conquest in the
seventeenth century, tended to promote better interethnic relations.
There was no living memory of the loss of the land. Also the two
communities were geographically well separated by a plateau above
the valley. In Ghazni the Hazaras were in part townsmen, accounting
for around one-third of the population, and had fought the Soviets
and the government troops side by side with the Sunnis throughout
the war. The personality of the local commanders also played a role:
Shah Jan of the Shiite Harakat-i Islami and Qari Baba for the Sunni
Harakat-i Enqelab favoured coexistence. The most savage conflicts
took place at Jaghori and at Anguri, where the Hazaras apparently
went on the offensive to recapture lands confiscated under Abdul
Rahman Khan. In the south and in Uruzgan the flight of Pushtun
families was reported, especially in the spring and autumn of 1990.
Meanwhile Hazaras travelling to Pakistan were often seized for ransom by the Pushtuns.
14

The situation was less tense in previous years: see Alain Guillo, Jean-Jos Puig,
Olivier Roy, La guerre en Afghanistan: modifications des dplacements traditionnels et mergence de nouveaux types de circulation, Ethnologica Elvetica
(Berne) 7, 1983.

The evolution of the political map

223

Relations between Sunnis and non-Hazara Shiites were on the

whole better than those between Pushtuns and Hazaras. The nonHazara Shiites, especially in the towns, had developed sophisticated
relations with the Sunnis which tended rather towards cooperation
than confrontation. When conflicts took place, the political affiliations of the commanders were crucial because although the commanders generally took decisions according to local circumstances,
the directives of the parties often had direct consequences for local
inter-party relations. Of the Sunnis Hezb-i Islami (Khales) and
Ettehad were the most anti-Shiite, although the Sunnis generally had
good relations with Mohsenis Harakat-i Islami, which was mainly
established in urban areas or on the fringes of towns. By way of illustration, the local Harakat-i Islami in Herat (formerly Rad) maintained good relations with Jamiyat-i Islami, but Ismail Khan several
times attacked the Hezbollah militias, whom he accused of having
links with his enemies.

Inter-party cooperation
The parties in Peshawar repeatedly attempted to coordinate their activities, usually in response to pressure from American or Pakistani
donors, but up to the fall of Kabul this cooperation was in name
only: it had no effect on the structure of the parties, which remained
independent of each other, and had hardly any impact on the situation on the ground.
In April 1981 an assembly of ulema summoned the party chiefs.
After some days of discussion an alliance was established bringing
together Jamiyat-i Islami, the two branches of Hezb-i Islami, Ettehad,
Jebhe and the fragments of Harakat-i Enqelab, especially that of mawlawi Mansur. However this alliance, entitled Ettehad-i Mujahidin-i
Afghanistan, proved ephemeral. After the failure of this initiative two
separate coalitions were established on the basis of ideological affinities and personal relations between leaders. On 2 June 1981 an alliance
sympathetic to royalist ideas came into existence which included
Harakat-i Enqelab, Jebhe and Mahaz.15 Following the appeal by
15

Mahazs military weakness led it to adopt the jirga (the tribal assembly) as its
mode of political-diplomatic action. These ploys never succeeded, in particular
as the result of the position of the Pakistanis and the Islamists, who were opposed
to the return of the king. The first came in the spring of 1980 at the initiative of

224

The Guerrillas

Zahir Shah in Le Monde on 22 June 1983 this coalition appealed for

the creation of a united resistance front coordinated by the king.
Meanwhile the remaining parties also made attempts to institutionalise their cooperation. On 2 May 1982 at Rawalpindi the
announcement was made of a complete unification of the movements. On 25 May 1983 Sayyaf was elected president in a second
round against mawlawi Khales. Rabbani was entrusted with military
affairs and Hekmatyar became vice-president. The president was
henceforth to be elected for a period of two yearsthere had initially
been a rotating presidencybut the profound hostility between the
groups involved precluded any joint strategy from developing. On
7 March 1984 the alliance was hit by a crisis when Sayyaf, in the
absence of Rabbani and Khales, arranged for his cronies to be elected
to positions of responsibility in charge of financial and military affairs.
In the end the two alliances disintegrated, and on 1 April 1985 the
parties proclaimed the formation of the Alliance of Seven for
which Hekmatyar became the spokesman.16 The earliest initiatives
towards the formation of a para-governmental structure date from
this period. On 22 July 1986 the Alliance announced the establishment of offices in New York and in Jeddah (to liaise with the
Organisation of the Islamic Conference). On 17 January 1987 this
alliance set up a High Council of Mujahidin bringing together the
seven Sunni parties. They participated for the first time together at
the Islamic Conference in Kuwait, and on 29 January 1987 Sayyaf
addressed the conference but failed to obtain the status of observer
which the mujahidin had been seekingunsuccessfullysince 1980.
Afghanistans seat remained vacant, as it had been since 1980.
Omar Barakzai, and therefore fell outside the sphere of influence of the Gaylani
family, who opposed it. The second, in the autumn of 1981, was planned by a
committee of former ministers and parliamentarians. This jirga, which was
planned to take place in Quetta, ended in disarray after being moved to Peshin
because of the opposition of the Pakistani government. The third jirga, which
never got beyond the planning stage, was the result of an appeal by the king on
22 June 1983. Later there were repeated attempts to convene a jirga, but these
met with limited success due to alterations in the balance of forces within the
country which closed the door to this kind of procedure.
16 The precise name of this alliance appears to have been Shura-yi Ettehad Islami
Haftgane (Council of the Islamic Union of the Seven).

Inter-party cooperation

225

As for the Shiite parties, all attempts to unite had taken place under Irans auspices, and were translated into an impetus towards unity
in the face of the Sunnis which existed at the level of diplomacy
rather than on the ground. In 1982 the first attempt at alliance
brought together various small parties within the Jebhe-yi Mottahid-i
Enqelab-i Islami, which in 1984 became the Etilaf-i Chaharganah
(Union of Four Parties). In 1985 an alliance based at the Iranian
city of Qom under the patronage of Ayatollah Montazeri brought
together Nasr, Sepah, Niru and Harakat-i Islami. Two years later the
principal constituents of another alliance, Shura-yi Etilaf-i Islami, were
Nasr, Sepah and some small groups of Khomeini supporters, excluding Shura-yi Ettefaq and the Mustazaffin. Relations between the
parties continued to be hostile on the ground, so that this alliance
operated only in the diplomatic sphere, and specifically in negotiations with the Sunnis over the setting-up of a parliament in exile.
The announcement of the Soviet withdrawal meant that the issue
of cooperation needed to be faced once more, this time as a matter
of urgency. On 11 May 1988 a provisional constitution was adopted
and on 19 June the Sunni parties announced the formation of an
interim government. This assisted the first efforts of the mujahidin to
establish a diplomatic presence. Rabbani was delegated by the alliance to make a tour of Africa. In addition he was received by President Reagan on 9 November 1988, and by the UN secretarygeneral. From 3 to 5 December he was at Taef, where he negotiated
with a Soviet delegation led by Vorontsov, the new Soviet ambassador in Kabul.
However, this interim government did not represent the resistance movement as it existed in reality. Each party was allocated two
ministers, which did not reflect their presence on the ground and
entailed an over-representation of the Pushtuns. In a government of
fifteen members there were eleven Pushtuns, two Tajiks, one sayyed
and one Hazara. The Pushtuns were mostly Durranis from Kandahar, including one with no party affiliation. The two Tajiks were
members of Jamiyat-i Islami. The process of choosing representatives exacerbated the ethnic opposition between the parties.17 An
additional issue was that the Shiite parties were not represented.
17

The Kabul government simultaneously announced the formation of a new cabinet, led by Hasan Sharq, of which the ethnic make-up was much more balanced.

226

The Guerrillas

The Soviet withdrawal accelerated the setting-up of a provisional

government, which was already present in embryo in the interim
government. The presumed imminence of victory led the Pakistani
leadership to promote the establishment of a body to represent the
mujahidin. The convocation of a shura announced by Mujaddidi in
January 1989 was the prelude to the provisional governments formation. The members of the shura were mainly Pushtuns, since each
party automatically sent sixty delegates. Presence on the ground was
not taken into account, and the over-representation of eastern Pushtuns was blatant. Disagreement immediately broke out among the
parties on the role of the Shiites and the part to be played by the king.
All Shiite attempts to obtain a place within this government failed.
The crucial disagreement was over the number of seats to be allocated
to the Shiites within a common shura. Notwithstanding, Mujaddididelegated by the Sunni partiesreached a preliminary agreement in Teheran on 5 February with Muhammad Karim Khalili, the
spokesman for eight Shiite parties, which were allocated eighty seats
in a shura of 539 members. However, this agreement was not endorsed by the Sunni parties, who offered the Shiites sixty seats out
of 440, in contrast with the 100 seats for which the Shiites were asking. Opposition to the Shiites from Hezb-i Islami, Hezb-i Islami
(Khales) and Ettehad precluded agreement, and these parties were
supported by Pakistan and Saudi Arabia, with the aim of keeping
Irans influence to the minimum.
Despite the absence of the Shiites, the shura, composed of 400
members, including some twenty intellectuals living in Europe, opened
its deliberations in the Pakistani city of Rawalpindi. On 23 February
Mujaddidi became president, with 174 votes, prevailing over the
other leaders. Sayyaf was appointed Prime Minister. On 9 March the
provisional government was recognised by Saudi Arabia, followed by
Sudan, Malaysia and Bahrain. A number of countries, including Britain and France, sent ambassadors. On 6 April 1989 Peter Thompson
was appointed the US representative to the provisional government
with the rank of ambassador, and on 16 April the provisional government was accorded Afghanistans seat at the Islamic Conference.
The provisional government, financed by Arab and western benefactors, set up its headquarters in Peshawar. Corruption and inefficiency were its two most striking characteristics. The appointment

The Guerrillas

227

of provincial governors was followed neither by the establishment

of an administrative structure within the country, nor even by enhanced coordination between the parties. Overall, the government
showed little cohesion. For example, the dispatch in January 1991
of a contingent of 2,000 mujahidin to assist the coalition against
Saddam Hussein was carried out against the wishes of some of the
parties, notably Hezb-i Islami. Developments on the ground were
in any case destined to make an irrelevance of this government, set
up as it was in anticipation of the rapid collapse of the Kabul regime.

The mujahidin after the withdrawal

The period from 1989 to 1992 was one of transition. The jihad of
the early years was over, but the survival of the Kabul regime continued to obstruct the establishment of the emerging new political
configuration. The rise of the mujahidin was especially swift during
the Soviet withdrawal. The morale of the fighters, who were aware
of having won a historic victory, was very high. In addition arms
flooded in from Pakistan. During the summer of 1988 a large number of military positions and towns fell into the hands of the mujahidin,
including in June Meydan Shahr (later retaken by the government),
Muhammad Agha and Qalat-i Ghilzai; in August Kunduz, Dara-i
Suf and Bamyan; and in early September Imam Saheb and Spin
Boldak. Asadabad fell on 11 October, and 21 October saw the capture of Mahmud Raqi, the provincial capital of Kapisa. By the time
the Soviet withdrawal was complete the Kabul regime held only
some sixty urban centres, and six provinces out of twenty-five had
fallen entirely into the hands of the mujahidin.
However, these victories were partly illusory, since they were the
result more of the redeployment of the governments defences than
of any improvement in the coordination or logistics of the mujahidin.
In fact, in the context of the post-withdrawal situation, Pakistani
strategy was aimed at the capture of the towns in order to precipitate
the collapse of the regime. However, the guerrillas did not possess
the human or military resources to bring to completion a campaign
which necessitated the military coordination of thousands of men. A
further factor was the suspension of US aid to the resistance from
December 1988 to July 1989: a crucial moment, just as the Soviets
were stepping up their arms shipments to the Kabul regime.

228

The Guerrillas

After the withdrawal of the Soviet forces from Jalalabad in May

1988, this city became the main objective of the Pakistani military,
which believed that its fall would result in the fall of Kabul, enabling
Hekmatyar to seize power. In this sense the fall of Jalalabad represented a risk for the other parties, and Masud is said to have allowed
the passage of a number of large convoys in March and April 1989 to
save the government forces from being starved of supplies.18 The
Kabul regime, aware that its future was at stake, sent 15,000 men to
Jalalabad, which had already been reinforced by troops from other
positions the government had abandoned. On 6 March 1989 Hezb-i
Islami (Khales) and Mahaz-i Melli mujahidin began their offensive
with the acquisition of Samarkhel, to the east of the city, which they
captured by bribing the officer in command. The mujahidin then
attempted to take the citys airport but suffered heavy losses; in particular, Scud missile strikes threw the attack into chaos, while at the
same time government reinforcements advanced along the KabulJalalabad road. Meanwhile, on 10 March the guerrillas attacked from
the north and then in the following days also from the south. However, government air support prevented the guerrilla fighters from
massing and the mujahidin were driven back. Fighting went on into
April but by now the government was on the offensive. A further
convoy came in by road from Kabul to Jalalabad and a counterattack at the end of April led to the guerrillas being driven out of the
outskirts of the city. Resistance losses were heavy, probably totaling
several thousand men, because of the lack of air cover on the mujahidin side and of a lack of discipline and effective command among
the Mahaz forces. The presence of Pakistani officers and Arab fighters on the mujahidin side was insufficient to transform a guerrilla
force into a regular army. The massacre of seventy soldiers and mutilation of their bodies by Khaless men during the capture of a frontier post also stiffened the government troops.19
This notable setback was to have a number of consequences. First,
there was a certain distancing by the United States, and later the morale of those mujahidin who no longer believed in a military victory
began to suffer. As Ismail Khan admitted, The battle of Jalalabad lost
18

A charge which Masud denied, while attacking Pakistans strategy (according to

an interview in October 1991).
19 Les Nouvelles dAfghanistan, 43, June 1989.

The mujahidin after the withdrawal

229

us the credit won in ten years of fighting.20 In consequence the

strategies of the parties began to lean towards a coup dtat, as the
prospect of a military overthrow of the regime receded.
During the two years which followed the defeat at Jalalabad, the
situation appeared to ossify, and in some areas the resistance even lost
ground. The capture of Khost on 31 March 1991 was an unusual
military success for the opposition, but this was a town which had
been completely surrounded since the beginning of the war and was
not of great strategic significance, although its capture did facilitate
the journey to Pakistan. During this period, as the fighting became
less intense, Masud was the only dynamic element; in both 1990 and
1991 he launched a spring offensive, and its success, though limited,
allowed him to extend his influence in the north. On 11 April 1990
he took Dasht-i Archi, and in June 1991 at Khwaja Ghar he inflicted
a defeat on the government militias, who withdrew towards Mazar-i
Sharif. Masud strengthened the security of his supply routes on the
Pakistani frontier with the capture on 15 July 1991 of Zibak, and on
21 July of Ishkashem in the province of Badakhshan.
Generally, the military stagnation aggravated the tensions between
the commanders and the leaderships in exile, whose shortcomings
were the subject of lively criticism. The establishment of parties
based in the country and the setting up of shuras by the commanders
were both expressions of this difference of views. Shuras set up by
commanders with assistance from the Pakistanis and the Americans
were an attempt to bypass the party leaderships as the levels of aid
diminished.21 Two shuras involving the commanders took place in
20
21

Interview, autumn 1989.

In 1987 a bid to coordinate the commanders, taken at the initiative of Ismail
Khan, took place with the aim of coordination at grassroots level to distance the
commanders from the political intrigues of Peshawar. From 12 to 23 July 1987
commanders from all over Afghanistan and from all the parties went to Sholgar
in Ghor province. Several thousand mujahidin attended, and around 1,200 commanders were represented, coming principally from Herat, Badghis, Ghor, Faryab, Farah, Helmand, Kandahar, Uruzgan, Wardak, Logar and Kabul. Among the
leading commanders were mullah Naqibullah, Dr Fazlullah and Amanullah
(from Hezb-i Islami in Helmand). Decisions taken related to improved coordination of the commanders (through radio contacts, joint training and strategic
bases, and the common administration of justice) as well as to the rejection of
any cease-fire or compromise with the government. The second meeting, set for
six months later, failed to take place due to the opposition of the various headquarters in Peshawar.

230

The Guerrillas

Paktya, one from 7 to 9 May and another from 22 to 25 June 1990.

A third meeting presided over by Masud was convened between
9 and 13 October at Shah Salim on the Badakhshan border. Among
the major commanders attending were mawlawi Haqqani, Qari
Baba, Abdul Haq, Amin Wardak, Sayyed Hasan jaglan and mullah
Malang, acting as a representative of Ismail Khan. The main points
on the agenda were military coordination, the free movement of
guerrilla groups and the administrative division of Afghanistan into
nine districts. Masud even paid his first visit to Pakistan since 1979
and in Islamabad met the army chief-of-staff General Beg and the
head of the ISI, Lieutenant-General Asad Durraniit appears that
these meetings had no outcome. This initiative was undertaken
with American support, but failed because of the divisions between
the participants; in fact the plan lacked coherence because of the desire of some commanders for military cooperation while others,
especially in Hezb-i Islami (Khales), wanted greater autonomy. In
this context it was easy for the party leaderships in exile to connive
at the failure of any such initiatives, which could have challenged
their monopoly of representation.
This challenge to the parties also paved the way for the emergence of a local movement in the province of Kunar. Other than in
the Nuristani regions, which followed their own line, the Kunar valley at the start of the war was largely Hezb-i Islami territory.22 The
amir-welayati (provincial commander) of Hezb-i Islami was Jamil urRahman, whose real name was mawlawi Husein.23 He was at that
time an alem and a former student of the madrasa of Panj Pir, which
was renowned for its fundamentalist teaching. The geographical position of Kunar and Jamil ur-Rahmans background meant that in
the early years he benefited from Arab aid. This had been a significant factor from 1985 as the result of the involvement of Arab militants in the camps of the Bajaur Agency (in Pakistan), where refugees
22

In the first phase of the rebellion before Hezb-i Islami took charge the khans
were active. In the Pesh valley the revolt was led by Samiullah Safi, born in 1940
in Murchel, a notable (son of Sultan Mohammad Khan) who was a former
schoolteacher and a member of the Wolesi Jirga from 1969 to 1973.
23 Born around 1939 in Pesh, Jamil ur-Rahman, who was from the Safi tribe, became a member of the Islamist movement before the war and took part in the
rising of 1975. He later joined Hezb-i Islami at the beginning of the war, and
took part in the rising of July 1978.

The mujahidin after the withdrawal

231

from Kunar were present in large numbers. Many volunteers, in

particular Egyptians, fought under Jamil ur-Rahman, who received
money directly from King Fahd of Saudi Arabia. Jamil ur-Rahman
left Hezb-i Islami in 1986 or 1987, whenwith Arab aid broadly
guaranteeing his independencehe headed the Jamaat-i Dawaat-i
al-Quran wa-Sunna-yi Afghanistan (Society for the preaching of the
Quran and the Sunna in Afghanistan). This party was ideologically
close to the Ahl-i Hadith.24
The evacuation of Soviet troops and the militias in 1988 left the
field free for Jamil ur-Rahman, who adopted an expansionist strategy, either crushing or buying off the other commanders in Kunar.
At the end of 1989 he and his only serious competitor Hezb-i Islami
reached an agreement providing for the election of a shura,25 which
in practice met in the early days of March 1990. However, rivalries
quickly re-emerged and in January 1991 Jamil ur-Rahman proclaimed the establishment of an Islamic emirate of Kunar.The emirate
was constituted as a government of fourteen ministers in the framework of a Central Council (Shura-yi Markazi), elected by an Executive Council (Shura-yi Ejray). All the ethnic groups of the province
and the various political parties were represented, especially the
Daulat-i Enqelab-i Islami-i Afghanistan Nuristani.26 In the spring of
1991 Hezb-i Islami lost its main bases, and its leader mawlawi Faqer
Muhammad found himself encircled by Jamil ur-Rahmans troops.
Jamil ur-Rahman, who was himself absent in Saudi Arabia, sent an
order to spare the life of mawlawi Faqer Muhammad, who was sent
to Peshawar. At this moment in early 1991 the virtual elimination of
Hezb-i Islami from Kunar prompted Hekmatyar to launch an offensive involving several hundred men, in cooperation with Hezb-i
Islami (Khales) and Harakat-i Enqelab. In August the provincial capital, Asadabad, was captured and sacked (the offices of the humanitarian organisations were not spared) and some fifty Arabs were killed.
After losing the battle Jamil ur-Rahman fled to Bajaur in Pakistan.
24

There was never any collective takfir towards the population, as B. Rubin contends in The Fragmentation of Afghanistan, New Haven, CT: Yale University Press,
1995, p. 86.
25 Hezb-i Islami always supported general elections, a point on which it differed
from the other six parties of the provisional government.
26 See the article by Paul Castella, Convoitises rgionales sur lAfghanistan.
Lexemple du Kunar, Les Nouvelles dAfghanistan, 55, 1, 1992.

232

The Guerrillas

Pakistan observed these developments attentively, and went as far

as to send the Bajauri Scouts to re-open the Nawa pass, at one point
closed because of the fighting. The presence of this unit at Jamil urRahmans side was viewed in different ways. Some saw it as a signal
that the ISI had dropped Hekmatyar, but in fact the recapture of
Asadabad by Hezb-i Islami could not have been carried out without
the tacit approval of the Pakistani intelligence services. In the following weeks the governor of the North West Frontier Province mediated between the vanquished Jamil ur-Rahman and Hezb-i Islami.
On 30 August 1991, during these negotiations, Jamil ur-Rahman was
assassinated by a twenty-four-year old Egyptian journalist, Abdullah
Rumi, who immediately committed suicide.27 Jamil ur-Rahmans
successor, mawlawi Samiullah, remained in Pakistan.
This conflict may be interpreted in two complementary ways. First,
it could be seen as the liquidation of a dissident Hezb-i Islami commander. Jamil ur-Rahman was not typical of a commander belonging to this partyhis links with the fundamentalist network of the
madrasa Panj Pir provided him with an opportunity to remain autonomous which he seized after the Soviet withdrawaland this tends
to confirm the general responsibility of the ulema for splits in Hezb-i
Islami. The other interpretation lies in the context of the Gulf War.
As it happened, Hezb-i Islami had adopted an explicitly pro-Iraqi
and anti-American position, as had the majority within the Muslim
Brotherhood. The struggle between Jamil ur-Rahman and Hezb-i
Islami therefore set pro-Saudis and pro-Iraqis against each other.
In contrast to the fragmentation observed among the Sunnis, several Shiite movements merged in 1990 in the Hezb-i Wahdat (Party
of Unity). This development, an augury of the impending regionalisation and ethnicisation of the parties, is analysed in the following
chapters.

27

Jamil ur-Rahmans assassin was close to the Muslim Brotherhood and especially
to Abdullah Azzam. He had from some months been a journalist with AlBunian-i Makhtus, after previously working for the magazine Jihad. He left no indication of the reasons for his action.

Part IV. THE TALIBAN

In the period after 1992 both the actors concerned and the ideological positions they adopted were substantially different from those of
the preceding era, to the extent that there are good reasons to speak
of a second Afghan war. The conflict became politicisednot depoliticised as is sometimes claimed.1 The crucial distinction was that
between depoliticisation and demobilisation. After the fall of the
Kabul regime religious duty could no longer be adduced as a justification for the war, and the spirit of jihad, the basis of popular solidarity, gradually dissipated. This development had been in evidence
since the Soviet withdrawal. The distancing of the mujahidin from
the people found expression in the language employed: the term
jang-i dakheli (civil war) was now in use, and jihad was no longer
mentioned. Mujahidin of all parties were accused of oppressing civilians, kidnapping and looting. Brigandage was also targeted at humanitarian aid operations, and attacks on convoys and hostage-taking
became more frequent.
The aggravated rivalry between the partieswhose behaviour,
far from being irrational, resulted from the fight for survivaldiscredited them in the popular view. In addition to the demobilisation
of the population, this rivalry had two further consequences. First,
the political forces became territorially based, controlling areas which
showed a tendency to political homogeneity. Secondly, only those
groups which had been able to put in place organised systems capable of taking control of resources were able to survive. These groups
were the educated class in the north and the ulema in the south and
in the Harzarajat. The process of concentration of resources therefore had both a communal and a social dimension.
The political actors, driven by competition, were obliged to develop
strategies to mobilise their supporters, since the unchallengeable and
essentially apolitical language of the jihad was no longer available.
Some of the parties opted deliberately to appeal to communal solidarity, an opportunity presented to them as the result of regionalisation, while others were involuntarily obliged to follow suit.
1

Olivier Roy, Afghanistan. Le retour des vieux dmons, Esprit, October 1989.

233

234

The Taliban

The conflict was therefore open to interpretation on three different

levels. It could be seen as being between ethnic groups; or between
ideological tendencies (which might be communist, fundamentalist
or Islamist); or between social groupsin particular the ulema and
the educated class. The claim of legitimacy for one or other of these
ways of reading the conflict was part of the strategy of the actors
themselves, since it had its impact on the parties policies and therefore in the end also on political loyalties. At first these lines of divisions did not coincide, which gave the war its particular character.
For example, the educated individuals who dominated the north
were either Jombesh communists or Jamiyat Islamists. The Hazara
ulema of Hezb-i Wahdat were opposed to the Pushtun ulema of the
Taliban, and the ethnic parties, Hezb-i Wahdat and Jombesh, were
controlled respectively by Shiite fundamentalist ulema and by educated communists. However, as the conflict developed, the various
divisions were gradually drawn into line with each other, ultimately
creating a single confrontation between a rural and fundamentalist
Pushtun movement and an educated and Islamist Tajik opposition.
As a prelude to the detailed examination of the issues of ethnicity
and the confrontation between the ulema and the educated class, we
now analyse the mechanism which led to the elimination of the majority of the actors and the coming into existence of a quasi-monopoly, to the benefit of the Taliban.

7. Competition and the Impetus

towards Monopoly (19922001)
In the 1990s three basic constraints governed the pattern of rivalry
between political groups. First, the confrontation took place within
a closed and unchanging geographical area. In the modern world
frontiers are more permanent than states, sometimes preceding them
and generally surviving their demise. No one in Afghanistan or in
the neighbouring countries sought to challenge them. The possibility that some part of Afghanistans territory might be annexed by
neighbouring countries, which had arisen repeatedly since the beginning of the war, was never likely to became actual. There was no
internal frontier, either administrative or communal, which could
conveniently have served as a basis for partition. In addition, the state
constituted a potential resourcefor international recognition,
access to aid and established legitimacyand became a principal
objective of the struggle, even for those parties which were against
its reconstruction. Competition between actors was therefore acted
out within the framework of the concentration of resources. This
process of coalescence through competition, with the difference that
the frontiers of the state were defined and the process was much
swifter, recalled the theories of Norbert Elias concerning the formation of the European states.2
Secondly, the war changed in nature. Guerrilla warfare to some
extent disappeared, to be succeeded by more conventional conflicts.
Operations, sometimes with air support, often involved tens of thousands of men. Their aim was to eradicate the enemy and capture the
towns, where wealth and population were concentrated. Forces were
increasingly territorialised, and recognisable front lines made their
appearance. To some extent these conflicts remained geographically
2

Norbert Elias, op. cit.

235

236 Competition and the Impetus towards Monopoly (19922001)

limited, in contrast to the period of the Soviet occupation, and outside the war zones the process of reconstruction went ahead. The
actors who survived were therefore those whose level of organisation enabled them to accumulate resources. Commanders and parties
which were too small or organised on the patrimonial model were
eliminated, since they were unable to compete with large organisations. A particular phenomenon was that the figure of the commander as an autonomous power gradually disappeared from the scene.
Thirdly, it is impossible to ignore the role played by neighbouring
states, which intervened on a large scale in the conflict through their
Afghan proxies. Such proxies were Jombesh, which acted for Uzebkistan; Hezb-i Wahdat, which served the purposes of Iran; and
Hezb-i Islami, and later the Taliban, which acted for Pakistan. Meanwhile, Jamiyat-i Islami developed its relations with India, Iran, Russia
and the western countries. Relations between outside states and
their Afghan clients were based on mutual instrumentalisation. The
three regional powers had their own divergent interests, and rarely
exerted any pressure towards a peaceful solution. However, their
intervention altered the balance of power, and influenced the movement towards the concentration of resources. For example, Rashid
Dostum was able to return to Mazar-i Sharif in 1997 thanks to help
from Uzbekistan, and after the fall of Kabul in 1996 the survival of
Masud depended largely on Russian, Iranian and western aid. However, although all the Afghan actors benefited from external assistance, Pakistan was the only state to have invested, first in Hezb-i
Islami and then in the Taliban, to an extent which was in the end
crucial in precipitating the transition towards the concentration of
resources. Without the support of Pakistan the presidential coalition
around Rabbani would probably have prevailed in 1992. Pakistans
logistical assistance played a key role in all the Talibans military operations; For example, the counter-attack which led to the capture
of Herat in 1995 would not have been possible without its material
support. In addition a significant section of the Taliban fighters
about a third during the offensive in the spring of 1999were foreign
fundamentalists, either Pakistani or having passed through Pakistan.
The process of centralisation took place in several phases. The collapse of the state in 1992 created a fluid situation which accelerated
the competition for resources and their accumulation in fewer hands.

Competition and the Impetus towards Monopoly (19922001) 237

Between 1992 and 1995 the political distribution on the ground
was dramatically simplified through the process of regionalisation of
forces. The rise of the Taliban brought with it a polarisation between it and the other parties, which were unable to do more than
delay their own decline.

The collapse of the state

In August 1991 the attempted coup against Gorbachev irreversibly
destabilised the Kabul regime, which could no longer rely on Russian aid.3 Yeltsin was all the more anxious to distance himself from
the Soviet legacy because of the support he derived from the afghantsy
(veterans of the Afghan war): men such as Alexander Rutskoy, who
opposed all support for Najibullahs regime. Inside Afghanistan perceptions also changed. The diplomatic efforts of the UN ceased to
be relevant, and within the government many began to anticipate
the post-Najibullah period. The royalists, Hezb-i Islami and Jamiyati Islami all increased their contacts with the communists. Competition between rivals got under way, to see who could profit most
from the change of regime.
In the end Najibullah fell victim to the growing autonomy of the
militias, which had become the principal military force in the north:
ironically this was a development he himself had fostered when he
was the head of the KHAD. The immediate reason for conflict was
his encouragement of the Pushtuns, a strategy aimed at sidelining
Karmals supporters, who were mainly Persian-speakers. At the end
of February the appointment of a Pushtun to a command in the
province of Balkh provided the pretext for a revolt. Mumin, the
commander of the base at Hayratan near the Amu Darya, refused to
obey the governments orders. The garrison at Mazar-i Sharif, and
then the north-west provinces, rose in a rebellion headed by Rashid
Dostum, who declared himself the leader of a new party, the Jombesh-i
Melli (National Movement). It quickly became apparent that Rashid
Dostum was in collusion with Masud and the Hezb-i Wahdat to
overthrow the regime.
On 18 March 1992 the mujahidin entered Mazar-i Sharif without
a fight, and on the same day Najibullah announced that if a neutral
3

In September 1991 the Soviet government announced that from 1 January 1992
it would deliver no more arms to the Kabul regime.

238 Competition and the Impetus towards Monopoly (19922001)

government could be installed he would resign. With the backing
of the rebels, Masud then began his advance on the capital, where
Rashid Dostums militia, already installed, would give him crucial
support. On 24 March 1992 Hezb-i Watan split up, and power devolved into the hands of a number of party officials: Farid Mazdak
(deputy Defence Minister), Abdul Wakil (formerly Foreign Minister), Nabi Azimi (head of the Kabul garrison) and Rashid Dostum
(head of the northern militias, who remained in Mazar-i Sharif). On
14 April Masud took Charikar without a fight, then Jabal ul-Seraj
and Bagram air base. The road to Kabul was now open.
The following day, 15 April, Najibullah attempted to flee to India,
but was prevented by Dostums militia, who were in control of the
airport. He took refuge in the UN offices, which were close to the
French embassy.4 On 16 April he was stripped of his power, and a
council composed of members of Hezb-i Watan was set up to cover
the interim period. On the same day the Interior Minister Gholam
Faruq, Yaqubi the director of the WAD, and his deputy General Baqi
were all killed by unknown assailants.
As the old regime collapsed the race to Kabul gathered speed. On
17 April Wakil, the former Foreign Minister, held a meeting with
Masud at Charikar to discuss the transfer of power. On 17 and 18
April government aircraft bombed Hezb-i Islami to the south of
Kabul, aiming to slow its advance. On 18 April Masud surrounded
Kabul to block the approach of Hezb-i Islami, which was coming
closer to the capital in spite of a rearguard action by Dostums militiamen. On 22 April Benon Sevan met Masud at Charikar, but it was
by then too late to agree on a transitional government. On 23 April
the mujahidin entered the capital without a fight, but several hours
later there were clashes between Hezb-i Islami and other groups,
while in Peshawar the parties attempted to reach agreement on the
formula on which a new government could be based.
On 24 April the Peshawar accords,5 agreed by the seven parties,
provided for the setting-up of an interim government of fifty-one
4

Najibullahs family was already in India, which said it would accept him as a refugee without granting him political asylum. Though he had several opportunities
to leave Afghanistan, he remained until he was murdered by the Taliban after the
fall of Kabul in September 1996.
5 The text of these accords can be found in an appendix to A. Saeeds article Afghanistan, Peshawar and After, Regional Studies, (Islamabad) 11 (2), 1993, pp. 10358.

The collapse of the state

239

members under the leadership of Mujaddidi, to hold power for a

period of two months; after which Rabbani would head a leadership
council for a further four months. The Prime Minister-designate was
Hekmatyar (Hezb-i Islami), while the Defence portfolio was assigned
to Masud ( Jamiyat-i Islami), Foreign Affairs to Gaylani (Mahaz),
Justice to Muhammad Nabi (Harakat-i Enqelab), Education to
mawlawi Khales (Hezb-i Islami Khales), and the Ministry of the Interior to Sayyaf (Ettehad). The Shiites rejected the agreement because their demand for a quarter of the ministerial positions was not
met. On 28 April power was transferred from the former government to the interim council. Mujaddidi became president of the
Islamic Republic of Afghanistan whose new black, white and green
flag bore at its centre the shahadat, the Muslim profession of faith.
The new regime was recognised by Pakistan on the same day, followed by the European Economic Community and the United
States. The following day, 29 April, Masud entered the capital with
an army of several thousand men and a hundred tanks.
During this transitional period the real loser was Hezb-i Islami.
Under pressure from the other parties it rapidly lost its vantage
points. On 29 April it was dislodged from the Ministry of the Interior that had been handed to it by the Khalqis. On 1 May an interparty organisation, the Committee for the Security of Kabul, put in
place a force of 1,600 men, ostensibly to prevent inter-party conflict
and banditry but in reality to block any offensive moves on the part
of Hekmatyar. On 2 May 1992 Hekmatyar demanded the withdrawal of Dostums men, who werewith justificationaccused of
looting. Dostums refusal gave Hekmatyar the pretext to launch the
first bombardments of Kabul, which by 1996 would claim 40,000
lives and were to destroy much of the capital, which till then had
been intact.6 Meanwhile, on 5 May the parties formed a government
of thirty-two members, not in conformity with the Peshawar accords,
in which Hezb-i Islami was not represented. On 7 May Masud, who
still held the Defence portfolio and had been appointed a general,
held a press conference in which he violently criticised Hekmatyar.
6

US Committee for Refugees, World Refugee Survey, 1997, pp. 1245. Hezb-i
Islami, though it in fact initiated these clashes, was only one of the groups which
in due course bombed Kabul, though it bombed more systematically due to its
inability to enter the city.

240 Competition and the Impetus towards Monopoly (19922001)

An attempt by commander mawlawi Haqqani to reconcile the two
adversaries failed.
At the same time, relations between the parties worsened. Dostum,
named as a four-star general by Mujaddidi, did not withdraw from
Kabul, where he continued to hold strategic positions, namely Bala
Hisar fort, the airport and a number of ministries. In addition, on
21 June Mujaddidi asked for the extension of his mandate, although
the basis for his argument for this was dubious, and his request was
refused by the other leaders. In the end he returned to the provisions
of the Peshawar accords and stepped down as planned on 29 June,
handing over to Rabbani. However, Rabbani breached the letter of
the accords and hung on after December. The absence of consultation between the parties, the antagonism between the commanders,
and the failure to include the Shites all conspired to set the scene for
a new phase of the civil war.

The regionalisation of the forces (19921995)

During the 1980s parties such as Hezb-i Islami (Khales), Jamiyat-i
Islami and Ettehad enlarged their geographical base, but after the Soviet withdrawal a contrary phenomenon was in evidence. Regionalisation accelerated after the fall of Kabul, when the headquarters of
the various parties were transferred to Afghanistan, since the scene
was set by the capture of the towns for a significant concentration of
resources. Distribution on the ground take increasingly the character
of political regions, each of which was dominated by a single actor,
whether a party, a commander or a coalition of commanders. In some
instances there was instability, as in Kunduz, or even anarchy, the
situation in Kandahar, but clashes were localised and involved only
limited forces. No organisation was truly represented at the national
level, since party discipline was generally less strong than local solidarity. For example, Masud and Ismail Khan belonged theoretically
to the same party, but for the most part behaved as autonomous or
even competing powers. Finally, the southern regions were more
fragmented than those of the north.
The situation in Kabul developed in a very different way from the
provinces. At the provincial level a new political distribution emerged
after the entry of the mujahidin into the towns, between 17 and
22 April. In Herat, Ismail Khan controlled the local administration at

The regionalisation of the forces (19921995)

241

the head of a coalition of commanders. Those militias which were

still represented in the town were dissolved some months later.
Civil order was consolidated, and the process of reconstruction was
embarked on with some success, helped by the return of refugees
from Iran. In Kandahar the town was partitioned between commanders of different partiesMahaz, Jebhe and Jamiyat-i Islami
who took concerted action to expel Hezb-i Islami from the urban
area. Afterwards the situation degenerated into anarchy, which prevailed till the arrival of the Taliban. In the north the town of Mazari Sharif was partitioned between Hezb-i Wahdat, Jombesh and
Jamiyat-i Islami, an unstable situation which continued until the alliance between Jombesh and Jamiyat collapsed in January 1994.
Jamiyat, which was militarily much reduced, was later driven out of
the town. In Jalalabad, Hezb-i Islami (Khales) established itself in a
position of strength. haji Qadir, the governor of the town, avoided
becoming involved in national affairs, probably because of his involvement in drug trafficking, together with the internal tensions
within the tribal coalition from which he drew his support. Harakat-i
Enqelab was dominant in Ghazni. Hezb-i Islami succeeded in keeping strongpoints only in Laghman and Charasiab, south of Kabul.
Hezb-i Wahdat controlled most of Hazarajat.
Between 1992 and 1994 the capital was the objective of most of
the fighting. With the exception of Ismail Khan, all the actors of
national importance maintained a military presence there with the
intention of influencing future developments. Two alliances came
into being, of which the dominant one supported the government
constructed by Rabbani and Masud around Jamiyat-i Islami. The
other, in the minority, grouped together opposition figures, with
Hezb-i Islami as its rallying point. The two alliances were differently
structured since only the one built around Jamiyat-i Islami was capable of governing, largely as the result of its dominant position in
Kabul. The advantage possessed by Jamiyat-i Islami, in addition to its
strength on the ground, was its acceptability in the eyes of the other
parties except for Hezb-i Islami, which was thus fated to remain in
opposition. Hezb-i Islami was used by other movements which
allied themselves to it if they were unable to find a place within the
governing coalition, or to strengthen their bargaining positions, but
left as soon as they were able, since its hegemonic ambitions meant

242 Competition and the Impetus towards Monopoly (19922001)

that nobody wished to see Hezb-i Islami actually achieve power.
The other major opposition was that between Hezb-i Wahdat and
Ettehad due to the latters extreme anti-Shiite position.
In addition to the hostility of Hezb-i Islami, a further basic factor
holding back the reconstruction of the state consisted of the policies
of Jombesh and Hezb-i Wahdat, each of which aimed to prevent any
coalition achieving sufficient power to reconstitute a central authority. For differing reasons each of these two parties feared political stabilisation. On the one hand, the former communists of Jombesh
knew that they lacked the legitimacy that would have enabled them
to survive beyond the end of the war. On the other hand, Hezb-i
Wahdat demanded recognition for an autonomous entity in Hazarajat, and therefore a de facto federalism. This rejection of a unitary
state arose from the desire for political emancipation by a minority
which for a century had been a pariah in Afghan history, as well as
from the fear of domination by the anti-Shiite fundamentalists of
Ettehad and Hezb-i Islami (Khales).
Without going into the detailed chronology of the events of
19924, three phases may be distinguished. The first saw the formation of a coalition between Jombesh-i Melli, Jamiyat-i Islami and
Hezb-i Wahdat. The common interest of these three actors was to
forestall the settlement plan taking shape under the auspices of the
UN. For Masud control of the capital enabled him to oust Hekmatyar and ensure that Jamiyat-i Islami would play a central role in any
new balance of power. In Dostums estimation Najibullahs disappearance would allow him to take power in the northwest provinces and win a place in the new political system. Meanwhile, Hezb-i
Wahdat was taking the opportunity to gain a foothold in Kabul
where there was a substantial Hazara community.
The breakup of this first alliance came in the autumn of 1992,
when Masud launched an operation in Kabul to disarm Hezb-i
Wahdat, whose substantial depredations against the population served
as a pretext for Masud to break an alliance which prevented him
from bringing the Sunni parties into government. As soon as Hezb-i
Wahdat ceased to be in the coalition, support for the government
was forthcoming from Harakat-i Enqelab, which had hitherto been
undecided, as well as from Ettehad, which at this point broke its links
with Hezb-i Islami. On 11 February 1993 Jamiyat and Ettehad

The regionalisation of the forces (19921995)

243

launched a joint assault on Hezb-i Wahdats headquarters, sited in

the Polytechnic School in the suburb of Afshar. Hezb-i Wahdats
forces retreated, and the area was looted for twenty-four hours, with
many civilians being murdered.
In the spring, as the result of pressure from Pakistan, Saudi Arabia
and Iran, the parties reached a further agreement, which was signed
on 7 March 1993. Rabbani was to remain President till June 1994
and Hekmatyar regained the position of Prime Minister. However,
this agreement did not outlive the launch of assaults against the government by Hezb-i Wahdat, as well as an attack on Rabbani as he
went to meet Hekmatyar. In April there was a major onslaught by
Hekmatyar on government positions at the very moment when the
parties were negotiating a new accord, ratified on 18 May, which officially installed Rabbani as Minister of Defence in place of Masud.
For the rest of the year military activity was less intense, though increasing pressure was put on Hekmatyar by the successes of Ismail
Khan in the south, where he took Helmand, and of Masud in the
Shamali plain. However, the growing imbalance in favour of the
government unsettled Rashid Dostum, who concluded during a
visit to the capital in July that the governments increasing power
would reduce its need to rely on his militia.
The third phase began in January 1994 with a change of allegiance
of Jombesh-i Melli from the government to the opposition side. For
some months the position of neutrality adopted by Jombesh in confrontations between Hezb-i Islami and Jamiyat-i Islami in the vicinity of Kabul had foreshadowed this reversal. In addition Jombesh had
been involved in fighting in the autumn with Ettehad in Kunduz
and with Ismail Khan in Badghisclashes that had exposed the lack
of consistency in Jamiyat-i Islamis policy. Attacks were in fact launched on Dustom at precisely the moment when Masud needed his
aircraft around Kabul to assist in an offensive against Hezb-i Islami.
In the end Jombeshs decision was prompted by the attitude of Hezb-i
Wahdat, which in the autumn moved closer to Masud and even had
the use of Masuds helicopters to move its headquarters to Kabul.
The prospect began to loom of a real marginalisation of Hezb-i
Islami in the vicinity of Kabul, which negated all justification for the
presence of Dostums men inside the capital.
On the opposition side Hezb-i Islami failed entirely to halt its
marginalisation. In the summer of 1992 the offensive by Dostum

244 Competition and the Impetus towards Monopoly (19922001)

and Masud came close to sweeping Hezb-i Islami from the south of
Kabul, but it was temporarily rescued by the ceasefire imposed by
Pakistan on an indecisive Rabbani; it then bombarded Kabul, hoping to weaken the government coalition. The rockets fired at the
capital had no military objective, but were intended solely to prevent a return to normality which could have included the installation of an effective government.
To break out of its isolation Hezb-i Islami attempted to reach
agreements with other parties. The first alliance was with Ettehad,
which had been on bad terms with Hezb-i Islami during the war,
and was purely a matter of expediency; Ettehad had launched more
attacks against the Shiites, their primary enemy, than against Masuds
forces. Although its bombardments were destabilising for the governing coalition as a whole, its links with Jamiyat-i Islami were never
completely broken. Hezb-i Islami and Ettehad also had different objectives with each in effect fighting its separate war. For Ettehad the
main priority was to dislodge Hezb-i Wahdat,in order to take its place
in the governing coalition; it was under no illusions as to Hezb-i
Islamis ability to take power, and did not wish to remain in the opposition. The second alliance, between Hezb-i Islami and Hezb-i
Wahdat, was even more temporary and was never formalised; it could
appropriately be viewed not as a political agreement between the
two parties but as a cease-fire with some logistical cooperation. This
alliance had no long-term future since the Shiites retained a significant enmity against Hezb-i Islami. A third alliance, between Hezb-i
Islami and Jombesh, resulted from the rapprochement between
Hezb-i Wahdat and Masud in the autumn of 1993 which precipitated Jombeshs defection. This alliance was also informal and ideologically incongruous, and lasted until the point was reached when
Hezb-i Islami, which had been politically isolated for some months,
lost its military and political power.
The sapping of Hezb-i Islamis strength in Kabul prompted a shift
in Pakistani policy, and in an unexpected way resulted in a radical
upset in the war. Hekmatyars failure at Kabul was also a failure for
the Pakistani intelligence services whose head, Lieutenant-General
Naved Nasir, had been appointed by Nawaz Sharif. After the deposition of Nawaz Sharif by President Ghulam Ishaq Khan in 1993,
Naved Nasir was dismissed along with several dozen other officers,

The regionalisation of the forces (19921995)

245

and General Nasrullah Babar,7 Benazir Bhuttos new Minister of

the Interior, took charge of Pakistans Afghan policy. At this stage,
though Hezb-i Islami no longer enjoyed the status of a privileged
ally, Nasrullah Babar had not yet clearly formulated his Afghan policy. In an interview in the Frontier Post on 30 April 1994, for example, he said he was in favour of the return of the king, but in fact the
primary objective of the Pakistan government was to establish a land
link to Turkmenistan in order to open up Central Asia to Pakistans
economic and political influence. In September 1994, at a ceremony
attended by Nasrullah Babar as well as the Pakistani ambassador in
Afghanistan and Ismail Khan, the governor of the town, a Pakistan
consulate was opened in Herat. On the practical side the Pakistan
government dispatched a convoy of thirty trucks to Turkmenistan
in November. However, it was halted near Kandaharat Ghar Killi
in Takhta Pulby several commanders who included notably amir
Lalay and Sarkateb from the Hezb-i Islami.
Some months before Nasrullah had stationed a small military force,
known as the Taliban, on the Afghan-Pakistani frontier to keep the
road clear in the event of any commander presuming to obstruct the
freedom of movement to Herat. The Taliban had their origin in the
Jamiyat-i Taliban (Society of Taliban), a subsidiary of Harakat-i Enqelab, which came into being officially around 1990 but had earlier
existed informally.8 Since Harakat-i Enqelab was at that time in the
process of collapse, a group of its commanders, and in particular Qari
Baba and Rasul Akhundzada, gravitated towards Rabbani, while
others moved closer to Ismail Khan. The Taliban had very close relations with the Pakistani Jamiyat-ul Ulema, which at that time supported the government of Benazir Bhutto, and this connection
facilitated their link with Nasrullah Babar. In the spring of 1994,
after training in ISI camps, the Taliban launched a successful attack on
Spin Boldak,a frontier post on the Quetta-Kandahar road.It preached
respect for Islamic morality to passing travellers, and obliged some to
7

Nasrullah Babar, who is a Pushtun from the NWFP and a former governor of
that province, had been in the 1970s the chief adviser on Afghan affairs to
Zulfikar Ali Bhutto.
8 The first attempt to make use of religious students may have been made as early as
19923 when General Durrani was head of the ISI. Benazir Bhuttos assumption
of power had ended this experiment.

246 Competition and the Impetus towards Monopoly (19922001)

cut their hair. They also stood out against the extortion practised by
the other parties.
The emergence of the Taliban coincided with a serious social crisis in southern Afghanistan. In Kandahar between 1992 and 1994
aimless quarrels between gangs broke out in the bazaar, while armed
men openly smoked opium in the streets. No administration was in
control of the town, and the schools were closed. Politics was in disrepute, to the profit of religious activists, since adherence to the
shariat appeared preferable to the dissension ( fitna) resulting from the
strife between the various parties. In such a situation of moral and
social crisis, allegiance to a charismatic person or movement enabled
social relations to be rebuilt especially in a segmentary society where
mobilisations around charismatic personalities were historically commonplace.9 In this situation the Taliban appeared to present an alternative to disorder. Its programme was based on the restoration of
order: freedom of movement and of trade, the end of banditry, a ban
on drug use, and so on.10
The commandeering of the Pakistani convoy gave the Taliban the
opportunity to mount a major operation aimed at Kandahar itself.
On 5 November, after a number of clashes, it succeeded in occupying the town, and by the end of November entered the neighbouring provinces of Helmand and Zabul. At this point the Taliban were
still regarded as allies by Rabbanis government, which may explain
the attitude adopted by the Jamiyat commander Naqibullah, which
in practice enabled it to take Kandahar. In January 1995 the Taliban
made a deal with Ghaffur Akhundzada for the cession of Helmand,
after which it found itself in contact with Ismail Khans troops.11 The
movement then pressed on towards Kabul, sweeping aside commanders who had been in place since the beginning of the war
such as Amin Wardak, who fled to France, and Qari Baba, the leader
9

For an anthropological analysis of charisma in politics in the Pakistani tribal

zones, see Charles Lindholm, Frontier Perspectives: Essays in Comparative Anthropology, Oxford University Press, 1996, pp. 73 ff.
10 This did not prevent the Taliban at that time from authorising the production of
opium, justifying it with the sophistry that only its consumption is forbidden by
the Quran.
11 In 1994 the Taliban sent a mission to Herat to request Ismail Khan to apply
Islamic law. A group of ulema from Herat went to Kandahar to reassure the
Taliban that the shariat was highly respected in Herat.

The regionalisation of the forces (19921995)

247

Simplified political map, 1992.

248 Competition and the Impetus towards Monopoly (19922001)

The regional distribution of political forces, 1996.

The regionalisation of the forces (19921995)

249

Territory controlled by Masud, March 2000.

250 Competition and the Impetus towards Monopoly (19922001)

of the shura at Ghazni, who rallied to Masud. The fall of Ghazni
came at the end of January 1995 following a Hezb-i Islami offensive
aimed at halting the Talibans advance. During his resistance to this
attack the governor, Qari Baba, actually made an alliance with the
Taliban, at a point when government aircraft were also attacking
Hezb-i Islamis positions.
In February 1995 the Taliban took Meydan Shahr, Pul-i Charkhi
and most significantly Gulbuddin Hekmatyars principal base at Charasiab, near Kabul. Within a few months its advance had brought
about a major simplification of the political map, with the disappearance or marginalisation of the predominantly Pushtun parties
Mahaz, Jebhe, Ettehad, Harakat-i Enqelab and Hezb-i Islami. The
town of Kandahar had been the only important stronghold belonging to Mahaz, for which Nangrahar was of less significance. Mahaz,
the only party favouring the return of the king, was of only marginal
importance in the overall balance of forces, and would not henceforth play any significant role. Similarly Harakat-i Enqelab was also
merged into the Taliban. Hezb-i Islami, cut off from Pakistani aid,
was reduced to a remnant, as was Ettehad. The regionalisation of forces was now complete and a new phase was set to begin.

Towards monopolisation
In due course it became clear that regionalisation was only an intermediate step in the process of the concentration of resources, as the
Taliban, by reason of its demographic preponderance and the divisions among its adversaries, as well as its access to aid from Pakistan,
came to dominate virtually the whole territory.
However, in Kabul itself, devastated as it was after three years of
fighting, Masud made his strength felt, and a reunification of the city
under his authority appeared a possibility. To the south of Kabul,
Hezb-i Islami was in full flight before the Taliban advance. Inside the
capital the Shiites of Hezb-i Wahdat were on the point of collapse,
under pressure from the coalition of Masuds forces with Harakat-i
Enqelab and Ettehad. In early March 1995 the government coalitions shelling of the Shiite areas was at its height, and the western
part of the city was largely razed to the ground. Hoping to escape
defeat, the Hezb-i Wahdat leader Mazari allowed the Taliban to
enter the city, but on 19 March clashes broke out between the two
groups. Mazari was taken prisoner but died in unexplained circum-

Towards monopolisation

251

stances while being transferred to Kandahar.12 Hezb-i Wahdat then

split into two factions. One, under the leadership of Akbari, entered
an alliance with the Rabbani government and established itself
at Bamyan. The other, headed by Karim Khalili, allied itself with
Dostum in June 1995. In southern Kabul, after a brief period of
fraternisation, the presidents men and the Taliban clashed when the
Taliban rejected Masuds demand that it should disarm. Masud won
back some territory, while the Taliban, which had suffered heavy losses,
fell back towards the south. Masud was then in sole control of the
capital, and pressed forward as far as Meydan Shahr, which became
for some time the front line.
After pulling back from Kabul, the Taliban then faced a threat
from the west, where Ismail Khan launched in early March an attack
which carried him to within 200 km of Kandahar, the Talibans principal stronghold. At the beginning of April 1995 Dostum launched a
successful attack against the province of Badghis, north-west of Herat,
as part of a pincer movement against Herat. However, the Taliban
lost a battle at Shindand where fierce fighting left hundreds dead. Its
only achievement during the spring of 1995 was in May on the
Kabul front when it ousted Sayyaf from his base at Paghman.
In the summer Ismail Khan launched a new offensive aimed at
Kandahar. In August he took Girishk, but the military situation underwent a spectacular reverse on 28 August when the Taliban recaptured the town and then the air base at Shindand. Ismail Khans
demoralised mujahidin fell back to the neighbourhood of Herat and
in the ensuing chaos dispersed into small groups. On 4 September
Ismail Khan relinquished his command to his deputy, Allauddin
Khan, before crossing over the Iranian border. During the night the
town was abandoned. On the following day, 5 September, Herat fell
into the hands of the Taliban without a fight.
There were many reasons for Ismail Khans collapse.13 The principal factor, which was entirely fortuitous, was the death in early Sep12

The helicopter taking Mazari to Kandahar stopped at the Baraki Barak hospital
to drop off wounded men. According to a witness, Mazari was at that point unchained and did not appear to have been beaten. The Taliban versionthat
Mazari was killed at Ghazni while trying to escapeseems therefore to have
some credibility.
13 See Muhammad Zahir Azimi, Taliban chegune amadand? (The Taliban: how did it
arrive?), no publisher, 1998. Muhammad Zahir, who formerly belonged to the

252 Competition and the Impetus towards Monopoly (19922001)

tember of an important commander Dr Nasir Ahmad Yar during
the fighting for Girish. There were also the operations against Herat
by Rashid Dostums militias, aimed at helping the Taliban to victory.
In addition Ismail Khans forces were fragmented and out of control. The mujahidin sent by the Kabul government in the spring of
1995, most of them Panjshiris, were especially reluctant to accept
Ismail Khans authority. In addition, Ismail Khans autonomous ambitions put him in opposition to Rabbanis government. For instance,
shortly before the fall of Herat a commission sent out by Kabul failed
to achieve a compromise, particularly over financial issues, with Herat
refusing to pass money on to the central government in spite of the
substantial revenues it derived from the customs post on the Turkmen frontier.14 Another bone of contention was the nomination of
governors to the western provinces, who were appointed by Ismail
Khan without consultation with Kabul. Ismail Khans attack on
Dostum in the spring of 1994 ran counter to the interests of Masud,
whose aim was above all to maintain his control over the capital.
Similarly the meeting held at Herat in early August 1994 was the
preliminary to a shura scheduled for November. Rabbanis authority
was directly challenged when the meeting demanded the ending of
his term of office. In the last resort Ismail Khan may have wished to
present himself as a compromise candidate between the parties. It
must be added that in more purely military terms the offensive against
Herat was mainly the brainchild of the Pakistan intelligence services, from which originated the superior logistics and enhanced
mobility which foreshadowed the later offensives in the north.
This defeat was a turning-point in the war. The Taliban was no
longer trapped between Masuds forces and those of Ismail Khan,
and was able to deploy its forces with greater ease. With its capture of
Shindand the Taliban gained an air base of prime importance and
also some ten aircraft. Also the road from Pakistan to Turkmenistan was
Rad, was the commander of Harakat-i Islami in Herat and worked with Ismail
Khan throughout the war. Accusations of treason made at this time against
Ismail Khan by Masuds associates were unfounded.
14 One member of this commission, Delju Huseini, a Herati who had been a
member of the Islamic movement before the war, had notoriously bad relations
with Ismail Khan. This serves as an illustration of the antagonism between the
Islamic networks before the war and the officials of the Emarat.

Towards monopolisation

253

opened, which brought great economic and political benefits. Finally,

on 8 September 1995, the capture of Herat sparked off an unprecedented crisis in Afghanistans relations with Pakistan when the Pakistani embassy in Kabul was plundered. In December 1995 there was
a further wave of incidents in Pakistan, when 65 bombs left 78 dead,
for which the Pakistani government held Masud responsible.
The limits of Masuds strategy were exposed when the priority he
gave to maintaining control over Kabul prevented him from pushing
outward into the northwest,15 while at the same time he was unable
to secure a military victory in the south because of the Pushtuns unwillingness to put themselves under the control of a predominantly
Panjshiri army. Ismail Khans strategy, which was predicated on a victory over Dostum, was probably in the end more fruitful, even
though Jombesh could call on formidable military strength.16 Lastly
Masud could not have foreseen the Talibans success, while his initial
support for the movement, which was essentially verbal, was based
on the antagonism between him and Hezb-i Islami. Another consequence of the advance of the Taliban was the defeat of the Rabbani
governments Pushtun allies, and in particular Qari Baba in Ghazni.
Masud also paid the price for his expansionist strategy, which led
many commanders to group themselves together in opposition to
him, including even some commanders within Jamiyat itself. Analyses suggesting that Masud was a good soldier but a bad politician are
probably based on a misunderstanding of both the constraints which
limited his freedom of action and the fluidity of the situation. Whatever the truth of the matter, Jamiyat-i Islamis position was seriously
undermined. The internal balance of the party shifted, with Masud
becoming the sole leader, while Rabbani was left with only regional
support, principally in his own home province of Badakhshan.17
15

In theory this would have been possible with the assistance of the Jamiyat based
in Mazar-i Sharif, but it was militarily difficult and politically fraught, since earlier attempts by Masud to gain a foothold in this region had been badly received.
16 During his visit to Kabul in June 1994 Ismail Khan made no secret of his refusal
to make any contact with Dostum, although he agreed to meet Hekmatyar.
17 Rabbani was no longer able to play the Jamiyat commanders off against each
other as he had hitherto done, for instance by offering the post of Minister of
Defence to Ismail Khan in 1992 while he was in Herat: an offer which Ismail refused (personal information from Stphane Thiollier).

254 Competition and the Impetus towards Monopoly (19922001)

From then on Kabul became the Talibans target. The first stage
was the capture on 12 September 1996 of Jalalabad after a brief
battle.18 The Taliban takeover had in fact been negotiated with a
group of local commanders a few months earlier; the city was by
now difficult to defend and fell into the Talibans hands on 26 September. In order to conserve his strength, Masud refrained from resistance, retreating in good order to Panjshir, easily defeating at the
same time an attempt by the Taliban to pursue him. From this point
on, conquest of the north was within the grasp of the Taliban, which
already controlled around two-thirds of the country and the larger
towns, with the exception of Mazar-i Sharif. The new balance of
power obliged Jamiyat, Hezb-i Wahdat and Jombesh to form an alliance to counter the Talibans thrust. However, this did not imply a
political transformation since, far from being a homogeneous political
grouping, the alliance brought together parties which were opposed
to each other in every wayin their ideology, their constituency
and their political ambitions.19 Military coordination between them
was largely in the realms of theory. This coalition underwent its first
serious defeat with the capture of Mazar-i Sharif by the Taliban on
24 May 1997.20 The Taliban took advantage of the treachery of
the commander of Maymana, Malik, whose motive was vengeance
for the death of his brother, assassinated by Dostum on 25 June 1996.
Once within Mazar-i Sharif, however, the Taliban reneged on the
agreements it had made with Malik, in particular its request for
representation on an ethnic basis together with a decentralised system. Consequently Malik repudiated the terms of the alliance and
inflicted serious losses on the Taliban, massacring 23,000 Taliban
18

Haji Qadir, who led the Jalalabad shura, fell from power and retreated to Pakistan,
where he organised two offensives against the Taliban at the end of 1996 and in
February 1997. Pressure from Pakistan afterwards prevented him from continuing his operations.
19 On 21 August 1997 the death in an air accident of Abdul Rahim Ghaffurzai,
who had been designated Prime Minister a week earlier, was a further blow to
the northern coalition. Ghaffurzai was a former representative at the UN under
Najibullah and a member of the Muhammadzai clan. He would have been able
to bring greater cohesion to the Northern Alliance and could have put them in
touch with the Pushtuns.
20 Ismail Khan, who had returned from Iran to fight the Taliban, was at this time
taken prisoner. He later escaped after two years of detention.

Towards monopolisation

255

prisoners.21 In September a second Taliban attack would have been

successful except for a massive delivery of Iranian aid. On this occasion it used Kunduz as its main base, and was joined there by the
Pushtun commanders which enabled it to obstruct all attempts by
Masud to reach Mazar-i Sharif.
Although Dostum retook Mazar-i Sharif in November, Maliks
treason had weakened Jombesh. The decisive attack, which set the
seal on the Taliban victory, began in the spring of 1998. The capture
of Faryab province and then Samangan opened the way to Mazar-i
Sharif, which fell in August 1998. In revenge for the massacres of the
previous years, several thousand mainly Hazara civilians were killed
and their bodies were left lying in the streets for days. In a further
incident the murder of a number of Iranian diplomats stationed in
Mazar-i Sharif led to the mobilisation of Iranian troops on the Afghan
frontier. Another consequence of the capture of Mazar-i Sharif was
the disappearance of Hezb-i Wahdat as a major military force in the
north. Sporadic fighting broke out in May 1999, especially around
Bamyan and Sar-i Pul, but a significant part of Hezb-i Wahdat
joined up with the Taliban. Karim Khalili, however, stood by his alliance with Masud.
The confrontation between the Taliban and Masud had by this
time become a clash between a centralised power that was steadily
consolidating itself and a guerrilla force entrenched in the mountains.
Masuds control extended over less than 10% of the population and
none of the towns. It was difficult for him to enlarge his territory
since he had at his disposal only some 6,000 men, which precluded
operations of any size. The Taliban also had control in the air. However, Masuds organisational ability and tactical skill enabled him to
hold out. As a result of underestimating him the Taliban failed to
consolidate its summer offensive to the north of Kabul in 1999,
when Masud counter-attacked brilliantly. Masuds predicament was
the reason why he paid particular attention to the international dimensions of the conflict. For example, the recognition of the Rabbani
government by the UN was one of the last cards available for him to
play. Masud was also able to get help from the regional powers
Iran, Uzbekistan and Russiato which a Taliban victory appeared
21

Amnesty International, Afghanistan: report of mass graves of Taliban militia, 1997, AI

index 11/11/97.

256 Competition and the Impetus towards Monopoly (19922001)

as a threat to the stability of Central Asia. Meanwhile the coup of 12
October 1999 against Nawaz Sharif and the return of Pakistan to
military rule under General Pervez Musharraf brought no marked
alteration in Pakistans Afghan policy.
Finally, the elimination of the opposition was also being carried
out outside Afghanistan, with assassinations of Afghan personalities
who were either involved in the search for a negotiated solution or
opposed to the Taliban. The leftists, the nationalists and the royalists
were the first to be targeted.22 In July 1999 the assassination in
Quetta of Abdul Ahad Khan Karzai, a former diplomat and senator,
coincided with a new bid by the royalists to implement a compromise solution involving the king. A member of Sebghatullah Mujaddidis family was also attacked, and in addition Afghan women
working in the refugee camps were threatened. Similarly, the fear of
attack by the Taliban obliged the leftist feminist association RAWA
(Revolutionary Association of Women in Afghanistan) to cancel
planned demonstrations in Peshawar. Even commanders who had
once enjoyed Pakistani support were eliminated. The massacre of
the family of Abdul Haq in the spring of 1999 was a sign that the
Taliban was determined to finish off the remaining commanders. In
addition, Hekmatyar was refused a visa for Pakistan, in a clear signal
of the Pakistan authorities unquestioning support for the Taliban
and the Talibans rejection of any compromise.

22

See Human Rights Commission of Pakistan, State and Human rights in 1998,
Lahore, 1999, p. 293.

8. The Ethnicisation of the Conflict

As the result of regionalisation each of the parties was inevitably
driven to recruit from a constituency which increasingly comprised
a single ethnicity. Between 1994 and 1998 the political disposition
on the ground and that of the major ethnic groups largely coincided.
Hezb-i Wahdat was entirely Hazara, Jombesh was predominantly
Uzbek, Jamiyat-i Islami was for the most part Tajik, and the Taliban
drew its membership essentially from the Pushtuns. In this situation
some parties used the rhetoric of community to mobilise support;
the appeal to the jihad was no longer sufficient to guarantee support
for a group. The situation of Hezb-i Islami exemplified this transformation. To offset its marginalisation Hekmatyar increasingly turned
to the argument that Pushtun interests must be defended, an argument presented at first in less formal contexts but in due course
more overtly, as in the partys daily newspaper Shahadat.
This appeal to community solidarity was logical in view of the
much increased tension between communities manifested in massacres targeted against particular communities. The towns, and in particular the capital, had ceased to be ethnic melting-pots. In pre-war
times inhabitants of different ethnic origins had made their homes in
their own particular areas of Kabul, but the existence of the governing class and the effects of the urban way of life served to foster the
emergence of a Kabuli identity. This tendency lost its momentum
when the war began. In Mazar-i Sharif in the 1990s recurrent clashes had weakened the urban identity, so that relationships between
ethnicities deteriorated. Even in Herat, which continued to typify
the urban melting-pot where intercommunal tension was low, the
situation in the outskirts of the town steadily worsened. The collapse
of the state and of the educational system caused linguistic frontiers
to become more marked than before. In Kandahar, for example, the
Pushtuns spoke less Persian, while it became rare for Pushtu to be
257

258

The Ethnicisation of the Conflict

taught in the north. These processes were reinforced by the media;

on the radio the ethnic affiliation of the leaders was, at least implicitly, presented as a factor explaining their political positions. Such
leaders as Rabbani, who came from Badakhshan, and Masud, from
Panjshir, began to be described as Tajik.
Although the war certainly gave ethnic affiliation a political significance, this was a process with implicit limitations. Within the
operations of the parties the ideology of ethno-nationalism never
dominated, while local solidarities were always the determining factor. In any case two different processes, between which it is important to distinguish, were subsumed under the idea of ethnicisation.
In the case of Hezb-i Wahdat and Jombesh-i Melli, although ethnic
rhetoric was openly employed, the official party line remained ambiguous. Shiism and not Hazara nationalism remained the ideological foundation of Hezb-i Wahdat, while Jombesh maintained
somewhat vaguely that its doors were open to all the northern communities. On the other hand, neither Jamiyat nor the Taliban based
its strategy on ethnic affiliation but set itself instead the goal of reunifying the country. Ethnicisation was therefore an unintended and
counter-productive result of regionalisation, rather than a strategy of
mobilisation.

Ethnicisation as a strategy
Although Hezb-i Wahdat and Jombesh claimed respectively to represent the Hazaras and the Turks, the ethnic strategies of these two
parties were validated in two different ways. Hezb-i Wahdat was
constituted as a movement for the unification of the Hazara parties,
and reflected a strong popular demand for autonomy. However, for
Jombesh ethnic mobilisation was largely a political stratagem by the
Parchami communist functionaries wanting to survive the collapse
of the regime.
Hezb-i Wahdat. The launching in 1990 of Hezb-i Wahdat (Party of
Unity) was an event of critical importance not only for Hazarajat,
whose political aspect it profoundly altered, but also for the Afghan
political landscape as a whole. The ethnicisation of politics and the
process of regionalisation were already implicit within it.

Ethnicisation as a strategy

259

At the close of the 1980s popular discontent over the endless war
between the Shiite parties prompted the notables and religious
leaders to call for a general ceasefire. The palpable loss of confidence
in the parties created a climate propitious to political change, but the
crucial impetus came from Iran. The reason was that its leaders, not
having taken part in the Geneva negotiations, feared marginalisation
in the context of a Soviet-Pakistani agreement and had decided, to
make every effort to create a Shiite front able to exert its influence
in the field of politics. It was therefore no accident that the initial
preparatory meeting was held at the precise moment when the Soviet
withdrawal became a certainty, taking place at Panjao between 12 and
16 July 1988. At this meeting the decision was taken to set up a
Hasta-yi Wahdat (Nucleus of Unity), whose task would be to prepare the way for the unification of the parties. On 1 September 1988
delegates meeting at Lal o Sarjangal issued a twelve-point declaration which ratified the union between Sepah and Nasr and specified
Bamyan as seat of the future movement. Hezb-i Wahdat was officially inaugurated on 16 June 1990 in Teheran, with the participation of Shura, Nasr, Sepah and other smaller groups. Only Harakat-i
Islami persisted in staying outside the new party.
Hezb-i Wahdat set up its headquarters at Bamyan. Its leadership
consisted of representatives from Nasr, Sepah, Shura, Niru, Nehzat
and Dawat, the parties which had united. This council was unelected,
its members holding their positions ex officio. There was also a shura
with a membership of several hundred, delegated by the parties,
some of whom were elected by popular vote. With the exception of
Sayyed Hasan jaglan, who took charge of military affairs, all the leading figures were ulema, notably the president Abdul Ali Mazari, a
former Nasr leader; Muhammad Akbari, a former leader of Sepah;
and Muhammad Karim Khalili,the movements spokesman.Hazarajat
was administratively divided according to the various pre-war provinces and districts. From 1988 onward the old regimes buildings
were gradually rebuilt. In general the organisation recalled the early
days of the Shura. In theory there was military service, and taxation
was raised.
The creation of a Hazara party capable of overcoming partisan divisions and presenting a united front against the Pushtuns and the
Sunnis in general was greeted in many places by a wave of nationalist

260

The Ethnicisation of the Conflict

fervour. There was no doubt that popular support sprang from the
nationalist agenda rather than support for revolutionary Shiism.
Notwithstanding its religious rhetoric, the formation of Hezb-i
Wahdat was primarily an expression of the Hazara community, as
was demonstrated by the failure of attempts to expand the movements to all the Shiites, in particular the Qizilbash. Hezb-i Wahdat
never succeeded in enlarging its membership beyond its original
constituency of Hazaras, run by clerics;1 for example, Harakat-i
Islami refused to become part of Hezb-i Wahdat because its leader
ayatollah Mohseni was non-Hazara, as were many of its activists, and
Harakat-i Islamis relations with Iran were difficult. After fruitless
negotiations over a merger with Harakat-i Islami, Hezb-i Wahdats
leadership attempted to recruit Harakats commanders individually.
This strategy was of limited effectiveness, although it met with some
success in Ghorband and Behsud, and the two parties contrived a
modus vivendi. In addition Hezb-i Wahdat never really succeeded in
recruiting cadres other than Hazara clerics.
Jombesh. Jombesh-i Melli (National Front) was established and developed very differently from all the parties previously discussed,
since its basis was an alliance between the northern militias and the
Parchami cadres who had supported Babrak Karmal. In fact Dostum
remained close to Karmal up till his death in 1996. Fifteen years of
war had given a distinctive political personality to the northwest provinces, where the policy of national reconciliation had met with some
success. The region was structured on a framework of positions where
the government was well represented, while its political and to some
extent its economic centre was the town of Shibergan rather than
Mazar-i Sharif.The region was the only one to offer a haven to which
the militias and the members of Hezb-i Watan could withdraw after
the collapse of the government.
The militias of Rashid Dostum and of Rasul Palawan at Maymana
comprised the military basis of Jombesh.2 The Ismailis, who con1

However, some of the khans sons, with little political background, had some success in joining Hezb-i Wahdat,where their technical abilities made them welcome.
2 A list of the main leaders of Jombesh, with their ethnic affiliations is given by
Esedullah Oguz in Afghanistan, Istanbul: Cep Kitaplari, 1999, p. 26.

Ethnicisation as a strategy

261

trolled a crucial position on the road between Mazar-i Sharif and

Kabul, also gave their support to Rashid Dostum, but were inclined
to maintain their autonomy politically. In addition to the regional
militias, groups which had been militarily defeated tended to rally to
Jombesh. This was the case particularly with the militiamen of Herat
after their expulsion by Ismail Khan at the end of 1992. Numerous
mujahidin, in particular those from Hezb-i Islami, also joined. After
the disappearance of Harakat-i Enqelab in the north-west, the mujahidin mainly regrouped into the two opposing parties, Hezb-i Islami
and Jamiyat-i Islami. This polarisation encouraged Hezb-i Islamis
commanders to join Jombesh, although in Kabul these two parties
were in opposition. This was probably due more to the disintegration of Hezb-i Islami in the north than to a strategy of infiltration
prompted by Hekmatyar. Some Harakat-i Enqelab commanders,
especially in the province of Samangan, rallied to Jombesh for financial reasons. For example ras Abdul Rahman, who was appointed
governor of Faryab in 1992 without real power, was a former Harakat-i Enqelab commander. A splinter group such as that of Azad Beg,
who proclaimed his allegiance to Uzbek nationalism, also saw fit to
join Jombesh, although it had probably been created originally by
the Pakistan secret services to destabilise Soviet Uzbekistan. Jombesh
also attracted figures independent of the parties, in a direct continuation of the policy of National Reconciliation, particularly after the fall
of the Kabul government. Finally, in certain instances when offers
of membership failed, Jombesh undertook political assassinations, like
that of the responsible Jamiyat-i Islami figure in Almar, Afizullah
Fateh, while he was negotiating with Rasul Palawan at Maymana
in 1992.
Jombesh was an amalgam between the state administration and
the militias. The militia commanders were integrated into the army:
Rashid Dostum was a general, although he had no formal military
training. Within Jombesh any boundary between the regular army
and the militia lost its meaning, since the militia leaders were in the
midst of regular army officers and organised their men on the same
lines. Jombesh commanders were also established locally and in the
countryside. In practice the militia chiefs retained close links with
their villages of origin, where they had built up a clientele. There
was a sense of distance from the towns as a result of the militia com-

262

The Ethnicisation of the Conflict

manders local roots. This was naturally more true of mujahidin recruits such as Rasul Palawan than of Dostum himself, who regarded
Shibergan as his base more than he did his native village.
From the ideological point of view Jombesh was formally organised as an Islamic movement. However, the attitude of the party officials left little doubt about their real beliefs: their acceptance of Islam
was entirely pragmatic and tactical, within the continuation of the
policy of National Reconciliation. Their alliance with the Ismailis
was another measure of the sincerity of their affiliation to an Islamic
ideology. In reality Jombeshs Islam was non-political, and was restricted to law and to religious ritual. In the government madrasas the
mullahs wore western dresssuits and overcoatsand Soviet Islam
inspired the approach of the official clergy. In these madrasas future
ulema, entirely identified with officialdom, were under the strict
control of the authorities.
Although official announcements avoided all reference to any particular ethnic group, propaganda emphasised the Turkish personality of Rashid Dostum, the partys principal leader, and Dostums
travels in Uzbekistan and Turkey were represented as the result of
ethnic solidarity.3 Was such nationalism behind Uzbekistans support
for Jombesh? Uzbekistans support is explained rather, in terms of
the search for a reliable ally to guard its southern frontier and not as
ethnic solidarity between Uzbeks, which seems to have been a marginal factor. The construction of an Uzbek nationalism such as Jombesh envisaged was certainly not encouraged by Uzbekistan, which
would not have considered any challenge to its frontiers with Afghanistan. In practice, however, part of the armaments received by Jombeshs forces came from the Uzbek armys arsenal; the frontier was
relatively open and trade was lively. In addition Turkmenistan had
opened a new road to Andkhoy. Dostums aircraft sometimes made
use of the airbase at Termez, but Uzbekistans assistance had its limits:
for example fuel deliveries were not adequate.
In any case, ethno-nationalist mobilisation clashed with the complexity of communal affiliations, whence the ambiguous expression
3

In 1992 Rashid Dostum began a series of trips to Turkey which would put him
in touch with various Turkish politicians, such as the parliamentary deputy Ayvaz
Gkdemir, a former lkc (ultra-nationalist) then supporting Tansu Ciller, as
well as officials of the regime. See Esedullah Oguz, op. cit., p. 28.

Ethnicisation as a strategy

263

Peoples of the North found in some Jombesh documents. For example, because of the presence of a Turkmen minority it was the
broader Turkish rather than the Uzbek aspect on which stress was
laid. In addition Tajiks and Hazaras were represented in Dostums
militias, and this, added to the alliance with the Ismailis, necessitated
a degree of rhetorical prudence. The region of Turkestan, from Faryab
to Balkh, had never been politically unified, although the attempt at
secession made in 1888 by Ishaq Khan, governor of Turkestan, seems
to have been welcomed by the population.4 Ethnic distribution on
the ground in northern Afghanistan was extremely complex, and
did not lend itself to nationalist mobilisation.5 In essence the northwest provinces of Faryab, Jozjan and Sar-i Pul might be seen as predominantly Uzbek, with strong Turkmen minorities in Andkhoy
and along the frontier as far as the province of Kunduz. Pockets of
Uzbek population were also found in the eastern provinces of
Takhar and Badakhshan. However the idea of an Uzbek identity as
such was less of a mobilising force than the multiple tribal identities
of peoples of widely different origins and characters, making sense
only in relation to a contrasting overall identity. For example, Turkish and Mongol groups pronounced themselves to be Uzbeks when
in contention with Tajik or Pushtun neighbours.
For the Jombesh functionaries the problem was therefore to initiate nationalist mobilisation in a context where only local identities
existed. The intention was to move from communal solidarity, based
on qowms, to a more abstract loyalty oriented towards a macroethnicity identified with Jombeshs para-statal apparatus. This transition required both an ideological framework and methods of mobilisation. In the urban context the party continued to mobilise using
methods reminiscent of those of Hezb-i Demokratik-i Khalq-i
Afghanistan. Although the subject-matter might have changed, the
style of propaganda was still that of the former regime, especially as
its officials had mostly continued to carry out their functions.Jombesh
systematically used nationalist symbolism: in Shibergan the street4

Jonathan Lee, The History of Maymana in Northwestern Afghanistan, 1731

1893, Iran 35, 1987. Ishaq had a Pushtun father and an Armenian mother, but he
had been able to enlist the support of the northern population by promising
independence for Turkestan.
5 Pierre Centlivres, Lhistoire rcente de lAfghanistan et la configuration ethnique
des provinces du nord-est, Studia Iranica (Paris) 5 (2), pp. 25568.

264

The Ethnicisation of the Conflict

names were changed to demonstrate affiliation to Northern Afghanistan. The Pushtun village of Pushtun Kot was re-named Imam
Saheb, a sign of rejection of the Pushtun presence in the north. The
new street-names referred to Uzbek personalities or reverted to historical nomenclature, as in Khorasan. In education Central Asian
languagesUzbek and Turkmenwere adopted, which posed the
problem of converting to the Cyrillic alphabet, to open the door to
Uzbek and Turkmen literature. Pushtun was abandoned.
To strengthen its position Jombesh attempted to enlist the Uzbek
and Turkmen regions in northern Afghanistan to boost its ethnic
sphere and strengthen its control over the frontiers, especially in the
province of Kunduz. In this region the commanders were affiliated
mainly to Hezb-i Islami, and to a limited extent to Jamiyat-i Islami.6
Political activities such as the forging of alliances and transfer of allegiance from one party to another were undertaken for local community reasons connected with the village or the clan, without
reference to any solidarity to a macro-ethnicity. From 1992 relations
with Jombesh were established on the basis of material considerations such as gifts of petrol, vehicles and money; the commanders
maintained their link not for ideological reasons but rather to bolster
their local power with the help of a powerful ally. What led to the
emergence of an ethnic ideology was the reaction of other parties in
the province who were unsettled by the ascendancy of Jombesh.
From this point of view the clashes of the autumn of 1993when
Jombeshs new recruits, rapidly reinforced by Dostums forces who
arrived from Mazar-i Sharif, clashed with the commanders of the
provincial shurasolidified the communal boundaries. The attacks
mounted by the other parties and the looting which followed created tensions between the Turkmen communities on the one hand
and the Pushtuns and Tajiks on the other, which justified their affiliation to Jombesh after the fact. The Turkmen commanders were in
any case not taken in by this blackmail, which drove them somewhat
against their will into the arms of Jombesh, with which they had few
ideological affinities. Thus ethnically motivated enlistments to Jombesh
apparently concealed other processes more complex than nationalist
fervour.
6

In Imam Saheb Hezb-i Islami sought Uzbek rather than Pushtun recruits. Allegiance was based here on social rather than communal proximity.

Ethnicisation as an unintended consequence

265

Jombeshs nationalist mobilisation nevertheless ran up against a

series of obstacles. First, it was not able to draw on popular demands
as strong as those of the Hazaras. Even among its own functionaries,
Jombeshs nationalist rhetoric was a matter of expediency since the
majority were former Parchamis who were unlikely to have been
Uzbek nationalists by conviction. However, with the passing of years
tensions appeared, since ethnic affiliation became a determinant of
access to power. Secondly, even within Jombesh the motivating force
of the qowms continued to be stronger than either the language of
nationalism or party solidarity. The assassination of Rasul Palawan,
the Jombesh leader at Maymana, mobilised the solidarity of his
qowm, rather than strengthening Dostum, who has ordered it, so that
Rasul Palawans brother Malik entered an alliance with the Taliban
which led to the first fall of Mazar-i Sharif in 1997.

Ethnicisation as an unintended consequence

In contrast to Hezb-i Wahdat and Jombesh, the influence of the
Taliban and Jamiyat-i Islami was never intended to be limited to a
single geographical region or community. At the outset these two
movements sought to restore a central state, but political developments restricted them to recruitment in practice from a single ethnic
community, with consequences for their activities.
Jamiyat-i Islami. Although it was the most firmly entrenched party at
the point of the Soviet withdrawal, Jamiyat saw itself progressively
driven back into the northeast quarter of the country. Its decline began with the agreement between Masud and Jombesh, which had
the effect in the northwest of marginalising Jamiyat, led by mawlawi
Alam in Mazar-i Sharif. In 1994 the arrival of the Taliban in the
south led to Jamiyats loss of its Pushtun commanders, and the loss of
Herat in 1995 completed the regionalisation of the party. In addition
the establishment of Hazara and Uzbek nationalist parties meant that
in reaction some Tajiks were liable to identify themselves politically
with Jamiyat.
On the other hand, a number of factors militated in practice
against Jamiyats slide towards Tajik nationalism. The preponderance
of Tajiks within Jamiyat did not lead to a neglect of local identities in

266

The Ethnicisation of the Conflict

favour of a Tajik identity, which continued to be elusive. The leaders themselves rejected ethnicisation and turned instead to identities
which were either more local or broader. During the war Ismail
Khan refused to admit his ethnic identity precisely to avoid the
interpretation of his actions in this light.7 Until the fall of Kabul,
Rabbani relied during his presidency on a clientele based on a province rather than on an ethnic group, and surrounded himself with
Badakhshis rather than with Tajiks in the wider sense. Masud regarded himself more as a Panjshiri than as a Tajik, and the war reinforced the identity of the inhabitants of the Panjshir valley to the
point where it virtually became an ethnicity.
Further, the creation of a coherent political region was never possible, and indeed would not have responded to a demand, as in the
case of Hazarajat, since the complexity of the distribution of ethnicities in fact made the formulation of a demand for a Tajik territory
impossible. The Tajiks, in the event of a demand for a federal structure, would not have been able to lay claim to a homogeneous and
viable territory. In any case the principal leaders of Jamiyat-i Islami
refused to contemplate a federal state; rather, despite the Tajik presence, most of the larger towns tended to favour attachment to the
Afghan state rather than to any ethnic loyalty. In the last resort Jamiyat was constructed by Islamist networks established in the pre-war
period on a political rather than a communal basis. What was seen
in the end was a reinforcement of local solidarities, which guaranteed both loyalty and access to resources, rather than a Tajik ethnonationalism. The effectiveness of these networks was balanced by a
progessive exclusivity of Jamiyat-i Islami which, at any rate in Kabul,
seemed increasingly to be in the hands of the Panjshiri qowm.
The Taliban. The initial success of the Taliban was astonishing, but
after the south of Afghanistan had fallen into its hands it encountered
structured organisations that controlled mainly non-Pushtun populations, for example in Herat. As a result it recruited only among the
Pushtuns, and although it never identified itself as such, it was seen
7

In fact Ismail Khan, like his deputy Allauddin, was a Persian-speaker from the region of Shindand, but having lived in a Pushtun environment he spoke both languages fluently.

Ethnicisation as an unintended consequence

267

by many as enabling a return to the traditional Pushtun domination

of Afghanistans national territory.
In this sense the Taliban is sometimes described as a tribal or nationalist Pushtun movement, two concepts which were not always
distinguished from each other. However, it was a movement dominated by a group of religious individuals, who established their authority with no reference to tribal institutions, which in any case barely
existed in the region of Kandahar where the movement had its base.
Taliban law was greatly at variance with tribal customs. The hypothesis that it was a tribal movement cannot be sustained, since neither
tribes nor clans were represented in their own right within it.
In addition the Taliban movement was founded on a fundamentalist ideology opposed to all nationalist pretensions. Its official goal
was the reunification of all Afghans under an Islamic government.
It rejected all national or tribal justifications and took satisfaction in
drawing attention to the presence within the movement of nonPushtuns such as mullah Ghaysuddin Agha, who was a member of
the shura of Kabul and came originally from Badakhshan. On two
occasions ideological tendencies alien to the Taliban attempted to
appeal to a putative Pushtun solidarity in an attempt to make use of
the movement, and its failure is in itself informative. Former communists believed at one point that they could insinuate themselves
into the Taliban movement on the basis of Pushtun solidarity. At first
the Taliban accepted them since their officers were particularly important for a relatively unstructured movement without military experts other than some Pakistani officers. After the capture of Kabul,
however, most of the former communists were ousted or even physically eliminated. Also, the royalists wanted to make use of the Taliban
to put the king or his heir back on the throne, making their appeal to
the origins of the Afghan dynasty in Kandahar. Here again harsh disillusionment followed since the religious and charismatic nature of
the Taliban was incompatible with the royalist project.
Within the movement local solidarities were a stronger mobilising force than affiliation to a Pushtun ethnicity as such. For example,
mullah Omar is from Tarin Kot in the province of Uruzgan, and a
strong representation of his countrymen was to be observed in positions of authority, e.g. mullah Abbas. In addition, mullah Omar was a
Ghilzai Pushtun Hottak, a group which Pierre Centlivres has estab-

268

The Ethnicisation of the Conflict

lished was over-represented in the Taliban government.8 Other solidarity networks existed, such as that between mujahidin who had
previously served under the same commander, such as mawlawi Jalaluddin Haqqani, formerly of Hezb-i Islami (Khales). On a more personal level the ministers and the governors surrounded themselves
with members of their own familiese.g. mullah Muhammad Abbas
and his family in the Ministry of Health.
Still, the Taliban retained a distinct Pushtun sensibility, especially
in cultural affairs, tending to reject the Persian culture which was the
basis of the training of ulema in pre-war times. Teaching in the
madrasas of the North-West Frontier Province was traditionally carried out in Pushtu and Arabic. The Pushtuns had also continued to
nurture the desire once more to achieve domination within Afghanistans national territory. This attitude did not necessarily demonstrate
the existence of a nascentor re-nascentnationalism,but rather the
wish to perpetuate an ethnic hierarchy dominated by Pushtuns. Further, the Taliban did at some points employ ethnic arguments to
mobilise support. For instance, in 1995 as Ismail Khan advanced
towards Kandahar, it appealed to Pushtun solidarity against a Tajik
aggressor. Similarly in the north its preference for alliances with Pushtun minorities gave credence, with hindsight, to the ethnic prejudices of both sides. The Taliban was also led by confrontation and
suspicion to institute discriminatory practices, for example against the
Panjshiris of Kabul who were suspected of assisting Masud.

The challenge to the ethnic hierarchy

The civil war and the collapse of the state undermined the informal
ethnic hierarchy according to which relations between groups had
been organised. Such political transformations had as their consequence that the pre-war ethnic hierarchy no longer appeared natural, i.e. it was not accepted by all sides. The Pushtuns had in fact lost
the leverage of the state, which had once given them a considerable
advantage, while the other communities armed themselves for the
struggle against the Soviets and in the process acquired their autonomy. Several groups underwent a change of status during the war, in
particular the Hazaras and the Uzbeks who in the past had often
8

Personal information, March 1999.

The challenge to the ethnic hierarchy

269

been mistrusted and excluded from the army and political authority.
These changes had tangible effects on relations between ethnicities
examples were a ban on Pushtun nomads from pasturing their flocks
in Hazara territory, and pressure on Pushtun minorities in the north
due to competition for land. There was an asymmetry between
ethnicities and regions which should not be overlooked. The Pushtun south was ethnically homogeneous, and tensions were therefore
centred on the north. In these provinces the Pushtun minorities
were often subjected to pressure by their neighbours and therefore
generally supported the Taliban. In particular the tensions arising
from this challenge to the position of communities caused a number
of massacres. Few of these explicitly targeted an ethnic group, and
the majority were directed against the Hazaras. On 11 February
1993 Sayyaf and Masuds troops launched an attack on the Shiite
party Hezb-i Wahdat, which lost control of Afshar, a district west of
Kabul. The victorious troops were given their head, and perhaps as
many as 2300 civilians were massacred by Sayyaf and Masuds men
in an episode which continued till 14 February. Witnesses claim the
mujahidin went into houses and carried out various atrocities, including rapes, killings and the mutilation of bodies.9 In 1998 during the
second capture of Mazar-i Sharif the Taliban massacred hundreds of
Hazara civilians. Witnesses agree on the facts: the Taliban took the
town and during the following three days slaughtered the Hazaras,
going into houses to uncover and kill men of fighting age. There
were probably several hundred victims. The bodies were left in the
streets for several days, after which the Taliban made members of
the public pick up the bodies and bury them in the common graves
used the year before for the bodies of Taliban prisoners massacred
by Maliks forces after the first capture of the town in 1997.10 On the
evidence of eye-witnesses, the executions were carried out on the
basis of ethnic and religious affiliation, since in the north Hazaras were
not easily distinguishable. It is possible that Pakistani fundamentalistswho are also supposed to have been responsible for the death
of the Iranians who were in Mazar-i Sharifmay have played a role
9

See Etienne Gille, Crimes Afchar, Les Nouvelles dAfghanistan, no. 60, 1993,
p. ii.
10 Eye-witness accounts collected in Mazar-i Sharif, April 2000, and in Pakistan in
1999.

270

The Ethnicisation of the Conflict

in these killings. This would explain why the Shiite Hazaras, who
did not bear the main responsibility for the massacres of 1997,
should have been the sole targets. Nothing on a similar scale took
place in relation to the Uzbeks.11 There were to be no further massacres on this scale, but various operations in Hazarajat were notable
for their brutality, including on several occasions the killing of civilians and in particular old men. In June 2001 the Taliban took and
destroyed the town of Yakaolang, Hezb-i Wahdats fief in the heart
of Hazarajat.12
How does one explain such massacres? The first point to emphasise is that they cannot be accounted for by a desire to exterminate a particular group or carry out ethnic cleansing of the kind seen
in the Balkans. There was in fact no ethnic cleansing on a major
scale in Afghanistan in the 1990s, although local minority groups
were the victims of pressure from neighbours who coveted their
lands. The survivors of victim groups were able later to return to the
same spot, since there was no systematic attempt at the transfer of
populations.13 For example, after the clashes in Bamyan, the Hazaras
returned to their villages, and similarly there was no flight of Hazaras
after the massacres in Mazar-i Sharif.14
The Taliban rebuilt a centralised authority at a local level, exploiting its alliances with local solidarity networks, a circumstance which
militated against the politicisation of identity construction on the
base of local ethnicities. In November 1998 Muhammad Akbari joined
up with the Taliban, on the grounds of its recognition by the Shiite
ulema; the Taliban in turn recognised them as legitimate ulema, with
the aim of making use of them as arms of the central state.15 In Bamyan
11

However, the Uzbek village of Zari (Balkh) suffered a massacre after being retaken by the Taliban in May 2001.
12 Afghanistan, Paying for the Talibans Crimes: Abuses Against Ethnic Pushtuns
in Northern Afghanistan, Human Rights Watch, April 2002, vol. 14, no. 2.
13 Nevertheless mention might be made of at least two places, the Bangi valley
(Takhar) and Robatak (Samangan), where Pushtun or Gujjara populations
moved in after the eviction of Hazaras, Tajiks or Uzbeks.
14 The Taliban took responsibility for a number of war crimes at Bamyan, including cutting the throats of several dozen old men, and the summary execution of
dozens of civilians.
15 This was in striking contrast to the Pakistani Deobandis, who were violently
anti-Shiite.

The challenge to the ethnic hierarchy

271

the Taliban made a local alliance with the Tajiks in order to overcome the Hazaras. Their standpoint was in practice a return to the
pre-war ethnic hierarchy, which was incompatible with ethnic cleansing. There was no wish to put a halt to multi-ethnic coexistence,
but such tolerance was only made possible because the communities
stood in a hierarchical relationship, a situation which offers a brutal
explanation of the massacres to which the Hazaras were subjected.
The Hazaras were not recognised as equal interlocutors, since this
would have breached the ethnic hierarchy. The massacres can therefore be seen as in essence part of a move towards internal reconquest,
and were not without echoes of the campaigns conducted by Abdul
Rahman Khan at the close of the nineteenth century.
In contrast, as developments that followed the fall of the Taliban
confirmed, the parties in the northJombesh, Hezb-i Wahdat and
Jamiyat-i Islamitended in the direction of driving the Pushtuns
out of the north. The conflicts brought about a transformation of
national feeling, by way of a new connection between national identity and communal affiliation. In exile the only Afghan national
identity acknowledged by the Pakistani government and the humanitarian aid agencies manifested itself paradoxically to the detriment of regional and tribal identities.16 Exposure to a quite different
way of life also gave rise to a feeling of common identity between
exiled Afghans, who at least shared similar interests if not the same
opinions. In the same sense the presence of hundreds of thousands of
Pushtuns in Pakistan often gave rise to tensions between the newlyarrived population and the host country, which tended to reinforce
the feeling of being Afghan to the detriment of a trans-border Pushtun identity. Although 70% of the refugees came from the Afghan
provinces bordering Pakistan, there was no general integration between the local population and the refugees.17 The settlement in the
frontier regions of several hundred thousand Afghans with Pakistani
identity cards had not up to that point resulted in assimilation.

16

Pierre Centlivres, Micheline Centlivres-Demont, Afghanistan Info, 29, March

1991.
17 Richard English, The Economic Impact of Afghan Refugee Settlement on the
Tribal Areas of North-West Pakistan, UNHCR, unpublished, 1989, p. 35.

9. The Clerical State

The Taliban regime can be interpreted in several different ways.1
The hypothesis defended here is that its power was of a religious or,
more precisely, a clerical nature. A comparison with revolutionary
Iran in terms of three key factors will help to illuminate the unique
nature of the Afghan situation. These factors are (1) the identity of the
actors; (2) the institutional model adopted; and (3) the plan of social
organisation.
(1) While in Afghanistan it was a group of ulema which took power,
the political actors who emerged from the revolution in Iran were
much more varied. The Iranian clergy had been largely inactive during the twentieth century except at certain moments of crisis, and
were far from unanimous in their backing for Khomeini.2 The urban
population, who were the major agents of the revolution, did not regard Khomeini solely or even principally as a cleric. A further element of difference here was that the relationship between the ulema
and the Islamists was different in Iran and Afghanistan. In Iran the
ulema and the Islamists acted in collaboration, while in Afghanistan
the opposition between these two groups was deepening. In the Iranian revolution also the mobilisation of urban groups was a crucial
factor, in contrast to the rural origins of the Talibans rank and file.
(2) As for the institutional model, although the Iranian regime is
authoritarian and the constitutional principle of the wilayat al-faqih
placed serious restrictions on the freedom of the administration, the
government nevertheless emerged from contested elections. In con1
2

See in particular William Maley (ed.), op. cit.

On these issues see Farhad Khosrokhavar, Lutopie sacrifie. Sociologie de la rvolution
iranienne, Paris: Presses de la Fondation nationale des sciences politiques, 1993;
and Paul Vieille, Lorientalisme est thoriquement spcifique? A propos des
interprtations de la rvolution iranienne, Peuples mditerranens 50, Jan.March
1990, pp. 14961.

272

The Clerical State

273

trast to the situation in Afghanistan, Iranian political life was complex, with opportunities for the exercise of liberty.
(3) Although religious puritanism was a consideration in both countries, the social models in the two cases were very different. The Iranian revolution was primarily a phenomenon of modernisation,
while the aims of the Taliban were reactionary. We consider these
issues further in relation to the Afghan case by means of a sociological study of the personnel of the revolutionary movements, an examination of state structures, and an analysis of the puritanical order
imposed by the regime.

The party functionaries

The Taliban was not a political party in the classic sense, since the
movement had no clear structure and soon became indistinguishable
from the state.3 The social origin of its officials showed a high level
of uniformity, and this consistency of membership was the source of
the movements unity. The party functionaries were generally Pushtuns, poor and of rural origin, with no education except what they
had received in the madrasas. Brought up during the warmullah
Omar was born around 1961they had fought as members of various parties, but had seldom held positions of responsibility since they
had little in the way of education, religious charisma or notability to
recommend them. The new ruling class was therefore entirely alien
to the bourgeois and urban world from which Afghan lites had
hitherto emerged. Their status as outsiders partly explained their
coherence as a group.
Their rise to power, the result of a collective mobilisation, enabled
them to occupy positions of prestige and authority out of proportion to what they could have expected within the Afghan society of
the 1970s, when the prestige of the ulema had been in decline. Mullah Omar himself, the son of a poor family, did not complete the
studies which would conventionally have given him the right to the
title of alem. His position as head of the Taliban represented a new
departure in the political and religious field. Under the Taliban the
3

Membership of the movement was signified by the fact of having fought in

Taliban groups or of carrying out administrative duties. This is the conclusion
from the decrees of mullah Omar, where he calls for the expulsion of corrupt
members of the movement (Decree 49, Gazette 788).

274

The Clerical State

great families lost most of the prestige on which they relied both as
ulema and as brotherhood leaders.4 They were not absorbed into the
Taliban, although the ulema linked to the Mujaddidi family in the
Ghazni region did join it after having already installed locally a fairly
similar fundamentalist political system themselves. The ousting of the
dominant families was the result of a twofold process. Although the
heirs of these families combined religious knowledge as ulema with
hereditary charisma as pirs, they were too implicated in political manoeuvres, corruption and the associated violence to appear as a credible alternative. An instance of this phenomenon was the erosion of
the legitimacy of the two great family networks, the Gaylani and the
Mujaddidi, both directly linked to political parties whose influence
was waning and which were notorious for their nepotism. Subsequently, without competent officials and lacking support from Pakistan, these parties were unable to play a significant role on the ground.
Positions of authority were held entirely by ulema belonging to
the closed circle of historic Talibanthose who had joined in the
early days. The social homogeneity of the Taliban leadership and the
unchallenged authority of mullah Omar may explain why there were
never internecine armed clashes or moves towards defection as a result of internal tensions, even after resounding defeats such as that at
Mazar-i Sharif in 1997. In this the Taliban differed from all other
Afghan parties. However the opacity of its administration has meant
that it is difficult to analyse its internal decision mechanisms.5 For
instance, there was never any serious basis for the supposition of
hostility between mullah Omar and mullah Rabbani, to which attention was once regularly drawn. There were probably some disagreements between leaders who were keen to achieve international
recognition, such as that between mullah Abdul Muttawakil (Minister
for Foreign Affairs) and mullah Muhammad Abbas Istanekzai (Minister of Health), on the one hand, and certain commanders, including
4

In addition, ulema who emerged from the government madrasas, who if they were
politicised were in general modernisers and Islamists, tended to take the side of
Masuds opposition, where they were not given positions of authority within his
organisation.
5 An informant in Kabul in 2000, who was himself a judge and an alem, verified
that information on the workings of the Taliban government was unavailable except to a few hundred Taliban ulema.

The party functionaries

275

Dadullah, on the other. However, mullah Omar continued to be the

unchallenged arbiter.
Although the Taliban movement was coherent it was not monolithic, and the networks of which it consisted may be differentiated,
in the first instance geographically. Under more detailed analysis
three groups of ulema can be distinguished: first, those from the region of Kandahar, who were directly aware of mullah Omar; secondly, the ulema of Ghazni and Logar, south of Kabul, who were
historically close to the Mujaddidi family and enlisted in 1995 and
1996; and finally the ulema in the east of the country, such as mawlawi
Haqqani. The significance of local solidarities, sometimes of tribal,
has already been examined. The existence of separate networks did
not imply ideological differences, and there is nothing to suggest irreconcilable tensions. The reality was that access to the resources of the
administration was shared,an outcome achieved through compromise.
The key experience shared by the group was the education of its
members in the Pakistani madrasas in the 1980s.6 Most of the ulema
at the level of the leadership had emerged from the madrasas of the
NorthWest Frontier Province and Baluchistan, for example the
Dar ul-Ulum Haqqaniya madrasa at Akora Khattak in NWFP.7 From
this madrasa came haji Ahmad Jan, Minister of Mines, mawlawi
Qalamuddin, head of the religious police, and mawlawi Arefullah
Aref, deputy Minister of Finance. One of the few former commanders to become a leader of the movement, mawlawi Haqqani from
Hezb-i Islami (Khales), spent several years at the madrasa Dar ul-Ulum
Haqqaniya, first as a student and then as a teacher. An address by mullah Omar was read from the platform at the ceremony marking the
6

When considering the following facts one must bear in mind that membership of
the Taliban was different in 2001 from what it had been on its first appearance in
1994. The Taliban was at first made up of a few hundred theology students, but
their numbers were substantially boosted by the membership of mujahidin from
the former parties, and from the government itself, as well as by the recruitment
of young men with no prior experience at a time when many of the Taliban old
guard had been killed in the fighting of recent years. In addition, the opportunistic adherence of former officials complicates the analysis. Finally, Pakistani nationals
sometimes played a significant role in the expansion of the movement, although
the leadership was undisputedly Afghan and stable from 1994 onwards.
7 Kamal Matinuddin, The Afghan Phenomenon: Afghanistan 19941997, Oxford
University Press, 1999, p. 17.

276

The Clerical State

graduation as ulema of Afghan and Pakistani students at this madrasa.

Also three close advisers of mullah Omar were graduates of the
madrasa Jamiyat ul-Ulum ul-Islamiyah, so that it was no surprise
when he sent a message of support after the assassination of two of
its teachers in November 1997.
The numbers of Afghans at the Pakistani madrasas was a phenomenon related to two considerations: first, religious institutions in
Pakistan were anyway increasing in size and numbers, and secondly
the substantial influx of Afghan refugees meant that more Afghans
were available as students. From the 1960s both the number of madrasas in Pakistan and the number of students (taliban) had expanded dramatically. Between 1960 and 1983 the number of taliban increased
tenfold, from 7,500 to 78,500, and of teachers from 321 to 2,217.8
There was no slackening in this increase, which was much more rapid
than that of the population. In 1988 there were 1,320 madrasas in
Punjab, but by 1997 there were 2,512 with 220,000 students.9 In
Karachi there were twenty-nine madrasas which educated an average
of 2,000 students each year.10 The acceleration in the 1980s was partly
a consequence of the policy of Islamisation undertaken by the regime
of Zia ul-Haq (197788). Two particularly important measures taken
by Zia were the recognition of diplomas from the madrasas by the
universities and the introduction of obligatory zakat, of which part
was given to the madrasas.11 Zia also encouraged madrasas to open in
the NWFP in order to sustain the Afghan jihad, which explains a larger increase in the number in this province than the national average.
The madrasas were attached to different religious tendencies:
Deobandi, Barelwi and Ahl-i Hadith. The teaching of the Deobandi
and Barelwi madrasas was generally very conservative. One of the
fundamental texts, the Dars-i Nizamiyya, dates from the eighteenth
century, and Aristotles logic was still taught. Some Deobandi madrasas exerted their influence on the national scale, in particular the
Jamiyat al-Ulum al-Islamiyah, founded by Yusuf Binari at Binari
Town close to Karachi, which had 8,000 students, including those at
its twelve affiliated mosques. Though fewer students are at present
8

Jamal Malik, op. cit., p. 178.

The News International, 28 May 1997.
10 The Herald, December 1997.
11 However, some madrasas refused the zakat to preserve their independence.
9

The party functionaries

277

taken by the madrasa Dar al-Uum Haqqaniyah at Akora Khattak in

the Peshawar district, founded in 1947 and now run by maulana
Sami al-Haq, in the past it produced a third of the Deobandi ulema
in Pakistan. Other madrasas, attached to the Ahl-i Hadith tendency
or linked to Jamaat-i Islami, displayed aspects of a more modernising
tendency, for example in their teaching of information technology
and English. Jamaat-i Islami opened a significant number of madrasas; especially in NWFP, where its forty-one, including nineteen set
up after the Soviet invasion, made up more than a third of the new
foundations in this province. Elsewhere Saudi Arabia financed the
construction of the Jamiyat Imam Bukhari madrasa at Peshawar,
which was opened in June 1999 at a ceremony addressed by a representative of the Saudi Ministry of Religious Affairs, Muhammad
Abdul Rahman.12 It was in the Ahl-i Hadith tradition and headed by
an official of the Jamaat al-Dawa lil-Quran wa lil-Sunna, a movement
which before the arrival of the Taliban had a presence in Afghanistan in the province of Kunar.
What was the position of Afghan students in these madrasas? When
the war began, the proportion of Afghan students rose markedly.
After 1982 some 9% of the students of religion in NWFP were
Afghans.13 For example Afghans made up the vast majority of the
750 students of the Jamiyat Imam Bukhari madrasa mentioned above.
In the 1960s, 15 per cent of the students at the madrasa Dar ul-Ulum
Haqqaniyah were Afghans, but by 1985 this proportion had risen to
60 per cent.14 Afghan students mainly attended Deobandi madrasas,
since these were in the majority in the frontier region and since the
Afghan ulema had historic links with the madrasa Dar al-Ulum
Deoband in India, though the links between this institution and the
Pakistani madrasas which claimed to follow the same teaching had
weakened. Under the generic term Deobandi there are in fact
movements of sometimes quite varied tendencies, and neither the
coherence of this school of thought nor the level of education of its
ulema should be over-estimated.15 Education at these madrasas had
12

The News International, 25 June 1999.

Jamal Malik, op. cit., p. 206.
14 Jamal Malik, op. cit., p. 207.
15 Concerning the penetration of the reformist movements in Afghanistan and
Central Asia, see the special issue of La lettre dAsie centrale 2, autumn 1994.
13

278

The Clerical State

two major consequences for their Afghan graduates: the cementing

of strong solidarities, emerging from a shared experience and a common world view; and (an issue to which we return) the appearance
of transnational networks.

State structures
The idea that the Taliban was not interested in the state and left it to
decay can be dismissed. Between 1996 and 2001 it gradually reconstructed various institutions, especially the administrative structure
and the judicial system. Still, Taliban ideology limited the remit of
the state to security, justice and the observation of religious regulations. It intervened little in economic matters, since there were few
taxes and little investment, or in the social sphere, which was largely
reduced almost entirely to the action of religious charities. Analysis
of state structures under the Taliban begins with the position of the
ulema,and continues with the administration and finally the judiciary.
The position of the ulema in state institutions. The legitimacy of the new
Afghan state was based neither on nationalist ideology nor on popular sovereignty. In particular, the Taliban rejected political parties and
the idea of democratically contested elections as a foundation for
political legitimacy. This is a theme common to many Islamic or
fundamentalist movements, which take issue with the idea that it is
not within the scope of a majority of electors to alter the law of God.
The shariat, interpreted by the ulema, was seen as the only legitimate
source of law, while religious scholars dominated the judiciary, the
executive and legislative activity, with distinctions drawn in practice
between these three functions. The members of shuras were mostly
ulema, since religious qualification was preferred to technical expertise or being socially representative as a criterion for membership. In
addition, several of mullah Omars decrees tended to reinforce the
special role of the ulema, in particular the preservation of Muslim
heritage through religious education. The ulema maintained vigilance over morals and identified transgressors, for whom reform was
compulsory.
For the individual, religious affiliation was a determining factor in
political status. This was tantamount to a return to the form of legitimation of the Afghan state at the time of its establishment at the end

State structures

279

of the nineteenth century, before nationalist ideology came into

play. A consequence was that non-Muslims, which meant in practice
some thousands of Hindus, were obliged, at least in theory, to wear
distinctive marks and were subjected to the status of dhimmi (those
under protection), a provision which rested on a 1924 amendment
to the 1923 constitution.16 As for the Shiites, the Ismailis and the
Twelver Shiites faced different situations. The Ismailis were not regarded as Muslims, and were subjected to forcible conversion to the
Sunni creed. Many were driven into exile, but they were not numerically important. In contrast, despite some localised incidents,
mainstream Shiites were not forcibly converted and were able, for
example, to observe ashura in public in Kabul, including self-flagellation, which was nevertheless criticised by many Sunnis.
Mullah Omar was the cornerstone of the political structure, since
he brought together the legislative, executive and judicial powers.
His decrees, signed amir al-muminin mujahid mullah Muhammad
Omar Akhund, covered all issues, including the status of women,
the treatment of prisoners, the obligation to say prayers, and the appointment of administrative officials and judges. The authority of
the various shuras, including the government in Kabul, did not exceed
what was delegated to them, and their power was entirely advisory.
The charismatic authority of mullah Omar did not arise from the
depth of his religious knowledge but was seen as a gift from God,
and his dreams were interpreted by his disciples as signs of the divine
will.17 Accounts of him convey the stereotypical image of a man of
religion with an inclination towards mysticism. He was said to be
16
17

This measure, which the Taliban promulgated, was later abandoned.

For a biography of mullah Omar see Kamal Matinuddin, op. cit., p. 223. He was
born in 1961 in the village of Naudeh (in the district of Panjwai, Kandahar
province). His father afterwards lived at Dehrwut (in the district of Tarin Kot,
Uruzgan province), where he was the mullah of a small mosque. Mullah Omar
fought as a member of various parties (Harakat-i Enqelab, Hezb-i Islami (Khales)
with Nik Mohammad and Jamiyat-i Islami) and was wounded a number of
times. He speaks Persian and had studied theology, but did not officially hold
the title of alem. After the Soviet withdrawal he continued to live in Nashke
Nakhud, in the village of Sinsegar in the district of Maywand, Kandahar province, rather than move to Pakistan, in contrast to many other future officials of
the movement. There have been very few interviews with mullah Omar; however see LIslam au bout du fusil Entretien avec Mohammad Omar, Politique
internationale 74, winter 19967, pp. 13543.

280

The Clerical State

humble and soft-spoken, and to derive his authority from an ability

to listen, from lack of personal ambition, and from the rigour of his
application of the shariat. His election as amir al-muminin (commander of the faithful) in Kandahar in April 1996 by an assembly of
1,527 ulema from Afghanistan, Pakistan and Iran was the first step in
the routinisation of his charisma. At the moment of his election the
cloak of the Prophet was displayed at a public gathering in Kandahar. This cloak was a relic preserved at a zariat in Kandahar and had
last been displayed in 1935 during a cholera epidemic. This ceremonial, reflecting the ritual previously used by Abdul Rahman Khan,
was an indication that the amir was regarded as God-sent rather than
elected in the usual sense. Mullah Omar, incidentally, had been criticised by the Gaylani and the Mujaddidi who correctly perceived
him as challenging the established religious hierarchy.18 Official recognition by the body of ulema enabled a link to be established between the charisma of the leader and the governmental institution.
Administrative structures. Mullah Omar did not personally head the
government in Kabul, which he delegated to his faithful followers
while himself remaining in Kandahar. However, his keeping a deliberate distance from the capital and from the government institutions
did not detract from his absolute authority. The structures of power
evolved after the capture of Kandahar. In the initial phases mullah
Omar was supported by shurasadvisory councilswhose membership was made public on a number of occasions, but once the Taliban had strengthened its grip, following the capture of Kabul in 1996,
the failure to distinguish between the structures of the state and the
Taliban movement became more marked.
A number of phases may be distinguished as a simplified account
of events will show. In the first place, mullah Omar was backed up by
a ten-strong shura mainly concerned with military activities. After
the capture of Herat the shura was expanded in size, rising from
thirty members in 1995 to 100 at the time of the capture of Kabul.19
The shura then established in the capital progressively became the
18

This appellation had, already been applied to mullah Omar, see The News International, 27 January 1995. Rabbani had also received it from the Shura Ahl-i Hal wa
Hakd in 1994, as had Jamil ur-Rahman in Kunar, from a local religious assembly.
19 According to the Frontier Post of 24 February 1995, the shura of Kandahar included mawlawi Muhammad Rabbani, mawlawi Esanullah, mawlawi Abbas, mawlawi

State structures

281

effective government of the country, led by mullah Rabbani, who

was known as the Ras-i Shura-yi Sarparast (director of the council
of guardianship). Another shura continued to function in Kandahar
with some ten ulema who made up mullah Omars immediate entourage. Some members of this shura were also members of the shura
of Kabul. The consolidation of the military situation and the seizure
of the capital created the opportunity for more elaborate formal
structures to be set up. In October 1997 Afghanistan was renamed
the Islamic Emirate of Afghanistan, and the administrative divisions
which had existed under Daud were mostly re-established.20
At the beginning the Taliban did not have sufficient numbers to
occupy the countryside in depth and controlled only the towns and
the principal roads. The outcome was basically the establishment of
a limited level of control. For instance, in Logar and Ghazni ulema
from Kandahar took over the posts of district officials (uluswal) while
the rest of the administration remained in place. Later the Taliban
did set up a local administrative system. In the towns, especially Kandahar, each of the various sectors was placed under the authority of a
commander responsible for security and respect for Islamic morals.
There were also local shuras which provided liaison between the
Taliban and the population, and looked after the details of everyday
life such as the organisation of the bazaar and the supervision of traffic. These councils were in principle elected by villagers. The guardians of the bazaars were elected in an analogous way.21 Similarly the
local mullah was normally chosen by the inhabitants rather than by
the uluswal.22 In practice the mullahs had a decisive influence over
these councils, tending to exclude the khans and the educated class.
Muhammad, mawlawi Pasaband, mawlawi Muhammad Hasan, mawlawi Nuruddin, mawlawi Walik Ahmad, mawlawi Shir Muhammad Malang, mawlawi Abd alRahman, mawlawi Abd al-Hakim, mawlawi Sardar Ahmad, haji Muhammad
Ghaus and Masum Afghani. The shura of Kabul included mullah Muhammad
Rabbani, mullah Muhammad Hasan, mullah Muhammad Ghaus, mullah Seyyed
Ghaysuddin Agha, mullah Ghazil Muhammad and mullah Abdul Razaq.
20 See decree no. 12 of 7 March 1996, Official Gazette no. 783, 1997, p. 21, which
sets up the same ministries which existed under Daud. The original date of this
decree is given as 1375/12/16, but it is noteworthy that from 1999 the Taliban
applied the lunar Islamic calendar (under which 1999 coincided broadly with
1419) abolishing the solar calendar where 1999 was equivalent to 1378.
21 Decree no. 1981 of 20 March 1995 (1374/12/29).
22 Undated decree, Ministry of Justice, Kabul.

282

The Clerical State

To avoid rivalry between communities, the Taliban administrative system was organised on the principle of a rapid rotation of officials, and their appointment in principle outside their home regions,
except in Kandahar. The provincial governors (the walis) were generally changed over within a few months, and brought their own
teams of officials, often drawn from their own qowm, together with
their own equipment, including vehicles. This avoided the establishment of strong regional powers, and prevented internal struggles
from developing into armed clashes. Rotation became less rapid as
the regime became more stable, but the principle of appointing governors from outside any given province remained the rule. The same
considerations resulted in the efficient and swift collection of weapons, since the authorities gave the appearance of being external to
local conflicts. Apparently the Taliban used fragmentation as a means
of conflict management, and situated itself in the position of a referee. Retribution against former party activists was not pursued on
condition that they refrained from further political activity; this
meant an escape from the cycle of vendetta.
The principal ministers, who took up office one by one, were able
only with difficulty to carry out their responsibilities since some
100,000 educated Afghans had fled the country, creating a crucial
lack of officials. A further difficulty was the lack of money, resulting
from the war and from the effects of the international embargo imposed in 1999. After a cut in staff of 40% in April 2000, 130,000 staff
were employed by the regime. The reduction in staff numbers was
concentrated in Kabul, with women particularly affected. With a
personnel and an outlook which scarcely fitted it for government,
the Taliban made the discovery that it was in practice dependent on
humanitarian aid. Although there were moments of strain, especially
when the NGOs temporarily withdrew in 1998, NGO activity was
perceived by the authorities as vital. The reality was that in 1998 the
NGOs spent $113 million in Afghanistan, and provided work for
around 25,000 Afghans.23 One of mullah Omars edicts decreed
five years imprisonment for anyone attacking foreigners working in
Afghanistan.
In some parts of the country humanitarian aid was essential to
stave off famine and avert a further popular exodus. In the district of
23

United Nations, Afghanistan Outlook, December 2000.

State structures

283

Bamyan, in Bamyan province, around a quarter of the income of the

population in 2000 was derived from humanitarian programmes.24
This was an exceptional situation, but the drought in 2001 caused
malnutrition and even famine. There was a general deterioration in
relations between the NGOs and the Taliban, who seem to have
wanted to increase the level of participation of Islamic NGOs.
The judiciary. In theory at least the judicial system was independent of
the executive. The qadis did not answer to the provincial governors
(the walis) or to the district officials (uluswal), although all ultimately
fell under the authority of mullah Omar. There were judges in all
districts and three levels of jurisdiction were in place in each province.25 In principle no punishment or sentence of imprisonment
could take effect without the sanction of a qadi.26 An edict was published by mullah Omar which decreed that detainees should not be
beatentorture had been widespread under all Afghan regimesand
that their money should not be taken.27 However, the conditions of
prisoners were not improved while the Taliban was in power, and
ill-treatment remained widespread, especially when the secret police
were involved.
All laws were subject to the approval of a shura of ulema which
scrutinised their conformity with the shariat and with Hanafi jurisprudence. The existing constitution, as well as a section of the criminal and civil law, was declared to be in conflict with the shariat and
was consequently abolished. Over criminal law and personal rights
Hanafi jurisprudence was applied, especially in the case of punishments such as amputation and stoning. An examination of juridical
texts issued by the Taliban shows that these generally followed the
precedent of the most traditionalist Afghan ulema, with no notable
24

This estimate is based on the reports of the NGO Solidarit Afghanistan.

Decree no. 105 of 10 September 1999 (1419/5/18), which distinguished between the lower courts (eptedai), then the court of appeal (morafea) and finally the
higher court (tahmiz).
26 See in particular the decrees of mullah Omar no. 4265 of 4 March 1999 (1419/
11/16) and no. 234 of 29 January 1999 (1419/10/11), available at the Ministry
of Justice, Kabul.
27 Decree no. 141 of 27 April 1997 (1376/6/7), available at the Ministry of Justice,
Kabul.
25

284

The Clerical State

innovations. Most of the measures promulgated derived from the

hadith collected by Ismail al-Bukhari (who died circa 870), which
served as the basis of Islamic law in Afghanistan. There was therefore little novelty except for the re-adoption of certain punishments
and other judicial measures which had fallen into disuse in Afghanistan and in the majority of other Muslim countries. Thus amputations for theft became frequent, although this practice had been
abolished in the 1950s. A number of women were stoned for adultery in spite of the disapproval of many of the ulema. The practice
had been abolished before the war and been infrequent since, with
only one reported case in Badakhshan.
Finally, the judgements of the qadis frequently went against tribal
customs, particularly with the outlawing of vendettas. In a case of
murder, the relatives of the victims were invited themselves to kill
the murderers, although a financial settlement was also provided for.
In all cases vendetta was forbidden.

The social backlash: the puritan order and social resistance

The intention of the Taliban was to institute a moral order such as it
imagined had existed before the modernist reforms of the 1950s. It
also largely readopted the measures taken in 1929 by Habibullah
Kalakani, who prohibited the wearing of European clothes, banned
the cutting of beards, and forbade women to move about without
the permission of their wali (male representative).28 Although the
media apparently discovered these measures only in 1996, they were
not radically new since most of the commanders had already implemented them during the war, including such measures as bans on
music and cigarettes and the compulsory wearing of beards. Such
fundamentalist parties as Ettehad and Hezb-i Islami (Khales) were
perfectly in sympathy with this aspect of Taliban policy.
On this basis the religious police, al-Amr bil-Maruf wal-Nahi an
al-Munkar, became a feature of the mechanism of social control.29
28

On the measures adopted in 1929 see Muhammad Naser Kemal, op. cit., p. 152.
For comparison with the Taliban, see for example decree no. 3409 of 25 December 1996 (1375/10/4), Official Gazette no. 783, 1997, p. 9.
29 The name refers to a verse in the Quran (sura 3, verse 104) which enjoins the
enforcement of virtue and the suppression of vice.

The social backlash: the puritan order and social resistance

285

This institution was organised at the national level with the status of
a ministry, and was especially active in Kabul, although a force probably numbering no more than 23,000 was unable to control the
population. It was to some extent a re-invention of a former institution, the muhtaseb, whose members were appointed by the local qadi,
with the responsibility of preventing petty crime, checking weights
and measures in the markets, and maintaining public morality through
such matters as attendance at the mosque. Under Nadir Shah the
muhtaseb had the power to demand the recitation of prayers from
passers-by, whipping them if they did not know the words. The
Taliban reintroduced this practice, and in addition made absence
from daily prayers punishable by up to ten days imprisonment.
In daily life the implementation of the shariat implied a degree of
puritanism hitherto unknown in Afghan society, as well as an intrusive police presence.30 Personal appearance was regulated. The beard
was to be of a precise lengtha fists width below the chinand
those failing to comply were jailed until their beards reached the appropriate length. In contrast, hair was to be kept short. As a general
rule all entertainment was forbidden. The festival of Nawruz, the
New Year, on 21 March, customarily marked by parties and picnics,
was declared unlawful because of its non-Muslim origin. Similarly
music-making was banned. To this end houses were searched, usually as the result of a denunciation, and those in contravention were
arrested and sometimes beaten. This brought radical change to the
atmosphere of a town such as Kandahar. Those caught selling audio
cassettes were liable to several days in prison and a lashing. The ownership of pigeons and other birds was no longer allowed as a hobby.
Pictures of living beings were forbidden,31 so that decorative designs
on lorries were erased, while television and the cinema were prohibited. In common with the rest of the press, the Taliban newspaper
Zarb-i Muminin (The Onslaught of the Faithful) strictly adhered to
this prohibition. School textbooks were censored to ensure the removal of pictures of people and animals.
Although fundamentalism had been increasing in Afghanistan for
twenty years, the Taliban regime, because of its radicalism, clashed
30

See Peter Marsden, The Taliban: war, religion and the new order in Afghanistan, London: Zed Books, 1998.
31 The sole exception to this rule seems to have been identity photographs for official documents.

286

The Clerical State

head-on with a significant section of the Afghan people. However,

the people had no way of giving direct expression to their discontent, since all demonstrations were ruthlessly suppressed. In early
December 1998, medical students protesting at the lack of resources
at the University of Jalalabad were brutally dispersed by the Qita-i
Muntazeri (Forces of Intervention), with the deaths of two demonstrators. Sometimes the Taliban was obliged to withdraw minor measures. For example, after the ban on Nawruz celebrations at Mazar-i
Sharif, it made a number of concession in the light of popular dissatisfaction. A further consideration was that there were no independent media or other independent organisations which could act as a
channel for discontent.
In such a situation, where open protest had become impossible,
the only alternatives were exile, for those who had the financial means
to leave, or oblique forms of popular resistance. In a city such as Kabul
discontent with the agents of the Taliban system was clearly evident,
with remarks sometimes being made even in their presence when
they were unable to understand Persian.32 Dissidence was also expressed through humour and through pressing aspects of physical
appearance to the limit of legality. In the towns the beard might be
worn a little short, with haircuts longish and romantic. Such popular
inertia and resistance meant that the regulations were far from being
applied to the letter. Music was still heard in private and even in
shared taxis once out of town, while birds were bought and sold in
spite of the ban. Books were sold under the counter in a trade which
was illegal in theory but tolerated in practice. Prostitution flourished
in spite of official puritanism since many women, especially widows,
were unable to carry on working. Street children were, and still are,
also easy prey.
Resistance to the regime was not unanimous. The degree of dissidence varied in different social groups and communities. The Taliban
enjoyed a certain level of support in the Pushtun countryside, where
the moral code was largely that of the local people, and because the
Taliban regime was a guarantee of security. Another factor was that
the ulema gave unstinting support to the new regime, which had spec32

On this kind of phenomena see in particular James Scott, Domination and the Arts
of Resistance, New Haven and London: Yale University Press, 1990. See also
Michel de Certeau, Linvention du quotidien, Paris: Gallimard, 1990.

The social backlash: the puritan order and social resistance

287

tacularly enhanced their prosperity and raised their social standing.

A further issue was that the rise of the Taliban gave expression to
the desire of rural people to avenge themselves on the towns. Many
young unemployed peasants from the south had joined the Taliban
both to earn a living and to obtain social status. Finally, some members of the merchant class had been favourable to the Taliban from
the start. The early operations on the Pakistan frontier in 1994 had
been partly financed by Kandahari merchants whose aim was to secure
freedom of movement. In this sphere the Talibans attitude was permissive; roads were kept open, while controls and taxes on trade were
minimal. The bazaars were also rebuilt, particularly in Kandahar.
As a general rule, however, the non-Pushtun population and the
Shiites were opposed to the new authorities. Communal affiliation
was not a determining factor; membership of the educated class was
more prescriptive of a persons attitude to the Taliban than his ethnic
origin. To illustrate in practice the diversity of reactions towards the
regime, we now examine in detail the circumstances of three very disparate social groupsmen of religion,the educated class and women.
The official role of the Ulema. The transformation of the ulema into
state officials had already been a long-term trend in most Muslim
societies, for example in Turkey and Pakistan. For more than a century the Afghan state had attempted to gain control of the ulema by
bringing them into the administration. Paradoxically the Taliban regime, which was infested with ulema, succeeded in divesting them of
their autonomy. The great majority of the ulema willingly came over
to the new regime,which provided them with a highly effective means
of exerting influence over Afghan society. The process of officialisation took place by way of the establishment of an administrative
hierarchy, so that ulema were ranked on a salary scale according to
their qualifications. The madrasas, in common with all teaching institutions, were henceforth controlled by the Ministry of Education.
Another issue was that the role played by the mullahs in the administrative process reinforced their relationship with the state apparatus.
In particular, they were entrusted with the evaluation of those taxes
which were obligatory on the citizens (ushr and zakat);33 they appropriated part of the revenue for their own needs, forwarding the rest
33

See decree no. 32, Official Gazette 788 and Official Gazette 789, p. 42.

288

The Clerical State

to the Ministry of Finance. Clearly such responsibilities placed a village mullah in the position of a representative of the state, playing a
role which had hitherto been the prerogative of functionaries or local
notables. The only circumstance which in theory allowed relative
autonomy to the religious establishment in its relations with the
authorities was the election of the mullah by the local community.
Still, although events observed in Wardak and Kabul might lead
to contrary conclusions, continuity tended to prevail. The available
choices were restricted, and the identification of the mullahs with
the machinery of state was not in reality called into question by this
procedure.
Therefore, in contrast to the previous situation the ulema wielded
an authority which depended on their relationship with the state.
This was a total reversal, since the ulema had been bastions of opposition to the state throughout most of the twentieth century, especially
to measures of modernisation. However, under the Taliban they were
assimilated to the state and therefore subject to criticism on the
grounds of their inefficiency and sometimes their corruption.
The alienation of the educated class from the state. Discontent was concentrated in the towns, especially Kabul, since the Taliban, in common with rural Pushtuns in general, particularly detested the urban
culture which it saw as anti-Islamic; in the town, they believed, they
were in danger of assimilation and therefore loss of identity.34 The
Talibans seizure of power was among other things a class struggle, in
which the urban bourgeoisie were for the moment the losers.
The urban lites, who valued modernity and their personal liberties whatever their communal affiliation, were unanimously critical
of the Taliban. Although their social influence was significant, they
were numerically in the minority and were politically marginalised
after the fall of Kabul in 1996. The privileged triangular relationship
which had existed between the state, the process of modernisation
and the educated class disappeared with the clericalisation of the state.
The educated class depended for its position on access to statelegitimised educational resources, while the state was at the same time
been its main employer. In consequence the majority of the edu34

Jon. W. Anderson, Social Structure and the Veil: comportment and interaction
in Afghanistan, Anthropos, 77, 1982, p. 416.

The social backlash: the puritan order and social resistance

289

cated class fled Afghanistan in successive waves.35 Those who remained were obliged to find jobs outside state institutions in either
humanitarian organisations or private institutions.
It is difficult to estimate the number of educated individuals working for NGOs in Afghanistan under the Taliban, a total to which
may be added these doing similar work in Pakistan who frequently
return to Afghanistan. On the basis of figures provided by the NGOs
themselves, an initial estimate might be some 10,000, taking into account that job descriptions were relatively fluid and that many employees were over-qualified in relation to the positions they occupied.36
More than 250 NGOs were officially registered, which at least enabled enterprising educated individuals to find niches in the humanitarian sphere. The numbers were relatively limited, but the NGOs
were probably as significant an employer as the Afghan state itself for
the educated class. The state actually employed its officials only parttime and paid extremely low salariesless than $10 per month for a
teacher, including bonusesand dismissed between a quarter and
half of its employees in 2000. Additionally, the most motivated and
highly qualified graduates tended to prefer working in NGOs, for
both financial and ideological reasons.
In fact parastatal organisations such as NGOs offered an organisational model and an ideological framework which were compatible
with the preconceptions of the Afghan educated class. The modernising role was thus transferred from the state to the NGOs, which
operated in a decentralised style, maintaining contact directly with
local circumstances. A further factor was that the NGOs constituted
a template for organisational modernity and bureaucratic rationality,
so that posts within them were much sought after by the educated.37
35

Research at a number of private schools at Mazar-i Sharif in April 2000 revealed

that between one-third and half of the boys aged 812 who were asked about
their future said they wished to go abroad. Though without statistical validity,
these results probably show that emigration was the predominant goal for those
seeking education.
36 This rough estimate is based on the 2001 handbook of organisations working in
Afghanistan produced by ACBAR, which serves as a coordinating body for the
NGOs and is available at the headquarters of ACBAR in Peshawar.
37 The evidence suggests, however, that the relations between Afghan-educated individuals and expatriates were not without complications. The growing distance
between the expatriates and the Afghan population had ramifications. Issues

290

The Clerical State

Consequently, because of their ethos, they became in practice a focus

of opposition to the Taliban government. The Taliban was not unaware of the situation, and the ambiguity of its attitude revealed the
two constraints under which it operated: there was its need to collaborate with the NGOs, but on the other hand it wished to keep
them under control, especially over issues relating to womens right
to work or education.
In the educational sector itself the evident deficiencies of the regime led the educated class to break free from the state. Educated
Afghans attached particular importance to schools, since education
provided a field for the transmission of their values and resources. At
the same time the ulema, themselves taught within the madrasas, displayed overt mistrust of the educational system. During the war fundamentalists frequently accused schoolteachers of being leftists, and
many schools were destroyed for this reason.
The states dereliction of this key sector resulted in a proliferation
of private classes, and those for girls were conducted in secret. Although
there seem to be no figures relating to this phenomenon, it was of
real significance and showed a failure by the state to respond to an
urgent social demand which extended far beyond the educated class
itself. However, because of the lack of financial resources the level of
teaching was low. There was one exception: five Turco-Afghan schools
of a very good standard operated in Afghanistan. Though mullah
Omar had issued a decree that banned foreigners from running
schools and required strict monitoring of financial donations, these
schools were nevertheless tolerated, since their personnel were Turkish, and hence Muslim, and they were ideologically close to Fethullah Glen, a disciple of the Islamist Nurcu movement in Turkey.38
This movements objective was to provide a basis of training for the
entire educational systemjust as the schools funded by the western
countries had done in the pastand to produce an Islamist lite
open to ideas of modernity.
The confrontation between these two social groups was not confined to Afghan territory. The ulema and the educated class both
included the influential role of educated Afghans as intermediaries, as well as
tensions between locals and expatriates.
38 Bayram Balci, Missionnaires de lislam en Asie centrale, Paris: Maisonneuve et Larose,
2003.

The social backlash: the puritan order and social resistance

291

operated transnational strategies for the mobilisation of support.

Associations run by educated Afghans, not all of whom were necessarily favourable to Masud, opted for transnational strategies based
on human rights, and were thus able to link up with a vibrant international network. In particular Afghan feminist groups, whose ability
to organise inside the country had always been limited, succeeded in
obtaining publicity for their cause.
Women. Why has the status of women become a central issue in the
wests perception of Afghanistan? This question may well appear
either nave or provocative, since self-evidently the oppression Afghan
women have suffered from the Taliban is a justification for international solidarity. Without wishing to ignore these arguments, one
should place in perspective a number of other issues which may lead
to some modification of this point of view. First, the degradation of
womens status did not originate with the Taliban, but dates back to
the fall of Kabul in 1992 or even to the beginning of the civil war.
Nevertheless, western public opinionor at least the media (a valid
distinction)was hardly interested in Afghan women before 1996.
There was some study of the issue before the Soviet invasion, but
Afghanistan was seen primarily as a traditionalsociety,living by timehonoured principles which were not conducive to critical analysis of
the position of women. Further, following the Soviet invasion arguments relating to the liberation of women were appropriated by the
Kabul regime, and these lessened their legitimacy in western eyes.
In this context the fall of Kabul came as a turning-point because
of the visibility of the Talibans actions in an urban environment relatively well covered by the media. However, revulsion towards the
monstrous Taliban did not lead to a better understanding of local
issues. The veil was probably the clearest example of the perverse nature of media coverage. Before the Taliban the wearing of the chador
(the Iranian-style veil) was widespread in the countryside, while the
burqa was worn both in the towns and everywhere in the south of
the country. Social pressures ensured the continuation of this practice, including in Panjshir, where Masud wished to present himself as
more modernised. In Herat the mujahidin had already threatened to
attack unveiled women in the towns. The burqa became obligatory
from 1992 onwards.

292

The Clerical State

The imposition of the burqa by the Taliban thus mainly affected

the educated class, particularly in Kabul, where it had been in disuse
for a generation. Despite the relative marginality of this question in
comparison with such issues as economic difficulties, violence against
the person and access to education and medical care, it nevertheless
became basic in international campaigns of mobilisation, which for
the first time included mass-circulation womens magazines. It was
symptomatic that there was little media coverage of the report of
Physicians for Human Rights in 2001, a considerable modification
of its report of 1998 which had caused a considerable stir.39 This may
have been because, counter-intuitively, it pointed out that the veil
was not a central issue for most Afghan women, emphasising instead
the obstacles which limited access to health care. The fall of the
Taliban has led to the virtual disappearance from the media agenda
of the issue of the veil, and indeed of Afghan women in general.40
Thus the situation of Afghan women in its diversity and its development needs to be presented in a broader historical perspective.
Masculine domination and the diversity of womens situations. A number
of social norms serve to legitimise the subservient position of
women in Afghan society. The dominant values are malenamely
activity, strength and independence. These are set in opposition to
such feminine characteristics as physical weakness; susceptibility to
influence, especially from magical spirits; and impurity, connected
with menstruation. The differentiation of tasks by gender is extremely marked, with a clear separation between mens and womens
spheres of activity.41 Inequality is difficult to challenge since it is seen
as originating in biology or the natural order, and inextricably
linked to the religious principle by way of interpretations of the
Quran which confirm the weakness of women. Boys are celebrated
from birth, while the birth of a girl is perceived as a problem. The
39

Physicians for Human Rights, The Talibans War on Women: a health and human
rights crisis in Afghanistan, Boston, MA, 1998, p. 27.
40 In captions to newspaper pictures the burqa was once more being described as a
traditional veil.
41 Many points of similarity may be observed with Kabyle society; see Pierre
Bourdieu, Le sens pratique, Paris: Editions de Minuit, 1980.

The social backlash: the puritan order and social resistance

293

hyper-mortality of female infants is a result of the lower level of care

accorded to them by their mothers. However, the marriage of boys
is an expensive matter, which symmetrically entails the fact that girl
children may bring substantial sums to their families. For instance, the
dowry of young women of the Turkmen community, who weave
carpets, may be the equivalent of several years income.
This inferiority, which is socially constructed and interiorised by
both sexes, has legal implications. In tribal custom, especially among
the Pushtuns, women have no inheritance and under Islamic law
their inheritance is half that of men. Marriage is arranged in almost
all cases, including among the bourgeoisie, and in most cases is an
expression of family strategies and economic interests. A further factor is that polygamy is a legitimate practice, although, being for the
rich only, it is relatively rare, involving only some 5% of marriages.
Consequently the couple is not a basic element in social life, and for
a woman her husband is not necessarily the most important man.
Despite relatively egalitarian legislation from the 1960s, divorce is
practised solely at the mans initiative, but the importance of the
dowry and the risks of conflict between family groups act in general
as a barrier.
Women have very restricted access to public space, since they are
the repository of the honour of the group.42 In this matterwhere
the dictates of honour often run counter to Islamic lawAfghanistan resembles the pattern of Mediterranean societies, as in southern
Italy and Spain. Womens behaviour is monitored by the men of the
household, their husbands or brothers, a social obligation which is a
source of psychological tensions to individuals of both sexes. What is
prohibited is not merely adultery in the strict sensealthough this
would in theory entail the death penaltybut any act that allows
the slightest suspicion to fall upon a woman and therefore on the
family honour. This is a system very different from that of the shariat
which demands proof of individual culpability. Another issue is that
the women do not necessarily perform only a passive role in the
42

In the view of some authors the veil, rather than providing protection for
women, safeguards men against feminine sexuality, which is unconsciously perceived as uncontrollable and anti-social. On this issue see Inger W. Boesen,
Women, Honour and Love: some aspects of the Pushtun womans life in eastern
Afghanistan, Afghanistan Journal, no. 2, 1980.

294

The Clerical State

protection of the honour of the group. They also play an important

role in the vendetta as bearers of the memory of the insult. A man
who refuses to take vengeance runs the risk of being openly despised within his own house.
Children or older people may serve as intermediaries for women
in the public sphere, in the transmission of news, making purchases
and so on.43 A woman exercises a certain degree of economic autonomy over the disposal of her own property, but employs a representative (wakil) to manage her affairs. In general womens freedom of
movement is very restricted, except among nomadic groups where
they go unveiled, a special status which they would lose if they become sedentarised. There are noticeably few women in the bazaars,
and many have never been outside their villages except for brief visits to their relatives or to ziarats. Contact with strangers is not generally feasible, and guests do not normally meet women. Children
serve at table and the guest is received in a special room which is
only an ante-chamber to the truly private areas where only relatives
in whose presence the women are not obliged to be veiled are permitted to enter.
Finally, there is the issue of the specific influence of Islam in the
pre-war period. Rather than the Middle Ages with which it is
fashionable to draw comparisons when discussing Afghanistan, rural
France of the nineteenth century may provide a more meaningful
parallel.44 Which religion is in play is not crucial, so that Catholicism
could in reality serve as well as Islam as a justification for classifying
women as inferior in terms of either biology or religion.
Other than in the these general aspects, women are not in any
case a homogeneous group, and three principal dichotomies can be
observed. First, they live in markedly different circumstances in the
town and in the country. In the villages women make up communities which work and seek their entertainment together. When the
men have left for the fields they are often alone in the village and
43

Women do not pray at the mosque but, in principle at least, do so at home.

A woman does not therefore have access to the basic location for village sociability. Meetings at the well or work jointly undertaken are their main opportunities for social exchange.
44 Eugen Weber, Peasants into Frenchmen: The modernization of rural France 1870
1914, Stanford University Press, 1976, esp. chapter 11.

The social backlash: the puritan order and social resistance

295

enjoy great freedom of movement.45 In the towns women are much

less free to move about; they are easily confined to the house and
have no independent economic activities and hence no separate
source of income. The wearing of the burqaapparently introduced
during the nineteenth centuryis a symbol both of social superiority and of a stricter seclusion.
Secondly, age and especially maternity is one decisive factors in a
womans situation within the family. The young wife is transferred
from the authority of her father to that of her mother, and falls under the domination of her mother-in law, for whom she in effect becomes an assistant or even a servant. Only if she bears male children
is she gradually able to acquire some degree of authority. As she grows
older she comes to occupy a position of power and influence within
the household, since she often chooses wives for her sons and, although
usually illiterate, she plays a decisive role in the management of goods
and in financial decisions. In some cases the sons will ally themselves
with her against their father, especially if he is persuaded by his children to divide up the land.46 Finally, as an older woman she is more
easily able to speak in public, sometimes with considerable authority.
Thirdly, the status of women differs from one community to another. The main distinction is that between the Pushtuns, situated
mainly in the south, and the rest. Among the Pushtuns the code of
manners and honour is extremely strict, and the separation between
men and women is especially rigid.47 In the Uzbek, Turkmen and
Hazara communities, mostly in the north of Afghanistan, relations
between men and women follow more flexible rules.48 Among the
45

See Michelle Robin, La socit des femmes dans un village afghan, Communauts, no. 79, JanuaryMarch 1987, pp. 95101.
46 On this issue see Cherry Lindholm, The Swat Pukhtun Family as a Political
Training Group in Charles Lindholm (ed.), Frontier Perspectives: Essays in comparative anthropology, Oxford University Press, 1996, pp. 1727. The author analyses
the structures of Pushtun families, bringing out in particular the often violent
nature of family relationships.
47 Jon W. Anderson, Social Structure and the Veil: comportment and interaction
in Afghanistan, Anthropos 77, 1982. The author particularly stresses the point
not well grasped in the westthat the veil is reciprocal and acts as an organising
principle in the avoidance behaviour of both men and women outside the family circle.
48 Among the Aymaqs of central Afghanistan relations between men and women
are relatively egalitarian. See Bernard Dupaigne, La femme dans lconomie

296

The Clerical State

Uzbeks nocturnal visits by a fianc to his future bride are accepted.

Among the Hazaras women go outside the house without difficulty
and are able to speak to guests. A traveller may thus take tea with a
couple, the wife participating in the conversation on an equal footing with the husband, which would be wholly unthinkable among
the rural Pushtuns. Womens economic situations or, to be more
precise, their occupations also vary according to the community
in question. For example, Nuristani women cultivate crops, while
the men devote themselves more exclusively to pasturage. In areas
where weaving is a significant activity, especially in the north, it is
the women who weave carpets.
Apprenticeship to modernity. The status of women is a symbolic issue
which arouses political and religious passions and becomes a kind of
testing-point for the authorities. Amanullahs government (1919
29) was the first to attempt to modernise or westernise it, especially
through such dramatic measures as the unveiling of women at court,
and the establishment in 1921 of the first school for girls. In the same
year the reaffirmation of the abolition of slavery, a measure which
had already been proclaimed at the beginning of the century, set free
hundreds of women who had been kept as concubines in Kabul.49
Some decades later, between 1953 and 1963, Daud reintroduced
a similar reformist project when he decreedor, more precisely, allowedthe unveiling of women. This was pushed through in the
face of protests from conservatives, who mobilised forcefully against
it. His initial measures were followed by others, particularly in education. In 1960 co-education was established in higher education.
Female students and scholarship-holders were still very few, with
only 7% of the scholarships in higher education allocated to women
in 1970, but from then onwards a trend towards modernisation was
firmly established. The obligation placed on both boys and girls to
wear uniform at secondary schools in effect introduced western dress.
It was a decisive factor that these decisions were taken mainly by
the political authorities, with no significant support from the popururale in La femme afghane travers lhistoire de lAfghanistan, Actes du colloque
Unesco, Paris 11 December 1998, p. 22.
49 Aman-i Afghan, Kabul, Dalwa 18, 1299/7 February 1921, quoted in Afghan
Women and Women in Islam: parallels and divergences in La femme afghane
travers lhistoire, Actes du colloque Unesco, Paris, 11 December 1988, p. 45.

The social backlash: the puritan order and social resistance

297

lation, and this factor determined the direction of future developments. In the countryside women did not openly challenge the
patriarchal order. Although some established values, such as arranged
marriages, were challenged by the traditional Pushtun womens
songs (landais), these also celebrated the dominant code of behaviour, particularly honour and virility. Reforms were initiated from
above, and their failure to take root explains why women remained
largely excluded from the public sphere, even in bourgeois circles.
There were no womens associationsleaving out of account for
the moment the feminist movementwhich could have lobbied
effectively for legislation, and few women were members of political movements. Only the Maoist groups and the communists, with
certain exceptions, took notice of issues relating to women.50 Even
in the towns the presence of women in public was frowned on. As
late as the 1970s a womens demonstration was attacked by men who
threw acid over the demonstrators.
In contrast to the communist regimes in China and the Soviet
Union, which at certain points envisaged the abolition of the conventional family, the communist regime in Afghanistan showed no
wish, or was not able, to intrude into private life. It adopted progressive policies concerning women, but also took traditional values
into account, recognising in particular the role of the shariat as complementary to state law. The continuity between the communist regime in Kabul and previous projects of modernisation was marked.
The regime maintained the proportion of women members of the
party at around 15% although many of these were the wives or
daughters of militants since many marriages took place within the
party. In addition, there were women members of the party militias,
especially in Kabul and in some of the northern towns. The most
marked changes were in public education, especially at the middle
and lower levels. In Kabul half of the holders of public teaching posts
were women, as were the majority of the staff of the Ministries of
Education and Health. Similarly, 55% of the students were girls.51
50

In Herat in the spring of 1989 an informant recalled the agitation of the Maoist
groups in favour of women, and the unenthusiastic response to these efforts from
the highly conservative local population.
51 See Micheline Centlivres-Demont, Les femmes afghanes aujourdhui, Afghanistan Info, Neuchtel, no. 23, November 1998, pp. 1718.

298

The Clerical State

Finally, especially in Kabul where the regime was relatively well

established and able to guarantee a degree of security, the wives of
the middle class carried on the modernising tendencies of previous
regimes. For instance, dress codes showed the beginnings of a break
with traditional practices, although these innovations were mostly
restricted to the modern areas of the capital and to a lesser extent of
Jalalabad and Mazar-i Sharif. This process of modernisation contrasted with the situation in the countryside, where the mujahidin
imposed an order of a kind that was much more conservative or
even fundamentalist.
The confrontation with fundamentalism. The mujahidins prohibition of
womens participation in public activities became stricter. The resistance parties had generally not called on women to mobilise, although
there were some womens groups, particularly within Harakat-i Islami
and Hezb-i Islami. The collapse of the structures of the state brought
education more or less to a halt, while opposition from fundamentalists, including those among the population at large, considerably
restricted the educational opportunities for girls. In some areas the
commanders who controlled the countryside continued to permit
education for girls, for example in Herat and in a large part of the
north, but this happened less in the Pushtun regions.
Women were a majority among the refugees (28% of women against
25% of men were refugees).52 The seclusion to which refugee women
were subject was even stricter than that prevailing in the villages, because of their exile in an unknown situation, where they were under
the sway of the rigid conceptions of honour of the tribal Pushtuns
among whom they found themselves. Because of overcrowding, it
was difficult in the camps to protect private space. On the other hand,
the health conditions were an improvement, so that infant mortality
fell substantially to a level even lower than that among the rural
Pakistani population. The Islamist parties attempted to improve the
education of girls in the refugee camps, although they were separated
from the boys, while at the same time the fundamentalist movements continued to discourage the education of girls. Womens lack
of political influence, without recognised representation, obstructed
52

Micheline Centlivres-Demont, op. cit., p. 18.

The social backlash: the puritan order and social resistance

299

them from struggling for recognition of their rights when the Kabul
regime fell and movements of the most fundamentalist type gradually took hold of the reins of power.
The arrival of the mujahidin in 1992 inaugurated a range of restrictions, from the wearing of the veil to the ban on women appearing on television. In August 1995 Afghan women were refused
authorisation by Rabbanis government to go to the fourth World
Womens Conference in Beijing, which was viewed as un-Islamic.
The representation of women in those areas of the administration
still functioning fell markedly. The practices of different political
groups varied, but the fundamentalist influence was universally felt.
Some commanders had a more open attitude regarding womens
education and kept the secondary schools open, for example in
Herat and in the north. The former communist militias in the north,
which remained in power till 1998, altered few features of the dayto-day operation of the administration. Other commanders ordered
the girls schools to be definitively shut and insisted on the strict application of the shariat. Finally, in some regions insecurity and fighting reached such a pitch that the schools were shut anyway, while
widespread violence had particular consequences for women: for
example, at the time of the capture of Jalalabad in 1992, the nomadic
tribes carried women off and enacted marriages which were no
more than enforced prostitution. In Kabul all the armed groups, especially Dostums militias, were guilty of rapes and kidnapping, leading
sometimes to the suicide of young girls who had been dishonoured.
The very few women who dared to dress in the western style in the
modern part of Kabul were harassed by the mujahidin. All this was a
new departure, and a contrast, since Afghan women had seldom before been threatened with deliberate acts of violence, and certainly
not with rape.
In this respect the Talibans victory represented the triumph of the
most fundamentalist tendency. The status it imposed on women was
above all a rejection of urban and bourgeois culture, which stood
opposed to the idea of women held by the rural and fundamentalist
Pushtuns. Its earliest victims were educated women, who were mainly
in Kabul and numbered around 165,000.53 Aggression against women
53

See Carol Leduc, The Impact of Conflict on Womens Lives post-1992 in La

300

The Clerical State

was one of the major factors which led to the reluctance of the urban bourgeoisie to cooperate with the regime and their flight into
exile. The situation was different in the towns, in which the religious police were especially strict, from the countryside where daily
life was little changed. Outside the towns women continued to
wear the chadora veil over the headrather than the burqa. For
country women, who for practical or other reasons did not have access to schools or paid work, the principal effect of the arrival of the
Taliban was an end to insecurity.
Taliban legislation installed the shariat as the basis of both civil and
criminal law, and the status of women declined, especially in the
towns. However, in some cases the measures put in place by the Taliban were more benign than the tribal customs normally enforced
among the Pushtuns. For instance, a decree issued by mullah Omar54
forbade the frequent tribal practice of obliging a widow to marry a
brother of her deceased husband, as well as outlawing the gift of a
woman as a compensation for a killing.55 These measures marked an
advance over Pushtun tribal law, but a step backwards compared to
the positive legal situation before the war.
The separation of the sexes, which was already the rule in many
situations, was carried to the extreme. Women were obliged to sit at
the back in buses, where they were henceforth separated from the
men by a screen, while the fare collector had to be under ten years
old.56 The Taliban also required the inhabitants of Kabul to paint
over windows to a height of 1.80 metres and forbade women to
wash clothes at the riverside. Women were also obliged to wear the
burqa, while white socks and noisy heels were banned. They were officially allowed out only in the company of a mahram, a male member
of their family who acted as their representative and protector
although groups of unaccompanied women were in fact to be seen in
femme afghane travers lhistoire de lAfghanistan, Actes du colloque Unesco, Paris,
11 December 1998, p. 63.
54 Decree no. 103, 10 September 1999 (1419/5/18).
55 See especially Pierre Centlivres, who sets out in detail the contradictions between Taliban law and tribal custom regarding the rights of women. Pierre
Centlivres, Le mouvement Taliban et la condition fminine, Afghanistan Info,
March 1999, pp. 1114.
56 See Choong-Uyun Paik, Final Report on the Situation of Human Rights in Afghanistan, United Nations, 1997 (www.reliefweb.int).

The radicalisation of the Taliban

301

the bazaar in Kabul during the Taliban regime. The Taliban also banned women from working outside the home, particularly in international organisations, a measure which had been in place in Jalalabad
since 1994, though an exception was made for hospitals. The major
defeat for women was their dismissal from all state administrative
posts and their exclusion from the universities, where they had
achieved virtually equal status during the war.
The issue of education and access to health care reveals the inability of the Taliban to include women in the educational and health
services. The Taliban was not formally opposed in principle to the
education of women, but wanted to set up an Islamic system whose
details were never clarified, although it was to be based on the separation of men and women. While there had sometimes been provision on a local basis, women had no access to education after the age
of ten or twelve. Their use of hospital services was not forbidden in
principle, contrary to what has sometimes been alleged, but was
strictly supervised. In principle, women were to be examined only
by women. In this area the crucial obstacles were economic, and
were also related to the attitudes of male family members. There
were significant differences between regions in the way these measures were applied.

The radicalisation of the Taliban

The installation of the Taliban government should have effectively
removed the barriers to its international recognition, thus depriving
the opposition of the last element of leverage it might have had.57
The Taliban earnestly desired such a development, which would have
entrenched its regime. The United States was initially in favour. The
capture of Kabul in 1996 was described as a positive step by the
State Department, which however gradually shifted its position after
the appointment of Madeleine Albright as Secretary of State, and as
a result of the influence of campaigns for womens rights. Oil companies, especially UNOCAL, also attempted to engage positively with
the Taliban up to 19989 in the hope of exploiting Afghanistan as a
channel for the export of Turkmenistans energy resources. Initially
57

Only Saudi Arabia, Pakistan and the United Arab Emirates recognised the Taliban regime.

302

The Clerical State

Zalmay Khalilzad, who was close to Paul Wolfowitz and a consultant to UNOCAL, gave his blessing to a dialogue with the Taliban,
but he later fell in with the majority view and declared himself in
favour of their destabilisation.58 Finally, the split with the United
States was precipitated more by the presence of radical groups on
Afghan soil rather than human rights violations or drugs.
The Taliban had inherited the networks and training camps
which had been established in the 1980s, when collaboration between Islamic movements and the Afghan parties had been actively
encouraged by the United States.59 The majority of the Afghan factions, including those which belonged to the Northern Alliance, were
at that time in contact with groups based in Peshawar which provided financial assistance and a flow of volunteers for the Afghan jihad.
For thousands of militants, who afterwards returned to their own
countries, their stay in Afghanistan was a significant or even decisive
experience. The militants were progressively radicalised, and dozens
of the splinter groups based in Peshawar took up increasingly antiwestern positions towards the end of the 1980s. In due course the
Gulf War offered the pretext for an open break with the United
States, especially over the presence of its troops in Saudi Arabia.
The fall of Kabul in 1992 deprived the commitment of foreign
militants of much of its point, but also opened the door to various
organisations to install themselves on Afghan territory, particularly
in the east, thus enabling them to escape the influence of Pakistan. In
particular, some Pakistanis set up camps in Afghanistan to train their
mujahidin for service in Kashmir.60 This applied particularly to Hezb
ul-Mujahidin, the military wing of Jamaat-i Islami, and to Harakat
ul-Ansar, which was still known by this name after it officially reentitled itself Harakat ul-Mujahidin. The Harakat ul-Ansar militants,
led by mawlawi Jabbar in Afghanistan and Qari Fazlur Rahman
Khalil in Pakistan, came mainly from the Punjab and underwent a
six-month military course, usually at Darwanta near Jalalabad, in a
camp initially set up by Hekmatyar for the Arabs, and then closed by
the Taliban after the capture of Khost in September 1996.61 Camps
58

See, in particular, Washington Quarterly, winter 2000.

On this issue, see John Cooley, Unholy Wars, London: Pluto Press, 2002.
60 Since these training camps were therefore outside Pakistans territory, this in theory at least, exempted them from Indias strictures.
61 The News International, 23 August, 1988.
59

The radicalisation of the Taliban

303

were also maintained at Muzafarabad and near Khost by two splinter

groups of Harakat ul-Ansar, namely Jamiyat ul-Mujahidin, led by
Mufti Bashir, and Harakat Jihad Islami, led by Qari Saif ul-Islam
Akhtar. However, the origins of the Pakistani fundamentalists and
their evolution had little to do with the Afghan crisis, and Afghanistans role must be seen in perspective, although Zia ul-Haqs policy
of establishing madrasas was an indirect consequence of the Soviet
invasion. In particular, sectarian violence between Shiites and Sunnis
was neither a result of war in Afghanistan nor a by-product of the
success of the Taliban.
The Pakistani madrasas were directly linked to the Afghan conflict
through the participation of their taliban in the jihad, which was the
natural extension of their teaching. While the volunteers were mainly
Afghans, there were also Pakistanis: the majority of these were Pushtuns from the NWFP and Baluchistan, but there were also Sindhis
and Muhajirs. In September 1994 when Kandahar was captured,
the first taliban were the product of madrasas throughout NWFP
and Baluchistan.62 According to the Deputy Chief of the Citizens
Police Liaison Committee of Karachi, between 600 and 700 taliban
were sent to Afghanistan in May 1997.63 By the end, during the
offensive of August 1999, Pakistani and Afghan taliban came in their
thousandsperhaps as many as 5,000to reinforce mullah Omars
troops.
The presence of these militants in Afghanistan gave rise to problems both with their countries of origins and with the United States.
During a journey to Egypt in November 1993, for example, it was
significant that Rabbani signed an extradition agreement with the
Egyptian government, committing himself to the expulsion of extremist groups. The perpetrators of various anti-American incidents
had been trained in the Afghan camps, including Ramzi Yusuf, one
of those who carried out the 1993 attack on the World Trade Centre, and Mir Aimal Kainsi, a Pakistani national who was accused of
machine-gunning offices belonging to the CIA in January 1994.64
62

The madrasas included, among others, Maulana Nur Muhammad Saqib, at Katcha
Garhi Camp, Zia ul-Madaris at Peshawar, Hashmia Madrasa at Bara, and Dar ulUlum Haqqaniya at Akora. See The News International, 11 December 1994.
63 Owais Tohid, The Herald, December 1997.
64 He was later arrested in Pakistan, in June 1997, and was extradited to the United
States.

304

The Clerical State

In response to such criticisms Pakistan decided at this point to deport jihadi fighters from Pakistan to Afghanistan, a move whose principal long-term effect was to displace the problem and to place these
groups still further beyond control. Pakistan was able to escape inclusion in the State Departments 1994 list of countries supporting
terrorism, which would have led to the cutting off of international
financial aid essential for Pakistans economic survival.
After the fall of Kabul to the Taliban, various foreign but nonPakistani radical groups, whose numbers had dropped after the fall of
Kabul in 1992, returned to Afghanistan. These included in particular
the militants of the Uzbek Islamic Movement, numbering 2,000,
and around 3,000 Arabs of various origins who set up bases with the
Talibans consent.65 The foreign fighters were organised by Osama
Bin Laden, who had come to wage the jihad in Afghanistan at the
beginning of the 1980s, and had been Abdullah Azzams deputy when
the latter had been the head of the Maktab-i Khidamat-i Mujahidin
(Mujahidin Services Office). In contrast to Abdullah Azzam, Bin
Laden expressed a desire immediately to expand his field of action to
include the United States, perhaps under the influence of the Egyptian radical Ayman al-Zawahiri. In 1987 he declared himself independent of Abdullah Azzams group, and established Al-Qaida (The
Base). Azzam was assassinated in 1989.66
Bin Laden, the heir to a large family fortune, employed both his
organising ability and his money to develop his group, and above all
to seek a safe haven from which he could launch his global jihad. In
order to carry on his struggle, he had set up a network which was
basically Arab but was also genuinely transnationala rare event
and had for some years been looking for a base. After a failed attempt
to establish himself in Sudan in 1989, he returned there in 1992 and
developing a civil engineering business in parallel to his clandestine
activity. In spite of his good relations with Hasan Turabi, he was expelled from Sudan in 1996 in response to pressure from the United
65

Anthony Davies, Foreign fighters step up activity in Afghan civil war, Janes
Intelligence Review, vol. 13, no. 8, August 1, 2001.
66 Rohan Gunaratna in Inside Al-Qaida, London: Hurst, 2002, p. 25, pursues the
theory that Azzam was assassinated on Bin Ladens instructions, but the weakness of this theory is that it is founded on the evidence of renegades. For a more
reliable account of the relations between Bin Laden and Azzam see Stephen
Engelberg, Holy Warriors, New York Times, 14 January 2001.

The radicalisation of the Taliban

305

States, and subsequently based himself once more in Afghanistan,

where he took advantage of the absence of an organised state and of
the Talibans blunt political style. The process of convergence between two movements which were very different in their origins
and their membership was never easy. Personal relationships later
came into existence between Bin Laden and mullah Omar:67 Bin
Ladens marriage to mullah Omars daughter cemented the ties between the two movements.
The fatwa of 23 February 1998, signed by various Al-Qaida
officials including Bin Laden himself, conveys an idea of his view of
the world and of his objectives. His grievances were precise and political; namely the presence of American forces in Saudi Arabia, the
destruction of Iraq by the sanctions, and the occupation of Palestine.
There followed an appeal to launch indiscriminate attacks against
the Americans, in the name of jihad.68 In most respects, Bin Ladens
claims do not differ greatly from those of the Arab nationalists.
More broadly, Bin Ladens religious rhetoric is a response to American financial, military and cultural domination that has been perceived by Muslim populations, and particularly by the Arabs, as a
threat to their identity, opening the way to the adoption of a religious vocabulary by opposition movements69. The Middle East, by
reason of its proximity to Europe and its richness in crucial commodities, falls within what might be called the exploitable hinterland of the Western powers. In the event, after the fall of the Soviet
Union, the United States and its allies, especially Turkey and Israel,
enjoyed unchallenged hegemony. However, the United States has
been incapable of fulfilling the role of an arbiter in the Middle East,
67

It is unlikely that financial considerations came much into play. Actually Bin
Ladens financial resources were certainly less than the profits from smuggling
across the Pakistani frontier, especially opium.
68 For a translation and a commentary on the text, see Magnus Ranstorp, Interpreting the Broader Context and Meaning of Bin-Ladens Fatwa, Studies in
Conflict and Terrorism, 21, 1998, pp. 321330.
69 Franois Burgat rightly lays stress on the affirmation of identity by contemporary Islamic movements, which have often taken up the programmes of nationalist and Marxist movements. See Franois Burgat, Lislamisme au Maghreb, Paris,
Karthala, 1988; and, in relation to Turkey, Gnter Seufert, Politischer Islam in der
Turkei: Islamismus als symbolische Reprsentation einer sich modernisierenden Muslimischen Gesellschaft, Istanbul, Franz Steiner Verlag, 1997.

306

The Clerical State

especially in the Israeli-Palestinian conflict. In addition, the US has

supported authoritarian regimes, often unpopular, such as those in
Egypt and Saudi Arabia. The Gulf War of 1991 resulted in the emasculation of the only Arab country which had the ability to formulate
an independent policy, while ending the alliance between the United
States and the fundamentalist movements, which had come into being in Afghanistan during the 1980s.70 Following the Gulf War, and
in breach of the initial agreements, the United States kept its forces
in Saudi Arabia, which provided a casus belli for the Islamic movements, and foremost among them that of Bin Laden. The chronology testifies to the expansion of anti-American actions after the Gulf
War. The first attack on the World Trade centre took place in 1993.71
Bin Ladens originality is to plan and act at a global level. This was
a striking departure from the practices of other radical groups, for
example the Egyptians, whose membership and objectives were local.
Bin Ladens strategic innovation, which seems unlikely now to cease
with his death or even with the destruction of Al-Qaida, was to strike
directly at the United States, and at its embassies and warships, even
before 11 September 2001. In an interview in 1998, Bin Laden developed these arguments and elaborated on his opinion of his adversary.72 A crucial issue, which offers a partial explanation of his strategic decisions, was his perception of the United States as a morally
weak power, which was unable to accept military losses in a conflict.
This view was based directly on events in Somalia, where Al-Qaida
appears to have played a part in the attacks on the Americans, which
led to the withdrawal of American forces after the humiliating incident on 3 October 1993 when 18 American soldiers were killed. It
is likely that the plan to target the World Trade Centre was already in
gestation by this time. However it was in Africa that Bin Laden
would strike his first truly spectacular blow against the Americans.
70

John K. Cooley, Unholy Wars: Afghanistan, American and International Terrorism,

London, Pluto Press, 2002 (1999).
71 The reader should not leap to the conclusion that anti-American operations
multiplied in this period. In fact, in the 1990s, the number of attacks registered as
terrorist by the State Department fell by 29 per cent in relation to the 1980s.
During the Clinton presidency, in spite of alarmist language, the number of dead
averaged 14 per year. Andrew J. Bacevich, American Empire, Harvard University
Press, 2002, p. 118.
72 The interview was conducted with an ABC reporter. See www.pbs.org/wgbh/
pages/frontline/.

The radicalisation of the Taliban

307

On 7 August 1998, eight years after the entry of US troops into

Saudi Arabia, the US embassies in Nairobi and Dar es-Salaam came
under simultaneous attack, resulting in the deaths of 247 people,
including twelve Americans. On 20 August US missiles struck a
number of camps in Afghanistan as well as a pharmaceutical factory
in Sudan. The military value of this response was doubtful: twenty
people died in Afghanistan, but none of them was from the Al-Qaida
leadership, and the attack in Sudan, based on faulty intelligence, destroyed only a genuine pharmaceutical plant.
US strategy, as it was deployed up till 2001, was not aimed at the
overthrow of the Taliban but at the maintenance of sufficient pressure on them to bring about the expulsion of Bin Laden. However,
the 1998 missile attacks had a political result which was actually the
reverse of the US objective, namely that Bin Laden became a popular
figure in Pakistan and the Gulf, in the eyes of Afghan and Pakistani
fundamentalists, and more generally to anyone with anti-American
views. Posters of Bin Laden went on sale in shops in Peshawar, where
they were openly displayed, while the given name Osama became
unprecedentedly popular in the frontier provinces. However, as has
been seen, it was on the Pakistani fundamentalist movements that
the Taliban relied, especially for the organisation of large-scale offensives. This, no doubt, is why Prince Turki, then the head of the Saudi
secret services, who had a relatively encouraging reception when he
visited Kandahar in June 1998 to ask for Bin Ladens expulsion, saw
his initiative turned down at the end of the summer.
The United States had two viable strategic options after the attacks
in Africa. The first was to combat the Taliban by supporting Masud
while putting pressure on Pakistan. However, its relations with Masud
had never been good, and it did not want to confront Pakistan. The
second option was to recognise the Taliban and speed up the reconstruction of the Afghan state, which would have strengthened the
hand of those who opposed the presence of radical factions within
the Taliban. However, this was a strategy which required time, and was
not politically acceptable at home given the growing wave of antiTaliban opinion in the US media. In fact the United States chose a
third option, namely to put increasing pressure on the Taliban. In the
event this strategy failed, since diplomatic isolation only increased its
radicalisation while at the same time, thanks to assistance from Pakistan, its ability to act on the ground was undiminished. The rational

308

The Clerical State

approach of the Americans failed probably because of their misunderstanding of the ideological constraints under which the Taliban
operated, as well as the virulent mistrust inspired in general by US
policies. In addition, the Taliban believed at this point that it was
invulnerable, both because a US intervention was very unlikely and
because it was convinced that it could withstand an American invasion, just as it had fought off the Soviets.
With no convincing military option, the United States then took
the decision to place its faith once more in the sanctions embodied
the UN Security Councils resolution 1267, unanimously adopted
by the Council on 15 October 1999, which had been reinforced by
resolution 1333 on 19 December 2000. Specifically, these sanctions
provided for an embargo on the importation of weapons into Afghanistan, the reduction of the size of foreign diplomatic missions in
Kabul, and the closure of Taliban offices abroad. Another measure
froze Taliban financial resources abroad, while the airline Ariana was
no longer permitted to fly outside Afghanistans frontiers. These sanctions achieved only a marginal economic effect, without the depth
of those which had inflicted grave damage on Iraq in the 1990s. In
fact Afghanistans infrastructure had already been largely destroyed,
while for political reasons it was difficult to halt the work of NGOs
in a country on the verge of famine because of a drought of historic
proportions. Smart sanctions were particularly difficult to apply because of the scale of illicit cross-border trafficking. Nevertheless, the
report issued by the UN on the effects of the sanctions demonstrates
that the population believed they were more comprehensive than
they were in reality.73
On the Taliban side the attitude of the west was not seen as a preliminary to negotiation, as the US government would have wished it
to be, but rather as an existential threat. After the closure in January
2001 of its office in New York, the Taliban was no longer even recognised as an interlocutor, which in practice brought to a halt the
peace talks set up by F. Vendrell within the framework of the UNs
mediating role. Contacts with US representatives in Pakistan continued to the end, but remained unofficial.
73

United Nations, Vulnerability and Humanitarian Impact of UN Security Council

Sanctions in Afghanistan, prepared by the Office of the humanitarian coordinator
for Afghanistan, Islamabad, Pakistan, 16 September, 2000.

The radicalisation of the Taliban

309

The position of the Taliban concerning Bin Ladens extradition remained virtually unchanged, and was also supported by the government of Pakistan before 11 September 2001.74 The Taliban put
forward compromise proposals, having ruled out both the option of
bringing Bin Laden to justice in Afghanistan, and that of his direct
extradition to the United States. One suggestion was that after an
initial hearing by ulema from Afghanistan, Saudi Arabia and a third
country, Bin Laden could be deported to a Muslim country. The
Minister of Foreign Affairs, mullah Abdul Muttawakil, made a further suggestion, proposing a mechanism whereby he would be placed
under surveillance by the Organisation of the Islamic Conference.
It appears that the foreign Minister had additionally proposed that recognition might be extended to the Taliban government in exchange
for Bin Ladens expulsion. Whether because mullah Omar did not
back this suggestion, or because the United States rejected it, this
proposal came to nothing. The impasse reflected Bin Ladens growing
influence over the regime, the nationalist response to the US attacks
of 1998, and the importance to the Taliban of transnational solidarities, at a time when the war was not yet over.
In addition to its support for fundamentalist movements, there
was a further issue which caused problems for the Taliban. This was
that Afghanistan had become the worlds leading producer of opium,
overtaking Burma. According to the UN Drug Control Programme,
half the worlds heroin was produced in Afghanistan, while 80%
of the heroin consumed in Europe was of Afghan origin.75 In its approach to this issue the Taliban displayed unaccustomed efficiency,
dramatically cutting down opium production. This was a demonstration of the degree of control it exercised in the countryside, but
worsened the situation of the rural population who were already
affected by the drought. However, these measures brought the Taliban no benefit in terms of diplomacy, since the United States was
interested only in Bin Laden.
By 2001, following the renewal of the sanctions in December
2000, the radicalisation of the regime was undeniable. In this context
the destruction of the Buddhas of Bamyan was symbolic of a defini74

See in particular the interview with President Musharraf in the Washington

Times, 21 March 2001.
75 The Economist Intelligence Unit, 1999, p. 34.

310

The Clerical State

tive break with the international community. The act was evidently
political, since previous decrees of mullah Omar, issued in July 1999,
had offered protection to works of art in general and specifically to
the Buddhas. Following the decree of 26 February 2001 ordering
their destruction, they were finally dynamited in March after repeated attempts to dissuade the Taliban from the destruction of a
unique monument.76
Relations with the NGOs also worsened in 2001 as a result of the
overall deterioration of the situation. The policy of religious conversion implemented by some organisations such as Shelter Now was
the cause of a further crisis. The practice of carrying out conversions
under the cover of humanitarian action was condemned by the
majority of the NGOs because of its intrinsic dishonesty and its
exploitation of the inferior economic status of the population. The
increasing presence of Protestant Christian preachers in a Muslim
country represented a return to a nineteenth-century practice, leading
probably to similar results.77 In the case of Afghanistan the security
of the great majority of NGOs whose operations were above-board
was put at risk by these organisations. As another consequence the
Islamic NGOs came to play an increasing role, whether they were of
fundamentalist inclination such as the Rasheed Trust, or non-fundamentalist such as Fetullahci and the Canadian Relief Foundation.
After the UN vote to re-impose sanctions Pakistan was the last
ally left to the Taliban, and its support largely accounts for the continued offensive capacity of the regime up to the summer of 2001.
However, questions began to be asked about the rationale for Pakistani assistance. The installation of a Taliban government in Kabul
had been intended to open up Central Asia to Pakistans economic
and political influence, but in practice the outcome had been the reverse, with the result that Pakistan found itself faced by hostility from
the Central Asian countries, except for Turkmenistan. The undisguised presence of Taliban mujahidin in Tajikistan fighting alongside
the Islamist opposition, and the links between the Uzbek Islamic
Movement and the Taliban, were seen as part of a deliberate bid to
destabilise Central Asia.78 After the capture of Kabul in September
76

See Pierre Centlivres, Les Bouddhas dAfghanistan, Paris: Favre, 2001.

This presence was reinforced after 11 September 2001; see Time, 4 August 2003.
78 The Taliban commander in Uzbekistan, Namangani, was put in charge of military
operations in the north of the country during the war against the United States.
77

The radicalisation of the Taliban

311

1996 Russia and the countries of Central Asia immediately organised

a summit meeting on security issues, whose tone was clearly hostile
to Pakistan. During a further summit in 1997 accusations were made
against Pakistan over its Afghan policy, in spite of the mollifying remarks of Nawaz Sharif. Finally, the offensives of 1999 and 2000
aroused protests from the regional powers and the UN against both
Pakistan and the Taliban.
In reaction to Pakistani policies some states developed a strategy
of containment of both Afghanistan and Pakistan. Iran represented
no political threat due to its broadly conservative position, but took
the decision to make economic overtures to Central Asia. The construction of a railway from Sarakhas in Turkmenistan to the Iranian
port of Bandar Abbas, together with the development of the port,
offered Turkmenistan an alternative outlet to that offered by Afghanistan. Turkmenistan and Iran also established a free trade zone at
Sarakhas, and in 1999 India signed an agreement with Iran to permit
the transit of goods from central Asia.
Finally, Pakistan lacked the resources to provide sufficient aid for
Afghanistan, particularly for the infrastructural repairs required for
the transport of goods to Central Asia. Even the section of road from
Kandahar to the Pakistani frontier was in bad repair, and was thus a
considerable obstacle to trade with Quetta. The National Highway
Authority, charged with the rebuilding of the road to Turkmenistan,
failed to produce any practical results, despite negotiations begun in
December 1996; while the so-called Afghan Trade Cell, established
in July 1997,79 was never allowed a budget sufficient to make any real
impact on cross-border trade.

79

This was the successor to the Afghan Trade Development Cell set up within
the Ministry of the Interior by Nasrullah Babar.

Part V. THE AMERICAN INVASION

AND THE RETURN TO
FRAGMENTATION
Perhaps as a result of the absence of TV stationsbanned by the
Talibanthe attacks of 11 September 2001 were not immediately
seen in Afghanistan as a major event, and the displays of emotion seen
elsewhere did not occur. The population, especially in the countryside, were preoccupied with another and more immediate problemthe drought. Much of the country was free of conflict, with
fighting localised in a few provinces. Although the Taliban was internationally ostracised after its destruction of the Buddhas of Bamyan,
it was nevertheless in control of most of Afghanistan, with the result
that journeys between the large towns could be undertaken without
hindrance. Certainly the regime had become more radicalised, but
Bin Ladens role in domestic Afghan politics was minimal. Real ideological differences separated him from the Taliban, and he was regarded with distrust and hostility by some members of mullah Omars
circle who suspected,not without justification,that he would not hesitate to sacrifice them in the interests of his wider strategy.1 In spite
of potentially grave consequences, mullah Omar refused to yield to
American pressure to expel Bin Laden. For Afghanistan the 9/11
phenomenon really began with the American bombing and the
rapid sequence of events which led to the Talibans downfall.
During the American intervention and throughout the following
months the debate hinged on two questions. Would an anti-American
guerrilla movement emerge after the military defeat of the Taliban?
And would a return to political fragmentation be a more probable
outcome than the re-establishment of a state? According to the most
widely-held theory, stabilisation would rapidly ensue. For example,
Olivier Roy argued, It is unlikely that a Taliban resistance will come
1

Hence the intelligence passed on a little before 11 September 2001 by the Taliban Minister for Foreign Affairs, Mottawakil, concerning a planned attack in the
United States.

312

The American Invasion

313

into being, and that the Talibans defeat would signal the probable
disappearance of the radical Islamists in Afghanistan.2 The opposition represented a viable alternative which would preclude a return
to civil war, and the end of the war was at hand: There is no longer
an Afghan issue, except on the symbolic level.3 According to other
interpretations, which took a different line, a strong fundamentalist
tendency would continue,4 while it was presumed that the American intervention would result in a return to the political fragmentation of the 1990s, with a possible return to civil war.5 These two
analyses also led to contradictory positions over the success of Bin
Ladens strategy. In the first case he was viewed as having lost the
battle: The military campaign of October and November 2001 has
considerably weakened the transnational Islamic networks.6 But in
the second case, because of the triangular nature of the conflict, the
defeat of the Taliban will not in itself guarantee the victory of the
United States.7
Developments on the ground have disproved the first of these hypotheses. The collapse of the Taliban regime in the event gave rise to
a resurgence of locally-based power centres, independent in practice
of the Kabul government, although lip-service might be paid to it.
International aid and the presence of foreign troops have not so far
led to reconstruction of the state. The government, subject to strong
internal divisions, exercises no authority outside the capital and is
unable to control recurrent clashes between warlords. The policy of
liberal modernisation supported by the UN faces strong resistance,
even inside the new regime, and the continued existence of a strong
fundamentalist tendency is manifest. Able to fall back on sanctuary
within Pakistan, the neo-Taliban continue to control the Pushtun
territories, while at the same time US troops carry out military operations against an openly antagonistic Pushtun population, inflicting
2

Mariam Abou Zahab and Olivier Roy, Rseaux islamiques, Paris: Autrement, 2002,
p. 70 and p. 68 (published in 2004 as Islamic Networks: the Afghan and Pakistani
Connection, London: Hurst).
3 Olivier Roy, Les illusions du 11 septembre, Paris: Seuil, p. 32.
4 Interview with Pierre Centlivres, Le Monde, September 2001.
5 Gilles Dorronsoro, Aprs les Taliban. Fragmentation politique, hirarchie communautaire et classes sociales en Afghanistan, Cultures et Conflits, January 2002.
6 Mariam Abou Zahab and Olivier Roy, op. cit., p. 71.
7 Gilles Dorronsoro, op. cit., p. 1.

314

The American Invasion

a concomitant burden of collateral damage. An American withdrawal would unleash chaos, and probably lead to the recapture of
Kandahar by mullah Omar. Finally, Al-Qaida has discovered a real
social base as well as sanctuary on the Afghanistan-Pakistan frontier,
where Bin Ladens popularity remains undiminished, confirming the
largely illusory character of the US victory in the autumn of 2001.

10. A Splendid Little War?*

Bin Ladens strategy, which was both audacious and logical, has so far
been successful. The evidence militates against any account based on
religious fanaticism. Other massacres in recent history, whether
comparable or even larger in scale, have not led to the presumption
of irrationality in those responsible. Leaving aside moral considerations, the attacks of 11 September 2001 were a strategic operation,
executed with great economy of means; and it is historically rare for
a small group to achieve such a massive impact on the international
situation. The attacks employed few men and little money but struck
a devastating political and economic blow. They left 3,000 dead, resulted in direct and indirect costs of several hundred billion dollars,
and last but not least brought about a significant and probably lasting
restriction of personal liberty in the United States. One of the most
unsettling effects of this strategy has been its ratchet effect, where
the perception of threat rules out a return to the pre-11 September
situation.
The objective of Al-Qaida was radically to change popular perceptions, in order to make the distinction between Muslims and nonMuslims a central element in political mobilisation, leading in turn
to a diminution of western influence in the Islamic world. This was a
strategic tactic similar to that of some anti-colonial struggles, where
radicalisation was achieved through the politicisation of religious and
communal identities. Seen from this standpoint, it was necessary that
the attack should be spectacular and cruel in order to inflame antiMuslim sentiments in western opinion. The growing gap between
western and Muslim views suggests this aspect of the strategy has been
effective. The wild statements persistently uttered by those American
personalities who present the conflict as a confrontation between
*

A phrase used in a letter from John Milton Hay to President Theodore Roosevelt
of the Spanish-American war of 1898.

315

316

A Splendid Little War?

Islam and the Judaeo-Christian camp tend to reinforce the accepted

perception of a clash between adversaries. The situation of Muslims
in the United States has been undermined by programmes of surveillance, which are resented as harassment. Another factor is that visits by
Muslims to the United States have become more problematic.8
Furthermore, the attacks guaranteed Bin Laden a global audience
as well as access to media which have permitted him to exploit symbols relating to Islamic legitimacy, such as his cave and his Quranic
quotations. This scenario ensured that a myth would spring up, and
that there would be regular reappearances. Bin Ladens death, whenever it takes place, might never be proved for certain. In any case he
did not address himself to the western public but exclusively to Muslims, who were unaware of the American message. Clumsy attempts
to censor Al-Jazeera during the US military operations, as well as the
mobilisation of the media industry and of communication consultants, changed nothing. The reality of American domination led to
resistance. The refusal to subscribe to any form of negotiation reinforces the integrity of Bin Ladens message, which served as a symbolic rejection of the present balance of power between the West
and the Islamic world.
In consequence Muslim views took a different slant on the events
of 11 September, in which Bin Laden was seldom seen as responsible
for the attacks, while paradoxically he was admired at the same time
for his boldness in opposing the United States. An opinion poll conducted in Pakistan showed that half of those questioned believed
Mossad was responsible for the attacks.9 Further, widespread credence has been given in the Middle East, even among educated and
westernised sections of society, to rumourswhich it is scarcely
necessary to say are falsethat Jews in New York did not go to work
on the day of the attacks.10 Bin Ladens popularity is startling, even in
8

The number of Arab students in the United States went down by a third between 1999 and 2002 (Report on human development in the Arab world,
October 2003, cited in Le Monde, 23 October 2003).
9 Of the persons asked 12% said Bin Laden was responsible for the attacks; 24%
saw them as an action which was part of the jihad; 48 % thought that Mossad was
responsible; and only 6% of the population thought that Bin Laden was a terrorist. Gallup International, quoted by www.tns-sofres.com, 17 October 2001.
10 The phenomenal success in Europe of a book denying that a plane crashed on
the Pentagon shows that this kind of fantasy is not exclusively Middle Eastern.

The strategy of the United States

317

the light of the diplomatic and psychological errors made by the

United States. In a survey carried out in Indonesia, Morocco and
Pakistan and among the Palestinians, Bin Laden was nominated as
one of three personalities who would be trusted to do the right
thing.11
In addition to such political and symbolic issues, Bin Ladens aim
was to lead the United States into a prolonged deployment of its
military force in various parts of the Middle East, and thus to transform a cultural and economic hegemony into direct military domination, which would be both more costly and more liable to provoke
resistance. Bin Laden, convinced that the United States was politically fragile, hoped for an American invasion of Afghanistan which
would draw it into a trap. Faced with this challenge, what was the
American response?

The strategy of the United States

Far from opening a debate in the United States on the countrys hegemony and the resistance to which it gave rise, 11 September was
the signal for a firmer commitment to imperialist policy.12 From this
moment the goal of the United States was global military hegemony,
arrogating to itself a unique position in international relations based
on what it saw as its exceptional status.13 Americas political and military doctrine thus represented an open break with international law.
A central provision of the UN Charter, the restriction of the use of
force to self-defence, was in effect flouted by the adoption of a docSimilarly, a majority of Americans, no doubt influenced by the ambiguous
language adopted by their government, believe that Iraq was implicated in the
attacks of 11 September.
11 Pew Global Attitudes Project, consultable at www.people-press.org, quoted in
International Herald Tribune, 4 June 2003.
12 See in particular the analyses by Andrew J. Bacevich, American Empire, Harvard
University Press, 2002. It could also be asked whether the increasing militarisation of American politics is not primarily a reflection of the relative decline of
the United States in the international scene. On this issue see Charles Kupchan,
The End of the American Era: US Foreign Policy and the Geopolitics of the TwentyFirst Century, New York: Random House, 2002.
13 The National Security Strategy of the United States of America, Washington, DC: the
White House, September 2002. The central concept was to deter any possible
future challenge.

318

A Splendid Little War?

trine of preventive action. This development had in fact already begun to take shape under President Clinton, but had not hitherto
been proclaimed in so clear and coherent a fashion. In practice the
American doctrine implied that states on the perimeter of the West
would enjoy only limited sovereignty.14 For instance, on 6 November 2002 a missile fired by a CIA plane hit a vehicle in Yemen, killing six persons suspected of participation in the attacks on the USS
Cole in 2000.
Although this hegemonic project attracted a wide consensus among
American lites, the specific policies of the Bush administration attracted opposition. In the event, the United States distanced itself
from its traditional allies, especially its NATO partners, since it came
to view them as militarily insignificant and politically untrustworthy.
Alliances were perceived as constraints and, especially under the influence of the Defense Secretary Donald Rumsfeld, military doctrine
prioritised the mobility of forces and relatively light operations. This
unilateralist posture was widely rejected abroad, so that the United
States found itself diplomatically isolated, as was to be seen some
months later in the Iraq crisis.15
The result of the attacks of 11 September had been to impose a
new international agenda, where the issue of terrorism took a central position. Semantically, the expression the war against terror was
in itself odd, since terrorism is not an adversary as such but a method
of action. However, the slogan has many advantages. In fact, the linkage between a doctrine of preventive action and a fluid definition of
the enemy permitted the legitimation as defensive of all kinds of steps,
even thosesuch as the measures taken against Iraqwhich were
most overtly hegemonic. In the name of these new priorities the
United States continues to extend its military involvement throughout the worldin Central Asia, in the Caucasus and in the Middle East.
Although the license to characterise movements as terrorist has
its advantages, the drawing up of lists of countries and organisations
14

Which did not deter the United States from acting to reinforce their internal
strength, especially for the repression of dissident movements.
15 A major problem for the United States was precisely the increasing impossibility
of reconciling its image with that of the benevolent empire, to quote Robert
Kagans expression, and with the overt antagonism of which it was the object
around the world.

The strategy of the United States

319

may rebound dangerously through the creation of common interests between previously disparate groups.16 In fact Al-Qaida is the
antithesis of regionally based movements such as Hezbollah and Hamas, whose objectives are limited and negotiable. While these groups
strongly denounce Bin Laden, as for example in the declarations of
Shaikh Fadlallah, the construction of lists of terrorist organisations
creates in practice a degree of solidarity between such organisations.
Instead of playing on the essentially contradictory interests of such
movements in order to target those which directly threaten the western countries, the strategy of the war against terror, by promoting a
degree of cooperation between networks, actually increases the risks
of bringing into being an anti-American front.
Even before the redefinition of American strategy in the months
that followed the attacks of 11 September, the first priority was the
elimination of Bin Laden, who was soon seen as the instigator of the
attacks. To deal with the Afghan aspect of the problem, the United
States was able to call upon a range of capabilities and alliances. However NATO, in spite of a historic declaration of solidarity based on
Article 5 of its founding treaty, was completely ignored by the
United States, a move which upset various sensibilities. Britain and
certain other countries offered some marginal assistance, but the war
remained in essence an American affair.
In the weeks following 11 September, the United States moved
with great efficiency to put in place the necessary diplomatic and
military preparations. Afghanistan is a landlocked country, and there
were political difficulties in basing a significant number of troops in
Pakistan, or even in Central Asia. In any case, the United States was
quick to rule out a land operation of any significant size. In fact the
Kosovo conflict was adopted as the strategic model, with the corresponding strategy of air strikes combined with the mobilisation of
local allies. The vast air power of the United States and the employment of Special Forces to coordinate operations on the ground were
therefore much in evidence. The number of troops committed never
exceeded 50,000, and losses were limited to a handful, often the victims of friendly fire.
At first the United States brought its logistical capability into play to
set up bases in Afghanistans immediate neighbourhood. On 5 Octo16

In December 2001, the State Department published a list of thirty-nine organisations viewed as terrorist.

320

A Splendid Little War?

ber a first contingent of 1,000 men arrived at Khanabad in Uzbekistan. At the same time, although the US presence at the Jacobabad
base remained separate and limited, Pakistans military cooperation
was confirmed, and its army attempted to close the frontier to prevent the flight of Al-Qaida militants. A further factor was that Iran,
which had been on the point of recognising the Taliban regime,
hastily entered an agreement on the possible recovery of US pilots.
Aircraft-carriers were deployed in the Indian Ocean and extra resources were given to the CIA.17 In particular, the Egyptian, Jordanian and Algerian intelligence services received subventions from
the United States, and the task of interrogating suspects was delegated to them, often with the use of torture.
On the diplomatic level the international isolation of the Taliban
was completed within a few weeks. At the end of September Saudi
Arabia broke off relations with Kabul, shortly after the United Arab
Emirates had taken the same step. Pakistan offered a channel for negotiations up till the outbreak of war, when the Afghan ambassador
in Islamabad was arrested and handed over to the United States.
Within a few weeks the United States was militarily ready to attack
the Afghan regime. However, the complexity of the situation dictated that its strategy was developed with caution.
The initial strategic issue was relatively easy to formulate, but its
resolution was highly complex. The war in Afghanistan, the first
conflict of its kind, was exceptional because of the confrontation of
three protagonists of different kinds. These were a state (the United
States), a quasi-state unrecognised internationally (the Taliban), and a
transnational movement (Al-Qaida). In addition, the Bin Laden
networks were active in the south and east of the country, where the
Taliban had a real power-base, while the opposition, which was militarily weak, was based in the north. Theoretically, the only effective
stratagem for the elimination of the radical networks would be to
make an alliance with the Taliban, which was in undisputed control
of the terrain, as its eradication of opium production in eastern and
southern Afghanistan had demonstrated. However, the Taliban was
difficult to talk to, and time was short. A variation on this approach
would be to destabilise the Taliban from within and retrieve the
so-called moderate elements, in other words those who would
agree to an alliance with the United States.
17

Robert Woodward, Bush at War, New York: Simon and Schuster, 2002, p. 77.

The strategy of the United States

321

The second possibility was to use the Northern Alliance to overcome the regime by military means, but it was politically divided
and militarily weak and could provide no access to the south, since
Masuds troops would never fight in Kandahar, and therefore was not
a means of overcoming Al-Qaida. In addition a northern victory
would involve the rejection of the Pushtuns. According to Charles
Tenet, head of the CIA, there would be opposition from the Pushtuns if the Northern Alliance took power.18
American strategy was largely improvised and led to clashes between the State Department and the Pentagon. The first option was
preferred by the State Department, which wanted to make use of
Pakistans assistance; while the second was the choice of the Pentagon, which wanted more rapid results. The first strategy, which would
have had the advantage of avoiding the fragmentation of the country,
did not pay off quickly enough, and the second, though militarily viable in the short run, opened the way to fragmentation.
Up to the end of September the United States, in parallel with the
expansion of its military capacity, opted to negotiate through Pakistan in the hope of persuading Bin Laden to give himself up to the
United States. To this end a delegation of Pakistani religious figures
together with Mahmood Ahmed, the director of the ISI, paid a fruitless visit to Kandahar on 17 September. A second visit on 28 September was equally unproductive, as were numerous subsequent contacts.
The situation nevertheless remained fluid since on 20 September
the Afghan ulema, after a meeting in Kabul, declared that Bin Laden
should leave of his own accord. Pakistans President Musharraf afterwards asked for Mahmood Ahmeds resignation, replacing him with
Lieutenant-General Ehsanul Haq, whose brief was to counter proTaliban elements within the ISI. Two elements played their part in
this failure. One of these was the ambiguous part played by the Pakistani ulema, who rather tended to support the position of mullah
Omar.19 The other factor was the political situation within the United
States, which was such that the American government preferred to
give the Taliban an ultimatum rather than negotiate, hence the refusal to provide any kind of proof of Bin Ladens implication.
18
19

Woodward, op. cit., p. 123.

The Nation, 1 October 2001.

322

A Splendid Little War?

In the last week of September there was a significant shift in the

position of the United States, which abandoned the search for a diplomatic solution and by the same token changed its objectives. On
20 September the Secretary of State Colin Powell again said that he
did not exclude a dialogue with the Taliban if it dismantled the radical camps and networks within Afghanistan,20 and added that the aim
of the US administration was to put pressure on the Taliban rather
than eliminate it entirely.21 Some days later Condoleezza Rice, the
National Security Advisor, issued a contradictory declaration: from
now on the goal was to be to overturn the Afghan regime.22
On 7 October the initial bombing raids took place and Operation Enduring Freedom was under way. During the first weeks of
the war, however, the United States continued to look for a Pakistani solution to the conflict, that is to say the replacement of the
Afghan regime by allies of Pakistan.

The failure of the Pakistani solution

The initial American plan, proposed by the State Department, was
to take place in two stages: first the overthrow of the Taliban by subversive measures and the liquidation of the radical networks by commando operations; then the installation of a pro-Pakistan regime in
Kabul. Under this plan the Northern Alliance was to be largely
ignored in order not to upset Pakistan. This was the reason for the
relatively superficial nature of the initial bombing, which irritated
the Northern Alliance. To the great displeasure of India in particular
the United States largely relied on Pakistan, while its strategy was predicated on the idea that only Pakistan could destabilise the Taliban,
to whose rise to power it had so greatly contributed. The military
side of the operation consisted of putting troops in position around
Afghanistan to facilitate in-depth raids, while relatively light bombing was used to destroy certain symbolic targets in the aim of sowing
panic among the Taliban. According to this proposal, the United
States would not clash with the Afghans but would operate directly
against the Al-Qaida network.
20

The Statesman, 21 September, 2001.

Woodward, op. cit., p. 124.
22 The Times, 24 September, 2001.
21

The failure of the Pakistani solution

323

In an attempt to sow division among the Taliban, the United

States and Pakistan on a number of occasions proposed the inclusion
of moderate Taliban in the future government. This was explicitly
mooted during a joint press conference conducted by President
Musharraf and Colin Powell at the end of September. In early October Powell once more took up the same idea, and expressed his opposition to the proposal that the Northern Alliance should take
Kabul in order to avoid a setback for the Pakistanis.23 Western diplomats were thus working towards the establishment of a coalition
government, in spite of the complete opposition of the Northern
Alliance to Taliban participation. An initial attempt was organised by
Gaylani, with Pakistani support, but it did not succeed in attracting
any of the acknowledged leaders or apparently even any significant
tribal chieftains. At the end of September Karzai held a meeting
with tribal leaders at Quetta and suggested setting up a loya jirga with
the participation of a section of the Taliban.24 Finally, on 1 October
negotiations between the royalists and the Northern Alliance produced a compromise, but excluded the Taliban. With the capture
some weeks later of Kabul by the Northern Alliance, this compromise solution became irrelevant.
While diplomatic exchanges gathered pace, Pakistan made a bid
to destabilise the Taliban. The CIA distributed $70 million in the
south to bring over the tribes, while infiltration of Taliban-held territory intensified.25 However, this strategy soon failed. Attempts to
destabilise the eastern region of Afghanistan were inconclusive. The
Taliban appointed as commander-in-chief Jalaluddin Haqqani, who
had substantial influence in the eastern provinces. In October a shortlived revolt in Khost was rapidly crushed. The tribal chiefs and commanders sent to persuade the Pushtun tribes to change sides had no
success. Abdul Haq was arrested and executed by the Taliban. Hamid
Karzai, the future President of the provisional government, was rescued from a perilous situation after a clandestine excursion to the
Kandahar region.
The reality was that the American operations on Afghan soil had
revealed an unsophisticated but courageous adversary, who did not
23

Woodward, op. cit., p. 192.

The News, 29 September, 2001.
25 Woodward, op. cit., p. 317.
24

324

A Splendid Little War?

give ground and whose morale at that point was excellent. An attack on 29 October against the buildings used before 11 September
by mullah Omar collapsed, and the US troops found themselves in a
trap: US spokesmen conceded that some thirty had been wounded.
The mystique attached by the western media to Special Forces operations should not obscure the fact that they require a different environment from that of southern Afghanistan, where there were
few military targets and western intelligence was exiguous.
The issue is: could the American approach have been effective?
The USPakistani strategy was based on the conviction that the Taliban was a transitory phenomenon, with neither a local basis nor an
ideology. Viewed in this way, the ulema were taken to be a marginal
force, who could be dislodged with the help of the khans. This was
particularly unrealistic, since it presumed rapid and efficient cooperation between the tribal chiefs, virtually a contradiction in terms. In
addition the strategy had been impeded by the US bombing, which
came too soon; this was the view taken by Abdul Haq in the last
interview he gave before his death, when he denounced the bombing for having hampered his approaches to tribal leaders, who were
obliged to support the jihad. Eye-witnesses report there were largescale mobilisations by the mullahs in the mosques, which considerably reduced the scope for the activities of non-religious lites.26 Before the tribal leaders could change sides it would be necessary at the
very least to wait for the fall of the Taliban.27
All that this crisis achieved was to illustrate the rapid changes in
the Taliban movement over the space of a few years, something that it
had hitherto found difficulty in seeing. The politicisation of the
movement, which was already apparent, had gathered pace. The professionalisation of its leadership was partly due to the involvement of
foreigners, but was also related to an intake of relatively well-educated
younger men. Outdoing even Bin Ladens still at communication,
the Talibans representatives played the media game well, although
media coverage was highly unfavourable towards ita measure of
the success of American efforts to prevent its message from getting
out. In contrast to the practice of Saddam Hussein in 1991, the Taliban showed its civilian casualties and exploited them.
26
27

Various personal reports communicated in Peshawar in November 2001.

Gilles Dorronsoro, Aprs les Taliban. Fragmentation politique, hirarchies communautaires et classes sociales en Afghanistan, Cultures et Conflits, January 2002.

The failure of the Pakistani solution

325

The fall of the Taliban

By early November 2001, almost a month after the start of the bombing, there was no observable result on the ground.To get results before
the beginning of Ramadan, and avoid the impression of dragging its
feet, the United States adjusted its strategy towards giving greater support to the Northern Alliance.28 There were few practical obstacles to
this option, and the United States faced no particular difficulty in
establishing a home territory for the opposition in northern Afghanistan, although the Taliban put up an honourable fight at Mazar and
in Kunduz.
Special forces were already on the ground with the Northern
Alliance, but the essential change in early November was the use of
air power. Extraordinary firepower was deployed by the B-51 bombers, which operated in conjunction with Stealth B-1s flying from
Diego Garcia and from the United States itself. In addition four
15,000-pound (7,000 kilo) Daisy Cutter bombs were dropped. Fragmentation bombs were also dropped, which continue to cause frequent casualties. By the end of December 12,000 bombs had been
dropped on Afghanistan, of which 6,700 were so-called precision
weapons.
On 9 November forces led by Dostum, Muhammad Atta and
Ustad Mohaqqeq took Mazar-i Sharif. The entire northeast collapsed rapidly, with the fall of Maimana on 11 November and Herat
the following day. On the same day the Shura-yi Nazar attacked
Kabul, taking the city in spite of warnings from Pakistan. Jalalabad
fell the day after Kabul. On 22 November Pakistan closed its Kabul
embassy. In the north the resistance was especially strong in Kunduz
because of the presence of foreign fighters, especially Arabs and
Pakistanis, who had no avenue of retreat. Some of these were evacuated to Pakistan by air, with US approval, in a nocturnal operation
lasting two weeks. On 9 December Kandahar fell after one last apocalyptic speech from mullah Omar.
In the shorter term this remarkable series of successes was not
fully carried through to victory. Due to its unwillingness to commit
28

At a meeting in Uzbekistan on 30 October 2001, General Tommy Franks, the

commander-in-chief of CENTCOM, is reported to have promised that US air
power would henceforth be placed at the service of the Northern Alliance. See
Michael Kranish, Turning the Tide, Boston Globe, 31 December, 2001.

326

A Splendid Little War?

ground forces, the United States was unable to capitalise on its successes and decapitate Al-Qaida and the Taliban movement. In late
November US ground troops first arrived in Afghanistan in significant numbers, but were put to guard prisoners. When Kandahar fell,
mullah Omar was able to flee to Uruzgan after negotiations between
Afghan ulema and pro-US commanders. Undoubtedly it was the militias who let him go, and American inaction at this juncture was later
seen to have been a grave error. Another factor was that even after
the campaign to capture Bin Laden had been launched, the Americans were unable to mount a decisive attack on the Tora Bora caves
to which his retinue had retreated. The Special Forces failed to seal
off the mountain complex, leaving operations on the ground to
Afghan commanders who were at odds with each other. The US officer in charge is reported to have refused offers of surrender from
Al-Qaida militants,29 while the final assault was not a success since
most of the Taliban fighters had bribed the Afghan commanders to
allow them to escape to Pakistan. The Pakistan government closed
the frontier, with the deployment of 40,000 men, but although several hundred fighters were captured as they crossed over, no senior
Taliban official was detained. Nevertheless Bin Laden, who had
been wounded, is reported still to have been in this region around
16 December, after which he went to the frontier.
No account was kept of civilian casualties, which the US army refuses in principle to estimate, but they were probably amounted to
several thousand. Enemy military losses also went unrecorded.30 However, a number of war crimes were committed by allies of the United
States. For example, on 25 November hundreds of Taliban prisoners
were killed in the prison at Mazar-i Sharif, after a revolt in which a
CIA agent who had been interrogating prisoners was killed. Apparently many prisoners were summarily executed once they had been
recaptured. The most serious incident concerned the deaths of
Taliban and foreign prisoners who were suffocated inside containers.
According to a meticulous inquiry,31 around 3,000 Taliban prisoners
29

Renaud Girard, Le Figaro, 19 June 2002.

By comparison the Iraq conflict of 2003 caused between 5,000 and 10,000 civilian deaths and 30,000 wounded, in a campaign also described as surgical.
31 See the account given by Jamie Doran in his filmed documentary, Massacre
at Mazar, with witness statements concerning the participation of American
30

The fall of the Taliban

327

were massacred by Northern Alliance forces, an atrocity which by

some accounts was perpetrated in the presence of American soldiers.
Despite the gravity of these reports and the known locations of
communal graves, the UN declined to carry out an inquiry in order
not to embarrass the Afghan and US governments.
The campaign lasted nine weeks. The loss of the north, which had
always been likely and had probably been foreseen by the Taliban,
deteriorated into a rout, with each defeat contributing to the panic.
The Talibans error was to have chosen to wage a conventional war
rather than making preparations for the inevitable transition to guerrilla warfare, a course of action it took anyway some months later.
Although the Taliban was never solely the instrument of Pakistans ISI (Inter Service Intelligence) it was in fact highly dependent
upon them for its military strategy. The Talibans principal vulnerability was military, resulting from the command of its troops by a coalition of commanders, a situation that reflected historic realities and
sprang indirectly from the controlling influence of the Pakistanis. As
a result of the ferocity of the US bombing the Talibans military
structure collapsed. Such disintegrations were not peculiar to the
Taliban but had been a recurrent factor in Afghan movements in defeat. For example, Dostums troops had scattered in 1997 and 1998,
as had those of Ismail Khan in 1995. Masud alone had been able to
retreat in good order after the loss of Kabul in 1996, thanks to the favourable terrain and his resolute professionalism. Another issue was
that the foreign fighters, who numbered several thousand and were
better organised, had never succeeded in organising the Afghans,
whose relations with the Arabs had always been difficult. It was the
collapse of the military apparatus which brought about the collapse
of the Taliban regime in the Pushtun region rather than any political
or social challenge, which was made impossible by the relative
breakdown of society.
The defeat of the Taliban was not therefore the result of a suicidal
strategy, founded on the idea of sacrifice but, as in the case of Iraq in
soldiers in the torture and execution of the prisoners, and also Jamie Doran,
Afghanistans secret graves. A drive to death in the desert, Le Monde Diplomatique, September 2002. See also the report of Physicians for Human Rights,
www.phrusa.org/research/afghanistan/report_graves.html. The United Nations halted
the inquiry even though Dostum had killed or imprisoned a number of witnesses.

328

A Splendid Little War?

1991, stemmed from an insufficient understanding of the balance of

power. The Taliban was probably also hampered by its high level of
political centralisation,32 while mullah Omar overestimated the level
of his popular support, and the degree of resistance his men would
show to the bombing.

32

In particular, there was no split between the hardliners and the moderates,
which demonstrated the cohesion of the networks of ulema which provided the
structure of the movement.

11. The Return to Political

Fragmentation
The accords of 5 December 2001, signed at Schloss Petersberg in the
German city of Bonn, paved the way for the establishment of the
Afghanistan Interim Authority. The composition of this body was a
triumph for the Northern Alliance leadership, Shura-yi Nazar, which
obtained three key ministriesDefence, the Interior and Foreign
Affairsas well as control of the secret services. Meanwhile the presidency went to Hamid Karzai, a Pushtun, whose appointment was
intended by the negotiators to endow the Interim Authority with
legitimacy in Pushtun eyes. However, the remainder of the groups
making up the Northern Alliance, other than Shura-yi Nazar itself,
were united in condemning what they saw as the Interim Authoritys unbalanced membership. In particular haji Abdul Qadir walked
out of the conference in protest at the inadequate representation
accorded to the Pushtuns. A further issue was that Dostum, Ismail
Khan and Sayyaf roundly condemned the monopoly over power
enjoyed by the Panjshiri membership of Shura-yi Nazar.
The issue of an international buffer force was bitterly contested.
Shura-yi Nazar initially opposed the presence of any foreign force in
Kabul, and objected a fortiori to the demilitarisation of the city. A
compromise was reached in due course providing for an international force of limited size, and Shura-yi Nazar retained its troops in
the capital. The negotiators accepted this formula because of Shura-yi
Nazars entrenchment within the city and because of the potential
threat it represented to the future government. For these reasons an
agreement seemed preferable to the risk of immediate destabilisation.
From the institutional standpoint the Bonn accords provided for
the convening of a Loya Jirga within six months of the Interim Authority being established on 22 December. In the first stage of this
process some 400 district assemblies selected representatives, and in
329

330

The Return to Political Fragmentation

the second stage the chosen representatives gathered in provincial

centres, under the scrutiny of foreign observers, for a secret ballot to
elect the actual delegates to the loya jirga. To the 1,501 delegates
elected by this process 450 more were added to represent women,
nomads, refugees and civil society, a term that included the religious and intellectual establishment. Some observers criticised the
method by which the delegates were appointed, which was open to
undue influence from the commanders.1
The role of the loya jirga, inaugurated by the former king Zahir
Shah on 11 June 2002, was the transfer of the interim administrations authority to the Afghanistan Transitional Authority. There was
inevitably some confusion, but the positions of the various delegates
gravitated towards the exclusion of former commanders and to
support for Zahir Shah. However, the crucial decisions, and in particular the choice of Hamid Karzai, had already been taken by the
Americans, at whose behest Zahir Shah was obliged to step aside.
Actually, a majority of the delegates appeared to be prepared to cast
their votes for Zahir Shah, a development which would have blocked
the election of the Americans candidate. For his part the king let it
be known that he was ready to assume any responsibility which the
Loya Jirga might wish to confer upon him, but in spite of this, shortly
after the Loya Jirga opened, the US envoy Zalmay Khalilzad publicly
denied that Zahir Shah intended to put himself forward, and confirmed that he would give his support to Hamid Karzai. Some hours
later Zahir Shah fell into line with the US position at a press conference, where the only diplomatic observers present were Americans,
and irrevocably renounced anything other than a ceremonial role.
In the two years which followed the formation of the Transitional
Authority a new constitution was meant to be adopted, a national
assembly elected, and a government appointed. However, it would
be possible to hold elections only if peace was restored to the coun1

Despite the best efforts of the Special Independent Commission for the Convening of the Emergency Loya Jirga (Special Commission) and the United
Nations, warlords have infiltrated and manipulated the process for selecting the
meetings delegates, and will attend the meeting in large numbers or act through
proxies. Human Rights Watch Briefing Paper, Afghanistan: Return of the Warlords,
June 2002, p. 2. During the deliberations the secret services attempted to influence some of the delegates.

The Return to Political Fragmentation

331

try, and if the state institutions were rebuiltgoals which are still far
from being achieved.

The reconstruction of the state

On the face of things, the US intervention had the effect of restoring
the political situation of the later 1990s in the sense that the Taliban
became once more the dominant political element in the Pushtun
regions. The Kabul government has no power outside the capital,
and little effective progress is being made towards the construction
of a central state. However, the presence of the western forces has
made a major difference. Control of the capital is no longer the objective of armed groups, while the survival of the government is
guaranteed by the foreign support it receives. The current situation
is marked by two opposing tendencies: on the one hand the impetus
towards the reconstruction of the state, together with attempts to
bring the regional forces under control, and on the other the resurgence of a Taliban guerrilla force, now well entrenched in the Pushtun areas.
The distribution of political power in the provinces. The immediate result
of the US invasion was the demolition of the still tenuous political
base established by the Taliban. In the south and east the results of
US action have tended towards fragmentation. As events have unfolded, the United States has not been willing to commit a significant number of troops and as a substitute has opted to finance local
militias, a policy inimical to any move towards centralisation. Mounting insecurity and the re-emergence of opium production have been
the two immediate consequences of the erosion of central authority.
Unofficial checkpoints have re-appeared on some roads. The level
of insecurity remains high even in Kabul, where the international
forces are stationed.
At the time of writing no open challenge is mounted against the
Kabul government other than by the Taliban (an issue still to be considered), and there are no armed clashes in the capital itself. However,
the regional authorities are able to protect their autonomy against
encroachments, whether by the central authority or by rivals. One of
the first announcements by the shura of Jalalabad was its rejection of
any Northern Alliance presence within the city; similarly, at the mo-

332

The Return to Political Fragmentation

ment when the Taliban fell, the Pushtun tribes in Kandahar rejected
external interference, which heralded a brief outbreak of conflict
against Ismail Khan in the Farah province. There are today five significant regional authorities, together with dozens of smaller groups.
In the north the fall of the Taliban opened the way for the Northern Alliance parties to reoccupy the positions they held in the 1990s.
The regional authorities are based in the principal towns. Thus Jombesh is based in Shibergan; Jamiat-i Islami, locally led by Muhammad
Atta, is established in Mazar-i Sharif; Ismail Khan holds Herat; and
Kabul serves as the base of Shura-yi Nazar. Finally Hezb-i Wahdat,
lacking control over a town, has established itself in the village of
Bamyan, although it also maintains some armed forces inside Mazar-i
Sharif. The degree of regional coherence in the north should not be
exaggerated: the seizure of Mazar-i Sharif by a number of parties
Jombesh, Jamiat, and Hezb-i Wahdatprovides a clue to the instability of the region, in which fighting is endemic. In Hazarajat tension runs high between different tendencies within Hezb-i Wahdat.
The Panjshir valley itself is divided between various rival commanders, as humanitarian aid workers based there are aware.
In the east and south the political balance is more complex. The
towns are politically less homogeneous and do not exercise control
over the countryside. In practice the balance of power often depends
on relationships between tribes. For example, three factions vie for
control of Kandahar. Gul Agha Shirzai,2 has recovered his position
as governor but is challenged both by mullah Naqibullah, a former
Jamiat-i Islami commander who has some support from Shura-yi
Nazar, and by Ahmad Wali Karzai, the Presidents brother. These
factional struggles are reinforced by the tribal oppositions between
the Barakzai (the tribe of Gul Agha Shirzai), the Alikozai (that of
mullah Naqibullah), and the Popolzai (that of Hamid Karzai). In the
east the degree of fragmentation is still greater. For example, the
shura of Jalalabad is divided into a number of factions and exercises
no authority further than a few kilometres from the town.
A divided government. The fragmentation of political factions might
almost have been of benefit to a central government seeking to extend
2

Gul Agha Shirzai was formerly governor of Kandahar, while his father haji Abdul
Latif was a powerful commander.

The reconstruction of the state

333

its influence, but for the fact that the government was itself split between the presidency and Shura-yi Nazar. The government,3 whose
membership was announced on 24 June 2002, was dominated by
Shura-yi Nazar, which retained the portfolios of Defence and Foreign Affairs and the headship of the secret police. Muhammad
Qasim Fahim, who had been both vice-president and Minister of
Defence in the Interim Administration, retained those portfolios.
After long negotiations one of Karzais associates, Taj Muhammad,
took the post of Minister of the Interior after his predecessor Yunus
Qanuni had reluctantly agreed to accept the education portfolio
combined with a position as special adviser on national security. The
Ministry of Finance, headed by Ashraf Ghani, was the other significant ministry which eluded Shura-yi Nazar.
Political conflict within the government broke out during 2002 as
various attempts were made to subvert it. It was not always clear who
was responsible for them, but they served to underline the absence
of trust and collaboration between the presidency and Shura-yi
Nazar. In February 2002 the assassination of the Minister for Civil
Aviation, Abdul Rahman,4 was the first event to spark off a crisis.
3

President, Hamid Karzai; vice-presidents, Muhammad Fahim, Karim Khalili,

Abdul Qadir; Minister of Education and special adviser for security, Yunus Qanuni;
Minister of Defence, Muhammad Fahim; Minister for Foreign Affairs, Abdullah
Abdullah; Minister of Finance, Ashraf Ghani; Minister of the Interior, Taj Muhammad Wardak; Minister for Planning, Muhammad Mohaqiq; Minister of Communications, Masum Stanakzai; Minister for Frontiers, Arif Nurzai; Minister for
Refugees, Inayatullah Nazeri; Minister of Mines, Juma M. Muhammadi; Minister
for Light Industry, Muhammad Alem Razm; Minister of Public Health, Dr Sohaila
Siddiqi; Minister of Commerce, Sayyed Mustafa Kasemi; Minister of Agriculture,
Sayyed Hussain Anwari; Minister of Justice, Abbas Karimi; Minister for Information and Culture, Sad Makhdum Rahim; Minister for Reconstruction, Muhammad Fahim Farhang; Minister of Mosques and the Hajj, Muhammad Amin
Naziryar; Minister for Urban Affairs, Yusuf Pashtun; Minister for Public Works,
Abdul Qadir; Minister for Social Affairs, Nour Muhammad Karkin; Minister for
Water and Electricity, Ahmad Shakar Karkar; Minister for Irrigation and the Environment, Ahmad Yusuf Nuristani; Minister for Martyrs and the Handicapped,
Abdullah Wardak; Minister for Higher Education, Sharif Faez; Minister for Civil
Aviation and Tourism, Mirwais Saddiq; Minister of Transport, Sayyed Muhammad Ali Jawad; Minister for Rural Development, Hanif Asmar; president of the
Supreme Court, shaikh haji Shinwari.
4 Abdul Rahman was from Nuristan and had occupied the same positions under
the presidency of Rabbani (19926), but then left Shura-yi Nazar in order to
support Karzai.

334

The Return to Political Fragmentation

Abdul Rahman lost his life in an incident at Kabul airport, possibly

at the hands of pilgrims enraged by having to wait to set off on the
hajj. Hamid Karzai accused Shura-yi Nazar directly of responsibility
for the killing. A number of officials were arrested, but later released.
Tensions within the government were also revealed by a further incident. In August 2002 Shura-yi Nazar supporters demonstrated against
the removal of a Shura-yi Nazar member, Abdul Hafiz Mansur, as
head of the state radio and television organisation, to be replaced by
Makhdum, a long-time exile in the United States and an associate of
Karzai. Mansur had banned television broadcasts of singing or dancing by women, and his replacement was an indication of Karzais
determination to appoint ministers who were ideologically more
liberal and closer to him politically. Finally the assassination in Kabul
on 6 July 2002 of haji Abdul Qadir, Vice-President and Minister for
Public Works, has never been satisfactorily explained.
The centralisation process and peripheral autonomy. The history of Afghanistan in the twentieth century has demonstrated that it is possible for some part of the resources necessary for the formation of a
state to be derived from external aid. In 2002 the situation was such
that it was once more in the interest of external powers to establish a
central authority in Afghanistan. In January 2002, as UN sanctions
were about to be lifted, an international aid conference was held in
Tokyo where $4.4 billion in donations and $1.4 billion in loans were
promised; however, these figures were not high per capita compared
to other conflicts, such as that in Bosnia.5 A further issue was that
only a modest proportion of these funds (16%) was given directly to
the Afghan government, with a priority set on emergency aid,
which left only 29% to be used for reconstruction. Finally, international aid was subject to the vagaries of global politics, and as the
Iraqi situation evolved, US priorities increasingly lay elsewhere. The
US budget for 2004 provided only a minimal appropriation for
Afghanistan by comparison with that for Iraq. For the year 2004
Congress was asked to approve $1.2 billion of civilian aid, of which
5

The sum envisaged for Afghanistan was $42 per annum per inhabitant for the
period 20026 as against $326 for Bosnia in the period 19969. See Analysis of
Aid Flow to Afghanistan, www.af.

The reconstruction of the state

335

$400 million were funds remaining from the 2003 appropriation.

By comparison $11 billion were earmarked for military operations
in Afghanistan.
The part played by international aid was all the more significant
because of the Kabul governments inability to mobilise resources of
its own. The economic state of the country scarcely permitted taxes
to be levied on agriculture, while the most profitable sectors were already either taxed by local authorities or, as in the case of opium and
smuggling, illegal. Customs duties were a major source of revenue
for powerful regional organisations such as those of Dostum and
Ismail Khanthis issue had been a cause of contention between the
commanders and Kabul since 1992. Arguing that customs duties had
in the past been made over by Kabul to the frontier provinces, the
governors refused to pass them to the treasury, although the surpluses were large, at least in the case of Herat, because of the redirection of the transit routes to Iran. In the north customs duties were
simply shared out among the leading commanders, Muhammad Atta
and Dostum, together with Muhaqqiq, who represented Hezb-i
Wahdat. Lastly Afghanistan had once more become the worlds leading opium producer,a development from which some groups,including in particular the shura of Jalalabad, Jombesh and the commanders
of the north-east, were able to derive profit. In some areas opium
again became the principal crop, especially in Badakhshan, where
some peasants seem to have abandoned their traditional cultivation
of wheat.
External funding of development is a phenomenon reminiscent
of the 1960s and 1970s, although the political context has become
more problematic. The international organisations and the NGOs
continue to play important roles in public administration, with all the
negative results for the government that such a situation may involve.
The impact of humanitarian aid on the process of strengthening the
central government is also mixed because of the tendency of the
NGOs to replicate the functions of the state, sometimes with the
effect of undermining them, especially when they by-pass the government to conclude agreements directly with local authorities. A
further issue is that the NGOs attract the best-educated staff away
from the government because of the higher salaries they are able to
offer. Finally, the high level of corruption in the government remains

336

The Return to Political Fragmentation

such that it discourages investment by Afghans abroad and militates

against the logical use of resources.
Despite such obstacles, the reconstruction of the state apparatus
has hitherto made some headway through the incorporation of local
power-brokers who emerged when the Taliban collapsed. These recognised the Kabul government which was in turn obliged to do
business with them. Karzai gave official positions to most of the major
commanders, thus effectively legalising their power. In some instances
such commanders, or their proxies, were given ministerial posts, haji
Qadir became Vice-President and Minister for Public Works, while
Karim Khalili, the leader of Hezb-i Wahdat, was also made VicePresident. Mirwais Siddiq, one of Ismail Khans sons, was appointed
Minister of Civil Aviation, and Ahmad Shaker Kargar, an associate of
Dostum, was made Minister for Social Affairs.
In further instances of the same phenomenon, Ismail Khan was
appointed governor of Herat, Gul Agha governor of Kandahar and
Din Muhammad governor of Jalalabad. Centralisation of administrative and military authority had the effect of enabling commanders to
extend their control over regions comprising several provinces, as
was shown by Dostums appointment of governors in Jozjan and
Faryab after the fall of the Taliban. Similarly, the province of Ghor
was in practice run from Herat. In general, governors appointed heads
of districts (uluswal) without reference to Kabul. However, armed
groups were brought nominally under the authority of the Ministry
of Defence. Amalgamation of units mirrored the pattern of distribution of the militias. For example, the 7th and 8th corps in Mazar-i
Sharif and Shiberghan could not be merged, as had previously been
customary, because of the rivalry between Rashid Dostum and Muhammad Atta, each of whom was appointed a general.
In 2003, in reaction to the increasing strength of the regional
power brokers in 2002, the government strove in various ways to assert its authority. In February 2003 Ali Ahmad Jalali, the new Minister of the Interior, moved to appoint two new governors in Jozjan
and Faryab as soon as he had taken office, but the resulting outbreak
of violence revealed the fragility of the local political equilibrium.
Events in Khost also underlined the complexity of local factors and
illustrated the difficulties facing the central authority as it attempted
to identify locally acceptable candidates for office. At the time of the

The reconstruction of the state

337

Bonn conference Padshah Khan, a leader of the Jadran tribe who

had supported Karzai, had been appointed governor of Paktia, at
Gardez. His brother Amanullah Khan was given the Ministry for
Frontiers, while his other brother Kamal was appointed governor of
the province of Khost. However, in spite of attempting to impose
himself by force in the spring of 2002, Padshah Khan was unable to
make good his position as governor of Gardez because of opposition
from a local rival, and at the same time his own relations with government deteriorated. In June 2002, during the Loya Jirga, Padshah
Khan actually supported the king, thus opposing Karzai and Shurayi Nazar.6 In the end an alternative governor was appointed.
The outward symbol of the reconstruction of the state was the rebuilding of national armed forces. However, fusion of the regional
militias was difficult to achieve, owing to the absence of any mechanism which would guarantee the neutrality of a future army. The
Ministry of Defence was viewed as the fief of the Panjshiris,who almost
entirely made up the garrison in Kabul and for that reason were not
able to present themselves as the credible nucleus of a national army.
The appointment in September 2003 of officers who did not belong
to Shura-yi Nazar brought no immediate or decisive results. A salient
feature of the agreement between the Afghan Ministry of Defence
and the UN Assistance Mission in Afghanistan was its provision for
the demobilisation of the militias, beginning in October 2003, in the
context of the Afghanistan New Beginnings Program announced
by Karzai in April 2003. The United States earmarked $60 million
to be used for this purpose during 2004. The mandate of ISAF (International Security Assistance Force) was extended to the provinces,
but there are still at present insufficient forces to curb commanders
who do not keep their promises by military means. Incidents continue to be frequent, in spite of the establishment in May 2002 of a
Security Commission for northern Afghanistan.
Overall, little progress has been made up to the present in the formation of a national army. Ten battalions, comprising around 6,000
6

Padshah Khans collaboration with the United States came to an end following
the supply of false intelligence which led to the bombing of a mosque in December 2001 and then to an attack on a convoy of officials on their way to support
Karzai in Kabul.

338

The Return to Political Fragmentation

soldiers, had been formed by the spring of 2003, but the soldiers
receive only around $17 a month, and desertions run at 3050%,
especially when men are posted outside their home provinces.7
Absenteeism and even criminality are common in most units.
The drive to reconstruct the state has therefore clashed with the
efforts of the regional powers to remain autonomous. The real obstacle, however, has been the re-emergence of the Taliban movement, which has obstructed all efforts at pacification.

The new guerrilla movement

Had the Taliban been an organic outgrowth of the Afghan society,
they would have survived much longer, and a campaign of harassment of the new power holders would have begun immediately.8 By
the winter of 20023 there was no longer any question that the Taliban had re-formed, an issue over which in the summer of 2002 there
remained some doubt. In Kandahar, in March 2003, the United
States deployed more than 1,000 men with strong air support in a
single operation against a Taliban guerrilla force which was already
beginning to carry out daily operations against the US forces and
their Afghan allies.
After some months of reorganisation, the Taliban had resumed its
leading role in the Pushtun provinces. As far as can be ascertained,
there had been no significant change of personnel among the movements activists, except for the replacement of those lost during the
war. At the head of the movement there seems to have been no challenge to mullah Omars leadership. Most of his former charisma had
probably dissipated by the time of the Taliban regimes military collapse, but he remains the leader of a disciplined group. The role of his
associates also continues to be significant. These include mullah
Baradar, from the same district as mullah Omar; Akhtar Muhammad
Osmani, the former military governor of Kandahar; and mullah
Abdul Razak, the former Minister of the Interior. Meanwhile the
prestige enjoyed by Jalaluddin Haqqani continues to boost the guerrillas in the east of the country.
7

International Crisis Group, Disarmament and Reintegration in Afghanistan, Asia

Report 65, September 2003, p. 5.
8 William Maley, The Afghan Wars, London: Palgrave, 2002, p. 267.

The new guerrilla movement

339

The alliance with Hekmatyar has entailed further radicalisation

among the Taliban and more stress on ideology.9 In particular its cooperation with Hekmatyar after his return to Afghanistan, when the
CIA made its attempt to assassinate him, won the Taliban supporters
in the towns, especially Kabul. Training camps for foreign recruits
once more appeared on the frontier, in the region of Asadabad,
taking the form of small mobile teams of some twenty individuals.
The Taliban seems to have changed its strategy after a period of
some months during which guerrilla actions were limited to ineffectual mortar attacks against US positions. Taliban military operations
became more effective, especially its ambushes, although by mid2003 western losses since the beginning of the war had not exceeded
fifty men. Another factor was the adoption by the guerrillas of new
tactics, hitherto unseen in Afghanistan. In June 2003 four German
soldiers were killed near Kabul in a suicide attack, probably organised by Al-Qaida, or perhaps by Hekmatyars network. However,
the Taliban has generally avoided direct confrontation with western
forces and is digging in for a long-term struggle, for which it possesses the necessary economic resources and manpower. The availability to it of a refuge inside Pakistan, an issue discussed in further
detail below, has helped them replenish their arms supplies and elude
their pursuers.
The aim of the Taliban is to prevent the reconstruction of the
central state. Afghan government troops have therefore become a
priority target, along with the UN and the NGOs. The Taliban now
maintains guerrilla forces several hundred strong in a number of
locations in the south and has the ability to seize government positions of moderate size. Following the hit-and-run operations of
2002 the new trend is to base groups of fighters inside Afghanistan.
In Ghazni the Taliban is able to move freely within the town, though
without arms, and has killed a UN employee in a daylight attack.
Only US air power has so far hindered the Taliban from starting the
process of reconquest with the capture of small towns.
There is uncertainty over the ability of the Taliban to extend beyond the Pushtun regions, despite its efforts to give the conflict a reli9

The assassination in cold blood in April 2003 of an ICRC representative illustrates this issue. The likelihood of such incidents occurring was heightened because of the increasing integration of humanitarian aid workers into the western
strategic deployment.

340

The Return to Political Fragmentation

gious dimension, Afghanistan is today the scene of a veritable propaganda battle.10 Taliban tracts denouncing the occupiers and their
allies, couched in primarily religious and nationalistic terms, are distributed, even in Kabul. In one such pamphlet American soldiers are
pictured body-searching a young woman.
The Taliban is now an acknowledged force in the south, to the
extent that in the province of Zabul, where it enjoys overwhelming
support from the population, the government has negotiated a ceasefire. The new governor of the province, Hafizullah Tukhi, initiated
talks in the summer of 2003. It seems likely that Hamid Karzai intends to put pressure on Shura-yi Nazar by leaving open the door
for the rehabilitation of the Taliban. Alternatively these events may
also be interpreted in military terms, as a consequence of the governments inability to overcome the guerrillas, who have forces of
several hundred fighters in some districts, for example in Dayshopan,
while the western military presence remains insufficiently strong.
The western forces number around 13,000, of whom 8,000 are
American, and 5,000 men of the International Security Assistance
Force (ISAF) stationed in Kabul (March 2004). In addition, logistical
support for the US troops involves more than 50,000 troops in the
surrounding region, particularly in Central Asia. The cost to the
United States of these operations is around $1 billion each month.
Because of the restricted size of its forces, it relies on air power and
embarks on few large-scale ground operations, whose risky nature
was illustrated by the fate of Operation Anaconda, launched near
Gardez on 1 March, 2002, which was intended to dislodge Taliban
and Al-Qaida fighters from a mountainous region with numerous
caves. The Americans lost two Chinook helicopters as the result of
poor planning and the under-estimation of their adversary; fifty soldiers were wounded, and if they had not worn body-armour the
number would have been much higher. The guerrillas appear to
have withdrawn into Pakistan territory with no significant losses.
From the spring of 2003 the increase in infiltration by the Taliban
10

Nor are US efforts on an insignificant scale. In April 2002 US aircraft dropped

10,000 Afghani banknotes in the Spin Boldak and Shaman areas. Pamphlets
were attached or dropped separately showing a woman begin beaten by the
Taliban. The effect of other leaflets depicting mullah Omar as a dog was probably
ambiguous when distributed by non-Muslim forces.

The new guerrilla movement

341

prompted the Americans to step up the number of hunt-and-destroy

missions, in which they provided air support for Afghan troops.
However, the Taliban avoided contact with the Americans where
possible and could always disappear into the population if the pressure became too strong. In this sense the initiative remains with the
guerrillas, who have learned the importance of mobility in the face
of foreign forces whose intelligence-gathering is still, after two years
in the country, surprisingly poor.
The US forces are unwelcome, especially in the Pushtun areas,
where the civilians have complained of harrassment. Regularly and
predictably, military operations result in civilian casualties, which are
all the more resented since the United States neither acknowledges
them nor pays compensation. For instance, forty-two Afghans died
and 118 were wounded on the night of 30 June1 July 2002 when
four villages near Kakrakai in the province of Uruzgan were bombed
during a marriage ceremony. In early July the military authorities
acknowledged that B-52s and AC-130s took part in the attack and
claimed that the planes retaliated after being targeted. An internal report by the UN contradicted this allegation, stressing the subsequent
appearance of US troops at the scene to remove traces of the bombing. The treatment of prisoners of war also does not measure up to
international standards. In a communiqu on 28 January, 2003 the
World Organisation Against Torture stated that Taliban detained by
the Americans had been subjected to torture in CIA interrogation
centres, particularly at Bagram air base in Afghanistan and on the
island of Diego Garcia.11 A further point is the continuing uncertainty over the status of more than 600 prisoners detained at Guantanamo Bay in contravention of international law, pending a decision
by the US Supreme Court.
In early 2002 both the UN and the Karzai government appealed
for further US forces to be sent to Afghanistan. According to F. Vendrell, deputy to the UN envoy Lakhdar Brahimi, 30,000 troops would
be necessary to guarantee security in the major towns alone. For
11

The WOAT report was based principally on an article in the Washington Post for
26 December 2002, U.S. decries abuse but defends interrogations. Prisoners in
these detention centres are said to have been subjected to physical and psychological torture. Three detainees are said to have died at Bagram airbase following
ill-treatment.

342

The Return to Political Fragmentation

months this request went unanswered, owing to a shortage of available troops and to a lack of commitment on the part of the United
States.
From the autumn of 2002, however, the US government became
increasingly conscious of the resurgence of the Taliban and of the
generally deteriorating situation. There were two indications that its
position had changed. First, it backed ISAFs transfer to NATO control in the autumn of 2003, and then endorsed the extension of its
mandate to cover areas outside Kabul, giving it the capacity to intervene in any of the major towns. Second, at the end of September
2003 the United States appointed Zalmay Khalilzad as its ambassador in Kabul, replacing Robert Finn. With Khalilzads appoint came
an increased willingness to give higher priority to the reconstruction
of the state. In particular the US administration wished to put advisers into all the ministries, in the same way as its advisers already led
Afghan military units. In addition the United States put pressure on
Shura-yi Nazar to relinquish its control over the Ministry of Defence,
to pave the way for the re-formation of a national army. Finally, the
objective of waging an efficient counter-insurgency campaign led to
the formation of Provincial Reconstruction Teams (PRA). These
units of fifty to 100 men, situated in a dozen key provinces including
in particularly Gardez, Bamyan and Kunduz, were given the task of
coordinating military operations with the humanitarian effort. This
plan was completely rejected by the NGOs working in Afghanistan,
which preferred to be dissociated from the military, fearing for their
security if they were not.

Ethnicity, social groups and ideologies

Two significant changes have taken place in the social balance of
power in Afghanistan. First, there has been a shift of power away from
a group of religious figures and towards new and more diverse lites.
Second, in parallel, the religious hierarchy has been overthrown.
The consequences of the lack of an accepted ethnic hierarchy
from the outset of the war have already been analysed. The fall of the
Taliban led to a new imbalance in representation in the national
institutions. The Uzbeks are able up to a point to regard themselves
as represented by Dostum, while the Hazaras are represented by
Hezb-i Wahdat and the Persian-speakers of the northeast by Jamiyat-i

Ethnicity, social groups and ideologies

343

Islami.12 However, the Pushtuns, the most numerous ethnic community, have unquestionably been marginalised by the new regime.
The ethnic cleansing to which they have been subjected in the north
has been tolerated or even connived at by the two principal backers
of the government, Jamiyat-i Islami and Jombesh, while Hamid
Karzai has had neither the will nor the ability to interfere.13 The provinces most concerned have been Faryab, Jozjan, Badghis, Kunduz,
Baghlan and Takhar. In February 2002 some 20,000 people fled
from northern Afghanistan because of ethnic persecution. At Dast-i
Arshi, where the Pushtuns fell victim to aggression by the Uzbeks,
they fled en masse. There are today several tens of thousands of Pushtun refugees in the south of the country. A further issue is that there
is an increasingly direct correspondence between ethnicity and political affiliation, particularly in the north. After the fall of the Taliban the Uzbek commanders joined up with Jombesh so as to be able
to resist the pressure of Shura-yi Nazar, a fact which explains Jombeshs recent appearance in the province of Takhar, especially at
Taloqan, Masuds former base.
The balance between social groups has also been upset. The ulema
no longer occupy a significant position in national politics, as was
demonstrated by their virtual absence from the interim government
formed in December 2001. On the other hand, their exclusion from
politics has enabled them to resume their earlier position as critics of
the government. Still, the elections planned for 2005 could bring a
return to power by the ulema in some areas. However, in this context
Hezb-i Wahdat is a regional exception within Hazarajat. The Shiite
ulema never left the political arena in Hazarajat, both because they
form part of the Northern Alliance and because there is no alternative to them. A further point is that, in the absence of trained personnel, the ulema continue to play a significant role in the fields of
primary education and justice. Finally, the disappearance of the Taliban government may permit the religious families to return, espe12

Pierre Centlivres and Micheline Centlivres-Demont remark on having heard

with surprise a mythical tale of Masud being of Pushtun descent, an indication
that he may be accepted by some Pushtuns as a national hero and martyr who is
above ethnicity (personal information, April 2003). An imaginary genealogy of
this kind could also be interpreted as a sign of the downgrading of the Pushtuns.
13 Afghanistan: Monthly Review, April 2003.

344

The Return to Political Fragmentation

cially the Gaylani, who are attempting to rebuild their networks of

clients, partly for political purposes.
As for the educated class, impoverishment or exile was their only
prospect under the Taliban, since the prospects for employment
offered by the NGOs were limited. It is evident today that this situation has been reversed. Political figures at the national level are mostly
university graduates or have at least been educated to secondary
level. Foreign aid has led to the development of bureaucracies, such
as those within the NGOs, the Afghan state and international organisations, which are able to offer employment. The reopening of the
girls schools and jobs in NGOs have opened up employment opportunities to workless Afghans. In addition, other opportunities for
modern kinds of activity have emerged in the major towns, especially Kabul itself.
However, the lites display only minimal coherence in their social
and ideological position, and distinctions based on social origins remain marked. The current President himself, Hamid Karzai, a typical
product of the traditional lites, is the son of a distinguished Pushtun
family from Kandahar, related to the royal family, which includes a
large number of high officials.14 This family enjoys the support of
the Popolzai of the Kandahar region, but it was his exile in the
United States which enabled Karzai to gain the backing of the US
government and therefore achieve his present position. Karzai tends
to rely on the support of men who have had similar careers. Thus
Ashraf Ghani, the Minister of Finance, is a former employee of the
World Bank. In contrast, the officials of Jamiyat-i Islami or Ettehad
derive their legitimacy from their participation in the jihad. Various
factions have emerged since the fall of the Taliban, mostly from the
erstwhile parties in Peshawar.
The end of the Taliban did not automatically imply the end of the
fundamentalist tendency, which continues to flourish even within
the present government. Opposition to the west is not a monopoly
of the armed opposition, and there is strong resistance to the liberalising measures initiated by the UN and the NGOs. A campaign of
intimidation has been waged against the Minister for Womens Affairs,
Dr Sima Samar, who was accused of blasphemy in June 2002, and
14

For biographies of members of the Karzai family see Ludwig Adamec, Whos
Who in Afghanistain, Graz: Akademische Druck- und Verlagsanstalt, 1987.

Ethnicity, social groups and ideologies

345

later acquitted. The head of the Afghan Supreme Court, mawlawi

Shinwari, follows an interpretation of the Quran close to that of the
Taliban. In September 2003 the council of ulema banned women
from working in the NGOs. This had previously been authorised
by mawlawi Shinwari although he declared that they should be separated from men in the workplace. For example, he ruled that women
should not be permitted to make journeys of longer than three hours,
or travel abroad, unless accompanied by a male member of their family.
Finally, marked regional distinctions, which had been to some extent suppressed by the Taliban, have re-appeared and seem set to
persist even after the hypothetical construction of the state. The
provinces stand in opposition to Kabul, where the insulating effect
resulting from the presence of the humanitarian organisations has
led to an accelerated process of westernisation among a section of
the younger people. However, even in the capital the burqa generally
continues to be worn, except in only a few areas. At the same time
everyday customs have been little changed in the Pushtun south,
where few girls go to school and a return to customary law is in evidence. Hazarajat is dominated by Hezb-i Wahdat, whose officials are
all men of religion, and whose ideology is clerical and fundamentalist. Herat, under the authority of Ismail Khan, is in an intermediate
position. Strict rules are maintained and women are obliged to wear
the burqa, which they have always done, even during the Soviet occupation, but on the other hand girls are permitted to go to school.

12. The Policy of Pakistan

Only a decade after the fall of the Soviet Union, the United States
set up military bases in Central Asia, thus radically altering the parameters of regional security. Russia has been dislodged from its position of regional supremacy, and the United States is now able to exert
direct influence over the internal political balance of the Central
Asian states. In Uzbekistan the presence of US troops is primarily of
internal significance, allowing all opposition to be suppressed in the
name of the war against terrorism. Elsewhere the United States has
denounced Iran as part of the Axis of Evil, openly encouraging the
overthrow of the current regime. A consequence of the invasion of
Iraq was that both of Irans immediate neighbours were occupied by
US troops. This situation explains the extreme caution of the Iranian
government, which refrains from significant interference in Afghanistan. In any case Iran no longer has a local proxy, since Hezb-i Wahdat
has opted to cooperate with the United States.
Ultimately the position of Pakistan will be decisive for the future
of Afghanistan. The United States needed Pakistans help, primarily
to seal the frontier during its military operations, but also for logistical reasons.1 Although Pakistan is a nuclear power, it has been unable
to resist the demands of the United States because of its parlous financial position. The Americans have neatly saved Pakistans face by presenting its acquiescence as a negotiated agreement. Pakistans change
of direction was disclosed by President Musharraf in a major speech
on 19 September 2001,2 in which he justified Pakistans alignment
with the United States and its abandonment of the Taliban on the
grounds of national interest and of the necessity of preventing the
installation of an anti-Pakistan government in Kabul. He backed up
1

These were the use by the US forces of the bases at Dalbandin (in Baluchistan),
Jalalabad and Pasni.
2 The full text in English is given in The News, 20 September 2001.

346

Trial of strength with the pro-Taliban movements

347

his position with a Quranic text which endorses the principle of tactical alliance with non-Muslims.
At first glance Pakistan had succeeded in a dmarche similar to that
which it had employed at the time of the Soviet invasion; Musharraf s
government used an international crisis to re-establish its credibility.
Following Pakistans humiliation at Kargil in 1998, the military coup
and the international economic sanctions triggered by the nuclear
tests, relations between Washington and Islamabad were at their lowest ebb. However, the diplomatic dtente did not signify a return to
the golden age of the 1980s. The objective of the United States was
to ward off the collapse of Musharraf s government by affording him
the scope to suppress radical movements within his territory. However, the level of US economic aid was limited, amounting to around
$6 billion,3 while Pakistan continued in reality to suffer from a degree
of diplomatic isolation. Above all, Pakistan could not count on US
support over the Kashmir issue.In the last resort the pro-Indian orientation of the United States was based on a real long-term commitment,
the result of a pro-Indian pressure group active in the US Congress,
at a moment when Pakistan had lost much of its support there.4
The abrupt realignment of Pakistani policy resulted in a confrontation with the religious movements, hitherto instrumentalised by
the state. There follows an analysis of pro-Taliban mobilisation, and
of developments in human and economic movements between the
two countries.

Trial of strength with the pro-Taliban movements

Pro-Taliban forces in Pakistan foresee an Iraq-like invasion of Afghanistan led by the United States, Russia and their allies. Neither Iran
nor Pakistan would be in a position to support Islamic warriors when
the doomsday arrives. Perhaps, realizing such an eventuality, the religious groups in Pakistan and their leadership are engaged in necessary homework to save their friends in Afghanistan. (Shaikh Akram,
A Storm is Brewing, Viewpoint, May 2001, p. 15)
3

In the same period Turkey and Argentina, each with a much smaller population,
had credit facilities with the IMF of over $30 billion and $22 billion respectively.
4 Arthur G. Rubinoff, Changing Perceptions of India in the US Congress, Asian
Affairs, 411, 2001.

348

The Policy of Pakistan

As this quotation demonstrates, the Pakistani states abandonment

of the Taliban was foreseen by some fundamentalist movements
which never had any illusions concerning the permanence of its support. Although public opinion in Pakistan was flatly opposed to American intervention, pro-Taliban mobilisation at the national level in fact
remained relatively low. The attacks of 11 September were viewed as
terrorism by 64% of Pakistanis, while 24% saw them as justifiable in
terms of the jihad. On the other hand, Bin Laden was well regarded,
with 82% seeing him as a mujahid, while to 6% he was a terrorist.
Only 3% of Pakistanis declared themselves pro-American as against
83% who took the side of the Taliban; 75% were opposed to the use
of Pakistani air bases by the United States. Nevertheless, Musharraf s
position enjoyed the sympathy of much of the population: 51% were
favourable to cooperation with the United States, no doubt hoping
not to be isolated in the confrontation with India, as against 41%
who actively disapproved.5
Popular support for the Taliban only rarely went as far as active
participation, and mobilisation in favour of it was geographically restricted and on a limited scale. This can be explained in a number of
ways. At the international level there was no pro-Taliban mobilisation comparable to that seen in favour of Iraq at the time of the Gulf
War, probably because Arab solidarity did not apply. At the same time,
poor coordination between the parties opposed to the American intervention, together with the lack of support from the two major
national parties, the PPP (Pakistan Peoples Party) and the PML (Pakistani Muslim League), deprived the pro-Taliban movement of active
supporters. Those parties which did actively protest, Jamaat-i Islami
and Jamiyat-i Ulema, were relatively unpopular. Also there was a
tough government clamp-down. After a purge of the army ridding
Musharraf of potential opposition, leading fundamentalist figures
were placed under house arrest and the police did not hesitate to
open fire on demonstrators, killing fifteen. The government took
prudent steps to forestall escalation, notably by outlawing the use of
mosque loudspeakers for political statements.
Support for the Taliban took different forms, including street
demonstrations, humanitarian aid and the dispatch of volunteers to
5

Figures taken from a Gallup poll, in www.sofres.com, 17 October 2001.

Trial of strength with the pro-Taliban movements

349

Afghanistan. Islamic humanitarian organisations attempted to supplant the western NGOs whose relations with the Taliban staggered
from one crisis to another. For instance, Islamic Relief gave $3 million to the Taliban government, while the Rasheed Trust made a
bid to replace the World Food Program, which had withdrawn from
Afghanistan in June 2001. In a related phenomenon the various fundamentalist and Islamist organisations regularly demonstrated after
Friday prayers and launched strike appeals in the bazaars, with varying degrees of success.
The most significant level of mobilisation was seen in the tribal
zones on the Afghanistan-Pakistan frontier, where the Pushtun population was committed to Bin Ladens cause and that of the Taliban,
due to both ethnic and religious solidarity. At the end of October
armed tribesmen mounted a five-day blockade of the Karakorum
highway linking Pakistan and China. A further development was the
organisation by fundamentalist movements of the transfer of volunteers to Afghanistan.6 In November it was estimated that there were
thousands of fighters waiting at the Bajaur Agency, where the inhabitants provided them with food. Afghan refugees fleeing from the
bombing were being welcomed in the same spirit as the muhajjirin of
the early 1980s at the time of the Soviet invasion.
The level of mobilisation might have appeared as a setback for the
pro-Taliban parties, but these groups, which had previously done
badly at the polls, made major advances in the elections of November 2002, especially in the provinces bordering Afghanistan. In the
NWFP the election was won by Muttahida Majlis-i Amal (MMA),
a union of six pro-Taliban parties including Jamiyat-i Ulema and
Jamaat-i Islami, which set the scene for a rampant Talibanisation of
the province.7 The emergence of provincial administrations which
overtly supported the Taliban contributed to establishing the frontier regions as a haven for the guerillas, who embarked on an undisguised propaganda campaign and were able to make preparations for
the military operations they intended to carry out inside Afghanistan.
Entire quarters of Quetta fell under de facto Taliban control.
6
7

The News, 2 November, 2001.

The expression is not too strong: the authorities clamped down on musicians,
male journalists were banned from covering womens sporting events, and in
broader terms there was a shariatisation of the law.

350

The Policy of Pakistan

Despite the authoritarianism of President Musharraf s government there was little it could do directly against movements which
remained within the law and enjoyed parliamentary representation.
The most radical fringe groups had always been the particular object
of repression, a tendency which increased in the wake of the attack
against the Indian parliament on 12 December 2001. In January 2002
a wide-ranging operation against jihadist movements resulted in
thousands of arrests, while hundreds of the offices of such movements were closed down across Pakistan. Several armed groups were
proscribed, including Lashkar-i Taiba and Jaish-i Muhammad, those
accused by India of responsibility for the attack on its parliament,
and two Pakistani movements implicated in sectarian conflict, namely
the Sunni movement Sipah-i Sihaba Pakistan (SSP) and the Shiite
Tehreek-i Jafria Pakistan, were also banned. A pro-Taliban organisation which had dispatched thousands of volunteers to Afghanistan,
Tehreek-i Nifaz-i Shariat Muhammado (TNSM), was dissolved.
After the American intervention, the ensuing political crisis
prompted President Musharraf to re-emphasise two policies he had
mooted at the time of his seizure of power. These were the extension
of state control over the religious institutions, and the re-imposition
of government authority in the frontier regions. On 18 August 2001
the government promulgated the Pakistan Madrassah Education
Board Ordinance, which was in effect a first step towards the integration of the madrasas into the educational system, but which encountered stiff resistance from the ulema.8 However, little changed in
the period following 11 September, since the government did not
have the funds to finance profound changes in the educational system. In Punjab alone there were more than a quarter of a million
students in the madrasas,most from poor families.9 The trial of strength
between the government and the religious movements exacerbated
the existing polarisation between social groups, in which the modern urban class, which was antipathetic to the ulema, found itself in
opposition to the population at large, over whom the ulema still exercised substantial influence, especially in the frontier provinces.
8

Another factor was that the American administration earmarked 100 million
dollars to extend surveillance over madrasa students, which was an indication of a
US intention to take a direct hand in intelligence operations in Pakistan.
9 Herald, November 2001, p. 50.

Trial of strength with the pro-Taliban movements

351

In parallel, there was also a movement to weaken the autonomy of

the Tribal Areas, although it was only with great difficulty that control could be maintained over heavily-armed tribesmen who depended for their economic survival on contraband. After the fall of
the Taliban the Pakistan army was able for the first time to enter the
FATA (Federally Administered Tribal Areas),10 and specifically occupied the Mohmand Agency. However, the army undertook not to
interfere in the contraband trade, and the sale of firearms was allowed
to continue. In May 2002 the arrival of a small US force at Miram
Shah was regarded as a provocation,especially when the troops searched
a madrasa, while joint Afghan-Pakistani operations met with little
success. Islamabads lack of control was clear, even though it seemed
beyond doubt that Bin Laden had taken refuge in the FATA.
A final factor was that even within Pakistan government circles
policy towards Afghanistan continued to be a subject for debate. At
the behest of the Northern Alliance, backed by Russia and India,
Pakistan had been entirely excluded from negotiations on the future
of Afghanistan. The hope of an Afghan-Pakistani alliance against
India faded, as Pakistan saw a pro-Indian and anti-Pakistan government re-established in Kabul. Meanwhile diplomatic relations between Pakistan and Afghanistan were less than warm, and the lack of
an agreed frontier line was an additional source of friction. At this
point some elements in Pakistan toyed with the idea of destabilising
the Afghan government, to open the way for the return to power of
a pro-Pakistan faction, though for the time being the presence of
American troops precluded open Pakistani support for the armed
Afghan opposition. For its part the Afghan government wanted the
United States to put pressure on Pakistan to dismantle the guerrilla
groups, but the options for the United States were limited from the
moment when it became apparent that any political alternative to
the regime of General Musharraf in Pakistan would be to its disadvantage. The continuing arrests of Al-Qaida militants in Pakistan,
together with occasional military operations, allowed Pakistan for
10

The NWFP proper is distinct from the FATA (Federally Administered Tribal
Agencies) which are administered by a political agent appointed by Islamabad,
on the British model of indirect rule. There are seven Tribal Agencies (South
Waziristan, North Waziristan, Mohmand, Kurram, Orakzai, Bajaur and Khyber)
as well as five Frontier Regions in the interior of the NWFP.

352

The Policy of Pakistan

the time being to make some response to Washingtons demands

without prejudicing the possibility of it intervening in Afghanistan
in future.

Commercial and human movements

One of the results of twenty-five years of war has been the opening up
of Afghanistan, especially as a result of two socio-economic phenomena, namely the Afghan refugee communities and cross-border trade.
During the war cross-border trade, previously limited, developed
to a striking extent, especially in the 1990s when the Afghan state
ceased to exist as such, while the intensity of the fighting diminished.
The development of towns close to the frontier, such as Jalalabad and
Khost, can be explained by the suppression of the checkpoints on
Afghan roads, of which there were more than fifty between Kandahar and the Pakistan frontier. Smuggling was especially profitable,
since the thirty-year Afghan Transit Trade Agreement (ATTA) signed
in 1965 permitted Afghanistan, as a land-locked country, to import
goods in sealed containers through the port of Karachi. This mechanism allowed goods to be illegally re-exported to the Pakistani tribal
areas. Far from being locally restricted, trade between Pakistan and
Afghanistan therefore brought wider networks into play, especially
involving the Gulf countries, where merchant families were based.11
Pakistan, which was losing considerable sums in customs duties,
attempted on a number of occasions to renegotiate the ATTA. Since
1995, when the 1965 accords came to an end, it has striven to restrict
the flow of Afghan imports through Karachi, but these attempts have
resulted in the re-routing of trade channels, so that goods now come
to Herat by way of Iran before, as previously, entering Pakistan as
contraband. Customs duties from the Herat region have correspondingly increased, while those from Kandahar and Nangrahar have
shrunk. About three-quarters of Afghanistans commerce now passes
through Iran.
11

Afghanistans exports were diverse. Wood for the construction industry was a
constant export, contributing to Afghanistans deforestation. However, vehicles
and electronic items brought from Asia by way of Dubai made up the bulk of the
goods involved. For 1997, the World Bank estimated that the volume of smuggling, excluding the drugs trade, amounted to 2.5 billion dollars. See Ahmed
Rashid, Nation, 21 January 1998.

Commercial and human movements

353

The second significant element in Afghanistans relationship with

Pakistan has been the return of the refugees. By the end of the 1990s
more than 3 million Afghans were registered as refugees in Pakistan.
A large proportion of these were settled in NWFP and Baluchistan,
where their presence was initially accepted by Pakistan, although requests for their return to Afghanistan were heard increasingly as time
went on. However, many refugees spontaneously returned to Afghanistan.12 By the time of the Talibans fall only some 1.2 million remained out of an original 3 million. Their return was largely due to
the improvement of Afghanistans internal security. However, the
figures must be treated with caution because of the phenomenon of
simulated return undertaken to obtain UNHCR grants. It is evident today that hundreds of thousands of Afghan refugees, of whom
many now have Pakistani citizenship, do not intend to return permanently to Afghanistan.

12

Iran put very real pressure on the refugees to go back to Afghanistan, but several
hundred thousand remain in Iran.

Conclusion
The Afghan war is not over, and there is no indication that a rapid
pacification of the country can be expected. A generation of Afghans
including many well into adulthood have experienced nothing but
war. As a result of such traumas of the most profound kind the prewar society has largely ceased to exist.
A primary consequence of the war has been abrupt and rapid
politicisation. Rebellion undertaken in the name of the jihad fell
progressively under the control of increasingly sophisticated organisations. In different regions the new political lites were either ulema
or members of the educated class, the two groups which alone could
organise themselves sufficiently. Political struggles were therefore
linked to the confrontation between these two social groups, the educated class and the ulema, a phenomenon which has explosively expanded the diversity of ideological and social positions found within
political Islam. While antagonism between the two groups became
entrenched as the 1990s went on, ideological differences on certain
issues became more flexible. The Islamists largely renounced their
modernising project and toughened their attitude on social issues, at
the same time as traditionalist movements such as the Taliban became more radical as a result of their contact with transnational
organisations such as Al-Qaida, and embarked on an abortive plan
to set up a fundamentalist state. Here the Taliban demonstrated that
fundamentalists are not necessarily unaware of ideas relating to the
state. This in turn means that any notion that such a movement must
necessarily be committed solely to re-Islamisation from below should
at least be qualified.
In the French edition of this book which appeared in 2000 the
author stressed the unique nature of the Taliban regime, which did
not seem to offer a model adaptable to other circumstances by reason
of the very unusual conditions necessary for a religiously motivated
group to be in a position to impose a new kind of political structure.
354

Conclusion

355

However, the Taliban model had then, and still has today, an ideological impact in Pakistan and in the Gulf, although it is highly unlikely that the ulema could exercise similar influence in a more
differentiated society, such as that of Pakistan. The persistence of a
strong fundamentalist strain after the fall of the Taliban regime demonstrates in any case the resistance of a substantial section of Afghan
society to the liberal model presented under the auspices of international assistance.
From the 1990s onwards one aspect of this accelerated politicisation, namely ethnicisation, has taken on a growing importance. The
redefinition of national feeling through its attachment to ethnic
identities has been another major result of the war. Without a recognised ethnic hierarchy, and in the absence of mechanisms capable of
guaranteeing the equitable representation of different groups within
institutions, the present situation is characterised by the marginalisation of the Pushtuns, the numerically most significant ethnic group.
In this sense, the present situation could lead to mass slaughter, particularly in the north, where the position of Pushtun population
groups which are locally in the minority could become especially
perilous if international forces are withdrawn.
The re-establishment of the state has been the key issue in Afghanistan since the Soviet withdrawal and the victory of the mujahidin in
1992. The Afghan government exists solely as a consequence of international military and financial aid, the continuation of which depends
on considerations beyond the control of the Afghans themselves.
Afghanistans subjection to international intervention, resulting from
the position of the Taliban as an anomaly within the international
order, has led to an experiment in social engineering with effects
that are sometimes unpredictable. The need to buttress the central
authority could easily generate an authoritarian regime, and constitutional plans at present under consideration are not reassuring on
this point. Another issue is that there is nothing to indicate that this
nascent state could survive the discontinuation of international aid,
while much will ultimately depend on the capacity of the United
States to oversee the process of reconstruction over the long-term.
Having exploited the Afghan resistance to weaken the Soviet
Union with some success, the United States subsequently had no
coherent policy for Afghanistan. Only the existence of its radical

356

Conclusion

movements obliged the United States once again to become involved, but this involvement was concerned solely with security
considerations. Even in this context the outcome has not necessarily
been positive for the Americans. The radical militants are today able
to seek refuge in a region on the Afghan-Pakistani border in which
they can move unhindered. In additionand this is a key point
they are not under the control of any state since this zone largely
falls outside both Pakistani and Afghan control. A further issue is
that the continued presence of US forces in Afghanistan entails the
disadvantage that the Americans present a permanent target for small
groups of fighters. The presence of these forces perpetuates the link
between the Afghan crisis and a more global confrontation. The dynamic resulting from this interaction of heterogeneous actorsstates,
networks, transnational organisations and partieshas effects which
are largely unpredictable.
In any case, there is nothing to suggest that even with its overwhelming technological superiority the United States will in the
short term achieve the stabilisation of Afghanistan. The guerrilla
movements, though regional, have the capacity to survive for years,
placing their faith in an eventual disengagement by the United States.
Failure in Afghanistan would have significant consequences for
America as a power, and thus the issues involved in this interminable
war extend far beyond the destiny of Afghanistan itself.

Select Bibliography
Abadi, Basir Ahmad Daulat, Hazb va jaraynt-i sisi-yi Afghanistan (Parts
and political events in Afghanistan), Tehran: Moallef, 1982.
Adamec, Ludwig W., Whos Who of Afghanistan, Graz: Akademische Druckund Verlagstanstalt, 1975.
, A Biographical Dictionary of Contemporary Afghanistan, Graz: Akademische Druck- und Verlagsanstalt, 1987.
, Historical Dictionary of Afghanistan,Metuchen:Scarecrow Press,1991.
, Afghanistan,Les Temps Modernes,JulyAugust 1980 (special issue).
AFRANE (Amiti Franco-Afghane), La femme afghane travers lhistoire de
lAfghanistan, Actes du colloque Unesco, Paris, 11 December 1998.
Ahmad, Qeyamuddin, The Wahabi Movement in India, Islamabad: National
Book Foundation, 1979.
Ahmed, Akbar S., Millennium and Charisma among Pathans: a Critical Essay in
Social Anthropology, London: Routledge and Kegan Paul, 1976.
, Pukhtun Economy and Society, London: Routledge and Kegan Paul,
1980.
, The impact of Afghan refugees on ethnicity and politics in Baluchistan, Central Survey 9 (3), 1980.
, and D. Hart (eds), Islam in Tribal Societies, London: Routledge and
Kegan Paul, 1984.
Allan, P., and A. Stahel, Tribal guerilla warfare against a colonial power,
Journal of Conflict Resolution 27 (4), December 1983.
Amstutz, J. B., The First Five Years of Soviet Occupation, Washington, DC:
National Defense University Press, 1986.
Anderson, Jon W., Tribe and community among the Ghilzai Pashtuns,
Anthropos 70, 1975.
, There are no Khans anymore: economic development and social
change in tribal Afghanistan, Middle East Journal 32, 1978.
, Social structure and the veil: comportment and interaction in
Afghanistan, Anthropos, 77, 1982.
Arjomand, Sad Amir, The Turban for the Crown, Oxford University Press,
1989.
Arnold, Anthony, Afghanistan: the Soviet Invasion in Perspective, Stanford, CA:
Hoover Institute Press, 1981.

357

358

Select Bibliography

, Afghanistans Two-Party Communism, Stanford, CA: Hoover Institute Press, 1983.

, The Fateful Pebble: Afghanistans Role in the Fall of the Soviet Empire,
CA: Presidio, 1993.
Aziz, Abdullah, Essai sur les catgories dirigeantes de lAfghanistan 19451963,
Frankfurt-am-Main: Peter Lang, 1987.
Azoy, Whitney, BuzkashiSo that his Name Shall Rise, Philadelphia: University of Pennsylvania Press, 1982.
Bahadur, Kalim, The Jamaat-i-Islami of Pakistan, Lahore: Progressive Books,
1978.
Baitemann, Olga, NGOs and the Afghan War: the politicisation of humanitarian aid, Third World Quarterly 12 (1), January 1990.
Balleau-Lajoinie, La condition des femmes en Afghanistan, Paris: Editions Sociales, 1980.
Barry, Mike, Afghanistan, Paris: Seuil, 1971.
, Rpression et guerre sovitiques, Les Temps modernes, July-August
1980.
Barry, Mike, Le royaume de linsolence, Paris: Flammarion, 1984.
, LAfghanistan entre Marx et le Coran, Politique internationale 31,
spring 1986.
, La deuxime mort de lAfghanistan, Politique Internationale 50,
1991.
Barth, Fredrik, Political Leadership among Swat Pathans, London: Athlone
Press, 1959.
, Pathan Identity and its Maintenance in F. Barth (ed.), Ethnic
Groups and Boundaries: The Social Organisation of Culture Difference, Bergen,
Oslo: Universittsforlaget; London, Geo. Allen and Unwin, 1969.
, The Last Wali of Swat, Oslo: Norwegian Universities Press, 1985.
Bennigsen, Alexandre, and Chantal Lemercier-Quelquejay, Lexprience
sovitique en pays musulman. Les leons du pass et lAfghanistan,
Politique trangre 4, 1980.
, Le soufi et le commissaire, Paris: Seuil, 1986.
Broxup, Marie, The Soviets in Afghanistan: anatomy of a takeover, Central
Asian Survey 1 (4), 1983.
Canfield, Robert L., Hazara Integration into the Afghan Nation: Some Changing Relations between the Hazaras and the Afghans Officials, New York:
Asian Society, 1972, Occasional Paper no. 3.
and M. N. Saharani (eds), Revolutions and Rebellions in Afghanistan,
Berkeley: University of California Press, 1984.
Carr, Olivier, Mystique et politique, Paris: Editions du Cerf, Presses Sc-Po,
1984.

Select Bibliography

359

, LIslam laque ou le retour la grande tradition, Paris: Armand Colin,

1993.
Centlivres, Pierre, Un bazar dAsie centrale, Wiesbaden: Ludwig Riechert,
1972.
, Paysannerie et pouvoir en Afghanistan. De la fin de la monarchie lintervention sovitique, Geneva: Institut universitaire de hautes tudes internationales (Occasional Papers), 1985.
, Les tulipes rouges dAfghanistan. Anctres, matres spirituels et
martyrs dans une socit musulmanein J. Hamard,R. Kaehr and F. Sabelli
(eds),Les anctres sont parmi nous,Neuchtel:Muse dethnographie,1988.
, Les trois ples de lidentit afghane au Pakistan, LHomme 8/4
(108), 1988.
Centlivres, Pierre et al., Afghanistan, la colonisation impossible, Paris, Les Editions du Cerf, 1984.
Centlivres, P., and Micheline Centlivres-Demont,, Et si lon parlait de lAfghanistan?, Neuchtel, Editions de lInstitut dethnologie, Paris, Editions
de la Maison des sciences de lhomme, 1988.
Centlivres, P., and Micheline Centlivres-Demont, Hommes dinfluence et
hommes de partis. Lorganisation politique dans les villages de rfugis
afghans au Pakistan in Erwin Grtzbach (ed.), Neue Beitrge zur Afghanistanforschung, Liestal: Bibliotheca Afghanica, 1988.
Centlivres Demont, Micheline, Types doccupation et relations interethniques dans le Nord-Est de lAfghanistan, Studia Iranica 5 (2), 1976,
pp. 6977.
, Les rfugis afghans au Pakistan. Gestion, enjeux, perspectives in
R. Bocco and D. Mohammad-Reza (eds), Moyen-Orient. Migrations, democratisation, mdiations, Geneva: Institut Universitaire des Hautes Etudes
Internationales, 1994.
, Les femmes dans le conflit Afghan, Socit suisse, Moyen Orient et
Culture Islamique, no. 2, May 1996, pp. 1618.
Chishti, Nighat Mehroze, Constitutional Development in Afghanistan, Karachi:
Royal Book Company, 1998.
Daoud,Zemaray,LEtat monarchique dans la formation afghane,Bern-Frankfurt/
Main: Peter Lang, 1982.
Digard, Jean-Pierre (ed.), Le fait ethnique en Iran et en Afghanistan, Paris: Editions du CNRS, 1988.
Dikshit, P., 1993: Afghanistan policy, Strategic Analysis XVI (8), November
1993.
Dorronsoro, Gilles, Laide humanitaire en Afghanistan, Cultures et conflits,
autumn 1993.
, La politique de pacification en Afghanistan in Grard Chaliand
(ed.), Stratgies de la gurilla, Paris: Payot, 1994.

360

Select Bibliography

, Politique et ethnicit. Les Tdjiks dAfghanistan, CEMOTI 18,

1994.
, Afghanistan. Des rseaux de solidarits aux espaces rgionaux in
Franois Jean and Jean-Christophe Rufin (eds), Economie des guerres civiles,
Paris: Hachette, 1996.
, Les oulmas afghans, Archives des sciences sociales des religions,
November 2001.
, Pakistan and the Taliban: State Policy, Religious Networks and
Political Connections in Christophe Jaffrelot (ed.), Pakistan: Nationalism
without a Nation?, Delhi: Manohar, 2002.
Dupre, Louis, Mahmud Tarzi: the forgotten nationalist, American Universities Field Staff Report VIII (1), January 1964.
, Afghanistan under the Khalq, Problems of Communism, July
August 1979.
, Afghanistan, Princeton University Press, 1980 (1st edition 1973).
, Tribal Warfare in Afghanistan and Pakistan: a Reflection of the
Segmentary Lineage System in A. S. Ahmed and David Hart (eds), Islam
in Tribal Societies, London: Routledge and Kegan Paul, 1984.
Dupre, Nancy Hatch, Femmes dAfghanistan (special report), Les Nouvelles
dAfghanistan, no. 2930, Paris, 1986.
, Afghanistan: Women Society and Development, Journal of Developing Societies, 8 (1), 1992, pp. 3042.
Edwards, David Busby, Charismatic leadership and political process in
Afghanistan, Central Asian Survey 5 (34), 1986.
, The Evolution of Shii Political Dissent in Afghanistan in
J. R. I. Cole and N. R. Keddie (eds), Shiism and Social Protest, New
Haven: Yale University Press, 1986.
, Summoning Muslims: print, politics, and religious ideology in
Afghanistan, Journal of Asian Studies 52 (3), 1993.
Elphinstone, Mountstuart, An Account of the Kingdom of Caubul, Oxford
University Press, 1972 (1st edition 1815).
Etienne, Gilbert, LAfghanistan ou les alas de la coopration, Paris: PUF, 1972.
Farhang, Mir Muhammad Sediq, Afghanistan dar panj qarn-i akhir (LAfghanistan dans les cinq derniers sicles), Peshawar: Derarshesh, 1988.
Ferdinand, Klaus, Nomad expansion and commerce in central Afghanistan, Folk (4), 1962.
Gaborieau, Marc, Typologie des spcialistes religieux chez les musulmans
du sous-continent indien. Les limites de lislamisation, Archives des sciences sociales des religions 55 (1), 1983, pp. 2951.
, Le no-fondamentalisme au Pakistan: Maududi et la Jamaatislami in Olivier Carr (ed.), Radicalismes islamiques, Paris: LHarmattan,
1986.

Select Bibliography

361

Gellner, Ernest, Saints of the Atlas, London: Weidenfeld and Nicolson, 1969.
Gentelle, Pierre, LAfghanistan et laide internationale, Tiers-monde, 80,
1979.
Ghani, Ashraf, Islam and state building in a tribal society: Afghanistan
18801901, Modern Asian Studies, 12 (2), 1978.
, Disputes in a court of Sharia, Kunar Valley, Afghanistan, International Journal of Middle Eastern Studies, 15, 1983.
Gregorian, Vartan, Mahmud Tarzi and the Saraj-al-Akhbar: ideology of
nationalism and modernisation in Afghanistan, Middle East Journal, 21 (3),
1967, pp. 34568.
, The Emergence of Modern Afghanistan, Stanford University Press,
1969.
Hauner, Milan, One man against the Empire: the Faqir of Ipi and the British in Central Asia on the eve of and during the second world war, Journal of Contemporary History, 16 (1), 1981.
Lobato, Chantal, Femmes afghanes, femmes musulmanes. Islam, le grand
malentendu, Autrement, no. 95, Paris, 1987.
, Kaboul 19801986. Un islam officiel pour lgitimer le pouvoir
communiste, Central Asian Survey, 7 (23), 1988.
Lorentz, J. H., Afghan aid: The role of private voluntary organizations,
Journal of South Asian and Middle East Studies, XI (12), autumn-winter
1987.
Majrouh, Sayd Bahodine, Petite anthologie de la posie fminine, Pashto
Quarterly, Kabul, 1, 3, spring, 1978.
Maley, William (ed.), Fundamentalism reborn? Afghanistan and the Taliban,
London: Hurst, 1998.
Malik, Jamal, Colonization of Islam: Dissolution of Traditional Institutions in
Pakistan, Lahore: Vanguard Books, 1996.
Marsden, Peter, The Taliban: War, Religion and the New Order in Afghanistan,
London: Zed Books, 1998.
Marwat, Dr Fazal-ur-Rahim, The Evolution and Growth of Communism in
Afghanistan, 19171979, Karachi: Royal Book Company, 1997.
Matinuddin, Kamal, The Afghan Phenomenon: Afghanistan 19941997,
Karachi: Oxford University Press, 1999.
Metcalf, Barbara Daly, Islamic Revival in British India: Deoband 18601900,
Princeton University Press, 1982.
Newell, Richard S., The government of Muhammad Mussa Shafiq: the last
chapter of Afghan liberalism, Central Asian Survey, July, 1982.
Poullada, Leon, Reform and Rebellion in Afghanistan 19191929: King Amanullahs Failure to Transform a Tribal Society, Ithaca, NY: Cornell University
Press, 1973.
Roy, Olivier, Afghanistan. La guerre des paysans, Rvoltes logiques, 13, 1981.

362

Select Bibliography

, Intellectuels et ulmas dans la rsistance afghane, Peuples mditerranens, 21, 1982.

, Sufism in the Afghan resistance, Central Asian Survey, 2 (4), 1983.
, LAfghanistan, Islam et modernit politique, Paris: Esprit-Seuil, 1985.
, Le double code afghan. Marxisme et tribalisme, Revue franaise de
science politique, 35 (5), December 1985.
, Etat et socit en Afghanistan, Revue franaise de science politique,
36 (6), December 1986.
, Lchec de lIslam politique, Paris: Esprit-Seuil, 1992.
, La guerre dAfghanistan. De la guerre idologique la guerre
ethnique, LHomme et la socit, 107108, 1993.
Rubin, B. R., Lineage of the State in Afghanistan, Asian Survey, XXVIII (11),
1988.
, The old regime in Afghanistan: recruitment and training of a state
elite, Central Afghan Survey, 10 (3), 1991.
, Political elites in Afghanistan: rentier state building, rentier state
wrecking, International Journal of Middle East Studies, 24, 1992.
, The Fragmentation of Afghanistan, New Haven, London: Yale University Press, 1995.
Shalinsky, Audrey, Ethnic reactions to the current regime in Afghanistan:
A case study, Central Asian Survey, 3 (4), 1984.
Sivan, E., Ibn Taymiyya, father of the Islamic revolution, Encounter, May
1983.
Sliwinski, Marek, Afghanistan 19781987: War, demography and society,
Central Asian Survey (Incidental Papers Series), 1988.
Stahel, Albert A., Simulation of Guerilla Warfare in J. Halin (ed.), Simulationstechnik, Springer Verlag, 1987.
Tapper, Nancy, Bartered Brides: Politics, Gender and Marriage in an Afghan
Tribal Society, Cambridge University Press, 1991.
Tapper, Richard, Holier than You: Islam in Three Tribal Societies in
Akbar S. Ahmed, David M. Hart (eds), Islam in Tribal Societies, London:
Routledge and Kegan Paul, 1984.
(ed.), The Conflict of Tribe and State in Iran and Afghanistan, London:
Croom Helm, 1983.
Yousaf, Mohammad, The Silent Soldier, Lahore: Jang Publishers, 1991.

JOURNALS
Afghanistan Info,Comit suisse de soutien au peuple afghan,Neuchtel,1980.
Les Nouvelles dAfghanistan, AFRANE (Amiti franco-afghane), Paris, 1980.

Index
Abbas, mullah Muhammad, 267, 268,
174, 280
Abdul Rahman Khan, 11, 25, 26, 28,
29, 30, 31, 32, 33, 34, 37, 40, 41, 43,
49, 56, 61, 62, 63, 120, 177, 271, 280
Abu Hanifa (madrasa), 76, 152, 165
Achekzai tribe, 187
Afghan Mellat, 71, 72, 212
Afghan Transit Trade Agreement
(ATTA), 352
Afzali, Hafizullah, 127
Aga Khan, Sadruddin, 201
Ahl-i Hadith, 51, 52, 53, 231, 276, 277
Ahmadzai tribe, 58, 89, 164, 193, 194
Akbari, Muhammad, 140, 151, 153,
219, 251, 259, 270
Akhundzada, Rasul, 245
Alam, mawlawi, 165, 265
Al-Amr bil-Maruf, 284
Al-Azhar University, 50, 69, 76, 134,
152
Albright, Madeleine, 301
Alikozai tribe, 332
Alizai tribe, 135, 202
Al-Jazeera, 316
Almar, 261
Al-Qaida, 4, 5, 304, 305, 306, 307, 315,
319, 320, 321, 322, 326, 339, 340,
351, 354
Amanullah, Shah, 8, 25, 27, 28, 29, 30,
33, 35, 36, 37, 38, 39, 43, 46, 47, 49,
50, 55, 57, 58, 59, 61, 62, 65, 77, 152,
179, 182, 229, 296, 337, 361
Amin, Muhammad Anwar, 139
Amin, Qazi, 82, 165
Amin, Hafizullah, 1, 2, 30, 74, 79, 81,
85, 89, 90, 91, 92, 95, 96, 104, 112

Andar tribe, 109, 114, 167

Andarab valley, 65, 128
Andkhoy, 262, 263
Andropov, Yuri, 91, 178, 189, 193, 199
Arghandab valley, 167, 184
Argu, 149
Aymaqs, 12, 295
Azad Beg, 261
Azzam, Abdullah, 133, 134, 232, 304
Baba, Qari, 126, 222, 230, 245, 246, 250,
253
Babar, Nasrullah, 245, 311
Badakhshan province, 52, 69, 73, 76,
78, 81, 82, 97, 107, 113, 119, 120,
122, 125, 129, 135, 136, 149, 163,
164, 185, 210, 216, 229, 230, 253,
258, 263, 267, 284, 335
Badakhshi, Taher, 70, 86, 185
Badghis province, 43, 45, 100, 114, 229,
243, 251, 343
Baghran, 126, 135, 167
Bahes, Badruddin, 86
Balkh province, 134, 214, 215, 237,
263, 270
Balkhi, Ismail, 48, 55, 70, 105, 160
Baluch, 12, 15, 95, 112, 120, 122, 135,
171, 180, 181, 275, 303, 346, 353,
Bamyan, 44, 97, 103, 112, 140, 141,
155, 162, 163, 218, 227, 251, 255,
259, 270, 283, 309, 312, 332, 342
Bank-i Melli, 32, 62, 63, 65
Baraki Barak, 122, 179, 251
Barakzai tribe, 54, 332
Barayalai, Mahmud, 89, 174
Barelwi, 52, 276
Basmachis, 2, 177

363

364

Index

Beheshti, sayyed, 55, 118, 140, 151, 153,

218, 221
Behsud, 118, 120, 140, 155, 166, 218,
220, 260
Bhutto, Benazir, 144, 245
Bhutto, Zulfiqar Ali, 81, 85, 143, 218,
245
Brahimi, Lakhdar, 341
Brezhnev, Leonid, 3, 90, 91, 92
Bush, George W., 320
Casey, William, 2
Charasiab, 241, 250
CIA, 2, 144, 188, 208, 303, 318, 320,
321, 323, 326, 339, 341
Clinton, Bill, 306, 318
Cordovez, Diego, 199, 200, 201
Dar al-Ulum Haqqaniyah (madrasa), 52
Dara-yi Suf, 166, 221, 227
Dasht-i Archi, 229
Daud, 1, 28, 30, 31, 36, 38, 40, 56, 57,
65, 69, 74, 76, 79, 80, 81, 83, 84, 85,
86, 87, 90, 93, 95, 96, 103, 152, 164,
166, 194, 218, 281, 296
Deh Kundi, 123, 166, 221, 222
Deoband, 50, 51, 52, 277, 361
Dost Muhammad, 28, 34, 45
Dostum, Abdul Rashid, 236, 237, 238,
239, 240, 242, 243, 251, 252, 253,
254, 255, 260, 261, 262, 264, 265,
299, 325, 326, 327, 329, 335, 336,
342
Dubs, Adolph, 86
Durand Line, 24, 30
Durrani, Lieutenant-General Asad,
230, 245
Durrani, 7, 23, 26, 29, 31, 37, 41, 68,
109, 120, 135, 162, 167, 169, 182,
202
Emat, Nurullah, 127
Etemadi, Nur Ahmad, 96
Ettehad-e Islami, 139, 150, 151, 159,
161, 215, 216, 223, 224, 226, 239,
240, 242, 243, 244, 250, 284, 344

Farah province, 100, 135, 166, 201, 214,

229, 332
Faryab province, 229, 255, 261, 263,
336, 343
Fazlullah, commandant, 112, 122, 229
Federally Administered Tribal
Agencies (FATA), 351
Gardez, 91, 337, 340, 342
Gaylani family, 53, 54, 119, 138, 151,
152, 153, 155, 156, 160, 162, 163,
164, 223, 239, 274, 280, 323
Ghaffurzai, Abdul Rahim, 254
Ghani, Ashraf, 49, 56, 333, 344, 360
Ghazniprovince, 44, 49, 53, 54, 59, 71,
86, 98, 99, 106, 112, 113, 114, 120,
123, 126, 140, 162, 163, 164, 167,
212, 215, 221, 222, 241, 250, 251,
253, 274, 275, 281, 339
Ghilzai tribe, 7, 23, 29, 30, 31, 34, 40,
41, 42, 49, 59, 60, 75, 86, 89, 101,
109, 120, 162, 163, 164, 167, 169,
202, 227, 267, 357
Gholam Faruq, 156, 194, 238
Ghor province, 44, 222, 229, 336
Ghorband valley, 260
Gorbachev, Mikhai, 3, 4, 180, 191, 192,
195, 199, 200, 237
Glen, Fethullah, 290
Habib ur-Rahman, 82
Habibullah, amir, 152
Habibullah, Bacha-yi Saqao, 8, 38, 59
Haq, Abdul, 148, 155, 230, 256, 323,
324
Haqqani, mawlawi Jalaluddin, 114, 155,
166, 230, 240, 268, 275, 323, 338
Harakat ul-Ansar, 302, 303
Harakat-i Enqelab, 102, 109, 137, 138,
150, 151, 154, 159, 162, 163, 164,
165, 166, 167, 169, 210, 214, 215,
216, 217, 219, 222, 223, 231, 239,
241, 242, 245, 250, 261, 273
Harakat-i Islami, 140, 141, 143, 146,
150, 152, 160, 162, 167, 219, 220,
221, 222, 223, 225, 251, 259, 260,
298

Index
Hazarajat, 44, 45, 46, 47, 48, 55, 57, 69,
71, 98, 103, 112, 114, 115, 117, 118,
119, 120, 121, 123, 129, 140, 141,
147, 155, 162, 166, 182, 211, 212,
217, 218, 220, 221, 222, 233, 241,
242, 258, 259, 266, 270, 332, 343,
345
Hekmatyar, Gulbuddin, 73, 75, 81, 82,
83, 128, 138, 139, 145, 146, 151, 152,
153, 154, 156, 157, 163, 164, 165,
169, 171, 202, 204, 217, 223, 228,
231, 232, 239, 242, 243, 244, 250,
253, 256, 257, 261, 302, 339
Helmand province, 63, 78, 112, 114,
122, 126, 135, 165, 167, 184, 210,
215, 229, 243, 246
Herat, 4, 13, 23, 44, 54, 68, 69, 72, 73,
76, 82, 91, 97, 98, 99, 100, 101, 102,
107, 112, 117, 123, 125, 126, 127,
134, 142, 158, 162, 164, 165, 167,
178, 180, 187, 189, 191, 194, 196,
201, 205, 210, 214, 215, 216, 223,
229, 236, 240, 245, 246, 251, 252,
253, 257, 261, 265, 266, 280, 291,
297, 298, 299, 325, 332, 335, 336,
345, 352
Hezb-i Demokrat-i Mottaraki, 70, 72,
81
Hezb-i Demokratik-i Khalq-i Afghanistan, 1, 70, 72, 73, 74, 79, 85, 86, 91,
157, 174, 193, 203, 263
Hezb-i Enqelab-i Melli, 81
Hezb-i Islami (Khales), 138, 150, 151,
155, 159, 162, 166, 215, 216, 223,
226, 228, 230, 231, 240, 241, 242,
268, 273, 275, 284
Hezb-i Islami, 75, 81, 83, 111, 137, 138,
139, 143, 144, 145, 146, 149, 150,
151, 152, 154, 155, 156, 157, 158,
159, 162, 163, 164, 165, 166, 167,
168, 169, 171, 185, 202, 203, 204,
209, 212, 214, 215, 216, 217, 223,
226, 227, 228, 229, 230, 231, 232,
236, 237, 238, 239, 240, 241, 242,
243, 244, 245, 250, 253, 257, 261,
264, 268, 273, 275, 284, 298

365

Hezb-i Wahdat, 232, 234, 236, 237,

241, 242, 243, 244, 250, 251, 254,
255, 257, 258, 259, 260, 265, 269,
270, 271, 332, 335, 336, 342, 343,
345, 346
Hezb-i Watan, 203, 238, 260
Hottak tribe, 169, 267
Ibn Abdul Wahhab, 18
Ibn Taymiyya, 51, 158, 362
Ibrahim Mujaddidi, 56, 69, 138, 163,
166
Ibrahim Qazimi, 140
Imam Saheb, 75, 163, 227, 264
Iran, 1, 3, 4, 10, 12, 39, 44, 49, 54, 55,
65, 71, 78, 85, 89, 98, 100, 106, 110,
118, 123, 134, 135, 137, 139, 140,
142, 143, 146, 147, 160, 169, 170,
181, 198, 219, 220, 221, 225, 226,
236, 241, 243, 254, 255, 259, 260,
263, 272, 280, 311, 320, 335, 346,
347, 352, 353, 359, 362
ISAF (International Security Assistance
Force), 4, 337, 340, 342
ISI (Inter-Service Intelligence) 82, 83,
144, 145, 146, 219, 220, 230, 232,
245, 321, 327
Ismail al-Bukhari, 284
Ismail Balkhi, 48, 55, 70, 160
Ismaili, 185, 201
Istanekzai, 274
Istiqlal lyce, 65, 95, 128
Jadran tribe, 114, 337
Jalalabad, 4, 43, 52, 54, 68, 86, 134, 164,
178, 181, 187, 195, 201, 203, 215,
228, 229, 241, 254, 286, 298, 299,
301, 302, 325, 331, 332, 335, 336,
346, 352
Jamaat-i Islami, 82, 154, 157, 277, 302,
348, 349
Jamil ur-Rahman, 230, 231, 232, 280
Jamiyat Imam Bukhari (madrasa), 277
Jamiyat ul-Ulema, 179
Jamiyat-i Islami, 76, 81, 83, 102, 111,
137, 138, 149, 150, 151, 152, 154,

366

Index

156, 157, 158, 159, 162, 164, 165,

167, 168, 171, 202, 209, 215, 216,
217, 223, 225, 236, 237, 239, 240,
241, 242, 243, 244, 253, 257, 261,
264, 265, 266, 271, 273, 342, 343,
344
Jamiyat-ul-Ulum ul-Islamiyah
(madrasa), 276
Jan, Shah, 162, 167, 222
Jebhe-yi Melli, 152, 155, 156, 159, 160,
166, 223, 241, 250
Jeghatu, 122
Jombesh-i Melli, 237, 242, 243, 258,
260
Jozjan province, 44, 55, 184, 263, 336,
343
Kabul, 38, 13, 21, 23, 26, 27, 32, 38,
408, 54, 56, 59, 63, 65, 68, 70, 71,
72, 74, 75, 76, 79, 82, 85, 86, 89, 90,
91, 92, 95, 98, 100, 103, 104, 108,
110, 113, 122, 127, 128, 129, 134,
137, 140, 145, 152, 162, 163, 165,
17385, 193205, 210, 213, 214,
215, 218, 219, 221, 223, 225, 227,
228, 229, 233, 236, 237, 238, 239,
240, 241, 242, 243, 244, 246, 2507,
261, 266, 267, 268, 269, 274, 275,
279, 280, 281, 282, 283, 285, 286,
288, 291, 292, 296, 297, 298, 299,
300, 301, 302, 304, 308, 310, 313,
3205, 327, 329, 33140, 342, 344,
345, 346, 351, 361
Kalakani, Abdul Majid, 213, 284
Kandahar, 5, 23, 31, 34, 41, 42, 49, 53,
55, 65, 71, 83, 97, 100, 110, 120, 121,
141, 142, 162, 16670, 178, 182,
183, 184, 187, 189, 191, 194, 201,
202, 209, 213, 216, 217, 225, 229,
240, 241, 245, 246, 250, 251, 257,
267, 268, 273, 275, 280, 281, 282,
285, 287, 303, 307, 311, 321, 323,
325, 326, 332, 336, 338, 344, 352
Kapisa province, 201, 227
Karmal, Babrak, 2, 30, 70, 72, 74, 76,
85, 86, 88, 89, 92, 106, 152, 169, 173,

174, 175, 177, 178, 179, 181, 182,

185, 187, 193, 194, 195, 204, 218,
237, 260
Karzai, Ahmad Wali, 332
Karzai, Hamid, 4, 5, 323, 329, 330, 332,
333, 334, 336, 337, 340, 341, 343,
344
Keshem, 81, 135, 149, 164
Keshmand, Sultan Ali, 74, 89, 194, 197
KGB, 63, 86, 90, 91, 92, 176, 177, 178,
193
KHAD, 174, 177, 178, 179, 193, 194,
195, 197, 203, 204, 205, 237
Khales, Yunus, 138, 145, 150, 151, 152,
155, 159, 162, 163, 165, 166, 167,
202, 215, 216, 223, 226, 228, 230,
231, 239, 240, 241, 242, 268, 273,
275, 284
Khalili, Khalilullah, 76
Khalili, Muhammad Karim, 226, 251,
255, 259, 333, 336
Khalilzad, Zalmay, 302, 330, 342
Khalq, 68, 70, 72, 73, 75, 85, 86, 88,
141, 160, 175, 195, 217, 360
Khan, Khan Abdul Wali, 204
Khan, Allauddin, 100, 102, 127, 251
Khan, Aref, 112, 168
Khan, Ismail, 100, 102, 125, 126, 127,
133, 158, 167, 196, 210, 211, 223,
228, 229, 230, 240, 241, 243, 245,
246, 251, 252, 253, 254, 261, 266,
268, 327, 329, 332, 335, 336, 345
Khan, Khan Abdul Ghaffur, 181
Khan, Muhammad Ibrahim, 47
Khan, Padshah, 337
Khano (militia of), 184
Kharuti tribe, 75, 96, 163
Khayber, Mir Akbar, 85
Kheyradmand, mawlawi, 149
Khodam ul-Forqan, 56, 69, 98, 137, 215
Khoei, ayatollah, 141, 160, 220
Khomeini, ayatollah, 160, 166, 225, 272
Khost, 25, 27, 58, 229, 302, 303, 323,
336, 337, 352
Khwaja Ghar, 229

Index
Kunar province, 52, 56, 70, 72, 139,
170, 230, 231, 277, 280, 360
Kunduz province, 43, 62, 75, 112, 163,
168, 178, 227, 240, 243, 255, 263,
264, 325, 342, 343
Laeq, Suleiman, 174, 194
Laghman province, 56, 82, 93, 121,
165, 209, 217, 241
Lal o Sarjangal, 123, 166, 221, 259
Lashkargah, 135, 184
Lashkar-i Taiba, 350
Latif, Abdul, 110, 332
Logar province, 54, 95, 98, 106, 114,
122, 167, 179, 215, 229, 275, 281
Loya Jirga, 5, 29, 35, 38, 39, 44, 62, 81,
153, 173, 194, 197, 323, 329, 330,
337
Mahaz-i Melli, 138, 150, 151, 215, 217,
228
Majrouh, Sayd Bahodine, 361
Majruh, Shamsuddin, 65, 164, 171
Maktab-i Khidmat-i Mujahidin, 133
Malik, commandant, 52, 112, 254, 255,
265, 269, 276, 277, 361
Mansur, mawlawi, 82, 127, 215, 223, 334
Maoists, 68, 70, 102, 104, 128, 213, 214,
215, 218, 220, 221
Masud, Ahmad Shah, 3, 4, 8, 81, 82,
108, 122, 123, 125, 126, 127, 128,
129, 133, 148, 164, 185, 188, 196,
207, 208, 209, 210, 211, 217, 228,
229, 230, 236, 237, 238, 239, 240 ,
241, 242, 243, 244, 250, 251, 252,
253, 254, 255, 258, 265, 266, 268,
269, 274, 291, 307, 321, 327, 342,
343
Maududi, 51, 154, 157, 360
Maywandwal, Hashim, 65, 70, 71, 72,
79, 81
Mazar-i Sharif, 45, 47, 72, 98, 114, 116,
122, 125, 140, 153, 165, 169, 177,
178, 183, 185, 189, 190, 210, 229,
236, 237, 238, 241, 253, 254, 255,
257, 260, 261, 264, 265, 269, 270,

367

274, 286, 289, 298, 325, 326, 332,

336
Mazari, Abdul Ali, 140, 151, 153, 166,
250, 251, 259
Meydan Shahr, 222, 227, 250, 251
Mir Yazda, 45
Mohammad Ibrahim, Bacha-yi Gaw
Sawar, 47, 70, 103
Mohseni, shaikh, 55, 104, 141, 142, 147,
152, 153, 160, 223, 260
Muhammad, Din, 155, 162, 167, 336
Muhammadzai, 31, 60, 193, 254
Mujaddidi, family, 53, 54, 55, 58, 59,
60, 69, 73, 76, 96, 113, 118, 137, 151,
152, 153, 156, 162, 164, 165, 215,
226, 239, 240, 256, 274, 275, 280
Musharraf, General Pervez, 256, 309,
321, 323, 346, 347, 348, 350, 351
Muslim, Ismat, 187
Mustazaffin, 139, 140, 141, 150, 152,
153, 154, 160, 163, 225
Muttawakil, mullah Abdul, 274, 309
Nabi, mawlawi Muhammad, 137, 138,
151, 152, 153, 163, 166, 215, 238, 239
Nadir Khan, 60
Naim Khan, 47, 48
Najibullah, Muhammad, 30, 83, 89,
173, 177, 178, 193, 194, 195, 196,
197, 200, 202, 203, 204, 218, 237,
238, 242, 254
Najm ul-Mudaris (madrasa), 52
Najmuddin, 125, 210
Nangrahar province, 56, 82, 135, 136,
163, 178, 250, 352
Naqibullah, mullah, 167, 216, 229, 246,
332
Naqshbandi, 54, 76, 119, 152, 164
Nasr, 110, 139, 140, 141, 147, 150, 151,
155, 157, 160, 162, 166, 219, 220,
221, 222, 225, 259
National Reconciliation, 173, 185, 194,
195, 196, 197, 201, 202, 203, 204
Niazi, Abdul Rahim, 69, 72
Niazi, Gholam Muhammad, 56, 69, 82
Nili, 114, 221

368

Index

North West Frontier Province

(NWFP), 112, 136, 232, 245, 275,
276, 268, 277, 303, 349, 351, 353
Nur al-Modares (madrasa), 163, 215
Nur, Nur Ahmad, 74, 173, 174
Nuristani, 3, 15, 57, 96, 97, 103, 128,
139, 140, 180, 181, 230, 231, 296,
333
Nurzai tribe, 333

212, 214, 216, 223, 226, 229, 231,

238, 239, 240, 256, 277, 289, 302,
303, 307, 324, 344, 360
Popolzai tribe, 332, 344
Powell, Colin, 322, 323
Provincial Reconstruction Team, 342
Pul-i Charkhi, 86, 96, 104, 250
Pul-i Khumri, 63, 169, 185
Pushtun Zargun, 99, 100

Omar, Dr, 81, 82, 164

Qader, Abdul, 89
Qala-i Naw, 45, 71, 109
Qazimi, Ibrahim, 140
Qizilbash, 29, 32, 42, 44, 70, 72, 104,
162, 220, 260
Quetta, 71, 121, 167, 215, 218, 219,
221, 223, 256, 311, 323, 349
Qutb, Sayyed, 51, 75, 76, 152, 157, 158

Paderwatan, 180, 182, 183, 185, 187,

196
Pakistan, 3, 4, 6, 24, 29, 42, 52, 58, 65,
69, 71, 76, 78, 81, 82, 83, 84, 85, 86,
89, 98, 108, 118, 121, 122, 123, 127,
128, 134, 135, 136, 137, 138, 139,
143, 144, 145, 146, 147, 148, 152,
154, 158, 169, 170, 171, 181, 182,
189, 191, 198, 199, 200, 202, 204,
208, 209, 210, 212, 213, 214, 215,
216, 219, 220, 222, 226, 227, 228,
229, 230, 231, 232, 236, 239, 243,
244, 245, 252, 253, 254, 256, 269,
271, 273, 274, 276, 277, 280, 287,
289, 301, 302, 303, 304, 307, 308,
309, 310, 311, 313, 316, 319, 320,
321, 322, 323, 325, 326, 339, 346,
347, 348, 349, 350, 351, 352, 353,
355, 358, 359, 360, 361
Paktia province, 337
Panj Pir (madrasa), 230, 232
Panjshir valley, 3, 45, 71, 82, 122, 123,
125, 127, 128, 184, 188, 189, 196,
208, 254, 258, 266, 291, 332
Parcham, 68, 70, 72, 73, 74, 75, 77, 85,
86, 88, 89, 175, 185
Parwan province, 178
Pashais, 180
payluch, 110
Perez de Cuellar, Xavier, 198, 199
Peshawar accords, 238, 239, 240
Peshawar, 24, 38, 59, 62, 67, 69, 71, 75,
82, 97, 122, 123, 132, 133, 134, 137,
138, 139, 1 48, 152, 156, 193, 198,

Rabbani, Burhanuddin, 30, 76, 81, 82,

83, 128, 151, 152, 153, 157, 162, 163,
164, 165, 177, 216, 223, 225, 236,
239, 240, 241, 243, 244, 245, 246,
251, 252, 253, 255, 258, 266, 274,
281, 299, 303, 333
Rabbani, mullah, 274, 281
Rad, 141, 142, 162, 223, 251
Rasul Palawan, 260, 261, 262, 265
Ratebzad, Anahita, 74, 89, 174, 194
RAWA, 215, 256
Reagan, Ronald, 2, 208, 225
Rumsfeld, Donald, 318
Saddeqi, Mir Hussein, 140, 219, 221
Sadozai tribe, 54
Sadruddin Aga Khan, 201
Safi tribe, 31, 32, 42, 43, 217, 230
SAFZA, 70, 185
SAMA, 213, 214, 221
Sama, Dr Sima, 344
Samangan province, 201, 255, 261, 270
Sami al-Haq, maulana, 277
Sar-i Pul, 255, 263
Sarkateb, commandant, 245
Sarwari, Asadullah, 90, 142, 174

Index
Saudi Arabia, 85, 133, 143, 226, 231,
243, 277, 301, 305, 306, 307, 309, 320
Sayyaf, Abdul Rasul, 139, 151, 152,
159, 163, 165, 223, 224, 226, 239,
251, 269, 329
Sepah-yi Enqelab, 184
Sepah-yi Pasdaran, 140, 221
Setam-i Melli, 70, 73, 86, 180
Sevan, Benon, 201, 238
Shafaq, Azizullah, 140, 166
Shafiq, Musa, 65, 76, 78, 79, 96
Shah Shoja, 106
Shah Walihullah, 18
Shahr-i Bozorg, 185
Shariat Madari, 100
shariat, 35, 41, 57, 58, 88, 152, 179, 246,
278, 280, 283, 285, 293, 297, 299, 300
Sharif, Nawaz, 144, 244, 256, 311
Sharq, Hasan, 225
Shiberghan, 184, 336
Shindand, 86, 100, 127, 201, 251, 252,
266
Shinwari tribe, 42, 59, 333, 345
Shir Agha Shongar, 100
Shirin, mawlawi, 164
Shirzai, Gul Agha, 332
Shura-yi Arbabah, 220
Shura-yi Ettefaq, 218, 225
Shura-yi Nazar, 210, 329, 332, 333,
334, 337, 340, 342, 343
Spin Boldak, 183, 227, 245, 340
Suleiman Khel tribe, 43, 163, 167
Tajikistan, 12, 310
takfir, 51, 158, 159, 231
Takhar province, 45, 180, 215, 263,
270, 343
Tanai, general, 174, 195, 204
Tanzim Nesl-i Naw-i Hazara-Moghol,
71, 218, 219
Taraki, Nur Muhammad, 30, 65, 70, 72,
74, 85, 86, 88, 89, 90, 91
Tarzi, Mahmud, 36, 37, 66
Tashqurghan, 12, 210, 211
Tawala, commandant, 220
Tenet, Charles, 321

369

Termez, 176, 262

Tora Bora, 326
Turkestan, 45, 108, 263
Turki, prince, 307
Turkmeni, shaikh Mohaqeq, 219
Ulumi, Nur ul-Haq, 202
UNHCR, 271, 353
United Nations, 4, 63, 132, 192, 198,
201, 225, 242, 255, 282, 300, 308,
309, 311, 313, 317, 326, 327, 330,
337, 339, 341, 344
UNOCA, 201, 202
UNOCAL, 301, 302
Uruzgan province, 44, 48, 120, 222,
229, 273, 326, 341
Uzbeks, 9, 13, 34, 40, 41, 43, 106, 135,
180, 182, 183, 211, 262, 263, 268,
270, 296, 342, 343
Vambry, Armnius, 101
Waez, ayatollah, 104, 141
Wahdat-i Melli, 76
Wahid, ras Abdul, 126, 167, 210, 215
Wakhan corridor, 24
Waras, 55, 114, 118, 140, 218, 221
Wardak province, 44, 96, 112, 122, 135,
148, 155, 164, 211, 222, 229, 230,
246, 288, 333
Wardak, Amin, 112, 122, 148, 155, 211,
230, 246
Watanjar, Aslam, 86, 88, 89, 90, 174,
195
Wilson, Charles, 208
Wolfowitz, Paul, 302
Yakaolang, 121, 140, 155, 166, 220,
221, 270
Yar, Khaleq, 194
Yusuf Beg, 47, 48
Yusuf, Ramzi, 303
Zabihullah, comandant, 122, 125, 126,
156, 164, 189

370
Zabul province, 246, 340
Zahir Shah, 1, 30, 39, 40, 59, 78, 79,
80, 103, 142, 166, 172, 193, 200,
223, 330
Zahir, Abdul, 78

Index
Zawahiri, Ayman al-, 304
Zeary, Saleh Muhammad, 173, 195
Zia ul-Haq, 85, 135, 143, 144, 200,
276, 303
Zindajan, 99

REVOLUTION UNENDING

The CERI series in

Comparative Politics and
International Studies
Series editor CHRISTOPHE JAFFRELOT
This series consists of translations of noteworthy
publications in the social sciences emanating from the
foremost French research centre in international studies,
the Paris-based Centre dEtudes et de Recherches
Internationales (CERI), part of Sciences Po and associated
with the CNRS (Centre National de la Recherche
Scientifique).
The focus of the series is the transformation of politics
and society by international and domestic factorsglobalisation, migration and the post-bipolar balance of power on
the one hand, and societal dynamics, ethnicity and religion
on the other. States are more permeable to external
influence than ever before and this phenomenon is
accelerating processes of social and political change the
world over. In seeking to understand and interpret these
transformations, this series give priority to social trends
from below as much as the interventions of state and
non-state actors.
Founded in 1952, CERI has fifty full-time fellows
drawn from different disciplines conducting research on
comparative political analysis, international relations,
regionalism, transnational flows, political sociology, political
economy and on individual states.

GILLES DORRONSORO

Revolution Unending
Afghanistan: 1979 to the Present
TRANSLATED FROM THE FRENCH BY JOHN KING

Columbia University Press

New York
in association with the Centre dEtudes et de
Recherches Internationales, Paris

Columbia University Press

Publishers Since 1893
New York
Copyright 2005 Gilles Dorronsoro
Library of Congress Cataloging-in-Publication Data
Dorronsoro, Gilles.
[Rvolution afghane. English]
Revolution unending: Afghanistan: 1979 to present/by Gilles
Dorronsoro; translated
from the French by John King.
p. cm.
Includes bibliographical references and index.
ISBN 0-231-13626-9 (cloth: alk. paper)
1. AfghanistanHistorySoviet occupation, 1979-1989.
2. AfghanistanHistory1989-2001. 3. AfghanistanHistory2001I. Title.
DS371.2.D69313 2005
958.1045dc22
2004057131
Columbia University Press books are printed on permanent and durable
acid-free paper.
c 10 9 8 7 6 5 4 3 2 1
Printed in India