The Astrological Origin of Islamic Geomancy: 1. Geomancy and Its Distribution in Time and Space

The astrological origin of Islamic geomancy
Wim van Binsbergen1

© 1996-2004 Wim van Binsbergen
1. GEOMANCY AND ITS DISTRIBUTION IN TIME AND

SPACE
Perhaps the best way to introduce the geomantic family of divination systems,
and their enormous cultural historical significance, is by pointing out that it, and
it alone, provides the answer to that unfortunately overlooked, yet inspiring,
question of modern scholarship: What do a nineteenth-century CE German
farmer, a twentieth-century CE typing-girl from Botswana, and a late first-
millennium CE Arabian sage have in common?
First introduced into West European intellectual life in the 11th century CE,
when numerous Arabic texts were translated, geomancy as a divination method
became associated with the most prominent representatives of the occult
sciences in medieval and Renaissance times, including Bernardus Silvestris,
Roger Bacon, Albertus Magnus, Cornelius Agrippa and Robert Fludd. As in the
1
This paper was read at The SSIPS/ SAGP 1996, 15th Annual Conference: ‘Global and Multicultural
Dimensions of Ancient and Medieval Philosophy and Social Thought: Africana, Christian, Greek,
Islamic, Jewish, Indigenous and Asian Traditions’, Binghamton University, Department of
Philosophy/ Center for Medieval and Renaissance studies (CEMERS), October 1996. The
anthropological field-work by which this paper was originally inspired, was conducted in 1988-1994
in Francistown (Botswana) and surrounding regions, occasionally extending into adjacent parts of
Zimbabwe. I am indebted to my main teachers of Southern African geomancy, the late Mr Smarts
Gumede and the late Mrs Rosie Mabutu, as well as to my mentrix in sangoma mediumship, Mrs
Elizabeth Mabutu (MmaShakayile); to the African Studies Centre (ASC), for research funds and for
granting me the freedom to venture rather beyond the research programme of the Department of
Political and Historical Studies; to the Royal Netherlands Academy of Sciences and its Netherlands
Institute of Advances Studies in the Humanities and Social Sciences (NIAS), Wassenaar, for offering
a most stimulating environment where the present paper was written; to the ASC and the Department
of Cultural Anthropology and Development Sociology, Free University, Amsterdam, for granting me
leave of absence during the academic year 1994-95; to the ASC, Tervuren and NIAS library staff, for
untiring assistance; to the members of the NIAS 1994-95 theme group on Magic and religion in the
Ancient Near East, for accommodating me in their midst, and for useful criticism (especially by Shaul
Shaked) of the present paper, which was presented at the group’s weekly seminar, 2 december 1994;
finally to my friend Rafat Badawy for reading al-Zanātī with me, which resulted in our manuscript
entitled: Treatise on the principles of sand-science by Sidi al-Shaykh Muhammad al-Zanātī, to which
has been added: Treatise on the ‘Jafr wa Qar’at’ by the Sayyid Ja‘far al-Sadiq, translation R. Badwy
with the assistance of Wim van Binsbergen.
Wim van Binsbergen: The astrological origin of geomancy (1996) – p. 2
Arabic originals, by Muhammad al-Zanātī, cAlī b. Umar, Fadhl b. Sahl al-

Saraksi, Ahmad b. cAli Zunbul, etc., the system was popularised through ‘books
of fate’, and ‘punctuation’ (as it was called after its characteristic patterns of
dots) became a self-help oracle even in the rural peripheries of European life
‘Also farmer Hinrich Fehse must have looked up things in such a book, one of the
characters in the Theodor Storm’s [1817-1888 – WvB] novelette ‘Draussen im
Heidedorf’ [‘ Out there in the village on the moors’]:
‘‘For this purpose he produced my mother’s punctuation book from the sewing box.
He set himself opposite me and then started to produce dots with chalk on the table.
He did this with such urgency and got such a blush in the process that I asked him:
‘Hinrich, what is it that you are punctuating on, after all?’ ‘Shut up, shut up’, he said.
‘Keep to your sewing job!’ – But without him noticing I leaned over the table and
saw at which number in the book he held his finger. – So the question was if the sick
person would get better? – I kept quiet and went on with my work, and he stipulated
on, counted ‘even’ or ‘odd’, and finally punctuated the figures with chalk on the
table.’’ ’2
From a very different corner of the Old World: Francistown, Botswana,

Southern Africa, hails the following account of Jane Sinombe’s oracular
consultation in 1991:
‘As one of the many daughters of Mr R. Sinombe, the high priest of the ancient and
widespread Mwali territorial cult, Jane (aged 21), an unattached typist living in
Botswana’s booming Francistown, has grown up in a milieu where divination,
mediumistic cults and herbalism were taken for granted. A prominent place among her
adolescence memories is occupied by the annual meetings of the Kwame/Legwame
Traditional Association – one of the several Botswana guilds of traditional healers, and
the one through which her father, as founder-president, pursued the local continuity of the
Mwali cult in a formal organisational shape well within Botswana’s Societies Act – with
scores of traditional healers discussing, celebrating and officiating at her father’s
spacious Nata homestead at 200 km north-west from Francistown. Only the previous year
she was instrumental in escorting one of the novices3 of MmaShakayile, a particularly
prominent cult leader, to Nata for official registration with her father’s guild. By that time
she was still enrolled in a typing course. Meanwhile she has finished the course and had
managed to find work as a secretary at a contractor’s office. Her euphoria over this hard-
to-find job has not lasted long: within a few months she finds herself the centre of a
2 Helm, E.M., 1985, Het grote orakelboek [: kaartleggen – koffiedik kijken – handlezen – astrologie
en vele andere wegen om met elkaar in de toekomst te kijken], Utrecht/Antwerpen: Spectrum, p. 72f,
my English translation of this passage; the book is a Dutch translation of Das große Orakelbuch,
München: Ehrenwirth, 1983.
3 The present writer. For an extensive treatment of my apprenticeship and practice as a Southern
African geomancer, cf. van Binsbergen, W.M.J., 2003, Intercultural encounters: African and
anthropological explorations towards a philosophy of interculturality, Berlin/Münster: LIT.
frightening whirlpool of backbiting and slander, in the course of which her financial
integrity is questioned and several candidates have presented themselves to take over her
newly-gained job. Her father’s towering position in the cult does not mean that he
engages in divination, he is not staying in Francistown, and ever since her parents’
divorce Jane’s attitude towards him has been very ambivalent. So it is to MmaShakayile
that she now turns for divinatory clarification and advice. This cult leader’s urban yard in
Monarch, an upgraded former mine compound, in its various rambling structures houses
a large number of junior relatives and lodgers who have no dealings with the cult, but the
yard is foremost, in popular perception, a cultic lodge, where junior adepts (amathwaza,
singular uthwaza) are being purified and trained in order to assume, ultimately, the office
of sangoma spirit medium, which includes trance divination, officiating at laymen’s
sacrifices to the latter’s ancestors, herbal therapy, and operating the region’s
characteristic four-tablet oracle (hakata or ditaola) – the local variant of geomancy.
Rarely does the old lady throw the tablets herself these days, but her senior adepts are
fully capable of conducting a divinatory session without even her intervention.
Upon her arrival at the yard, one evening, Jane presents herself to Molly, senior
adept and MmaShakayile’s granddaughter, whom she finds chatting in front of the main
house. She briefly explains the purpose of her visit, and is taken inside where she is
seated on a mat, the legs stretched in front of her. Jane produces her five-Pula note,4 and
MmaShakayile’s sacred hakata (four small flat wooden tablets each with distinct
markings) are produced. Molly is to preside over the session, with MmaMleya and
Johannes5 – both amathwaza about to graduate – in attendance in order to complement
Molly’s interpretations and to gain further experience and confidence. After Jane has
imparted her own life essence to the tablets by handling them and blowing over them,
Molly throws them down on a gaudy cloth featuring pictures of lions in the sacred
colours red, white and black. She names their first fall as Chilume6 (one of the sixteen
configurations that are possible when each of four tablets can end up face up or face
dawn), and interprets this fall in terms of an inquiry about sorcery. Then she throws
again, naming and interpreting this fall again, and so on, occasionally handing the tablets
to Jane to let her throw, and persuading her, through interjections, questions, hesitations,
to formulate specific questions and to respond to the ever more detailed and dramatic tale
that gradually shape up out of the oracular pronouncements of Molly and her two junior
colleagues: a tale of competition between colleagues at Jane’s place of work, and of the
sinister, occult weapons which her adversaries are employing. In the final part of the
session, which all together comprised twenty-eight throws, Jane is exhorted to rely on her
ancestors, and on MmaShakayile’s counter-medicine (for which an additional P10 is to
be charged), in order to fight back her colleagues’ attacks, and she is reassured that she
will keep her job and even gain a supervisory rank in the near future. (This came true,
incidentally.) Momentarily relieved, Jane leaves the yard, and stops a taxi to take her all
4 Pula, the Botswana currency. At the time, P1 ≈ US$0.50, and the legal minimum wage was just
below P1 per hour; Jane earned about 50% more than that.
5 The present writer.
6 , i.e. the configuration where all tablets are upside down except for the senior
male tablet. For further details, see: van Binsbergen, W.M.J., 1993, ‘Divinatie met vier tabletten:
Medische technologie in Zuidelijk Afrika’, in: Sjaak van der Geest, Paul ten Have, Gerhard Nijhoff en
Piet Verbeek-Heida, eds., De macht der dingen: Medische technologie in cultureel perspectief,
Amsterdam: Spinhuis, pp. 61-110; van Binsbergen, W.M.J., 1995, ‘Four-tablet divination as trans-
regional medical technology in Southern Africa’, Journal of Religion in Africa, 25, 2: 114-140.
the way back to her room in the Block VII residential area, a distance of some fifteen
kilometres all across the sprawling town.’7
And finally, as one of the earliest documentary attestations of geomancy in an

Arabian context:
‘Ce procédé [i.e. geomancy], considéré par Doutté8 comme une modernisation d’at-tarq
bil-’hasat9,10 est déjà décrit par Ibn al-A’rabi 11 en ces termes: ‘‘Le hazi s’assied et fait
tracer, par un jeune garçon à son service, des lignes sur le sable ou de la poussière; il les
trace avec agilité et promptitude, de manière qu’il soit impossible de les compter. Puis,
sur l’ordre du maître, il les efface deux à deux, tout en disant: ‘Vous deux, témoins
oculaires [de la volonté des dieux], faites apparaître promptement l’évidence’’12! S’il n’en
reste à la fin que deux lignes, c’est signe de succès; mais s’il n’est reste qu’une, c’est
signe d’échec et de malheur.’’13 14
What is the formal nature and what is the inner working of this geomantic
divination system, whose ramifications stretch across three continents and
several millennia? What are the historical antecedents that explain its
emergence, apparently out of the blue, in Islamic Mesopotamia by the end of the
first millennium CE? These are the central questions of the following argument.
7 Author’s field-notes.
8 Doutté, E., 1909, Magie et religion dans l’Afrique du Nord, Alger: Jourdain.
9 ‘Casting of pebbles’, i.e. lithoboly.
10 Doutté, o.c. p. 378. [Fahd’s original footnote]
11 ‘Mort vers 230/844, à l’âge de 81 ans’; cf. Brockelmann, C., 1937-42, Geschichte der arabischen
Literatur, Leiden: Brill, Supplementband I, 179; cité ap. Alusi, loc. cit. [Fahd’s original footnote; al-
Alusi, i.e.: Mahmud Shukri al-Alusi, Buludj al-arab, III, 323; cf. Fahd 1966: 195 n. 4] . This 9th
century CE figure is not to be confused with the famous mystic by the same name, nearly half a
millennium later (1165-1240 CE; cf. Chittick, W.C., 1989, The Sufi Path of Knowledge: Ibn al-
c
Arabî's Metaphysics of Imagination, Albany: SUNY Press.
12 Here Fahd gives a long philological footnote citing passages from the above work of al-Alusi, and
from: al-Zabidi, Muhammad Murtadah, Taj al-‘arus min jawahir al-Qamus, i-x, éd. du Caire
1286/1869-1287/1870; and refers to: Goldziher, I., 1902, ‘Einige arabische Ausrufe und Formeln’,
Wiener Zeitschrift für die Kunde des Morgenlandes, Wien, 16: 131-146, esp. 139f.; all of which is
suggestive of an intimate link between geomancy, belomancy and ornithomancy.
13 Cette ligne est appelée āsham, ‘noire’ (Taj al-‘arus min jawahir al-Qamus, 5, 129, l. 12 sq, cf. ib.
l. 13 sqq. (...) Comp. Taj al-‘arus min jawahir al-Qamus s.v. khatt, traduit par A. Jaussen, Les
coutumes des Arabes au pays de Moab, 183, n. 1. [Fahd’s original footnote, with shortened references
expanded by me]
14 Fahd, T., 1966, La divination arabe: Etudes religieuses sociologiques et folkloriques sur le milieu
natif de l’Islam, Leiden : Brill, p. 197f.
2. ISLAMIC GEOMANCY AS A DIVINATION SYSTEM
Essentially, any inductive divination system consists of two15 elements: a

random generator producing a-select numbers, and an interpretative catalogue
whose entries can be selectively consulted by means of these numbers.16 In
formal terms, the geomantic family of divination systems is characterised by
random generators (often – but not always – involving the use of sand or earth,
hence the common designation cilm al-raml,17 ‘sand science’) producing 24
equiprobable different configurations, all of which have been named and
specified in the attending catalogue.
In the standard geomantic practice sixteen different configurations are
arrived at by the construction of four horizontal rows, each element consisting
of one dot (for the outcome ‘odd’ of a particular chance procedure, e.g. making
an uncounted number of dots on the ground (cf. Plate 1 below) or on a piece of
paper; throwing pebbles, shells, sticks etc.) or two dots (for ‘even’). In this way
patterns of the typical and well-known geomantic figures are produced:
Two dots on the same horizontal line may also be connected so as to form
horizontal lines (an obviously secondary usage common in Arabic contexts):
15 Or rather three, if one distinguishes, between the random generator and the interpretative
catalogue, the coding procedures by which the numbers produced by the former, are turned into
meaningful entries (e.g. geomantic figures, of the characteristic shape etc. – see below)
whose meaning is then to be looked up in the interpretative catalogue; cf. van Binsbergen, W.M.J.,
1997, ‘Rethinking Africa’s contribution to global cultural history: Lessons from a comparative
historical analysis of mankala board-games and geomantic divination’, in: van Binsbergen, W.M.J.,
1997, ed., Black Athena: Ten Years After, Hoofddorp: Dutch Archaeological and Historical Society,
special issue, Talanta: Proceedings of the Dutch Archaeological and Historical Society, vols 28-29,
1996-97, pp. 221-254. For our present purpose, these coding procedures may well be regarded as an
intrinsic part of the interpretative catalogue.
16 Cf. van Binsbergen 1994, o.c.; van Binsbergen, W.M.J., 1995, ‘Four-tablet divination as trans-
regional medical technology in Southern Africa’, Journal of Religion in Africa, 25, 2: 114-140; van
Binsbergen, W.M.J., 1996, ‘Transregional and historical connections of four-tablet divination in
Southern Africa’, Journal of Religion in Africa, 26, 1: 2-29;
17 In Arabic pronunciation, the article’s ‘-l’is assimilated to the initial ‘r-‘ of raml.
Dots may be connected by lines also vertically and diagonally, and the resulting
figures embellished at liberty, so as to produce the fantastic glyphs of geomantic
magic (cf. Plates 2 and 3 below). Each of the sixteen possible configurations is
identified by name, and for each a number of standard interpretations is given in
an interpretative catalogue which in principle is in written form, although in less
literate peripheries it may simply have been committed to memory. Naturally, in
the course of their transmission over vast expanses of space and time the
contents of these catalogues may have come to vary considerably, but an
underlying identical structure and contents remains detectable. Below (Tables 1,
3) we shall look at some of these catalogues in detail. In the Southern African
case, the chance procedure leading to the selection of one out of the sixteen
configurations consists not of uncontrolled, spontaneous production of a
repetitive series followed by elimination and assessment in terms of ‘odd’ or
‘even’, but, as we have seen at Jane’s session, by the fall of four tablets which
can each take two values, notably ‘face up’ and ‘face down’; in this far corner of
the Old World, where the presence of geomancy and Islamic influence in
general so far has gone unnoticed, the interpretative catalogues as applied by
individual diviners tend to be very idiosyncratic yet, as I have shown
elsewhere,18 derive from the same medieval original geomantic corpus as the
catalogues found throughout the Islamic world and in Europe.
A central feature of classic, Arabic geomancy is the construction of a
‘complete geomantic theme’: from the original four tetragrams (the four
‘mothers’) twelve other configurations (four daughters, four nephews, two
witnesses and two judges, of which the last one is often called ‘arbitrator’) are
derived by a special form of addition, based on the following computational
rules:
even + odd = odd, i.e. oo +o = o,

odd + even = odd, i.e. o + oo = o,
even + even = even, i.e. oo + oo = oo,
odd + odd = even, i.e. o +o = oo.
In passing we note the strange mixture of a kinship idiom, with a legal idiom
of witnesses and judges.
There are indications however that the mothers here do not primary stand for
a kinship reference, but for fundamental cosmological entities: planets or
18 van Binsbergen 1996, o.c.

elements.19 In very rare variant it is fathers and sons, rather than mothers and
daughters, that are being constructed.20
If we represent each geomantic element in formula as the bilocal relation:
E (a, b)
where
E = a geomantic element (one out of sixteen)
a = serial number of element E in the ordered series 1 through 16
b = line number
whereas
∑(a1+a2) = the computational result of the value on line b, of elements numbered a1
and a2,
then the computational rules for the derived figures (i.e.: all figures except the
four mothers) can be summarised in formula format as follows:
E(5, 1...4) = E (1...4, 1)

E(6, 1...4) = E (1...4, 2)
E(7, 1...4) = E (1...4, 3)
E(8, 1...4) = E (1...4, 4)
E(9, 1...4) = E (∑
(1+2), 1...4))
E(10, 1...4) = E (∑
(3+4), 1...4))
E(11, 1...4) = E (∑
(5+6), 1...4))
19 Carra de Vaux, B., 1974, ‘al-Sabi’a’, in: Gibb, H.A.R., & J.H. Kramers, 1974, Shorter
Encyclopaedia of Islam, Leiden: Brill, reprint of the first edition of 1953, pp. 477-78; mothers as
elements also among the Ikhwan al-Safac, cf. de Boer, T., 1921, De wijsbegeerte in den Islam,
Haarlem: Bohn, p. 90; also cf. de Boer, T., 1967, The history of philosophy in Islam, trans. by E.R.
Jones, first published 1903.
20 The point is not without significance. In an Arabic environment, it would elicit comment if the
dominant patrilineal and patriarchal idiom is not observed; the descent series mothers/ daughters/
nephews is distinctly matrilateral. Should this be read as evidence that the system was imported from a
non-patrilineal environment? And what milieu in the ancient and medieval Near East would qualify as
such an environment? Or, somewhat more likely, could it be taken – within the generally patrilineal
context of the Near East – as another example of the widely attested general rule, first expressed by
Meyer Fortes (cf. Fortes, M., 1953, ‘The structure of unilineal descent groups’, American
Anthropologist, 55: 17-41.) that such symbolic domains as religious specialism, spirit possession,
magical expertise etc. tend to be inherited not in the dominant but in the submerged line of descent,
i.e. matrilaterally in a patrilineal society. Another widely attested application of this rule is the fact
that when, in West Asian including Arabian magic, a client is to give a parent’s name, it is usually the
name of the mother, not the father, which is required.
E(12, 1...4) = E (∑
(7+8), 1...4))
E(13, 1...4) = E (∑
(9+10), 1...4))
E(14, 1...4) = E (∑
(11+12), 1...4))
E(15, 1...4) = E (∑
(13+14), 1...4))
E(16, 1...4) = E (∑
(1+15), 1...4))
The example in diagram 1 should suffice to illustrate this apparently

complex algorithm.
For those not familiar with geomancy it may be useful to point out that the
appearance of a series of sixteen elements in two related but essentially different
contexts may at first present a source of confusion:
(a) mathematically, there are 16 possible different configurations which

can be formed out of four lines, having in each line one element
which can take two values (‘even’ or ‘odd’, ‘oo’ or ‘o’). Whenever a
geomantic figures is constructed by means of a chance process, one
and only one of these sixteen will be produced;
(b) The construction of a full geomantic ‘theme’ (i.e. horoscope’) out of

four original mothers resulting from chance operations, necessarily
produces a series of sixteen and only sixteen figures, several of which
may occur more than once, as is clear from the example in diagram 1,
where only and occur only once.
Figure 1. An example of the production of a full geomantic theme of 16 figures,

starting from the four ‘mothers’
For the mathematician, these procedures offer all sorts of possibilities for
further combinatorial and (since they involve all kinds of symmetries)
topological manipulation,21 which would largely seem to be empty mental
exercises not conducive to increased insight in the geomantic system. However,
there is one important, and unexpected, implication of the mathematical
properties of the generation of the complete geomantic theme: although all
configurations have an equal chance of occurring among the first four elements
(the mothers), the computational rules in combination with the rotational
symmetry of the tetragrams are such that out of the sixteen geomantic figures
(16 possible ‘mothers’), only eight can be produced as judges. This means that
oracular pronouncements produced with classic (= ‘complete theme’) geomancy
are not truly equiprobable, and that in fact pessimistic outcomes are less
frequent. This effect does not occur with the simpler versions (without derived
geomantic theme), nor with the attenuated African versions.
In the simpler form, the fifteenth element, the ‘judge’, decides the entire
outcome of the oracle; the specific configuration it displays is interpreted
primarily by reference to a catalogue listing the specific oracular meanings of
the sixteen named configurations. Here the literal semantic contents of the name
(in Arabic, Greek or Latin) of the configuration usually serves as a point of
departure for the interpretation. The interpretation may be enriched by taking
into account the planetary and zodiacal associations of each of the sixteen
names. A provisional summary is presented in Table 1.
In the more complex versions like al-Zanātī’s22 the first twelve of these
sixteen configurations are identified with the twelve astrological houses and
interpreted accordingly, while taking into account the usual correspondences
and conventional meanings of the twelve houses, the planets and the zodiacal
signs.
A sample from the beginning of al-Zanātī’s highly influential Treatise on the
principles of sand-science may convey the general tone and the highly elliptical
style of this author:
21 Jaulin, R., 1957, 'Essai d’analyse formelle d'un procédé géomantique', Bulletin IFAN, xix, 1-2,
série B: 43-71; Jaulin, R., 1966, La géomancie: Analyse formelle, Cahiers de l’Homme, Ethnologie–
Géographie–Linguistique, N.S., iv, Paris: Mouton; Jaulin, R., 1968, ‘Sur la géomancie’, in: Caquot,
A., & M. Leibovici, eds., 1968, La divination, tome second, Paris, Presses Universitaires de France,
pp 473-552.
22 al-Zanātī, Muhammad, 1320 H./1923 CE, al-Fasl fi usūl c
ilm al-raml, Cairo; al-Zanātī,
Muhammad, 1341 H./1902 CE, al-Aqwal al-mardīyya fi ma rifat al-acmal al-ramlīyya, Cairo; both in
c
the University of Leiden library.

‘In the name of ALLAH,

the Compassionate, the Merciful
and peace be upon Our Lord Muhammad,
the most exalted of all prophets,
and upon his kith and kin and all his companions
Now this book is the treatise on the principles of sand-science according to the bases of
the Idrīsian facts. Shaykh Zanātī – may the forgiveness of Allah be upon him23 – has said
that every science has a beginning and every work an introduction and we want to
introduce the bases of sand-science which give satisfaction to beginners and foundations
to who is already advanced in this science. We begin in the name and with the help of
Allah, saying that sand-science goes back to that prophet of Allah by the name of Idrīs,
may the forgiveness of Allah be upon him and upon our prophet Muhammad and all other
prophets before him. Idrīs has seen facts in his dreams and the scientists and wise people
have found that all the creatures consist of four elements: Fire, Air, Water, Earth24 and
then four results heat, cold, moist, dry, and four directions east, west, south, north. And
they have looked at people’s activities finding that there are four measures: weighing,
counting, planting, and farming. And the name of Allah is four letters and cIsa is four
letters and Muhammad is four letters25 and each house26 must have four corners. Then
everything is stabilised on four corners, four figures, four extremities [of the human
body], four elements. They have built configurations up from them designating them
‘Mothers of the houses’. The first house is the future and it is enunciative, because it is
the first you see when the mystery of the world is explained in terms of reality and
practice. That is why they call it the future of the client,27 moreover it is the soul’s house,
the ‘self of men’. Then they have formed a second house called the house of capital. It
informs you of the conditions of capital and money. And they have formed a third house
called house of movements because when a person gets capital he can move and act....
[etc. the fourth house is introduced, as well as the formation of up to sixteen houses – i.e.
configurations – basically by regrouping of elements in the first four configurations,
along the lines set out in my diagram 1, above] ...and they have formed a new
configuration called ‘sand ruler’ and it is the ‘friendship balance’ and sand too, and it can
never be a person. It is the fifteenth house. Then they have formed from the first and the
fifteenth house a new configuration called ‘the end result’. It is the sixteenth. From that
house you get sand from sand. When you see happy configurations in the sand you
pronounce good news and success. When you see distressful or tragic configurations in
the sand you pronounce bad news from whatever is the object of the consultation. When
you see in the sand a mixed configuration you pronounce middle results ‘difficult in the
23 I.e. ‘the late’; Zanātī is here implied not to have written this book himself.
24 Arabic: trobiyya, an adjective referring to both ‘earth’ and ‘dust’, and hence coming close to
‘sand’ (Arabic: raml) on which the sance-science (cilm al-raml) revolves.
25 Sc.: ‘four letters in Arabic’. Muhammad is written in four letters but the central ‘m’ has a
duplication sign. cIsa (the initial cain also counts as a letter) is the Arabic name for the founder of
Christianity, called Jesus by Christians.
26 Arabic: bait, meaning ‘human dwelling’, but also the established astrological technical term
‘house’ as a section (usually 1/12) of the zodiak specific to a particular person, place, and moment of
time.
27 In the present translation, this word designates the one who consults the oracle.
beginning and easy at the end’. The happy configurations are five in number:
. The strong configurations for happiness are three: . The
distressful negative in the sand are four: .28 The mixed configurations are five:
...’ etc.29
Figure 4. Opening page of a popular Cairene edition of al-Zanātī, 1320

H./1923 CE.
28 Repetition in the original.

29 Badwy & van Binsbergen, o.c.
The planetary correspondences attributed to the zodiacal signs in subsequent

sections of Zanātī’s argument, as well as those attributed by other Arabic
geomantic authors, are fairly standard considering the astrological tradition as
codified by Ptolemy and Manilius in Imperial Antiquity, but the geomantic
zodiacal correspondences often so deviate from this tradition that – given the
present, well developed, state of our knowledge of the history of astrology in the
Babylonian, Greek, Hellenistic, Hebrew, Arabic, Indian and Chinese contexts –
they may help to situate the emergence of geomancy in time and place.
Ibn Khaldun’s Muqaddima,30 written by the end of the 13th century CE,
explains the emergence of geomancy as resulting from a situation when would-
be astrologers, typically poor, under-educated and urban, had no longer access
to the astronomical tables and the complex techniques necessary for calculating
a proper professional horoscope, and therefore replaced the empirical input (the
actual, astronomically absolutely correct, position – in most cases simplified to a
mere longitude – of the heavenly bodies at the particular moment which the
horoscope seeks to interpret) by the mock-astrology of geomancy: a series of
chance outcomes of simple manipulations with a stick on sand, with pebbles,
beans, shells or with pen on paper, but subsequently interpreted in the light of a
conventionalised or better ossified astrological idiom deprived from all spatio-
temporally specific astronomical input. This status as a ‘poor man’s astrology’
is confirmed by the fact that only the first four astrological houses receive an
independent input, while the contents of the eight others is merely secondarily
derived – which is comparable with a hypothetical situation in which some
astrologer would only calculate the planets and zodiacal signs for the first four
houses, and ignore the others, or (and this as been widely attested practice
astrological since Antiquity)31 as if the horoscope would be based not on the
30 Ibn Khaldoun, 1863-68, Prolégomènes historiques, traduits en français et commentés par [W.]
M[acGuckin baron] de Slane, Paris: Imprimérie Impériale; photomechanical reprint, Paris 1934-38;
Ibn Khaldûn, 1980, The Muqaddimah: An introduction to history, translated from the Arabic by F.
Rosenthal, 3 vols, second printing of second edition, Princeton (N.J.): Princeton University Press,
1980; first edition Bollingen Series XLIII, New York: Bollingen Foundation Inc, 1958.
31 A full discussion falls outside our present scope; cf. van Binsbergen, W.M.J., in preparation (b),
Four tablets: A Southern African divination system in its transregional and historical context. Let me merely
mention as an example the once enigmatic Tabula Bianchini, which was effectively and lucidly
interpreted in the following book: Boll, F.J., C. Bezold & W. Gundel, 1966, Sternglaube und
Sterndeutung: Die Geschichte und das Wesen der Astrologie: 5. durchgesehene Auflage mit einem
bibliographischen Anhang von H.G. Gundel, Darmstadt: Wissenschaftlichte Buchgesellschaft; first
published Leipzig 1926: Teubner Verlag, p. 60, 191f.
painstaking calculation of the heavens at the time of birth, but by the mere
casting of dice.32
The distribution area of geomancy encompasses the entire Islamic world,
most of Africa (including large non-Islamic sections), most of Europe, and large
sections of a para-Islamic periphery in Asia and the Indian Ocean region.
Moreover, with the Atlantic slave trade the geomancy-based divination systems
of West Africa crossed into the Caribbean, Central and South America.33
Geomancy must be one of the most impressive examples of ‘globalisation’
avant la lettre, i.e. before the modern means of transport and communication
made us take for granted the instant world-wide diffusion of material and
intellectual products. Although often practised by illiterate diviners and their
clients (e.g. in West and Southern Africa and on Madagascar), and partly rooted
in simple chance procedures (like hitting the earth, throwing tablets, beans,
shells etc.) which do not require recourse to writing, the system in its subsequent
processing and interpretation of the chance outcomes has unmistakably a literate
basis. Hence the most usual designation of the geomantic system of divination is
khatt al-raml, i.e. ‘sand calligraphy’, with a definite reference to writing,
literacy.
It is no accident that the prominent anthropologist Jack Goody concludes his
introduction34 to a seminal collection of papers on Literacy in traditional
societies with an extensive reference to geomancy:
32 In fact, geomantic dice are being used in India and Iran: Culin, S., 1890-91, ‘East Indian Fortune-
telling with dice’, Proceedings of the Numismatic and Anthropological Society of Philadelphia, 1890-
91: 65. Of course, the Southern African Hakata four-tablet oracle could be regarded as consisting of
geomantic dice, but it is difficult to regard them as modifications of some South or South West Asian
geomantic dice: the latter are cubic and four are pivoted together on a rod, so that each can
independently show any one of its four available faces; the Hakata are unstrung when in use, and each
can show only two faces. There are however many mixed sets of Hakata dice which include four-
sided dice made out of quadruped’s hoofs or astragali.
33 Bascom, W., 1980, Sixteen cowries: Yoruba divination from Africa to the New World,
Bloomington: Indiana University Press; Bastide, R., 1968, ‘La divination chez les Afro-Américains’,
in: Caquot, A., & M. Leibovici, red., 1968, La divination, tome second, Paris, Presses Universitaires
de France, pp. 393-428, espec. 423f.
34 Goody, J., 1968, ‘Introduction’, in: Goody, J., ed., Literacy in traditional societies, Cambridge:
Cambridge University Press, pp. 1-26.
Table 1. The sixteen geomantic figures, their names in Arabic and Latin, their
meaning, planetary and zodiacal associations, and their overall benefic or
malefic nature.35
35 Source: le Scouézec, ‘Géomancie’, o.c.; al-Zanātī, o.c.

‘The significance of writing varies widely among the societies discussed [in this edited
collection of papers]. But even among pastoral peoples like the Somali, even in societies
long cut off from the mainstream of literate cultures, like the Merina of Madagascar, even
in religiously very mixed areas, like the western Sudan, the book is an important feature
of social life, because it provides a reference point for individual and social behaviour,
especially that kind of verbal behaviour we think of specifically as ‘symbolic’, magico-
religious, mythopoeic or cosmological – though our given categories do us a great
disservice here. So that when Hébert36 offers us a ‘structural analysis’ of divinatory
systems in Madagascar and in Africa (it is even less easy to perceive the force of
‘structural’ here than in most usages of the term), he is pointing to certain features of a
system of divination, crystallized in writing, that is carried out from Kano to Calcutta,
from Tananarive to Samarkand, a ‘symbolic’ system that has little or no intrinsic
connection with the myths, beliefs and categories of the peoples among whom they [sic]
are found. Even supposing that all the societies in question were equally committed to
‘orthodox’ Islam and acknowledged the Islamic elements in the system of divination
(such as the use of the names of caliphs and archangels and the sacred names of God),
many elements in the system (e.g. the 9-cell squares) have quite a different derivation, a
much wider distribution, which makes them as much (and as little) an intrinsic part of
specifically Gonja or Hausa symbolic structures as the mathematical theory of groups is
of specifically Japanese or Belgian thought.’37 (Goody 1968: 25f)
Not only is geomancy thus an early example of globalisation, it is also a

powerful reminder, especially for anthropologists and then again especially for
the Africanists among them, that even conspicuous and central intellectual and
symbolic products are not necessarily contained within the idiosyncratic
linguistic and cosmological horizons of a localised culture, but may reflect
interconnections and transmissions across vast expanses of space and time,
which may render them totally incapable of explanation by reference to a local
society and culture. Implied are fundamental questions of cultural meaning: in
so far as the outcomes of geomantic divination have meaning, as they clearly
have for farmer Hinrich and typist Jane, such meaning, although mediated
36 The reference is to: Hébert, J.C., 1961, ‘Analyse structurale des géomancies comoriennes,
malgaches et africaines’, Journal de la Société des Africanistes, 31, 2: 115-208; and thus primarily to
geomancy. Yet Goody’s argument here does not entirely limit itself to this form of divination, as is
shown by his references to ‘the names of caliphs and archangels and the sacred names of God’, which
are absent at least from the versions of geomancy known to me, as are ‘the 9-cell [] squares’.
Obviously, the well-known magical squares are meant, that feature in the magical sciences from
Graeco-Roman Antiquity, Islam, South and East Asia, and whose distinctive feature is that the sum of
the numbers in all cells in one row (whether horizontally, vertically or diagonally) is identical.
Goody’s ‘9-cell square’ would have rows and colums of three cells each. However, although common
and wide-spread in its simplicity, this is by no means the standard shape of such squares, since they
exist for all numbers from 1 (or, more properly, 2) to 7, each number specific to a particular planet
(including the great luminaries Sun and Moon.
37 Goody, J., o.c., p. 25f.
through local concepts of e.g. illness and sorcery, can ultimately derive from
and be expressed in, other forms than just the local symbolic system, and may
partly refer to conditions and cosmological connections way beyond the reach of
the actors using the geomantic system in a particular spatio-temporal setting. Is
it this very remoteness – the actors’ awareness of drawing from a fount that is
conspicuously not local and that therefore has no resonances in their everyday
experience – which adds credibility and legitimation to the system? Is it, in other
words, the boundary-crossing nature of the system, its absence of recognisable
cultural specificity or perhaps its having absorbed influences from a great many
diverse intellectual sources, which explains its power, rather than its power
explaining its boundary crossing? What is, among rival divination systems at the
local scene (there is nearly always a diversity in this field) the unmistakable
competitive power of geomancy?
Figure 2. Distribution of the geomantic family of divination systems.38

1. Sikidy of Madagascar and Comoro Islands
2. Ramala-shastra of South Asia
3. Ifa and ‘Sixteen Cowries’ of West Africa and the New World
4. Southern African four-tablet system
5. simple geomancies of the African interior
38 For a fuller discussion cf. van Binsbergen in: ‘Rethinking’, o.c.

6. I Ching system of China

7. khatt al-raml and European derivates since the Middle Ages
These are momentous questions, and they are being confronted in the course
of my ongoing research on geomancy. However, in the present paper my focus
shall be not on transmission, reception and transformation, but on origin. Where
did the incredibly successful geomantic family of divination systems come from?
In order to answer this question, I shall first summarise the state of the art in
this field (section 3); then discuss some of the methodological problems
involved and suggest how an anthropologically-inspired typological approach
may help to overcome them – but after demonstrating how anthropology can
also be a poor guide in approaching these problems (section 4); and finally I
shall l apply that approach to one of the central features of the geomantic
system: the bafflingly obscure nomenclature of its sixteen configurations. This
will highlight the tremendous contribution which classical astrology has made to
the geomantic system; I shall have to assume in my reader a basic knowledge of
astrology, some key notions of which are brought together in the Appendix.
3. THE HISTORY OF GEOMANCY AS A PROBLEM
If the history of science can be said39 to be Clio’s step-child (and by

implication: neglected) this applies a fortiori to the history of geomancy, which
yet for centuries constituted one of the most prominent occult sciences.
In this case the situation is particularly marked since all traces of the genesis
of Islamic geomancy appear to have been blotted out. It appears to have sallied
forth like Athena, a young adult clad in full armour (and a cleromantic deity, at
that!), from her father’s forehead.
When in the second half of the nineteenth century scholarly translations became
available of major Arabic works such as Ibn Khaldun’s Muqaddima and Shaykh
Muhammad ibn Umar al-Tunisi’s Voyage au Darfour,40 and when also the
39 Dijksterhuis, E.J., 1990, Clio’s stiefkind, ed. K. van Berkel, Amsterdam: Bert Bakker.
40 Mohammed ibn Omar el Tounisi, 1845, Voyage au Darfour par le cheykh Mohamed ebn-Omar a-
Tounsy ou l’alignement de l’esprit par le voyage au Soudan et parmi les Arabes du centre de
systematic study of African societies made a beginning,41 the result was a –

regrettably ephemeral – interest in the distribution and history of the geomantic
family of divination systems. Within decades, offshoots of the family were
described for the medieval classical Arabic culture, West Africa, East Africa and
throughout the Indian Ocean, Byzantium, medieval Hebrew literature; also,
classic geomantic texts from the occult tradition of Europe since the Middle
Ages became available outside the narrowest circle of philological specialists
and esoteric dilettante, and began to appear in a new light. On the strength of his
phenomenal knowledge of manuscript sources (by then still almost entirely
unpublished and scattered) Steinschneider devoted a perceptive early study to
the Malagasy ‘sikidy’ or geomantic figures,42 and established their relationship
to the main geomantic family; in his other works, on medieval translations
between Hebrew, Arabic, Latin and Greek, and on the Indian naxatra or lunar
mansions,43 he repeatedly came back to geomancy as a genre of intellectual
production. Another major researcher in this field was that great historian of
science Paul Tannery, who already in 189744 made a fundamental statement
concerning the origin of the term ‘geomancy’, identifying it as the literal
l’Afrique, publié en arabe par le Dr Nicolas Perron, Paris: Duprat; Mohammed ibn Omar el Tounisi,
1851, Voyage au Darfour par le cheykh Mohamed ebn-Omar a-Tounsy, ou l’alignement de l’esprit
par le voyage au Soudan et parmi les Arabes du centre de l’Afrique, traduit de l’arabe par A. Perron,
ed. M. Jomard, Paris: Duprat.
41 For the study of geomancy in Africa, the work of Richard Burton was important: virtually all his
books contain a passage or footnote devoted to this theme: Burton, R.F., 1856 [check 1858], First
footsteps in East Africa, London, p. 55f; Burton, R.F., 1893, A mission to Gelele, King of Dahomey, i -
ii, London: Tylston & Edwards, Memorial edition, first published in 1864, i 330f; Burton, R.F., 1893,
The book of Thousand nights and a night, London, p. 55 n. 1;
42 Steinschneider, M., 1877, ‘Die Skidy oder check geomantischen Figuren’, Zeitschrift der
deutschen Morgenländischen Gesellschaft: 762-765.; of course he made the point that the desgination
s[i]kidy is from the Arabic shikl, figure; also cf. Ferrand, G., 1891-1902, Les musulmans à
Madagascar et aux Iles Comores, 3 vols (1891, 1893, 1902), Paris: Ernest Leroux, i, 73f.
43 Steinschneider, M., 1956, Die europäischen Übersetzungen aus dem Arabischen bis Mitte des 17.
Jahrhunderts, rééd. à Graz: Akademische Druck- und Verlagsanstalt; Steinschneider, M., 1864, ‘Über
die Mondstationen (Naxatra), und das Buch Arcandam’, Zeitschrift der Deutschen Morgenländischen
Gesellschaft, xviii: 118-206. Also cf. Steinschneider, M., 1862, Zur pseudoepigraphischen Literatur
insbesondere der geheime Wissenschaften des Mittelalters aus hebräischen und arabischen Quellen,
Berlin; Steinschneider, M., 1893, Die hebräischen Übersetzungen des Mittelalters, Berlin, Preisschrift
der Pariser Académie des Inscriptions, Paris: Académie des Inscriptions – both of which I have not yet
seen.
44 Tannery, P., 1897, ‘[short untitled note on geomancy]’, Comptes Rendues de l'Académie des Insc.
xxv, 1897, p. 519 [check 529] ; reprinted in Tannery, P., 1922, ‘La géomancie (Hugo Sanctelliensis)’
[Extrait des Comptes rendus de l’Académie des Inscriptions et des Belles-Lettres, 1897, xxv: 529], in:
Heiberg, J.-L., ed., 1922, Paul Tannery: Mémoires scientifiques, vol. v: Sciences exactes au Moyen
Age, Toulouse: Privat/ Paris: Gauthier-Villars, p. 356.
translation, by the pioneer medieval translator Hugo Sanctalliensis, of the

Arabic cilm al-raml, literally, ‘sand science’.
The point is of some interest, since Hugo by this designation implicitly
associated the newly imported divinatory science with the concept of
geomantia, divination by the earth, which in a world thought to consist of four
elements including earth had to exist if only for systematic reasons; but whereas
the term itself is attested45 in Antiquity there is not the slightest information
concerning the procedures constituting this geomancy. In fact Hugo’s was an
apt misnomer, since the earth or sand only provides one of the many possible
forms the random generator necessary for the system can be produced, and once
the numbers of the relevant catalogue entries have been produced by chance, the
catalogue itself is interpreted in a framework which is not so much an
‘astrologia terrestris’,46 but simplified astrology tout court, as we have seen
above for al-Zanātī.
45 The word ‘geomantia’ is traditionally associated with the Roman writer Varro (116-27? BCE), cf.
Cardauns, B., ed., 1976, M. Terentius Varro, Antiquitates rerum divinarum, Teil I Die Fragmente,
Mainz: Akademie der Wissenschaften under der Literatur/ Wiesbaden: Franz Steiner Verlag. The
standard reference to the – otherwise undefined – ‘Varronic geomancy’ in the early middle ages is in
Isidorus’ encyclopaedia: Lindsay, W.M., ed., 1957, Isidori Hispalensis Episcopi Etymologiarum sive
Originum Libri XX, Oxford: Clarendon; also Migne, J.P., ed., [year] Patrologia latina, LXXXII, cols.
73-728 (Isidorus hispalensis). There is not the slightest evidence that Varro had sophisticated and
formalised randomising and computational techniques in mind, as would have foreshadowed the
Islamic geomancy (cilm al-raml). Rather, he seems to have dealt with simple and unprocessed natural
phenomena such as cracks in the dessicating earth, the fissures created by earth quakes, etc. If so, his
geomancy would be another instance of ‘divination by natural, randomly formed cracks’, of which the
most famous example comes from the eastern end of Eurasia: reading the natural cracks that appeared
in sheep’s scapulars after heating them over a fire was the standard Chinese divination technique
during the Shang dynasty, 2nd millennium BCE. Cf. Keighley, D.N., 1978, Sources of Shang history:
Oracle bone inscriptions of Bronze Age China, University of California Press, Berkeley/Los
Angeles/London. Although Rome was in distant, indirect contact with China in Late Republican and
especially Imperial times, Chinese scapulomancy had long become obsolete by that time; if there is a
historical connection (which given the remarkable uniformity of divination systems throughout
Eurasia is quite likely) it would go back to a common cultural fond dating to the Neolithic or even the
Upper Palaeolithic, probably in Western Asia. The same fond may be encountered, in fragmented and
eroded form, in the well-known genres of omina literature of, especially, the Ancient Near East and
South Asia.
46 A designation used by the Arabic author ‘Ali, e.g.: Volkommene geomantia etc., 1715 -16,
Volkommene geomantia, deren ersten Theil die aufs neue revidierte...Punctir-Kunst in sich begreiftt...
Der andere Theil aber des berühmten Arabers Abu-Hali-ben-Omar niemals vorhin gedruckte
Astrologia Terrestris order Irrdische Stern-Kunde, Freystadt; ‘Ali was popularised by de Cattan’s or
Catani’s version, also in German editions cf. Steinschneider 1956 o.c., and as: de Cattan, C., 1572 La
géomancie du Signeur ChristofeCattan, Gentilhomme genevois: Livre non moins plaisant et récréatif,
que d’ingénieuse invention, pour sçavoir toutes choses présentes, passées et advenir, avec la Roue de
Pythagoras..., Paris: Jean Corrozet; also cf. Skinner, S., 1980, Terrestrial astrology: Divination by
geomancy, London: Routledge & Kegan Paul.
Tannery continued to work on the topic, and after his death a section of more
than a hundred pages on the topic was published in volume IV of his
posthumous Mémoires scientifiques, including a contribution by Carra de Vaux
on ‘la géomancie chez les Arabes’.47
Tannery concentrated on the Byzantine geomancies. Thus al-Zanātī’s
geomantic treatise (written by an author about whom no biographical
information whatsoever – not even a patronymic – is known, but who by his
name could be considered to have begun, or ended, his life as an member48 of
the maghrebine tribe of the Zanāta) was first introduced into the Byzantine
milieu when the monk Arsenius49 translated it in 1266 into Greek verse, from
the Persian. This was however by far not the earliest geomantic text to reach the
West, as the 11th century CE translations by Hugo Sanctalliensis and others
demonstrate.50 Tannery appreciated that the word rabolion does not exist in
classical Greek and can only be understood as a Greek mere phonetic rendering
of an untranslated Arabic (raml, ramlīyya) , with the -ml- assimilated to -b- and
the foreign root embellished with a Greek morphological ending -on. Like this
pivotal word, also Hugo’s translation in general became a model for the
assumption that, as far as geomancy was concerned, there could only have been
a one-way traffic from Arabic to Byzantine and Latin sources.
Neither Tannery, nor even his Arabist co-worker Carra de Vaux, managed to
proceed beyond the presence of geomancy in medieval classical Arabic culture.
47 Tannery, P., 1920, ‘Le Rabolion – Traités de géomancie arabes, grecs et latins’, in: Heiberg, J.-L.,
ed., Mémoires scientifiques [P. Tannery], vol. iv: Sciences exactes chez les Byzantins (1884-1919),
Toulouse: Privat/Paris: Gauthier-Villars, pp. 295-411; Carra de Vaux, Baron, 1920, ‘La géomancie
chez les Arabes’, in: Heiberg, J.-L., ed., 1920, Mémoires scientifiques [P. Tannery], vol. iv: Sciences
exactes chez les Byzantins (1884-1919), Toulouse: Privat/Paris: Gauthier-Villars, pp. 299-317.
48 In view of the very unusual absence of the patronymic, one might venture to suggest that he was
only an adoptive member, from unknown origin.
49 I have not been able to identify a text containing Arsenius’ translation. Steinschneider 1956, p. 7,
in his typical stenography, refers to a manuscript source which I would not know how to locate:
‘Arsenius, ein Mönch, übs (126) die arab. Geomantie von al-Zanātī aus dem Persischen (daher Z.
als 'Perser' bezeichnet in griechischen Versen, Probe aus ms Wien in Lambec, Comm. ed. Kollar
p. 554 n. 140; Nicoll, Catal. p. 327; vgl. HÜb 855)’.
The latter reference is to Steinschneider 1893 o.c. Note the designation ‘‘Persian’’ for al-Zanātī,
which might also be explained in a more direct fashion (that he had actual Persian connections; cf. his
appeal to that other authority with eastern connotation, Tumtum al-Hindi) than the use of Persian as an
intermediary in translation.
50 A rather full if cumbersome catalogue of such geomantic translations can be extracted from
Steinschneider 1906; also cf. Carmody, F.J., 1956, Arabic astronomical and astrological sciences in
Latin translation: A critical bibliography, Cambridge & Los Angeles, which I have not yet seen.
The system’s inner workings and variations were described in moderate detail,
but the historical questions asked concerned diffusion of the geomantic corpus
once already established, not the origin itself of that corpus. Perhaps one
considered the problem already solved; for, as Steinschneider51 had already
claimed, it was clear that cilm al-raml could only have come ‘from the desert’ –
for where else could one possibly find the few handfuls of sand necessary for
this divinatory practice...? But even if we were to take the message implied in
the name cilm al-raml literally, and situate its origin in a sandy environment,
then it is still true that there are plenty of different deserts (including that of
Syria in West Asia, the Kalahari and Sahara in Africa, and the Gobi and the
Tarim deserts in East Asia) within the distribution area of the geomantic family
of divination systems, which after all encompasses the better part of the Old
World.
Since, there have – again, to my knowledge – only been very few serious
scholarly contributions to the earliest history of geomancy.52
A. Delatte and E. Delatte, Belgian classicists and historians of science, must
have realised that sand was not all that rare in Classical, i.e. Hellenic and
Hellenistic Antiquity, even outside deserts. As is generally known, sand was the
standard medium in which to work out mathematical problems. Plutarch tells us:
‘Just so at Syracuse, it is said, after Plato had arrived [ca. 387 BCE], and an insane ardour
for philosophy laid hold on Dionysius, the king’s palace was filled with dust by reason of
the multitude of men that were drawing their geometrical diagrams in it...’53
51 1877 o.c.
52 A popular New Age type of account, without references and largely absurd, is: Pennick, N., 1992,
‘Ancient secrets of the earth: The oracle of geomancy’, in: Matthews, J., ed., The world atlas of
divination: The systems – where they originate – how they work, Boston/Toronto/London: Bulfinch
Press/Little, Brown & Comp., pp. 195-201. Pennick includes a map (p. 196) which postulates a
possible origin in Greece, subsequent spread to Egypt 8th century CE, Sudanic belt 10th century,
which produces Ifa and ‘Sixteen Cowries’, the dominant West African versions of geomancy; from
Egypt 9th century alleged spread to Madagascar through Ethiopia etc., and along the same route to
India; 12th century western Europe from North Africa via Spain; and finally, 13th century as rabolion
from the Arabic world back to the Balkan and Eastern Europe in general. Of course, this misses the
essential question as to the origin of Arabic geomancy. It fails to realise that Hellenistic science
constitutes an absolutely essential link between Hellenic types of divination and Arabic ones. Now
Hellenistic science had its centre, never in Greece but in Alexandria – so that whatever Hellenic
elements towards a Hellenistic proto-geomancy must have crossed the Mediterranean not in 700 CE
but nearly a millennium earlier. Moreover, against all evidence from astrology – I shall come back to
this point – Pennick’s map links Hellenistic and Indian science by an unnecessary detour via Ethiopia
and Madagascar (although some nautical, coastal Indian Ocean connection is likely, as we know since
the rediscovery of the Periplus), and sees the role of North-west Africa in the development of
geomancy as merely passive.
There is also the vignette – a widely circulating and clichéd iconographic theme
– of Archimedes working on a mathematical problem in the dust during the sack
of Syracuse (212 BCE):
‘ “What an ardour for study, think you, possessed Archimedes, who was so absorbed in a
diagram he was drawing in the dust that he was unaware even of the capture of his native
city!” ’54
Some time earlier Archimedes had found in the grain of sand his inspiration for
that amazing mathematical treatise on finity and infinity, Psammitēs arithmos,
Sand Reckoning.55 What is important here is not whether these admiring
accounts written several centuries after the event are biographically correct, but
the fact that they testify to the common use of dust, soil or sand for scholarly
purposes in the Hellenic and Hellenistic period. Not unlikely, sand was also not
only mathematical drafts but also for scribbles of a more occult nature.56
In A. Delatte’s Anecdota Atheniensia, vol. i57 and – from both Delattes – in
their comment on yet another Byzantine geomancy,58 connections were traced
53 E.g. Plutarch, Quomodo adulator ab amico internoscatur, 52, D/7; in: Plutarch’s Moralia: In
sixteen volumes, vol. I, F.C. Babbitt, tr., Loeb Classical Library, no. 197, Cambridge (Mass.): Harvard
University Press/London: Heinemann, p. 283.
54 Cicero, 1971, De finibus, tr. H. Rackham, Loeb Classical Library no. 40, London/
Cambridge(Mass.): Heinemann/Harvard University Press, p. 451: book V 50 (xix). The famous adage
‘Noli turbare circulos meos’ [‘please, do not disturb my circles’], attributed to Archimedes on this
occasion, can only be found in Valerius Maximus, Factorum et dictorum memorabilium libri, 8, 7,
and is probably merely legendary.
55 Cf. Hultsch, 1965, ‘Archimedes’, in: Paulys Realenzyclopädie der classische Altertums-
wissenschaft, ed. G. Wissowa, Stuttgart: Druckenmüller, 3. Halbband, repr. of the 1895 edition, col.
507-539, espec. 515f, 537.
56 Cf. Jesus writing with his finger on the ground ( ‘eis tēs gēs’, using the same word ‘gē’ as is
constitutive of the term ‘geomancy’), in the puzzling, little commented upon passage in John 8: 6-8,
which is often discarded as corrupt, probably because, in ways unwelcome to confirmed Christians, it
seems to confirm Celsus’ allegation that the founder of Christianity was, after all, a magician.
57 Delatte, A., 1927, Anecdota Atheniensia, vol. i.: Textes grecs inédits relatifs à l’histoire des
religions, Bibliothèque de la Faculté de Philosophie et de Lettres de l’Université de Liège, fasc. 36,
Liège: Vailant & Carmanne/Paris: Champion.
58 Delatte, A., & Delatte, E., 1936, ‘Un traité byzantin de géomancie – codex Parisinus 2419’,
Mélanges Franz Cumont, Annuaire de l’Institut de philologie et d’histoire orientales et slaves, t. iv,
Bruxelles, pp. 575-658. Incidentally, in the Byzantine context the word laxeutērion (stone mason’s
chisel, often specified as Pythagoras’) is often used as a synonym for rabolion. I suppose that a
particular graphic notation (cf. the hammer-like example in Tannery 1920) of the full, 16-figure
geomantic theme in a wedge-shaped grid produced, in connection with the precise and sharp nature
imputed to geomancy as a divinatory instrument, the association with a stone-worker’s tool; or was it
with clearly similar forms of Hellenic and Hellenistic divination based on lot-
casting, numerology, onomamancy, and the divinatory manipulation of beans,
grains and wheat. These publications are welcome sources for divinatory
systems as circulating in the Byzantine environment and they do exemplify
some of the basic procedures also found in geomancy; we shall come back to
them below. However, from an analytical and historical perspective it does not
seem unfair to say that the correspondences pointed out by the Delattes are far
too superficial, too general and too commonly known as standard features of
Hellenic and Hellenistic occult sciences59 than that they can simply account for
the emergence of such a highly specific and elaborate system of divinatory
procedures and concepts as geomancy is.
Even less of an analytical or systematic treatment is given to geomancy in T.
Fahd’s monumental study of La divination arabe.60 Among the hundreds of
divinatory methods discussed over hundreds of pages, cilm al-raml occupies
only a minor place. Fahd’s great command of primary sources both in Arabic
and in other languages allows him to trace the earliest attestation61 of a very
the dust that is associated with both? Also cf. Desrousseaux, A.M., 1886, ‘Sur quelques manuscrits
d’Italie’, in: Mélanges d’archéologie et d’histoire de l’École française de Rome, 6: 483-553, who
presents similar Greek texts but apparently without yet realising their significance for the history of
geomancy. No doubt modern scholarship has, since the Delattes, unearthed similar texts which have
escaped my attention so far.
59 Cf. Betz, H.D., 1986, The Greek magical papyri in translation: Including the Demotic spells,
Chicago/London: University of Chicago Press; Bouché-Leclercq, A., 1879-1882, Histoire de la
divination dans l’antiquité, 4 vols., Paris: Leroux; 1975, 2 vols., New York: Arno Press, reprint of the
original edition in four volumes, 1879-1882; related texts have also been published in the volumes of
the Catalogus Codicum Astrologorum Graecorum, ed. F. Cumont, F. Boll et al., vols. 1-12, Bruxelles,
1898-1951.
60 Fahd, T., 1966, La divination arabe: Etudes religieuses sociologiques et folkloriques sur le milieu
natif de l’Islam, Leiden : Brill; also: Fahd, T., 1978, ‘Khatt ’, in: C.E. Bosworth, E. van Donzel, B.
Lewis & C. Pellat, eds., The encyclopaedia of Islam, new edition, volume IV, fasc. 77-78, Leiden:
Brill, pp. 1128-1130.
61 Another apparently early reference to geomancy is given by Morony, M.G., 1984, Iraq after the
Muslim conquest, Princeton (N.J.): Princeton University Press, ch. 13.
‘Conditions in the early seventh century are summarised by Djahiz [this author lived c. 776-
868/9, so rather in the ninth century – WvB], who tells how Muslaylima toured the marketplaces
of Ubulla, Anbar, and Hira, where Arabs and Persians met to buy and sell their goods. In these
places he learned magic, sleight of hand, and the tricks of astrologers, of false prophets, of the
keepers of idol temples, of soothsayers who practice ornithomancy and geomancy [italics added],
of fortunetellers, of magicians and of those who claimed to be inspired by jinn’.
Djahiz’s text (cf. al-Djahiz, Abu Utman ‘Amr b. Bahr, Kitab al-hayawan, 1408 H/ 1988 CE, ed.
A.M. Harun, Cairo, 7 vols; earlier edition 1356-64/ 1938-45; the reference is in vol. IV, p. s369) has
khatt and although this might be read as ‘calligraphy’, the context strongly suggests that this is a
(probably anachronistic) reference to geomantic practice in the 7th century CE; that, however, would
make it a real, and very welcome and revealing, reference to such practice in the 9th century, when
crude form of cilm al-raml to al-A’rabi of the late 8th and early 9th c. CE
(already cited above), to reiterate the claim of a semantic and generic link with
the, admittedly often sandy, desert environment, and to remind us of the
system’s kinship with other cleromantic procedures in Islam (notably,
belomancy and especially ornithomancy – i.e. divination by shooting of marked
arrows, and by the behaviour of birds, respectively). These however in no way
display the complexity of geomancy as a fully-fledged formal system, and
therefore can only have the most general genetic link with geomancy. Fahd’s
stress on al-Zanātī and the latter’s better documented, presumable contemporary
al-Munadjdjim (whose name means ‘astrologer’ in the sense ‘star-gazer’, cf.
nadjm, ‘star’) merely confirms what we already learned from Ibn Khaldun’s
discussion of the topic: the system’s very close links with astrology. Ibn
Khaldun’s position is ambivalent. For although he exempts geomancy from the
severe criticism which he applies to related systems current at his time, and
although he himself consults the system62 so as to elicit from it statements on its
own origin (finding as several others63 have claimed before him in the Arabic
Djahiz wrote. The summing-up of occult practices in itself is telling: such series form a standard
feature of medieval literary, didactic, judicial and theological texts, both in Arabic and in the
European languages, but I have the impression that it is only around the turn of the first millennium
CE that geomancy was included in the list.
62 Of course, like Ibn Khaldun before me in his own explorations into the history of geomancy, I
drove this point to extremes by using both classic geomantic, and Southern African geomancy-
derived, divination methods to produce statements on the origin of cilm al-raml itself. The results were
most entertaining and detailed. They by and large confirmed the historical analysis that was shaping
up in my mind at the time of the experiment and which I have since greatly revised. This supports my
growing conviction (cf. my Intercultural encounters) that, when all is said and done, divination is a
technique intended to facilitate the practitioner’s uninhibited (and occasionally inspired) introspection,
often at normally inaccessible layers of the individual and/or collective consciousness. Thus
divination remains highly subjective, despite the quasi-objective complex procedures it may involve
and which try to persuade the client, and even the practitioner, of the contrary view.
63 E.g. the strongly astrological treatise by al-Zanātī, al-fasl fi usul cilm al-raml, o.c., p. 7, cites as
the conventional chain of traditional authorities: Gabriel, Idrīs, Enoch, Tumtum, Ptolemy, Hamidani,
and cAli cAbd Allah al-Tamimi:
We have been told that the first to conduct divination by drawing lines in the sand [litt.: khatt al-
raml] was Jibril – peace upon him – . Then he taught this science to the prophet Idrīs. The latter
in his turn pased it on to his son Ikhnuakh. From there this science was kept in the shed of secrets
until the time of Tumtum. And then Batlimus added more and divided the paragraphs, elements,
and zodiacal signs. Then Idrīs was the first wise man of geomancy, the second his son ‘N’,
Ptolemy ‘B’, cAbd Allah al Tamimi ‘T’, Hamadani ‘H’, Tamtam ‘T’, cAbd Allah al-Fakir’
No unequivocal information as to time and place of the origin of geomancy can be derived from these
mythical attributions, although there may be a certain Asian (Persian or Indian) suggestion; not only
in the name of Tumtum al-Hindi, but also in the person of Idrīs, who is often identified not only with
Enoch but also with Alexander the Great’s Persian cook; Wensinck, A.J., [year] ‘Idrīs’, in:
Encyclopaedia of Islam. Leiden: Brill.
tradition, that the system was first revealed by the Angel Jibrīl to the Prophet
Idrīs), his basic view is that geomancy is merely adulterated astrology.
To my knowledge, the most serious attempt to confront problems of origin
in the context of Islamic geomancy was made by Klein Franke,64 whose
sixteenth-century CE vintage point (when Ahmad b. ‘Ali Zunbul flourished)
however leads to perspective distortion. He insufficiently appreciates the
mythical, arbitrary and charter-like nature of the legendary authorities
(including Tumtum and Idrīs) cited in the context of geomancy, tries to derive
geographical and historical clues from these conventional embellishments. Of
course, the listing of intellectual forbears is a standard passage in many
medieval geomancies, whether in Arab, Hebrew, Greek, Latin, Provençal or
French. But these are lines of transmission, not stages of creation, and most of
the authorities cited in such chains of transmission are legendary, the fictitious
authors of pseudo-epigraphic works, rather than historical: Idrīs (Enoch) and
Tumtum al-Hindi feature most prominently. Another shortcoming of Klein
Franke’s contribution is that he has little grip on the intricacies of astrology.
Even so he is essentially right in presuming that the most likely context for the
emergence of geomancy was that of the Ikhwan al-Safac.
Finally, mention should be made of my friend and colleague Louis Brenner’s
attempt65 to situate geomancy, and its success especially in West-Africa, against
a sound background of Arabist scholarship, drawing on additional manuscripts,
trying to interpret the significance (an oriental connection?) of such mythical
attributions as those to Idrīs and to the mysterious but frequently cited Tumtum
al-Hindi, but above all adding further relief to the idea first launched by
Becker66 at the beginning of the twentieth century: that the success of Islam in
Africa depended less on deliberate surrender to the One who is Compassionate
and Merciful, than on the converts’ hopes of a greatly increased access to a
superior yet familiar form of magic. Brenner’s is an argument about geomancy
64 Klein Franke, F., 1973, ‘The geomancy of Ahmad b. ‘Ali Zunbul: A study of the Arabic corpus
hermeticus’, Ambix, 20: 26-35. Also cf. Savage-Smith & M.B. Smith, 1980, Islamic geomancy and a
thirteenth-century divinatory device, Studies in Near Eastern Culture and Society, Malibu (Cal.):
Udena, which however is less a contribution to the study of the origin of geomancy, than a description
of an intricate mechanical device – directy related to astronomical and astrological apparatus, from
astrolabes to the zairdja – producing geomantic figures.
65 Brenner, L., 1985, ‘La géomancie pratiquée par des musulmans’, ch. IV, in: idem, Reflections sur
le savoir islamique en Afrique de l’ouest, Bordeaux: Centre d'Étude d’Afrique noire, pp. 78-98.
66 Becker, C.H., 1911, ‘Materialien zur Kenntnis des Islam in Deutsch-Ostafrika’, Der Islam, 1: 1-
48; Becker, C.H., 1913, ‘Neue Literatur zur Geschichte Afrikas’, Der Islam, 4: 303-312.
in Africa and although he reminds us of the mythical traditions concerning the

ultimate, and ultimately divine, origin of cilm al-raml, he does not tangibly add
to our understanding of its origin and earliest history in human, specifically
Arab hands.
This is the question which I seek to address in my current research and
specifically in the present paper. But how to arrive at a method by which it can
be answered?
4. IN SEARCH OF A METHOD
Struggling for a feasible method by which to approach our central question as to

geomancy’s origin, I was tempted, as basically an anthropologist by training, to
look at symbolic products as being rather specific to distinct societies and
cultures, each identified in space and time. Therefore, my naïve initial strategy
aimed at identifying the precise, and presumably unique, conceptual, spiritual
and intellectual milieu in which geomancy as a complex package – but still
retaining the fundamental features of that milieu – would have been conceived
and launched for the first time. In an earlier stage of the project, reading through
whatever materials (mostly of a secondary nature) on divination and related
topics in the Ancient Near East, Graeco-Roman Antiquity, and the early Middle
Ages came my way, I would be constantly shaken from one infatuation to the
next. One day I would know for sure that geomancy with its basic dichotomy
between even and odd and its insistence on a foursome (the four vertically
arranged lines) must be an Iranian invention: reflecting the four primary gods
one of which, Zurvan, produced the central antagonists Ahura-Mazda and
Ahriman. The next day this euphoria would be supplanted by another, seeking
the underlying dichotomy in the Chinese concepts of Yin and Yang, the
foursome as an adulteration of the five Chinese elements or the six lines of the
duplicated trigrams that constitute the I Ching formalism, and interpreting the
geomantic kinship idiom of the construction of daughters and nephews out of
mothers, as a transformation of Taoist cyclical schemes of the transmutation of
the five elements, and relying67 on the Silk Road or the Indian Ocean trade to
67 Not unreasonably; cf. Needham, J., with Wing Ling, 1961, Science and civilization in China, vol.
1. Introductory orientations, Cambridge: Cambridge University Press; first edition 1954, which
take all to the West. One week later, I would give all this up for an exploration
of the extent to which the Egyptian world-view, dwelling on the sacred soil as
fertilised by the sacred river, with in the background the perennial struggle
between riverside Osiris and desert-based Set, as well as the alternation between
favourable and unfavourable days in the calendar, could account for the
geomantic repertoire, either directly in the form of ancient Egyptian religion, or,
even more attractive, the latter as it was filtered through Hellenistic, Alexandria-
based magic, mystery cults, Hermetism etc. Only a few days later to realise that
the dichotomy between even and odd, male and female numbers was already
Pythagorean, and that Pythagoras himself (although alleged to have been
unfavourably disposed towards divination and beans) reputedly exhorted his
followers to swear by the tetraktus,
•
••
•••
••••
the holy foursome whose four lines of dots constitute a likely prototype for
geomancy if ever there was one...! Or should I explore, for an explanation of the
foursomes so prominent in geomancy, the doctrine of the four elements which
Empedocles, Pythagoras’ near-contemporary, introduced in Western thought but
whose antecedents ramify far away from Graecia Magna (Graecised Southern
Italy) into Ancient Egyptian, African and East Asian cosmologies? Or was the
foursome, after all, primarily an expression of the – from a geomantic
perspective so aptly named –tetragrammaton, , with a fundamental
dichotomy at least hinted at by the division between the Urim and the Tummim
as mysterious oracular instruments, or the Jachin and the Boaz (the two pillars
flanking the entrance to the temple), so that the origin of geomancy might
primarily be sought in early medieval Jewish circles?... Did not the puzzling
discussion, in the Babylonian Talmud,68 of the odd and even number of cups of
wine (with emphasis on four) to be taken or avoided, suggest as much? Was
such a suggestion not reinforced by the Jewish connotations of the Maroccan
Zanāta tribe?
discusses in detail the interaction between China and the West in the course of human history; also cf.
Needham, J., with Wing Ling, 1956, Science and civilization in China, vol. 2. History of scientific
thought, Cambridge: Cambridge University Press, for an excellent introduction to the Chinese
trigrams and their role in the Taoist worldview.
68 Goldschmidt, L., ed., 1906-1935, Die Babylonische Talmud, Leipzig/Berlin: Jüdischer Verlag;
the reference is: Pesahim, X, i, Fol. 108a ff.
Or should my main attention rather follow hints from Ancient Coptic magic,
as to the magical significance of the act (reminiscent of that of the slaying
pharaoh – for three millennia a dominant iconographic theme in Ancient Egypt)
of hitting the ground with a stick, so central as a random generator (the number
of indentures made in one stroke is assessed in terms of even or uneven, and this
yields one of the horizontal lines for the geomantic figure) in many attested
forms of Islamic geomancy? For if earth is the opposite of heaven, then
geomancy is divination not by the stars but by the earth, while the earth is the
typical place where magicians, by hitting the very ground with a stick or a wand
(e.g. Exodus 7: 8-12 on Aaron’s rod, and Exodus 17 on water from the rock),
assert their autonomous right to divine status and power, and by implication
their kinship with Satan, as in the following Coptic formula for love magic (first
millennium CE):
‘...Shurin, Shuran, Shutaban, Shutaben, Ibonese, Sharsaben,... Satan the devil, who beat
with his staff upon the earth saying: ‘‘I am a god also’’...’69
Here the magician is speaking, in all likelihood, not only to the earth but also on
behalf of the earth, thus asserting the latter’s powers as the Great Mother, which
despite repeated attempts in many later religions was never completely deprived
of its divine nature and never completely subjugated to ethereal gods who
literally represent the more advanced levels of sublimation.
Having started to think about the earth as a fundamental symbol the
emphasis on the number four, in geomancy, perhaps no more than a celebration
of the earth (after which it got its name, both in Greek and – although more
specifically limited to sand – in Arabic), and must we not situate the Islamic
invention against the entire background of chthonic cults throughout the Old
World? This opens up a veritable Pandora’s box. We are dealing here with one
of the most powerful symbols that Man has developed. Its antiquity is suggested
not only by its ubiquity,70 but also by the way in which the earth appears as a
69 Papyrus Berlin 8320 (Koptische Texte), as quoted in: de Jong, K.H.E., 1921, De magie bij de
Grieken en Romeinen, Haarlem: Bohn, pp. 238f; further brief reference to this text in: Meyer, M., &
Smith, R., 1994, Ancient Christian magic: Coptic texts of ritual power, San Francisco: Harper Collins,
p. 367, n. 75, l. 18, cf. p. 161; also cf. Isaiah 14: 13-14; Ezekiel 28: 2.
70 E.g. it is a major aspect of religions throughout Africa (cf. Schoffeleers, J.M., 1979, ed.,
Guardians of the Land: Essays on African territorial cults, Gwelo: Mambo Press) and around the
Mediterranean (van Binsbergen, W.M.J., forthcoming (a), Religion and social organisation in north-
fundamental, independent moral category in myths, oaths and rituals from

Mesopotamia, Greek and Roman Antiquity, West and East Africa, while even
the pacifist marabouts of North West Africa, administrators of collective oaths
by saints associated with local shrines, may also be considered land priests in
Islamic garb.71 Many layers of reference are piled up here, making for a multi-
referential coding whose co-ordinates in space and time are typically complex
and confused, but together elucidating the implied symbolism and the possible
original inspiration of geomancy in particular, and more implicitly of board-
games. There is the maternal (and psychoanalytically oedipal), nurturative,
agriculture-related symbolism of unfathomable and ungraspable earth as the
source of life. But there is also the symbolism of fragmented and tangible earth,
dust, mud, dirt, pebbles, as the lowly (psychoanalytically anal) origins of man
and of life in general. There is the combination of these two themes in the ‘black
and red’, the fertile alluvial soil (symbolised by Osiris) as against the barren
desert (symbolised by Set), which was how Ancient Egyptians conceptualised
their country – and an inspiration (besides the moon’s phases and the
hemerology of lucky and unlucky days) for the binary symbolism underlying
e.g. geomancy. There is earth with its four cardinal directions (essentially
derived from the specific symmetry structure of the human body: left and right,
back and front, which makes for square representations (reproduced in many
square game-boards), rectangular grids as subdivisions, and in general a
preponderance of the figure 4 or higher powers of 2.72 There is earth as the time-
less repository of the dead, as the underworld or the gate to the underworld,
hence the alternative source of power, knowledge, legitimacy of political and
ritual office...
western Tunisia, Volume I:Kinship, spatiality, and segmentation, Volume II: Cults of the land, and
Islam; and references cited there.
71 Cf. Gellner, E.A., 1969, Saints of the Atlas, London: Weidenfeld & Nicholson.; van Binsbergen
& Wiggermann, 2000; Fontenrose, J., 1980, Python: A study of Delphic myth and its origins,
Berkeley etc.: University of California Press; paperback edition, reprint of the 1959 first edition;
Fortes, M., 1945, The Dynamics of Clanship among the Tallensi, London: Oxford University Press for
International African Institute; Fortes, M., 1949, The Web of Kinship among the Tallensi, London:
Oxford University Press for International African Institute; Simonse, S., 1992, Kings of disaster:
Dualism, centralism and the scapegoat king in southeastern Sudan, Leiden etc.: Brill.; van Binsbergen,
W.M.J., forthcoming (a).
72 Pennick, N., 1992, Secret games of the gods: Ancient ritual systems in board-games, York Beach
(Maine): Weiser, reprint of the 1989 ed; first published London: Century, 1988; this is an utterly
unscholarly text which yet makes many good points.
Was all this merely New Age, Coca Cola-ized scholarship, the flotsam of avid
but amateur reading, or did it contain shimmerings of true insight? What could
be the context where so many, apparently similar and converging, themes could
be absorbed, digested, mediated? In the process, how could I be sure that I was
making any progress, or even, that I was comparing the comparable? Was it
enough to have – somewhere in a Near Eastern or Mediterranean cosmology or
religion – some disconnected geomantic traits, such as a fundamental dichotomy
and a foursome, to jump to historical conclusions, or was it necessary that the
milieu displaying such features had used them specifically for the construction
of, specifically, a divination system? In the absence of a rich corpus of transitory
and preparatory forms of geomancy, such ideas (and the predictable alternatives,
toying with possible contributions from Marconites, Manichaeans, Mandaeans,
Sabaeans, Harranites proper, Marcionism, Talmudic and other Jewish literature,
the extensive Hellenic and Hellenistic body of magical and divinatory texts and
procedures, real and pseudo-epigraphic references to the Pythagoreans and Neo-
Pythagoreans, etc. etc.), brought home the fact that not only my knowledge and
skills were utterly deficient to cut my way through this primal forest of
possibilities, but that the real problem was that of method. The hit-and-run
method that had first presented itself, had to be replaced by something rather
more systematic and controllable.
4.1. A heuristic typology of geomancy
The only way out of this confusion seems to consist in the adoption of a method
which rigorously defines a typology of geomancy as a complex package, and
which seeks not so much for the unique context in which this total package may
have emerged, but for the composite history of the constituent items in the
typology. This means rephrasing our central research question as follows:
where, then, does one find a divinatory and cosmological tradition containing all
the following characteristics:
a. A focus (conceptual, often also material) on sand (Arab. raml) or earth (geomantia,
astrologia terrestris)
b. The use of multiple unmarked elements.
c. Simple procedures of casting or tracing, and subsequently sorting or eliminating these

elements
d. A fundamental dichotomy (cf. two-valued tablets; standard lines of one or two dots)
e. Mathematical representation of that dichotomy in terms of ‘odd’ and ‘even’
f. A propensity for the number 4 (cf. four tablets; four geomantic lines forming a
tetragram; therefore, 4 is the power to which 2 (as derived from the fundamental
dichotomy underlying the geomantic system, i.e. one dots or two) has to be raised in
order to create 16 configurations; also cf. the number of four tablets, to produce
essentially the same result73
g. A propensity for the numbers 16, 64 and 256, clearly resulting from the repeated
duplication of 2 or 4 as basic numbers in geomancy
h. A graphical representational system featuring dots, which may be combined or
connected so as to form lines
i. A graphical representational system featuring tetragrams, e.g. symbols consisting of
four horizontal lines, each line consisting of one or two dots (or a line)
j. A basic kinship idiom (with its classification in terms of age, gender and biological
reproduction) serving to derive certain configurations from others, in the production
of a full geomantic theme or horoscope
k. The distinction between a primary level determined by chance (the ‘mothers’ in
geomancy), and secondary levels (‘daughters’, ‘nephews’, ‘witnesses’ and ‘judges’)
produced by a simple derivative calculation and not involving chance inputs in their
own right.
l. A catalogue of the names of all mathematically possible configurations, with their
associated oracular interpretations.
m. That catalogue is consistently built so as to accommodate various dimensions between
which the interpretation may toggle at libitum.
n. Extensive use of astrological notions.
Table 2. An exhaustive descriptive definition of Islamic geomancy
4.2. Getting focused as to time, place and topic
Formulated with such detail, we now have a finely-meshed net in which to catch
whatever historical evidence on the origin of geomancy.74 Of course, these
items need not be present simultaneous but may turn out to be gradual
73 Considering the ubiquitous emphasis on ‘4’ as the earth’s number throughout Old World
cosmologies, it is unlikely that the prominence of foursomes in geomancy is simply based on a
duplication of two and nothing more.
74 The typological notions developed above now enable us to formulate the research question with
such precision that superficial parallels from Graeco-Roman Antiquity (e.g. Fortuna’s oracle at
Praeneste, Italy; Heracles’ shrine at Bura, on the Peloponesus; or Varro’s ‘geomantia’ (see above), can
be ruled out as immediat, let alone unique, ancestors of Islamic geomancy. The actual geomanctic
package that materialsed by the end of the first millennium CE in Abassid Mesopotamia is far more
specific and complex.
accretions from different sources. It is only the combination which makes for a
specificity identifiable as geomantic. Some elements may in themselves be so
general as to be almost universal and even biologically anchored, such as the
fundamental dichotomy which is found in all cultures as a recognition of gender
differences and of left/right symmetry (yet only a few cultures have made
dualism a really decisive feature of their world-view), the binary opposition
which informs much or human thought anyway, the ubiquitous Old World
feature of the four directions (which may well result from the combined
left/right symmetry and front/back asymmetry of the human body).
My intention is to try and explore the history of geomancy as a composite of
the histories of each of these fourteen items before the high-tide of Arabic
geomancy in the beginning of the second millennium CE. This is not to claim
that these items have developed in total independence from one another, but
each can be traced – to a certain extent – in its ramifications across linguistic,
political, religious and cosmological boundaries far further back then that we
have attestations of geomancy as an integrated, finished product.
Depending on the audience, some of my fourteen items would be more
interesting than others to concentrate on. One would like to see, e.g., an
argument about the emergence of formal compendia, probably emerging from
Old Babylonian or perhaps even earlier prototypes, and soon tending to
standardised formats, internal consistency, and the addition of hypothetical
cases meant to complete and balance the records of actual omens taken in
concrete situations.75 If ‘books of fate’ form the standard format for later
geomancies, how much did such texts, from Astrampsychus and the Sortes
Ovidianae or Homericae to medieval Arabic and Hebrew versions, owe to the
Ancient Mesopotamian prototypes of omen compendia? How did they become
streamlined, and amenable to numerological, onomamantic and cleromantic use,
through the imposition of a numerical or graphical classification upon the
entries – so that they could be addressed by the outcomes of chance processes
producing numbers (through dice) or visual patterns. And how and when did
they become saturated, like almost all divinatory and occult sciences in Late
Antiquity,76 with astrology? Alternatively, one would like to see the range of
75 Jeyes, U., 1989, Old Babylonian extispicy: Omen texts in the British Museum, Istanbul:
Nederlands Historisch-Archaeologisch Instituut.
76 On this problem, cf. Bouché-Leclercq, A., 1879-1882, Histoire de la divination dans l’antiquité,
4 vols., Paris: Leroux.; 1975, 2 vols., New York: Arno Press, reprint of the original edition in four
cleromantic and otherwise stochastic procedures for divinatory and ludic

purposes, even if cleromantic divination is scarcely attested in Mesopotamia in
the earlier millennia, and was never considered worthy, by those who practised
it, of a formal, written compendium spelling out its interpretational catalogue.
However, given the astrological emphasis in the Arabic and later
geomancies, it is to be expected that considerable insight can be gained by an
exploration of the extent to which the origin of geomancy can be understood
from an astrological perspective. I shall devote the remainder of this paper to an
exploration of the origin of the sixteen geomantic names, since this is the one
item77 in my typology of geomancy where the astrological dimension can be
studied in some depth.
Of course there is a danger in taking apart a working intellectual product like
the geomantic divination system: although the various constituent parts of
course each have their history, and perhaps a rather heterogeneous history (e.g.
hitting the sand with a stick appears to belong to a very different intellectual
idiom and history than interpreting the configurations, once found, in an
astrological format of houses and planetary correspondences), they only work
after they have been assembled. In evolutionary biology this kind of problem is
often illustrated by reference to the very complex eye of the vertebrates:78
admittedly one can imagine some process of evolution to produce, little step by
little step, some tissue modification ultimately leading to that complex optical
organ, but how is it possible that such blind steps – in the double sense – can in
the end suddenly produce something which can actually see? Tracing the history
of geomancy cannot just be reduced to the history of its components: if possible,
we would like to capture the magical moment when the ingredients were put
together and a new, competitively powerful, divinatory tool was created. And, if
at all possible, we would like to understand what suddenly brought about this
creative transformation of material which, in all likelihood, had been available
in discrete form for centuries: astrology, numerological operations reminiscent
of the generations of daughters and nephews, arithmetical and mystical juggling
with even and odd, etc. But without the one felicitous find that shows us
volumes, 1879-1882; and the more recent but patchy: Tester, S.J., 1989, A history of western
astrology, New York: Ballantine, repr. of 1987 first edition. Also cf. van Binsbergen & Wiggermann,
o.c.
77 See table 2 above: ‘l. A catalogue of the names of all mathematically possible configurations,
with their associated oracular interpretations’.
78 Cf. Vroon, P., 1992, Wolfsklem: De evolutie van het menselijk gedrag, Baarn: Ambo.
geomancy in statu nascendi (perhaps we shall encounter just that towards the
end of my argument), we have to contend ourselves with circumstantial
evidence – and not pass a final verdict.
In general, what we are looking for is the complex history of origin of a
complex, internally heterogeneous system of magical representations and
practices – the result, no doubt, of an amazing bricolage by some school of
occultism, using elements of great specificity and heterogeneity.
If the geomantic system can be truly said to have come into the world fully-
fledged, this simply means that this particular concerted combination of
elements was never seen before. This in itself is a historical fact of considerable
significance. For here we are not charting some illiterate cultural domain in the
shady beginnings of history – like in the reconstruction of the pre-colonial
history of African religious systems including divination – but a literate field of
scholarly production, whose intellectual links with the philosophical and
magical traditions of Ancient Egypt and Mesopotamia, Graeco-Roman
Antiquity, Judaism, early Christianity, Iran and India are in principle open to
detailed investigation – even if the relevant texts have often gone lost and many
of those that survive still await formal publication. If in this context a composite
bricolage of exiting and newly-invented elements produces something
unprecedented, like geomancy, this strongly suggests that we can put the date
for the first formation of the geomantic system not too long before the first
attestation and the rapid spread of the system in the early second millennium
CE. In other words, geomancy emerged in the late first millennium AD.
This also provides a clue as to where to look for the location in which the
system was first formulated. This must of necessity be sought among the few
main centres of the preservation and innovation of scholarly knowledge at the
time – which for the date mentioned almost inevitably points to Mesopotamia
under Abassid rule, with the academies of Mosul, Baghdad and Basra.
The composite nature of the geomantic system also means that the obvious
acceptable way of tracing its early history is to break down the system into its
arguable constituent components, and trace the history of each of these
components. These histories may be very different, and point to very different
geographical and intellectual milieus in very periods of time. But then again,
magical systems display traditional, accumulative, conservative inward-looking
tendencies just as much as they are prone to innovation and experiment. The
documented history of magic, divination, cosmology, and philosophy in general
from Sumer and pre-dynastic Egypt right through to Neo-Platonic, Sassanid,

Coptic and Arabian times on the eve of Islam, presents the picture of basically
one widely ramifying and internally contradictory system in motion, not of as
many discrete systems as cultural and political history chooses to define
separate cultural, ethnic and statal units. Therefore it is likely that some of the
components meeting in geomancy have already been brought together and
amalgamated earlier on in the history of magic. One cannot trace the history of
geomancy without tracing the entire history of magic, and that, clearly, is way
beyond my scope in the present paper, and beyond my competence in general.
When appreciating the composite nature of geomancy and dissolving its
history in the part histories of its components, much depends on making the
right kind of distinctions when identifying the various components, and giving
each its proper weight. This can only be done – and then of course only partially
and conjecturally – on the basis of a broad, historically involved, comparative
approach to magic, astrology, cosmology and philosophy in general, in the
centuries preceding the emergence of the system and in the cultural
environments known to have influenced the presumable intellectual milieu
where geomancy was created. In some cases merely the possibility of
intellectual contact is established, while the fact of borrowing remains to be
argued; this for in stance in the case for the contributions from China and North
West Africa79 to the geomantic system. In other cases continuity is so
unmistakable that the real question is whether there has been any continuity but
whether that continuity allowed for any change and variation; this is for instance
the case with regard to the dominant astrological component of the geomantic
system, which may not be totally main-stream Ptolemy-derived, still is very
much along the accepted lines of astrological orthodoxy as established since late
Antiquity.
In fact, astrology does occupy a central position in this kind of argument,
since this particular, astronomically-based divination technique, which –
whatever its history in Egypt – up to the first millennium BCE was merely
ancillary to other Mesopotamian divination techniques especially extispicy,
from the middle of that millennium imposed itself on all other magical arts, a
79 China is touched upon at various points in the present paper. The North West African link is
argued briefly in: van Binsbergen, W.M.J., ‘Rethinking’, o.c.
process that was to be continued in Hellenistic and Imperial times.80 Therefore

we may expect that by the end of Late Antiquity, when the scene is set for the
emergence of geomancy, astrology was not just one ingredient among many –
much more than that, it henceforth was to be and remain the dominant
component of any magical art in the Near Eastern and Western tradition.
Although Hebrew contributions to geomancy as a finished products, and
although there are Hebrew geomancies, there is no reason to assume that
geomancy was an invention of medieval Hebrew scholars.81 I have no evidence
yet of Iranian, Pahlavi etc. versions of geomancy. Inspection of such Indian
sources as may exist is unnecessary since David Pingree, the greatest living
authority on Indian astrology and related sciences, tells us that geomancy in
India is an Arabic import, which cannot even boast a local Indian name but has
retained the word raml:
‘XIII AND XIV. RAMALA AND TAJIKA

Finally, Yavana is an authority on ramalashastra (from the Arabic raml, ‘geomancy’)
and tajikashastra (the science of Tajiks); here, of course, Yavana means ‘foreigner’
rather than ‘Greek’. Several ramal texts are ascribed to Yavana.’82
Thus for geomancy we can rule out a history parallel to that of chess: from an
origin in India, the latter invention spread to Iran, was there transmitted to the
Arabic culture of the early Middle Ages, and thus reached Southern Europe.83
80 Cf. W.M.J. van Binsbergen & F.A.M. Wiggermann, 2000, ‘Magic in history: A theoretical
perspective and its application to Ancient Mesopotamia’, in: T. Abusch & K. van der Toorn, eds.,
Magic in the Ancient Near East, Groningen: Styx; and extensive references cited there.
81 Medieval Hebrew geomancies are referred to by Steinschneider 1864, 1877 o.c., and in:
Trachtenberg, J., 1939, Jewish magic and superstition: A study in folk religion, New York ; my
colleague at NIAS, Shaul Shaked, who catalogued the famous texts from the Cairo Geniza (cf.
Shaked, S., 1964, A tentative bibligraphy of Geniza documents, Paris/ The Hague: Mouton) tells me
that he encountered several geomantic fragments among this material, in addition to a very extensive
geomancy claiming to be composed by Abraham ibn Ezra (c. 1089-1164 CE). However, as we
ascertained in a provisional reading (for which I am greatly indebted to Professor Shaked) this
geomancy is little original and scarcely deviates from the pattern established in Arabic texts a century
or longer earlier. Steinschneider still considered the attribution of a geomancy to ibn Ezra pseudo-
epigraphical, but in the face of this Geniza text there is little reason to doubt the author’s identity,
even though it was a later copy. When ibn Ezra wrote the great diffusion of geomancies had already
started, and whatever his great qualities in other domains, here he did not lead, but follow.
82 Cf. Pingree, D., 1978, The Yavanajåtaka of Sphujidhvaja, Harvard Oriental Series 48, 2 vols.,
Cambridge (Mass.)/London: Harvard University Press, i 38. In corroboration, Pingree mentions six
ramal manuscripts of which three are dated as to the time when they were copied: the oldest in 1662
the most recent in 1837. Also cf. Steinschneider, M., 1877, ‘Die Skidy [sic] oder geomantischen
Figuren’, Zeitschrift der deutschen Morgenländischen Gesellschaft ( Leipzig), 31: 762-765
All the various clues which I gradually detected, can be accommodated in

the eclectic and intellectually eager environment of the Ikhwan al-Safac, where
the Hellenistic and Imperial astrological corpus including Ptolemy was received
by the middle of the tenth century CE, while substantial Indian influences (in
themselves highly indebted to the same Graeco-Roman sources)84 were on their
way. Here, due to the Indian Ocean trade which actually brought Chinese ships
to the Persian Gulf, knowledge on China was available in sufficient detail so as
to produce, already in the 9th century CE,85 a fairly factual travelling guide. The
persecution which soon was the fate of the Brethren makes it very plausible that
specific traces of their presumably vital contribution to the origin of geomancy
were effectively concealed or destroyed, especially since within a century
geomancy rose to be the dominant divination technique throughout the world of
Islam – from which it was translated and emulated, in many versions, into
Persian, Hebrew, Greek and Latin. Such a postulated concealment could very
well have produced the remarkable effect of geomancy emerging suddenly and
without documented preparation, fully-fledged, like Athena sallying forth in full
armour from her father’s skull.
Arabist scholarship has considerable experience in tracing such attributions
on the basis of mere circumstantial evidence, e.g. in Kraus’s brilliant analysis of
the DJabir corpus,86 or Halm’s approach to the Umm al-Kitab.87 Further work
along this line, for which I lack the specific competence, will greatly help to
bring in focus the creative moment when Islamic geomancy was actually
formulated. Meanwhile, we can only appreciate the originality of the actual
creation once we have sorted out the various building materials that went into it.
And it is here that the present paper seeks to make a very specific substantive
83 Murray, H.J.R., 1913, A history of chess, Oxford: Oxford University Press; Trend, J.B., 1947,
‘Spain and Portugal’, in T. Arnold & A. Guillaume, eds., The legacy of Islam, Oxford: Oxford
University Press, first ed. 1931, p. 1-39, on chess p. 32f.
84 This indebtedness has especially been established by Pingree, o.c.
85 Ferrand, G, ed. & tr., 1922, Voyage du marchand arabe Sulayman en Indie et en Chine, Paris:
Brossard; Sauvaget, J., ed., 1948, Relation de la Chine et de l’Inde: ’Ahbar as-sin wa -hind: rédigée
en 851, Paris: Les Belles Lettres.
86 Kraus, P., 1935, Jabir ibn Hayyan: Contribution à l’histoire des idées scientifiques dans l’Islam,
vol. 1, Cairo; vol. 2. Jabir et la science grecque, Paris: Belles Lettres; photomechanical reprint of
earlier edition, 1942.
87 Halm, H., 1982, Die islamische Gnosis: Die extreme Schia und die ‘Alawieten, Zürich/München:
Artemis.
contribution, beyond the comparative and methodological points made so far in

the course of my argument.
5. THE GREATEST PUZZLE OF ALL: THE ORIGIN OF THE

SIXTEEN GEOMANTIC NAMES, AND THE CASE FOR AN
ASTROLOGICAL SOLUTION ON THIS POINT
Since the whole geomantic exercise revolves around identifying which of the
sixteen configurations apply in a particular context, it is clear that the names and
attributions of each of these configurations should occupy a strategic position in
any attempt to explore the history of geomancy. Now the amazing thing is that
all modern scholars writing on geomancy have summed up the names of the
sixteen configurations in Arabic, Latin, Greek, various West African or
Malagasy languages etc. but that – to my knowledge88 – the meaning of these
puzzling terms in their puzzling combination has never been investigated from a
historical or systematic point of view. Assuming, apparently, that these terms
were arbitrary and meaningless anyway, the analysts have contented themselves
with offering literal translations, instead of coherent systematic interpretations.
Tannery’s tabulation of the names of the sixteen geomantic configurations in
one standard Arabic version,89 two Greek manuscripts versions, a medieval
Provençal version and three Latin versions, may serve as a good introduction to
the puzzles geomantic nomenclature represents, and to some possible – if far
from equiprobable! – solutions. For this purpose, let us first inspect the semantic
referents of the names given to the sixteen standard geomantic figures in Arabic.
88 I have so far seen most of the extensive and often obscure scholarly literature on geomancy, but
not yet, e.g., Blier, S., 1983, ‘African geomancy: Art and the idiom of knowledge revelation’, paper
presented at the African Studies Association meeting, Boston.
89 Which corresponds with al-Zanātī, o.c.
5.1. Semantic heterogeneity in the Arabic geomantic nomenclature
Within each series (column) in Table 3, the names appear to come in clusters, as
far as their semantic connotations are concerned. For the Arabic names (also cf.
Table 1) we can distinguish90
• human physical features indicative of gender and age (‘beard’; ‘scarce beard’91 ;
‘with smooth cheeks’)
• a widespread human artefact symbolic of group or religious identity (‘flag’)
• emotions (‘joy’)
• colours (‘white’; ‘red’);
• abstract conditions describing logical relationships between abstract elements
(‘link’, ‘reversion’);
• abstract conditions describing social relationships between people or groups
(‘meeting’; ‘victory when going out’; ‘victory when coming in’, ‘accord’)
• abstract conditions describing legal or pragmatic relationships between animate
subject and inanimate or animate object (‘taking when coming in’; ‘taking when
going out’; ‘agreement’; ‘link’; perhaps again: ‘meeting’)
• concepts marking either boundaries in space (‘threshold when coming in’,
‘threshold when going out’, ‘internal/external’) or the connection between points
in space (‘road’)
Here I would like to invoke an argument which over the years I have used
time and again to make history out of apparently history-less, synchronic
distributional cultural phenomena:92 cultural phenomena, when passing through
the hands of symbolic innovators (like the creators of religious cults, divination
systems etc.), tend to a certain unity and internal systematics; if we find that the
internal structure of a symbolic system displays very heterogeneous elements
which have scarcely accommodated to one another within an overall framework,
we have there indications not just of synchronic social cleavages (between
genders, classes, different residential and productive section of the population
etc.) but also of a long and diverse history, which the assessed heterogeneity
90 I largely limit the discussion to the Arabic terms, although a perusal of the translations in various
European languages (columns 5 to 8) could be argued to reveal semantic implications which at the
time of the transmission of the geomantic corpus to Europe existed also in the Arabic corpus.
91 I do not think that Tannery’s addition ‘growing up with’ or even ‘having grown up with’ is
strictly implied in the Arabic terms.
92 van Binsbergen, W.M.J., 1981, Religious change in Zambia, London/Boston: Kegan Paul
International; van Binsbergen, W.M.J., 1992, Tears of Rain: Ethnicity and history in central western
Zambia, London/Boston: Kegan Paul International.
Table 3. The geomantic nomenclature according to Arabic, Greek and Latin

sources
(from: Tannery 1920, o.c.; for notes see next page).
invites us to explore and for which it suggests specific clues and leverage. If
thus a mere sixteen configurations are described in terms of at least six different
clusters of concepts, there can hardly be an overwhelming suggestion of a
systematically produced, coherent semantic domain, in the Arabic nomenclature.
Tannery based his analysis in part on the Greek manuscript 2424. This
however is not going to make an independent contribution to our understanding,
since the Greek ‘words’ contained there are merely transliterations of the Arabic
terms, making up meaningless sounds in Greek; e.g. Greek nakhoulkhat, istima,
tarikh, which follow the Arabic to the letter.
This cannot be said for the Greek manuscript 2419, whose terms all belong
to one consistent semantic field, as I shall presently demonstrate
5.2. The astrological contents of the nomenclature in the Greek

manuscript 2419
In an attempt to establish the technical astrological nature of the geomantic

nomenclature, let me concentrate on just one column, that of the Greek
manuscript 2419. In fact, it is the only set of names of configurations, albeit
incomplete, to display semantic consistency. A detailed discussion will
demonstrate that here, by contrast with all other versions in Tannery’s table, the
nomenclature does not consist of a multitude of different, mutually unrelated
clusters which each produce their own independent, complex semantic space,
but that they all spring from one principle: astrology, more specifically the
doctrine of planetary dignities and planetary interrelations in general.
The astrological interpretation of the standard geomantic nomenclature turns
out to be exhaustive, as far as the Greek manuscript 2419 is concerned.
@
) Table 3: Originally Tannery listed in his second column, under ‘nom arabe’, the names of the
configurations in Arab script, and in his fourth column the translitteration in European script inspired
by French orthographic usage. Since it is difficult to combine Arabic and European script here, I have
combined Tanery’s two column’s here, but following a different transliteration system.
±
) Table 3: For reasons given in the main text I leave out Tannery’s fifth column, describing the
names of the configuration according to Greek manuscript 2424.
#
) Table 3: I leave out Tannery’s right-hand column, which only contains six entries largely identical
to the Hugo Sanctalliensis column.
Table 4. The astrological contents of the geomantic nomenclature in the Greek

manuscript 2419
notes continued on next page
*
) Table 4: This interpretation still needs to be underpinned with textual evidence from the medieval
Arabic astrological literature.
It is clear that the references, in the geomantic nomenclature, to thresholds

and lintels constitute technical terms of an astronomical/astrological nature.
They cannot refer to everyday concepts: the literally Janus-like nature of the
ordinary threshold of a room does not allow for the existence of two thresholds,
one for going in and one for going out. Pingree93 calls our attention to an Indian
linear planetary theory which goes back to a Babylonian prototype and
**) Table 4: This interpretation still needs to be underpinned with textual evidence from the medieval
Arabic astrological literature.
°) Table 4: Although there is no Greek term and therefore no astrological interpretation for this
configuration in the present manuscript, all the other manuscripts including the Arabic one list here
the term for an accumulation of several planets in one sign or house.
°°) Table 4: In this sense Ptolemy uses thēlukos as a technical term in Tetrabiblos 20, 33.
#
) Table 4: Liddell, H.G., & R. Scott, 1968, A Greek-English lexicon, ed. H.S. Jones with R.
McKenzie, with a supplement, Oxford: Clarendon, reprint of the 1940 9th ed. Dorotheus Sidonius’
Carmen Astrologicum V 27, 1-13 applies the same notion specifically to the position of the Moon, and
reserves the idiom of imprisonment for this specific case. Cf. Dorotheus Sidonius, 1976, Dorothei
Sidonii Carmen Astrologicum: Interpretationem Arabicam in lingvam Anglicam versam vna cvm
Dorothei fragmentis et Graecis et Latinis, ed. Pingree, D., Leipzig: Teubner: 264f.
&
) Table 4: Cf. the astrological terms terms ‘Cazimi position’ and ‘combust’ for the planet Mercury
(17 seconds of arc, and 5°, respectively, from the sun).
%
) Table 4: There is an extensive astrological doctrine about the colour of planets (e.g. Bouché-
Leclercq o.c.: 313-315; Pingree, D., 1978, Yavanajåtaka, o.c., ii: 248f): ‘White’ (leukos) in general
refers to Jupiter although it is also used for Venus, leukotaton (Plato, Republic), leukē (Teucer,
=Valens and ‘Rhetorius’), ksanta (Ptolemy), alba (George of Antioch); and ‘red’ normally refers to
Mars (although it may also be seen to refer to the Sun or Venus (in the dubious ‘Palchus’), while
George of Antioch uses it for the Sun at dawn. If the other geomantic names in the table could be
construed to refer to individual planets rather than to relations between planets, I would interpret the
colour references in this light instead of looking for interplanetary clues. However, this is a rather fine
point, given the emphasis on planetary correspondences (i.e. links with individual planets) of the
geomantic figures. The important point is that there is a choice of possibilities for attributing a specific
astrological meaning to the term leukotēs.
$
) Table 4: This point needs further evidence.
@
) Table 4: Although there is no Greek term and therefore no astrological interpretation for this
configuration in the present manuscript, all the other manuscripts including the Arabic one list in this
row the standard term for conjunction (coniunctio, ijtima‘), the 0° aspect.
§
) Table 4: Although there is no Greek term and therefore no astrological interpretation for this
configuration in the present manuscript, the ‘burned path’ is very much an astrological concept; cf.
Pingree’s rendering of Dorotheus Sidonius’ Carmen Astrologicum V 5, 8 (Pingree 1976 o.c.):
‘If the Moon in its motion is in the path which the learned call ‘‘the burned path’’ (the burned
path is the middle of the equator, which is Libra and Scorpio) and if the Moon is in the last
degrees of a sign, then it is according to this in the term of Saturn or Mars, and none of the terms
which are at the end of the signs are harder than the terms of these two.’
93 Pingree 1978: ii, 411 (1959); cf. Neugebauer, O., 1969, The exact sciences in Antiquity, New
York: Dover, 2nd edition; 1st ed. 1957, Providence (R.I.) [check] : Brown University Press.
which, considering the influence of Indian (largely Hellenism-derived, as he

has conclusively shown) astrology on Arabic astrology,94 almost certainly has
an Arabic counterpart. It contains the following parameters or boundary
conditions:
‘G[amma] = first visibility in the East

Ph[i] = first stationary point = sthitva [Sanskrit]
Th[ēta] = opposition
Ps[i] = second stationary point
Ō[mega] = last visibility in West = asta [Sanskrit]
Phi –>Psi = retrogression = vakra [Sanskrit]
This would largely make astronomical/astrological sense of the puzzling

‘thresholds’: Phi would be the point where the planet becomes retrograde;
probably the ‘entrance threshold’ i.e. ‘stationary point when coming in’; while
Psi would be the point where the planet becomes direct again, i.e. the ‘exit
threshold’ i.e. ‘stationary point when going out’.
Alternatively, the thresholds might be taken to refer to the ‘orb’: since the
planets are in constant motion, it is only during an infinitesimal fraction of time
that they make a specified angle vis-à-vis each other, e.g. the exact angle by
which a particular aspect is defined; assuming that the actual influence of that
condition is likely to last considerably longer, astrology reckons not only with
the exact angle but with a sector O:
O = A ± B,
where A = the exact aspect angle as defined, and

B = a specific number of degrees (many astrologers take as much as 10° in the case of
aspects involving the Sun).
B and -B then constitute boundary values for the aspect to become effective. In
other words,
orb = | B |
Also there are in-going and outgoing aspects, as well as increasing and
decreasing ones.
Even where the present Greeks manuscript fails to specify a geomantic name
for some of the sixteen configurations, there is overwhelming evidence for an
94 Ramsay Wright, R., ed. & tr., 1934, al-Biruni’s Elements of astrology, London; Alberuni, 1910,
Alberuni’s India, tr. Sachau, 2 vols
astrological reading in the other columns, referring to the Arabic, Latin and
Provençal manuscripts.
This concludes, in my opinion, the case for a primarily astrological reading of

the basic geomantic nomenclature. It means the demise of the ‘Girl’-with-the-
‘Beard’: both concepts, however featuring in the final geomantic corpus and
however conducing to juicy anecdote of the kind on which divinatory
repertoires thrive, dissolve in the austere abstractions of planetary motion. We
see that even the geomantic term ‘prison’, so becoming in the soap opera series,
means actually a simple astrological term referring to dejection, i.e. when a
planet finds itself opposite the sign or house of its rulership. The various
hurdles, upper and lower thresholds, to be taken in life, when entering and
(could it be less fair) again when going out, and all the other puzzling anecdotal
terms of the geomantic repertoire, turn out to be reduced to very technical
characteristics of planetary positions. Even if some of my astrological
explanations may turn out to be capable of improvement and replacement by
others, the overall message is clear: in its least adulterated form the geomantic
nomenclature is astrology of a particularly technical nature.
5.3. Semantic erosion in the Arabic nomenclature
With the extremely positive results of our analysis of the Greek geomantic
nomenclature, let us now go back to the Arabic version.
Let us not forget that geomancy as such was (if we may accept Ibn
Khaldun’s view) already an attempt to simulate the effects of astrology without
having to command the very complex skills needed for the astrological
apparatus itself. As one came to be further and further removed from an
understanding of the underlying astrological meanings of the nomenclature,
more and more space opened up for systematic bricolage on the basis of
iconographic fantasy. This bricolage could all the more satisfactorily converge
towards a consistent pattern (e.g. a remarkable fit between nomenclature and dot
pattern) because one was not longer impeded by the kind of genuine astrological
knowledge which would have made it only too clear that the link between dot
pattern and nomenclature was merely conventional, arbitrary, and not
substantial.
It is striking that already in the Arabic nomenclature itself we can detect this
same process of lessening understanding of the underlying astrological
meanings. As the Arabic geomantic nomenclature stands, it is extremely
heterogeneous (appealing to six different semantic domains) and rather
meaningless (certainly from an astrological point of view), but it does generate
associations: with profound emotions (success, loss, glory, joy, union), with
gender and generational oppositions, with a life’s road full of hurdles which
have to be taken. It is not surprising that such a semantic hotchpotch produces a
successful divination systems which managed, in all sorts of transformations, to
spread across half the world: vagueness, and absence of internal systematics,
render it possible for each diviner and each client to project new and topical
personal and local circumstances as well as deep-seated archetypes onto the
system. The cilm al-raml is a divination system with the semantic structure of a
late twentieth-century soap opera! But how could it ever have been designed
that way? The answer is that it was not: it was designed in a different, far more
specifically astrological way, and became what it is under the impact of the
eroding and streamlining effects of the human mind – it was popularised, turned
into symbolic and conceptual pulp.
There is no denying that besides the predominant astrological reference there
is in Islamic geomancy a secondary, conventionalised source of nomenclature:
the imagery based on the strictly graphic appearance of the geomantic figures. If
the figure is associated with incarceration, this can hardly be unrelated to its
closed, cyclical form. The same applies when is called ‘path’, ‘inward
threshold’ (where the road ends – even more conspicuous in the Arabian
notation, ).
The ‘beard cluster’ is quite suitable to argue the erosion and the secondary,
graphical re-interpretation of the geomantic system already in the Arabic form
as presented in Tannery’s table:
5.4. The puzzling beard complex
It is noteworthy that the Arabic column in Tannery’s table contains multiple

entries for the following configurations:
al-lahyan (the beard, a sign of masculinity and seniority but not an established
astrological term; contrary to what was probably the original word here, al-ahyan:
‘propitious times’) and
al-djaudala, ‘downy, feathery’; kausadj, ‘with a scarce beard’; farkh, ‘chick’; farkh
could also be a Persian loan word95 meaning ‘joy’ and in that case it does refer to a
specific astrological concept, see above; and
rayat farah, ‘flag of joy’; al-catabat al-dakhilat, ‘threshold when entering’
It should be noted that the orthographic difference between farkh, ‘chick’ and
farah, ‘joy’ is merely one dot on the in the former word, and that dot’s
presumably spurious addition might well have been the humble origin of the
whole series revolving on beard and varieties of beardlessness: ‘joy’, contrary to
beard, is a proper astrological concept, referring to a situation when a planet is
in the sign or house of its rulership, e.g. in in , etc.
Such sheer textual corruption seems to offer the best explanation for the
appearance of the beard/ scarce beard/ clean cheek complex in the Arabic
series.96 Could it be a mere accident of a confusion between farkh, ‘chick’ (i.e.
‘downy cheek’) and farah, ‘joy’ i.e. planetary rulership? It is noteworthy that in
Hugo’s translation from the Arabic the name of the configuration retains, in
‘barbatus-ridens’ (‘the laughing bearded man’), both the beard and the rulership
aspect. Would this be an indication that the Arabic geomantic texts, even if the
subsequent transmission of geomancy over a large part of the Old World is
largely due to them, are not the most original version but in themselves are
translations – perhaps from the Latin, Hebrew or Greek, or even, like Arsenius’
Greek rendering, from the Persian? An analysis of this translation against the
various, far from identical, Arabic versions that exist of al-Zanātī might be very
revealing on this point, but this must wait till I have located Arsenius’ text.
Meanwhile we have to admit that the three multiple entries in Tannery’s
table, including the beard complex, may also be attributed to a secondary,
merely graphical reading of the signs by those employing them in the early
95 As the iranologist S. Shaked tells me.

96 Of course, the beard as a sign of male authority and symbolic purity (as against the threats from
adolescent men, and from women) is a widely ramifying theme in the cultures, mythology and
iconography of the Ancient Near East, Graeco-Roman Antiquity, and the Islamic World. Further
research is needed on this point. I would also suggest that the beard complex, once established in
geomancy, in its turn probably produced the classification of Southern African geomancy-related four
divination tablets according to gender and generation; cf. van Binsbergen, 1994 and in press (a), (b).
second millennium CE. Such a secondary reading tallies very well with Ibn
Khaldun’s historical explanation of the rise of geomancy as a poor man’s
astrology, in the hands of those who had no longer access to the detailed and
complex astrological science with its sound foundation in computational
astronomy. Above I have demonstrated that most geomantic nomenclature has
an astronomical (or astrological) background. If this background fades away in
the hands of ignoramusses, they will find themselves deprived of an systematic
aids to remember the names and oracular implications of the sixteen
configurations, and instead will resort to far-fetched aids to memory.
If we imagine a human face with the top (the single dot) of the ‘al-
lahyan’/‘beard’ sign projected onto the place of the mouth, the
appropriateness of the term ‘beard’ will be immediately apparent. The
‘jaudala’/’downy’ sign is linked with the ‘beard’ sign through two simple
transformations: the beard sign is rotated 180°, and then the contents of the two
bottom rows is reduced from two dots to one dot each – as if the beard had been
thinned out or had not yet reached it full maturity! Now that this rather
convincing case reminds us of the fact that the geomantic symbols may have
been rotated in the process of naming or re-naming, we may look at the ‘flag of
joy’ with different eyes. The sign could be seen as a flag on a pole – although
this effect is belied by the ‘flag’s symmetrical placement on the top of the three
vertical dots forming the pole. Rotating the symbol 180° produces the sign for
‘al cataba al-kharidj’/‘outward threshold’, for which Tannery records no dual
meaning. A hidden obscene implication of the sign becomes now apparent: it
resembles an erect penis supported by two testicles, – surely a ‘flag of joy’ (the
top corresponding with the swollen gland, symmetrically placed on the shaft)
from the point of view of male chauvinism which was taken for granted in the
Ancient Near East, Graeco-Roman Antiquity and medieval Islam. If specific
knowledge is lost, obscenities (involving bodily parts and functions which are
always close at hand) form excellent aids to memory.
Meanwhile this suggestion of the signs having been rotation prior to
receiving the names under which they were subsequently mainly known, implies
the possibility of a much longer process of preparation, experimentation,
transmission and erosion for geomancy than would correspond with the image
of
having sprung forth fully-fledged and practically without preparation.
5.5. Is the Arabic nomenclature truly older than the Greek one?
Our analysis so far has brought us to a point where we may begin to question
the conventional wisdom that geomancy starts with the Arabs, and to consider
the existence of hypothetical earlier forms, less eroded en more consistently
astrological, than the Arabic versions at our disposal.
There is an extensive literature concerning the translation, into Greek and
Latin, of pre-existing Arabic geomantic texts; therefore, the overall historical
situation would assign greater age to the Arabic texts, so that the Greek and
Latin versions would be secondary derivations. But the case is not so
straightforward as all that. When Tannery97 describes ‘La géomancie chez les
Latins’, he deals with MSS which, through an oversight or deliberately, he fails
to assign a date to; as a non-Latinist, I am certainly incapable of assessing
whether these texts belong to either the early second millennium CE (which
would make translations from the Arabic) or to Late Antiquity (which would
make them parallels developments to, or possibly sources for, the Arabic texts).
By the same token, the accepted wisdom is that the Greeks texts, marked by
their neologism rabolion which does not mean anything in Greek but
phonetically refers, directly or via a Persian corruption, to the Arabic raml, are
necessarily secondary derivations.
Then how could we find in a Greek manuscript a geomantic nomenclature
that is less eroded than the standard Arabic version?
The easy answer would be to hypothesise the existence of an early,
unadulterated, consistently astrological, Arabic geomancy from which both the
later adulterated Arabic and the Greek unadulterated manuscript 2419 would
derive.
But there is a shocking alternative to this attractive view. One Greek
manuscript – the Codex Harleianus 5596, British Museum, from the 15th
century CE – 98 presents a simplified geomancy based not on tetragrams but on
trigrams, i.e. only three lines of one or two dots each – out of which of course
only eight different configuration can be constructed:
97 1920 o.c.
98 A. Delatte (1927, Anecdota Atheniensia, vol. i.: Textes grecs inédits relatifs à l’histoire des
religions, Bibliothèque de la Faculté de Philosophie et de Lettres de l’Université de Liège, fasc. 36,
Liège: Vailant & Carmanne/Paris: Champion, pp. 387-396.
At first glance one would consider this to be merely a simplified, adulterated

geomancy derived from an richer Arabic original. But why could it not be the
other way round: a more original, proto-geomancy inspiring a subsequent
Arabic development, and then dating back to perhaps the third quarter of the
first millennium CE, or even earlier?
Delatte describes this manuscript in the following terms:
‘‘Cléromancie astrologique: ‘Ce petit traité de cléromancie décrit la valeur prophétique de

vingt quatre combinaisons (skhēmata ou klēroi ) de sept demeures (oikoi) ou signes figurés
par des points. Ces signes, à leur tour, sont formés vraisemblablement par les diverses
dispositions de trois nombres, pairs et impairs, de même que, dans le traité arithmomantique
qui précède, les figures sont formées des combinaisons de quatre nombres pairs et impairs
(cf. infra, ad. codic. Paris 2494). En outre, chacun d’eux représente une planète et des signes
du Zodiaque, comme on le voit dans l’introduction et dans une figure placée au f. 4v. Voici
quels sont ces rapports:99
Could this be a proto-geomancy providentially preserved to reveal this

system’s true, Greek-language i.e. Hellenistic, origin?
The number of eight configurations is interesting in the light of the fact that,
given the astrological orientation of geomancy, the emphasis is on planets not
signs. Well, one does not need four lines, yielding 16 different configurations, to
accommodate all seven planets known in Antiquity: ; after
allocation of these seven planets we would still have to fill the eighth slot ad
hoc, as in the case discussed by Delatte, where 5 has received two figures
instead of only one.
This tallies with the stress on planets, rather than on zodiacal signs, in
geomancy, and marks a rather early phase in the development of Hellenic and
99 Delatte does not give the astrological glyphs but presents exactly the same information in
discursive French prose.
Hellenistic astrology.100 As a non-specialist I also wonder whether the names of

Greek deities would have been used – as usual in Antiquity – to designate the
planets, if this text in the Codex Harleianus 5596 hailed from post-geomantic,
Christian Byzantine times (the historically less controversial interpretation)
instead of from Late Antiquity or earlier.101
Where would the idea have come from, of graphically representing up to
eight different options in a divinatory system, by three lines of single or double
dots? To my knowledge, such a system does not seem to have any precedents in
Greek and Roman Antiquity, Egypt or the Ancient Near East. It could however
very well have come – through whatever detours, like so many other Chinese
intellectual products – from China, where our Greek geomantic trigrams have
obvious counterparts in the eight sacred trigrams (pa kua) that permeate the
entire Chinese cosmology. Making use of the fact that, in the Chinese
symbolism, a broken line means odd and an unbroken line even, one could
match the Greek trigrams in the Codex Harleianus 5596 with Chinese ones in
the following way:
The Chinese trigrams are of great antiquity,102 communications between

Europe and China were relatively open during certain periods in Late Antiquity
when e.g., from the 1st century CE onwards, silk was the main luxury article
100 Cf. Bouché-Leclerc, o.c., and Tester, o.c.

101 A quick glance at the Catalogue, o.c., which contains much Byzantine astrology, should suffice to
demonstrate the futility of this argument, but I do not have that publication available right now.
102 In the Chinese tradition, they are attributed to a prehistoric culture hero king Fu Hsi, and they
actually appear on artefacts from the first millennium BCE; the oldest tangible, archaeological
evidence for the I Ching ‘omen compendium’, featuring hexagrams traditionally considered to be
constructed out of combinations of pre-existing trigrams, dates from the 1st century CE: Ngo Van
Xuyet, 1976, Divination, magie et politique dans la Chine ancienne: Essai suivi de la traduction des
“Biographies des magiciens” tirées de l’ “Histoire des Han postérieurs”, Bibliothèque de l’Ecole des
Hautes Etudes, Sciences religieuses, vol. lxxviii, Paris: Presses Universitaires de France.
among the Roman elite. After a chance diffusion of the idea of the Chinese
trigrams (most probably without any of the attending Chinese cosmology and
divination techniques) one can imagine how only later such a proto-geomancy
was diversified and complicated by replacing, under whatever influence or
inspiration, the three lines by four, which then necessitated twice the number of
configurations to be filled with meaningful correspondences and names. At least
two possibilities seem to present themselves for accommodating 16
configurations:
(a) A doubling of the presumed original planetary attributions, so that each planet takes
care of a pair of configurations with minor differences in meaning; while this only
fills 14 configurations, one needs an additional 2 to fill, e.g. by Caput Draconis ( )
and Cauda Draconis ( ), abstractly calculated astronomical points which
significantly each occur only singly, not (like the planets) doubly – even although
they are further treated at a par with the other planets. (It would look as if this was
finally implemented, so as to produce classical or advanced geomancy)
(b) The introduction of a fresh, if related, basis for all sixteen configurations; this basis
could not be found in the otherwise so prominent lunar mansions,103 but it could be
found – and the extensive evidence from Greek manuscript 2419 suggests it has been
found – in the very rich astrological doctrine on the various interrelations between
planets, and between planets and signs, such as defined under the heading of
planetary dignities.104 (This would then presumably constitute the first astrological
form of geomancy.)
103 Of which there are 27 or 28, and not 16; but one could imagine yet a further development towards
complexity in geomancy, working with five lines, which yields 32 configurations, i.e. sufficient to
accommodate all lunar mansions, which a few more attributions to be added ad hoc. In fact, a five-line
geomancy does exist, not only in the standard commentaries to Dante, but also in one of the most
popular modern ‘books of fate’, the pseudo-epigraphical Napoleon’s Book of Fate, – which admittedly
does not seem to be older than the 19th century CE, and which has hardly penetrated outside the
British Isles; cf. Napoleon’s Book of Fate and Oraculum, ca. 1925, London: W. Foulsham & Co;
Parker, B., ed., 1988, Napoleon’s Book of Fate: Ancient Egyptian Fortune-Telling for Today, by H.
Kirchenhoffer, 1822, revised and edited by B. Parker, Wellingborough: The Aquarian Press; Skeat,
T.C., 1954, ‘An early mediaeval Book of Fate: the Sortes XII Patriarcharum: With a note on “Books
of Fate” in general’, Mediaeval and Renaissance Studies, iii: 41-54.
104 Further research would have to extend the detailed analysis of the nomenclature of the Arabic and
Greek manuscript 2419 version to the Latin and Provençal versions. Here the Hermetic manuscript
7349 might turn out to occupy a key position, especially if it can be analysed against the wider family
of Hermetic astrological texts, and suggestions can be made as to the likely date of its production
and/or copying. Cf. Gundel, W., 1936, Neue astrologische Texte des Hermes Trismegistos, Abh.
Bayer. Ak. Wiss. Philos.-his. Abt., N.F. 12.; Festugière, R.P., 1943, La révélation d’Hermès
Trismégiste, I: L’astronomie et les sciences occultes, Etudes bibliques, Paris: Lecoffre/Gabalda, 2e
éd., 1950; Massignon, L., 1950 (1943), ‘Inventaire de la littérature hermétique arabe’, appendice iii,
in: Festugière, o.c., pp. 384-400.
This exploration into the possibility of a Greek or Latin (or, for that matter,
Pahlavi, Hebrew, or Mesopotamian) Chinese-influenced proto-geomancy in
Late Antiquity or the early middle ages preceding any Arabic version, is highly
hypothetical. For the time being I do not think we should try to proceed beyond
the insight into the astrological basis of geomancy, and into the eroded nature of
even the standard Arabic version of that occult science.
In a next phase of the research I hope to develop an argument according to
which certain astrological elements in the various geomancies can be used as
indicators for periodisation. E.g. although and were already available in
Hellenistic astrology105 they are reputed, like other immaterial points such as
,106 to have only become a major topic in astrology with the Arabs. Or again,
while we now take the correspondence between certain planets and signs for
granted, different attributions have been made in the long history of astrology;
now since, as the prominent astrologer Hamaker-Zondag remarked,107 the
correspondences as found in the medieval and Renaissance body of geomantic
texts in the European tradition go against the pattern as agreed on in mainstream
astrology, this may help to pinpoint the period of origin of geomancy as a
breakaway from mainstream astrology.
6. CONCLUSION
An analysis of one of the key features of the geomantic family of divination

systems, the nomenclature of the sixteen configurations, has yielded important
insights for the history of the system before it appeared, fully fledged, in
documentary evidence towards the end of the first millennium CE.
The question as to which particular nation, ethnic group, language group,
religion or sect would have been the system’s cradle, turns out to be rather
105 Bouché-Leclercq, A., 1899, L’Astrologie grecque, Paris: Leroux, p. 122-123 and passim; caput
draconis: anobibazōn; cauda draconis: katabibazōn.
106 The astrological symbol of the earth, and of the Pars Fortuna; its iconography of course shows the
chthonic foursome in the form of a cross.
107 Hamaker-Zondag, K., 1981, Oude Europese voorspelkunst, Amsterdam: Schors; a useful
summary of these correspondences in European geomancies since the middle ages (virtually all based
on Arabic sources) is also offered in: le Scouézec, G., 1965, ‘Géomancie’, in: Le Scouézec, G., H.
Larcher & R. Alleau, 1965, Encyclopédie de la divination, n.p.: Tchou, pp. 392-459.
irrelevant. Although further detailed analysis of the many other features of the
system may reveal certain culture-specific contributions, it is clear that as an
intellectual development geomancy situates itself primarily within the
international free floating, increasingly universal repertoire of advanced
astrology, which emerged as a product of Hellenism and already then, as in its
later development, absorbed and transformed within a new unity the many
heterogeneous inputs of the rival yet converging intellectual, cosmological and
religious idioms of Antiquity and the early middle ages. The point is not to deny
that identifiable intellectual traits specific for a national culture, a philosophical
school, a creed, have found their way into the body of advanced geomancy, but
that such contributions were invariably filtered or mediated through astrology:
e.g. if there was ever a Pythagorean contribution to geomancy in terms of
number symbolism, this would only be indirectly reflected in geomancy to the
extent to which the number four (which, and not by accident, the number of the
earth, which gave geomancy its name) found expression in the astrological
doctrine of the four cardinal, fixed and cadent signs, and in the doctrine of
aspects which, for each different aspect, each of the sacred numbers featuring as
divider of the total circle of 360° – thus yield 360° (or 0°), 180°, 120°, 90°, 72°,
60°, 45° etc. as aspects to be considered; cf. Appendix 1.
This view also helps to bring the astounding achievement of geomancy in
perspective: it was as an adulterated form of astrology, that geomancy
conquered the world – greatly facilitated by the fact that a large part of that
world had already given in to astrology as an international, boundary-crossing,
intellectual framework in which to discuss fate, character and the vicissitudes of
life and relationship in general.
While earlier, more primitive forms of ‘striking the sand’ have been attested,
and while my research in progress suggests an important contribution from
north-west Africa in towards the chance procedures and iconography of the
geomantic system as a whole, in the case of geomancy at least astrology yet
cannot be said to be the secondary imposition it was in the case of chiromancy,
numerology, oniromancy, herbalism, mineralogy, etc. Our decision, in the
present argument, to concentrate on the history of the geomantic nomenclature
was strategic: it has begun to persuade us that the central corpus of the
geomantic family can be largely understood on the basis of the state of astrology
in Late Antiquity, possibly with some modifications from subsequent
developments in Indian and Arabic astrology. It remains a matter of academic
interest, and national pride perhaps, to pinpoint the specific milieu (Harranite,
Arabic, Jewish, Persian) in which the true, astrological geomancy was first
formulated as a total package, but this must be seen as merely a particular
intellectual innovation within the well-established tradition of astrology.
And the latter field is no longer terra incognita, since the twentieth century
has seen a host of studies (including Neugebauer’s, Franz Cumont’s, the
Catalogus108 of the Greek astrological corpus whose publication was initiated by
the latter, more recently Pingree’s studies) which have made the history of
astrology one of the best documented fields of ancient studies, and one in which
continuity can be traced and demonstrated not so much by the centuries, but by
the millennia.
I have broken down the standard geomantic package into fourteen
constituent items. When the detailed analysis of merely one of these items, the
geomantic nomenclature, already offers so much that is exciting and that
questions established wisdom in this field, one wonders what the results will be
of similar analysis of the remaining items.
108 Catalogus, 1898-1951, Catalogus Codicum Astrologorum Graecorum, ed. F. Cumont, F. Boll et
al., vols. 1-12, Bruxelles.
Les seize figures du sikidy.
Fig. I
Fig. II
Fig. III
Fig. IV
Plate 1. An illustration from a pioneer, 19th century study of geomancy on Madagascar:

Ferrand, G., 1891-1902, Les musulmans à Madagascar et aux Iles Comores, 3 vols (1891,
1893, 1902), Paris: Ernest Leroux, p. i 76. It nicely visualises how ‘striking the sand’ (darb
al-raml ) with a stick produces a series – of change length – of dots; counting subsequently
yields, for this series, the value ‘even’ or ‘odd’, which then defines the entry ‘two dots’ or
‘one dot’ on one of the four lines of a geomantic tetragram to be formed. For each tetragram
therefore one has to strike four times. For complete geomantic themes one has to strike
sixteen times, but of course this only yields the four initial mothers; the remaining twelve
figures have to be calculated following the algorithm set out in my argument.
Ms. arabe 2631, f° 64 v°:
TALISMAN CONTRE LES MALADIES QUI PEUVENT ATTEINDRE

LES DIFFÉRENTES PARTIES DU CORPS DE L'HOMME
Plate 2. ‘Geomantic man’, from a Paris manuscript attributed to Tumtum al-Hindi.

Melothesia (the doctrine of specific correspondences between parts of the human body on the
one hand, celestial bodies, zodiacal signs and immaterial astrological points on the other)
formed an integral part of classical astrology (cf. Bouché-Leclercq 1899 o.c.; Pingree 1978),
and from there this approach was copied into geomancy.
LIBER SECVNDVS CXCV

CHARACTERES VENERIS.
Ab amissione
A puella
CHARACTERES SOLIS.
A fortuna
maiore,
A fortuna
minore.
Charasteres Martis.
A rubeo
A puero
Characteres Iouisi
Ab acquisitio
one.
A laetitia
Characteres Saturni
A carcere
A tristitia
Characteres capitis Draconis.
Characteres causa Draconis.
Plate 3. Geomantic glyphs (from Agrippa, as copied in le Scouézec, o.c.)

APPENDIX 1: AN OVERVIEW OF WESTERN (≈

PTOLEMAEAN) ASTROLOGY
For a good understanding of the preceding argument, a basic knowledge of the Western (more or less:
Ptolemaean) tradition in astrology is needed. The following pages offer an overview in tabulated form.

The Astrological Origin of Islamic Geomancy: 1. Geomancy and Its Distribution in Time and Space

Uploaded by

Copyright:

Available Formats

The Astrological Origin of Islamic Geomancy: 1. Geomancy and Its Distribution in Time and Space

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Astrological Origin of Islamic Geomancy: 1. Geomancy and Its Distribution in Time and Space

Uploaded by

Copyright:

Available Formats

The astrological origin of Islamic geomancy

Wim van Binsbergen1

1. GEOMANCY AND ITS DISTRIBUTION IN TIME AND

Arabic originals, by Muhammad al-Zanātī, cAlī b. Umar, Fadhl b. Sahl al-

From a very different corner of the Old World: Francistown, Botswana,

And finally, as one of the earliest documentary attestations of geomancy in an

2. ISLAMIC GEOMANCY AS A DIVINATION SYSTEM

Essentially, any inductive divination system consists of two15 elements: a

even + odd = odd, i.e. oo +o = o,

18 van Binsbergen 1996, o.c.

E(5, 1...4) = E (1...4, 1)

The example in diagram 1 should suffice to illustrate this apparently

(a) mathematically, there are 16 possible different configurations which

(b) The construction of a full geomantic ‘theme’ (i.e. horoscope’) out of

Figure 1. An example of the production of a full geomantic theme of 16 figures,

the University of Leiden library.

‘In the name of ALLAH,

Figure 4. Opening page of a popular Cairene edition of al-Zanātī, 1320

28 Repetition in the original.

The planetary correspondences attributed to the zodiacal signs in subsequent

35 Source: le Scouézec, ‘Géomancie’, o.c.; al-Zanātī, o.c.

Not only is geomancy thus an early example of globalisation, it is also a

Figure 2. Distribution of the geomantic family of divination systems.38

38 For a fuller discussion cf. van Binsbergen in: ‘Rethinking’, o.c.

6. I Ching system of China

3. THE HISTORY OF GEOMANCY AS A PROBLEM

If the history of science can be said39 to be Clio’s step-child (and by

systematic study of African societies made a beginning,41 the result was a –

translation, by the pioneer medieval translator Hugo Sanctalliensis, of the

in Africa and although he reminds us of the mythical traditions concerning the

Struggling for a feasible method by which to approach our central question as to

fundamental, independent moral category in myths, oaths and rituals from

4.1. A heuristic typology of geomancy

c. Simple procedures of casting or tracing, and subsequently sorting or eliminating these

Table 2. An exhaustive descriptive definition of Islamic geomancy

4.2. Getting focused as to time, place and topic

cleromantic and otherwise stochastic procedures for divinatory and ludic

from Sumer and pre-dynastic Egypt right through to Neo-Platonic, Sassanid,

process that was to be continued in Hellenistic and Imperial times.80 Therefore

‘XIII AND XIV. RAMALA AND TAJIKA

All the various clues which I gradually detected, can be accommodated in

contribution, beyond the comparative and methodological points made so far in

5. THE GREATEST PUZZLE OF ALL: THE ORIGIN OF THE

5.1. Semantic heterogeneity in the Arabic geomantic nomenclature

Table 3. The geomantic nomenclature according to Arabic, Greek and Latin

5.2. The astrological contents of the nomenclature in the Greek

In an attempt to establish the technical astrological nature of the geomantic

Table 4. The astrological contents of the geomantic nomenclature in the Greek

It is clear that the references, in the geomantic nomenclature, to thresholds

which, considering the influence of Indian (largely Hellenism-derived, as he

‘G[amma] = first visibility in the East

This would largely make astronomical/astrological sense of the puzzling

where A = the exact aspect angle as defined, and

This concludes, in my opinion, the case for a primarily astrological reading of

5.3. Semantic erosion in the Arabic nomenclature

5.4. The puzzling beard complex

It is noteworthy that the Arabic column in Tannery’s table contains multiple

95 As the iranologist S. Shaked tells me.

At first glance one would consider this to be merely a simplified, adulterated