Dark Tongues

The Art of Rogues and Riddlers

Daniel Heller-Roazen

Z O N E B O O K S • N E W Y O R K

2013
© 2013 Daniel Heller-Roazen

ZONE BOOKS
1226 Prospect
Avenue Brooklyn,
NY 11218

All rights reserved.

No part of this book may be reproduced, stored in a

retrieval system, or transmitted in any form or by any
means, including electronic, mechanical, photocopying,
microfilming, recording, or otherwise (except for that
copying permitted by Sections 107 and 108 of the U.S.
Copyright Law and except by reviewers for the public,
press), without written permission from the Publisher.

Printed in the United States of America.

Distributed by The MIT Press,

Cambridge, Massachusetts, and London, England

Library of Congress Cataloging-in-Publication Data

Heller-Roazen, Daniel.
Dark tongues / Daniel Heller-Roazen.
pages cm.
ISBN 978-1-935408-33-8
1. Jargon (Terminology) 2. Cant. 3. Languages,
Secret. 4. Poetics. 5. Historical linguistics. 6.
Anthropological linguistics. 1. Title.

P409H35 2013
417-dc23
2012047271
 Contents

I Forkings 9

II Coquillars 19

III Principles of Cant 31

IV Signs 45

V Knowings 55

VI Riddles 65

VII Nomina Diving 83

VIII Nomina Sacra 95

IX Anaphones 109

X Patterns 131

XI Secrets of Tristan Tzara 153

Notes 185

Bibliography 215

In d e x 2 3 3
 C H A P T E R O N E

Forkings

That human beings are speaking beings is an old and oft-stated

belief. Aristotle was perhaps the first to make of it the basis of a
definition when, in a famous passage of the Politics, he declared that
among all animals, "man is the only one that has language" (logon
de monon anthropos ekhei ton zoon).' "Language," however, was and
remains an obscure word. The fact that the Greek philosopher's
animal that has language could have been renamed, in Latin transla-
tion, the "rational animal" (animal rationale), affords one example of
how a term for the fact of speaking can lend itself to several inter-
pretations. Aristotle's logos, in Greek, signified a thicket of notions
usually distinguished today: "word," "speech," and "discourse," to
be sure, but also "reason," more generally, and more specifically,
arithmetical "ratio" and musical "interval."' The Aristotelian argu-
ment, therefore, can be recast in more ways than one. Yet its gram-
mar is also significant in itself. Aristotle's words suggest that what
human beings possess in contrast to all others is an ability that may
be named by a singular noun. It is the faculty of speaking. However
self-evident it may seem today, that claim runs up against a reality
more perplexing than the ancient philosopher and many of his suc-
cessors seem to have been willing to admit. It can be simply stated:
speaking beings never speak language, except in speaking languages.
English has one word for two linguistic beings that may clearly be
distinguished: on the one hand, "language," insofar as it designates
the generic fact of speaking, and on the other, "language," insofar
as it points to a variety of speech, such as Armenian, Japanese, or

9
 PARK TONGUES FORKINGS

-Viable. Some foreign tongues are more perspicuous. The Romance that claim their roots in the disciplines of Greece and Rome. More
languages. tor example, regularly admit a lexical distinction between than once, it has been observed that the Greeks and Homans dis­
an abstract term tor speech (such as lenguaje, linguagem, langage, or played relatively little interest in the idioms that, as they knew well,
linguaggio) and a specific term for a language, with its words and surrounded them. Scholars have advanced several hypotheses in
rules (idioma, lengua, langue, lingua). That there is a relation between explanation o f this omission. Perhaps the Greeks and Romans did
the ideas expressed by these two sets of terms, of course, can hardly not trouble themselves with foreign tongues because they held the
he doubted. It points to an epistemological circle that, explicitly peoples they called “ barbarians” to speak idioms absolutely unlike
or implicitly, sustains a practice of definition by abstraction. The their own, being both unknown and essentially unknowable. Per­
single faculty o f language can be found nowhere but in languages, haps, instead, the Greeks and Romans deemed foreign languages
which are plural by definition; yet languages, for their part, cannot unworthy objects of study because they held them to be essentially
be considered as the members of a class unless one has presupposed like their ow n, differing only in their vocabularies.5 It is remark­
this concept: language.3 Depending on one’s interests and one’s per­ able, in any case, that the great exponents of classical disciplines as
spective, one may wish to consider the notion or its instantiations, diverse and sophisticated as philosophy, geography, historiography,
the faculty in general or its varieties. Yet the point o f departure, and grammar all concurred in seeing no need to treat the plurality
among speaking beings, remains this first forking. Whenever there of languages as a fact in need of special comment.
is language, in the definite singular, there are, in truth, languages, The Homeric epics, the earliest monuments of Greek literature,
in indefinite and indeed innumerable multiplicity; whenever there portray a w orld that hardly requires interpreters, and in which
are languages, in the plural, one may detect the shadow o f a speak­ important speaking subjects, whether Achaean or Trojan, all con­
ing faculty, no less perceptible for remaining distinct, by definition, verse freely in one tongue. Certain passages, to be sure, indicate
from every tongue. The predicament may be extolled or decried, but that the Homeric poet was well aware of some foreign idioms: thus
it cannot be denied. “ Languages,” Mallarme observed: “ imperfect in the Iliad evokes the Carians, “ barbarian” (barbarophonoi) on account
that, being many, the highest of them is lacking.”4 of their tongue, and the Odyssey dwells briefly on the Cretans, said
If one considers the history of investigations into the nature of to be “ o f many roots and tongues.”6 But such non-Greek languages
speech, it is difficult to avoid the impression that more often than would seem, in the Homeric world, to be signs of distant marvels.
not, the discourse on language, in its simplicity, has left little room The philosophers said little more. There is no doubt that Plato,
for the multiplicity of tongues. That human beings are speaking for example, was familiar with the fact of linguistic diversity. Yet
beings has been interpreted as implying that, by nature, they confer when he dedicated a dialogue to the nature and formation of names,
with one another about the good and the bad, and do not only signal he wasted no words on the differences between Greek and other
the pleasant and the unpleasant; that they aim to im part to each tongues, nor did his Socrates see fit to investigate the reasons why
other, as best they can, their ideas and conceptions, for whatever forms o f speech clearly differed among the parts of Greece, to say
ends; that they designate, act, reason, calculate, or communicate. nothing o f different peoples. The world of the Cratylus is one of a
The possibilities are many. The “ theory of language,” in any case, single language, in every sense. Aristotle, who defined the human
has tended to consider its object as a single being. peoples as the sole possessor of logos, proposed an elaborate theory
This much may be an inheritance from antiquity, at least with of language and logic, which he developed in treatises bearing on

regard to those forms of knowledge, such as philosophy and grammar, numerous subjects, such as signification, deduction, poetry, rhetoric,

10 11
 DARK TONGUES FORKINGS

politics, ami biology. Yet everywhere Aristotle proceeds as if his an influence on later thinking about the nature of the speaking ani
logos could he treated as one. mal. According to the author of the Book of Genesis, there was an
One might expect the classical historians to have had a livelier age in which “ the whole earth was of one tongue.” In sacred time,
interest in differences of speech, and to a certain degree, this is if not in history, language could, therefore, be purified of the differ­
indeed the case. Herodotus, the first in the tradition, noted with ence of languages. Confusion could come later.10
curiosity that in different regions, the earth bears not one but three Today, there is, of course, a form of inquiry and knowledge that
names, each of which appears to evoke a woman: Europe, Asia, takes the diversity of languages to be fundamental. It is the sci­
Libya.' He observed, moreover, that the same divinities seem to ence of language. Linguistics must admit, as an axiom, that there
reappear from culture to culture, named each time differently.8 is not only a distinction between language and nonlanguage but
f et despite his commitment to studying the testaments to human also that there is a distinction between one language and another.
diversity, Herodotus saw no need to wager any explanation for them, Admittedly, linguists may define this distinction in several ways,
nor did he offer a commentary on the proliferation of apparent syn­ accepting, for example, existing sociological determinations, such
onyms among the peoples of the world. One might well have won­ as labels of national language and dialect, or seeking to ground this
dered why it is that different communities call the same things by distinction in the consciousness of groups of speaking subjects.
so many different names, even if one did not go so far as to pose this But linguistics must admit this much: there are systematic formal
fundamental and unavoidable question: What sense may one give to differences between languages. Any grammatical analysis, in the
the fact that the human capacity to speak finds expression only in a traditional sense, will illustrate this fact.
multiplicity of tongues? Yet if linguistics differs from grammar in the old sense, it is in
It would be untrue, of course, to maintain that the thinkers of passing from the diversity of such idioms to more general consid­
the classical world simply ignored the problem of linguistic differ­ erations. Presupposing that there are various languages that share
ence. References to variation in dialect and language are never hard properties that, once abstracted and combined, define the faculty
to find, and occasionally, the question of plurality even emerged in of speaking, linguists may, for example, establish historical and
philosophical form. Democritus, the early atomist, seems, for exam­ genetic relations between languages: derivations and divergences,
ple, to have treated the multiplicity of languages as a phenomenon resemblances and dissimilarities. At times, linguistics can thus offer
in need of scientific analysis. According to Diodorus, the materialist a clear historical answer to the riddle of linguistic diversity: multiple
addressed the question in a book now lost to us, arguing that varia­ tongues may be shown to derive from a single language. Indo-Euro­
tions in speech result from distinctions of geography and climate.9 pean philology presents perhaps the most powerful example of this
Yet his discussion, if it did exist, would have represented more the kind. Through close examination of the distinctive properties of a
exception than the rule. The evidence suggests that the classical host of European and Asian languages, the scholars of the nineteenth
Greek and Roman thinkers, by and large, believed speech to be a century succeeded in revealing a series of striking correlations that
thing from which multiplicity could be, as it were, subtracted, if not suggested a common source now lost to us: “ Indo-German,” as it
in reality, then certainly for the purposes of theoretical speculation. was once named, or “ Proto-Indo-European,” as its current practi­
They accorded only slight importance to the fact that language is tioners in English choose to call it.
always distributed in languages. To this degree, at least, their view The rigor of such scientific inquiry, however, derives from the
was not inconsonant with that of the Bible, which was to exert such limits it sets itself. No serious comparative linguist ever sought to

12 *3
 PARK TONGU ES FORKINGS

argue that all tongues emerged from a single source, for reasons no conscious knowledge of them: rules of syntax, which dictate
both methodological and material. Comparative linguistic analysis the arrangement of phrase structures, independently of their con­
rests on the supposition that languages, in general, are distinct in tent; rules of morphology, which determine the possible shapes that
their rules and in their elements. From this perspective alone are expressions can acquire within sequences of discourse; rules, finally,
correlations and analogies significant. For example, if the charac­ of phonology, which bear on a restricted set of sounds that possess
teristics uniting Greek and Sanskrit, or Old Irish and Latin, are no meaning in themselves but that every speaker of a language some­
remarkable and in need of explanation, this is because they are, in how knows to order, combine, and understand.
principle, unexpected. Only where the natural diversity of idioms This much, however, is no less enigmatic today than it was in
seems lacking will one think to argue for a single source. And, it antiquity: speaking subjects speak only languages, and their basic
should he added, some languages alone seem to be related in such a element is opacity. Of course, for an individual, study and acquain­
form. There are many European and Indian tongues, such as Basque, tance can dispel, in part, the obscurity of unknown tongues, but in
Hungarian, and the multiple Dravidic languages, that appear clearly general, foreign languages appear to speaking subjects as those that
irreducible to the “ Indo-European” set. There are, more impor­ are resistant to comprehension and appropriation. One may even
tantly, entire groups and “ families” of idioms that seem to bear no find in the fundamental perception of unintelligibility the simplest
substantial genetic links among each other. The Afro-Asiatic or index of the difference of languages. Two tongues can be considered
“ Hamito-Semitic” languages, for example, do not appear to derive distinct when their respective speakers, each employing their own
from the same “protolanguage” as the Indo-European, nor can one idioms, fail systematically to understand each other. One may well
establish that they derive from the same roots as the Altaic, the Sino- hold, therefore, in accordance with an old tradition in philosophy,
Tibetan, or the Iroquoian tongues, to choose only some from many that there is language from the moment in which there is significa­
possible examples. In linguistic investigation, grammatical diversity tion, reasoning, and articulated intention. One may also concede,
remains a fact that must be presupposed. Only exceptionally can it granting the validity of linguistic research, that there is language
be explained. from the moment in which one may detect, in a single idiom, a finite
If it is to possess a single object, the science of language, there­ grammatical system that speaking subjects unknowingly respect in
fore, must pass by abstraction from languages to language: the fac­ producing an infinity of utterances. Nonetheless, one may be certain
ulty of speech. That transition may well recall the inference by that there are languages, in the plural, if such modes of understand­
which the philosophers of antiquity moved from a single language, ing can come systematically to a halt, as rules of forming correct
such as Greek, to a general principle, such as logos. Yet linguistics utterances in one tongue run up against those in another. Language,
has made of this procedure a first step in the constitution of a new in its singularity, may be grasped as a mode of intelligibility common
mode of inquiry, which has led to a major discovery concerning the to human beings as a rational species, or to communities united in
fact of speaking. It is worth recalling today, if only because it seems following the rules of grammars. Languages, in their plurality, point
to be increasingly forgotten. Since the emergence of comparative to impenetrability and incommensurability, which incessantly sow
grammar in the nineteenth century, the science of language has division among those who speak.
established, with increasing precision, that the utterances of speak­ The most obvious illustrations of the fracture produced by lan­
ing subjects, as a whole, systematically respect a limited set of formal guages are to be found whenever speaking communities, coming
rules of grammar, even where speaking subjects themselves have into contact with each other, discover that their idioms, otherwise

15
H
 PARK TONGUES FORKINGS

huh reliable moans of communication, forbid all understand­ subjects, albeit within limits, to take apart ami rebuild a single
ing It has always boon known that in such circum stances, little is tongue. One language, in itself, can form a fork, not only by nature,
loss eloquent than discourse; little is more obstinately unintelli­ but also by will and art. In private or in public, those who speak a
gible—less, in short, like language—than a language. What could be language retain the capacity to draw from their knowledge of its
more unfathomable, in itself, than the significations enclosed in an grammar the elements of a new and cryptic variety of speech. Such
unknow n tongue, in sentences, words, or even such small soundings an idiom may be playful or serious, a secret shared by children in
as a change in vowel quantity, a heightening or lowering of a pitch, their games or by adults in their work. It may affect individual words
an addition, to a string of consonants and vowels, o f an aspiration, or phrases, phonemes or flexions, formulae or sentences, whether
such as that of the letter h? Those who speak a language somehow taken in isolation or in coordination. It may appear to be as alien
know that such elements can all decide meanings that are irreduc­ as a foreign language, only slightly different from the idiom from
ible to the physical properties of speech. Authorities of various kinds which it springs, or almost indistinct from the language from which
have always recommended, for this reason, that, when confronted it has been made, its traits as imperceptible as its hidden sense. At
with the plurality of tongues, one set languages aside. It is wiser, the limit, the existence of such an occult idiom may itself become a
then, to turn to forms of expression unencumbered by the subtleties doubtful thing, a hypothesis of a hidden object that one might affirm
of grammar: forms such as gesture, “ the discourse common to all or deny. The ways of splitting speech are many, surely no fewer than
men” (omnium hominum communis sermo) , 11 as Quintilian once wrote, their occasions. Yet each time a tongue, by effort and by skill, divides
“pantomime,” which Rousseau held to be older than and prior to or even merely appears to divide in two, the same perplexing fact
individual tongues,12 dance, which can convey meanings without may be discerned. It seems that human beings not only speak and
words, as Lucian believed,13 or music, so often extolled as a universal speak languages. They also break and scatter them, with all their
means of expression. reason, in the sounds and letters of tongues made multiple and dark.
Yet the truth is that one need not look to encounters between
different linguistic communities for evidence of the impenetrability
brought upon speaking subjects by their tongues. There are occa­
sions when the confounding force of languages exerts itself within
the limits of what would, in all other respects, appear to be a single
grammatical system. A new and unexpected divisiveness then can
be perceived.
This is not merely a consequence of the continuous nature of lin­
guistic change, which dictates that in speech, as Dante wrote, “ every
variation varies in itself,” 14 such that from one time to another, a
single language grows ever more opaque to those who would make
use of it. One must also take into account a fundamental aspect of
the power to speak that has not received the attention it deserves
from philosophers or linguists. Among the capacities implicit in the
faculty of language there is the ability, possessed by all speaking

16

%
 CHAPTER TWO

Coquillars

In 1890, there appeared in the Memoires de la Societe Linguistique de

Paris an article by the twenty-three-year-old Marcel Schwob, who
had yet to publish the works of poetry and prose for which he would
one day be known. The essay was a study of a medieval manuscript
belonging to the Archives of Dijon that Schwob had transcribed,
introduced, and annotated. He had learned of the existence of the
codex from a bibliographical note that Joseph Gamier, archivist
of the Cote-d’Or in Burgundy, had devoted to it in 1848. Gamier
himself had chosen not to comment in detail on the materials he
had made known, and the eminent French philologist Francisque
Michel, to whom Gam ier mailed one of the forty fascicles of the
manuscript, had also not seen fit to discuss it. Schwob thus became
the first to investigate a set of documents that had most likely
remained unstudied since the time of their redaction in the fifteenth
century, and it was his essay that revealed their curious contents to
the public.
The manuscript contained a legal dossier, drawn from the
records of the Burgundian capital, concerning a company of ban­
dits arrested, tried, and condemned in the year 1455. The deeds
the wrongdoers were accused of having committed seem, at first
glance, unremarkable in themselves. The authorities alleged that in
Dijon and the neighboring countryside, the vagabonds had commit­
ted various acts of violent depredation: picking and breaking locks,
plundering coffers, holding up hapless travelers on their way, rob­
bing unsuspecting merchants who shared rooms in inns. It was said

19
 OAK TONGUES COQUILL ARb

that some of the idlers were skilled in the arts of petty cheats and
merchant, bemoan their misfortune with the robbed men bant, while
swindlers. it was claimed that others would not hesitate to murder
having a man working for them with whom they then share the booty
to bring themselves profit. t here was no doubt that the company
Others cheat at dice games, and they win all the money of those with
of ro gues possessed no wealth of its own, But poverty being the
whom they play. They know subtleties such as card games arid hop
mother of invention, the brigands had devised an ingenious tech­
scotch, and no one can win against them. And what is worse, many of
nique that enabled them to appropriate one thing that they and their
them are spies, and attack people in the woods and on the roads, bring
countrymen otherwise possessed in common. This fact lent their
thieves and murderers. One may presume that it is there that they lead
mans crimes some originalitv. Before seizing any precious goods
their dissolute life. When they have spent all their money, they go away
from the burghers, clerics, and aristocrats among whom they lived,
without anything at all; sometimes they abandon even their clothes,
the fifteenth-century Burgundian bandits had taken hold of the lan­
and they return soon afterward, mounted, dressed, and full of gold and
guage. They had made of the common speech of their time a jargon,
silver, as has been said.4
“ an exquisite language,” as the city authorities relate, “ that other
people cannot understand” (un langaige exquis, que aultres gens ne It may be safely supposed that all these vicious acts had been
scerent entendre) . 1 repeatedly attested before the appearance of the depredators tried
Chief among the expressions of obscure meaning that the vaga­ in 1455. Doubtless for this reason, the municipal papers dwelled less
bonds had forged w as the name by which they were known to each on the fact o f the Coquillars’ crimes than on the special means that
other: “ Coquillars," “ People of the Shell,” or, as one might also say they employed to commit them. The legal documents leave no doubt
today, “ Shellsters.” 2 The extracts from the manuscript published by that this means was linguistic and, more exactly, nominal in char­
Schwob opened with a note on this one term. “ It is true that the acter. Jean Rabustel, public prosecutor and clerk of the court of the
aforementioned persons have among themselves a certain language viscountcy o f Dijon, made this much clear at the opening of his sum­
in jargon [certain langaige de jargon] and other signs by which they mary of the trial: “ Every trickery of which they make use has its name
come to recognize each other; and these bandits call themselves in their jargon, and no one could understand it, were he not of their
Coquillars, which should be understood in the sense that they are number and compact, or if one of them did not reveal it to another.” 5
‘the Companions of the Shell,’ who, it is said, have among them a As a source of evidence, Jean Rabustel cited “ Perrenet le Fournier,
King, called ‘the King of the Shell.’” 3 Only after establishing this barber dw elling in Dijon, of the age of thirty-four years, more or
preliminary identification did the manuscript relate in detail the less.” Perrenet testified that he willingly sought out the People of
various crimes for which the Coquillars had been condemned: the Shell, not, o f course, to join their perfidious company in earnest,
but to “ see certain secrets, so as to guard against deception were
It is also true, as has been said, that some of the aforementioned
he to find h im self where evil seed was sowed.”6 Thus the barber
Coquillars are breakers of locks, cases, and coffers. Others are trick­
could bear witness to the fact that the Coquillars “name properly,
sters and rob people by exchanging coins for gold, or gold for coins, in their language, all the matters of their sect, which matters were
or by buying items of merchandise. Others make, bear, and sell false revealed to him by many of them, who did not distrust him, since
threads and goldlike chains. Others bear and sell or make false stones, he feigned to be as subtle as they.” 7 His findings were confirmed by
claiming they are diamonds, rubies, and other precious stones. Others other sources, such as “Jean Vote, known as ‘the Auvergnac,’ quar-
sleep in an inn with a merchant, steal from themselves and from the ryman dwelling in Dijon, of thirty-six years of age, more or less.”8

21
2o
 DARK TONGUES COQUILLARS

The Auvergnac testified to the tact that to each o f the custom ary
a murderer. A desrocheur is someone who leaves nothing to the person
names tor their tricks and crimes there corresponded, in the special
he robs.... A blanc coulon is someone who sleeps with a merchant or
language of the Coquillars, a hidden name, whose sense the brigands
someone else and robs him of his money, his clothes and everything
took pains to keep to themselves.
he has, and throws it from the window to his companion, who waits
As a rule, the sources relate, “ when they speak in their afore­
below. A baladeur is someone who rushes ahead to speak to a church­
mentioned jargon, and one of them says a bit too much in a place
man or someone else to whom he wants to offer a fake golden chain or
\\ here it seems that there might be people who could harm them or
a fraudulent stone. A pipeur is a player of dice and other games in which
who could betray them, the first among them who takes notice w ill
there are tricks and treachery----Fustiller is to change the dice. They
start to spit, in the manner of a man w ho suffers from a cold, unable
call the court of any place the marine or the rouhe. They call the ser­
to call forth his saliva." As one act of expression turns to another,
geant the gaf f res___A simple man who knows nothing of their ways is a
“ each of the Companions of the Shell then falls silent, and changes
sire or a duppe or a blanc__ A bag is a fellouse To do a roy David is to
the subject, and speaks of something else."9
open a lock, a door, a coffer, and to close it again__ To bazir someone
Yet at times, their precious terms did nonetheless escape them,
is to kill him__ Jour is torture___When one of them says, “EstoJJel ” it
and then onlookers and eavesdroppers could penetrate into their
means that he is asking for his booty from some earnings made some­
secret tongue. The Burgundian manuscript lavished the greatest
how from the knowledge of the Shell. And when he says, “Estof fe, ouje
attention on the lexical information that could then be gleaned.
faugereyl” it means that he will betray whoever does not pay his part.10
After a summary description of the idlers and their violent ways,
Jean Rabustel offered a veritable glossary of hidden terms and This list, w hile selective, suffices to suggest one fundamental
phrases, filling folio after folio with definitions of the expressions fact about the C oquillars’ hidden speech: namely, that it consti­
that the Coquillars had sought to conceal. He conceded that his tuted a w illfu l deformation of the language generally in use in late
efforts were of a synthetic nature and issued in a compendium of medieval northern France. The words contained in the legal docu­
terms that no single Companion o f the Shell, strictly speaking, ments would have been known in large part to the inhabitants of
would have employed. “ The aforementioned and other men, who Burgundy, but not in the meanings lent to them by the Companions
are of the company of the Coquillars,” he explained, “ have in their of the Shell. The rogues, in other words, had selected certain words
language many names, and they do not know all the sciences or and phrases, w ithdrawn their usual sense from them, and conferred
trickeries of which mention has been made,” for while some are upon them a new and impenetrable significance. The procedures by
“ skilled in doing one thing,” others are trained in doing another. which they altered the meanings of terms were multiple. Several ties
First the prosecutor offered an enumeration of the professional might link a M iddle French expression to its homonym in jargon. In
activities into which the class of rogues could be divided: a catalogue certain cases, one may detect a direct lexical tie, as when the croche­
o f noms de metier of the most obscure and iniquitous sort. Then teur, or “ picker,” comes to name “ celluy qui sait crocheter serrures,”
he furnished a lengthy key, proceeding to explain phrases as well “ he who knows how to pick locks.” In other cases, the Coquillars
as names: forged their term s by subtler rhetorical means. Sometimes they
proceeded by metonymy, as when they named the court of justice
A crocheteur is someone who picks locks. A vendegeur is a snatcher of
the “ w heel” (rouhe), alluding to the tool that, if not avoided, might
bags. A beffleur is a thief who draws fools into the game. An envoyeur is
well prove the ultimate instrument of their punishment. Elsewhere,

22 23
 COQUILLARS
DARK TONGUES

incomprehensible tongue.15 In Richard li biaus, an anonymous Old

they w orked by metaphor, as when they called the innocent victim o f
French romance dating from the thirteenth or fourteenth century,
then deceptive arts a dupe, or in other w ords, an huppe, or “ p ig eo n ,”
the word appears in a meaning close to that in which the Burgundian
that »s, a bird to be killed and consum ed w ith re lish .11 Yet in many
officials employ it, indicating a language that is proper to thieves and
cases, the connection betw een the ordinary m eaning o f a term and
accessible only to those skilled in their deceitful arts.16
us technical sense, w hile com prehensible in hindsight, could not be
That M arcel Schwob was drawn to the Burgundian brigands
anticipated by any given ru le . The possibilities o f suggestion, substi­
above all for the speech that they devised may be inferred from
tution, and abbrev iation w ere too num erous. Consider the exclam a-
the title he gave to his essay of 1890: “ The jargon of the Coquillars
tor\ utterance “ Estoffel” in the sense provided by Rabustel. To hear
in 14 ^ 5.” A year earlier, in collaboration with his friend Georges
that single word as shorthand for a proposition, as w ell as a dem and,
Guieysse, he had contributed to the Memoires de la Societe Lmguis-
one would need to apply to it principles o f allusion and truncation,
tique de Paris one o f the first works to bring the methods of modern
according to a pattern that it w ould d ifficu lt, if not im possible,
linguistics to bear on the peculiar idioms cultivated by society’s
to divine.
dangerous classes: a paper titled “ Reflections on French Argot.” 17
The m edieval legal dossier, to be sure, did not take pains to
Schwob presented his study o f the medieval manuscript from the
analyze the various m echanism s that the C oquillars had exploited
archives in Dijon as a natural sequel to that essay. In his introductory
in their inventions. Yet it named the type o f speech to w hich they remarks, he observed that historians of French argot possessed only
gave rise in no uncertain term s, for it designated the bandits’ tongue
a small number o f fundamental texts. “ Until now, the sources of jar­
as a langaige exquis, a discourse literally recherche (the verb exquerre gon” consisted essentially o f a few sets of literary and historical texts
m eaning o rigin ally “ to seek o u t” ) and, by extension , “ n o te w o r­ containing isolated words or phrases, such as the term marie, found
thy,” “ refin ed ,” and “ cultivated.” More repeatedly, the m anuscript in m edieval chronicles, “ which has been incorrectly interpreted,
referred to the idiom o f the Com pany o f the Shell by a richly am big­ and which means h a n g e d ";the word dupe, recorded and glossed, in a
uous noun: “ jargo n .” That m edieval French term and its related police document from 1426, as the name of the prey of scoundrels’
form s (jargoun, gargon, ghargun, gergon, gorgon) had long designated tricks; formulae contained in late medieval farces, songs, and ballads
not so much m eaningful speech as the m aking o f sounds resem bling by such w riters as Eustache Deschamps and Charles d’Orleans; and
it, from anim al noises to babble, prattle, and chatter.12 In the earliest works by the didactic humorists who preceded Rabelais, such as Jean
attestation o f the w ord, which dates to the end o f the tw elfth cen­ M olinet. In his b rief account o f the old founts of jargon, Schwob
tury, gargun, for exam ple, names an act o f intelligible, yet inhuman then named three exceptional sets of texts: the manuscripts copied
com m u nication: chirping. In a fable, M arie de France evokes the by Raoul Tainguy before 1425, which contain works by Deschamps,
counsel at w hich birds “ all spoke in their jargon, affirm in g w ith rea­ Jacques de Cessoles, and Froissard; six ballads in jargon long attrib­
son” (Tuit diseient en lur gargun / e afermoent par resun) . 13Three cen­ uted to Francois Villon; and five ballads in jargon copied in a manu­
turies later, C harles d ’O rleans still em ployed the noun in a related script belonging to the Royal Library of Stockholm, recently pub­
sense, w ritin g that “ there is no beast or bird that does not sing and lished by Auguste Vitu, who believed them to be also by Villon.18
c ry out in its jargon, once the season has cast o ff its mantle o f w ind, When com pared with the legal papers concerning the Coquil­
cold and rain ” (II n y a ne beste n oyseau / Q u'en son jargon ne chante et lars, however, all the documentary sources receded into relative
crie: / Le tems a laissie son manteau / De vent, defroidure et de pluye ) . 14 insignificance, as Schwob him self was the first to observe. “ Thanks
Later, in a secondary sense, the term cam e to denote a foreign and
25
 PARK TONGUES
COQUILLARS

to 4 criminal trial preserved in the regional archives of the Cote- unmistakable, apostrophe to the Company of the Shell: “ Coquil-
d'Or," he declared in his preface, “ 1 can now publish an independent lars enruans a ruel.” 21 And for centuries, few had doubted that the
source, the most important of all those of the fifteenth century strange idiom of these six ballads meant to evoke the calculatedly
(with the exception of the ballads), both for the number o f the impenetrable jargon of ruffians. Yet the status of this curious poetic
terms treated and for the details given over the course of the pro­ language had been the subject of considerable debate.
ceedings."1'* The legal records were unrivalled in their linguistic Villon’s early readers appear to have been inclined to treat the
wealth and, as Schwob indicated, they were comparable only to the discourse of the ballads as a faithful rendering of the speech of
six ballads traditionally attributed to Villon. Those Middle French medieval rogues, despite the possibly troublesome consequences
poems certainly contained at least as many evocations of the obscure that such a reading implied for the biography of the poet. When he
language of idlers as the Burgundian court papers. But it was Jean undertook to produce a comprehensive edition of Villon’s works in
Rabustel, the prosecutor of Dijon, who had taken the trouble to the sixteenth century, Clement Marot, for one, praised the “antique
gloss each obscure name and phrase, providing commentaries and and noble language” of Villon’s great poems, while taking excep­
reformulations of the bandits’ expressions. The prosecutor’s gram ­ tion, for this reason, to the six ballads, which he refused to edit.
matical scruples, in this sense, were unmatched, even by the author “ Concerning the jargon,” he commented, “ I leave all corrections
who possessed the skill to set the secret words of jargon in line and commentaries to Villon’s successors in the art of the claw and
and stanza. crowbar” (Touchant le jargon, je le laisse a corriger et exposer aux suc-
Schwob did not remark upon the unmistakable, if surprising, cesseurs de Villon en l ’art de la pinse et du croq) . 22 Geoffroy Tory, in
coincidence between the interests of the prosecutor and the phi­ the same generation, was no less severe. On the opening page of
lologist, the lawyer and the linguist. It would seem that for him, it his treatise Le champ fleury , he condemned “ three types of men who
was the link between the poet and the bandits that mattered most. struggle and compete to corrupt and deform language”: “skinners of
Editing the papers concerning the trial of the Coquillars, Schwob Latin, jokesters, and jargonneurs” Yet he immediately conceded that
insisted above all on these two points: the antiquity of this source among the third variety, there had been one poet: “ When jargonneurs
of jargon and the fact that it cast a new light on the one poet who, hold forth in their malicious jargon and nasty language,” he wrote,
for him, “achieved the greatest poetic glory in his century.”20 Yet “ it seems to me that they show only that they are destined for the
the truth was that Schwob’s findings bore above all on the works gallows, yet it would be better were they never born. I admit that
by Villon that had traditionally been considered the least glorious Master Francois Villon, in his time, was wonderfully ingenious; yet
of all: not the Testament, the Lais, or any of the memorable single he would have done better had he attended to doing some better
lyric pieces for which Villon was so loved, but the six (and perhaps thing.”23 Villon’s subsequent readers often expressed a similar dis­
eleven) ballads set in the strange and apparently hermetic tongue satisfaction with the unruly ballads, which few editors deigned to
of malefactors. translate into more standard French. Some denied that these poems
Already in 1489, in the first printed edition of the poet’s works, were, in fact, the genuine work of the author of the Lais and the
Pierre Levet had separated these texts from the others, dubbing Testament. Others held that their language was less the record of
them “ Ballades en jargon et jobelin.” It was apparent that they were any historically attested form of speech than a bizarre and Baroque
all addressed to criminals o f different kinds and specialties, not invention: “ some imagining of the poet,” as Lucien Schone argued in
least the second of their number, which opened with an opaque, if 1888, “ rather than the unique monument of a vanished language.”24

26 27
 D ARK TONGUES
COQUILLARS

Schwob $ edition and study of the records of 145^ settled that

than a fortuitous occurrence, which brought one artificial language
debate. In a lecture given before the Academic des Inscriptions et
into proximity, if not in unity, with another. One may, however, also
Belles-lettres in he produced a strictly philological demon­
reason otherwise, if at the risk of seeking a truth where there is only
stration that the idiom of the ballads was, in large part, that o f the
chance. Could there be some hidden link between the two hermetic
Coquillars condemned in Dijon. One might, of course, continue
forms of speech, which makes of verse a kind of idlers’ talk, or jargon
to interpret historical and biographical indices in the poet’s work, some variety o f poetry? It is difficult to elude this question, even if
hypothesizing about the dealings that Villon himself may have had Schwob, for his part, seems to have done exactly that. Because of
with the Company of the Shell. Schwob himself sought to do as his scholarly method, his w riter’s craft, or because few more years
much in his “ imaginary life” of the fifteenth-century poet. But would be left to him, Schwob did not draw out the consequences
this much would henceforth be certain: a fact of linguistic identity of the discovery he made. Today, over a century later, they remain,
could be ascertained. Schwob showed that no fewer than twenty- therefore, still to be unfolded.
four of the obscure terms in the poems had been recorded in the
Burgundian proceedings, where they had been explained, trans­
lated from one region of the language into another. “ These words,”
Schwob continued, bringing his proof by lexical units to a higher
sum, “ being repeated many times, form a total of fifty-eight terms,
with respect to which, from now on, discussion can concern only
matters of nuance and etymology, the meanings having been estab­
lished.”25 “ The fact that such a large number of words are common
to the language of Villon and those of the Coquillars,” he concluded,
“suffices to show that these pieces were truly written in the jargon
of malefactors.”26
There now arose, however, one fundamental question, which
strictly historical and philological forms of evidence, on their own,
could not resolve. Schwob had unearthed the oldest major witness
to the secret idiom of rogues; he had shown, moreover, that it was to
become the medium of a poet’s work. From his research, one could
conclude that “ the art of the claw and crowbar” and the art of verse
had become strictly inseparable, not slowly or gradually, but almost
immediately and within the first years in which, it seems, thieves’
cant was born. That marriage of the two arts could hardly have
been more rapid. Yet what sense might one give to it? One could,
of course, consider the meeting of the idioms of the poet and the
vagabonds to have been as singular as Villon’s oeuvre, the expression
of a single biography or vocation, or both. Perhaps it was no more

28 29
 THE CHAPTER THREE

CANTING P r in c ip le s o f C ant

Academy? O R ^
V il l a n ie s D is c o v e r e d .
W H EREIN IS S H E W N
The Myfterious and Villanous Practices
Of that wicked Crew, commonly known by the Our language contains several words for lesser varieties of speech.
Names of Hecto r i, Trapanner*, Gilts, (ic.
“ Slang” refers to words and phrases that, within a single tongue,
W it h se v e r a t N e w C a t c h e s belong to particular segments of society defined by place, class, and
AND association. Such expressions tend to live short lives, their fates being
tied to fashion. “Jargon,” in its common English usage, alludes, by
S O N G S 1 contrast, to the terms and formulae employed by people who share
A L S O
a certain practice, activity, or knowledge. Those in possession of
A Compleat Canting-Dictionary, both of
old Words, and such as are now most in use, crafts and skills—artisans and doctors, students and teachers-tend
A Book very useful and necessary (to be to use such a type of speech, either knowingly or unknowingly. Yet
known but not practis’d) for all People, there is also a third variety of discourse, distinct from both slang
and jargon. Since the Renaissance, it has been known in English as
“cant,” a term that seems to represent an Anglicized form of the
T h e S e c o n d E d itio n .
Latin word for church song or chant, cantus.1 This word refers to an
London, Printed by f . Leach for Mat. Drew, and obscure form of speech, which is employed above all for the com­
are to be Sold by the Booksellers. 1674 . mission of crime. “ Cant,” as an historian of the subject has recently
written, “goes one step further than jargon. Its primary purpose is
to deceive, to defraud, and to conceal. It is the language used by beg­
gars and criminals to hide their dishonest and illegal activities from
potential victims. It is sometimes used as a disparaging synonym for
jargon and applied to other groups, particularly religious hypocrites,
and occasionally real estate agents.”2
Such summary considerations run the risk of suggesting that as
a rule, within a single language, all three varieties of special speech
are discernible. This seems, however, not to be the case. If one con
siders the ancient cultures of Greek and Rome, it is striking that

31
 DARK T O N G U E S PRINCIPLES OF CANT

trounces of the first two types of lesser discourse can be found. crooks and robbers.6 These are the first signs of the emergence of
Classical literature contains many slang expressions, in Greek and the artificial idiom of malefactors. Philologists have noted others in
Latin, long known to readers of Aristophanes, Petronius, and Plau­ the fourteenth century, for example in fabliaux .7 Yet it is only in the
tus, That the ancient branches of learning, from philosophy to math­ fifteenth century that one encounters a more complete account of
ematics and medicine, developed their own jargons, moreover, has cant. The age of the Coquillars’ trial was that in which the “exqui­
never been in doubt. Many of their technical terms, from “ energy” site languages” of cheats and scoundrels began, for the first time, to
and “square root” to “ interval” and “ melancholia,” remain in use make their strange sounds heard in public.
to this dav. As a rule, whenever a new field of knowledge arises, it Early modernity proved to be an epoch favorable to rogue
quickly develops a distinct lexicon, which in turn provides the mat­ tongues. A few decades after the Burgundian bandits were con­
ter for an unfamiliar jargon. The classical examples, like all others, demned for their nefarious acts, plotted in “ a jargon that no man
illustrate this principle. Yet the case of cant is different. Among the understands,” one Teseo Pini of Urbino wrote his Speculum cerrata-
remnants of the ancient world, one looks in vain for signs of a speech norum, a treatise exposing the ways of beggars who defraud and prey
crafted so as not to be commonly understood. No classical bandits on those so credulous as to believe in their apparent misery.8 Pini
seemed to have forged a secret tongue, no antique ruffians a hidden identifies thirty-nine distinct classes of false mendicants and, in the
idiom. The Amazonians may have invented a language all their own, conclusion to his book, offers a glossary of the many obscure terms
as Herodotus, for one, reports, but no ancient author maintains that that they employ, pairing each curious Italian word or phrase with
they did so to conceal their ideas from those hostile to their com­ a standard Latin expression.9 Only a few years later, there appears
pany.3 The least one may say is that if ever there did exist classical the most famous of all early modern accounts of the ruses of men­
idioms similar in form to thieves’ cant, their traces are lost today. dacious idlers: The Book of Vagabonds and Beggars (Liber vagatorum,
The first references to a secret tongue of malefactors date to the also known as Der Betler Ordern), thought to have been composed
end of the twelfth century, when various authors begin to write of sometime after 1509 by an author who, withholding his full name,
obscure “jargons.” The Old Occitan grammatical treatise Donats signed simply as one “expert in roguery”: “ Expertis in Truffis.”10 His
proensals, thought to have been composed in the second half of the treatise contains both a typology of cheats and a lengthy dictionary
century, includes one passing, pregnant reference to gergons, which of their expressions. Its anonymity was no hindrance to its circula­
it glosses in Latin as vulgare trutanorum, “ beggar’s” or “ thieves’ tion, for it was soon reproduced in verse as well as in prose, and in
talk.”4 Only a few years later, Jean Bodel’s play Le jeu de saint Nico­ 1527-15-28, it appeared in print, with a preface by no less an authority
than Martin Luther.
las, from the beginning of the thirteenth century, sets on stage two
“ I have thought it a good thing that such a book should not only
robbers who, conversing in a tavern with a disreputable young man,
be printed, but that it should become known everywhere, in order
exchange phrases that, while clearly French, have long defied inter­
that men may see and understand how mightily the devil rules in
pretation. They may constitute the earliest attestations o f criminal
this world,” Luther explains in his introduction. He then advances a
jargon in French, although historians of the language continue to
hypothesis concerning the lexicon of malefactors: “ Truly, such Beg­
disagree about their relationship to modern argot.5 In the same
gars’ Cant has come from the Jews, for many Hebrew words occur in
years, the unknown author of the romance Richard li biaus recounts
the Vocabulary, as any one who understands that language may per­
how its eponymous hero, stepping into a lair o f thieves, sought the
ceive.” " While conceding that there do exist genuine paupers whom
advice of a knowledgeable esquire who “ knew all the gargoun” of

33
32
 DARK TONGUES PR IN C IPLES OF C AN T

to
w n $ and villages should help, Luther declares that “ outlandish and
the very first hour, that they take upon themselves the names of knichin
Strang beggars" employing unknown words ought not, under any
coes, till they are grown rufjlers, or uprightmen, which are the highest
circumstances, to be tolerated. By way o f conclusion, he reveals that
in degree among them .15
he himself has had bitter experience of their ways: “ O f late years," he
writes, “ 1 have myself been cheated and befooled by such tram ps and In English, D e k k e r’s w ork on the ru ffle rs ’ “ new speech or lan ­
liars more than I should like to confess. Therefore, w hosoever hear gu age” w as h ard ly to be the last. From R ichard H ead’s The Eng­
these words, let him be warned, and do good to his neighbor in all lish Rogue o f 16 6 5 and Canting Academy o f 1673 to the books on
Christian charity, according to the teaching o f the com m andm ent. cant published in the last years o f the seventeenth century by one
So help us God!” 12 “ B .E .G e n .,” from Fran cis G ro se ’s Classical Dictionary o f the Vulgar
In the following centuries, the interest in rogue tongues con tin­ Tongue o f 1731 to the essays on “ cant and flash language” published in
ued to grow. In England, the literature on the subject begins in 1557, Britain and A m erica in the late eighteenth and early nineteenth cen­
when Thomas Harman publishes his Caveat or Wareningf o r Common turies by such authors as H um phrey Tristram Potter, Jam es Hardy
Cur$etors.n Here Harman dates the appearance o f the secret language Vaux, Peirce Egan , and Jo h n B ee, an entire literature on the special
of criminals-which he calls “cant” or “ peddler’s F ren ch ” —to thirty speech o f crim in als arose and cam e to m atu rity.16
years before the time when he composed his book. He claim s that The other cu ltu res o f E u rop e w itn essed sim ilar developm ents
cant was the invention of an individual, although he also suggests in these centu ries. In Fran ce, H enri Etienne m arvels in 1566 at the
that the impenetrable tongue might be attributable to the G ypsies, refinem ent o f the new “ jarg o n ” ; he speaks o f the perfection o f a lan­
those “wretched, wily, wandering vagabonds callin g and nam ing guage long ru dim en tary and suggests that V illon h im self would have
themselves Egyptians, deeply dissembling and long hid ing and cov­ had much to learn from his crim in al contem poraries. He declares
ering their deep deceitful practices.” 14 h im self “ ce rta in ” that the language in w hich thieves converse in his
Thomas Dekker continued Harman’s research in a series o f publi­ day “ was never yet o f such p erfe ctio n ___W ithout having taken any
cations beginning with The Bellman o f London, which dates from 1608. great pains, these m en have so enriched their jargonlike language,
Dekker distinguishes between “ Cursitors, alias V agabonds,” “ Fay- and so w ell studied it, that w ithout any fear o f being uncovered by
tors,” “Robardsmen,” “ Draw-latches” and “ Sturdy B eggars,” yet he those not o f their profession, they can negotiate calm ly am ong them ­
argues that they all share a single language, cultivated for the com m is­ selves.” 17 The first extended study o f the subject in French dates to
sion of their crimes. This, he writes, is a matter o f material necessity: 15 9 6 .18 It w as fo llow ed , a few decades later, by an influential w ork,
Jargon, or The Language o f Reformed Argot, as it is Now in Use among
It was necessary that a people (so fast increasing, and so daily practic­
the Good Poor and by a host o f diverse discussions o f the cryptic lan ­
ing new and strange villainies) should borrow to themselves a speech,
guage o f various tricksters (not least fisherm en and prisoners) that
which (so near as they could) none but themselves should under­ continued to appear throu ghout the seventeenth, eighteenth, and
stand. And for that cause was this language (which some call Peddlers’ nineteenth cen tu ries. In 1837, E. F. V idocq, once an infam ous crim i­
French) invented, to the intent that (albeit any spies should secretly nal and later the fam ous ch ief o f police in Paris, set a new scholarly
steal into their companies to discover them) they might freely utter standard in the field. He published a philological monument to the
their minds to one another, yet avoid that danger. The language there­ w orld o f thieves, Les Voleurs, w hich soon proved an indispensable
fore of canting, they study even from their infancy; that is to say, from source to both Balzac and V ictor Hugo.

35
34
 DARKTONGUES PRINCIPLES OF CANT

The interest in thieves’ cant in these centuries traversed Western can here clearly be discerned. It has been observed that scholars of
Europe. In Iberia, the hidden language of wrongdoers acquired the language tend to approach their objects as if they were immutable,
special name of germania. It resounded, in different ways, in fictional even when they are in fact in variation; linguists thus treat living
as well as scholarly works, marking the literature of the picaresque languages as if they were dead, grasping their rules in the artificial
with particular force. Hence the special terms employed and inter­ fixity of a single moment, if only to write their grammars. When
preted in the narrative of Quevedo’s hapless Swindler; hence, too, it so happens that scholars do speak the languages that they study,
the secret expressions learned by Rinconete and Cortadillo in C er­ they regularly approach them, for reasons of scientific method, as if
vantes's exemplary tale of initiation to the world of roving thieves. they were foreign to them. In each case, the procedure remains the
In Portugal, the term of choice was calao, which evoked the special same: knowing subjects must exclude themselves from the realm of
tongue of the Gypsies of Alentejo.19 In German, one would speak of what they wish to know, precisely to bring it into focus in the light of
“red Welsh,” Rotwelsch, or, alternately, to employ a word of differ­ scientific objectivity. This systematic self-exception of the linguist is
ent origins, “thieves’ language,” Gaunersprache. 20 The great monu­ perhaps nowhere as apparent as in the study of cants, which has the
ment to the study of this obscure idiom dates to 1858. This was the peculiar consequence that each increase in the scholar’s knowledge
year when Friedrich Christian Benedict Ave-Lallem ent, scientist, brings about a corresponding decrease in the effectiveness of the hid­
explorer, and chief of police in the city of Lubeck, adduced new den tongue. Hence the peculiar animus that writers on jargon tend
philological evidence in favor of Luther’s old claim, aiming to dem­ to express, which insistently, if awkwardly, accompanies the scholarly
onstrate that the German language of criminals was, in fact, simply tone in the analysis of thieves’ speech. From Jean Rabustel to Martin
Yiddish.21 Italian authors write of an idiom “ w ily” by nature, f urbesco Luther, Ave-Lallement, and Vidocq, the interest in the idiom of cheats
being the long-standing name of the special tongue in which cheats is in turn denunciatory and scientific, aggressive and neutral, for rea­
plot their illicit acts. The literature on the subject stretches from sons involving the epistemology of the knowledge of secret languages.
the Renaissance to the twentieth century, when Alfredo Niceforo Several centuries of research have made it possible to grasp the
published a series of influential works, in French as well as Italian, fundamental characteristics of such forms of speech. They are less
on the language of the pernicious fringes of society. They are well languages in the full sense of the word than “ special languages,”
represented by his comprehensive treatise o f 1897, Jargon among that is, forms of discourse that result from the modification of given
the Normal, the Degenerate, and the Criminal, whose well-ordered tongues according to rules known only to a select few. It suffices to
chapters consider the types of argot in use in “ normal associations,” apply certain special principles to the vocabulary and grammar of a
“the amorous couple,” “the lesbian, onanist, and pederast couple,” language to obtain such idioms, which, while altogether dependent
“lower social stratifications,” “ feminine milieus,” and “ between the on existing linguistic systems, may be effectively impenetrable to
prostitute and her souteneur! ' 22 the untrained ear. Strictly speaking, they constitute not so much
From early to later modernity, one basic structure in the study of secret languages, therefore, as secret uses of languages. At times,
they have been called “anti-languages,” for they serve the interests
cant has remained in place. Publications on secret tongues can only
of “anti-societies” that strive to protect themselves from threat­
be at odds with them. Since the jargons of cheats demands secrecy,
ening majorities.23 From a linguistic point of view, they may be
any attempt to illuminate their elements, rules, and mechanisms
dubbed “ parasitic languages,” as Marcel Cohen long ago maintained.
undermines their force and validity. Despite the curiosity of the
“ Given any linguistic unit (ordinary French, the dialect of an Alpine
case in question, a fundamental characteristic of all linguistic study

37
36
 PARK TONGUES PRINCIPLES OF C A N T

valley)," Cohen reasoned, one may conceive the existence of idioms of linguistics, Jean-Claude Milner, for these reasons, has gone so
that will borrow “ their phonetics, their morphology, and their syn­ far as to make of this an axiom in the science of language: absolute
tax from it,” using them in their own way and for their own ends, synonyms do not exist; or, positively stated, whenever two terms are
at times submitting them to an “abnormal transformation.” 24 Such in use, their lexical meanings must be distinct.26
idioms may he “ incomprehensible to the noninitiate,” being con­ Slang, jargon, and cant would all seem to revoke that principle of
ceived as the instruments of defense or attack.25 Yet it is certain that speech. Pointing to sets of terms signifying the same acts, objects,
such linguistic parasites, like all others, possess their own means affects, and states, Schwob and Guieyesse argue that in special lan­
of subsistence. These involve both the sense and the sounds of the guages, every word can produce “an artificial double” that may, in
languages by which they live. fact, become interchangeable with it. It suffices to deform an ordi­
The simplest and most obvious modification from which a “ spe­ nary expression by the rules of the parasitic idiom and, in a second
cial language” may be born is that by which a given lexicon comes step, to insert it into a hitherto unknown “synonymous series.” An
to be increased, diminished, or deformed. Cryptic jargons, like all example may be of help. In nineteenth-century argot, the French
others, are made of new words and phrases, and the first step in their word bonneteau, the name of the “ three-card trick” or “monte”
study consists in the enumeration and examination of formulae. At played by cheats, came to be reduced, by apocopation, to bonnet, a
times, the usual vocabularies of languages w ill be reduced, such that noun already in use in the language in the sense it also possesses
a single expression comes to convey more meanings than it would in in English: “ hat, cap, usually soft and without a brim.” Henceforth,
its ordinary usage; at other times, by contrast, vocabularies w ill be the speakers of jargon could take this word and forge from it a set of
augmented by new expressions. Words and phrases may be forged by doubles that, while produced on the basis of a metonymic associa­
tion with the French word bonnet, would all share the lexical sense
the most varied of rhetorical means: metaphor, metonymy, antith­
of the word bonneteau. Schwob and Guieyesse report, for example,
esis, and epithets, most obviously, can all have roles to play.
that bonneterie (“ hosiery” ) and lingerie could function as semanti­
Among such lexical elements, moreover, one may discern a
cally equivalent to bonnet in its secret sense, meaning “three-card
relation found in such special tongues alone. Marcel Schwob and
trick”; moreover, bonneterie and lingerie would, in this hidden usage,
Georges Guieyesse were perhaps the first to identify it when, in
also function as semantically equivalent among themselves.27 The
their 1889 “ Study of French Argot,” they maintained that slang
two terms were no more and no less than artificial doubles of
admits of a curious mode of signification that they baptized “ deriva­
the phonetically altered term from which the “synonymous series”
tion by synonymy” (derivation sy nony mique). To grasp the full force
had begun: bonnet, as abbreviated form of bonneteau. Moreover, the
of their argument, one must recall that implicitly, if not explicitly,
doubles, in turn, could lead to other synonyms and so to other series
scholars of language often assume that no two words in a language
of artificial, yet equally regular forms of designation. At any point
possess exactly the same signification. Admittedly, the dictionary
along the chain of interchangeable terms, a new series might begin.
and the thesaurus indicate that within a single tongue, one may
“A metaphor can cause a cloud of synonyms to fall about it, like the
define the sense of a given word by means of other expressions. Yet
mushrooms that, exploding, produce a haze of spores, destined to
the synonymy that such procedures establish is only ever relative. perpetuate their own kind.”28
If the lexicographic analysis of a certain term succeeds, it reveals The principles of cant extend, however, beyond the level of
the traits by which that term may be distinguished from those of an vocabulary. The secret use of a language may also involve the
apparently similar semantic value. In his analysis of the principles

39
38
 DARK TONGUES
P R IN C IP L E OF CANT

alteration of its sound shape. It suffices to think o f operations of syl­

“ special languages,” built from different grammars, follow similar
labic scrambling, suffixation, infixation, prefixation , apocopation,
principles.34 Each time, such forms of speech can prove impenetrable
ami truncation, all of which may he conjoined in different ways and
to the unadvised speakers of the idioms of which they nonetheless
in accordance with rules too numerous to anticipate.
constitute regular, if subtle, deformations.
Restricting oneself to the special varieties o f the French language,
Yet the special alterations of given languages can he more com­
one might recall the idiom called verlan, w hich rests upon the inver­
plex. Hidden plays o f meaning and sound may also he conjoined.
sion, according to fixed patterns, o f the phonological and syllabic
Pierre Guiraud has called attention to the practice of what he calls
structures defining common words. Envers (“ backw ards” ) becomes
“ homonymous substitution” in twentieth-century French argot,
verlan; feuille (“ leaf” ), y euffe; arabe, beur, and, by virtu e o f a further
whereby one expression comes to be replaced by another of a similar
inversion, which does not return the transform ed term to its initial
or identical phonological form .3S It is this operation that allows ore
form, reube.^ One may also cite the exam ple o f javanais, the cryptic
to apply the French word blanchisseuse (“ female launderer” ) to those
special tongue of nineteenth-century French prostitutes, invented,
clients who, leaving a shop without making a purchase, promise the
if one believes the historians, for protection from m ale clients. This
seller that they w ill return. A fact of pure homophony justifies this
idiom arises from the insertion, after each syllable o f a common
semantically unmotivated usage, for in French, the statement “ I will
French word, of an infix that begins w ith the consonant v and that
return” (je repasserai) is phonologically identical to the sentence, “ I
usually assonates with the preceding vow el. U nder these artificial
w ill iron” (je repasserai), and this assertion, in turn, may be taken as
constraints, entire sentences, with all their elem ents, may be trans­
synonymous to the declaration, “ I will launder” (je blanchirai). On
formed.30 Hence Queneau’s “ Javanese” exercise in style: “ Un jour,
account o f an identity o f sound, two units of differing sense thus
vers midi-vingt, sur un autobus de la ligne S, j ’apercu s un jeune
come to be substitutable: “ returning” may be expressed as “ laun­
homme avec un long coup et un chapeau entoure par une ficelle au
dering” and, by extension, as “ ironing.” The artifice at work in the
lieu du ruban” becomes “ Unvin jovu r vevers m ividin suvur unvin
special tongue in this case is that of the pun, which sets phonological
vautobovus deveu lava livigneve esseve, jeveu vapervercuvus unvin
equivalence against unexpected semantic divergence.36
jeveunovomme vavec unvin lonvong couvu evet unvin chavapo-
One may also think of the related English-language procedure
veau envantouvoureve pavar uvune fivicevelle oveau lieuveu deveu
commonly referred to as “ rhyming slang,” which has variously been
ruvubanvan.”31 Or, in the “ back slang translation” proposed by Bar­
attributed to East London Cockneys and Australian and United
bara Wright: “ Unway ayday aboutayay iddaymay onyay anyay essyay
States prisoners, among others. This curious operation consists in
usbay Iyay oticednay ayay oungyay anmay ithway ayay onglay ecknay
two basic steps. First, one replaces a word by an entire phrase whose
andvav ayay athay enyayircledcay ybay ayay ortsay ofyay ingstray
ending rhymes w ith the word in question; then, an element of
invayeadstay ofyay ayay ibbonray.” 32 Then there is largonji de louch- the rhyming phrase may be omitted. The remnant of the rhyming
erbeme, the secret jargon of nineteenth-century Parisian butchers, sequence w ill be grasped as synonymous with the substituted term.
the rules of which are more com plex. H ere, in all w ords begin­ In a first step, for instance, the word “ stairs” may be replaced by
ning with consonants, the first phoneme is erased, replaced by 1, the phrase “ apples and pears,” on the basis of the rhyme between
and moved to the end of the phonological sequence; then, to that “ stairs” and “ pears.” In a second step, “ and pears” may be dropped.
new element, a new and meaningless m orphem e is appended. Thus Through a procedure o f suggested rhyme and partial deletion, a new
jargon becomes largonji, and boucher, loucherbeme. 33 Countless other synonym w ill be forged: “going up the apples” will signify “going up

40 41
 PARK TONGUES
PRINCIPLES OF C A N T

the stairs."17 So. too, given the rhyme betw een “ m oney” and “ bees
idioms may be datable, in Europe, to the mid-fifteenth century. Yet
and honey.‘* one may say simply “ bees,” for in its truncation, the part
as appropriations of language that set sound and sense against each
will summon the whole from which it was d raw n .38 Im perceptibly,
other, simultaneously revealing and concealing what they suggest,
the rhyme between “ honey” and “ m oney” still sounds. As a fact of they are far more ancient. Before the rise of late medieval rogues,
phonological identity enables a lexical substitution that is partially before the emergence of Renaissance outcasts and more modern
elided, in turn, by a principle of syntactic truncation, the procedure outliers, secret languages played a crucial role in the work of a dif­
quickly gives rise to the most enigmatic o f equivalences.39 ferent company of speaking subjects. Their old and foreign names
Like all arts, such procedures lend them selves to m ore than a today are largely forgotten: aeidioi, brahmana, vates, troubadours
single use. Techniques o f appropriating languages are appropriable and trobairitz, skalds, riddlers, and scribes. Their titles varied from
in many ways. One may imagine why historians have w ished, at dif­ epoch to epoch, culture to culture, vocabulary to vocabulary. Yet in
ferent times, to attribute such forms o f cryptic speech to as diverse a each case, if in different ways, these versifiers and scriveners were
cast of socially marginal characters as beggars, butchers, fisherm en, perhaps the first to seek to make one speech divide in two, and to
prostitutes, and prisoners. Yet the principles o f cant, no m atter how summon, within the limits of a shared language, an insistent and
varied, subtle, and complex their historical instantiation, are united startling opacity.
in the minimalism of their material. Secret languages do nothing but
recapitulate the various units that are operative in a gram m ar, from
the phoneme to the syllable, the word to the phrase and the com ­
plete sentence. As arts of speaking that order and reorder the parts
of language to w illful ends, the procedures o f such herm etic idioms
may call to mind the techniques o f rhetoric. Yet the uses o f special
languages far exceed the aims of “ speaking w e ll” (bene loquendi) that
rhetoric, in its classical definition, strives to serve. W ith greater
justice, one might view the opaque expressions o f “ anti-societies”
as close, in kind, to literary discourse in the sense given to it by
Viktor Shklovsky. Like the w riter’s speech, the tongue o f cheats and
scoundrels troubles the patterns that define known expressions. It is
speech become unfamiliar.40
Yet one may also take a further step. To the degree to which
thieves’ cants join and disjoin the phonological and semantic levels
in their procedures, they come close, in structure, to the v a ri­
ety o f literary discourse that Valery once defined as a “ prolonged
hesitation between sound and sense.”41 That variety is poetry. This
proxim ity points to an aspect of criminal jargons that may not have
received the attention it deserves. As arms or shields employed by
the dangerous classes o f modernity, gram m atically regular cryptic
43
42
 CHAPTER FOUR

Signs

The vernacular literatures of medieval Europe are in several respects

perplexing objects o f critical study. They are written in idioms com­
monly taken to constitute the first attestations of the Romance, Ger­
manic, and Slavic languages that are still in use today. In their forms,
they are generally believed to represent early examples of what
later became modern literary genres. One might, therefore, infer
that medieval literary texts possess a clarity of language and form
that renders them more immediately intelligible than the works of
Graeco-Rom an antiquity. Yet the literature composed during the
medieval periods o f modern languages remains, in several respects,
impenetrable to the contemporary eye. This curious fact has been
noted several times. It has led some scholars to write of the simulta­
neous “alterity and m odernity” of medieval writing; it has prompted
others to go so far as to assert that the vernacular works of the
Middle Ages, despite their relatively recent date, are obscurer than
those of antiquity.1 “ Medieval poetry!” Paul Zumthor exclaimed in
1963. “ Let us admit it: in its most perfect forms, the civilization of
the tenth, eleventh, and twelfth centuries is more foreign to ours
than that of classical Rom e.”2 Buried in the tongues of the first poets
of the modern world lie signs that resist even the most erudite of
interpretations.
One may take, as a first example, the oldest Romance literature
in verse. This is the corpus of largely lyric works attributed to poets
of the eleventh, twelfth, thirteenth, and fourteenth centuries who,
in a geographical range stretching from northern Iberia to southern

45
 p a r k TONGUES SIGNS

France and northern Italy, called themselves "troubadours”: accord­ account of whom he has begun to sing and the poem, therefore, has
ing to the most accepted etymology, "finders” or "inventors,” named begun to be, he refrains from offering of her anything resembling a
after the Old Occitan verb trobar, "to find.” Their works have often description. From the earliest attestations of the troubadour tradi
been considered to define, with the simplicity proper to beginnings, tion to its conclusion, from poet to poet, style to style, and school
features that would soon become the hallmarks o f the modern to school, this fact remains constant. Despite the variety of terms
brie: verse structure, measured by a regular number o f syllables, and tones with which they conjure their love in words and music,
ith rhymes; grammar, characterized by a discourse in the voice
w the medieval “ finders” deny their public any representation of the
of a first person singular; and topics, involving a being who speaks physical, social, historical, or biographical characteristics of their
and sings to evoke the joy and pain of his amorous passion. Yet as ladies. From a modern perspective, it has been observed that the
Alfred Pillet observed long ago, "the troubadour texts, in them­ medieval treatments, for this reason, exhibit a somewhat dissatisfy­
selves, present... more difficulties than any poetic texts of the old ing uniformity. “ One might think,” Andre Moret commented of the
Romance languages.”3 troubadours’ Middle High German followers, the Minnesanger, “ that
This fact is all the more striking if one considers that it cannot they all wrote in praise of the same woman.”9
he explained by reference to problems of an exclusively textual or Similar claims could be made about the being who speaks in
linguistic kind. The manuscripts containing the troubadour songs, these poems. Although the troubadours sometimes evoke the condi­
which date mainly to the late thirteenth and early fourteenth centu­ tions in which they have come to meet their ladies or learn of their

ries, are relatively few in number and have been edited, sometimes renown, and although they depict the pleasure and sorrow that their
love has brought upon them, they systematically avoid naming them­
repeatedly, with great care and exactitude. Scholars possess a reli­
selves in the principal portions of their poems. Many troubadours,
able critical repertory of the entire corpus, estimated to consist of
indeed, never introduce themselves by name at all.10 Those who do
2,542 lyric works.4 Their language, once called Old Provencal and
so restrict their self-identifications to the abbreviated stanzas with
now known as Old Occitan, has been subjected to thorough philo­
which they conclude their works. Here, in the lesser units of poetry
logical and linguistic analysis, as a whole and in its details. Verse
that the medieval theorists of Occitan poetry call tornadas, some
forms have been identified and catalogued.5 Yet despite the scholarly
authors, such as Bernart de Ventadorn, reveal their names occasion­
grasp of the troubadour lyric, there are elements in the medieval
ally. Other troubadours, such as Marcabru and Cercamon, identify
craft that remain to us opaque. To invoke a rhetorical and stylistic
themselves more often. There are a small number of poets, finally,
term used by the troubadours for some of their compositions, one
who name themselves in these stanzas almost regularly. Sometimes
may say that in certain respects, the entire corpus seems sealed, as
they join their names to carefully crafted epithets. The exemplary
if in accordance with an art of trobar clus, or “ closed finding.”6
case is that of Arnaut Daniel, whom Dante’s Guinizelli calls “the
If anything in this literature is certain, it is that at its center lies better craftsman of the mother tongue” (miglior fabbro del parlar
the figure of the poet’s beloved. She is the “ point,” as one literary materno) . 11 This twelfth-century troubadour concludes a song by
historian has written, “ toward which the troubadour constantly naming him self in three famously enigmatic lines: “ I am Arnaut,
turns his eyes.”7 For her, the troubadours generally employ the who hoards the wind / And chases the hare with the ox / And swims
Old Occitan word domna: a term derived from the Latin domina, against the swelling tide” (Eu sui Arnautz gamas l ’aura / e catz la lebre
which, in feudal parlance, denotes a married lady.8 Yet it is strik­ ab lo bou / e nadi contra suberna) . 12
ing that although the troubadour always evokes such a beloved, on
47
46
 D ARK T O N G UE8 SIGNS

As arule, it is in these concluding portions o f the poem that female being. W ithin these limits, the troubadours form a vast array
the Occitan poets also refer, in a single, elusive invocation, to their of elliptical titles and abbreviated descriptions. The poet may name
beloved ladies.1' Vet it is striking that even when the concluding the addressee o f his poem by a formula of admiration, such as Be;
units of the poem seem to constitute “ stanzas of address” (Addressen- Esgar, “ Fair Loo k,” Bel Rai, “ Fair Ray,” Bel Vezer, “ Fair Sight,” or
rernjJenh the medieval authors refrain from naming their domnas, Plus Avinen, “ More Charm ing.” He may, instead, build a name from
at least by any ordinary nominal means.14 The troubadours avail the expression o f an affective state, evoking Mon Desir, “ My Desire,”
themselves, instead, of openly cryptic turns o f phrase. The first Fin Joy, “ Pure Joy,” Mais d ’Amic, “ More than Friend,” Mielhs de Be,
modern scholars of the literature immediately observed this fact. “ Better than Good,” Gai d Amors, “ Drunk with Love.” He may allude
They commented that the Occitan poets systematically substitute to his own wishes and expectations, calling out to Bel Esper, “ Fair
for the name of their beloved “ an allegorical name o f the lady” Hope,” Bel Respeit, “ Fair W aiting,” and Conort, “ Comfort.” He can
{nomen alleqoricum dominae), as Alfred Kalischer observed in 1866, also craft a name from the designation of his pleasure, addressing Bel
introducing, in place of an existing designation, a “ pseudonym” or Gazanh, “ Fair G ain,” Mon Plazer, “ My Pleasure,” and Gen conquis,
“ disguise name” (Versteckname).15 Guilhem de Peitieus, the earliest “ Nobly Conquered.” More rarely, he will summon a figure named
recorded troubadour, also known as W illiam o f Aquitaine, thus through bitterness or reproach, alluding, for example, to a lady
evokes his lady twice. Once he calls her Mon Esteve, “ My Stephen” ; named Tort n ’avez, “ You Are Wrong.” 18
once he names her Mon Bel Vezi, “ My Fair Neighbor.” 16 Adm ittedly, In the fourteenth century, the principles of the troubadour poetic
not all the troubadours choose to designate their loves in such a craft came to be recorded by a company of poets and grammarians
form. The twelfth-century poet Cercamon, for exam ple, appears from Toulouse. In a monumental treatise of Occitan poetics titled
to avoid the practice altogether; none of his poems offers any name The Laws o f Love (Las leys d ’amors), or The Gay Science (El gay saber),
for the lady around which, nonetheless, they turn. Yet the Occitan these medieval scholars gave a name to the troubadour procedure of
minstrels who designate their domnas follow Guilhem. In the final substituting one name for another. Forging a technical term from
lines of their work, they offer a noun, title, or phrase as a cipher of the common vocabulary o f the Old Occitan language, the members
the name that they withhold. of the Gay Science called the “allegorical names of the lady” senhals.
In its form, the procedure obeys certain grammatical constraints. That term may be translated in several ways, not least by the words
Edoardo Vallet has shown that as a rule, the designations adhere to “ insignia,” “ sign,” “ m ark,” “ seal,” “ proof,” “ signal,” and “ portent.” 19
one of four linguistic forms. They may consist of a single noun, such In the section o f their treatise devoted to the formal structure
as Dezirat, “ Desired One,” or a proper name drawn from myth or of troubadour verse, the fourteenth-century waiters explain that

literary history, such as Tristan; a nominal phrase, consisting o f a poems may be divided into stanzas (or cobias) of equal numbers of

noun and qualifying adjective, such as Bel Cavalier, “ Fair K night,” lines that end, for the most part, with a shorter strophic unit that is

or Fin Joi, “ Pure Joy”; a comparative prepositional phrase, such as called tornada. “ Let it be known,” they declare, “ that one can place
and apply a senhal in the tornada and that everyone must choose one
Mielhs-de-Domna, “ Better-than-Woman”; or, finally, a verbal phrase
for himself, without doing any wrong to another, that is, that he
(usually based on a present-tense verb, conjugated in the second
must not put into his work or appropriate from another any senhal of
person, singular or plural), treated as a single name, such as Dreit
which he knows someone else has made use” (devetz saber. .. pauzar
n’avez, “ You Are Right.” 17 It is remarkable that such appellations, as
et aplicar a so senhal. lo qual senhal cascus deu elegir per si. sesfa r tort J
a rule, are masculine in gender, even when they appear to denote a

48 49
 PARK TONGUES
SIGNS

a autre, soes que no vuelha en sos dictatz metre et apropriar aquel senhal
senhals invoked in the Old Occitan songs remain opaque. The vidas
que saubra que us autres fa).20
and razos suggest, moreover, that the medieval designations should
One learns more about these “signs” from the anonymous works
not be understood as referring solely to the poets’ beloved ladies,
in prose that often accompany the troubadour poems in medieval
for at least some appellations are of another kind. The “ Life” of the
manuscripts. These texts date from the thirteenth and fourteenth
twelfth-century troubadour Bertran de Born, for example, explains
centuries and are known to the literary tradition as vidas and razos,
that this poet devised a set of titles to designate some of his noble
“ Lives” and “ Reasons.” They comment, in different ways, on the
contemporaries. “ He called the count of Brittany Rassa, the king of
origins and sense of the Occitan poems. Often they furnish the
England Yes-and-No [Oc e no] and the young king, his son, Seaman
modern reader with precious information about senhals. In the one- [Marinier] . ” 26 Building on such suggestions, critics have hypothesized
page “ Life" of the twelfth-century poet Raimbaut de Vacquerias, that it is likely that other senhals in the corpus are of this nature.
for instance, we learn that this troubadour “ fell in love with the In an important essay on Peire Vidal, Ernest Hoepffner thus raised
sister of the marquis, who had the lady’s name of Biatritz, and who the question of the meaning of the name “ Castiat,” “ Punished,”
was the wife of Enric del Caret. And in his songs, he called her which one finds often in the work of this troubadour from Toulouse.
‘Fair Knight’ [et appelava la en sas cansos Bel Cavaliers]”2' In a razo or Through a careful reading of the poet’s work and the extant sources
“ Reason” relating to Bernart de Ventadorn’s famous song about a relating to his life, the philologist argued that the Castiat against
lark rising and falling in the air, we read that Bernart “ called his lady whom the troubadour rails in verse was, in truth, the count of Tou­
Lark...and she called him Ray ” (apelava la B[ernart] Alauzeta.. . e louse, the poet’s own lord, whose wrath he had provoked.27
ela apelet lui Rai) . 22 These expressions are often repeated in the Certain passages in the medieval corpus suggest, moreover, that
medieval literature on the troubadours, which evokes practices of a senhal could also be employed by one troubadour poet to designate
“naming,” “calling,” and “appellation” (appelar) that seem techni­ another. Such an act of allusion might be direct or indirect. There is
cal in kind. At the court of the count of Provence, the thirteenth- no doubt that a minstrel could, for instance, cite a senhal invented by
century poet Sordello “ loved a beautiful and noble lady o f Provence, another, thereby evoking its referent, for “ once one troubadour had
and in his songs, which he made for her, he called her Sweet Enemy imagined a title for a baron or a lady, others could adopt the same
[et apellava la en sieus chantars, que elfazia per liei, Doussa Enemia].”23 designation for the same character.”28 Yet there is evidence that the
Richart de Berbezill, poet of the late twelfth century, “ fell in love troubadours also invented titles for each other. A razo to Guilhem
with a lady, wife of Master Jaufre de Taona.. . and the lady was noble de Baus thus explains that the mysterious name “ Engles” is the
and beautiful__ And when he began making songs of her, he called sign by which the poet chose to evoke Raimbaut de Vaqueiras.29 An
her Better-than-Woman in his songs [apellava la Meillz-de-Domna en anonymous note inserted in the troubadour songbook H explains
sus cantars].”2* Raimbaut d’Aurenga, twelfth-century lord of Orange that “ Desirat,” in a lyric by Arnaut Daniel, refers to the troubadour
Bertran de Born.30 Modern interpretations of such appellations must
and Aumelas “and great king of many castles,” “ for a long season
be more conjectural, although critics have not hesitated to formulate
loved a lady of Provence, who was called lady Maria de Vertfuoil;
their readings in terms of apparent certainty. According to Isabel
and he called her His Playing in his songs [appellava (la) ‘son Joglar’ en
de Riquier, the cryptic name “ Linhaure” in the poems of Guiraut
sas chiansos].”2S
de Bornelh, which has been interpreted in more than a single sense,
Yet the medieval “ Lives” and “ Reasons,” while crucial, do not
“ is, or was, or could be, Raimbaut d’Aurenga, thus designated by his
allow the critic to decipher all the medieval signs. Many o f the

Si
5°
 SIGNS
DARK TONGUES

colleagues and friends, who gave each other names by senhals, signs Whatever their origins, there can be no doubt that the senhals
of intimate recognition, which are often enigmatic for us.” 31 play a crucial role in a fundamental aspect of the Old Occitan craft:
Scholars have proposed several accounts o f the origin o f the what Bernart de Ventadorn, one of the greatest troubadours, calls
senhals. In his classic study of the troubadours, Alfred Jeanroy m ain­ bel celar, “ fair concealing.” 37 Yet critics have differed vastly in the
tained that the Old Occitan pseudonyms constitute an utter novelty readings they have given of this practice. At one extreme, one may
in the history of literature. “ This usage,” he com m ented, “ has, as interpret the troubadour names and phrases as willfully opaque:
tar as i know, not been observed in any literature. U nder C h ar­ titles intelligible solely to the poet and his addressees or, perhaps, to
lemagne, the members of the Palatine school gave each other the them and to those few who remained in their immediate vicinity. If
names of great characters of antiquity. In a more recent period, John one grants, however, that the Old Occitan names were formed and
of Salisbury and Gautier Map designated their contem poraries by used in accordance with the rules of an art, one may infer that they
the names Cornif icius and Porphyrius; but the senhals invented by the were also comprehensible to a public familiar with the conventions
troubadours a r e ...o f a completely different nature.” 32 This was a of the medieval craft. That a single senhal can be found in texts by
polemical claim, for at least some of Jeanroy’s colleagues had argued several authors seems telling. The “ allegorical name” may have been
the opposite. Several years before Jeanroy, V. de Bartholomaeis sug­ less than obscure to the medievals. Yet one can go further. If one
gested that “the use of the senhal is n ot. . . a novelty, since the poets grasps senhals as appellations of a public kind, titles such as “ Lack-
of the Palatine school had precisely the habit o f designating each land,” “ the Handsome,” or Court-Martel, one may also draw a pos­
other by means of pseudonyms.” 33 sibly discom forting conclusion: they may hardly have been secrets
One may recall that the classical Latin love poets, too, employed at all. At m ost, they could have been “ open secrets,” as Gerard
pseudonyms to designate their ladies. Catullus called his Clodia Gourian maintained, or “ functional elements” in a system requiring

“ Lesbia”; Propertius named his Hostia “ Cynthia” ; Tibullus evoked the pretense o f obscurity.38 The Old Occitan name for the poetic

his Plania as “ Delia.”34 The Arabic poets of m edieval Iberia also name is, in any case, richly ambiguous. “ Cipher” or “ insignia,”

cultivated an art of naming by dissimulation, availing themselves o f “mark” or “ proof,” the word senhal—signans and signatum—remains
decidedly equivocal.
poetic titles that could be elaborate. Long ago, A. R . N ykl argued
This much, however, is certain: the Occitan “ sign” possesses a
that the troubadour procedure might well derive from that o f the
role shared by none of the other constitutive elements of the trouba­
Arabic authors, who regularly gave their loved ones masculine “ ficti­
dour song. Whether openly or enigmatically, exoterically or esoteri-
tious names,” or poetic kinaya: amali, munyati, bugyati , gari, sayyidi.3S
cally, the “ allegorical name” points, by its nature, beyond the world
Attending to the more immediate historical and social surround­
conjured up by the poetic text. There is a simple reason for this fact,
ings of the troubadours, Martin de Riquer has advanced a simpler
from which several consequences follow. The “ signal” marks the
historical explanation. The troubadour practice o f poetic nam ing
bare condition o f the poem’s existence. One may give several names
could also be derived from that of the feudal courts at which the
to this condition. More exactly, one may give exactly as many names
poets presented their works. For there, the nobles themselves often
to it as there are accounts of the sense of the senhal. The elliptical
bore surnames. It suffices to think of Richard “ Lionheart” or John
title or phrase may indicate the lady who allows the poem to be
“ Lackland.” “Authentic pseudonymous” were also in use. Hence “ the
made, the noble patron at whose court it can be recited, or, finally,
viscountess of Chatellerault, friend of Guilhem de Petieu, called
the troubadour who w ill enable to it to be received and integrated
Dangerosa and La Maubergeonne, after the castle in which she lived.” 36

53
52
 DARK TONGUES

intv' a tradition. The “ sign” is in this sense ambiguous. Yet its func­ CHAPTER FI VE

tion remains constant. From the eleventh century to the fourteenth,

Knowings
the senhal continues to point, within the poetic w ork, to a reality
that precedes the poem and that enables it to be. The nam ing, for
this reason, shelters a hidden reference to the “ finding” at the origin
and the center of troubadour song. Whatever else it may designate,
a troubadour "insignia” always also alludes to the language in which
the poets craft their work, being an index to the prim e m atter from
which they draw the elements of their song. Among the troubadours,
that allusion remains, for the most part, only implicit. At least once,
however, a poet chose to identify that which these m edieval signs It can happen that words and phrases that are unknown to the
obliquelv, if obstinately, evoke. In the early thirteenth century, a majority o f those who speak a language sound not only in certain
trobairitz, Na Castelloza, noblewoman o f Auvergne, left her read­ portions o f poems but throughout them, becoming, in their dark­
ership a telling indication. The sole woman troubadour ever to ness, the signs o f literary speech. Old Norse literature, in this
employ a senhal, she called her beloved, in bright opacity, Bels Noms, regard, contains perhaps the most brilliant of examples. They are to
"Fair Name.” 39 be found only in part in the great prose works, the sagas, composed
in the tw elfth, thirteenth and fourteenth centuries, which contain
occasional citations in verse. Perhaps the most striking witness to
the secret idiom o f w riters in the Middle Ages derives from a period
in Old Norse literature several centuries before the emergence of
poetry in the Romance languages.
Like that o f all ancient literatures, the dating of the earliest
Old Norse verse is uncertain. This poetry was doubtless transmit­
ted orally before being committed to writing, and today, scholars
maintain that it most likely came to maturity before the tenth
century. Its corpus is diverse. Some o f the oldest poems in the
language record cosm ology in narrative form, being traditionally
considered to be “ the original storehouse of Germanic mythol­
ogy.” ' These are the works o f varying length contained in the thir­
teenth-century Codex regius discovered in Iceland in 1643, which
have been referred to since as the Elder Edda or Poetic Edda. 2 Other
Old Norse poems are o f a fundamentally different form and mat­
ter. They arose between the ninth and eleventh centuries at courts
in Sweden, D enm ark, and Norway, where authors wrote brief
poems, chiefly in praise o f their noble lords and kings. These are

55
S4
 KNOWINGS
DARK TONGUES

Toward the beginning o f this book, the god of poetry, Bragi,

the works of the “skalds,” the Scandinavian poets who crafted, as
instructs “ a certain man named AEgir, or Hler,” on the fundamentals
F .V. Gordon wrote long ago, “ a poetry more m elodious, more
of literary craft. Bragi explains that “ there are two categories into
ornate, and more artificial than any other type that g re w from
which all poetry is divided.”6These are “ language and verse forms”
Germanic tradition.”1
(Mal ok haettir ) . 7 O f the second, metrics, he says no more at this
In its rhymes and in its rhythms, skaldic poetry is exceptionally
point. Here the medieval author concentrates, instead, on the spe­
complex. The dominant verse form, the drottkvaett stanza, obeys
cial variety o f “ language” that is circumlocution, “ the paraphrasing
redoubtable formal demands. Each strophe is com posed o f eight
of one word by several.” 8 “ There are,” we learn, “ three categories
six-syllable lines that bear three stresses and that fall into tw o lesser
[grein] in the language of poetry.” Each consists of a type of nam­
units of four lines each. In the stanza, one observes a rigorous dis­
ing. The first is “ to call everything by its name; the second is the
tinction between the lines that fall in odd-numbered positions (1,
one called substitution [fornqfn]; and the third category of language
3, 5, 7) and those that fall in even-numbered ones (2, 4, 6, 8), for
is what is called kenning.” 9 The first “ category” appears to pose no
there is alliteration between two words in the odd-numbered lines,
interpretative difficulties, for it consists in designation by customary
tied to one alliterating word in the even-numbered ones. There are
names and nouns. The second “category” seems to involve the opera­
two internal rhymes or assonances within every line, varying also
tion by which the proper name of an individual comes to be replaced
in nature on account of their position within the stanza: in the odd-
by a nominal circumlocution, as in the classical figures of pronomina-
numbered lines, one may detect a partial internal rhyme, while in
tio and antonomasia. ' 0 It is the third category of poetry that interests
the even-numbered lines, by contrast, there sounds a full internal
the divine authority most, and in what follows, Bragi relates a host of
rhyme or assonance. Within these constraints of lengths and tim ­
periphrastic formulae to exemplify the technique to which the Old
bres, quantity and quality, the skaldic poet composed sentences that
Norse poets and theorists of poetry gave the name “ kenning” (or, as
fully exploited the syntactic capacities of Old Norse. Relying on the the figure is known in its original plural form, kenningar).
inflected grammar of the language, the court poet could subject the The Old Norse term kenning “ is probably derived from the
order of words to considerable variation, intertw ining and em bed­ expression kenna vid (to express or describe something).” 11 It is con­
ding entire sentences within each other.4 structed on the basis of the verb kenna, “ to ken,” “ know,” “perceive,”
Yet the work of the skalds is no less carefully wrought in its dic­ “recognize.” 12 Stretching a translation, one might conclude that the
tion and, for this reason, their poems tend to evoke the world in kenning implies a “ knowing” of a kind. Yet one would need to add
ways difficult to decipher. That the drottkvaett stanza poses a special that the knowledge sealed in the kenning relates first and foremost
challenge to its listeners and readers has been noted often since the to the structure o f a language and, more precisely, to the distinct
waning of the Middle Ages. But this fact was already observed and levels that, within it, tie one expression to another. If the kenning
discussed in the monument of medieval poetics that is the Younger opens onto the world beyond the literary work, it does so by sub­
Edda or Prose Edda of the Old Icelandic writer Snorri Sturluson. This stituting for one ordinary element of speech another that is more
thirteenth-century work contains at its center an extended account complex than it is.
of literary discourse: a treatise titled “ The Language o f P oetry” Snorri’s god announces that he will hold forth on kennings with
(,Skaldskaparmal).5 This work reveals, by way o f introduction, that didactic precision, thus assuring the future of an art that might
the evident obscurity of poetic speech is nothing less than the con­ otherwise be lost. “ These things,” Bragi explains, “ have nowr to
sequence of a divine decision.
57
56
 PARK T O N G U E S
K N O W IN G S

be told to young poets who desire to learn the lan gu age o f poetry
R an ,” and “ dau gh ter o f N io rd ,” or to a single or ch aracteristic deed,
and to furnish themselves with a wide vocabu lary u sin g traditional
such as “ Stran gely w ise R ave n -te ste r,” or “ Hanged G o d .” By contrast,
terms; or else they desire to be able to understand w hat is expressed
in its n a rro w e r d e fin itio n , the k e n n in g p ro per involves a “ p erip h ­
obscurely " ,J He continues: “ We speak o f O din or T h o r or T yr or
rasis, co n sistin g o f tw o or m ore substantive m em bers, w hich takes
one of the AEsir or elves, in such a way that w ith each o f those that
the place o f a n o u n .” 19 E xam p le s o f ken n in gs in this n arrow er sense
I mention, l add a term for the attribute o f another or m ake m en­
include such phrases as “ land o f h e lm e t,” or “ burden o f the n eck,”
tion of one or other of his deeds. Then the latter b ecom es the one
for “ head.”
referred to, and not the one that was nam ed.” 14 O n e m ay thus call
The skalds c arrie d the art o f nam in g by periphrasis to an extrem e
the god Odin the “All-Father,” “ the hanged G o d ,” the “ Strangely
poin t. In h e r c la ssic stu d y o f O ld N o rse stro p h ic co u rt poetry,
wise Raven-tester,” “ Vilir’s brother,” or “ M im ’s frie n d ,” ju st as one
R o b erta F ra n k cites as an e x e m p la ry te x t a passage from the late
may refer to his wife, Freyja, as “ daughter o f N iord , sister o f Freyr,
e le ve n th -ce n tu ry Ice la n d ic author M arku s Sk eggjason , w hich she
wife of Od, mother of Hnoss, possessor o f the fallen slain and of
renders, for the p u rp o ses o f p o etic and gram m atical analysis, w ord
Sessrumnir and tom-cats, of Brisingamen, V an-deity, V an-lady, fair-
for w ord into a “ v irtu a lly u n in fle c te d E n g lish ” :
tear deity.” 15So, too, by the art of kennings, things m ay be named
and renamed in seemingly unlimited sequences o f synon ym ous sub­ Fjardlinna od fannir
stitutions: “ On a man there is what is called a ‘ head.’ T h is shall be fast vetrlidi rastar;
referred to by calling it toil or burden o f the neck, land o f helmet hljop of huna gnipur
and hat and brains, of hair and eyebrows, scalp, ears, eyes, m outh.” 16 hvalranns iugtanni;
bjQrn gekk framm a fornar
How shall the sea be referred to? By calling it Ym ir’s blood, visitor to
flods hafskida slodir;
the gods, husband of Ran, father of AEgir’s daughters, whose names are
skurQrdigr braut skordu
Himinglaeva, Dufa, Blodughadda, Herfring, Unn, Hronn, Bylgia, Bara,
skers glym fjqtur bersi.
Kolga; land of Ran and of AEgir’s daughters and of ships and of terms
for sea-ship, of keel, stem, planks, strake, of fish, ice; sea-kings’ way O f the fjord snake waded through the snowdrifts
and roads, no less ring of the islands, house of the sands and seaweed Firmly the bear of the current;
and skerries, land of fishing tackle and of sea-birds, of sailing wind.17 Jumped over the peaks of the mastheads
O f the whale house the bear;
Modern scholars, following in Bragi’s footsteps, have sought to
The bear went forward on the old
refine the formal definition of the kenning. They have suggested that
O f the flood sea ski’s tracks;
the Old Norse theorists of the thirteenth century, such as Snorri, no
The storm-breasting broke through of the prop
longer distinguished between two varieties o f periphrasis whose dif­
The skerry’s clashing fetter the bear.20
ference one may still detect among the earlier court poets. O ne is the
figure of the “qualified name” or kend heiti, w hich consists o f “ any Frank com m ents that “ at first hearing, the stanza appears to be
circumlocution in which the base word literally refers to the concept about an acrobatic, if dam p, bear.” 21 The lines evoke, no few er than
designated by the whole.” 18 Examples include constructions that refer four tim es, a single m otion that sets on stage a subject, a verb, and
directly to relations of kinship, such as “ V ilir’s brother,” “ husband of its objective com plem ent. Each tim e, the poet names the elements

58
5 9
 DARK TONGUES NOWINGS
K

of that motion by the art of kennings. Thus, the subject is alternately argued that more is required for the Old Norse periphrasis than
"the bear of the current," "the bear of the m astheads," "the bear Meissner’s theory suggests. Norse kenningar, Heusler maintained,
of the flood." and "the storm-breasting bear of the prop.” All are are not merely condensed descriptions in several terms. These may
periphrases for "ship." The four verbs in the stanza (od, hljop, gekk, be found in all traditions: it suffices to think of such circumlocu­
brain), while lexically distinct, designate, by a rule o f synonymy, a tions as the “ Protector o f the Land” (Schutzer des Landes) for “ruler,”
single activity: despite their differences, they all sign ify a spatial “which,” Heusler remarked, “ one can find everywhere.”27 One might
displacement. Finally, the objects—"snowdrifts o f the fjord snake,” add such expressions for divinity as the Hebrew “ lord of hosts” (ado-
“ peaks of the whale house,” "old tracks of sea skis,” “ clashing fetter nai tzevaot), the Latin “ maker of the world” (creator mundi), and the
of the skerry"—are circumlocutions for “ sea.” In short, “ the poem Arabic “ master of worlds” (rabb ’ ul-‘alamin).
savs—four times—‘the ship sailed over the sea.’” 22 Yet it does so not Heusler’s criterion for the Old Norse kenning was more exacting.
by sequential repetition, but by the superposition and coordination He characterized the medieval circumlocution as “a playful type of
of sets of formulae, from which an attentive reader can also glean the metaphor,” being “ in spirit, the relative of the riddle and the joke,
season and the direction of the poet’s voyage. since it demands a solution and glimpses the similarity of the dis­
The structure of such Norse periphrases has been studied several similar.” In its simplest form, it rests on one structure: “metaphor
times. In what remains the single most comprehensive treatm ent of with derivation” (Metapher mit Ablenkung).28 The phrase can be sim­
the literary device, Rudolf Meissner, in 1921, defined the kenning ply explained. Metaphor operates by a rule of substitution, by which
in its fundamental form as a “ two-part substitute for a substantive a stated term takes the place of another that remains unmentioned.
of common speech.”23 He added that in this “ tw o-part substitute,” a In the kenning, Heusler suggests, one can identify such a substitu­
formal distinction must be made between the “ base w o rd ” (Grund- tion, yet one also finds something more: the index, namely, of its
wort) and the definer, which is the agent o f its “ determ ination” formulation, which functions as its “ derivation.” In the kenning
(.Bestimmung). When the tenth-century N orw egian skald Eyvindr “arrows o f the sea,” for instance, a metaphoric equivalence is estab­
Finnsson Skaldaspillir speaks, for example, o f “ slim arrow s o f the lished between the base word, “ arrows,” and the substituted term,
sea” (maevQrum saevar), meaning “ herring,” “ slim arrow s” is the base “ herrings,” yet this periphrasis also contains a sign of the ground
word, “of the sea” the determinant.24 Likewise, in the thirteenth- of its making in the definer “ of the sea.” One term is equivalent
century Icelandic writer Sturla pordarson’s kenning for the sea, to another, in other words, because of the determinant: A (“slim
“eel ground” (alfoldur), “ground” is the base w ord, w hile “ e e l” is arrows” ) may be taken in place of C (“ herrings” ), since A is said to
the determinant.25 Such an analysis of the kenning has the merits of be o f B (“ the sea” ). As John Lindow, building on Heusler’s argu­
clarity as well as simplicity. Since the time of its formulation, it has ment, suggests, the kenning may therefore also be understood as an
allowed literary historians to identify a number of formulae in Old abbreviated riddle. One may always expand Old Norse periphrasis
English poetic diction that seem to mirror the form o f the Old Norse into a question, followed by an answer: When are slim arrows like
kenningar. Examples include Anglo-Saxon “ sea stallion” (saehengest) herrings? When they are slim arrows in the sea. When is a road like
for ship, “whale road” (hron-rad) for sea, and “giver o f rings” (beaga the sea? When it is the road of whales. When is a bear a ship? When
g if a) for king or chieftain.26 it is of the current.29

Meissner’s account of the kenning, however, has not gone unques­ If one accepts Heusler’s thesis, one will exclude from the field
of kennings all circumlocutions that do not function according to
tioned. In an early review of his monumental book, Andreas Heusler

61
60
 DARK TO N G UES
NOWINGS

the rule of “metaphors with derivation.” It has been observed that is offended by their wholesale use of this figure, which is usually
this choice will significantly restrict the catalog o f such periphra­ regarded as mere frippery obscuring the more essential meaning of
ses in the literary corpus. Kennings w ill be distinct not only from the verse.” 35 Borges, who dedicated an essay to the kenningar in 1933,
“qualified names,” such as “ daughter of Niord, sister o f F reyr” ; these was considerably more equivocal. Admitting that he had amassed
periphrases will also be irreducible to such form ulae as “ giver of memorable examples o f the Old Norse circumlocutions “ with an
rings" and “ protector of the land,” which imply shortened descrip­ almost philatelic pleasure,” he nevertheless called the kennings, in
tions. In the West Germanic languages, such as Old English, ken­ his opening judgm ent, “ one of the most frigid aberrations recorded
nings in the strict sense, then, must be considered to be relatively in the history o f literature.” Yet he ended his considerations by stat­
rare.30 Yet they continue to proliferate in skaldic verse, despite the ing that no aesthetic evaluation of the periphrasis could truly be
fact that the Norse court poets used kennings to name only some conclusive. “ We no longer know its laws,” he wrote.
one hundred things.31 There is a simple explanation for this fact, and
The words left to us are so few.... It is impossible to know with what
it involves a basic feature of the medieval figure: the tw o-term ed
intonation of the voice they were pronounced, with what expres­
kenning is but the simplest of its kind. W hen an Old N orse poet
sions----What is certain is that they once fulfilled their function to
chose to replace a substantive by a circumlocution, he need not stop
astonish. Their gigantic impropriety enchanted the red-haired poets
when he had found two terms for one, or a single compound expres­
of the volcanic deserts and fjords no less than did their strong ale and
sion to substitute for a simpler noun or name. “ Any m em ber of a
their duels of raging stallions.36
circumlocution might be expanded by further kennings.” 32 In a pas­
sage of his Prose Edda, Snorri distinguished, for this reason, between More recent readers of the kennings have not resolved the ques­
varieties of periphrases, graded in terms of their rising complexity: tion. It cannot be denied that, while unmistakably obscure, the
“ First there are simple kennings [kenningar]; second, double [tvikent]; kennings remained a basic element of medieval Scandinavian poetry
third, extended [rekit].” 33 At a first degree of periphrasis, one may, for more than four centuries. That fact raises a question. How to
for instance, designate the sea by the kenning “ earth o f the fish”; understand the persistence o f this figure? Various accounts of its
at a second degree, one may expand “ fish” by means o f the cir­ literary and historical functions have been proposed. Social condi­
cumlocution “ snake of the fjord”; then, at a third degree, one may tions may have demanded that poetic diction remain unclear. “ Some
substitute for “ fjord” the phrase “ bench of the ship.” In three steps, drottkvaett stanzas might be deliberately cryptic,” Frank comments.
one thus reaches a single synonym for “ sea” that is as prolix as it is “ The syntactic am biguity in stanza 25 allows the commemorated
enigmatic: “earth of the snake of the bench of the ship.” 34 There is, loyal retainer to share in his lord’s glory; that in stanza 46 seems
moreover, no rule or reason why such circumlocutions must not be designed to protect its author from prosecution in a paternity suit.”37
carried beyond the third degree. The Old Norse kenning is, at least Other critics have evoked tendencies proper to the grammar of the
in theory, indefinitely expansible. Old Germ anic tongue that may, in some way, have suggested the
That such a procedure could lead to poetry lacking in beauteous art of periphrasis to the Scandinavian poets. Borges, for one, con­
proportions has been stated more than once. Am ong twentieth- cludes that “ the style codified by Snorri is the exasperation and, as
century critics, evocations of the “ Baroque” dimensions o f the Old it were, the reductio ad absurdum of a preference common to all Ger­
Norse kennings are legion. Discussing the skaldic penchant for manic literature: the preference, namely, for compound words.”38
periphrasis, E. V. Gordon, for one, declared in 1927: “ modern taste Linguists have pointed out that from a syntactic point of view, the

62 63
 DARK TON GUES

CHAPTER SIX
circumlocutions, despite Appearances, are a means to achieve conci­
sion, kennings possessing, as Gordon noted, "the m eaning o f a sub­
Riddles
ordinate clause in briefer space and with less em phasis.” 39 Rhetoric
and stylistics have also furnished explanations. It has been argued
that the periphrases are best grasped as Norse varieties o f metaphor
or, alternatively, as the elements of a poetic system articulated in
several levels and governed by rules comparable to those o f a single
language.40
The debates about the functions of kennings w ill doubtless con­
tinue. Yet it is certain that for them, a stratum o f crafted impenetra­
bility was crucial. It can be no accident that in his account of the Every secret institutes a division. Its effects can be detected in the
Old Norse poetic art, Snorri attributes to a god the insistence on individual who, know ingly or unknowingly, conceals something
the hermetic quality of poetic naming. Divine authority establishes from another. But they also extend to larger numbers of speaking
that literary periphrases are to be opaque. Bragi recounts to AEgir beings. Friends, fam ilies, associations, and communities can set
that when the giant Olvaldi died, his sons divided their inheritance, themselves against themselves by the silences they keep concerning
measuring out their father’s gold and “ in turn taking a mouthful, certain matters. Two steps alone suffice for the production of such
each of the same number.” Hence the rise o f a curious, yet divine dissension in dissimulation. Something, first, must be distinguished,
appellation. “ Now,” Bragi explains, “we have this expression among for whatever reason, as worthy o f concealment; it must, second, be
us, to call gold the mouth-tale of these giants, and we conceal it in withdrawn from the usual field o f perception. The etymology of
secret language or in poetry, by calling it speech or words or talk the word “ secret” suggests both acts. In English, the noun derives,
of these giants.”41 That gloss is well worth pondering. B ragi’s ken­ by means o f the French cognate secret, from the Latin secretum, past
ning, “speech of giants,” names a precious good hidden in "secret participle o f se-cernene, meaning “ to discern” and also “ to separate.” 1

language, or poetry” (i runum eda i skaldskap) . 42 In those few words, One might deduce from this fact that for secrets to be held, two

which interpret one circumlocution, another kn ow ing lies safely elements alone are therefore required: the thing set aside and the
means of its concealment. Yet if the secret is not only to be made and
sealed. Implicitly, yet surely, the god establishes a second substitu­
kept but also transmitted as such, a third element will be demanded.
tion: what is to be consigned to "secret language” (felum i runum)
There must be some mechanism capable of establishing and main­
reveals itself to be, whether by expansion or by reduction, synony­
taining the division that results from the existence of a secret: a test
mous with “poetry” (skaldskap). This is a language w ithin language
of sorts, which w ill determine where certain individuals stand with
in which one can rename anything in skilled obscurity. One must
respect to what has been “ discerned” and “separated.” The border
only leave some trace or trail for the reader—god, giant, dw arf or
between those who know and those who do not may, then, be both
human being—who may come, at a future time, to recover its riddle.
preserved and reproduced.
Defined in such term s, the mechanism of secrecy may seem
abstract. Yet it is illustrated in a single linguistic and literary shape
known to all. Riddles not only constitute examples of a minor type

64
65
 DARK TONGUES
RIDDLES

of folklore and literature that, like the proverb, exhibits a conven­ and “ enigma.” The second portion, which Blacking here (alls
tional and often formulaic structure. They also exhibit the form that “answer,” has been variously named “ solution,” “reply,” and “ key.”
secret knowledge takes in the moment that it passes from the field of Some scholars have wished to avoid the confusion that results from
the statement to that of the puzzle and the question. Any concealed so many terms of differing implications. In his work on African rid­
matter can he safely conserved and transmitted in the perplexing dles, Lyndon Harries, for one, proposes a more formal distinction:
shapes of enigmas, which only a few w ill know how to resolve. that between “ Precedent” and “ Sequent,” “ both of which are neutral
Hence the “virtual universality” that riddles would seem to pos­ terms indicating only that one expression precedes the other.”5
sess.2 They may be oral, being composed and transmitted between From Plato to Aristotle, from the Latin orators to Augustine
generations without writing. Folklorists have studied such riddles and the medieval theorists of rhetoric, riddles have been defined
across the world, from the Americas to Europe, Asia, and Africa. in many ways.6 Modern scholars of folklore have sought to offer a
They may also be written, either secondarily, in the sense that they firm foundation for earlier accounts by proposing a systematic and
are transcribed after being composed, or originally, as in the works comprehensive analysis of the form. In an important early contribu­
traditionally designated as “ literary riddles.” Certain traits, in any tion, Archer Taylor suggested that the riddle is essentially a device
case, mark the structure, function, and conditions of recitation of of comparison. “ The true riddle, or the riddle in the strict sense,”
all these cryptic forms. he wrote, “ compares an object to another entirely different object.”7
It is characteristic of riddles to express only established knowl­ The link between the “ Precedent” and the “ Sequent” would then
edge. They may test the ingenuity of those who hear them, but their have its basis in a principle of similarity. Some elements in the com­
primary aim seems less to prompt new responses than to determine parison, Taylor specified, will be positive, while others will be nega­
the competencies of speakers with respect to a traditional corpus of tive. As a rule, each type of characterization will possess a specific
special names and formulae. In a study of the practices of the Anang rhetorical valence. Positive descriptive traits, for instance, tend to
in southeastern Nigeria, John Messenger comments, in this sense, be metaphorical, while negative ones are in general to be understood
that “riddles are not told with the end in view of baffling the audi­ literally. One may take, as an example, the Irish riddled phrase for
ence and stimulating its members to provide correct answers.” 3 John “potato”: “ What has eyes, but cannot see.” The quality of “ having
Blacking, an ethnographer who has investigated the value of riddles eyes” must be understood figuratively, while that of “ being unable
among the Venda in the Northern Transvaal, advances a similar to see” is literally true.8 In “ Towards a Structural Definition of the
claim. “ I never encountered anyone who thought about the meaning Riddle,” Robert A. Georges and Alan Dundes, taking a further step,
of an unknown riddle and tried to reason out the answer. Whenever argued that a riddle may be more fully defined as “a traditional ver­
someone knew a riddle well he answered it pat, as if the answer bal expression which contains one or more descriptive elements, a
was an integral part of the question. Riddle and answer, therefore, pair of which may be in opposition: the referent of the elements is to
are learnt as one unit, and it is knowledge of the riddle that is more be guessed.”9 That thesis has been criticized for several reasons.10 Yet
important than the ability to work it out.”4 the characterization is helpful in its emphasis on the fixed form of
Such terminology is traditional in form but may be questioned. the riddle, the contrast of its descriptive elements, and the challenge
That a single riddle falls naturally into two units—one obscure, one that it provokes.
clarifying—is commonly accepted. The first portion has gone by That challenge, as Georges and Dundes indicated, is one of
many names, such as “riddle,” “question,” “ query,” “ proposition,” reference. The riddling phrase and its solution, “ Precedent” and

66 67
 PARK TONGUES
RIDDLES

"Sequent," point to one reality, whether it be natural, cultural,

the “ Precedent” does so only obscurely. Successful riddlers are the
mythological, or linguistic. Its two parts indicate a single object,
ones who know the “ Sequent” that will satisfy all the conditions of
though by two distinct means: indirectly, then directly. From a logi­
reference stipulated by the contradictory predicates named in the
cal point of view, one may note, moreover, that the riddle tends to
“ Precedent.” When faced with a seemingly impossible set of descrip­
recall and invert an old epistemological form: that o f the predicative
tions, they can find the one term to which they all can be attributed.
definition. For example, when Aristotle asserts that am ong animals,
Riddlers know, in other words, how to find a synonym for the odd­
man is "the only one that has language,” he names a predicate—pos­ est of long phrases. To resolve the enigmas put to them, they must
sessing language-that defines a single species: hum anity. When, call upon a special knowledge of names and predicates, terms and
by contrast, the Sphinx confronts the inhabitants o f Thebes with their descriptions.
her riddle, she proceeds along another path, albeit by denoting the Riddles, in this sense, test a kind of linguistic learning close,
same subject. She offers one complex predicative phrase composed in nature, to that possessed by the scholars who write dictionaries
of three seemingly incompatible elements: “ Being a creature that and thesauruses. The crucial difference between the two sets of
moves on four legs in the morning, two legs in the daytim e, and knowledgeable speakers is that lexicographers concern themselves
three legs in the evening.” Despite its apparent contradictions, this with the vocabulary o f a language that is (or that at least in principle
"Precedent” alludes to a single being, which the “ Sequent” must may be) generally known to those who speak it. Those who solve
name. Oedipus, to his misfortune, knew its nature, for it was his riddles possess a different cunning. It bears on a language shared
own: “man.” While the definition, in its classical form , passes from by many but employed, in technical ways, by few. The corpus of
the given thing to its constituent traits, the riddle, in this sense, riddles, in short, composes less a language in any ordinary sense
moves from a collection of distinct and apparently contradictory of the term than a special language, which arises in the process by
characteristics toward a single subject. D efinition, in short, leads which speakers make of their language a jargon incomprehensible to
from a name to its multiple, consistent predicates; riddling, by con­ the uninitiated.
trast, winds a way from a set of seemingly incompatible attributes to The oldest recorded Indo-European riddles confirm this hypoth­
the common term that can be substituted for them all. esis. The Rig Veda, whose most ancient stratum dates to the second
One may go further still in the analysis of the structure o f the millennium b .c ., contains a host of formulae suggestive of “ Prec­
riddle. Since the “ Precedent” and the “ Sequent,” by definition, edents” disjoined from their “ Sequents,” and among the ancient
possess a single referent, one semantic relation obtains between hymns, one finds a poem composed almost wholly of such expres­
them. It is that of synonymy. This term, to be sure, has several sig­ sions, known in Sanskrit as the “Asya Vamasya,” “ Riddle Hymn,”
nifications, and one may understand it in at least the tw o meanings or “ Riddle of the Sacrifice.” 12 In an influential study of these riddles
famously distinguished by Frege: in terms o f “ sense” (Sinn) and, published in 1925, W alter Porzig argued that they “clearly display
therefore, the semantic conditions under which an expression can a special term inology” with a tendency to denote certain objects
name a certain thing, or in terms of “ reference” (Bedeutung), that according to fixed patterns of description. “ Thus,” he explains,
is, the act of indicating, denoting, or “ picking out” one portion
mobile things (sun, moon, year, foot) are said to be wheel or wagon;
of reality.11 If one defines the word “ synonymy” in this second of
equally ordered things (days, months) are said to be brothers; things
these two meanings, it fits the riddle’s two parts w ell. The “ Prec­
that appear in the air (sun, sparks, lightning) are birds; things from
edent” and the “ Sequent” point to the same part o f reality, though

68 69
 DARK T O N G U E S
RIDDLES

which other things come (clouds, dawn, fire) are said to be a cow; or ‘figurative expression.’” 17 One may find echoes of such claims in
things that unite are called hub, navel, encompassing womb; that which scholarship on traditions as distant from Sanskrit as Old Norse, As
is underneath is a foot; that which is above is a head.13 early as 1923, Heusler called the kenning a “playful type of metaphor
[ ...,] related to the riddle and the joke, since it demands a solution
These facts are rich in implications. “ In the language o f riddles,” and glimpses the similarity of the dissimilar.18 Lindow later observed
Porzig writes, “we stand before a language that, once known, grants that the affinities between the riddle and the Old Norse periphrasis
belonging to a certain circle. The coherence o f this circle can be are threefold. “ 1) Both expressions are composed of two or more
seen in this language and in the worldview it expresses. The lan­ components. 2) The ‘meaning’ or referent of each expression may
guage of riddles is a special language [Ratselsprache ist Sondersrpache]. be (usually is) different from each of the components. 3) The refer­
The group to whom this special language belongs is, for the Veda, ent is found by identifying the nature of the relationship between
easy to grasp: it is the Brahmans, the priests separated off by their the expression’s components.” 19 Riddles, like kennings, make of the
caste."14 In an enigmatic passage, the ancient “ Riddle Hymn” itself words of one language the elements of a new lexicon that doubles,
seems to suggest no less: surely if impenetrably, the one commonly in use.
It would be a simplification, of course, to assume that any lan­
catvari vak pa'rimita padani guage contains solely two such registers: one general and known,
tani vidur brahmana ye manlsmah the other special and opaque. Multiple secret idioms may all be
guha trini m'hita ne'ngayanti lodged within a single tongue. One of the Elder Eddas, “ The Lay of
turlyam vaco manusia vadanti Alvis” or Alvissmal, sets on stage a learned dwarf who, to win Thor’s
daughter for himself, must first meet the challenge of an extended
Speech was measured out into four tracks.
examination in not one but five riddled languages, all somehow
Brahmins with insight know these. People
composed from the standard linguistic material of Old Norse. The
Do not set in motion the three that are hidden:
work opens as Alvis arrives in the dwelling of the gods, eager to
They speak (only) the fourth track of Speech.15
claim for himself the divine bride whom he believes has been prom­
Porzig maintains that the function of the Vedic riddles is in sub­ ised him. Thor, reluctant to remit his daughter to the knowledgeable
stance specific to classical India; in form, however, it is representa­ dwarf, demands that Alvis first prove to all that he is, in truth, as
tive of a more general fact. “A glance at the riddles of other peoples “all wise” as his name suggests. To this end, the god submits him
shows that these, too, rest on special languages and even that the to an examination in five varieties of synonyms. For thirteen Old
Norse words uttered by Thor, Alvis must give equivalent expressions
essence of riddles is everywhere special language.” 16
in the languages of zEsir divinities, the Vanir divinities, the giants,
Such “special languages,” as Porzig comments, appear to rest on
the dwarves, and the elves. The examination fills the entire Edda,
certain fundamental rhetorical operations, such as metaphor and
constituting, in Hugo Gering words, “a versified chapter from the
metonymy. “When we speak of the foot of the mountain, or the foot
skaldic Poetics.”20 The god’s question about “earth” is exemplary:
of a lamp,” he writes, “ then the word ‘ foot’ has a meaning of the
same kind as in the Rig-Vedic riddle of the dawn, which bears the (Porr kvad:)
sun ‘by foot,’ or the clouds, which drink water ‘by foot.’ These are
S^g mer jaat, Alviss,
processes of semantic modification that one tends to call ‘metaphor’

7i
70
 p a r k t o n g u e s

RIDDLE6

v'U ot' rvk fira phrase built from a chain of several disparate descriptions. Other
wrumk dvergr, at vitir, Old Norse works offer clear examples of this pattern. In the “ Lay of
hvo su )Qrd h^itir, Vafthruthnir,” or Vafprudnismal, it is a god, Odin, who must answer
or liggr fyr alda sonum questions. He has disguised him self as one “ Gagnrath” and resolved
hyimi hv^rjum i? to seek out the giant Vafthruthnir, so as to test his renowned knowl­
edge against his own. Accepting the challenge, the giant proposes
(Alviss kvad:)
description after description. In each case, the god must provide a
“J^rd h ^ it ir med m Q n n u m , proper name that w ill satisfy the conditions of reference stipulated
t^n m e d Q lfum fold, by the giant’s representation:
kalla v eg a vanir,
(Vafprudnir kvad:)
ig ro en jQtnar,

alfar g r o a n d i, S^g me'r, Gangradr,

kalla aur u p p r ^ g in .”21 Alls a golfi vill
}nns fr^ista frama,
“Tell me this. All-wise, since, dwarf, I suspect
hve h^stre h^itir
you know every creature’s whole history:
what the earth’s called, spread out before the songs of men, Sas hv^rjan dr^egr
in every world there is.” Dag of drottmQgu.

“‘Earth’ it’s called by men, by Tisir ‘ground,’ (Odinn kvad:)

the Vanir call it ‘ways’; Skinfaxi h^itir,
‘evergreen’ giants, ‘growing’ elves, es hinn skira drQgr
the lofty powers call it ‘mud.’”22 dag of drottmqgu;
h^sta baztr
Alvis succeeds in proving his mastery of the five secret languages.
Jaykkir med Hr^idgotum;
Yet in the end, he fails. As he lays bare his knowledge o f names by
£y lysir mQn af mari.25
night, day begins to dawn. The first rays of light strike his body and
for reasons that remain obscure, the dw arf turns, as the god had (Vafthruthnir said:)
foreseen he would, to stone.23
“ Tell me, Gagnrad, since you wish from the floor
The “ Lay of Alvis” is exceptional in the m inim alism o f its syn­
to make a test of your talents,
onyms. The Old Norse work tends toward establishing a term-for-
what’s the name of the horse who always drags
term identity of reference between words in the common language
the day over troops of men?”
and in the cryptic tongues of the gods, giants, elves, and dw arves.24
Formulaic riddles, by contrast, usually consist not o f tw o expres­ (Odin said:)
sions but of two asymmetrical linguistic parts. To a single word in
the common language, there tends to correspond a synonymous “ He’s called Shining-Mane who always drags
the day over troops of men.

n 73
 DARK TO N G U ES
RIDDLES

The glorious Goths think him the best horse:

That is fog; it passes over the earth, so that one cannot see be< duv of
His mane shines always aflame.”26
it, not even the sun; but it is gone, so soon as the wind gets up.”27

A chapter of the medieval prose Hervarar saga contains a contest

Similar contests can be found in many Indo-European traditions.
of riddles in a stricter form. A man named Gestum blindi must settle
Medieval European literature, for instance, contains parallels in
a dispute with his famously learned king by an act o f verbal battle. the Anglo-Saxon Solomon and Saturn dialogue and the Old Irish
The man, fearing for his fate, makes a sacrifice to O din, who in rec­ Colloquy o f the Two Sages.2* Old and Middle Iranian literary texts
ompense offers to disguise himself in mortal shape and to go “ bring have been said to contain traces of similar confrontations, in which
up riddles” (beta upp gatur) to the king. Repeatedly, O din, dressed learned minds evoke obscure and formulaic phrases, demanding that
as Gestumblindi, advances expressions in dark term s, w hich King their interlocutors interpret and explain them.29 Ancient Sanskrit
Heidrik must reformulate and explain in clear speech. “ Precedents” literature abounds in such examinations, which are known by the
are here in verse, “ Sequents” in prose: name of brahmodya. i 0 Ernest Renou was the first to offer a sustained
analysis of these simultaneously poetic and liturgical tests of riddled
Hverr er sa inn mikli,
speech. One o f the oldest Vedas, the Shukla Yajurveda, represented
er lidr mold yfir,
by the Vajasaneyi Samhita, dictates that the sacrifice of a horse be
svelgr hann vQtn ok vi6;
followed by a contest in which the two officiating agents, a presiding
glygg hann oast,
priest, or brahman, and a reciter of invocations, or hotr, sit on either
en gumna eigi
side of the sacrificial stick. From there, they are to engage in a ritual
ok yrkir a sol til saka?
dialogue o f question and answer. First the brahman asks and the hotr
Heidrekr konungr,
answers; next the hotr poses queries to which the brahman responds.
hyggdu at gatu!
Such exchanges, measured out in verse, obey a regular form:
‘God er gata j)in, Gestumblindi, getit er |)essar; J>at er myrkvi; hann ka'h svid ekaki carati
lfdr yfir jQrdina, sva at ekki ser fyrir honum ok eigi sol, en hann er af, ka u svij jayate punah
|>egar vind gerir a.’ kim svid dhimasya bhesajam

“Who is that great one kim u ava'panam mahat

Over ground passing, surya ekaki carati

Swallowing water and wood; candrama jayate punah
The wind fearing, agnir himasya bhesajam
But fleeing no man, bhumir avapanam mahat
And waging war on the sun?
Who is he who moves alone?
This riddle ponder,
And who that is born again?
O prince Heidrek!”
Who is the remedy for cold?
“Your riddle is good, Gestumblindi,” said the king; “ I have guessed it. And who is the great sowing?

7S
74
 d a r k t o n g u e s

R I D D L ES

Sun is he who moves alone. Such riddles, like all others, function within certain unstated
Moon is who is horn again, constraints. Formulaic expressions may demand corresponding syri
hire is the remedy for cold, onyms only if they exhibit two features: opacity and the hidden
harth is the great sowing.31 principle of their interpretation. That summary may appear willf ully
paradoxical, yet enigmas verify its truth. Were a riddle immediately
Longer brahmodya may display more “ form ulaic, repetitive and
intelligible, it would be none at all. Comprehensible in itself, it would
catechistic” traits, which may function also in the absence of an
neither call out to a synonym in ordinary parlance nor allow the
interrogative syntax:32
mind to linger on the passage from “ Precedent” to “ Sequent.” Yet
prchSmi tva pa'ram a'ntam prthivyah there is another possibility that would be no less threatening to the
prchami va'tra bhuvanasya nabhih form of riddles. Were an enigma solely solvable by means of knowl­
prchami tva vrsno asvasya re'tah edge extrinsic to its construction, it, too, could hardly be called
prchSmi vaca'h parama'm vi'oma an “ enigma.”
Mythology, folklore, and literature abound in examples of “ false
iyam ve'dih pa'ro a'ntah prthivya riddles,” which violate the conventions of contests of wisdom by
ava'm yajno' bhuvanasya nabhih presenting apparent “ Precedents” from which even the cleverest and
ayam so'mo vrsno asvasya re'to most learned o f hearers cannot deduce the correct “ Sequents.” Fear­
brahmaya'm vaca'h parama'm vioma ing that there is no enigma that King Heidrik cannot solve, Odin,
in the Hervarar saga, for example, poses an unanswerable query to
I ask you of the far end of the earth.
foil him. It appears, deceptively, in the same metrical form as his
I ask where the world’s navel is.
preceding questions: “ What said O d in /in the ear of Baldr,/ before
I ask of the seed of the stallion.
he was borne to the fire” (Hvat mcelti Odinn / i eyra Baldri, / adr hann
I ask of the highest heaven of speech.
vari a bal hafidr)?34 The clever king, however, will not be fooled.
This altar is the far end of the earth. Heidrik immediately sees that his interlocutor is Gestumblindi only
This sacrifice is the world’s navel. in appearance, for if he knows the answer to such a question, he can
This Soma is the seed of the stallion. be none but the god Odin himself, who accompanied Baldr to his
This Brahmin is the highest heaven of speech.33 death. Striking his interlocutor with a magic sword, Heidrik thus
exclaims: “ Only you know that, wicked creature” (pat veiztu einn,

However obscure they may at first glance appear, “ Prece­ rqg veettr) ! 35

dent” and “ Sequent” here refer to one event: the taking place of The angrily impious king is not unjustified in his act. In such
contests, linguistic knowledge that is accessible to a single speaker
the riddle as a performance. The formulaic construction “ I ask”
alone is inadmissible. A secret tongue can never be strictly private.
(prchami), like the response beginning with the dem onstrative pro­
To function within the conditions of conversation, even in the most
noun “ this” (iyamor ayani), alludes to the liturgical occurrence itself:
restricted form , it must, in principle, be decipherable, however
the conditions of questioning, as well as the occasion for correct
indirectly, by the one who is to hear it. “ Special languages” must,
response. The enigma can be solved only when the one who hears it
in short, be regular in structure, like the grammatical tongues from
knows that it turns, allusively, yet systematically, back upon itself.

77
76
 DARK TONGUES RIDDLES

which they arc forged. The rules in question, o f course, may be Only if one considers the terms of this “ Precedent” as essentially the
difficult to discern; they may be complex and ideally so. Yet every elements of a simile may one pass to the “ Sequent”: A TreeA1
enigma, no matter how hermetic, must contain within its form the The second variety of riddle may be dubbed “grammatical,” if
kev to its unlocking. Only then may it become productive o f expres­ one takes the term “grammar” as the designation of the linguistic
sions made in its mold. Those who know its hidden sense must be system as a whole. Such enigmas play one element or domain of a
able also to grasp the principles of its construction, such that they single tongue against another. Individual words, for example, may
can appropriate them, in turn, to devise new riddles similar in their be simultaneously yoked in multiple meanings; simple facts of lexi­
patterns to the old. cal ambiguity w ill work to cloud comprehension. Thus a host of
The true secret of the riddle, therefore, lies less in the “ Sequent” familiar forms of English riddle questions and answers: “ What turns
that is synonymous with the “ Precedent” than in the structural but does not move? M ilk ” “ What has a mouth but does not eat? A
bond that unites them. This bond may take a number of shapes. Yet river.” “ What has an eye but cannot see? Needle”™ Yet the ambigui­
research on riddling suggests that there are two fundamental means ties exploited by enigmas may also affect other domains of language.
of tying “ Precedents” to “ Sequents,” of which all others are, in Word structure may play a crucial role. Take the question and
essence, variations.36The first is rhetorical and relies on the arsenal answer: “ Why is coffee like the soil? It is ground.”™ The ambiguity
of tropes and figures familiar to orators. The Sphinx’s riddle is of is in this case morphological, rather than lexical. As W. J. Pepicello
this kind. To pass from its “ Precedent” to its “ Sequent,” Oedipus and Thomas A. Green observe, “ the lexical item / grawnd/ , a noun,
must distinguish between the literal and the figural: “ creature” is is homophonous with the past participle of the verb / graynd /.”40
to be taken in its usual meaning, while the distinctions between Phonology may play a similar role. The link between “ Precedent”
“morning,” “ daytime,” and “evening” are all metaphors for three and “ Sequent” can be a matter of as minimal a distinction as that
stages in life. Similarly, if one believes Porzig, the enigmas of the between one vowel and another. “ What is the difference between a
Rig Veda display regular rhetorical substitutions, such that one can baby and a coat? One you wear, the other you were” 41 Confusion may
establish correspondences between certain types o f objects and cer­ also be sowed at the level of the entire sentence: “ Why is a goose
tain figured shapes that signify them according to regular patterns. like an icicle? Both grow down” Here the two elements of the riddle
Riddles that rest on metonymy, rather than on metaphor, are also of may be understood to form a whole only if one succeeds in detect­
ing, in a single “ Sequent,” two fundamentally distinct syntactic
this kind. The Old Norse periphrasis for “ sea” as “ eel ground,” for
structures. In the first, “ down” functions as a direct object and thus
example, functions by such means: it replaces one term by another,
as a noun phrase; in the second, it is, instead, an adverb.42 Often
linking two expressions that are already united by association. Tra­
these domains, however, can be compressed and superimposed.
ditional enigmas built from implicit comparisons also belong to this
Phonology and syntax can collide. “ When is it hard to get your
set. Consider the English lines:
watch out of your pocket?” finds a resolution in a single phonological
In spring I am gay, chain, which is susceptible, by a fact of homophony, to two differ­
In handsome array; ent chains of words: “ When it keeps sticking there” or “ When it keeps
In summer more clothing I wear; ticking there ” 43
When colder it grows Rhetorical and grammatical conceits, of course, can also meet.
I fling off my clothes; Principles of carefully crafted speech may dictate that in the language
And in winter I quite naked appear.
79
78
 PARK TO N Q U E S
RID DLES

of the T n v e d t'n tth e “ Sequent” be concealed, albeit in ways dif­

can lie hidden in a “ Precedent.” Italian affords an example: “ I will
ficult to discern, literary riddles often follow such a rule. A poem
say to you, 1 repeat to you, and if you do not know what it is, you
b\ Cervantes, for instance, ends with a challenge to the reader: “ Tell will truly be donkeys” ( Velo dico, velo ripeto, se ancor non lo sapete,
me what is one thing and another!” (Decidme que es cosay cosa).44 The dei ben asini sarete). Within the repeated expression “ to you” (velo)
answer lies inscribed in the question: “ one thing and another” (cosay one finds the homophonous noun “ veil” (velo), which points to the
cosa) is a common expression for “ riddle.” The seventeenth-century enigma itself.49 Linguists who have studied folk riddles have found
Hebrew poet Moshe Zacuto devised a like enigm a: “ If a man is bewil­ that such procedures are more common and more deeply rooted
dered bv a secret sealed,/He should learn the concealed from the than one might think. Developing research begun by Osip Brik and
concealed.*'*5 The key to these lines is buried in one o f their words. Roman Jakobson, the philologist and chess composer Alexander
As Dan Pagis has observed, “ the reader assumes the w ord ‘secret’ Gerbstman has shown that one type of Russian folk riddle demands,
refers to the riddle as a genre, but here the w ord also implies that as a formal feature, phonological affinity between “ Sequent” and
the solution of this riddle is the ‘secret.’ . . . If you are bewildered by “ Precedent.” 50 The sounds of the solution, in other words, must in
a sealed secret, by the solution, deduce what is concealed (i.e., the principle be scattered throughout those of the enigma. Consider the
encoded solution, here: secret) from the concealed, from the lines example “ Net ni okon ni dverej. Posredine—arkhierej.—Orekh.” “It has
of the poem itself, which mention the ‘secret’ explicitly.”46 Writers neither windows nor doors. In the middle there is a bishop. A nut”
have carried such operations to an extrem e point. Riddles in one A phonological analysis o f this riddle reveals that the sounds of the
tongue have been written so as to sound aloud, in another, the solu­ solution are distributed across it, not once, but almost three times.
tions for which they call. The early modern Jew ish poets, who wrote Hence the choice o f the “ bishop,” as Tzvetan Todorov, comment­
in the language of the Bible while speaking European vernaculars, ing on Gerbstm an’s work, has remarked: “Among all the possible
practiced arts of this kind. They would embed, w ithin their Hebrew inhabitants o f this strange house, this one clearly follows from the
verse, words homophonous with others in Italian, Spanish, or Portu­ phonic relation o f orekh to arkhierej.” 51 Vladimir Toporov and Taty­
guese.47“ In the riddle Miyom nivra ‘olam by Jacob Vita Israel (Amster­ ana Elizarenkova have argued convincingly for the presence of such
dam, mid-eighteenth century),” Pagis relates, “anagrammatic” procedures in Vedic enigmas, and Vladimir Milicic
has shown that certain Macedonian and Bulgarian riddles follow
the hidden speaker says, “My friend in the hills, I say to y o u ,/if you
similar patterns.52 M ilicic has suggested that such procedures can
ask Shlomo he will reveal it to your ears/ Ask Esther as well, go then
also be found in English-language riddles, although the speakers who
to/Seek, and find me among her perfumes.” Shlomo and Esther are
transmit them may be unaware of this fact. A few examples suffice
the couple in whose honor the riddle was composed. “ In the hills”
to illustrate the principle o f such “ subliminal structures in folklore.”
alludes to Mt. Morah (har hamor); “among her perfumes” (tamruqeha)
“ It goes upstairs and all downstairs and never makes a track. A ir”
is hamor, myrrh; thus, in accordance with their pronunciation, we are
“ Two lookabouts, tw o hookabouts, and four big standers. A cow ”
led to Amor, love. The word omar (I say) suggests the solution as well,
“ Which is the biggest saw you ever saw? Arkansas.”53
inverting its vowels.48
Riddling here reaches a point of maximal abstraction and sim­
Spoken riddles are rarely so sophisticated. Yet w ithin the con­ plicity. The secret is hidden, with the greatest surety, precisely in
fines of one language, they also exploit patterns o f sound, word, being exposed. Two synonymous expressions are at once disjoined,
and sentence in ways no less surprising. In them , too, a “ Sequent” conjoined, and superimposed upon each other: one, in being uttered,

81
80
 DARK TO N G U E S

CHAPTER SEVEN
beckons to another, whose sounds lie quietly in it. To find the way
from “ Precedent" to “ Sequent,” it suffices, then, to listen closely.
Nomina Divina
The syllables, sounds, and letters that com pose an enigm a simul­
taneously, if inaudibly, spell out its certain solution. Those in pos­
session of the rules of the special language w ill know to grasp it.
To the question of the identity of these masters o f speech, several
ancient and medieval traditions suggest an answer. From Vedic India
to archaic Greece and the medieval North comes the same startling
suggestion: those who long ago appropriated a shared means of
communication, making of a common idiom an opaque jargon, were
those who claimed to act with the knowledge o f the gods. Before To know a language is to be in a position of mastery and uncertainty
the emergence of modern crime and its calculatedly cryptic cants, for reasons that follow from the nature of the faculty of speech. It
divinities demanded that a secret be kept and spoken, concealed is a widely known fact that before beginning to form utterances in
from some and transmitted, at the same tim e, am ong the members any given tongue, speaking beings possess an intuitive, yet complete
of a single company. Piety sowed the first seeds o f dissension, for it grasp of its gram m ar. They are somehow aware of the rules that
exacted what must not be said. Tradition follow ed. Priests and their define the set o f signifying sounds from which words can be com­
helpers—poets and scribes—obliged. posed, as well as the principles that determine the ways in which
words can follow each other in correct sentences. Yet it is also com­
mon knowledge that even the most learned of speaking subjects can
always, unexpectedly, come face to face with a linguistic barrier.
It suffices for them to run up against one or more impenetrable
expressions. These may consist of words or phrases that are new in
their basic shapes, yet they may also consist of ordinary locutions
employed in an unfam iliar manner. One might be tempted to attri­
bute their possibility to ignorance of a subjective kind; individuals,
after all, tend to move in circles in which the same special vocabu­
laries are in use, a fact that doubtless limits their knowledge of the
lexicon. Yet there is an objective reason for which speaking sub­
jects can always find themselves at a loss in the languages that they
would seem to master. It involves the special uses that individuals
and groups can make o f a tongue. From their unconscious knowl­
edge of a single grammar, speakers can draw the form and matter
of unattested expressions. To the degree to which they respect the
phonology, morphology, and syntax of a given grammar, such utter­
ances may be considered to be limited in scope. They will not, for

83
82
 d a r k t o n g u e s NOMINA DIVINA

example, involve sounds that cannot be pronounced by speakers of

involves the referent: it is one and the same, and the divine name grants
the language, or that cannot possibly be recognized by them. In this
it a cosmic dimension that the human name does not possess.2
sense, such formulae will be consonant with the languages in which
thev appear. Yet they may be utterly incomprehensible nonetheless. The Iliad contains the most important Greek examples of these
The threat would be minimal were the set o f such possible terms names. Four times, the Homeric narrator interrupts his discourse to
restricted. The truth, however, is that w ithin the terrain circum­ introduce an expression in need o f special commentary. The term,
scribed bv a grammar, obscure expressions can always be invented. in each case, appears to be an unusual Greek word for which the
They can, moreover, proliferate without end, for the parts of speech poet hastens to offer a synonym in the vocabulary that would have
can be recomposed in infinitely new wrays. been immediately intelligible to his audience. Book 1 contains the
The oldest literatures of the Indo-European traditions suggest, first evocation o f such a term in the scene in which Achilles con­
in striking unison, that the true masters o f such obscure expres­ verses with his divine mother, Thetis. Speaking of Zeus, “ the son of
sions are divine. A corpus of ancient sources leads one to believe Chronos,” the Achaean hero recalls: “ You came, goddess, and freed
that the gods of the ancient Greeks, Celts, N orsem en, Indians, and him from his bonds, when you had quickly called to high Olympus
Anatolians, in particular, employed a set o f special term s and phrases him of the hundred hands, whom the gods call Briareus, but all men
that were at once similar in form to those com m only recorded in Aegaeon; for he is mightier than his father [hon Briareon kaleousi theoi,
human tongues and noticeably distinct from them. Such expressions andres de te pantes / Aigaion, ho gar aute hiei hou patros ameinon].”3
appear to have been largely nouns and nom inal phrases, like the The other three examples in the epic set godly terms against human
elements of the argots that developed across Europe centuries later. ones in like fashion. “ Now ,” we learn in book 2, “ there is before the
In the Greek tradition, the examples begin w ith Homer. His two city a steep mound afar out in the plain, with a clear space about
epics contain six references to a special lexicon: a jargon of sorts, it on this side and on that; this do men truly call Batieia, but the
that, to use an irreverent yet entirely appropriate term , one might immortals call it the barrow o f Myrine, light of step [ten etoi andres
name “godly cant.” After Homer, Greek literature contains over a Batieian kikleskousin, / athanatoi de te sema poluskarthmoio Murines] ”4

dozen more allusions to such a form of hidden discourse that offer, Elsewhere, Homer evokes the “ clear-voiced mountain bird, that the
in each case, a term and its translation. From H esiod and Pindar gods call Chalcis, and men Cymindis [khalkida kikleskousi theoi, andres
to more minor authors, such as Pherecrates and M axim us of Tyre, de kumindin]” 5 and “ the great, deep-eddying river, wdiich gods called
no fewer than twenty precious lexical items have be found.' In Xanthus, and men Scamander [hon Xanthon kaleousi theoi, andres
the fullest treatment of the Indo-European language o f the gods, de Skamandron]”b
Fran^oise Bader has shown that these mentions generally display The Odyssey contains only two terms belonging to the divine jar­
two characteristics: gon. Once, the poet explains o f a special herb: “Moly the gods call it,
and it is hard for mortal men to dig; but w ith the gods all things are
From a formulaic point of view, the names of the language of gods and possible” (molu de min kaleousi theoi: khalepon de t’orussein/ andrasi ge
men form two couples (or give rise to more distinctions in more recent thnetoisi, theoi de te panta dunantai) . 1 Later, Circe, revealing wonders
texts), and they are accompanied by the author’s reference to the giving to her parting guest, explains that Odysseus, when at sea again, will
of the name: “among the gods (or men), X is called. . . ” or, in an active reach straits where “ on the one hand are beetling crags, and against
form, “the men (or gods) give the name X. . . . ” Another characteristic them roars the great wave o f dark-eyed Amphitrite; the Planktai

84 85
 PARK TONGUES NOMINA DIVIN

do the Mossed gods call those [Planktas de toi tas ge theoi mak- between them is more exactly intralinguistic, involving two levels
aro> Ltleousij,”8 It is striking that both these terms remain without of a single language. Hermann Giintert thus maintained in 1921 that
apparent svnonvms in the language of the Homeric epics. Homer’s the divine words constitute archaic expressions deriving from the
twelfth'centun commentator Eustathius of Thessaloniki explained field of religious terminology, while the human words, by c o n t r a s t ,
the good reasons for this fact. Moly and Planktai remain nameless in were in Homer’s time all terms of common use. Antoine Meillet, in
the discourse of men, he commented, for they are unknown among response, appealed to the possibility of a linguistic prohibition that
mortals.* While the double terms of the Iliad “ may suggest the rela­ could have caused certain terms to become unspeakable, Homer’s
tive superiority of divine to human knowledge,’’ Jenny Clay notes, divine names could then be explained as “circumlocutions of the
“the Planktai and mofy belong to a kind of knowledge available to the kind that are used to avoid words that have been forbidden by
gods alone.’’10 taboo.” 16 Romano Lazzeroni later developed Meillet’s argument in
Less evident is the nature of the words that belong to the godly greater detail. In an incisive essay, he recalled that the Homeric gods
argot. Plato already raises the question in the Cratylus. Socrates do not, as a rule, use a special language, either among themselves
enjoins Hermogenes to learn from “ Homer and the poets” about the or in conversation with men. Homer’s allusions to a divine idiom,
“correctness” (orthotes) of expressions by attending to the passages therefore, are in need of explanation. This Lazzeroni offered by
in which he indicates the ways in which “gods and men designate maintaining that whenever a specifically divine lexical item appears,
the same things.”11 In jest or in earnest, the philosopher suggests it functions to denote “ a natural object that, in a specific context, has
that the unfamiliar names of divine origin must be taken to be cor­ taken on a sacred character,” withdrawing it from the reach of mor­
rect, the human to be false.12 There remains the problem, however, tals. ‘“ Divine terms,” ’ he concluded, “ have the structure, the sense,
of the features that define divine words. Philologists and linguists and the value of tabooed expressions.” Yet he also conceded, follow­
have sought to resolve it in several ways. In one o f the first studies ing Gunther Ipsen, that the literary practice nonetheless remains
of Homer’s references to a godly speech, J. van Lee wen suggested something of a mystery: “ It is difficult to understand what stylistic
in 1892 that the immortals’ words could be distinguished from the effect Homer was aiming for in giving a ‘divine’ denomination for
mortals’ in being Hellenic in origin, rather than non-Hellenic.13 In an object.” 17
1910, H. Jacobsohn pursued the argument, holding that one can find Early in the development of modern philology, scholars noted
in the differences between the two types of designations the signs that Homer’s allusions to divine speech, however curious, are not
of a partition of an ethnic and therefore interlinguistic kind. While entirely unique. In his Germanic Mythology, first published in 1835,

the divine terms are Greek, he reasoned, the human terms recall Jakob Grimm remarked on this unexpected and suggestive fact:
“concordant information reaches us from our own antiquity, as
various tongues of Asia Minor.14 In a similar vein, Paul Kretschmer
well as the Greek, regarding the language of gods.” 18 The classic
suggested, a decade later, that one can recover in the two sets of
Germanic example, of course, was the Eddie “ Lay of Alvis,” which
words a distinction between “ the national [or ‘popular’ ] speech of
evokes not only the tongues of gods and men, but also those of
the poet” (Volkssprache des Dichters) and expressions deriving from a
dwarves, giants, and elves.19 Grimm’s remarks were to be the start­
“foreign dialect.”15
ing point for the first major study of the subject: Herman Giintert’s
Such claims have proven difficult, if not impossible, to demon­
1921 book, On the Language o f the Gods and the Language of Spirits:
strate. More recent scholars have held that human and divine words
Historical-Semantic Investigations into the Homeric and Eddie Divine
are equally Greek and foreign in origin and that the difference

87
86
 DARK T O N G U E S
NOMINA DIVINA

language* Conducting an extensive exam ination o f the Old Norse

in the Rig Veda in affirmative sentences, where it has the meaning of
and ancient Greek literary works, Giintert reached the conclusion
“ how,” “ just as if.” 2*
that what appears in both texts as an opposition betw een languages
reflects, in truth, two types of designations in one tongue. In the “ Somadeva 1.5 9 .6 4 contains a discussion of four languages: San­
Homeric epics, as in the Edda, godly term s are not foreign words, skrit, Prakrit, popular dialect and the language of demons.” There
or vives mysticae, or even, as some had held, artificial lexical con­ are even “ individual ‘ divine w ords,” ’ Giintert added, “ which
structions. They are, instead, “ poetic circum locutions o f a general confirm these findings; I not only mean such nomina sacra as svaha,
kind” (poetische Umschreibungen allgemeiner Art), b eing distinct in hin, om, vau$at and so forth, but also individual expressions, which
structure from the standard nonpoetic designations o f things in in their form recall the Homeric divine words.”24
Homeric Greek and Old Norse.21 In the closing pages o f his book, More recent scholars, following in Gunter’s steps, have studied
Giintert enjoined his readers to see in the language o f the gods one these terms in detail. They have thus been led to propose a new
variety of artistic creation, a variety whose origins might well lie interpretation o f the lines o f the Vedic “ Riddle Hymn” that state
in acts of inspired utterance, such as those o f speaking in tongues, that speech has been divided into “ four tracks,” of which only one is
that, when transcribed and studied at a later date, had engendered known among m ortals.25 The three hidden tracks, it has been said,
new expressions. Yet he also argued that such designations are are idioms known solely to the Rishi, the poets who composed the
close in nature to those employed by the m odern poets who have ancient hymns, and to the priests who, officiating at Vedic sacrifices,
invented new “spiritual languages.” “ Such poetic testim onies,” he employ a secret terminology. “ This esoteric knowledge,” M. L. West
wrote, “ bring us news of the most beautiful and the deepest spiritual has commented, in reference to the three secret tracks of speech,
language, which is close in its effects and its m ode o f expression “must have included the ‘ hidden names of the gods’ (RV 5.5.10,
to music, and to which the poet’s word alone, am ong all the arts, devanam guhya namani), needed for the successful ritual; the expres­
gives speech.”22 sion implies that the gods had names for themselves that differed from
As research in the fields of ancient literature and poetry contin­ those in common use,” yet that were known to the learned Vedic
ued, philologists became increasingly aware that languages of the poets.26 “ In 4.58.1 ‘the secret name of the ghee,’ understood to be the
gods are more extensively documented than G iin tert had recog­ holy Soma, is called ‘the tongue of the gods,’ apparently because it
nized. In an excursus, Giintert had alluded in his book to the exis­ causes the sacral vocabulary to be uttered at the sacrifice.”27 There are
tence of passages in Sanskrit literature that seemed to echo those in several other examples. One Sanskrit prose text “ describes a ritual in
Homer and the Elder Edda. The classical com m entators of the Rig which the priest takes a black antelope skin (carman-) with the words
Veda, he recalled, maintain that there are discernable differences ‘You are a defence (s'arman-),’ and it is explained that while carman- is
between words in the tongues of gods and men: its human name, ‘it is s'arman among the gods.’”28 Strikingly, the ritual
of the horse sacrifice, or As'vamedha, “ the principal Indo-European
Thus, for example, Mahidasa, the compiler of the first six books of
kingship ritual,”29 names the horse four times, according to a pattern
the Aitareya-Brahmana, often says,yad vai devanamneti tad esam o3 m iti,
reminiscent o f the Old Norse “ Lay of A Ivis”:
“what is ‘no’ to the gods, is ‘yes’ to human beings,” which, according to
B. Liebich, Introduction to Native Indian Linguistics (1919, II, 6), means hajo bhutva devdn avahad vajf gandharvan
that na, the particle that later was to mean “not,” clearly also appears arva asuran as'vo manusyan

88 89
 d a r k t o n g u e s
NOMINA DIVINA

as fkivd ho boro the gods, as vajm the gandharvas, a manuscript dating from the twelfth century. Yet there are reason*
as lirvun the Asuras, as as'va human beings.30 to believe it contains material that may bear witness to “a doctrine
of poetic learning” developed in an earlier epoch. In one precious,
The Vedic text offers three synonyms for the “ m ortal” word albeit obscure passage, the treatise justifies certain grammatical
for “ horse," diva: in M.L. West’s transposition, “ as it were, steed, features of the Irish language by appealing to the three successive
courser, racer."31 The suggestion is clear. To each type of being—gods, ‘“ legendary races’ of men and gods which occupied Ireland: the
Gandharvas, Asuras-there corresponds a different cryptic tongue Milesians (Sons of Mil), ancestors of the Gaels; the Fir Bolg invaders;
with its own vocabulary.32 and the ancient Celtic pagan gods, the Tuatha De Danann.”37
Signs of godly cants have also been found in several other archaic Such accounts can be interpreted in several ways. For some
Indo-European traditions, which do not appear to have influenced comparative philologists, they offer compelling evidence of a single
each other. The earliest records are those contained in the oldest Indo-European poetic technique transmitted in different languages
know n Indo-European tongue of all: the ancient Anatolian language, for millennia: in Watkins’ words, “ a conscious tradition of obscu­
Hittite, generally believed to have been in use between the sixteenth rantism, of secrecy, which serves like a cipher to protect the poetic
and the thirteenth centuries b .c . Bilingual tablets inscribed both message.” 38 Yet one may also question the unity of this “conscious
in Hittite and in Hattie establish a series o f parallelism s between tradition.” Whether by chance or by necessity, the documents sug­
two sets of names for gods: the human and the divine. “ For man­ gest less a single inheritance than a corpus in bits and pieces, not all
kind, thou (art) Tahattanuit, but among the gods, the Mother of the of which are incontestably Indo-European. That the twenty Greek
Spring, a queen [(art) thou]__ For mankind, Wasizzil, but [among] works in the language of the gods and the terms of “ Lay of Alvis” are
gods a Lion-King (art) thou, and [thou holdest] Heaven and Earth.... straightforward Indo-European witnesses can be granted. Yet the
For mankind, Tasimmit, but among gods Ishtar the Queen (art) Hittite texts are also transcribed in Hattie, a non-Indo-European
thou.”33 A formal “schema,” as Johannes Friedrich has observed, language; they have often been considered to bespeak Hattie reli­
is evident even in such a translation: “ first we are told what the gious rites.39 The classical Sanskrit testimonies of the jargon of the
divinity is called among human beings (ha-pipunan), which is its gods, while numerous, fail to conform, as a whole, to the models
everyday name; then follows the divinity’s name among the gods presented in the other examples; even as convinced a proponent
(ha-yashahipi), accompanied by an attribute.” 34 The first designation of the idea of an Indo-European hermetic tradition in poetry as
situates the divinities in the mortal realm, the second in the world Fran^oise Bader admits, in this regard, that “ it is perhaps for India
of the immortals. Literature in Avestan, the Iranian language of that the language o f the gods leaves the most to be desired.”40 The
the second millennium b . c ., may furnish related cases in texts that Avestan sources are few in number, and while evoking a partition
offer, for a single object, two sets of strictly synonymous names, one of tongues similar to the others, lack any explicit reference to the
of which is lowly, the other of which is elevated.35 Finally, Calvert language of the gods. There is the Middle Irish allusion to the form
Watkins has maintained that Old Celtic literature transmits a dis­ of speech of the “ ancient pagan gods,” to which Watkins accords
tinction that mirrors that of the archaic Greek, Indian, and Hittite considerable importance, yet it consists of one sentence, copied in a
traditions. His source is “one of the most curious o f all Irish texts,” medieval Christian intellectual milieu that can hardly be considered
the Auraicept na n-Eces, known in English translation as The Scholar’s to have been exclusively Celtic. Moreover, even if one accepts that
Primer or The Handbook of the Learned.™ This work survives solely in these various sources all bear witness to a “conscious tradition of

90 91
 DARK TONGUES NOMINA DIVINA

obscurantism" in poetry, one may still wonder about the degree to gods’ is derived from it.”44 One might add that the fact of synonymy
which the “tradition” in question is Indo-European in character. is also basic, for only where several terms share a single refer­
It could he so on account of its remotest historical and linguistic ent can they be distinguished solely by virtue of “ markedness” or
origins, which still survive in fragments in later form. It could also “unmarkedness,” lexical rarity or commonality. Watkins’s principle
he so on account of accidental borrowings made after the period in nonetheless remains valid, as well as impressively supple. “ From
which the Indo-European languages would have diverged. Finally, it these examples,” he w rites, having surveyed the Indo-European
could be so on account of the nature of sacred speech, which may, in instances of divine naming, “ it should be clear that the social groups
the most disparate of linguistic cases, demand linguistic obscurity.41 can be multiplied to fit the number of synonyms or quasi-synonyms
Certain common traits among these various forms of poetic in the lexicon to be contrasted. ”4S Poetic theology, in any case, will
“obscurantism” in any case merits close study, for they cast a bright recapitulate facts of grammar: the opposition of the speech of gods
light on the structure of jargons and cryptic tongues in general. In and men w ill mirror the distinctions implicit in the lexicon, and the
a penetrating paper, Calvert Watkins has identified a feature that poet’s “obscurantism” w ill draw its crucial matter from the relative
marks all the examples of Indo-European lexical doubling. He notes darkness of certain appellations in the language.
that in each case of opposition between an expression in the lan­ This hypothesis is seductive, but one may object that it is one­
guage of the gods and an expression in the language of men, one may sided. For it rests on the crucial presupposition that the grammar of
observe the same “metalinguistic poetic figure.” This figure func­ a language precedes and determines the special idioms evoked in its
tions to exhibit a semantic hierarchy that is already present in the poetry. Only if one grants that the lexicon of a language essentially
lexicon: that, namely, between terms of different levels that refer to preexists the literary act can one reason that given certain synonyms,
a single object. “ The lower level,” Watkins w rites, “ that of ordinary a poet may conceive certain expressions, multiplying “social groups”
language, is figured as ‘the language of men,’ the higher and more in numbers m irroring those of words. One might also entertain
restricted level of formal, poetic or otherwise exotic language is the possibility that makers of poetry, in crucial ways, always also
figured in this ancient metaphor as the ‘ language o f the gods.’ This invent aspects of their languages, not least their vocabularies. Their
metaphor represents a conscious signalization of an opposition exist­ discourse is not only given in advance; it is also forged, each time
ing in the lexicon, between the common, semantically unmarked term, anew. Into a set of existing synonyms, a poet can always introduce a
and a rarer, more ‘charged,’ semantically marked term.”42 hitherto unknown distinction of markedness and unmarkedess, not
Watkins’s thesis, in short, is that seemingly theological distinc­ least by assigning one expression to the idiom of the immortals and
another to that of mortals. Poetic theology, then, would determine a
tions between two expressions are linguistic and, more precisely,
feature of the common language, for composers of epics, hymns, and
lexical in nature. Gods and men are the elements o f a single figure
narratives would alter the sense of some names in distributing them
for the division of one vocabulary into several sets o f hierarchically
to the different beings who inhabit their worlds. Jargons can also
distinct, yet synonymous terms, among which a fundamental dis-
bring about changes in the tongues of which they are a part, and the
tirictive opposition can always be discerned: that between the “com­
cant of gods, no less than that of criminals, can exert an influence on
mon” and the “rare,” or, in the Jakobsonian terminology evoked by
the common idiom in which it is evoked. The possibilities are doubt­
Watkins, between the structurally “ unmarked” and the “ marked.”45
less many. It would perhaps be misguided to linger on questions of
As Watkins writes elsewhere, “ it is the semantic hierarchy that
precedence and consequence, for on this level, they can hardly be
is basic, and the metaphor of ‘ language o f men’ and ‘ language of

93
92
 DARK T O N G U E S

CHAPTER EIGHT
resolved. As Walter Porzig observed in his study o f Vedic enigmas,
‘ one should not ask which of the tw o is older: com m on language
Nomina Sacra
or special language.”4*' The truth is that they are strictly simultane­
ous. The philology of Indo-European languages has demonstrated
that even the most ancient of idioms had its jargons; already in the
second millennium b .c ., tongues were seized and altered in secret
bv “canting crews” no less hermetic for being priestly. Speech, one
might conclude, is always imparted in parts w aiting to be taken. It
is in the nature of the rules of grammar, however bounded they may
be, to leave a space for speakers to craft new expressions from them
and for writers, with or without w ritten signs, to make the idioms To devise an im penetrable expression, one need not compose an
thev inherit their own. entirely new word. It can suffice to alter those already present in
the language. The simplest level on which to intervene is that of the
smallest units of meaningful sound, which coincide, in certain writ­
ing systems, with the letter. Speaking beings possess a conscious­
ness, or credulity, so acute that they can be led astray by the altera­
tion of even one o f the phonological atoms that, when ordered and
combined with others, make up known words. O f course, if the need
to conceal a certain term or phrase is great, it may be wiser to effect
a systematic set o f changes to the chain of signifying sounds. Sueto­
nius reports that Julius Caesar, for one, did so with some frequency.
Whenever he had something o f a private nature to divulge in a letter
to a friend, he would “ write it out in signs [per notas scripsit], that is,
by so changing the order o f the letters, such that not a word could
be made out.” The key to his alphabetic procedure, nonetheless, was
rudimentary. “ If anyone wishes to investigate these letters and peer
into their m eaning,” Suetonius explains, “ he must substitute the
fourth letter of the alphabet, namely D, for A, and so for with the
others.” 1 Augustus, after him, employed a simpler system. “ When he
had occasion to w rite in signs [per notas]” his biographer recounts,
“ he put B for A, and so forth, and instead of Z, A A ” 2 The Biblical
practice known as atbash, recorded in the Book of Jeremiah, rests on
a similar conceit. Tav, the last letter of the Hebrew alphabet, takes
the place of aleph, the first, even as shin, the penultimate, takes the
place of bet, the second. The Hebrew name of Babylon, Bavel (?:□ ),

95
94
 DARK TONGUES NOMINA SACRA

thus becomes Sheshakh ( iw ) , according to a secret system whose fitting names of the things they designate and are, in structure,
exact function is no longer clear today.3 susceptible to a grammatical analysis so thorough as to leave no
Such procedures long outlived classical and Biblical antiquity. remainder, thereby justifying each of the linguistic traits perceptible
The curious seventh-century Latin grammarian Virgilius Maro, once in words. Yet it is also repeatedly reported that the gods, by nature,
believed to have been a Spanish Jew, later thought to be an Irish “ delight in what is concealed” (paro’ksakama hi devah) and “ detest
monk, included in his Epitomae a lengthy treatment, “ On the Art of what shows itself in plain view ” (pratyaksadvisah),6 “ This state­
Scrambling Words” (“ De scinderatio fonorum” ), in which he distin­ ment,” Jan Gonda has commented, “ implies that gods like to use,
guished several ways to disorder the letters of known expressions. among themselves, a vocabulary which is not intelligible to (most)
From classical Roman literary texts he made new and unintelligible men. They like to keep the names of important beings and objects
series of signs, grouped together as words. Virgilius listed, in didac­ secret.” 7 The ancient Sanskrit prose works known as Brahmanas
tic terms, at least three reasons for such practices: “ First, so that we amply confirm his interpretation. They furnish precious accounts
may test the ingenuity of our students in searching out and identify­ concerning how the gods, in a time almost before all human reckon­
ing obscure points; second, for the ornamentation and reinforcement ing, secretly altered the form of words in ways that might easily have
of eloquence; third, lest mystical matters, which should be revealed escaped all mortal perception.
solely to the initiated, be discovered easily by base and stupid peo­ Charles Malamoud has dedicated an exemplary study to these
ple.”4 In hindsight, one may say that such “ scramblings” were to be, changes made to the language shared by mortals and immortals. That
in a sense, only beginnings. Starting with the Renaissance, ever such alterations confound grammatical analysis can hardly be doubted;
more complex forms of artificial encipherment and decipherment the divine changes work both to disguise etymology and to cloud mor­
came to be devised. The modern cryptographers might begin, as did phology. “A great many passages tell us that the original form of this
the ancients and medievals, with the signs and expressions of natural or that word has been slightly modified, becoming a hidden form.”8
languages, but they quickly made of them new “ secrecy systems,” Human beings, for example, believe themselves to be familiar with a
often unspeakable as well as unrecognizable in form. god named Indra. But they reason with a term artificially made by the
Arts of dissimulation of this kind may be effective in concealing gods on the basis of a related word. The god’s true name, we learn,
speech. Yet they possess a weakness, and that is that they are mani­ is “ Indha, that is, ‘what lights,’ a name derived from the verbal root
festly occult. Confronted with sentences reordered by the Caesars’ INDH, ‘to light, burn,’ and justified by a fire function attributed to this
cipher or the prophet’s atbash, speakers of Latin and Hebrew could divinity.”9 Elsewhere, however, another origin and another disguise
hardly fail to observe that into their customary speech, an unheard are suggested: “ Indra,” we are told, derives in truth from idamdra, “ he
expression had suddenly been inserted. The craft is too apparent. who has seen this,” “ since as soon as he was born, Indra looked about
Subtler techniques of hiding efface their traces, becoming almost him and saw a man who was none other than the Brahman (that is,
undetectable to the ear and to the eye. The gods of ancient India, the Absolute), in all its extension, and he said, ‘I have seen this.’” The
according to the classical Vedic sources, cultivated such procedures. explanation could hardly be clearer: “ That is his name. We call the one
Tradition teaches that on earth, the bodies of these deities can cast who is Idamdra Indra, in a hidden fashion, for the gods delight in what
no shadows, being absolutely clear.5 They themselves speak the is hidden.”10 So, too, the name of the god commonly called Agni is, in
truth, for it is their nature. Their language is the “ refined speech” truth, “Agri”; it derives from the adverbial forms agram, “ in front,” or
known as Sanskrit (samskrta): an idiom in which words are the agre, “ in the first place,” though few mortals know it.11

96 97
 DARK TON GUES NOMINA SACRA

Common nouns have also undergone such processes of phono­ minute of changes to given words. Letters may be either added or
logical deformation. Thus, the Sanskrit word for sweat, sveda, can subtracted to the items in the vocabulary. Most often, however, they
he etymologically tied, by means of an explanatory myth, to a word are merely substituted for each other, one by one, in elementary
apparently unrelated to it, suveda, “ happily found.” “ What is suveda phonetic forms. That is enough to confuse our mortal minds.
is called, in secret, sveda, for the gods delight in what is concealed One might imagine that for such operations, writing would be
and detest what lies in plain view.” 12 The sources suggest that even crucial. How could one change Indha to Indra, Agri to Agni without
one of the oldest Indo-European designations for humanity has also decomposing these names into the minimal units now represented
been altered in such a way. When pierced by a three-pronged arrow, as letters? It is remarkable, however, that while according such
the creator god Prajapati was wounded, and his sperm descended to importance to details of a grammatical variety, the Vedic corpus
earth and formed a lake. “ The other divinities exclaimed: ‘May this refuses to admit the role of grammata, “ letters,” scripts, and books,
sperm of Prajapati not be spoiled (ma dusat)V Because the gods said in the transmission of revealed tradition. In classical India, Mala­
ma dusat, the lake of sperm came to be called madusa” But, once moud writes, “ knowledge is above all sonorous speech, vac, and
again, “ the gods like to express themselves in terms unknown to writing grants access to knowledge only if it leads to familiarity
men”: madusa thus “changes to manusa, ‘ human,’ ‘man.’” 13 with sounds.” 16 Not by accident does the Sanskrit term for “revela­
To a certain degree, the alterations recounted in these Vedic tion,” sruti, mean “ hearing.” Although the Vedic poet is considered a
sources can be seen to involve a cryptic tongue invented within “seer” (rsi), he is not thought to read or write. Nor are the guardians
an existing grammar. Yet this divine jargon works in ways unlike of Vedic tradition to frequent books. Commandments, both positive
those of other special languages. The gods, the texts relate, willingly and negative, are explicit on this matter. “ Vedas are to be learned
divulge their artificial vocabulary to m ortals; what they conceal orally,” declares the Gopatha-Brahmana.17“ It is held that no knowl­
is the common lexicon from which they departed. The “ secrecy” edge of dharma can result from a Veda known improperly, either
of this godly cant is therefore in some senses unique. Malamoud from a written text or studied by a servant [sudra].”18 “ Innumerable
observes: “ The word that results from the alteration inspired by the signs indicate that in any milieu dominated by Vedic religion or by
desire for esotericism is not occult in the sense that it belongs to a the Hindu religion that derives from it, writing is held to be vile
and suspect. At best, it is useful as a last-resort technique. But it is
language of a closed group. On the contrary, it is the falsified word
inadequate to the task of preserving the most prestigious bodies of
that belongs to ordinary language; everyone says Indra, and not
works, which, though they can be very voluminous, must be com­
Indha, Agni, and not Agri.” By an unexpected artifice, jargon, the
mitted to memory.” 19
special tongue, thus becomes “ the common language,” while “ inno­
More than once, the foreign visitors to ancient India expressed
cent and intact words” are veiled in secrecy.14 If human language
some surprise about this fact. Megasthenes, Greek ambassador to
appears obscure in its origins and form, one can then infer that it
King Kandragupta in the fourth century b . c ., reported, for example,
may be the work of divine minds, who distorted grammar accord­
that the Indians of his day “ have no knowledge of letters and regulate
ing to rules that they alone—with certain rare exceptions—master.
every single thing from memory” (oude gar grammata eidenai autous
What they guard to themselves is not a lexicon but a systematic
allapo mnemes hekasta diokeisthai) . 20 The soundness of that general
mechanism: “ The secret of the gods is the secret o f a fabrication and
pronouncement has been questioned, for in India writing was used
deformation of a product that is everywhere in use.” 15 It is a “ fab­ from an early date, and books were neither rare nor strange.21 As an
rication and deformation” that proceeds, at each step, by the most

99
98
 DARK TONGUES NOMINA SACRA

observation about the Vedic tradition, however, it is not implausible. memory.”25 Modern scholars have offered their own interpretations
The eleventh-century Persian geographer al-BTrunl visited India at of the druidic ban.26 Yet they have largely agreed with Caesar in
a time when sovereign administrative practices, in particular, relied holding that if the Celtic priests refused to commit their ancient
crucially on letters, hooks, and note keeping. Yet the classical Arabic poetry to writing, it must have been to preserve some knowledge of
author reports that the Brahmans would not allow the Vedas to be a hermetic kind that letters, being public, would destroy.
transmitted by any means other than oral recitation. “ They do not Such an opposition between writing and secrecy is all too appar­
allow the Veda to be committed to w ritin g, because it is recited ent. One might choose to doubt accounts of civilizations that, while
according to certain modulations, and they therefore avoid the use conceding the existence of writing, insist that it was unknown or
of the pen, since it is liable to cause some error, and may occasion an forbidden to priestly classes. Yet there is a more compelling reason
omission or defect in the written text.” 22 to question the familiar dichotomy of common techniques of writ­
This exclusion of writing from the domain o f priestly and poetic ing and secret practices of oral wisdom. It is that writing, no matter
practice, while striking, is not entirely w ithout parallels among how formal, may also be a means of concealment. Under certain
ancient cultures. In a famous passage o f his Gallic Wars, Julius Caesar conditions, it may even itself become a matter of some secrecy. The
leads one to believe that the company o f Druids played a role similar ancient world was also familiar with peoples who, while appar­
to that of the Indian priests. Druids, the Roman author explains, “are ently without writing, nonetheless employed arts of marking signs
engaged in things sacred, conduct the public and the private sacri­ in special circumstances. Tacitus reports in the Germania that the
fices, and interpret all matters of religion,” being “ supreme authori­ Germanic tribes, while illiterate, practiced an art of divination that
ties” among the Celtic people of Gaul.23 Yet Druids “ do not go to relied on inscribing “ marks” on bits of wood:
war, nor pay tribute together with the rest; they have an exemption
A little bough is lopped off a fruit-bearing tree, and cut into small
from military service and a dispensation in all matters.” They are,
pieces. These are distinguished by certain marks [notis], and thrown
moreover, entrusted with a large body o f poetry, which they must
carelessly and at random over a white garment. In public questions the
memorize: “ They are said to learn by heart a great number of verses;
priest of the particular state, in private the father of the family, invokes
accordingly, some remain in training for tw enty years.” Graphic
the gods, and, with his eyes towards heaven, takes up each piece three
arts, we learn, are expressly forbidden to the Druids, though the
times, and finds in them a meaning according to the mark previously
ancient Celts otherwise do avail themselves o f w riting: “ Druids do
impressed on them.27
not regard it as lawful [fas] to commit [their great number of verses]
to writing, though in almost all other matters, in their public and The nature of these “ marks” is to us, as perhaps it was to Tacitus,
private transactions, they use Greek letters.” 24 Caesar advanced two obscure. Yet it is difficult not to interpret them as the elements of
explanations for this habit, which clearly struck him as in need of some variety of writing.
commentary. The first is that the Druids do not “ desire their doc­ There are also examples in literary history of graphic arts of lin­
trines to be divulged among the mass o f the people.” Their teachings guistic representation that are thought to have once served hidden
are instead to remain esoteric. The second is that the Celts do not and magic ends. A famous example is that of the Northern European
wish to weaken the powers of their recollection, “ since it generally letters called “ runes,” whose related names in Old English, Old
occurs to most men, that, in their dependence on writing, they relax High German, Old Norse, and Old Irish evoke not only written
their diligence in learning thoroughly, and their employment of the signs but also “ mysteries,” “ whispers,” and “secrets.”28 They have

101
100
 DARK T O N G U E S NOMINA SACRA

boon id en tifie d w ith th e v e r y G e r m a n ic “ m a r k s ” e v o k e d by Taci­ name o f physical being, such as a “ rod,” or as the rune commonly
tus, although one can n ot e sta b lish th e lin k w it h c e r t a in t y .29 There called Rad (k), or, by extension , as the sign of the phoneme /r/ it
is, h o w ever, e v id e n ce that th o se w h o c o p ie d r u n e s in th e M iddle represents, w hich is com m only w ritten in Latin letters by “ r.” Only

Ages b elieved th e m se lv e s, at tim e s , to b e t r a n s m it t in g know ledge by conducting these various procedures can one reach the riddle’s

o f secret th in gs. A n in th -c e n tu ry s c rib e in E n g la n d w h o annotated most probable, if un certain solution.33 Even more hermetic are the

ru ne lists in L a tin , for in sta n c e , e x p la in s to h is re a d e rs in a note that conceits o f R iddle 19, in w hich a dozen runes appear in nine lines

“ these fo rm s o f le tte rs are said to h a v e b e e n in v e n te d am on g the of A nglo-Saxon poetry. The tw o types o f writing oblige the eye to

people o f the Northm anni [or N ordm an n ],” a d d in g th at “ th e y still use read in tw o directions at once, for to spell intelligible words in the

them to com m it th eir songs and in c a n ta tio n s to m e m o ry . T h e y give language, the runes must be sounded out from right to left, while

the nam e ‘ru n e -stave s’ [runstabas] to th e se le t t e r s , I b e lie v e , because the Roman letters continue to move from left to right. The solution,

bv w ritin g them th ey u se d to b r in g to lig h t s e c r e t th in g s .” 30 The snac, “ a light, sw ift ship,” is inscribed in the riddle itself, but as an

copyist, o f cou rse, m ay w e ll have b e e n m is ta k e n . Y e t w h e re runes acronym. Its four letters are distributed across those of four common

appear in A n g lo -S axo n lite r a tu r e , th e y b r in g to it a palpab le layer nouns. “ The w ords are spelled backward, so that one takes the ‘s’

o f opacity. from hors ‘horse,’ ‘n ’ from mon ‘m an,’ ‘a’ from wega ‘warrior,’ and ‘c’

T h is is due in part to the fac t th at in th e O ld E n g lis h tradition, from haofoc ‘ h aw k,’ to find the answ er.” 34

each ru n e b ears a sp e cial n a m e th a t, in a d d itio n to designating Such texts m ay w ell be exceptional in the arts of their obscurity.

the w ritte n sign, also has a fu n c tio n in th e la n g u a g e as a common Yet there are practices o f w ritin g that, while far more common, rest

noun den otin g a w o rld ly o b je c t. In th e e ig h th - o r n in th -ce n tu ry on techniques o f ellipsis and dissimulation that are not dissimilar

Rune Poem, w hich has p a ra lle ls in th e s o - c a lle d Icelan d ic Rune Poem in structure. O ne m ay take, as an exem plary case, the standard

and Norwegian Rune Poem, e v e ry stan za tu r n s a ro u n d a sin gle rune, methods o f tran scribing holy names in the Christian manuscripts of

whose various associations it e x p lo r e s in L a tin le tte r s . “ R ath er than antiquity and the M iddle Ages, which Ludwig Traube analyzed in
his path-breaking 19 0 7 book, Nomina sacra.1* Traube borrowed the
using ru n es,” R o b ert D iN a p o li has c o m m e n te d , “ th e p o em is gener­
expression in his title from his older British colleague, Sir Edward
ated by them , or m ore p re c ise ly b y th e ir n a m e s , w h ic h introduce a
Maunde Thom pson, w ho had employed it to refer, as he wrote,
note o f am biguity from the o u tse t, sin c e e ach is th e n am e o f both a
to “ sacred and litu rg ical con tractions” found in Greek and Latin
ru nic character on the page and an o b je c t, b e in g , o r c o n ce p t in the
manuscripts. Such “ contractions” or abbreviations are ubiquitous in
w o rld .” 31 Several o f the rid d le s in th e t e n t h - c e n t u r y A n glo-Saxon
the tradition and can render works obscure even to those who have
Exeter Book exp loit such a m b ig u itie s to an e x t r e m e d e g r e e . Here,
mastered both languages.
ru n es can be the keys to e n c r y p tio n an d d e c r y p t io n , even where,
Traube began by show ing that such abbreviations are fundamen­
s tric tly sp eakin g , th ey do n o t a p p e a r in th e t e x t . R id d le 58, for
tally o f two kinds. There are those, first, in which a word, title, or
exam p le , ends w ith the fo llo w in g p r o n o u n c e m e n t, u tte re d in the
name is shortened by the rem oval of its final letters: such is the case,
first-person sin gu lar o f the rid d le : “ T h e re a re th r e e rig h t runes in
for exam ple, o f Imperator, contracted as “IMP.,” Consul, abbreviated
m y nam e, the first o f w h ich is R a d ” (Pry sin d in nam an / ryhte runsta-
as “C.,” and Dominus, “ Lord ,” shortened as “ DOM.,” “DO.” or sim­
Ja s , para is R ad fo rm a ) .32 To solve th e rid d le , o n e m u st first recover
ply “D.” In such cases, a dot is placed after the last remaining letter
the “ th re e ru n e s” co n c e a le d in th e se lin e s o f p o e t r y in Roman
to indicate that an om ission has been made. The second variety of
le tte rs. T h e reader m ust d e cid e w h e th e r R a d is to b e taken as the

103
02
 DARK TONGUES NOMINA SACRA

shortening results from the decision to remove m iddle letters from deity, they had “ D O ” instead. Sim ilarly, “ SPU,” for spiritu, always
a word while retaining, in each case, the last. Then the final letter represented the Holy Spirit, although the Latin word spiritu might
or letters will be written next to the first. Dominus, in this form, can equally have signified profane “ breath.” Suddenly, it became dear
appear as “ OS.” In such cases, no dot w ill be em ployed; instead, a why an anonymous seventh-century corrector of an Oxford manu­
superscript stroke will hover over the entire shortened form. Traube script had emended every com plete mention of deus and dominus,
called the first variety of abbreviation “ suspension,” reserving the such that both these sacred names appeared solely in contraction.38
old name, "contraction,” for this second. W hile the first consists It was hardly to save space on the page, for the words were already
in a "setting aside” (Weglassen) of parts o f a w ord, the second, as in place, and nothing new w ould be w ritten over the areas from
he put it, brings about an "omission” or “ leaving out” (Auslassen) of which letters were deleted. Omission, in this case, expressed an act
letters. Traube showed that while examples o f “ suspension” can be of pious reverence. The art o f transcription reached a high point
found in pagan as well as Christian manuscripts, the first instances when the holy name came to be rew ritten, its letters blotted out in
of “contradiction” are no older than the third century a .d . and can such a way to give rise to a new and obscure sign. In time, however,
be found solely in Christian texts. W ith considerable clarity and the principles o f the scribal technique were forgotten. Abbreviation
precision, Traube posed, then, two basic questions: that o f the origin by contraction, once a basic feature o f manuscripts in Latin, Coptic,
of the nonclassical form of abbreviation and that o f the reason for its Gothic, and Armenian manuscripts, was by the ninth century a prac­
general adoption by Christian copyists. tice whose principles European scribes could no longer decipher.39
Traube argued that the curious technique o f shortening by con­ According to Traube, the reasons for this art of writing lay in
traction could be explained only with reference to a culture distinct a specifically Jew ish relation to the holy name. Yet the Christian
from those of both Greece and Rome. One must, he wagered, recall nomina sacra also illustrate a linguistic phenomenon that is by all
a basic belief of ancient Judaism: “ the four-letter name is taken to be accounts more general. There are certain words and phrases that

holy, rife with magic, and unspeakable.” 36 For this reason, biblical no one wishes openly to express, and in all human societies, there

Hebrew scribes avoided transcribing it in its entirety. That medieval are linguistic prohibitions that characteristically affect those areas

Latin manuscript practices should derive from an ancient Jewish of life considered to be subject to taboo. In some cases, silence is
formally imposed. Since, however, it cannot always be assured, a set
faith in the magical properties of Hebrew letters, to be sure, was in
of substitutions is also required. Antoine Meillet commented on this
1907 anything but self-evident. In an early appraisal, W. M. Lindsay
fact in a classic essay, first published in 1906. “ In various of the most
noted that Christian paleographers and classicists w ere bound to
diverse parts of the w orld,” he w rote, “ one may observe vocabulary
greet Traube’s thesis with a “ first attitude o f startled incredulity.”37
‘taboos.’ ”40 By way o f introduction, he mentioned South Africa,
Yet Traube marshaled compelling philological evidence in favor of
the Malayo-Polynesian languages, the Far East, Europe, and North
his thesis. He showed that only words related to the Christian deity
America, where “ certain words are forbidden by usage either to a
and holy things were ever subject to contraction. Some such words,
whole group o f people, or to certain particular individuals, or dur­
moreover, were at times also abbreviated by suspension; but in all
ing certain periods and in certain settings.” A taboo, he explained,
such cases, a theological difference explained the tw o techniques.
may concern “ the name o f the deceased, or a leader, or those from
Thus Deo, “ God,” might appear as “ D O .,” but only where it signified
whose family one takes a w ife, etc.” Yet prohibitions can also affect
a pagan god, such as in Latin transcriptions o f V irg il’s Aeneid; wher­
common nouns, “ which may or may not be identical with those
ever the manuscripts employed the word to sign ify the Christian

105
104
 DARK TONGUES NOMINA SACRA

names, which sound in some way identical or analogous, or even requires neither pen nor paper. Nor does it rely for its workings on
onlv partially analogous, to them.”41 any knowing representation o f the units of sound. Sacred words
Haifa century later, Emile Benveniste pursued M eillet’s reflec­ can be modified—and deformed—without a moment’s thought. It is
tions. Meillet had concentrated above all on individual terms that often said that once it is invented and established, a system of writing
are subject to taboo in Indo-European languages. He had observed transforms the sense that speaking beings possess of their language
that certain common names of animals, such as “ bear” and “ snake,” and its elements, from the sentence to the word, the morpheme,
appear to have been avoided by the speakers of languages as diverse and the phoneme. Yet such modes of linguistic representation have
as Russian, English, French, and Greek; as a rule, they substituted their foundation in the sensitivity to oppositions that is the condi­
for them words unrelated to them in their roots. Benveniste noted tion of the faculty of speech. Only beings to whom the distinction
that taboos not only dictated forms of lexical substitution. Subtler between “parbleu!” and “ par D ieu!” is evident—and, one might add,
modes of deformation could also respond to the risks of speech. of religious significance—w ill devise a means to represent it in the
One may, he observed, in technical terms, “ mutilate the word ‘God’ building blocks of signifying sounds. For the same reason, only they
through the aphaeresis of the final syllable” or by the substitution will act, whether in speaking or in writing, to alter the expressions
of an assonating term: in place of the exclamation “ by God!’’(“par of their languages—letter by letter, syllable by syllable, according to
Dieu!”), for instance, one may say “ pardi!” or “ parbleu!” although the occasion—at times to conceal them, at times to mutilate them.
neither phrase, strictly speaking, can be reduced to words present To this degree, at least, mortals act, by nature, like the gods of the
in the lexicon of the language.42 To avoid the infraction of a pro­ Vedic tradition. By substitution, suppression, addition, and contrac­
hibition, speakers may even go so far as to create an entire “ form tion, they work and play, modifying the constituents of speech to
of non-sense” produced by the scrambling of the sounds that make render the familiar unfamiliar. One might object that the changes
up a single, blasphemous phrase. Thus, in place of “ par le sang de wrought upon a language to avoid blasphemy or the infraction of a
Dieu!” (“ by God’s blood!” ), Benveniste reports, one says “ palsam- taboo proceed from a powerlessness that lies at the furthest extreme
bleu!” and instead of “ je renie D ieu!” (“ I deny G o d !” ), one will from the mastery with which immortal beings must act. The dif­
prudently, if somewhat meaninglessly, exclaim “ jarnibleu!”43 It is ference is certain, yet it may be of less consequence than it appears.
not difficult to find examples of analogous formulations in other One fact, in any case, cannot be doubted. Whether for piety or for
tongues. In 1944, H. L. Mencken recorded no fewer than ten pho- pleasure, mortals, too, like to keep names of important persons and
nologically related English euphemistic substitutes for the single objects secret. All speaking beings, swearers and scribes no less than
name “Jesus,” all of which were in use in the language spoken by gods and priests, share an attraction to what is hidden and a discom­
one inhabitant of Dalton, Pennsylvania: “ gee,” “ jeez,” “ jiminy” (or fort before “what lies in plain view.”
“jeminy” ), “Jemima,” “Jerusalem,” “ Jehosaphat,” “gee-whizz,” “gee-
whillikins,” “gee-whittaker.”44
Such distortions all involve the constituent elements of forbidden
words, being examples of reorderings of letters. Yet they presuppose
no craft of writing, at least in the usual sense o f a conscious practice
of transcription. The passage from “ par D ieu!” to “ pardi!” or from
“Jesus” to “gee,” “gee-whizz,” “gee-whillikins,” and “gee-whittaker”

106 107
 CHAPTER NINE

Anaphones

In the winter o f 1905, Ferdinand de Saussure, professor of Indo-

European languages at the University of Geneva, took a leave from
teaching. He meant to make a “ complete break” with his “customary
occupations and preoccupations,” as he recalled in a letter to his
former student, Antoine M eillet.1 In the company of his wife, Marie-
Eugenie, he traveled to Italy. Saussure spent a month in Naples
before moving on to Rome, where he lingered in the Forum, pon­
dering the remains of its monumental inscriptions. “ No need to tell
you,” he wrote to M eillet in a note from January 1906, “ that here I
am not doing much. The archaic inscription in the Forum is just the
diversion I need to rack my brains. Nothing to be drawn from it, of
course, but it is interesting to contemplate the enigmatic block and,
by sight, to confirm the transcriptions.”2 Among the texts chiseled
on stone were some of the oldest surviving examples of Latin poetry:
epitaphs and triumphal commemorations dating from as early as the
third and second centuries b .c . These brief and often fragmentary
texts, the oldest in the Roman literary tradition, are composed in
the Italic verse form commonly known as the “ Saturnian,’ which
the Roman authors employed before their successors adapted the
meters of Greek literature to suit the Latin language. It would seem
that in the Roman Forum, it was this archaic metrical form, more
than the contents of any one inscription, that unexpectedly caught
Saussure’s attention.
The Saturnian was, and still remains today, in large part, a
mystery.3 The Romans of the classical age may themselves not have

109
 AN APH ON ES
DARK TONGUES

understood it. Often they claimed that the old Roman form of verse simple Greek hexameter,” Saussure added, merely requires certain
possessed no rhythm whatsoever. Grammarians decried its linguis­ “ licenses,” which he went on to enumerate in detail.
tic liberties and lack of order; poets described it as belonging to a Soon that metrical explanation seemed to him inadequate. Later
crude and undisciplined epoch in history. Horace, for one, did not the same month, Saussure w rote again to Bally, correcting his
hesitate to call it “ horrid” (horridus).4 In Saussure’s own lifetime, account. “ Without abandoning what seemed to me true in metrics,”
however, a new interpretation of the meter had been proposed. he wrote, “ that is, that it is a question of a dactylic or rather spondaic
It was argued, particularly in Germany, that the Saturnian was meter of five or six feet, with an anapest or an amphibrach as an
not quantitative, like Greek and classical Latin poetry, but rather acceptable substitute of a spondee (since they too hold as — ), I see
accentual in nature, like verse in German, Russian, and English.5 Its now that one must at all costs resolve the problem of Alliteration.” 10
lines, in other words, would have been organized not by the regular This “problem” was not strictly speaking metrical, since it affected
alternations of long and short vowels but by the ordering of sets of neither vowels nor accents taken in quantitative proportions. “ The
accented and unaccented syllables. Saussure does not appear to have problem of Alliteration” involved, instead, other relations of words:
commented extensively on this hypothesis, either to assert it or to affinities and symmetries between the textures of vowels, con­
deny it, although having studied in Germany and having continued sonants, and whole syllables of Saturnian poetry that seemed to
to follow the research of his colleagues beyond the Rhine, he most Saussure to echo each other, both within individual lines and across
likely knew it well. As he examined the archaic Roman inscriptions, them, in almost systematic form. The linguist was now convinced
his attention was drawn, instead, to what seemed to be an odd lin­ that these striking symmetries of letters played a crucial part in the
guistic phenomenon: “something as bizarre,” as he later wrote to art of Saturnian verse.
Meillet, “as phonic imitation.”6 Elements in the sound structure of Saussure, however, also took a further step. He held that the
the archaic Italic poetry seemed to m irror each other, as if obeying traits he had noted in the Forum were not exclusive to archaic Latin
some unknown law of “ phonetic harmonies.” 7 literature in verse. Similar phonetic phenomena could also be found
Upon his return to Geneva, Saussure persevered in his Saturnian in later Roman poetry. Moreover, he now explained, even beyond
studies. He began to investigate the entire corpus, which has been the limits of the Latin literary domain, one can find examples of
transmitted in a small number of precious manuscripts as well as a systematic mirroring in the vowels, consonants, and syllables of
epitaphs. Soon he believed he had found the forgotten key to the the kind present in the Saturnian. In the hymns of ancient India,
Italic form of verse. “ By dint of turning the Saturnian upside down the songs of medieval Germany, and the epics of classical Greece,
in all directions,” he wrote in a letter to his former pupil, Charles Saussure detected something that had gone unnoticed by scholars: a
shared Indo-European poetic technique involving the sound shape
Bally, in July 1906, “ I have reached a solution.” 8 He added that it
of words. Comparative grammar had, by this point, established
was only provisional, and he entreated his correspondent to keep
that many of the ancient languages o f Latium, India, Germany,
it to himself. “ The solution,” he announced, “ is so simple that it
and Greece were structurally as well as genetically related. Now
makes me laugh when I think of the detours I had to go through
Saussure suggested that one might find analogous correspondences
to get there. For me, the Saturnian is no longer anything but the
between the poetic practices attested in these various languages.
pure and simple Greek hexameter, adapted in such a way so as to
“All this leads me to think,” he wrote, summarizing the path his
allow substitutions of spondees by anapests, <the tribrach>, and the
research had taken, “ that if the Saturnian, with its considerable
amphibrach as well as the dactyl.”9 This adaptation of the “ pure and

111
110
 DARK TONGUES ANAPHONES

metrical affinities for other ancient Indo-European types of verse, these notebooks was generally unknown at the time o f the linguist’s
concealed, moreover, an extraordinary system o f combinations of death in 1913, and it remained so for h alf a century after the posthu­
syllables, consonants, and vowels, all regulated by an unapparent mous appearance of the famous Course in General Linguistics in 1916,
law, perhaps there was something similar in the poetry of India, and which Charles Bally and Albert Sechehaye prepared on the basis
Greece itself?!”'1 of notes taken from Saussure’s lectu res.15 O nly in the m id-1960s
Saussure distinguished this “ unapparent law ” from the various did Saussure’s research into the “ unapparent law ” o f Indo-European
rules of language that define individual grammars and that linguists verse come to be known, thanks to the five articles on Saussure that
aim to lay bare. Those rules, he maintained, obtain in the absence Jean Starobinski collected in his 1971 book, Words upon Words: Ferdi­
of any clear awareness on the part of speaking subjects, in the sense nand Saussure’s Anagrams.16
that those who speak a language need not possess any conscious In deference to Starobinski’s subtitle and Saussure’s own sugges­
knowledge of its grammatical system to produce utterances that tions, the notebooks are today generally thought to contain a theory
will nonetheless conform to its many constraints. The “ bizarre” of anagrams. That term, however, is for several reasons imprecise.
principle of Indo-European poetry, he held, was of another kind. Strictly speaking, an anagram consists in the reordering o f the let­
Saussure wagered that it was a conscious inheritance that rested on ters in one name (or noun), such that they come to form another
an art knowingly, if secretly, passed on from poet to poet, genera­ expression. William Camden, one o f the first English writers to dis­
tion to generation. “ The results are no less confounding for their cuss the procedure, defined it in classic terms when, in his Remains
Indo-European significance,” he observed. “ It is no longer a ques­ concerning Britaine o f 160^, he explained that “Anagrammatisme, or
tion of something unconsciously transmitted, such as a language, or, Metagrammatisme,” is “ a dissolution o f a Name truely written into
in poetry itself, rhythm and visible form; it is, rather, necessarily a his Letters, as his Elements, and a new connexion of it by artificial!
secret of the poets, the Kavayas, the vates, which requires that a for­ transposition, without addition, substraction, or change of any let­
mula be learned and then transmitted from master to disciple over ter into different words, making some perfect sence applicable to
thousands of years.” 12 An archaic convention would have demanded, the person named.” 17 Examples o f such transpositions can already be
in other words, that poets acquire not only “ the traditional Indo- found in the literature of the ancient Greeks, who coined the noun
European versification” that was “ immediately grasped by the ear,” “anagrammatism” (anagrammatismos) and the verb “ to anagramma­
but also something else: “an obscure, almost secret thing, which is, tize” (anagrammatizein) for such “ artificial transpositions.” 18 In the
for us, absolutely astonishing, a thing such as the counting of pho­ second century b . c ., Lycophron, for exam ple, made o f the name
netic elements by ‘regular’ arithmetical addition and subtraction.”13 Ptolemaios “Apo melistos,” “ From Honey,” and turned Arsinoe into
In the years following his return from Italy, Saussure dedicated “Hera’s violet,” “ Ion Heras.” 19 The “ phonic imitations” that Saussure
the greatest attention to the study of this “ obscure, almost secret believed he had discovered in ancient Indo-European poetry are of a
thing” in poetry. Suffering from what he once termed an “ incur­ different nature. Their rules are more numerous, their elements are
able graphophobia,” he neither sought to publish his findings nor more complex, and their reality is less certain.
expressed any wish to see them in print.14 Yet between 1906 and Between 1906 and 1909, Saussure’s theory underwent several
changes. One may nonetheless present its basic lineaments in a
1909, he devoted no fewer than ninety-nine notebooks to elucidat­
single form. First, Saussure established, through a close study of
ing the principles of the Indo-European “ phonetic harmonies” that
Saturnian poetry, that, within a single line o f verse, individual
he first perceived in fragments of old Italic verse. The existence of

112 1F 3
 DARK TONGUES ANAPHONES

sounds—vowels as well as consonants—appear in groups of two:

phonological “couplings” are the rule. “ A vow el has the right to in sedeto: venis-se
appear in the Saturnian,” he explained, “ only i f it has a countervowel (manensse-: ven(i)sse
in some place whatsoever in the line.” 20 The only liberties that the
ENT: ENT: in carp-ento: veh-ent-e
ancient poets gave themselves for vowels were three “ compromises”
UMV: UMV in donicumvid: domuraven25
(transactions) bearing on timbre: “ ist, between short e—short 1, short
o—short u; 2nd, sometimes e : ei; 3rd, som etim es o: u.” 21 Similar Such a reading reveals several o f the fundamental principles
“couplings,” Saussure argued, dictate the quantity o f consonants. that underlie Saussure’s representation of the “ counting of pho­
Where the number of vowels or consonants in a line is odd, rather netic elements” in verse. The elements in question, first, are not
than even, there will naturally be a “ rem ainder” (residu), an ele­ letters but phonemes. They may therefore be spelled in different
ment without a corresponding partner in the line. Yet in such cases, forms with varying diacritics, their letters halved or doubled and
Saussure remarks, “one sees it reappearing in the next line as a new treated in isolation and in larger groups. Monophones, diphones, and
remainder, corresponding to the overflow o f the preceding line.”22 polyphones, second, are dispersed across both the single line and
Next, Saussure applied this principle to a higher level of verbal beyond it; there can then be parallelisms between not only sedeto
construction. As he explained to M eillet in a letter dated September and videbis in the same line but also between donicum and ddmum
23,1907, when one pushes one’s analysis beyond the level of the sin­ between two. Third, single elements can be counted in different
gle sound or “monophone,” one observes a more complex phenome­ forms according to their position. In Saussure’s representation, the
non that is at once “more easily grasped, yet more difficult to reduce single word venisse, for example, contains no fewer than six distinct
to a fixed formula.”23 Sets of two and three phonem es—“ diphones” units of phonic imitation: ve; nr, en; se; sse; and ensse (according to
and “ triphones” —will mirror each other in remarkable symmetry. a spelling nowhere present in the transcription). One might well
“ In fact,” Saussure notes, “ almost any Saturnian passage is no more understand such sonic poetic material in terms of alliteration. Yet
than a swarming of syllables or phonic forms echoing each other.” Saussure argues that such a traditional figure of sound play cannot
As an example, he cited two lines from Livius Andronicus: do justice to the “ unapparent law ” that commands the construction
of this form of verse. “ The wTiole phenomenon of alliteration (and

{
Ibi manens sedeto donicum videbis
also rhyme) that one may observe in the Saturnian,” he asserts in
Me carpento vehente domum venisse.24
1907, “ is nothing but an insignificant part of a more general, or,
There followed the following “ polyphonic” analysis: rather, an absolutely total phenomenon.” 26 Here one finds far more
than the alliterative echoing of certain letters with each other,
DE: DE in sedeto: videbis. for each and every portion of the poem can play several parts in
BI: Bl in ibi: videbis. this “extraordinary system o f combinations of syllables, consonants,
DO: DO in donicum: ddmum and vowels.”
VE: VE in vehente: venisse The formulation o f the system demands, however, a further prin­
TO: TO in sedeto: carpento ciple. Not only do the vowels and consonants in individual and adja­
NI: Nlin donicum: venisse cent lines correspond to one another, and not only do whole group­
EN: EN in man-en-s: v-en-isse ings of syllables also echo each other in fixed form. These elements,

US
H
 DARK TONGUES A N AP H O N ES

when conjoined, also spell out special nouns and nam es. “ As soon as
his vocation, dedicated himself to the phonic analysis of words. It is
the occasion is given," Saussure continues, “ the poly phones visibly
most probably this science of the sonic forms of words that from the
reproduce the syllables of a word or an im portant nam e for the text,
earliest Indo-European period lent superiority, a particular quality, to
then becoming diagrammatic polyphones.” 27 It seem s that Saussure
the Hindu Kavis, the Latin Votes, etc.30
was quick to reach this conclusion. In a le tte r to B a lly from 1906
evoking his Homeric studies, Saussure rep orts h avin g detected the Saussure’s “ thesis” is w orth unfolding. It holds that despite their
presence of "theme w ords” (mot-themes) or “ key w o rd s” (Stichworte), apparent diversity, the Indo-European poets all possessed a special
whose constituent phonic and syllabic elem ents are scattered across knowledge o f the properties o f sound structure, which they jealously
the epic texts. “ What is most absolutely certain for m e ,” he declares, kept to them selves: hence their tacit “ superiority” with respect to
the other m em bers o f their societies. Their special “ science” would
is that the entire text of the Homeric poems (or, if it is not entire, this
have allow ed the poets to order and reorder units of sounds, just as
will be one easy way to see which parts were added) rests on a secret
m athematical know ledge allow s those who possess it to submit indi­
law, <in which> the repetition of vowels and consonants of an abso­
vidual num bers to the operations o f “ ‘regular’ arithmetical addition
lutely fixed number, on the basis of a keyword [Stichwort], a THEME
and subtraction.” “ Every Latin poet [Vates] was above all a specialist
WORD, is observed, from line to line, with an admirable and total
in facts o f phonem es.” 31 For this reason, Saussure saw no difficulty
precision— One can <already> say that the whole text of Homer is
in holding that the Sanskrit Kavis and the Latin Vates also carried
nothing but a vast and continuous anagram, running on the key word
out com plex procedures o f phonological inversion and transposition.
[SricAivort] or key words [Stichworte] that are renewed every two, or two
Metathesis, he suggested, was, to the archaic poets, a familiar trans­
and a half, or three lines, without the least imprecision with respect
form ation. In reading and in analysis, one must keep its incessant
to the number of consonants, vow'els, and hiatuses demanded in this
possibility in m ind. EM, in an ancient text, can also be ME;32R T may
space by the key word.28
be deciphered as 7 7 1 ;33 TAE can become TA—E, or TE—A .34

Audible phonetic play, therefore, points to an unspoken noun If one approaches the Saturnian by means of “ this science of the
or name. Saussure did not hesitate to evoke an art o f circumlocu­ sonic form s o f w o rd s,” hidden “ theme words” and “ key words” sud­

tion. Vowels, consonants, and syllables might all compose a “ phonic denly appear. Saussure cites, as an example, a single line of a Roman

paraphrase” “according to regular form s .” 29 As long as he pursued epitaph that conceals the constituent elements the name SCIPIO in

his studies of Indo-European poetry, Saussure sought the rules that dislocated fo rm :3S

would define these forms. He believed them to be independent of

purely metrical considerations:

If all syllables can lay claim to phonic symmetry, it follows that what
dictates these combinations in no way depends on the line and its
rhythmic scheme. A second principle, independent of verse itself, was
joined to the first to constitute a specific poetic form. To satisfy the
condition of the “poem” (carmen), which is completely independent of
the feet or ictus, I state, in fact, (and this is my thesis) that the poet, in

116
 DARK TONGUES ANAPHONES

Vet this example runs the risk o f sim plifying the principle of Meillet, entire hymns are made from the names of gods.41 Evoking
phonic harmonies, for it suggests that the units o f a “ theme word” the first text in the collection, a poem in praise of the god Agni,
can he restricted to the space of a single line o f poetry. This need Saussure comments that “ this hymn positively declines the name of
not he the case. The elements of a certain “ key w ord ” can also be Agni, and it would be very difficult to think that the succession of
distributed over more than two lines. To be sure, Saussure began lines—some o f which begin with ‘Agnim ide,’ others with ‘Agnina
with the supposition that there would be, at m ost, one or two pho­ rayim a£navat,’ others with ‘Agnaye,’ ‘Agm ’ etc., mean nothing for
nic “ remainders” in each line; but he soon came to renounce that the divine name.”42 In other cases, however, Saussure suggested that
principle. He remained committed to the task o f establishing some the phonic shape o f texts might be crucially determined by words
criterion that would allow for the identification o f an “ anagramma­ that remain unsaid in them. In the opening of Lucretius’s De rerum
tized” term in a given work, yet the conditions that he demanded ran natura, Saussure thus found the many sonic elements of a Greek
the risk of becoming ever less exacting. He dim inished the number name nowhere mentioned by the Roman poet: “Aphrodite,” whom
of the phonic imitations he required and extended the textual space Lucretius, for this part, names only “ Venus.”43
over which they might be strewn. He began w ith the supposition Over the course of the four years in which he sought to recover
that the correspondence would be distributed over a single line; fail­ the secret art o f the Indo-European poets, Saussure gave sev­
ing that, he later added, one may study two, six, or even eight lines.36 eral names to the ancient practice of “ phonic paraphrase” that he
There is no doubt that the “ law” of sounds, then, grows increas­ believed he had uncovered. “ Anagram” did not fully suit him, for
ingly lax. Some have wondered pointedly w hether the entire theory several reasons. In addition to evoking the procedure of reordering
of anaphones does not become “ deeply questionable” as a result.37 elements within the limits of individual words, that term refers, in
Saussure himself appears to have noted this danger. As if to ward it its root, to the elements of writing: grammata, “ letters.” Saussure’s
off, he developed a further rule that was to allow him to assert the archaic literary corpus was in large part transmitted orally, and he
existence of a technique of “phonic paraphrase” w ith greater cer­ did not wish to suggest that the system of phonic paraphrase he had
tainty. He argued that at certain points in ancient literary works, one reconstructed rested essentially on a graphic practice. He sought,
can detect a passage in which the trace of the entire anagrammatized therefore, to dispel the misconception that might follow from his
word appears: “ a closed and limitable sequence” in the lines of the usage. “ Em ploying the word anagram” he specified, “ I certainly
poem that “one may designate as the place specially devoted to this do not mean to appeal to the intervention of writing.... Anaphony
name.”38 Saussure called this site the locus princeps. A “ key word” would be better, in my mind, but this last word, if one creates it,
can appear in several ways within its borders.39 Its first and last let­ seems suited to performing another duty, namely, that of designating
ters may coincide with the first and last letters o f this locus; in such an incomplete anagram.”44 Elsewhere, Saussure referred to “ hypo-
cases, it may be called a “ mannequin,” which “ encloses, within its grams,” evoking the Greek word for the inscriptions at the bottom
own limits.. . the complete syllabogram.”40 At other points, Saussure of steles, or the pigment used for the makeup that lies under (hypo)
granted that “ theme words” can be present in w orks by more and the eyelid.45 He also evoked the “ paragram,” a classical term for the
less visible means. Sometimes Stichworter may be explicitly stated, rearrangement o f letters in coded messages, and other neologisms
being the names of the fundamental figures o f the literary works and technical expressions: “ paramorph,” “ homogram,” “ logogram,”
that contain their dismembered parts. In the Rig Veda, a collection “antigram,” as well as “ homogram.”46For cases in which “ key words”
of poetry “ literally covered with anagrams,” as Saussure wrote to remained unspoken in the works that systematically recombine their

118 119
 DARK TONGUES

phonic elements, he wrote of "cryptographic” anagrams, for which,

The sign that the ancient poets systematically sought to dis­
at times, he also used the simpler term, "cryptogram .”47 Yet if one
simulate—be it name or noun, person or thing—remains, therefore,
m alls that the phenomena he conceived need not rest on techniques
obstinately opaque. Yet there is more. The reason for which the votes
of writing, one may consider them all to be, more simply, varieties
developed and employed their “ science” is also unclear. Saussure
of "anaphones."
suggests two possibilities: anaphones may have served a “ religious”
Saussure’s theory of the secret science o f sonic forms remained
or a “ poetic” aim. This possibly exclusive alternative already appears
equivocal in several crucial respects. Beginning w ith his first inves­
in the first of the linguist’s notebooks: “ From the moment in which
tigations into "phonic imitation,” he appears to have held that the
the poet was obligated, by religious or poetic law, to imitate a name
regularities he detected were essentially lexical in structure and
[or noun: nom], it is clear that having distinguished the syllables,
that the units of sound he tracked, although scattered across many he was, without having willed it, forced to distinguish its forms.”51
lines, composed important words. For Saussure, this appears to Elsewhere, w riting of the Saturnian, Saussure explicitly confronts
have been a point of fundamental im portance, which distinguished the question of the purpose that Indo-European “phonic paraphrase”
the poetic regularities he posited from traditional Indo-European was to serve. Underlining his uncertainties, he notes: “ The reason
features of versification, such as alliteration, assonance, and rhyme. may have been the religious idea that an invocation, a prayer, a
Those were, to him, strictly metrical phenomena, transmitted from hymn, is only effective on condition of mixing the syllables of a
poet to poet and generation to generation, like all the properties divine name into the text__ The reason may have been nonreligious
of individual languages and language groups, in the absence of and purely poetic, of the same kind that determines rhymes, asso­
conscious decision and awareness. “Anagram m atic” or "anaphonic” nances, etc.” 52 He could also present this hesitation in positive form:
regularities were, for him, of a different nature. They bore witness “ The function o f the anagram (as such) can be understood, with­
to a self-conscious knowing, a "science” possessed by poets that out contradiction, in many ways, just like its relation to the more
they could exploit as they saw fit, concealing precious terms. Yet general forms of phoneme play; thus, the question, in all respects,
precisely on this matter, Saussure seems to have left one question admits of many solutions.” 53
open. What are the precious words that poets wished to conceal: Most im portant to Saussure, it seems, was the principle that
names or nouns? Does the difference matter? Names could be read “ phonic imitation,” in any case, be an intentional procedure in the
as signs of individuals, nouns, by contrast, as allusions to concepts or making o f the poem. Only an anagram consciously willed, for him,
the things that they subsume. But the possibilities are many. In the counted as a “ real anagram.” 54 Yet the intention is susceptible to
notebooks, scattered names outnumber dispersed nouns. Writing to several readings, which may be not only distinguished but opposed.
Bally in 1906, Saussure goes so far as to assert that the “ main word” One can certainly imagine a case, such as that of the Vedic hymn, in
(mot principal) is “always a name” (un nom propre) . * 6 Yet Saussure also which the religious function includes the poetic; but one may also
uncovered nouns. O f the eighty-eight “ cryptogram s” the linguist conceive a case, such as that of Epicurean treatise De rerum natura,
identified in the De rerum natura, for example, only forty-three are in which the poetic function excludes any positive religious mean­
names.49 Sometimes he argued that both types o f words function in ing. The reason for the secret art, in any case, remains structurally
the same work: in his Lucretius notebooks, for exam ple, Saussure indeterminate: a datum as impenetrable as the grammatical object
tracked the movements of both the name “ A phrodite” and the noun of the hermetic technique. Just as it is necessary that the “unapparent
postscenia, “ backstage.” s0 law” of phonetic distribution conceal a word whose formal identity

120
 DARK TONGUES NAPHONES

not'd not bo revealed, so the effects of phonic paraphrase result from harmony” remained essentially untouched by grammatical varia­
an intention whose identity cannot be unveiled. tion, so he moves, without methodological scruples, between the
Yet the most manifest mystery of the anagram notebooks involves most distant of epochs. There is evidence that Saussure believed the
the place they occupy in Saussure’s new conception of the science of technique that he had recovered was practiced in the Middle Ages,
language. That Saussure’s study of the Saturnian and other ancient the Renaissance, and even the recent past. “ From the most ancient
Indo-European poetic forms is linguistic in character may seem an Saturnian monuments up to the Latin poetry written in 1815 or
obvious fact today. Saussure, after all, shows little interest in the 1820,” we read in one note, “ there has never been any other way of
interpretation of the works he reads; at each stage, he focuses instead writing Latin verse than by paraphrasing every proper name by the
on the phonological composition of the poems, as well as on the regular forms of the hypogram.” 57
nouns and names that they suggest. Saussure was, moreover, active Even if one sets aside the question of the distinction between the
in those years as a professor of linguistics. The period in which he synchronic and the diachronic, however, the contrast between the
worked on the “obscure, almost secret thing” of the archaic poets principles of the Course and the method of the notebooks is striking
was also, in part, that in which he gave the lectures in Geneva from and far-reaching. The lectures on general linguistics define the unit
which his students would later draw the Course in General Linguistics. of language as the sign. This fact entails, for Saussure, at least three
Saussure worked on cryptograms from 1906 to 1909, and he gave his basic theoretical consequences. The two faces of the sign—signifier
famous course on the science of language in 19 0 6 -19 0 7 , 1908-1909, and signified—are tied by a principle of association; their link is
and 1910-1911. Yet the relation between Saussure’s study of phonic “arbitrary” in nature and, as a two-faced whole, they form a being
harmony and his general linguistics is anything but simple. On sev­ defined in purely differential terms with respect to all others. 58
eral counts, it even seems that the theory of anaphones is at odds Jean-Claude Milner has shown that the notion of the Indo-European
with the fundamental principles of the Course. “anagram,” as Saussure treats it, excludes each of these three prin­
It is well known that in his systematic account of the science ciples. 59 Saussure considers the anagrammatized name or noun, first,
of language, Saussure teaches that the object o f linguistics is “ lan­ not as an element of a certain language, but as the name of a referent
guage” (langue), grasped as a system o f signs susceptible to syn­ or a part of the world: not, in short, as a “signified,” but as a “sense,”
chronic and diachronic study. The investigation into anagrams a linguistic expression that, for the purposes of analysis, has become
seems irreducible to both these forms of research. It is clearly not synonymous with a real thing. The anagrammatized term, second,
synchronic, for it does not examine a single state o f a language in cannot be considered to be arbitrary; on the contrary, “ its func­
the “static perspective” in which it presents itself to its speakers. tion consists in imposing a necessity on the phonemes in the poem,
Saussure showed little concern, for example, for the system of Latin withdrawing them from the contingency that affects lexical units. ” 60
anagrams in the archaic period of the Saturnian, in distinction from Finally, the “ anagrammatic” word is fundamentally nondifferential,
those that followed it. But the analysis of anaphones is obviously also for its identity can be established independently of its opposition to
not diachronic in character.” Saussure seems to have no interest any other words. The sign exists only in the system of differences
in any “evolutional fact” (fait evolutif) in phonic m irroring, for the that links it to and distinguishes it from all other signs in a given
notebooks contain no study of the process, whether “ prospective” or language; it has, that is, no positive self-identity. The Stichwort, by
“retrospective,” by which the “ unapparent law” developed.56 Just as contrast, requires no opposition to be recognizable as such. It is an
Saussure passes from one language (langue) to another, as if “ phonic axiom of Saussure’s method that wherever it appears in the literary

122 123
 DARK TONGUES ANAPHONES

work, no matter the degree of phonic decom position, the crucial on anaphones, he appears to have passed from a state of conviction to
term—whatever it may be—w ill continue to appear as itself. one of skepticism, then grave doubt. In 1906, writing of the Satur­
Yet the notebooks belie another linguistic principle that is funda­ nian, he had evoked the discovery of a “ solution... so simple that it
mental to the teachings of the Course. It is the principle o f the “ lin­ makes me laugh when I think of the detours I had to go through to
earity" of speech.MIn his general linguistics, Saussure accords great get there.” He already had announced the possibility that the “ pho­
importance to this point: “ The signifier, being by nature auditory, nic m irroring” he had observed might be common to other Indo-
unfolds in time alone, and it has the characters that it borrows from European poetic traditions, such as that of Sanskrit poetry. By 1907,
time: a) it represents an extension, and b) this extension is mea- he believed the Vedas to be “ literally covered with anagrams” and
sureable in one dimension alone: it is a lin e.”62 “ In discourse,” we Homer’s epics to be “ a vast and continuous anagram.”65 In January
read later in the Course, “ words, by virtue o f their sequence, main­ 1908, he wrote to Meillet: “ I have attained a state of certainty that
tain relations founded on the linear character o f language, which far exceeds, and that has no analogy with, the one I reached during
excludes the possibility of pronouncing tw o elem ents at once— these last months.”66 That, however, was the year that Saussure’s ana­
They always come after each other in the speech sequence.”65 Sau- grams began to “ proliferate.”67 He now concluded that the “ furor of
ssure’s anaphones openly flout both consequences o f this principle. phonic play” was more long-lived than he imagined.68 Not only did
Several of their parts, in the unfolding o f a poem , may well appear the linguist now claim to find its signs in the classical Roman poets,
“at once.” It suffices to consider a “ polyphone” that can be counted such as Ovid; he also announced that he had found examples of the
simultaneously in several forms: once, for exam ple, as TAE, once secret procedure in the neo-Latin writers of the Renaissance, such
as L 4 and £, and once as TE and A or TA + TE. The idea of “ phonic as Angelo Poliziano, and even in more recent Latin poets, such as the
mirroring,” moreover, suggests that “ sonic form s” can appear not late seventeenth-century English Latin author Thomas Johnson.69
only at once and in succession, but also in transposition: backward In an experimental science, confirmation by numbers is double-
and forward, in retrogression no less than in progression. edged. Saussure responded with elation at the discovery of increas­
That Saussure was aware of the troubling im plications of these ing examples o f the “ obscure, almost secret thing” he believed he
facts can be gleaned from several passages in his notebooks, where had discovered. Yet it also occurred to him that should he be able
he explicitly raises the question o f forms intelligible “ outside” the to find no examples of speech from which the effects of phonic play
order of time. “ Can one present TAE as TA + T E ,” he asks in one were absent, their presence would mean little. The notebooks sug­
manuscript, “ that is, invite the reader not to juxtaposition in con- gest that he sought a rigorous demonstration of his thesis. Strictly
secutivity, but to a means of acoustical im pressions outside time? speaking, the proof of his theory of anagrams, as Milner has written,
Outside the order that elements have in tim e? O utside the linear “would have to have the following form: establishing 1) that there
order that is observed if I present TAE as T A -A E or T A -E, but not if are texts without anagrams; 2) that all the anagrams observed are
I present it by TA + TE, amalgamated outside tim e, as I could do if the effects of a specific technique.” 70 Saussure did not prove the first
they were two colors?”64 of these two propositions. One may, of course, interpret the wid­
One may certainly wonder how Saussure would have reconciled ening of his corpus, from 1906 to 1909, from Saturnian poetry to
the methods and findings of his notebooks with the science taught in Vedic hymns, Homeric epics, and modern Latin verse, as an indica­
his Course. It is worth recalling that he chose to publish neither. Yet tion that he was searching for some index of falsifiability: evidence
it must also be said that in the four-year period in which he worked of literature that, precisely in lacking “ phonic mirroring,” might

124 125
 DARK TONGUES ANAPHONES

confirm its certain presence elsewhere. Yet Saussure appears to have was at its apogee. Readers found in the notebooks the signs of a “sec­
found no texts that did not obey the “ unapparent law” he discerned ond Saussure,” a “ dark double” of the thinker who had first defined
in the Saturnian. Even classical prose proved no exception. In works language as a system of signs.76 In Words upon Words, Starobinski
bv Plinv, Caesar, and Cicero, Saussure detected a “ constant preoc­ presented a critical account of Saussure’s research that determined
cupation” with the shapes of sounds that, when examined closely, much of its subsequent reception. Having retraced and commented
pointed “ hypogrammatically” to certain key words.71 upon the genesis of the theory of anaphones, Starobinski judged
Perhaps for this reason, Saussure finally chose to test the prin­ Saussure’s findings to be questionable, and he concluded that in any
ciple according to which anagrams, when present, are the signs of case, the linguist’s critical methods were unsound. He had mistaken
poetic intention. In 1909, he wrote to a living Latin poet, Giovanni unintentional literary effects for the signs of a willed system. To this
Pascoli, to ask whether “ certain technical details” in his verse degree, he argued, “ Saussure’s error” taught one “exemplary lesson”:
resulted from a special system. Saussure promised that, if Pascoli how difficult it is for the critic to avoid taking his own find for a rule
obliged, he would send him a full exposition of the regularities he followed by the poet. Having made a discovery, the critic has trouble
had observed in Latin poetry.72 The Saussure archives do not contain accepting that the poet did not consciously or unconsciously want what
the poet’s response, but it appears that he did answer Saussure. The analysis only s u p p o s e s . The critic has trouble remaining alone with his
linguist’s archives contain a copy o f a second letter to Pascoli, in discovery; he wants to share it with the poet. But the poet, having said
which Saussure permits himself to go into greater detail about the what he has to say, remains silent. 77
phonic procedure. This letter, however, seems to have met with no
response. Saussure appears to have abandoned his studies of “ phonic Those who followed Starobinski often echoed that judgment. The
mirroring” soon thereafter. It is possible that, as Starobinski main­ scholarly value of Saussure’s findings was generally considered null.
tains, Pascoli’s silence was the cause of the linguist’s decision to Critics had few hesitations about evoking a mental malady that would
“ interrupt his inquiry into anagrams.” 73 Yet by 1909, Saussure also have besieged the linguist: “ madness,” “ dementia,” or even “schizo­
had his own reasons to doubt the likelihood o f the system he had phrenia,” they variously argued, could be detected in the pages of
reconstructed. He had repeatedly confronted a troubling question the anagram notebooks.78 Such wishfully medical judgments of the
that he had no means to resolve: that o f chance, or contingency. linguist were once commonplace.79 The critics failed to note that
The “materiality” of phonic facts, as he admitted in one notebook, in issuing such pronouncements, they merely echoed the linguist’s
“might be due to chance.” In another passage, an imaginary oppo­ own doubts, while adding to them a diagnosis of mental illness. One
nent to his theory similarly declares: “ Chance can do anything in might also reason otherwise. The fact that Saussure chose to end his
three lines.”74 Repeatedly, Saussure suggests that only a calculus of inquiries into anaphones could also be taken as a sign of his scientific
the probable distribution of sounds in a language would be capable of rigor. John E. Joseph, the linguist’s biographer, has advanced this
determining whether “ anaphonic” effects are intentional or contin­ argument. “ Oddly enough,” he wrrites, “ the anagram research is one
gent.75 In 1909, however, the tools needed compute such statistical of the few projects Saussure pursued that did not fail,” in the sense
facts of language were lacking. that having tested his hypothesis and observed that it did not lead to
When, over half a century later, Starobinski first began to publish certain conclusions, he resolved to set it aside.80
portions of Saussure’s notebooks, they provoked an immediate criti­ Other views of Saussure’s investigations, however, have also been
cal response. The interest in the founder o f structural linguistics expressed. Upon their first publication, Roman Jakobson hailed the

127
126
 ANAPHONES
DARK TONGUES

anagram notebooks as the beginning of a second Saussurian revolu­ though it was neither general nor secret.” That “procedure” was
tion, after that of the Course, opening “ unseen perspectives for the linked, he explained, to the genre of the encomium and, more pre­
linguistic study of poetry.” “ If Saussure’s manuscripts of this massive cisely, to the poetic praise of a name. “ In order to establish properly
work had not been spurned for many decades as supposedly ‘ futile the glory of a name,” he wrote, the elements of a certain special
digressions,’” Jakobson later wrote, “ the international struggle for word can be “ mobilized in an exalted fashion... as if the reorder­
a science of poetics would have received beneficial incentives.”81 ing of the letters and the phonemes in a name better twisted the
That judgment was general in force; it bore on the conception of branches of the crown belonging to the subject of praise.” Within
poetry as involving a “ furor of phonic play” and patterning in sound the limits of this “genre,” he stated, “ the use of anagrams is regularly
structure. Yet it is worth noting that in the years that have passed present.”84 Without reasoning in terms of genre, Saussure, of course,
since the first appearance of Saussure’s research into “ anaphones,” had said no less.
details of his analyses have found more support among scholars than In this article, Starobinski cited only early modern texts: Latin
one might have first expected. In her important work on the poetic Renaissance inscriptions and Romance poetry, such as that of
structure of the Rig Veda, Tatyana Elizarenkova, for example, treats Petrarch and his followers, which extract and scatter the parts of
Saussure’s study of Sanskrit anagrams as a major contribution, which important names across the lines of many lyric forms, such as the
casts significant light on the ancient literary corpus. In Vedic hymns, song, sonnet, sestina, and ballad. Starobinski avoided confronting
she argues, the name of a lauded deity can be explicitly expressed, the question from which Saussure had begun. Could the archaic
yet it also “appears as a distinct pattern, embroidered on the can­ Roman poets not also have practiced such a “procedure” ? In the
vas, as it were, of the metrical scheme. It can also be indicated by modern period, authors of treatises of poetics explicitly refer to
sound-hints intertwined in the hymn’s texture. That this is a fully figures such as the anagram, William Camden’s “Anagrammatisme,
conscious effort, and not an accident, is suggested by those cases in or Metagrammatisme,, being but one example. Saussure had sought
which instances of surface sound-play share a common etymological indications of a similar kind among the authors who most interested
basis.”82 In different settings and for different reasons, several com­ him, but the ancient and even archaic sources were naturally more
parative linguists have similarly argued that Saussure’s notebooks, exiguous. He was the first to recognize, in any case, that he was
however problematic in parts, constitute a basic element in the study no scholar of the history of poetic theory. “ Without having made
of Indo-European poetic techniques.83 a special study of the writings of Latin prosodists,” he wrote in
Most remarkable is perhaps that some twenty years after publish­ one notebook, “ it would be difficult for me personally [to say] if
ing his influential study of Saussure’s “ anagrams,” Starobinski him­ there is some allusion in their writings to the necessity of hypo-
self suggested, albeit only implicitly, that his early appraisal may have grams. Since such an allusion has never been pointed out, one must
been inadequate. Without formally retracting the judgment he made suppose that the ancient theorists of Latin versification constantly
two decades earlier, Starobinski admitted in 1995 that his treat­ abstained from mentioning an elementary and primary condition of
ment had perhaps been peremptory. In an article bearing the tell­ this versification.” 85
ing subtitle “ Complement to a Reading of Ferdinand de Saussure’s Here, however, Saussure encountered a literary fact as impen­
Anagram Notebooks,” he stated that it would be an error to exclude etrable as the triumphal lines of verse inscribed on the tablets of the
from poetry the possibility of phonic play based on names. “ Today,” Roman Forum. “ Why they observed this silence,” he commented
he granted, one may “admit that there was once a real procedure, briefly, referring to the ancient specialists of poetry, “ is one problem

128 129
 DARK TONGUES

to which I have no solution.”86 That the system o f phonic doublings C H A P T E R TEN

and scattered polyphones had been the content o f “ an occult tra­

dition” was one axiom that Saussure did not question during the
Pat t er ns
entire period of his research into the techniques o f anaphones.87 Yet
how had the teaching passed from master to student for millennia?
Transmission was perhaps the final mystery that the linguist must
confront. “ How Naevius, Ennius, Pacuvius, Attius still preserved
a tradition that might have seemed inviolable in their imitative
epoch,” he wrrote, “ is something I can understand. But how could
Virgil, who, after all, possessed the breath o f original poetry, how
could Lucretius, with his intense preoccupation for the idea, how From 1909 till 1961, the existence of Saussure’s notebooks on the
could Horace, with his good common sense about everything, how sound system of Indo-European poetry remained almost entirely
could they all adopt this incredible relic o f another age?” To these secret. Saussure did not mention them in his publications, nor does
questions, Saussure offered no answer. “ There is something there, I he appear to have discussed them in the courses he gave at the Uni­
admit it, that escapes me altogether. I see no other thing to do than versity of Geneva. He undertook his research in private and, with
to present the riddle such as it presents itself.”88 Yet the riddle barely the notable exception of letters to certain friends and colleagues, he
“presented itself” at all, for Saussure alone perceived it. The art of appears to have kept his findings to himself. When, after his death,
anaphones was so secret that it almost effaced itself. What they wrere his former students arranged to turn his teaching into a book, to be
to guard, the reason they were to guard it, and the form by which called the Course in General Linguistics, they limited their editorial
the method had persisted—all had vanished. Saussure retained only activities to the materials pertaining to the lectures held between
the uncertain signs of a nameless linguistic technique, and the one 1906 and 1910. The notebooks went unnoticed. Half a century later,
hypothesis to which they led him during his four years of study: in i960, Robert Godel published a brief article, “ Inventory of the
that something, for some reason, had knowingly been scattered in Ferdinand de Saussure Manuscripts Held by the Public Library and
speech, almost inaudibly, almost invisibly, and that its hidden places University of Geneva” in which, after enumerating the unpublished
were in poetry. papers on general and historical linguistics, he designated a last set
of manuscripts, least, one might infer, in importance: “Ms. fr. 3963-
3969, ANAGRAM S (or HYPOGRAMS),” consisting, as he wrote,
of “ninety-nine notebooks” and a “ dossier of tables w ritten on large
sheets.” 1 Four years later, Jean Starobinski published an article on
these documents.2 He followed it in 1967 with a second essay bearing
what was to become, in 1971, the title of his book on “the anagrams
of Ferdinand de Saussure”: Words upon Words. The essay appeared in
a collective volume, To Honor Roman Jakobson.3Starobinski’s choice
of contribution was carefully weighed. When revealed, Saussure’s
inquiries into the law of poetic sounds met with almost unanimous

31
130
 DARK TONGUES PATTERNS

skepticism from linguists and literary scholars. Among major figures Jakobson singled out only one element in the theory of aria-
in the fields of language and literature, only Jakobson dissented. His phones that was in need of correction, being, as he argued, both
engagement with Saussure’s research on poetry was to be unique, questionable in itself and a methodologically superfluous supposi­
on account of both the aspects of the theory o f “ anaphones” that he tion. Saussure, however, had held it to be a fundamental point in his
retained and those that, by contrast, he set aside. investigations. It was the “ the hypothesis,” as Saussure had written,
Jakobson was hardly known as a commentator on Saussure. Yet that the Indo-European procedure of sound mirroring in poetry was
in 1971, he edited and published a document from the Genevan’s the object of “an occult tradition” that had remained, for the course
archive: a letter addressed to Antoine M eillet from November 12, of centuries, “a secret carefully concealed.”9Jakobson was quick to
1906, to which Jakobson gave the title “ Ferdinand de Saussure’s First point out that despite continuing to maintain this principle, Sau­
Letter on Anagrams.’’4 In the lengthy note that accompanied the ssure had alluded, albeit in passing, to the fact that one could also
text, Jakobson lamented the delay with which Saussure’s research set the postulate of secrecy aside, considering phonic structure in
had been brought to light. “ It is truly astonishing,” Jakobson wrote, itself. One could focus on the mere “materiality of the fact” of sonic
“that Saussure’s ninety-nine manuscript notebooks on ‘sounding symmetries, without advancing any theses about their subjective
poetics’ [poetique phonetisante], and in particular on ‘the principle causes.10 In a notebook on Homer, Saussure had even suggested—in
of the anagram,’ could have been hidden from readers for almost a terms Jakobson would cite repeatedly—that in certain conditions,
half century.”5 He recalled that as early as 1966, when Starobinski one might concede that anagrammatic sound play can mark poetic
published his first article on the “ anagram s,” Jakobson chose to language “whether the critic, on the one hand, and the versifier, on
express his admiration for Saussure’s hitherto neglected research the other, wishes it or not.” 11 This was the theoretical possibility that
into phonic poetics. “ The theory and analysis of the sound figures Jakobson deemed most worthy of exploration. Phonic mirroring, he
[jeuxphoniques],” Jakobson had written, “ particularly anagrams, and argued, need not be considered the object of a conscious will, on
their role in the diverse poetic traditions as elaborated by him the part of either those who compose poetry or those who transmit
simultaneously with his renowned courses o f general linguistics it. The intuition that poetry rests on a “ furor of phonic play” can
may certainly be counted among Saussure’s most daring and lucid be freed of any thesis about poetic intention. The secret, in other
discoveries.”6 Jakobson dismissed the idea that Saussure’s decision words, may be eliminated. The many varieties of anaphones can then
not to publish his research pointed to any particular dissatisfac­ be considered for what they truly are: patterns of language percep­
tion with its results. “ Must one recall,” he asked, “ what Saussure tible, to different degrees, in poetry.
himself says about his ‘gift’ of interrupting the publication of his By the time Jakobson advanced his reading of the theory of ana­
linguistic articles when they were ‘not only written, but also almost grams, he himself had devoted half a century to the study of sound
entirely composed’ ?” 7 Jakobson went so far as to suggest that in a structure in verse. In his youth, he had participated in the first
certain sense, it was in the study o f poetic textures, and not general years of the Linguistic Circle of Moscow, formed in the winter of
linguistics, that Saussure had made his most original discoveries, 1914-1915 to further, as its founding statement declared, the study
“ discoveries all the more striking,” he added, “ if one considers the of “ problems of linguistics, metrics, and folklore poetics.”12 Implicit
fact that on this path, Saussure had no signposts to follow, whereas in that research program was the principle that forms of discourse as
in the theses of his Course in General Linguistics, he was inspired by apparently diverse as literature and riddles exhibit certain common
several precursors.”8 characteristics, which are susceptible to a single linguistic analysis.

133
32
 DARK TONGUES PATTERNS

In W17. Jakobson became a founding member of the Society of the discourse in relation to other forms of language. He aimed, in short,
Study of Poetic language (known by the acronym o p o j a z ), which to situate the “ poetic function” with respect to “ the other functions
also counted among its members Victor Shklovsky, Boris Eikhen- of language.”
baum, and Yuri Tynyanov, as well Victor Mayakovsky, Osip Man­ To this end, Jakobson began by identifying “the constitutive fac­
delshtam, and Boris Pasternak.13 Jakobson would later recall how tors in any speech event.” He reduced them to six elements that may
his involvement in these two associations led to his work on popular be related in a simple form. An “addresser” (1) sends a “message” (2)
and literary verse forms. It also served to spur him on to what would about a “context” (or referent) (3) to an “addressee” (4), a message
become a lifelong search for a linguistic criterion for the distinction, that will be expressed in a certain “code” (5) and assured by a “con­
in any given tongue, between poetic and nonpoetic sequences. By tact” (6) (which Jakobson glosses as a “physical or psychological con­
1933, if not sooner, Jakobson had begun to envisage this search in nection” ).16 “ Each of these six factors,” he explained, “ determines
terms that derived directly from the research program of the Mos­ a different function of language.” All six may be present, albeit in
cow Circle. The crucial question, for him, was that of “ poeticalness” differing degrees, in a single utterance. The first five are most easily
or “poeticity,” defined as the function by which literary language defined. The “referential” function, which focuses on extralinguistic
is distinct from other types of speech. Unlike poetry, “ poeticity,” reality, depends primarily on “context.” The “emotive function,”
Jakobson argued, remains constant across cultures, traditions, and which involves the expression of affect, reflects above all the state
languages. It may be recognized by certain signs. “ Poeticity,” he of the “addresser.” The “conative” function draws attention to the
argued in a lecture given in Czech to an artistic society in Prague, “addressee.” The “ phatic function” involves the “contact” by which
“ is present when the word is felt as a word, and not as mere rep­ communication occurs, being displayed, for example, “ by a profuse
resentation of the object being named, or an outburst of emotion, exchange of ritualized formulas, by entire dialogues with the mere
when words and their composition, their meaning, their external purport of prolonging communication.”17 The “metalingual func­
and inner form acquire a weight and value of their own, instead of tion,” finally, focuses speech “on the code,” in science or everyday
referring indifferently to reality.” 14 discourse, for the sake of clarifying terms that would otherwise
That statement offers a key to Jakobson’s entire literary schol­ remain ambiguous or unknown.
arship. From his 1919 essay, “ The Newest Russian Poetry,” which The “poetic function,” as Jakobson defines it, approaches the
appeared as a preface to Velimir Klebnikov’s collected works, to his “metalingual,” in that it draws the attention of speakers toward
1921-1922 book on the comparative structure o f Czech and Russian language itself. The aims of metalanguage and poetry, however,
metrics, to his later studies of German, French, and English “ verse are fundamentally distinct. Taking the term “message” to signify
and its masters,” Jakobson sought to define the forms and rules not a particular semantic or propositional content but the utter­
by which, in the most diverse of languages, the “ w ord” comes “to ance as such, Jakobson asserts that “ focus on the message for its
be felt as a word and not as a mere representation o f the object.” own sake. . . is the POETIC function of language.”18 In content, if
His theory of the poetic function was perhaps never more fully not in form, this was a claim shared by several of Jakobson’s early
colleagues, Viktor Shklovsky not least. Jakobson’s original thesis is
and precisely presented than in the famous lecture “ Linguistics
diffferent, consisting in a proposition all the more remarkable for its
and Poetics,” which he gave at the “ Conference on Style” held at
formal simplicity: that there exists “an empirical linguistic criterion
Indiana University in the spring of 1958.15 Here he stated that his
of the poetic function.” 19
aim would be, as it had been decades earlier, to characterize literary

»35
134
 DARK TONGUES PATTERNS

To define this “ empirical linguistic crite rio n ,” Jakobson dis­ mnemonic series.” “ Combination,” instead, involves Saussure’s syn­
tinguishes “ two basic modes of arrangement used in verbal behav­ tagmatic type o f arrangement. It is “ based,” in Jakobson’s terms, on
ior, selection and combination.” Those “ tw o basic m odes” allude to not “ equivalence” but “ contiguity.” Certain features involving the
an opposition set forth in Saussure’s Course in General Linguistics: lexicon, for example, may decide which terms in a given language
that, namely, between “associative” and “ syntagm atic” relations.20 may be substituted for each other, while rules of syntax, among oth­
According to the Course, any linguistic expression may enter into ers, w ill determine which linguistic elements can follow or precede
these two formally distinct types o f arrangem ent. For example, each other in the articulation of grammatically correct phrases
the word “ relation” may be related to others by association: terms or sentence.
related to it by its root, such as “ relative,” “ relatively,” or “ relation­ Only if one grasps this distinction between “selection” and “com­
ship"; terms related to it by its form, such as “ anim ation,” “ station” bination” can one understand the “empirical linguistic criterion of
or “equation”; terms related to it by its m eaning, such as “ connec­ the poetic function” that Jakobson proposes as a major contribution
tion,” “ tie,” “ link,” or “ bond.” The possibilities o f association are to the linguistic study of poetry. He explains the criterion in the
many, if not infinite. “ Syntagmatic” relations are o f a different kind. following terms:
They consist in the relations that words bear to those that precede
The poetic Junction projects the principle of equivalencefrom the axis of
and follow them in fully articulated speech. M orphological and
selection into the axis of combination. Equivalence is promoted to the
syntactic rules bear on such relations to the degree that they define
constitutive device of the sequence. In poetry one syllable is equalized
the sequence in which words may be placed in correct sentences.
with any other syllable of the same sequence; word stress is assumed to
Saussure observed that syntagmatic and associative relations may be
equal word stress, as unstress equals unstress; prosodic long is matched
generally distinguished by their mode o f appearance in language.
with long, and short with short; word boundary equals word bound­
The first sequence is actualized in discourse; the second is only ever
ary, no boundary equals no boundary; syntactic pause equals syntactic
potential. “ The syntagmatic relation,” we read, “ is in praesentia; it
pause, no pause equals no pause. Syllables are converted into units of
rests on two terms that are equally present in an effective series. The
measure, and so are morae or stresses.23
associative relation, on the contrary, unifies tw o term s in absentia,
in a virtual mnemonic series.”21 When the principle of equivalence passes from the “axis” of
In defining the empirical linguistic criterion o f the poetic func­ selection to that o f combination, from the associative order to the
tion, Jakobson treats Saussure’s “ associative” relation as an arrange­ syntagmatic, and when such equivalence in the “effective series” of
ment of “selection,” which conditions and anticipates every par­ consecutive utterances dictates an entire sequence of speech, the
ticular choice of terms. If the “ topic” o f a certain message is to be a poetic function may be said to be operative. Hence the proximity
child, Jakobson explains, “ the speaker selects one among the extant, and distance, for Jakobson, of metalanguage and poetry. The “meta-
more or less similar, nouns like child, kid, youngster, tot, all of lingual function” focuses language on “ the code” in such a way as to
them equivalent in a certain respect, and then, to comment on this construct a single equation that clarifies the meaning of a particular

topic, he may select one of the sem antically cognate verbs—sleeps, term: “A=A,” Jakobson explains, giving, as a single example, “Mare is

dozes, nods, naps.”22 Selection is thus “ produced on the basis of the female of the horse.”24 By contrast, the “poetic function” concen­

equivalence, similarity and dissimilarity, synonym y and antonymy,” trates attention on the “ message for its own sake,” such that syntag­
matic equivalences come to form hierarchically stratified sequences
each of the terms that may be chosen being a m em ber of a “ virtual

137
136
 DARK TONGUES PATTERNS

of measured units. In a strophic poem, for example, syllables, at a Waugh, Jakobson favorably evoked this aspect of the Saussurian
first level, will be treated as equal in their durations or stresses. In theory of anaphones. The “constituents” of Latin verse are “exactly
their concatenations, they w ill compose substitutable metric feet. coupled, reiterated in even numbers.” Hence the “slogan” that Sau-
Such feet, grouped together in fixed patterns, w ill constitute lines ssure had transcribed in his notebooks in capital Latin letters and
of verse, which may be exchanged according to certain prosodic that Jakobson recalled: “ NUMERO DEUS PARI GAUDET,” “ God
rules. Lines, finally, in various commensurate relations, will form delights in even numbers.”29 With his theory of the “projection of
the strophes or stanzas that articulate the poem. the principle of equivalence from the axis of selection to the axis of
That equivalence is the fundamental formal principle of poetry combination,” Jakobson provided a linguistic foundation for such
was an idea Jakobson could find in several of the writers who mat­ phenomena. It is not so much God, he might argue, as the poetic
tered most to him. In “ The Rationale o f Verse,” Edgar Allan Poe function that dictates that in utterances considered to be poetic, the
made this one of his central points, declaring: “ Verse originates in many parts of speech tend toward equivalence, composing patterns
the human enjoyment of equality, fitness. To this enjoyment, also, that, however complex, may always be reduced, like the numbers of
all the moods of verse—rhythm, metre, stanza, rhyme, alliteration, classical arithmetic, to commensurate collections of equal units.30
the refrain, and other analogous effects—are to be referred.” In this In his reconstruction of the “occult tradition” of phonic mir­
sense, the idea of “equality,” Poe explained, “ embraces those of simi­ roring, Saussure had observed that anaphonic recurrences seem to
larity, proportion, identity, repetition, and adaptation or fitness.”25 be at odds with the consecutive character of linguistic relations. If
Perhaps most important, for Jakobson, was the recurrence of mea­ the phonological elements that make up a single noun or name can
sured linguistic units in the simplest form o f distinction, namely, appear out of their customary lexical order, their last letters, for
opposition. The patterns of such recurrence constitute, he wrote, example, preceding their first, while nonetheless evoking the word
the “much more general, we may even say the fundamental, prob­ that they all compose, one may deduce that signifiers can function
lem of poetry, namely, parallelism.”26 He repeatedly evoked Gerard outside “ juxtaposition in a sequence” and perhaps even, as Saussure
Manley Hopkins’s programmatic pronouncement, recorded in an had wondered, with some perplexity, “outside time.”31 To the theo­
entry of his Journals and Papers dating from 1865: “ The artificial part rist of the Course, who stipulated that the elements of language,
of poetry, perhaps we shall be right to say all artifice, reduces itself being linear, “ always follow each other in the speech sequence,”
to the principle of parallelism. The structure o f poetry is that of such a possibility was troubling. In itself, it sufficed to cast doubt on
continuous parallelism, ranging from the technical so-called Paral­ the theory of anagrams. Jakobson, for his part, drew the opposite
lelism of Hebrew poetry and the antiphons of Church music up to conclusion. He argued that the notebooks had intuited an aspect of
the intricacy of Greek or Italian or English verse.” 27 language that the lectures had dogmatically excluded. Against one
Jakobson grasped immediately that Saussure’s account of the Saussure and for another, Jakobson concluded that where equiva­
Saturnian offered an exemplary illustration o f such a “ continuous lence is “ promoted to the constitutive device of the sequence,” sym­
parallelism.” According to the Genevan linguist, in the archaic metry emerges beyond the apparent linearity of time.
Roman form, Latin phonemes always appear in pairs, a vowel, for As early as 1942, in the lectures that he gave at the Ecole Libre
example, “ having the right to appear in the Saturnian only if it has des Hautes Etudes in New York and that appeared as Six Lectures on
a countervowel in some place whatsoever in the line.” 28 In one of his Sound and Meaning, Jakobson had judged the theory of the linear­
last books, The Sound Shape o f Language, which he wrote with Linda ity of the signifier a “ serious obstacle” to progress in phonology.32

38 H9
 DARK TONGUES PATTERNS

Saussure’s Course famously declared that “ the linear character of order of speech also drew on his study of poetic language, in which
language [longue] excludes the possibility of pronouncing two ele­ he had observed formal correspondences between lines and metrical
ments at the same time.” 33 Jakobson objected, remarking that when positions separated in time. “ Parallelism at a distance” may be found
referred to signifying sounds, such a claim is either tautological or in many literary forms; it suffices to think of “compositions of the
incorrect. If one defines the “ element” in question as the minimal rondo type, which are based on a regular link between the end and
unit of sound articulated in an instant, the statement, by definition, the beginning of a piece.” 39 Poe himself explained, in “ The Philoso­
must be true; yet if one takes “ element” to refer to a single pho­ phy of Composition,” that he had begun “ The Raven” with the word
neme, it is misleading at best. The reason involves the nature of the that constitutes its terminal refrain, “ Nevermore.”40Jakobson added
phoneme as Jakobson conceived it. Linguistics before him had com­ that the words of the poem’s title and refrain could be related as
monly defined the phoneme as a minimal unit of signifying sound two lexical figures in retrogression, on condition of bracketing the
that exhibits a certain feature, such as being voiced or unvoiced or order of temporal succession. “ The never-ending stay of the grim
being occlusive or nonocclusive. Jakobson proposed a new concep­ guest is expressed by a chain of ingenious paranomasias, as we would
tion. He argued that the phoneme be defined not as a minimal unit, expect from such a deliberate experimenter in anticipatory, regres­
but as the representative of distinctive features more minimal than sive modus operand!, such a master in ‘writing backwards,’ as Edgar
the phoneme itself.34 He argued, in other words, that the qualities Allan Poe. In the introductory line of this concluding stanza, ‘raven,’
of being voiced, occlusive, and alveolar, for example, do not define contiguous to the bleak refrain word ‘never,’ appears once more as
the English phoneme / d/ in the way that certain predicates, in a an embodied mirror image of this ‘never’: /n.v.r./—/r.v.n./.”41
classical predicative definition, may be referred to a certain sub­ This reading could be viewed as an ingenious application of the
ject. Rather, the phoneme /d / is no more than an abbreviated sign theory of “ hypograms” to a literary corpus that Saussure did not take
of these three “ features” when clustered together. It is, in short, into account. The American poet would have scattered the phonic
strictly equivalent to them.35 The phoneme, in this sense, resembles elements of the noun—or name—“ Raven” in the bird’s own insis­
less an individual musical note than a single “ chord,” being “ unit and tent, exclamatory refrain: “ Nevermore!” Yet one might also reason
bundle at once.”36 otherwise, for “ Raven” could be read as a rewriting of “Never.” Jako­
Such an argument could be seen as announcing the “ nonlinear bson’s comments suggest both possibilities. They thus also allude to
phonology” that, starting at the end of the 1970s, would show that a third literary and linguistic interpretation. “ Raven” and “ Never”
speech patterns are not chains of individual elements but hierar­ may be grasped as two arrangements of a single sound sequence:
chically stratified systems that may be studied on several levels.37 “/n.v.r/” or “/r.v.n./.” Such an interpretation may seem improbably
Jakobson had, of course, drawn the elements o f his phonology from abstract. Yet it is worth recalling that Poe, for his part, alleged that
research in general linguistics that suggested to him that the “ linear­ the poem had arisen from even simpler phonemic matter. He recalls
ity” of speech is far from absolute. He cited the phenomenon known that once he had decided that his composition would be organized
to linguists as “regressive dissimilation,” by which, as if in a curious by a refrain consisting of a single word, the question “arose as to the
chronological distortion, “ in a sequence, a prior moment depends character of the word”:
on a later one”: “ for instance, when Spanish and English changed
Having made up my mind to a r e fr a in , the division of the poem into
the first / l/ of the world colonel into / r/ in anticipation of the final
stanzas was, of course, a corollary: the r e fr a in forming the close to
/l/.”38 Yet Jakobson’s critique o f the doctrine o f the consecutive

140 4i
 PATTERNS
DARK TONGUES

phonic elem ents that evoked particular nouns and names dissem i­
each stanza. That such a close, to have force, must be sonorous and
nated across lin e s o f verse. In addition to finding the “ Raven” in
susceptible of protracted emphasis, admitted no doubt: and these con­
“ N ev erm o re ,” he re c a lle d how V elim ir Khlebnikov, com m enting
siderations inevitably led me to the long o as the most sonorous vowel,
on his ow n poem years after com posing it, had been “ charmed by
in connection with r as the most producible consonant.42
the anagram ” he d iscovered hidden in a gerund. “ The word uskuj
“ Raven" and “ Never” would then be no m ore than the lexical prog­ (‘pirate ship,’ m eto n ym ically ‘p irate’ ) sits in the poem ‘as if in the
eny of the mournful, phonetically back diphthong, conjoined with Trojan h o rse’: K R Y L js K U J A ‘w in g in g ’ sK R Y L uSKUJA derevjdnnyj
the eminently “ producible” liquid consonant. Jakobson could inte­ kon ‘the w ooden horse concealed the pirate.’”43 In his “ Microscopy
grate this account, which he knew w ell, into his theory. Poe would o f the Last ‘ Spleen ’ in Les Fleurs du m a l” Jakobson argued that the
have begun his composition with a linguistic relation o f association: elements o f this p o e m ’s title —“ Spleen” —could be discerned in its
in this case, that of the vowel /o / (or /o u / ) . He w ould then have lines, as the scattered parts o f a concealed linguistic being that,
projected the equivalence it suggested from the “ axis o f selection” after Saussure, he referred to as a “ theme w ord” (mot-theme) . 44 In his
to the “axis of combination,” the m etric recurrences o f the vowel in paper “ Sh akespeare’s Verbal A r t,” he noted that “ in a few o f Shake­
the refrain word being the result. speare’s sonnets ( 13 4 - 3 6 ) his nam e W ill is inserted in a punning
The example illustrates the affinities and divergences between way and suggests the tentative question whether his signature is not
Saussure’s theory of anagrams and Jakobson’s idea o f the poetic func­ anagram m atized in 129 so that the poet’s rem ark—‘every word doth
tion. For Saussure, there exists a categorical distinction between alm ost tell m y n am e ’ (Sonnet 7 6 )—might be applied in its literal
forms of linguistic arrangement based on m etrical principles, such as m eaning to the poem under discussion.”45
quantity, alliteration, and rhyme, and the art o f “ phonic m irroring” Yet Jakob so n also argued that the texture o f poetic language
cultivated by the Indo-European poets. M etrical structures, for him, could be cru c ia lly d eterm in ed by sound structures that summon
are essentially of a piece with a language. They are passed on uncon­ no special w o rd s. E n tire section s o f the phonological systems o f
sciously from generation to generation, like the many phonological, languages cou ld be p ro je cte d , in their associative relations, from
morphological, and syntactic rules of individual gram m ars. The prin­ the axis o f sele ctio n to that o f com bination. They might involve
ciples of verse forms, in this sense, demand no special commentary. elem ents o f sound stru ctu re : vow els or consonants, for example,
Anaphonic procedures are of another nature, for they bespeak a rendered fu n ction al by th eir appearance or even by their absence.46
particular poetic technique that, to persist over tim e, must be con­ Sim ilarly, Jak o b so n aim ed in som e readings to show that larger
sciously transmitted. They have as their basic objects not grammatical gram m atical elem ents could in them selves become the material of
elements, such as syllables or accents, but rather certain nouns and poetic patterning. Parallelism s involving the disposition o f articles
names of particular importance. From Jakobson’s perspective, this in Yeats,47 m arkers o f gender in Baudelaire, “ differences in tense,
distinction is of no theoretical consequence. Poetic patterning can be aspect and m o o d ” in Sidn ey,48 all suggested to him the existence of
of any element of speech. What matters, above all, is that in each case a “ poetry o f g ra m m a r” w hose regularities were to be investigated
there be a projection of the principle o f equivalence, such that the by the new “ gram m ar o f p o e try ” that he proposed. One o f its fun­
poetic sequence have as its basic principle a system o f recurrences. damental principles w ould be that “ any noticeable reiteration of the
In his essays on poetry, Jakobson, for this reason, entertained same g ram m atical con cept becom es an effective poetic device.”49
many possibilities. At times, he detected m easured recurrences of Strictly speaking, any gram m atical unit whatsoever could become a

142 H3
 DARK TONGUES

unit of poetic patterning; it sufficed for it to becom e the element of Yet Jakobson also held that poetic sequences might be operative

svntagmatic recurrences. The linguistic study o f literature, Jakobson even where speaking beings themselves consciously perceive no

wagered, demonstrated as much: recurrences. He presented his argument most fully in an essay point­
edly titled “ Subliminal Verbal Patterning in Poetry,” marshalling
Any unbiased, attentive, exhaustive, total description of the selection, evidence in support o f the claim that a host of popular and literary
distribution and interrelation of diverse morphological classes and syn­
poetic forms can circulate unhindered by the nearly total lack of
tactic constructions in a given poem surprises the examiner himself by
awareness o f the equivalences that define their art:
unexpected, striking symmetries and asymmetries, balanced structures,
efficient accumulation of equivalent forms and salient contrasts, finally Phonology and grammar of oral poetry offer a system of complex and
by rigid restrictions in the repertory of morphological and syntactic con­ elaborate correspondences which come into being, take effect, and are
stituents used in the poem, eliminations which, on the other hand, per­ handed down through generations without anyone’s cognizance of the
mit us to follow the masterly interplay of the actualized constituents.50 rules governing this intricate network. The immediate and spontane­
ous grasp of effects without rational elicitation of the processes by
In treating sound play and anagram m atic play as o f the same
which they are produced is not confined to the oral tradition and its
nature, and in refusing to grant Saussure’s distinction between
transmitters. Intuition may act as the main, or not seldom, even sole
the level of merely metrical composition and that o f suprametri-
designer of the complicated phonological and grammatical structures
cal poetic “ technique,” Jakobson’s theory o f the poetic function
in the writings of individual poets. Such structures, particularly pow­
effectively dissolved one distinction that had been fundamental to
erful on the subliminal level, can function without any assistance of
the system of the ninety-nine notebooks: the distinction, that is,
logical judgment and patent knowledge both in the poet’s creative work
between unconscious and conscious literary procedures. Poetic pat­
and in its perception by the sensitive reader.53
terning, for Jakobson, remained a valid principle “ whether the critic,
on the one hand, and the versifier, on the other, wishes it or not.” Jakobson thus questioned one of the basic principles in the tra­
Defending his readings of Baudelaire against critics who claimed ditional study o f meter, which held that the obligatory features of
that the French poet could not have conceived as subtle a set of verse must be “ in some way audible.” 54 Formal structures, he argued,
grammatical parallelisms as he had uncovered, Jakobson recalled, may be present without needing to be heard.55 It would be an error,
for instance, Baudelaire’s own claim in his “ Poem o f Hashish”: however, to infer that Jakobson therefore entertained the possibil­
“ Grammar, barren grammar, itself becom e som ething akin to an ity that there might be no cause whatsoever of poetic regularities.
evocative sorcery. Words come back to life, dressed in flesh and “ Verbal patterning,” he maintained, excludes contingency. In every
blood: the noun, in its substantial m ajesty; the adjective, with the linguistic work, there is a “ deliberate veto of any game of chance.”56
transparent garment that clothes it and colors it like a glaze; and the For Jakobson, as for Saussure before him, statistical regularity was
verb, angel of movement, which causes the sentence to tremble.”51 therefore to be set aside. One might conclude that some subjective
Jakobson could also cite Poe, who likened the beauty o f metric form principle, for Jakobson, still motivates the poetic function. Yet the
to that of the crystal: “ the equality o f the sides pleases us; that of the important point in his theory is that this principle may also remain
angles doubles the pleasure. On bringing to view a second face in all indeterminate, in the sense that it does not matter whether pattern­
respects similar to the first, this pleasure seems to be squared; on ing is voluntary or involuntary. One is tempted, indeed, to go still
bringing to view a third it appears to be cubed, and so on.” ” further and to maintain that this much must remain indeterminate.

!44
 DARK TONGUES PATTERNS

Implicith\ Jakobson proceeded as if the neutralization of the differ­

were objects of plural verbs, and then one might subdivide each of these
ence between consciousness and its absence w ere the hallmark of
classes according to the tense of the verbs. This process of progressive
poetics. In striking contrast to Saussure’s notebooks, the “grammar
differentiation can produce an almost unlimited number of distribu­
of poetry” granted, in principle, the possibility that “ a semblance of
tional classes, and thus if one wishes to discover a pattern of symmetry
prosodic symmetries, sound repetitions, and a verbal substratum—les
in a text, one can always produce some class whose members will be
mots sous les mots (J. Starobinski’s felicitous expression)—transpire
appropriately arranged. 59
w ithout being supported by some speculative insight into the meth­
ods of the procedure involved.” 57 Such patterns would be “ ‘objectively’ present in the poem,” but “not
At the time he presented it, Jakobson’s theory o f the poetic func­ for that reason of any importance.”60 That objection, in its conclu­
tion elicited considerable scholarly interest, becoming perhaps the sion, draws close to the second question raised against the theory of
most prominent example of the linguistic method that structural­ the “grammar of poetry,” which was most famously formulated by
ist criticism seemed to promise. As such, it was once commented Michel Riffaterre. In a review of the essay that Jakobson and Levi-
upon in detail.58There was much to discuss. Jakobson’s essays, while Strauss dedicated to Baudelaire’s “ Les chats,” Riffaterre did not
ahvays circumscribed in their linguistic objects, ranged over a poetic contest the linguistic validity of their analysis so much as the intrin­
corpus seldom explored by a single critic. In addition to composing sic literary significance that grammatical features, for the authors,
studies of Russian, Czech, Polish, French, and English poetry, Jako- possessed. “ The method,” he observed,
bson worked with colleagues to offer new perspectives on literature
is based on the assumption that any structural system they are able to
in Italian, Gallego-Portuguese, Norwegian, Japanese, and Chinese.
define in the poem is necessarily a poetic structure. Can we not sup­
As one might expect, specialists of these fields contested details of
pose, on the contrary, that the poem may contain certain structures that
his readings. Yet literary scholars also raised certain basic questions
play no part in its function and effect as a literary work of art, and that
concerning the principles that underlay Jakobson’s conception of the
there may be no way for structural linguistics to distinguish between
poetic function.
these unmarked structures and those that are literarily active?61
These were fundamentally of two kinds. The first involved the
potentially unlimited set of parallelisms that might be pertinent to Jakobsonian poetics possesses the means to answer both objec­
“verbal art.” Commenting on Jakobson’s project to offer an “unbi­ tions. In response to the charge that the analysis of grammatical
ased, attentive, exhaustive, total description of the selection, distri­ symmetries must be arbitrary, Paul Kiparsky has remarked that from
bution and interrelation of diverse morphological classes and syn­ the linguistic point of view, it is mistaken to suppose “that gram­
tactic constructions in a given poem,” Jonathan Culler argued that matical categories can be multiplied ad libitum
it could lead to any poetic analysis whatsoever:
They are in fact a fixed set to be established within the theory of gram­
One can produce distributional categories almost a d lib it u m . One mar. To be sure, there are open questions about what the correct theory
might, for example, begin by studying the distribution of substantives of grammar is, including precisely how the set of categories should be
and distinguish between those which were objects of verbs and those delimited, but this much seems clear, that the pseudo-categories cited
which were subjects. Going one step further, one might distinguish by Culler, such as objects of plural verbs, have no linguistically justifi­
between those which were objects of singular verbs and those which able status whatever and therefore s h o u ld not play any role in poetic
patterning either. 62

147
146
 DARK TONGUES
PATTERNS

In the lengthy “ Retrospect” that appears in the third volume of sought “ to define and delimit the range” of poetic epit hets. They met
his Selected Writings, Poetry o f Grammar—Grammar o f Poetry, Jako- with an objection:
hson dwelt at some length on R iffaterre’s critique of his reading
The poet Majakovski rebuked us, saying that for him any adjective
of Baudelaire. He confronted the question o f the alleged need of
appearing in a poem was thereby a poetic epithet, even “great” in the
a criterion by which to distinguish between m erely grammatical
G reat B e a r or “ big” and “ little” in such names of Moscow streets as
recurrences and properly poetic features. R iffaterre argued that
B o l ’s a ja (“ big”) P r e s n ja and M a la ja (“little”) P resn ja. Cf. Majakovskij’s
such a distinction could depend only on the “ response” of individual
poem of 19 15 , “ I and Napoleon,” which begins with the words J a zivu
readers. Aware that an individual reaction to a text would neces­
n a B o l ’s o j P r e s n e , / 3 4 , 3 2 /: “I live on the Big Presnja, 3 4 , 4 2 . Apparently
sarily be partial, Riffaterre had sought to construct an imaginary
it’s not my business that somewhere in the stormy world people went
reader through the postulate of a “ special tool o f analysis.” Riffa­
and invented war.” And the poem ends: “The war has killed one more,
terre baptized this tool the “ superreader,” explaining that he had
the poet from the Big Presnja” (p o e t a s B o l ’soj P re sn i). Briefly, poetical­
constructed it by compiling a set o f individual responses that he
ness is not a supplementation of discourse with rhetorical adornment
took to be pertinent for the study of Baudelaire’s sonnet: statements
but a total reevaluation of the discourse and of all its components
made by the poet himself; remarks by two o f his contemporaries;
whatsoever. 65
suggestions proposed by several translators and critics; and impres­
sions gleaned, as the critic wrote, from “ students of mine and other This “ total reevaluation” does not admit the existence, in speech
souls whom fate has thrown my way.”63 It is not difficult to discern marked by the poetic function, of any opposition between the lit­
the arbitrary element in the construction o f such a “ superreader.” erary and the nonliterary. In the poem, not one, but every word
But it is equally clear that such a point of hermeneutic orientation, possesses the capacity to be “ felt as a word and not as mere rep­
however anonymous, introduces a subjective principle of interpreta­ resentation of the object being named,” for any element that can
tion of the kind that Jakobsonian poetics excludes. Yet in the theory be grammatically defined may enter into a system of recurrences.
of the poetic function, there was, in any case, neither a need nor Jakobson illustrated the poetic principle by means of an anecdote.
a place for the separation between “ linguistic and poetic actual­ “A missionary blamed his African flock for walking around with no
izations.”64 The reason is simple. W ithin the patterns of a poetic clothes on. ‘And what about yourself?’ they pointed to his visage, ‘are
sequence, Jakobson maintained, the distinction between poetic not you, too, somewhere naked?’ ‘Well, but that is my face.’ ‘Yet in
and nonpoetic cannot be made. “ If this organization of grammati­ us,’ retorted the natives, ‘everywhere is face.’”66Jakobson offered a
cal parallelisms and contrasts, a property characteristic of poetry, gloss, referring the remark to the ineluctable consequence of posit­
does not function as a poetic device,” Jakobson asked, respond­ ing the poetic function: “ In poetry, any element is converted into a
ing to the literary critics, “ for what purpose is this framework figure of poetic speech.”
introduced, scrupulously maintained, and remarkably diversified That statement clearly excludes the possibility that literary lan­
by poets?” guage be conceived as a domain in which only certain elements can
Before his work came to be discussed by historians and theorists play a poetic role. Yet it also casts an unexpected light on Jakob-
of literature, Jakobson had, in any case, encountered this question. son’s relation to Saussure’s hypothesis of a technique dimly discern­
In his presentation of his theory of the poetic function, he recounted ible in recurrent elements of sound. With respect to it, one might
read Jakobson’s dictum as imposing a law of disclosure. Where
how, in 1919, the members of the Moscow Linguistic Circle had

149
148
 DARK TONGUES PATTERNS

“everywhere is face,” where every “ component” —metrical or pho­ may be composed each time anew, in a different idiom and setting,
nic, stylistic or rhetorical—is as naked as the visage, nothing can be for the most diverse of reasons. The poetic function, by contrast,
hidden. There is no place in which to guard and transmit matters to is one, being but a consequence of grammar. In the symmetries o f
he kept “occult.” Jakobson’s opposition to Saussure’s hypothesis, to sound and sense that extend to every part of every tongue, it remains
this degree, is principled. Yet such a presentation is only partial, for discernible, if not discerned. It is the secret of no one that poems, in
Jakobson could also be said to set in place of Saussure’s art an activ­ sound or silence, knowledge or ignorance, lay bare.
ity that is far more obscure. Against the idea of a literary technique
developed and passed on in secrecy, Jakobson advanced his theory
of a function whose rules can, in principle, elude the speakers who
appropriate and make use of them, clear and distinct awareness
being no requirement for “subliminal verbal patterning in poetry.”
That suggestion amounts to the proposition of an intensely cryptic
thing. Jakobson’s poetic function would have been concealed not
only from the public but also from the poets, not recently, or even in
the distant past, but from the beginning. Poetic patterning, in short,
would have always emerged and persisted, almost without notice, at
once familiar and unknown to all.
The secret of this patterning extends across space as well as
across time. Saussure had investigated a procedure detectable in
certain literary traditions, being instituted and transmitted in cir­
cumstances that, however opaque in themselves, could be clearly
delimited. He thus restricted his attention to Indo-European antiq­
uities. Jakobson’s perspective, by contrast, is that of the general lin­
guist. His object is a possibility given with the very ability to speak.
Since the elements of poetry are nothing but the units of grammar,
set in recurring relations of equivalence, they are implicit in all lin­
guistic regularities. In this sense, too, poetry is no “ supplementation
to discourse”: nothing need be added to the grammatical system of
a language for verse to become possible in it. In every language, by
means of its structure, poetry remains, rather, potential. There is
no language in which, and of which, poetry’s varied forms cannot
be made. There is none in which patterns of equivalence may not
somewhere begin to proliferate, as if distinctly willed, for reasons
clear or dark, although their rules may be impenetrable to those
who compose, transmit, and perceive them. “ Poetry,” in this sense,

iso 1 5’1
 C H A P T E R E L E V E N

S e c r e t s o f T r i s t a n Tzara

On December 22, 19^9, the French newspaper Le Monde ran an

article bearing the striking, if tentative title: “ New Findings about
Francois Villon?” 1 A subtitle explained that enigmatic question in
more certain terms: “ Tristan Tzara Restores to the Poet of the Tes­
tament All the Lines Attributed to Jean Vaillant, Discovering 1600
Anagrams in Villon’s Work, Increasing ‘Master Francois’s’ Age by
Two Years.” The essay began by comparing Villon’s famously tumul­
tuous life to a dark forest that concealed within it not only trees but
also “ monks, archers, pickpockets, lawyers, poachers, ambushes, and
enchantments.” “ For three years now,” the article reported, “ Tristan
Tzara has been walking in this forest. He has left it dazzled, with
the feeling of having grasped hold of something rare: a treasure, or a
secret. In one month he will publish—or perhaps one should say, he
will detonate, since it will be more like a bomb—a book on Villon,
chock-full of explosive material.”2 The literary news quickly appeared
elsewhere in the French press. Only a day later, on December 23,
1959, the French weekly Les Lettresfrancaises published an article by
Charles Dobzynski whose title proclaimed: “ Five Centuries after the
Testament, Tristan Tzara Reveals The Secret of Francois Villon.” 3
“ In a few weeks, Fasquelle will publish a surprising book,” the
article related, “ a kind of challenge to the passing of time and the
mystery that envelops certain poetic works, that draws away the
depth of soul from attention to contemporary matters. The book is
signed Tristan Tzara and is called The Secret of Francois Villon”* Lest
Tristan Tzara, Villon’s Anagrams, TZR 698B (337) (courtesy of the Bibliotheque litteraire the readers mistake the identity of the author, Dobzynski recalled
Jacques Doucet, Paris, France).

153
 DARK, T O N G U E S S E C R E T S OF TR ISTAN TZARA

that the Tzara who would soon unveil the u n kn ow n secret of the revealed that a co m p lete ed ition o f Tzara’s work was underway arid
fifteenth-century poet was the same who had once founded Dada. that, “ at a later d a te ,” The Secret o f Villon would appear in print, “ so
that the re a d e r” co u ld at last “ form an exact and personal opinion”
Tristan Tzara is a man of surprises. Almost exactly half a century ago,
co n cern in g the p o e t ’s “ d e cid e d ly innovative thesis.” 10 The “ later
with Dada, he invented a terrible weapon, a kind of destructive ray,
date” turn ed out to b elo n g to a fu ture m ore distant than one might
that gave poets the right of life and death over language. Words were
have an ticip ated . O n ly in 19 9 1 did the sixth and final volume of
submitted to a strange fire that killed commonplaces and, at once,
Tzara’s Complete Works appear, disclosing the aged but still new “ New
obliterated meanings that had been on the map for centuries. For four
Findings” p ro m ised th re e decades earlier.
years now, in the quiet of his study, which is filled with mementos of
Tzara’s in te re st in the m edieval poet had long predated his last
Surrealism and Oceanic masks, all competing with the Phoenix of
book. As early as 19 4 9 , he graced P ierre Savin el’s edition of Villon’s
the unusual for perpetual rebirth, Tristan Tzara, like a scientist in his
poems w ith a p re fato ry essay in w hich he stated that modern poetry,
laboratory, has been preparing to reveal one of the most vertiginous
as a w hole, “ finds one o f the elem ents o f its functional mechanism in
literary surprises of the century. It will radically change our knowledge
the p o etry o f V illo n .” 11 “ Som e have wished to see in Baudelaire the
of the greatest poet of the end of the Middle Ages, F ra n c is Villon,
initiator o f m o d ern p o e try ,” he w ro te ,
whose Testament is, in fact, the birth certificate of modern poetry. This
time, therefore, Tristan Tzara has not confronted language in order to because o f his recognition o f the real world from which he draws his
destroy it, but to discover within it, under the heavy ash of the centu­ poetry. In his sincerity, he represents a reaction against Romanticism.
ries, the hidden sense, the fire that burned under the cold appearance In the same way, Villon is at the source of an equally modern current
of enigmatic allusions and cryptograms, encrusted in a text that has in poetry: the one that, in reacting against the Romantic and later
been studied a thousand times, interpreted, debated, in all its symbols. 5 conventional love o f the troubadours and against the unrealistic reli­
gious formalism of his time, announces the end of the Middle Ages. By
The Secret of Villon, however, was slow to be revealed to the pub­
this realistic position, and by moving from the elements of his life to
lic. The book did not appear in i960, as the tw o articles from 1959
reach a personal vision o f the world, Villon bestows a new criterion on
promised. At the time of Tzara’s death in D ecem ber 19 6 3, the work
poetic criticism . The authenticity of poetry, from this point on, will
had still not come to light. Christophe Tzara, the author’s son, would
be a quality residing in the variable and organic concordance between
later recall that his father had been “ correcting the typed manuscript
the apprehended fact and its expressed transposition. Poetry will be
in his last moments.”6 Yet it was not printed in the im m ediate after-
true if the sentiment that animates it has been intimately lived, not if it
math of his death. For further news o f Tzara’s final w ork, readers
results from some received formula. It will be necessary, in short, for
would have to wait till 1967, when Pierre Le G e n til, in a chapter
the poet to have lived it intensely enough for his poetic expression to
of a monograph on Villon, related what he had found in studying
be naturally adequate to it . 12
the manuscript of Tzara’s last book.7 M ore inform ation about The
Secret of Villon was disclosed eight years later in an article by Jean In this sense, T zara con cluded, V illo n ’s is not only a “ poetry of cir­
Dufournet, “ Tzara and Villon’s Anagram s.” 8 In the same journal in cum stance” (poesie de circonstance), marked by a particular historical
which Dufornet’s essay appeared, Henri Behar published a nine-page setting; it is, m ore fu n d am en tally, “ poetry of the circum stance”
fragment from Tzara’s book, titled “ The M eaning o f Anagram s.”9 (poesie de la circonstance), w hich draws from the world the materials
He also contributed to the issue a critical essay o f his ow n. Here he o f an im age u n iqu ely faith fu l to the poet’s encounter with it.13

154 155
 DARK T O N G U E S
SECRETS OF TRISTAN TZARA

In his preface, “ The M odernity o f V illo n ,” Tzara m ade no men­

tion o f anything that the m edieval poet m igh t have sought to hide By now someone else takes up the slack
from his readership. The V illo n o f 19 4 9 , it se e m s, possessed no Who’s filled to the brim on the gantry
secret, or still kept it to him self. T h e re velatio n o f The Secret o f Vil­ For the dance starts in the belly.16
lon, when made, would consist b efore all else in the disclosure of
C itin g these lines, Tzara draws the reader’s attention to the
this one tact: that there existed a “ secret o f V illo n ” that had lain con­
final cou plet, w hich follow s the poet’s allusion to the “ someone
cealed for centuries. As the French press re v e a le d , the cryp tic thing
else” w ho, having taken “ up the slack,” is “ filled to the brim on the
in question consisted in an art o f w ritin g , w h ich allo w e d V illon both
gantry.” “ Qui est ramply sur le chantier,” he notes, “ is an expression
to record and to conceal m atters re la tin g to the circu m stan ces of
o f the period, m eaning ‘ W ho has eaten well and drunk well,’ the
his poetry: allusions, suggestions, accu sations, and statem ents about
chantier or gantry, the piece o f wood supporting the barrel, being,
contemporaries, signaled above all by the p re se n ce o f proper names by extension, the cellar. The line Car la dance vient de la pance [For
hidden in a w eb o f anagrams. the dance starts in the belly] paraphrases a known expression. Here
Tzara began his book by re callin g that V illo n ’s Lais presents itself dance has the erotic m eaning implied by the context.” 17 That com­
as a “ romance” or “ novel” (roman) o f love that re te lls a m ajor episode m entary is supported by the scholarship on Villon’s language, which
in the life of the poet and, m ore p recisely, the reason s that “ forced Tzara kn ew w e ll.18 Q uickly, however, Tzara passes from the poet’s
Villon to flee Paris.” “ W hile relatin g the co n tin u atio n o f his adven­ statem ent to a level o f the text that lies beneath that of the words
tures, always in an ironic fashion, the Testament com pletes this first that com pose it. He evokes a reading of the passage proposed by the
work and retrospectively explains the o rig in s o f his m isfortunes.” 14 early tw entieth-century philologist and editor of Villon, Lucien Fou-
Villon loved a woman and w ould have con tin u ed to do so, had he not le t.19 The penultim ate line o f the stanza, Foulet suggested, contains
suffered the combined blows o f fate and p o ve rty. T h e twenty-fifth four groups o f sounds that, when combined, spell out the name of a
stanza of the Testament could not be m ore e x p lic it: contem porary o f V illon ’s evoked in both the Lais and the Testament.
It suffices to read “ Q ui est R A M plY sur les CHAN-TIERS” for the
Bien est verte que j’ay aime
figure to appear, concealed in an extended anagram: “ Itiers March-
Et amereoie voulentiers;
ant.” 20 Foulet had drawn a simple conclusion from this unexpected
Mais triste cuer, ventre affame
fact: Itiers M archant must be the rival “ someone” whom Villon
Qui n’est rassasie au tiers
evokes, w ithout explicitly naming, one line above.
M’oste des amouereux sentiers.
O nce noted in the line, the syllables of the name are difficult to
Au fort, quelqu’ung s 'en recompence, deny. B eing audible, their presence, however dislocated, is almost
Qui est ramply sur les chantiers! apparent. A note contained in Tzara’s unpublished papers suggests
Car la dance vient de la pance.15 that w hen he first encountered Foulet’s remark, he was immediately
persuaded o f its validity. Yet he also sought to take a further step. He
It’s very true I have loved
wondered w hether there might not be other names hidden in such a
And willingly would love again
form in the syllables o f Villon’s lines.
But a heavy heart and starved craw
Never full by more than a third The anagram of Ythiers Marchant is perhaps not an isolated case. The
Drag me down from love’s ways analogy of certain lines of the Lais with the stanzas of the Testament, the

157
 SECRETS OF TRISTAN TZARA
DARK TONGUES

(Le vers, qui de plusieurs vocables refait un mot total, neuf, etranger a
nomenclature of the legatees, arranged according to category or order
la langue).25 Finally, within this new “ word,” a center may be dis­
of importance, the correspondence of the nature of the inheritance left
cerned: an “ axis,” “ constituted by one or two alphabetic signs,”
to his tutor and his mother, are in no way gratuitous. It is in this direc­
with respect to which the letters of the hidden word or words are
tion, it seems to me, that one should pursue investigations.21
distributed in strictly symmetrical positions to the left and to the
By the time he composed The Secret o f Villon, however, Tzara admit­ right. For example, if a letter belonging to a secret expression can
ted that this first intuition had been mistaken. “ Persuaded that there be discerned two positions to the left of the center, another will be
must be other revelations in the work o f V illo n ,” he recounts, “ I legible two positions to its right; if, by contrast, no such component
looked for anagrams formed as indicated above—without the shadow letter may be observed one position to the left of the center, none
of success.”22 Pierre Le Gentil later confirm ed his conclusion. “Ana­ may be considered to lie one position to its right. Tzara adds that
grams analogous to the one cited by Foulet,” he asserted, albeit not when the hidden word is composed of an even number of letters,
unequivocally, “apparently exist nowhere else in lines by Villon.”23 “ the center is constituted by one or two letters; in the latter case,
Yet Tzara tells us that in time, his failure led to one major linguistic the letters may or may not be used in the formation of the anagram.”
discovery. He found that in the line commented upon by Foulet, the When the concealed expression, instead, possesses an odd number
name “ Itiers Marchant” could also be represented as present in a of letters, “ the center w ill be represented by a single letter, which
form different from the one suggested by the philologist, according must inevitably count among those that are necessary for the forma­
to anagrammatic rules more subtle, more exact, and more systematic tion of the anagram.” 26
than had ever been observed. For the poetic procedure to function fully, certain licenses are
Tzara explains in his book that these rules, when conjoined, required. “ While the numeric disposition of these anagrams is strictly
define a “procedure” that Villon adopted in the com position of a observed,” Tzara explains, “ Villon takes liberties, which concern
number of his poems, most notably the Lais. “ The procedure con­ only the spelling o f words.”27 Some reflect variations in orthography
sisted in including, in a line or in a portion o f a line involving the that are commonly to be found in medieval French manuscripts, not
anagram, a word or several words whose letters are systematically least those that transmit Villon’s poems. For example, pauvre may be
distributed with respect to a center, constituted by one or two spelled as povre, vieil as viel, vengeance as venjance.w Yet Tzara insists
alphabetic signs, the spaces between w ords counting as null.”24 that one also take into account other types of divergences in spelling.
Three points implicit in this summary are w orth retaining. The first Scholarship on Villon’s language, as it is recorded in manuscripts,
is that the anagrams detected by Tzara, unlike the cryptic figures suggested to him that the poet had taken as equivalent certain writ­
that traditionally go by that name, cannot be restricted to the space ten signs that in Middle French were otherwise held to be distinct.
of individual words, being in each case distributed across a single Louis Thuasne noted, for instance, that in some cases, Villon adds to
line of verse or one of its segments. The second point is that within words a terminal s (or withholds it), without effecting any change in
such a textual extension, the white spaces between words can play their meaning.29 But Tzara argues that one must also admit further
no role. To recover the names hidden in V illon's lines, one must licenses in writing, many of which were, and remain today, generally
begin by counting only letters. Implicitly, Tzara reasons in this sense unknown to specialists of Middle French scribal habits.
as did Mallarme: for both, albeit in different w ays, “ from many Tzara explains that, to grasp the words secretly scattered across
expressions, the line makes a total, new w ord, foreign to language” Villon’s lines, one must always keep in mind a complex system of

158 159
 d a r k t o n g u e s

orthographic imperatives: suppress a t at the end of a word ending

in -an or -«i, replacing it, if need be, by an 5; erase any * before *
inside a word; remove 1 before x; replace 1 by z or x or by c before 1

e, or a; take i, y, and j to be equivalent (or replace a y with two i ’ss);

exchange j and g; rewrite any double consonants as a single conso­
nant or rewrite, instead, some single consonants by two consonants;
treat any metrically elided or “ mute” e as either possibly included in
or excluded from the anagram; consider 0 to be equivalent to a; take
e as interchangeable with /; write ou as oe, or oue; replace c, when
necessary, with g ; and take m to be equivalent with n and n as sub­
stitutable for m. 30 Such rules enable the reader to find certain words
that, for the purposes of anagrammatic analysis, can be represented
in a number of different shapes. The priest Serm oise, whom Vil­
lon is reported to have killed on June $, 1455, for instance, appears
in Tzara’s readings in no fewer than tw elve distinct forms: “ Sar-
moies,” “ Sarmoie,” “ Sarmoye,” “ Sermoye,” “ Serm oie,” “ Sermoies,”
“ Cermoie,” “ Cermoye,” “ Cerm oies,” “ C arm oie,” “ Carmoie,” and
“Carmoies.” Yet Tzara also admits that it is possible that his ortho­
graphic principles are imprecise, in the sense that “ more general
laws” concerning the proper spelling o f V illon’s manuscripts may
still be found. “ I do not claim ,” he observes, “ to have exhausted
the question.” 31
The Secret of Villon suggests a graphic method for the illustration
of this procedure. It consists of tying individual letters on either side
of the central axis to each other by means o f “ curves” drawn below
the printed text . 32 The publishers o f Tzara’s final work include no
images of this method of representation, of which Tzara offers an
abbreviated description in his book. Yet one can find an example
of its form in an article published in 1960 in which Louis Aragon
announced that Tzara will soon unveil unseen anagrams in Villons
successor, Rabelais. 33
The publishers of The Secret o f Villon chose to present the ana
grammatic procedure by more traditional typographic means, which
are easily described. First, all the letters in a single line are set beside
each other, without spaces, as if they formed an entire word. Then Tristan Tzara,
Villon’s Anagrams, TZR 718 (courtesy of the Biblioth£que litteraire Jacques
Doucet, Paris, Prance).

160
 DARK TO N G U E S SECRETS OF TRISTAN TZAR

on two successive lines below, two more sets of signs are printed. phrases, being projected across a single metric space arid writ
The first will indicate, by means of plus and zero signs, which let­ ten, so to speak, upon each other. Sometimes entire sentences
ters plav a role in the hidden word. On the second line beneath the may be recovered in the space of a line. In the conclusion to the
text, the functional letters of the concealed expression will then be book, for instance, Tzara gives a thorough analysis of Villon’s
numbered according to the order of their appearance in the word or “ Ballade a s’am ie,” ending with an intense interpretation of its
words that they evoke. final address:
The first example printed in this form in T h e S e c r e t o f Villon

is the line from the T e s ta m e n t analyzed by Foulet, which appears Prince amoureux des amans le greigneur
Vostre mal gre ne vouldroye encourir,
as follows:34
Mais tout franc cuer doit, par Nostre Seigneur,
Sans empirer, ung povre secourir.35

Amorous prince and greatest lover

I don’t wish to call down your disfavor
Here Tzara recovers the same anagram as did Foulet—Itiers But every true heart must by the heavenly Father
M a rch a n t— without, however, admitting Foulet’s implicit principle Save a poor man before he sinks under.36
that the concealed name be present in the line through the rear­
rangement of its constituent syllables. Tzara reckons, in short, not Applying the rules of the procedure to the first line of this envoi,
in “anaphones,” as did Saussure in his hidden notebooks on poetry one discovers the hidden name of the “prince”: Charles d’Orleans,
but in gra m m a ta in the strict sense, letters distributed across each Villon’s friend and fellow poet. Yet one finds find six other names,
line. For him, the words concealed in Villon’s verse are composed which Villon has “ ironically” written over his: “ Noe Jolis,” “ Pardry-
not from sound clusters but from graphic signs, which function as ers,” “ Denise,” “ Itiers,” “ Perinnet,” and “ Sarmoie.”37 In his analysis
the indivisible elements of words. It follows that the hidden term of the penultimate line of the same envoi, Tzara goes further, finding
ceases to be perceptible to the ear. The play o f purely alphabetic seven names, among them that of the knight Jehan le Cornu, that all
units is too subtle, and too complex, to be heard as such. A syllabic point to the actors in the “ drama” Villon cannot forget:38
decomposition of one of Villon’s lines will issue in only eight or ten
elements, depending on the verse form employed by the medieval
poet. Admittedly, one may fashion more units by extracting the
occasional phonemic part from a syllable, as Foulet does in draw­
ing a Y from p ly in r a m p lY . Yet as long as one counts in syllables,
the set of functional units w ill remain small in number. If one
reckons in letters, at least twice as many elements will appear.
The possibilities of combination and recombination, accordingly,
will increase.
Tzara knew this well. More than once in T h e S e c re t o f Villon ,

he argues that several names can be found in the same words and

163
162
 SECRETS OF TRISTAN TZARA
T O N G U E S
DARK

Person ified , th is fin a l an agram w rite s o f itself, com m enting, in

testam en tary fo rm , on the reasons for its persistence. It is the
proof, Tzara w rite s, that V illon conceived the entire Lais as a con­
tinuous roman a clef. “ O n e could even say a double c l e he adds,
“ i f one k eep s in m in d that the key that serves to open the real
dom ain o f the d ram a and its characters is a function of the sec­
ond key, w h ich allo w s one to decipher the anagrams. Villon’s rich
and varied im ag in atio n is thus contained in the corset of a system
o f in te rp retatio n w h o se rig o r and lib erty reciprocally command
each other.” 39
L earn in g o f this late p ro ject, a reader fam iliar with the works of
Tzara’s youth m ight th in k o f that other polyphonic poetic form, the
poeme sim ultan, w h ich T zara, together w ith Richard Huelsenbeck
and M arcel Ja n c o , fam ou sly perform ed in Novem ber 1916 before
the public o f the C ab are t V oltaire in Z urich. Read “ in parallel”
in Fren ch, G e rm a n , and En glish , not successively, but in a single
stretch o f tim e , the p oem “ L’am iral cherche une maison a louer”
consisted in the stru ctu ral layering o f three distinct texts upon one
another.40 In the e xp lan ato ry “ N ote to the Bourgeois” that accom­
panied the o rig in a l pub lication o f this w ork, Tzara stated that it
responded, in its com position, to a particular pictorial practice. The
“ transm utation o f objects and colors” in “ Picasso, Braque, Picabia,
D uch am p-V illon , D ela u n a y ” had prom pted him to seek to “ apply
the same sim ultaneous principles in poetry.” The resulting creation,
Having revealed all the members of this company, Tzara explains as he w ro te , w as con ceived in such a fashion as to permit any mem­
that Villon also encrypted in the line a “ signature,” “ validating the ber o f the public to lin k the disparate associations o f the poem to
authenticity of his anagrams.” It consists of a sentence, its letters the elem ents ch aracteristic o f his own personality, m ixing them,
symmetrically placed as if in a single name: fragm enting them , etc., w hile rem aining in the direction that the
author ch an n eled .”41 D espite its apparent cacophony, the poeme, in
this sense, w as to be conducted, albeit in such a way as to suggest
different executions to its diverse listeners. “A contrapuntal recita­
tive, in H ugo B a ll’s w ords, it was to constitute a performance in
which three or m ore voices sim ultaneously speak, sing, whistle, or
do som ething o f the kind, such that their encounters constitute the
elegiac, hum orous, or bizarre shape o f the matter.”42

I64 16 5
 SECRETS OF T R I S T A N T Z A R A
DARK TO N G U E S

The simultaneity of text and anagram in The Secret o f Villon is “seems to have had any knowledge of the treasure hidden inside the
of another nature. In the Dadaist poem, three tongues coincide in verbal matter of his verse,” and why it is that to recover the buried
a temporal extension, sharing a single acoustical form: each is read gold, systematic orthographic adjustments must be made.48
aloud. Every language is therefore equally perceptible, or impercep­ At other times, however, Tzara suggests that the circle of those
tible. Tzara presents Villon’s poetry, by contrast, as animated by two who knew of Villon’s procedure was considerably wider than this
formally distinct levels of enunciation, “ superposed,” like images, summary would allow. Perhaps, he muses, Villon developed his
upon each other.43 While the poeme simultan, moreover, prompts technique with a view to the many friends who would be capable of
those who hear it to “mix, fragment, etc.” its elements in possibly detecting the recollections, thoughts, and reproaches encrypted in
incompatible ways, the two levels of the medieval work, for the late his verse. Tzara wonders even if Villon’s misfortunes may not be, at
Tzara, necessarily compose an ordered and coherent whole. Villon’s least in part, attributable to the reactions of contemporaries, who
poetry “contains” its anagrams in every sense. Not only are the signs could, and did, perceive the provocations that he had only faintly
of the secret words literally present in those that compose its lines; covered over in his poetry.49
what is concealed in them also reinforces, as if in a secondary role, Yet in the pages of Tzara’s final work, one also encounters the
what the text has openly said. “ It w ill be observed,” Tzara writes, possibility that the poet’s concealed communications may, in truth,
“that the meaning of the anagrams is often inscribed in the line that be “secrets” only for us. Tzara sought to show, in varying degrees
contains it, although neighboring lines or parts o f them may lend of detail, that poets from the troubadours to Charles d’Orleans
them a signification by im plication.”44 Anagram s do not so much and Rabelais all employed the “procedure.” At the limit, Villon’s
trouble the verse in which they can be discerned as offer them com­ cryptographic method may have been transparent not to some but
mentary: explanatory and revelatory notes for their interpretation. rather to all of his contemporaries. The content of the messages,
Tzara repeatedly calls them “ signatures,” which point to the reality then, would have been common knowledge. “ If the anagrams give us
from which the poems spring.45 Over the syllables of Villon’s songs revelations,” Tzara writes, “ we must nonetheless believe that in Vil­
and ballads, they sound softly, as “ a kind o f echo” or “ a voice in lon’s time, they were only ever ‘open secrets’ [secrets de Polichinelle].
mute” (une voix en sourdine) . * 6 He wrote the anagrams much more for the amusement of his friends
In several respects, the theory o f this dark speech remains itself than to unveil to them facts that they knew.”so
obscure. A first ambiguity involves the relation that is to obtain, for The equivocation is of great consequence for Tzara’s claims.
Tzara, between the medieval poet and his hermetic linguistic and Depending on the status of the “secret” revealed in the pages of his
literary procedure. The Secret o f Villon suggests several possibilities. book, the procedure may, or may not, be employed as an historical
At times, Tzara argues that the poet’s art o f embedding words within and philological tool. Repeatedly, Tzara suggests that the presence
words was never widely known. He reasons then that it allowed the or absence of anagrams in verse of Villon’s epoch will be a positive

poet to escape censure when inculpating his contemporaries, or criterion for the determination of authorship. One of the more pro­

when disclosing matters of an erotic nature, which medieval mores vocative theses of his work, as the French press immediately grasped,
was that all the poems attributed to the lesser fifteenth-century poet
required him to keep concealed. The “ secret,” he suggests, was
Jean Vaillant were, in fact, Villon’s work. Part of Tzara’s demonstra­
shared only with Villon’s closest friends, Catherine and Noel Jobs,
tion consisted in his proof that Villon’s procedure could be detected
whom he “ initiated” into the principle o f his art.47 This explains why
in the poems of Vaillant: the names “ Francois” and “Catherine,” in
“not one” of the medieval copyists who transcribed Villon’s poems

167
166
 SECRETS OF T R I S T A N TZARA
DARK TO N G U E S

“ Tabary” and “ Perrinet,” for example, since he knows the exact signs
particular, as he showed, are often scattered there. Yet if Villons
by which these names are written. Yet the principles of his method
technique was a common one, the conclusion hardly holds, since
dictate that a name possess no single form. Just as “ Sermoie” can
am one of his contemporaries could have woven those names into
be written as “ Sermoye” or “ £armoie,” so “ Tabary” can be spelled
his own lines. A similar ambiguity marks the substantial biographi­
“ Tabarye,” “ Tabarrie,” or Tabaries,” and Perrinet, as Tzara’s own
cal assertions that Tzara sought to dem onstrate by means of his
analyses demonstrate, may appear as “ Parynet,” “ Perenet,” “ Par-
anagrammatic decipherments. At some tim es, it seems that words
rinet,” “ Perynet,” or “ Parrenet.” An intractable difficulty follows,
found concealed in verse may be taken as biographical evidence of
for a simple reason: the very “ liberties” that allow Tzara to recover
an almost documentary kind. The most striking example, in this
regard, involves Tzara’s treatment o f V illon’s alleged date of birth. certain names in Villon’s work are those that forbid him from rec­
On the basis of his recovery o f hidden names, Tzara maintained ognizing them with certainty. Inevitably, the crucial element in Vil­
that Villon must have been born tw o years earlier than had been lon’s hermetic art of language—the secret of the secret—cannot but
thought.51 At other times, Tzara defends the pertinence of the ana­ escape him.
grams he has uncovered by anchoring them firm ly in historical facts In one of the earliest articles on Tzara’s last project, Aragon
that he takes to be certain. In certain passages o f his book, finally, wagered in i960 that The Secret of Villon would, at a future date,

Tzara suggests both possibilities simultaneously, as if unaware of provoke critical perplexity. “ It will one day be a great subject of
the vicious circle that such a reasoning would imply. “ One can only astonishment and study how it is that he who, almost half a century

establish with certainty a point in the biography o f Villon,” Tzara ago, founded the Dada Movement, has today become the researcher
who, in Villon and Rabelais, takes pains to show that the obscurity
then writes, “ if the indications derived from the anagrams confirm
of texts is essentially a result of our ignorance of both the social
the facts, and reciprocally.” 52 Yet if the first proposition and its recip­
conditions and the biography of writers.” 53 The exactitude of Ara­
rocal are both true, neither, strictly speaking, holds.
gon’s prophecy is undeniable. Yet it also raises several questions. It is
Perhaps the most perplexing am biguity in the argument of The
indeed remarkable that the author w ho once proclaimed that “there
Secret of Villon involves the status o f those elements most often
is a great destructive, negative task to accomplish” and who called,
treated as secrets: names. It cannot be denied that Tzara’s method
in his Manfesto of 1918, for the radical ‘Weeping away” of atrophied
of decipherment supposes that one know, before confronting Vil­
forms of artistic expression, should have later devoted such atten­
lon’s poetry, which special words it may conceal. Unknown terms,
tion to the recovery of an ancient art of w'riting for which he must
as Tzara himself admits, cannot be recovered by this method. One
devote himself, over the course of almost a decade, to mastering the
must draw the name for which one searches, therefore, from Villon’s
philology of medieval French.54
own poetry, concealed names being all, in this sense, recurrences
Beyond that formal observation, however, other differences
of apparent ones. Yet there is more. Even if one grants that the pro­
are worth noting. Beginning with his earliest work, Tzara defined
cedure consists in hiding what is known, so that it may, in time, be
“spontaneity” as the supreme feature of future art.55 In the “Lecture
found anew, a troubling fact remains. It pertains to the identity of on Dada” that he gave in Weimar and Jena in 1922, he declared:
the secret names. On the surface, Tzara would seem to accept a com­ “ We have had enough of reflective movements that have dilated,
mon premise in research into anagrams, which holds that only forms beyond all measure, our credulity in the benefits of science. What
with definite alphabetic shapes may be hidden and revealed in other we now want is spontaneity__ In art, Dada brings everything back
expressions. Tzara can recover the distorted shapes o f such names as
169
168
 SECRETS OF T R I S T A N TZARA
DARK TO N G U E S

to an initial, vet relative simplicity .”56Again, the contrast with his process of composition. The ultimate sequence of words in the
later conception of Villon’s metric lines o f words upon words seems “ Dadaist poem” w ill be determined not only by the initial choice of
sharp. “ They are the fruit o f a m eticulous assembly of disparate an article, with its length and lexicon, but also by the “soft shaking"
pieces, incessantly polished and polished anew so as to be able to in a bag of its words, followed by the extraction of each cutout, “one
follow each other,” The Secret o f Villon explains.” But it would be an after the other,” in an order that the poet cannot anticipate. The
error to infer that the late Tzara abandoned the literary quality he experiment has been compared to practices of artists close to Tzara,
once extolled. “ Granting that the term spontaneity has no meaning not least the ones Hans Arp named in his “ law of chance” (Gesetz des
when referred to poetry,” Tzara writes in his last book, “one must Zufalls).61 As Hans Richter long ago remarked, of all the Dadaists,
admit that Villon has succeeded in keeping intact the freshness of it was, however, Tzara who carried the principle of chance “to the
tone and the accents of his presence, despite the anagrams that lead most extreme point in literature.”62
their independent and swarming life on the inside of the lines that One might argue that Villon’s procedure, as the late Tzara
serve as their basis.”58 describes it, recalls to a certain degree the poetic method proposed
The most profound tension between the projects of the early and in the Manifesto of 1920. A “meticulous assembly of disparate pieces,
the late Tzara, however, lies elsewhere. It involves a force that the incessantly polished and polished anew so as to be able to follow
Dadaists, as perhaps none before them, summoned in the fabric of each other,” Villon’s verse, which resembles him like no other, arises
their art: chance, or contingency.59 In the “ Dada Manifesto on Weak from the combination of minimal units, abstracted and “cut out,”
Love and Bitter Love,” first read in Paris in 1920, Tzara offered this so to speak, from chains of signifying sequences. They are more
famous account of how “ To Make a Dadaist Poem”: minimal even than the elements that the Manifesto recommended
as building blocks, being neither words nor even metric parts—syl­
Take a newspaper.
lables, rhymes, or vowels—but mere phonemes, transcribed, for the
Take a pair of scissors.
most part, in accordance with the graphic conventions of fifteenth-
From the newspaper, choose an article of the length you would like
century France.
your poem to have.
There, however, the similarities between the two literary prac­
Cut out the article.
tices end. The Tzara of The Secret of Villon insists that the sequence
Then cut out each of the words that compose the article and put them
of names hidden in Villon’s verse could be no mere fruit of chance.
in a bag.
It must be willed. Like Saussure before him, Tzara could not, how­
Shake it softly.
ever, entirely exclude the possibility that certain stochastic processes
Then remove each cut-out, one after the other.
might be at play in speech. The “ keys” he had detected might par­
Carefully copy them out
take somehow in random realizations of some possible combinations
In the order in which they left the bag.
among others. “ One cannot deny,” he conceded in a fragment relat­
The poem will resemble you.
ing to his final work, “ that chance plays a relatively important part
And thus you will be an infinitely original writer of charming sensibil­ in the formation of anagrams made of discontinuous letters. The
ity, although still unknown to the common crowd.60 question that arises is that of knowing whether Villon and the other
These lines evoke a poetic practice in which the methodic anagrammatists, conscious of this chance, employed, amplified, or
destruction of a given text goes hand in hand with a partially random provoked it in intending to insert into their texts the greatest number

171
170
 DARK T O N G U E S

of anagrams and, furthermore, to place them in the places that I simply came to say to him, “ This is what is due to chance. l et us
best suited them."61 decide.” But faced with my figures, he responded with the conviction
Tzara had not been led to the “ question that arises” entirely on of a man who, having discovered a treasure, is told that some of his
its own. In July 1961, one M. Puissegur, teacher o f mathematics in pieces, if not all, are counterfeit. To my doubts, he responded with
Nevers, happened, during his summer holidays, to read a newspaper his faith. It was difficult to separate the wheat from the chaff. But was
article announcing an extraordinary cryptographic find: Tristan there any wheat? He, too, must have asked himself this question, and I
Tzara had discovered a forgotten method, once employed by poets, believe that for this charming, elderly man, I was the cause, that morn­
for embedding the letters of names in verse.64 Few details of the ing, of some pain.66
technique were offered, yet the indications sufficed to arouse the
scholar’s curiosity. On the basis o f the examples given, Puissegur Tzara’s papers suggest that he took the mathematician’s arguments
began to devise a formal method for calculating, as best he could, the seriously, at least with respect to Rabelais, who, having been the
probability that names of a certain length might be present in lines subject of the newspaper article of 1961, was also their primary topic
of poetry, their letters distributed according to the symmetrical of conversation. “ Employing most ingeniously the calculus of prob­
form that Tzara had identified. “ Som ething” in the announcement, ability,” we read in a posthumously published note, “M. Puissegur
he would later recall, “ had ruffled m e.”65 has obtained interesting results concerning the intrusion of chance
Once he completed his computations, Puissegur wrote to Tzara, in the formation of Rabelais’s anagrams.”67 Yet Tzara maintained that
through the intermediary o f the paper, acquainting him with the such an “ intrusion” could be curtailed—above all, with respect to
results of his findings. They could be sim ply summarized: for Villon. The “authenticity” of the anagrams could be defended. “The
the formation of discontinuous anagram s spelling out names in proofs by which I have sought to demonstrate the authenticity of
poetry, “chance,” he concluded, “ needs no help.” In short, he took the anagrams in Villon’s work are isolated and diverse, yet they are
the postulate of a poetic procedure to be superfluous, because concordant.”68 Tzara distinguished, broadly speaking, between two
the laws of probability alone suffice for lines to contain within types of demonstration. The first involved a hermeneutic principle:
them the letters of certain names, distributed in strictly symmetri­ in each case, he held, what is hidden in verse responds semantically,
cal patterns with respect to a central axis. In an essay published in some form, to what is explicitly said in it. That such a rule rests
in the Bulletin de l ’Asssociation des professeurs de matbematiques de on readings that would be difficult to verify is a point Tzara appears

l ’enseignement public in 1971, eight years after Tzara’s death, he sum­ to have conceded: “ One could object,” he admits, “that these mean­
ings all reflect a hypothetical interpretation that I give of the facts.”
marized the formal reasoning that had led him to these conclu­
Then he advanced his second proposition, whose implications far
sions. He also recounted how his first letter, years earlier, had met
outstripped the interpretation of Villon, bearing on several centuries
with a courteous reply. The famous poet invited the young math­
of literary history in the Romance languages. “ The second phase in
ematician to visit him at home. “ Tzara received me in his rue
my demonstration has consisted in situating Villon’s own activity in
de Lille apartment full o f Picasso draw ings and African masks,”
a historical movement in which the use of anagrams in poetry can
Puissegur recalled. He presented Tzara with his calculations. “1
be seen to stretch back to the twelfth century. Leaving aside the
brought him no certitude,” he would observe, “ since to express a
question of the origin and the history of anagrams, which I will, inci­
probability (unless it is equal to o or to 1) is to make a figured con­
dentally, take as a subject of study elsewhere,” he continued, “I have
fession of ignorance.”

172
 DARK TONGUES SECRETS OF TRISTAN TZARA

found the prolongation of this activity, whose principles are immu­ statistical data about the relations of letters in the language. Yet “the
table, even when the meaning given to anagrams and the means of science of linguistics at its present stage is not equipped to construct
employing them may differ from poet to poet, all the way up to the a comprehensive system of the interdependence of letters in any
work of Rabelais, who is one of the last writers to have used it.”69 natural language, let alone fifteenth-century French. Unfortunately
For Tzara, the two-part proof was sound. “ It follows,” he concluded, such a system is a necessary component of any formula that could be
“that the doubts concerning the authenticity of the anagrams in the used to predict accurately the number of symmetrical anagrams that
work of Villon must be definitely set aside.” 70 can be expected to appear in a text by chance.”73
Yet the doubts were to return. In 1976, Lynn D. Stults pub­ Having renounced the possibility of determining the exact likeli­
lished an article in the journal Romania, “A Study of Tristan Tzara’s hood with which hidden names might be lodged in Villon’s poetry,
Theory Concerning the Poetry of Villon,” in which she recalled the scholar chose to evaluate Tzara’s hypothesis by establishing a
that in 1959, it had been announced that a major literary-historical more modest literary fact: the frequency of anagrams contained in
discovery would soon be revealed: “Apparently the celebrated Dada­ the corpus of Villon’s verse when compared with those of two of
ist poet, Tristan Tzara, had been spending his late years developing his contemporaries. To this end, Stults availed herself of a computer
a theory concerning the writings of Villon. Tzara had intended to program, specially designed by Robert Stults, named Vilgram, that
publish his findings in a two-volume work about Villon, but he died would search for anagrams of forty-three names in Villon’s L a is,
before his plans could be realized.” 71 Stults recalled that Puissegur, Vaillant’s L e d e b a t d e s d e u x s o e u r s , and Charles d’Orleans’ ballads.74
in an earlier article, had subjected the hypothesis of “ the celebrated Stults tabulated the findings reached by her computer, commenting
Dadaist poet” to linguistic scrutiny based on the principles of the that “ from the breakdown of these results one fact clearly emerges:
mathematics of probability. Puissegur, she noted, devised a tw o-part while all but six of the forty-three names searched for do appear in
formula to test Tzara’s discoveries: “ In the first part the number of these six hundred lines of poetry, the majority of them do not appear
possible symmetrical positions for an anagram of a given number significantly more frequently in the lines by Villon than in the lines
of letters is calculated. In the second part the probability that the by Charles d’Orleans.” 75
letters occupying those positions be identical with the letters of a “ One could perhaps argue,” she conceded, “that Tzara’s theory
particular name is calculated.” 72 Although such reasoning is “math­ is strengthened by the rare occurrence of a long anagram, as for
ematically sound,” Stults wrote, “ its application to French poetry example in stanza 6 5 of V E m b u s c h e d e V a illa n t , le u r s f a i t z n e sont
is not.” The reason is simple. Puissegur’s formula accounts for the p o i n t v i c i e u x ”:

relative probability with which any letter will appear in the language
(based, of course, on the statistical frequencies of modern, rather
than medieval, French). The formula cannot, however, account for
the probabilities dictated by the interdependence of letters in the
language. One may, for example, calculate the relative probabilities
“ On the other hand,” she continued, “ it would be rather uncon­
that the letters Land R will appear in the language; but the relative
vincing to follow the same logic and yet explain how the name of the
probabilities that an F will appear after a R, or before it, or near it
Secretary of State of the United States crept into line 130 of Le L ais,
are another matter. The problem, Stults admitted, is not strictly
R e g n i e r d e M o n t i g n y t r o is c h i e n s ”:
mathematical. Linguistic research would need to provide crucial

OS
17 4
 DARK TO N G U E S SECRETS OF T RI STA N TZARA

one be certain that the program is truly capable of recovering all

the symmetrical anagrams in a given verse?”82 The second criticism
was one o f literary method. He recalled that Stults had compared
the number of anagrams of selected names in Villon’s poetry with
those in the work of two of his medieval contemporaries, Charles
Stults also noted a curious detail that Tzara seems not to have d’Orleans and Jean Vaillant. Having proven that the quantities of
observed. “ Often when the letters needed to form the anagram are hidden names were comparable in all three works, Stults judged it
present in a line, several symmetrical anagrams o f the same name unlikely that Villon possessed an anagrammatic art of his own. Ber­
appear in the same line.” 77 The name “ Noe Jo lis” thus appears not nard pointed out that such a line of reasoning was weak.
once but four times in the fifth line o f the Lais, and Stults revealed
Charles d’Orleans is not only a contemporary but also a friend of Vil­
the appearance of no fewer than “ twenty Perrenets in line 285:.”
lon’s, who could certainly have written under the name of the prince,
It would be difficult to imagine that the appearances of all 20 symmet­ or have entered into a contest of virtuosity with him, in the tradition of
rical anagrams of the name Perrenet in line 285 of L e L a is was caused by the g r a n d s r h e t o r iq u e u r s . The second corpus is even more problematic,
the manipulations of the poet, and yet it would be equally difficult to since L. D. Stults studies the anagrams in L ’em busche V aillant, which
determine which one of the tw enty anagrams he intended to appear. It Tzara, precisely, attributes to Villon! In theory, all three texts could
seems more reasonable to attribute these multiple anagrams to chance therefore be written by the same author, and when L.D. Stults shows
rather than to the efforts of the poet. 78 that there are even more anagrams in the works of Charles d’Orleans
than in those of Villon, she does not necessarily demonstrate the false­
Stults could only conclude that although Puissegur’s demonstration
hood of Tzara’s theses. 83
was in part faulty, its conclusion was sound. “ The results of my ana­
gram search strongly suggest that Tzara’s theory is invalid.”79 Bernard argued that a definitive refutation of Tzara’s claims
One might have thought the secret o f The Secret o f Villon dis­ would need to involve poems unrelated to the literary culture of
solved. More than fifteen years later, however, another scholar, fifteenth-century France. This he set out to provide. Availing him­
Michel Bernard, reopened the dossier in an article, “ The Secret oj self a computer program whose technical specifications he did not
Villon Put to the Test of the Computer: Tzara and the Anagrams,” conceal, Bernard presented a thorough analysis of the combinatorial
published in the pages of the same journal.80 Respectful of his prede­ possibilities of a single line by the nineteenth-century poet Gerard
cessors in the mathematical study o f linguistic regularities in poetry, de Nerval, “Je suis le Tenebreux—le Veuf—I’lnconsole.” He showed
he permitted himself nonetheless to restate Stults’s criticisms of that in that alexandrine, one may detect, arranged according to
Puissegur. To it he now added that Stults’s proof, too, was, from both the secret procedure identified by Tzara, a startlingly large set of
a technical and a literary perspective, unsound. Although Stults’s distinguished authors: not least Breton, Corneille, Celine, Crevel,
Ionesco, Roussel, and Villon.84
essay is “the more convincing of the tw o,” he wrote, it “ presents two
Yet Bernard also applied his program to the medieval corpus
failings that call her conclusions into question.” 81 The first involved
Tzara had studied. In the same 537 lines in which Tzara, by pains­
software: insufficient information had been given about the com­
taking analysis, had recovered 1,235 anagrams, Bernard’s program,
puter program, Vilgram, since the exact algorithm that had been
working for twenty-four hours, found almost three times as many:
employed had not been specified. Hence a troubling doubt: “Can

177
176
 S E C R E T S OF TRISTAN TZARA

3359 hidden names. “And not all Tzara’s orthographic lice , - 1.

added, “ were even used.”” Bernard had shown that the scholar '' h
had sought to dismiss the theses of The Secret of Villon were in
mistaken in their methods and, at the limit, that they wereunable to
offer a firm foundation for their claims. Yet their intuitions were' «
be definitively confirmed when Tzara failed his last statistical test'
Despite the differences between these variously scholarly
approaches to The Secret of Villon, certain critical elements in the
discussion have remained constant. The question to be resolved, as
Tzara himself, in the end, appears to have conceded, is not whether
or not there are names hidden in the lines of the Lais, but to what
degree it is linguistically probable that they are in fact present there.
One may infer from the research that has been conducted on Villon’s
poetry that the precise likelihood of encountering the symmetrically
placed letters of certain names in fifteenth-century manuscripts may
remain unknown forever. What can be determined, however, is the
approximate degree of probability that the letters of certain names
may be distributed, according to the procedure, in octosyllabic and
decasyllabic lines in the language.
It is here that one reaches what may be the most curious element
in the mathematically informed refutation of Tristan Tzara’s last
work. Explicitly or implicitly, all the critics of The Secret of Villon
orient their arguments with respect to one unstated axiom in Tzara s
investigation: the axiom, namely, that it is unlikely that there will
be names legible in Villon’s verse as the procedure would dictate,
or, positively stated, that the presence of such names, if established,
will constitute an inherently improbable occurrence. Tzara means
his unlikely hypothesis—that Villon embedded the letters of certain
names in his lines, beyond the threshold of auditory or visual per
ception—to be supported by this unlikely fact: the names are truly
there. The probability of an unlikely claim, in other words, isto e
b
verified by an improbable but true proof. To refute Tzara, it suf-
fices, then, to reject the validity of his evidence. If it is more ey
lik
than not that letters in a line, disjoined and placed insymetrical
positions, will spell out certain names, then the improb lecim
ab

179
 DARK TONGUES

will he revealed to be no more than what it appears to be: not only

improbable, but, indeed, even arbitrary. Hence the curious form of
disproof to which the critics ultimately submit Tzara’s argument,
which consists, in essence, in reasoning that the poet-scholar was
mistaken to think that he had recovered some one thousand two
hundred anagrams, because there are more than twice as many there
that he did not see. He was more right, in other words, than he ever
dared to hope. Therefore he was wrong.
Did the author of The Approximate Man anticipate such forms of
reasoning? “ It seems that it exists,” he had declared in 1920, “more
logical; very logical; too logical; less logical; illogical; truly logical;
fairly logical.”86By choice, by calling, or by chance, Tzara, at the end
of his life, was still declining these various possibilities, while also
suggesting a field in which they might be tested. Another element,
however, also lies hidden in the revelation of Tzara’s last work. A
final mystery, so to speak, remains concealed, although from 1959
to this day, it has never ceased to lie in plain view. It consists in the
simple fact that Tzara refused to consider the possibility that the
poet’s treasure might lie buried in the one place in which one might,
in all likelihood, have expected to find it. Villon, by all accounts,
wrote a series of poems in a form of speech meant to conceal cer­
tain matters. They are his so-called “ Ballads in Jargon,” which
evoke the obscure idiom of the Coquillars, the bandits who plotted
their unmentionable acts, as the authorities of fifteenth-century
Dijon alleged, in “an exquisite language, which other people cannot
understand” (un langaige exquis, que aultres gens ne scevent entendre).
These ballads remain, to a significant degree, impenetrable to this
day. Tzara, a learned reader of Villon and the scholarship devoted to
him, knew them well. He had also read the critical essays published
on their strange language, those by Marcel Schwob not least. Yet as
Henri Behar, Tzara’s editor, observes, “curiously, none of his drafts
shows any trace of an anagrammatic investigation carried out on the
ballads in jargon.”87 Tristan Tzara, V illon’s Anagrams, TZR 701 (courtesy of the Bibliotheque litteraire Jacques
Doucet, Paris, France).
At the beginning of The Secret o f Villon, Tzara appears to have
anticipated that his book might well, for this reason, provoke some

180
 DARK T O N G U E S SECRETS OF TRISTAN TZARA

surprise. As if to dispel the possibility o f the reader’s bewilderment Tzara w ill then implicitly maintain, lies in the brightest language
or disappointment, he w rites: “ These inquiries do not cover the of all, its darkness dimly discernible in the speech that would seem
domain of Jargon. The presence o f anagrams is all the more probable most crystalline. The true secret, like poetry for Breton, is there
there, since this language, being in part recreated by Villon, must where one would least expect it.90In the light of evidence, in words
have rendered their formation easier.” 88 His statement is worth pon­ manifestly free of the artifice of jargon, there are signs, names, and
dering. It suggests, in truth, two propositions. The ballads, first, will purposeful patterns, whispered in “a kind of echo” or by a “voice in
not be considered as belonging to the corpus o f poems containing mute,” waiting quietly, below each line of verse, for their knowing
anagrams; for anagrams, second, are most probably present there. reader. To take dictation from these unheard sources, to copy what
It is difficult not to pose the question o f the logical link that is to had not been written, Tzara silenced the doubts that his interpreta­
obtain between these statements. But Tzara offers another reason tion provoked and that he himself lived, in part, to perceive. Diving
for the exclusion of jargon from his book. Raising the question of ever deeper into the perilous waters of Villon’s songs and ballads, he
the opaque poems in a note that was to accompany an unfinished would not abandon the image of the treasure that he sought, even
edition of Villon’s work, he comments: “ Despite the many more or if it was to be at the risk that the language he discovered lay on the
less fantastical attempts that have been made at decipherment, one crystal’s surface and that the secrets he uncovered were his own.
can only agree with Sainean, the greatest specialist of jargon, that
‘all things considered, most o f the terms o f the argot or jobelin will
probably remain a closed book for us, and that, forever.’”89
The ballads in jargon, therefore, are probably hermetic and cer­
tainly obscure. In both cases, they can have no place in The Secret
of Villon. The conclusion may seem sudden, but the truth is that it
is far from arbitrary. One may even go so far as to define the object
of Tzara’s last book in punctual opposition to the two characteris­
tics that, in these passing observations, he attributes to the poems
composed in the bandits’ cryptic tongue. In distinction from the
obviously “closed book” o f the ballads in jargon, the texts of the
Lais appear to conceal no hidden message. Their obscurity, if one
grants it, is uncertain in the extrem e. Their secret, by that token, is
unlikely. Faced with this fact, Tzara, in a sense, does no more than
respect a principle that one might well wish to grant: were the trea­
sure likely, were its appearance obvious, it would be none at all. He
takes a further step when he effectively deduces the reality of a hid­
den thing from that of its unlikelihood, inferring that Villon’s poems
exhibit a secret procedure not despite but because of the improb­
ability that such a possibility is real. The “ explosive material” con­
tained in this most vertiginous o f literary finds ensues. The mystery,

182
 Notes

CHAPTER ONE: FORKINGS

1. Aristotle, Politics 125339.

2. See Johannes Lohmann, Musike und Logos: Aufsatze zur griechischen Philoso
phic und Musiktheorie.

3. On this and other basic principles in the epistemology of linguistics, see

Jean-Claude Milner, Introduction d une science du langage, esp. pp. 23-90.
4. Stephane Mallarme, “ Crise de vers,” in (Euvres completes, pp. 363-64.
5. See Jurgen Werner, “ Nichtgriechische Sprachen im Bewusstsein der anti­
ken Griechen” and “ Kenntnis und Bewertung fremder Sprachen bei den antiken
Griechen I. Griechen und ‘Barbaren’: Zum Sprachbewusstein und zum ethnischen
Bewusstsein im fruhgriechischen Epos”; Bruno Rochette, “Grecs et latins face
aux langues etrangeres: Contribution a l ’etude de la diversite linguistique dans
l’antiquite classique”; Vincenzo Rotolo, “ La communicazione linguistica fra allo-
glotti nell’antichita classica” ; Anna Morpurgo Davies, “ The Greek Notion of
Dialect.”
6. See Homer, Iliad 2, 867; Odyssey 19, i72sq.
7. Herodotus, Histories 4.45. 2-5.
8. Ibid., 2.50; for a discussion, see R ichm ond Lattim ore, “ Herodotus and the

Names of Egyptian Gods.”

9. See Diodorus 1.8.3, in Diels and Kranz (eds), Die F r a g m e n te d e r Vorsokra

fragment B 5.
10. See Genesis 11:1-9 .
11. Quintilian, Institutio oratorio 11.3.87.
12. See Rousseau, “ Essai sur l’origine des langues, ch. 1.
13- Lucian, De saltatione 64.

185
 NOTES
DARK TONGUES

14. Dante Alighieri, De vulgari eloquentia, 1.9-4* in Opere minori, p. 74. sur l ’argotfrangais, pp. 7 -6 0 .
18. Ibid., pp. 6 0 -6 4 .

CHAPTER TWO: C O Q U ILLARS 19. Ibid., p. 65.

1. Marcel Schwob, “ Le jargon des Coquillars en 14 5 5 ,” in Memoires de la societe 20. Marcel Schwob, “ Francois Villon,” in (Euvres, vol. 1, p. 10.

linguistique de Paris (June 1890), pp. 16 8 -8 3 and 2 9 6 -3 2 0 , reprinted in Marcel 21. See F ra n c is Villon, Poesies completes, p. 327.

Schwob, Etudes sur l ’argotfrangais, pp. 6 1- 15 0 and 83. 22. Clement Marot, (Euvrespoetiques vol. 2, p. 778. On Marot’s relation to Vil­

2. Such a translation supposes that the word coquillar derives, as the legal dos­ lon, see Jacqueline Cerquiglini-Toulet, “Marot et Villon.”

sier suggests, from the word coquille, “ shell.” It has also been suggested that the 23. Geoffroy Tory, Champ fleury, ou L’art et science de la proportion des lettres,

term derives from the word coq, “ rooster,” the bandits seizing their victims “with A VIII r°: “ Quant Iargonneurs tiennent leurs Propos de Ieur malicieux Jargon &

a powerful blow” like that by which roosters leap on chickens. See Pierre Guiraud, meschant langage, me semblent quilz ne se monstrent seullement ester dediez
Le jargon de Villon, ou Le gai savoir de la Coquille, p. 8. au Gibet, mais quil seroit bon quilz ne feussent oncques nez. Ja^oit que maistre
3. Schwob, Etudes sur l ’argotfrangais, p. 81. Fran5ois en son temps y aye este grandement ingenieux, si touteffois eust il myeulx
4. Ibid., pp. 81-82. faict d ’avoir etendu a faire autre plus bonne chose.”
5. Ibid., p. 82. 24. Cited in M arcel Schwob, “ Francois Villon et les compagnons de la
6. Ibid., pp. 82-83. Coquille,” lecture presented to the Academie des Inscriptions, April 2, 1890,
7. Ibid., pp. 83-84. reprinted in Marcel Schwob, Frangois Villon, p. 34. Cf. Lucien Schone, Lejargon et
8. Ibid., p. 91. jobelin de Maistre Frangois Villon, suivi de Jargon au theatre, p. 22.
9. Ibid., p. 92. 25. Schwob, Frangois Villon, pp. 29-30.
10. Ibid., pp. 86 -9 0. 26. Ibid., p. 34.
11. Ibid, p. 61. Cf. Pierre Guiraud, L’argot, p. 49, who claims that the term is
the first recorded word o f jargon. CHAPTER THREE: P R I N C I P L E S OF C A N T
12. See, for example, Frederic G odefroy, Dictionnaire de l ’ancienne longue 1. The Oxford English Dictionary relates that the term cant, while apparently
frangaise et tous ses dialectes du lXe au XVe si'ecle, s.v. “ jargon” ; and Adolf Tobler derived from the Latin cantus and its neo-Latin forms (Provencal and Norman
and Erhard Lommatzsch, continued by Hans H elm ut Christm ann, Altfranzo- French cant and French chant), acquired its English sense by a derivation and devel­
sisches Worterbuch, s.v. “ jargon.” For a fuller and more recent treatment, see John opment that remains unknown. The earliest attestation of the term dates to 1553.
Trumpner, “ Slang and Jargons,” esp. pp. 6 6 0 -6 2 . See Oxford English Dictionary, s.v. “cant.”
13. Marie de France, The Eables, 4 6 .5 5 .13 - 14 , pp. 132 -33. 2. Julie Coleman, A History of Cant and Slang Dictionaries, vol. 1 : 1567-1784, p. 4.

14. Cited in Godefroy, Dictionnaire d e l ’ancienne langue frangaise, s.v. “ jargon.” 3. For Herodotus’s remarks on the language of the Amazons, see Histories

15. Tober-Lommatzsch (Ahfranzdsisches Worterbuch, s.v. “ jargon” ) gives as book 4, 117.

an example Alexandre le Grand, D 72, p. 198: “ Quant sout par les esteilles la lour 4. See John Trumper, “ Romance Slangs and Jargons,” p. 660.

entencion, / Un ris gitat de joie e dist une oreisun; Une charme en chaldeu, ne sai 5. On the argot in the Jeu de saint Nicolas, see Michel Dubois, “Sur un passage

pas le jargoun.” obscur du Jeu de saint Nicolas"; Francisque Michel, Etudes de philologie comparee sur

16. See Richard li biaus, vv. 33 34 -53 . l ’argot, pp. viii-ix; Lazare Sainean, Les sources de l ’argot ancien, vol 1, p. 3.

17. Georges Guieysse and M arcel Schw ob, “ £tude sur I’argot fran5ais," 6. Richard li biaus, vv. 3333- 53. P- 92-
7. See Sainean, Les sources de l'argot ancien, p. 3.
Memoires de la Societe Linguistique de Paris 7 (1889), pp. 5 -2 8 , reprinted in Etudes

187
1 8 6
 DARK TONGUES NOTES

8. See Piero Camporesi, // libro dei vagabondi, pp. 17 9 - 2 4 6 . Camporesi dates 31. Raym ond Q ueneau, Exercices de style, p. 123. On Queneau s javanais exer­

the manuscript to 1484-86 (p. 13). c is e , see M arc Plenat, “ M orphologie d ’un ‘langage secret’: Le javanais de Queneau,

Ibid., pp. 241-46. Description dans un cadre autosegm ental.”

to. For an English translation, see Anonymous, The Book o f Vagabonds and Beg­ 32. Q ueneau, Exercises in Style, p. 158.

gars: ll’it/i a Vocabulary o f Their Language. 33. For the Q ueneau’s exercise in loucherbem style, see Exercices de style, p. 122:

11. Ibid., p. 3. “ Un lourjingue vers lidim ege sur la lateformeplic arriere d ’un Iobustotem, je gaffe

12. Ibid., pp. 4-5. un lypetinge avec un long loukem et un lapeauchard entoure d’un lalongifau lieu
13. Thomas Harman, Caveat or Wareningjor Common Cursetors. de lubanrogue.” For an analysis o f loucherbeme, see Marc Plenat, “ Morphologie du
14. Ibid., Epistle, A iii.v. largonji des lou cherbem s.”
15. Lanthorne, ch. 1, B3r. cited in Colem an, A History o f Cant and Slang Dic­ 34 . In English, one m ay cite “ pig Latin ,” discussed by Morris Halle, “ Pho­
tionaries, vol. 1, p. 31. nology in G enerative G ram m ar,” pp. 6 2-6 3. Cf. the case of the “pa-language of
16. See the first two volumes of Coleman’s History o f Cant and Slang Dictionaries. Mayalam,” described and analyzed by Karuvannur Puthanveettil Mohnan, “Lexical
17. Henri Estienne, Apologie pour Herodote, vol. 1, pp. 2 11- 12 . Phonology,” pp. 1 4 1 - 4 3 . The bibliography on secret languages is abundant.
18. See Pechon de Ruby, La vive genereuse des Mercelots, Gueuz et Boesmiens, 35. Pierre G uiraud, Largot, p. 59.
contenans leurfagon de vivre, subtilitez 8cgergon. 36. See G u irau d ’s analysis in ibid., p. 50.
19. See Francisco Adolpho Coelho, Os ciganhos de Portugal, com um estudosobre 37. See Joh n Ay to, Oxford Dictionary o f Rhyming Slang, p. 175.
0 calao. 38. Ibid., p. 213, w here the rhyme appears in its entire form.
20. The German word Gauner appears to derive from the Yiddish ganef, “ thief.” 39. On rhym ing slang in general, see Leonard R. N. Ashley, “ Rhyme and Rea­
On the Yiddish term, see Uriel W ein rich,,Modern English-Yiddish, Yiddish-English son: The M ethods and M eanings o f Cockney Rhyming Slang”; D. W. Maurer and
Dictionary, s.v. “ganef.” Sidney J. Baker, “ A ustralian R hym ing Argot in the American Underworld,”; St.
21. See Friedrich Christian Benedict Ave-Lallem ent, Das deutsche Gaunertum in Vincent Troubridge, “ Some Notes on Rhyming Argot”; and Ayto, Oxford Dictionary
seiner sozialpolitischen, literarischen und linguistischen Ausbildung zu seinem heutigen o f Rhyming Slang. On the history o f the study of rhyming slang, see Julie Coleman,
Bestande. A History o f Cant and Slang Dictionaries, vol. 3, pp. 91-9 6.
22. Alfredo Niceforo, II gergo nei normali, nei degenerati, e nei crimianli. 40. For Sh k lo vsk y’s program m atic statement, see “ Isskustvo, kak priyom”;
23. M .A.K. Halliday, “Anti-Languages." English in “ A rt as Technique.”
24. Marcel Cohen, “ Note sur la rg o t,” p. 137. 41. Paul Valery, “ Rhum bs,” in CEuvres, vol. 2, p. 637. Roman Jakobson long ago
25. See ibid, p. 139. Cohen also considers the related case o f playful jargons, drew attention to the “ realistic and scientific” aspect of this definition. See his
which are intended not to be secret, but may be so, in response to the “exigencies Selected Writings, vol. 3, p. 38.

of a game.”

26. See Jean-Claude Milner, Introduction a une science du langage, pp. 341-49- CHAPTER FO UR: SIGNS

27. Marcel Schwob, E tu d es s u r l a r g o t f r a n q a i s , p. 22. 1. See H ans-Robert Jauss, Die Alteritat und Modernitdt der mittelalterichen Lit-

28. Ibid., pp. 44-45. eratur: Gesammelte Aufsatze 1956-1976.

2. Paul Zum thor, I.angue et techniques poetiques a 1 'epoque romane (Xle-XIlIe
29. On verlan, see Natalie J. Lefkowitz, “ Talking Backw ards in French.”

30. There are other “ dialects” o f ja v a n a i s . See M arc Plenat, “ Le javanais: Con­ siecles), p. 208.
3. A lfre d P ille t , “ G ru n d lag e n , Aufgaben und Leistungen der
currence et haplologie.”

189
188
 DARK T O N G U E S NOTES

Troubadours-Forschung,” p. 342, reprinted in Der provenzalische Minnesang: Ein 20. In Adolphe Felix Gatien-Arnoult (ed.), La flors del gay saber, estier dichas
Qiierschnitt durch die neuere Forschungsdiskussion, p. 56. Las leys d ’Amors,vol. 1, p. 338. Cf. the explanation of the senhal in the verse version
4. See Istvan Frank, Repertoire metrique de la poesie des troubadours, vol. 1, p. xvi. of the Les leys d ’dmours, ed. Joseph Anglade, vol. 2, p. 176.
5. See Frank, Repertoire metrique; P. G. Beltrami, Rimario trobadorico provenzale; 21. Jean Boutiere and Alexander Herman Schulz, Biographies des troubadours,
Giovanna Santini, Rimario dei trovatori. LXXXIV, p. 267, 12-15.
6. On the expression trobar clus, see Ulrich M olk, Trobar clus, Trobar leu: 22. Ibid., XIV C, p. 28,1-2.
Studien zur Dichtungstheorie der Trobadors; cf. Aurelio Roncaglia, “ Trobar clus: 23. Ibid., XCIV B, p. 323,23-25.
Discussione aperta.” 24. Ibid., XC A, p. 310, 20-21.
7. Glynnis M. Cropp, Le vocabulaire des troubadours de Vepoque classique, p. 17. 25. Ibid., LXXXIII, p. 264,2-10.
8. See ibid., pp. 17-47. 26. Ibid., XVII A, p. 35, 8-10.
9. Andre Moret, Anthologie du Minnesang, p. 34, commented in Jacques Lacan, 27. See Ernest Hoepffner, “ Le Castiat du troubadour Peire Vidal.”
Vethique de la psychanalyse, 1959-1960, pp. 189 -9 0 . 28. Stanislaw Stronski (ed.), Le troubadour Folquet de Marseille, p. 33.
10. For an enumeration of the troubadours who give their name in their poems, 29. Boutiere and Schulz, Biographies des troubadours, XLVII, p. 151,23-24.
see F.M. Chambers, Proper Names in the Lyrics o f the Troubadours. 30. See Bertran de Born, The Poems of the Troubadour Bertran de Born, p. 55.
11. Purgatorio 26.117. 31. Isabel de Riquer, “Linhaure: Cent ans d’etudes sur un senhal,” p. 63.
12. For Guinizelli’s identification of Arnaut, see Pugratorio 26.117. For an earlv 32. Alfred Jeanroy, La poesie lyrique des troubadours, vol. 1, p. 317.
reading of the enigmatic phrase, see Jean Boutiere and Alexander Herman Schulz, 33. V. De Bartholomaeis, “ Du role et des origines de la tornade,” p. 462.
Biographies des troubadours: Textes proven$aux des XIlie et XlVe siecles, 10.43-45, p. 34. See Martin de Riquer, Los trovadores: Historia literariay textos (Barcelona:
15. Arnaut names himself in thirteen of his seventeen lyric poems. This signature Planeta, 1975), vol. 1, pp. 95-96.
appears in “ En cest sonet coind’e leri.” 35. Alois Richard Nykl, Hispanio-Arabic Poetry and Its Relations with the Old
13. For an exact account of the relative frequency of such invocations in the Provencal Troubadours, p. 271.
concluding stanza of the troubadour lyric, see Eduardo Vallet, “ II senhal nella lirica 36. Riquer, Los trovatores, vol. 1, pp. 95-96.
trobadorica,” p. 154. 37. See Bernart de Ventadorn 20.21-22.
14. On Addressentornaden, see Bernart de Ventadorn: Seine l.ieder mit Einleitung 38. Gerard Gourian, L’amour et la guerre: L’oeuvre de Bertran de Born, p. LXV;
und Glossar, p. XCIX. Luciano Formisano, La lirica, p. 69.
15. Alfred Christlieb Kalischer, Observationes in poesim romanensem provicialibus 39. Na Castelloza, “Ja de chantar no dreg’aver talen,” in Oskar Schultz-Gora
in primis respectis, pp. 25-26. (ed.), Die provenzalische Dichterinnen: Biographien und Texte, 8.59, p. 24. See Mari­
16. See Pasero (ed.), Poesie di Guglielmo IX, Mon Esteve, V II.47; Mon Bel Vezi, anne Shapiro, “ The Trobairitz and the Limits of Courtly Love,” p. 571.
X.26.
17. See Vallet, “ II senhal nella lirica trobadorica,” p. 114 . For the continuation CH APTER FIVE: KNOWINGS
and conclusion of Vallet s study, see his “ II senhal nella lirica trobadorica (con alcune 1. The expression is Henry Adams Bellows’s. See Bellows (ed. and trans.),

note su Bel/ Bon Esper in Gaucelm Faidit).” The Poetic Edda, p. xii. For a more recent appraisal, see Klaus von See, Beatrice

18. See Alfred Jeanroy, La p o esie ly r iq u e de s tro u b a d o u rs, vol. 1, pp. 317-18. La Farge, Eve Picard, Ilona Priebe, and Katja Schulz, Kommentar zu den Liedern

19. See the lexicographic analysis in Vallet, “ II senhal nella lirica trobadorica der Edda.
2. They have also been called the Edda Scemundar. For translations, see Bellows,
(con alcune note su Bel/ Bon Esper in Gaucelm Faidit)," pp. 283-85.

191
1 9 0
 DARK T O N G U E S NOTES

Troubadours-Forschung,” p. 342, reprinted in Der provenzalische Minnesang: Ein 20. In Adolphe Felix Gatien-Arnoult (ed.), Laflors del gay saber, estier dichas
Querschnitt durch die neuere Forschungsdiskussion, p. 56. Las leys d ’Amors,vol. 1, p. 338. Cf. the explanation of the senhal in the verse version
4. See Istvan Frank, Repertoire metrique de la poesie des troubadours, vol. 1, p. xvi. of the Les leys d ’dmours, ed. Joseph Anglade, vol. 2, p. 176.
5. See Frank, Repertoire metrique; P. G. Beltrami, Rimario trobadorico provenzale\ 21. Jean Boutiere and Alexander Herman Schulz, Biographies des troubadours,
Giovanna Santini, Rimario dei trovatori. LXXXIV, p. 267,12-15.
6. On the expression trobar clus, see Ulrich M olk, Trobar clus, Trobar leu: 22. Ibid., XIV C, p. 28,1-2.
Studien zur Dicbtungstbeorie der Trobadors; cf. Aurelio Roncaglia, “ Trobar clus: 23. Ibid., XCIV B, p. 323,23-25.
Discussione aperta.” 24. Ibid., XC A, p. 310, 20-21.
7. Glynnis M. Cropp, Le vocabulaire des troubadours de l ’epoque classique, p. 17. 25. Ibid., LXXXIII, p. 264, 2-10.
8. See ibid., pp. 17-47. 26. Ibid., XVII A, p. 35, 8-10.
9. Andre Moret, Antbologie du Minnesang, p. 34, commented in Jacques Lacan, 27. See Ernest Hoepffner, “ Le Castiat du troubadour Peire Vidal.”
L’ethique de la psychanalyse, 1959-1960, pp. 189 -9 0 . 28. Stanislaw Stronski (ed.), Le troubadour Folquet de Marseille, p. 33.
10. For an enumeration of the troubadours who give their name in their poems, 29. Boutiere and Schulz, Biographies des troubadours, XLVII, p. 151,23-24.
see F.M. Chambers, Proper Names in the Lyrics o f the Troubadours. 30. See Bertran de Born, The Poems of the Troubadour Bertran de Born, p. 55.
11. Purgatorio 26.117. 31. Isabel de Riquer, “Linhaure: Cent ans d’etudes sur un senhal,” p. 63.
12. For Guinizelli’s identification of Arnaut, see Pugratorio 26.117. For an early 32. Alfred Jeanroy, La poesie lyrique des troubadours, vol. 1, p. 317.
reading of the enigmatic phrase, see Jean Boutiere and Alexander Herman Schulz, 33. V. De Bartholomaeis, “ Du role et des origines de la tornade,” p. 462.
Biographies des troubadours: Textes provengaux des X llle et XlVe siecles, 10.43-45, p. 34. See Martin de Riquer, Los trovadores: Historia literariay textos (Barcelona:
15. Arnaut names himself in thirteen of his seventeen lyric poems. This signature Planeta, 1975), vol. 1, pp. 95-96.
appears in “ En cest sonet coind’e leri.” 35. Alois Richard Nykl, Hispanio-Arabic Poetry and Its Relations with the Old
13. For an exact account of the relative frequency o f such invocations in the Provengal Troubadours, p. 271.
concluding stanza of the troubadour lyric, see Eduardo Vallet, “ II senhal nella lirica 36. Riquer, Los trovatores, vol. 1, pp. 95-96.
trobadorica,” p. 154. 37. See Bernart de Ventadorn 20.21-22.
14. On Addressentornaden, see Bernart de Ventadorn: Seine l.ieder mit Einleitung 38. Gerard Gourian, L’amour et la guerre: L’oeuvre de Bertran de Born, p. LXV;
und Glossar, p. XCIX. Luciano Formisano, La lirica, p. 69.
15. Alfred Christlieb Kalischer, Observationes in poesim romanensem provicialibus 39. Na Castelloza, “Ja de chantar no dreg’aver talen,” in Oskar Schultz-Gora
in primis respectis, pp. 25-26. (ed.), Die provenzalische Dichterinnen: Biographien und Texte, 8.59, p. 24. See Mari­
16. See Pasero (ed.), Poesie di Guglielmo IX , Mon Esteve, V II.47; Mon Bel Vezi, anne Shapiro, “ The Trobairitz and the Limits of Courtly Love,” p. 571.

X.26.
17. See Vallet, “ II senhal nella lirica trobadorica,” p. 114 . For the continuation C H APTER FIVE: KNOWINGS
and conclusion of Vallet’s study, see his “IIsenhal nella lirica trobadorica (con alcune 1. The expression is Henry Adams Bellows’s. See Bellows (ed. and trans.),

note su Bel/ Bon Esper in Gaucelm Faidit).” The Poetic Edda, p. xii. For a more recent appraisal, see Klaus von See, Beatrice

18. See Alfred Jeanroy, La poesie lyrique des troubadours, vol. 1, pp. 317-18. La Farge, Eve Picard, Ilona Priebe, and Katja Schulz, Kommentar zu den Liedern

19. See the lexicographic analysis in Vallet, “II senhal nella lirica trobadorica der Edda.
2. They have also been called the Edda Scemundar. For translations, see Bellows,
(con alcune note su Bel/ Bon Esper in Gaucelm Faidit),’ pp. 283-85.

191
1 9 0
 DARK T O N G U E S NOTES

The Poetic Edda; Ursula Dronke, The Poetic Edda; Carolyne Larrington, The Poetic 19. Ibid.
Edda; Andy Orchard, The Elder Edda: A Book o f Viking Lore. 20. Frank, Old Norse Court Poetry, p. 46.
3. E. V. Gordon, An Introduction to Old Norse, p. xxxix. 21. Ibid., p. 47.
4. On the formal structure of the drottkvcett stanza, see Roberta Frank, Old 22. Ibid., p. 46.
Norse Court Poetry: The Drottkvcett Stanza; Hans Kuhn, Das Drottkvcett; Kari Ellen 23. Rudolf Meissner, Die Kenningar der Skalden: Ein Beitrag zur skaldischen
Gade, The Structure of Old Norse Drottkvcett Poetry. Poetik, p. 2.
5. For the Old Norse text, see Snorri Sturluson, Edda: Skaldskaparmal, ed. 24. See Eyvindr Finnsson Skaldaspillir, Die Dichtung des Eyvindr skaldaspilhr:
Anthony Faulkes, vol. 1; for translations, see Snorri Sturluson, Edda, ed. and trans. Edition-Kommentar-Untersuchungen, pp. 274-76.
Anthony Faulkes and The Prose Edda, trans. Arthur Gilchrist Brodeur. 25. Sturla bordarson, 3, in Kari Ellen Gade, Poetryfrom the Kings’ Sagas, vol. 2,
6. Snorri Sturluson, Skaldskaparmal 58, in Faulkes, Edda, vol. 1, p. 5; English From c.1035 to 1300, pp. 731-32.
in Faulkes, Edda, p. 64. 26. On Anglo-Saxon kennings, see Wilhelm Bode, Die Kenningar in der angel-
7. Ibid. sachsischen Dichtung; James Walter Rankin, “A Study of the Kennings in Anglo-
8. See the account in Heinrich Lausberg, Handbuch der literarischen Rhetorik: Saxon Poetry”; Hendrik van der Merwe Scholz, The Kenning in Anglo-Saion and
fine Grundlagung der Literaturwissenschaft, § 589-98. Old Norse Poetry; and Thomas Gardner, “ The Old English Kenning; A Character­
9. Snorri Sturluson, Skaldskaparmal 1, in Faulkes, Edda, vol. 1, p. 5.17-18; Eng­ istic Feature of Germanic Poetical Diction?”
lish in Faulkes, Edda, p. 64. 27. Andreas Heusler, Altgermanische Dichtung, p. 136.
10. The meaning of Snorri’s use o ffornqfn has been much discussed. See 28. Ibid., p. 137.
Faulke’s comments in Edda, vol. 1, pp. x x ix -x x x . Cf. Faulkes’s glossary, Edda, vol. 29. To the play of metaphor, which substitutes one term for another by “deno­
2, pp. 277-78; Snorri devotes a longer discussion tofornrfn in ch. 67 of the Skalds­ tation,” the kenning thus adds a play of metonymy, which functions by “connota­
kaparmal. Cf. Arthur G. Brodeur, “ The Meaning of Snorri’s Catgeories”; Margaret tion,” thus joining paradigmatic and syntagmatic processes. See Bjarne Fidjestol,
Clunines Ross, Skaldskaparmal: Snorri Sturluson’s ars poetica and Medieval Theories “ The Kenning System; An Attempt at a Linguistic Analysis,” esp. p. 22.
of Language, pp. 64-79. 30. See Gardiner, “The Old English Kenning.”
11. Roberta Frank, s.v. “ kenning,” Dictionary o f the Middle Ages. 31. Frank, Old Norse Court Poetry, p. 42.

12 .An Icelandic-English dictionary, s.v. “ kenna.” 32. Ibid., p. 45.

13. Snorri Sturluson, S k a ld s k a p a rm a l 1, in Faulkes, Edda, vol. 1, p. 5.25-28; 33. Snorri Sturluson, Edda: Hatttal, ed. Anthony Faulkes, 2, p. 5.11; English in

English in Faulkes, E d d a , p. 64. Faulkes, Edda, p. 168.

14. Snorri Sturluson, Skaldskaparmal 1, in Faulkes, Edda, vol. 1, p. 5.17-21; 3 4 . Ibid.

English in Faulkes, Edda, p. 64. Translation slightly modified. 35. Gordon, An Introduction to Old Norse, p. xl.

15. For the kennings of Odin, see S k a ld s k a p a rm a l 2 ; fo r the kennings of Freyja, 36. Jorge Luis Borges, “Las Kenningar,” pp. 42, 35, 53.

37. Frank, Old Norse Court Poetry, p. 28.

see Skaldskaparm al 20.
38. Borges, “Las Kenningar,” p. 50.
16. Snorri Sturluson, Skaldskaparmal 69, in Faulkes, Edda, vol. 1, p. 108.6-8;
39. G ordon, An Introduction to Old Norse, p. xi. Cf. Frederic Amory, “Towards
English in Faulkes, Edda, p. 153.
A Gram m atical Classification o f Kennings as Compounds,” which builds in large
17. Snorri Sturluson, Skaldskaparmal 25, in Faulkes, Edda, vol. 1, p- 36.24-29,
part on this hypothesis.
English in Faulkes, Edda, p. 91.
4 0 . O n the literary system of the kennings, see Edith Marold, K en n in g k u n st:
18. Frank, “ Kenning,” p. 230.

193
192
 DARK T O N G U E S
NOTES

Fin Beitrag tu ciner Poetik der Skaldendichtung. For an analysis of the kennings as 15. Rig Veda 1.164.45. Text and translation in George Thompson, “The hr*f
forming a linguistic system, see Fidjestol, “ The Kenning System: An Attempt at a modya and Vedic Discourse,” p. 15; for a different English translation, see Dom^t
Linguistic Analysis.” Cf. the synthetic perspective offered by Gary Holland, “ Ken­ The Rig Veda, p. 80.
nings, Metaphors, and Semantic Formulae in Norse drottkvcett.” 16. Porzig, “ Das Ratsel im Rigveda,” p. 654.
41. Snorri Sturluson, Skdldskaparmal 57, in Faulkes, Edda, vol. 1, p. 3.5-8; 17. Ibid., p. 656.
English in Faulkes, Edda, p. 61. 18. Andreas Heusler, Das Altgermanische Dichtung, pp. 131-32.
42. Ibid. The phrase “ i runum” could also mean “ in w riting.” On run, see Jan 19. John Lindow, “ Riddles, Kennings, and the Complexity of Skaldic Poetry "
de Vries, Altnordisches etymologisches Worterbuch, pp. 4 5 3 -5 4 . P- 312 .
20. Hugo Gering, cited in Henry Adams Bellows, The Poetic Edda, p. 183.
CHAPTER SIX: R ID D L E S 21. Alvissmal, 9-10 , in Finnur Jonsson, Degamle Eddadigte, p. 133.
1. See Oxford English Dictionary, s.v. “ secret.” 22. Andy Orchard (trans.), The Elder Edda: A Book of Viking Lore, p. 109.
2. Archer Taylor, “ The Riddle,” p. 141. 23. Several hypotheses have been advanced for the reasons of the transforma­
3. John C. Messenger, Jr., “Anang Proverb-Riddles,” p. 226. tion. See Paul Acker, “ Dwarf Love in Alvissmal,” in Paul Acker and Carolyne
4. John Blacking, “ The Social Value of Venda Riddles,” p. 5. Larrington, (eds.), The Poetic Edda: Essays on Old Norse Mythology.
5. Lyndon Harries, “ The Riddle in Africa,” pp. 379-80. 24. On the principles that govern this identity, see the illuminating paper
6. For Plato’s definition of the riddle, see Plato, Charmides, i6 ic-d ; cf. Aristo­ by Lennart Moberg, “ The Languages of Alvissmal.” Cf. Rudolf Meissner, “Die
tle, Poetics i458a26-3o. On the rhetoric of the enigma, see Eleanor Cook, Enigmas Sprache der gotter, riesend und zwerge in den Alvissmal.”
and Riddles in Literature, esp. pp. 27-63. 25. Jonsson, De gamle Eddadigte, p. 55.
7. Taylor, “ The Riddle,” p. 129. 26. Vafprudnismal, 11-12, in Orchard, trans., The Elder Edda, p. 41.
8. For the riddle, see Archer Taylor, English Riddles from Oral Tradition, iy)i\ 27. Christopher Tolkien (trans.), Saga Heidreks konungs ins vitra: The Saga of
for the commentary, see Robert A. Georges and Alan Dundes, “ Towards a Struc­ King Heidrek the Wise, (58), pp. 38-39.
tural Definition of the Riddle,” p. 112. 28. See Carolyne Larrington, “Vafjirudnismal and Grimnismal: Cosmic His­
9. George and Dundes, “ Towards a Structural Definition of the Riddle,” p. 113. tory, Cosmic Geography,” p. 63.
10. See, for example, Charles T. Scott, Persian and Arabic Riddles: A Language- 29. Anders Hultgart, “ The Wisdom Contest in the Vafj)rudnismdl,npp. 536-38.
Centered Approach to Genre Definition, p. 8. 30. On the brahomdya, see Ernest Renou and Lilian Silburn, “Sur la notion de
11. For Frege’s distinction, see Gottlob Frege, “ Uber Sinn und Bedeutung,” brah m an Thompson, “ The Brahmodya and Vedic Discourse”; and Tat’jana Jakov-
Zeitschriftfiir Philosophie undphilosophische Kritik, pp. 25-50 ; reprinted in Gottlob levna Elizarenkova and Vladimir NikolaevicToporov, “Zum vedischen Ratsel des
Frege, Funktion, Begrijf, Bedeutung, pp. 4 0 -6 5 . Typs Brahmodya,” in Eismann, and Grzybek (eds.), Semiotische Studien zum Ratsel:
12. Rig Veda 1.16 4 . 1° Wendy Doniger’s edition and translation, The Rig Veda, it Simple Forms Reconsidered II, pp. 39-73. On Sanskrit riddles more generally, see
is called the “ Riddle of the Sacrifice.” See pp. 7 1-8 3. On the “ Riddle Hymn,” see Ludwik Sternbach, Indian Riddles: A Forgotten Chapter in the History of Sanskrit
Walter Porzig, “ Das Ratsel im Rigveda: Ein Beitrag zum Kapitel Sondersprache”; Literature.
and Jan E.M. Houben, “ Ritual Pragmatics of a Vedic Hymn: The ‘ Riddle Hymn’ 31. Thompson, “ The Brahmodya and Vedic Discourse,” p. 14; cf. Renou, “Sur

and the Pravargya Ritual.” la notion du brahman,” p. 97.

13. Porzig, “ Das Ratsel im Rigveda," p. 654. 32. Thompson, “ The Brahmodya and Vedic Discourse,” p. 17.

14. Ibid., p. 655. 33. Vajasaneyi Samhita 23.61-62 (a quotation of Rig Veda 164.34-46), cited in

195
194
 DARK T O N G U E S NOTES

Thompson, “ The Rrahmodya and Vedic Discourse,” p. 17. may be misleading; it suggests letters (grammata), while many of the tradition*

34. Tolkien (trans.), Saga Heidreks konungs ins vitra, p. 44. The same question in question developed in the absence of systems of writing. On “anagrammatK

appears at the end of the Vaffrrudnismal (54), suggesting, as many scholars have riddles,” see also Vladimir Nikolaevic Toporov, “Das Anagramm im Ratsel,” in

pointed out, a link between the two literary contests in wisdom. Eismann and Grzybek (eds.), Semiotische Studien zum Ratsel: Simple forms Recon­

35. Ibid. sidered II, pp. 181-90.

36. See W.J. Pepicello and Thomas A. Green, The Language o f Riddles: New 53. Milicic, p. 75. Milicic gives eleven examples, which, as he indicates, he

Perspectives, who distinguish systematically, in similar terms, between the “meta­ culls from the set of thirty-four riddles published by David Evans in his “Riddling
phorical” and the “grammatical.” and the Structure of Context.”
37. Archer, English Riddles, 587 B, discussed in Pepicello and Green, The Lan­
guage of Riddles, p. 215. C H A P T E R SE V E N : NOM INA DIVINA
38. See the helpful treatment in Pepicello and Green, The Language of Riddles, 1. For a helpful summary of the corpus, see M. L. West, commentary to 831,
p.27. in Hesiod, pp. 368-88. On the Greek language of the gods, see Hermann Giintert,
39. Ibid., p. 37. Von der Sprache der Gotter und Geister: Bedeutungsgeschichtliche Untersuchungen
40. Ibid., p. 38. zur homerischen und eddischen Gottersprache; Alfred Heubeck, “Die homerische
41. Ibid., p. 35. Gottersprache”; Calvert Watkins, “Language of Gods and Language of Men:
42. Ibid., p. 45. Remarks on Some Indo-European Metalinguistic Traditions”; Romano Lazzeroni,
43. Ibid., p. 34. “ Lingua degli dei e lingua degli uomini”; Jose Caldereon Felices, “Lengua de los
44. Miguel de Cervantes Saavedra, La Galatea, p. 245. dioses/Lengua de los hombres”; Ann Suter, “Language of Gods and Language of
45. See Dan Pagis, ‘A l sod hatum: Le-toldot ha-hidah ha-‘Ivrit be-ltalyah uve- Men: The Case of Paris / Alexandros”; and Fran£oise Bader, La langue des dieux ou
Holand, p. 248. Thermetisme des pokes indo-europeens.
46. Dan Pagis, “ Towards a Theory of the Literary Riddle,” p. 87. 2. Bader, La langue des dieux, p. 191.
47. Ibid., pp. 87-90. 3. Homer, Iliad 1.403-404.
48. Ibid., p. 88. 4. Ibid., 2.813-14.
49. So, too, the similarly formed “ Me lo dico, me lo ripeto, me lo torno a dir 5. Ibid., 14.290-91.
di nuovo, e se non lo capiro, testa d’asino saro; indovina che cos’e?” (il melo), “Te 6. Ibid., 20.72-74.
lo dico, te lo provo, te Lo torno a dir di nuovo, e se tu non lo capirai, testa d’asino 7. Homer, Odyssey 10.305-306.
sarai; indovina che cos’e?” (il telo). 8. Ibid., 12.61.
jo. Osip Maksimovich Brik, “ Zvukovye povtory," in Two Essays on Poetic 9. Eusthatius 1658.53, cited in Jenny Clay, “The Planktai and Moly: Divine
Language, p. 131; Roman Jakobson, “ Subliminal Verbal Patterning in Poetry," in Naming and Knowing in Homer," p. 128.

Selected Writings, vol. 3, pp. 14 0-4 3; Aleksandr losivic Gerbstman, “ O zvukovom 10. Ibid.

stroenii naradnoj zagadki.” 11. Plato, Cratylus 39 ^ 4 .

51. Tzvetan Todorov, Lesgenres du discours, p. 245. 12. On the passage, see among others, John E. Joseph, Limiting the Arbitrary:

52. Elizarenkova and Toporov, “ Zum vedischen Ratsel des Typs Brahmodya"; Linguistic Naturalism and Its Opposites in Plato’s Cratylus and Modern Theories of

Vladimir Milicic, “ Subliminal Structures in Folklore: Anagrammatized Answers Language, pp. 39-58.

to Riddles,” pp. 63-76. The term “anagrammatic,” while commonly employed, 13. J. van Leeuwen, “ Homerica XVI: De lingua deorum.”

197
1 9 6
 NOTES
DARK TONGUES

14. H. Jacobsohn, “ Aeolische Doppelconsonanz: Zur Sprache und Verstechnik 35. See Herm ann G iintert, Uber die ahurischen und daevischen Amdriicke m

des Homerischen Epos,” pp. 81-84. Awesta: Eine semasiologische Studie, as well as the critical remarks in £mile Ben

15. Paul Kretschmer, review of Giintert, Von der Sprache der Gotter und Geister. veniste, “ Une differenciation de vocabulaire dans 1 'Avesta.” Cf. Bader, La langue

16. Antoine Meillet, review of Hermann Giintert, Von der Sprache der Gotter des dieux, pp. 2 4 6 -4 7.

und Geister, p. 56. 36. See George Calder (ed.), Auraicept na N-Eces: The Scholar’s Primer, Being the

17. Romano Lazzeroni, “ Lingua degli dei e lingua degli uomini,” pp. 14-15. Texts o f the Ogham Tractfrom the Book o f Ballymote and the Yellow Book of Pecan, and

18. Jakob Grimm, Deutsche Mythologie, vol. 1, p. 275. the Text o f the Trefhoculfrom the Book o f Leinster.

19. Ibid., p. 276. For special studies of the languages o f the Alvi'ssmal, see Lenn­ 37. Watkins, “ Language o f Gods and Language of Men,” p. 9, which cites the

art Moberg, “ The Languages of Alvi'ssmal,” and Rudolf Meissner, “ Die Sprache der long version o f the Yellow Book o f Lecan (4554-56).

gotter, riesend und zwerge in den Alvissm al.” 38. Watkins, How to K ill a Dragon, p. 181.

20. Giintert, Von der Sprache der Gotter und Geister. 39. See West, Indo-European Poetry and Myth, p. 160.
21. Ibid., pp. 110 -11; cf. p. 155. 40. Bader, La langue des dieux, p. 248.
22. Ibid., p. 170. 41. That there are elaborate non-Indo-European traditions of the “ language
23. Ibid. pp. 157-58. of gods” may also be worth recalling. See the comparative study, Marcel Detienne
24. Ibid., p. 158. and Gilbert Hamonic, et al. (eds.), La deesse parole: Quatrefigures de la langue des
25. Rig Veda 1.164.45. dieux.
26. M.L. West, Indo-European Poetry and Myth, p. 161. 42. Watkins, “ Language of the Gods and Language ofMen,” p. 2.
27. Ibid. 43. See Roman Jakobson’s influential formulation, “Signe zero,” in Selected
28. $atapatha Brahmana 1.1.4 .4 . For a translation, see West, Indo-European Writings, vol. 2, pp. 2 11- 19 , wdiich rests, as Jakobson makes clear, on Saussure’s
Poetry and Myth, p. 161. famous principle that language can content itself with the opposition of something
29. Calvert Watkins, How to K ill a Dragon: Aspects o f Indo-European Poetics, p. and nothing (Corns de linguistique generale, p. 124).
265. On the Asvamedha, see ibid., “ The As'vamedha or Horse Sacrifice: An Indo- 44. Watkins, “ Language o f the Gods and Language of Men,” p. 7.
European Liturgical Form,” pp. 265-76. 45. Ibid., p. n .

30. Satapatha-Brahmana 10 .6 .4 .1. See Vladim ir N. Toporov, “ Die indoeu- 46. Walter Porzig, “ Die Raitsel im Rigveda,” p. 656.

ropaische Poetik und ihre Urspriinge,” p. 204. See also Bader, l a langue des dieux,
pp.248-49. CHAPTER EIGHT: NOM INA SACRA

31. West, Indo-European Poetry and Myth, p. 161. 1. C. Suetonius Tranquillus, De vitae caeserum, DivusJulius, 56.6; English in The

32. See Watkins, “ Language of Gods and Language o f M en," pp. 5-6; Giintert, Lives o f the Twelve Caesars, p. 80.

Von der Sprachen der Gotter und der Geister, p. 132; Toporov, “ Die indocuropaische 2. C. Suetonius Tranquillus, De vitae caeserum, Divus Augustus, 88.3; The Lives of
the Twelve Caesars, p. 257. On the two Caesars’ “ciphers,” see Albert C. Leighton,
Poetik und ihre Urspriinge,” pp. 204-205; Bader, La langue des dieux, pp. 248-49.
“ Secret Communication among the Greeks and Romans,” pp. 152-53.
33. E. Laroche, “ Hattie Deities and their Epithets,” p. 188.
3. For Shesakh in place o f Bavel, see Jeremiah 25:26 and 51:41. Perplexed by
34. Johannes Friedrich, “ Gottersprache und Menschensprache im hethitischen
the reasons for the procedure, Wilhelm Gesenius already raised doubts about its
Schriftum,” p. 139. Cf. Lazzeroni, “ La lingua degli dei," pp. 18-23; Watkins,
existence in the prophetic text: “Assuming, but not conceding, that this secret
“Language of Gods and Language of Men,” pp. 7 -8 ; Bader, l.a langue des dieux,
language is really so old. one can still see no reason why in 51, 41 Babylon should
pp. 244-45.

1 9 9
19 8
 DARK T O N G U E S NOTES

be called once by its true name and the other time by a concealed name.” See books in classical India, see Malamoud, Le jumeau solaire, p. 128.
his Hebrdisches und chalddisches H a n d w o rte rb u c h iib e r d a s A lt e T estam en t , s.v. “ssk.” 22. Edward C. Sachau, Alberuni’s India: An Account of the Religion, Philosophy,
For a more recent discussion that argues in support of the antiquity of atbash, see Literature, Geography, Chronology, Astronomy, Customs, Laws and Astrology of India
Richard C. Steiner, “ The Two Sons of Neriah and the Two Editions of Jeremiah about A.D. 1030, vol. 1, pp. 125-26.
in the Light of Two Atbash Code-Words for Babylon,” pp. 8 1-8 4. On related 23. Julius Caesar, De bello gallico, 6.13; English in Commentaries on the Gallic
practices in the ancient world, see Aaron Demsky, “ A Proto-Canaanite Abecedary and Civil Wars, with the Supplementary Books Attributed to Hirtius, Including the
Dating from the Period of the Judges and Its Implications for the History of the Alexandrian, African and Spanish Wars.
Alphabet,” pp. 19-20. 24. Ibid., 6.14.
4. Virgilius Maro, E pitom ae 10 .4 -9 , in V irg iliu s M a ro G ra m m a ticu s: Opera omnia, 25. Ibid.
p. 213. For a discussion of Virgilius’s teachings on the subject, see Vivian Law, 26. See, for instance, Georges Dumezil, “ La tradition druidique et l’ecriture,”
Wisdom, A uthority a n d G ram m ar in th e S ev en th C e n tu ry , pp. 84-88. PP- 125-33.
5. See Charles Malamoud, C u ire le m o n d e : R it e et p e n s e e d a n s l ’ln de ancienne, 27. Tacitus, Germania 10.1, in Cornelii Taciti De Germania, ed. Henry Furneaux;
p. 242. English in The Complete Works of Tacitus, p. 713.
6. See ib id ., p. 243. 28. On the semantic field of Old English run and its cognates, see R.I. Page,
7. Jan Gonda, N otes on N am es a n d th e N a m e o f G o d in A n c ie n t In d ia , p. 81. An Introduction to English Runes, pp. 106-107.
8. Malamoud, C uire le m o nde, p. 243. 29. Ibid.; for a less skeptical position, see R.W.V. Elliott, Runes, pp. 65-66.
9. Ibid. On runes and magic, see Karl Martin Nielsen, “Runen und Magie: Ein forschun-
10. Ibid ., p. 244. A ita re y a -A ra n y a k a 2.4.3; A it a r e y a - U p a n is a d 5.1 jff. gsgeschichtliche Uberblick”; and, for Anglo-Saxon runes in particular, LA. Page,
11. Ibid. Text in S a ta p a th a -B ra h m a n a , 6 .1.1.11. “Anglo-Saxon Runes and Magic.”
12. Ib id ., p. 245. 30. Rene Derolez, Runica manuscripta: The English Tradition, pp. 354-55.
13. Ib id ., p. 246. A ita re y a -B ra h m a n a 3.33, English p. 219. 31. Robert DiNapoli, “ Odd Characters: Runes in Old English Poetry,” p. 148.
14. Malamoud, C uire le m o n de, p. 246. 32. Riddle 58, lines 14-15, in Craig Williamson (ed.), The Old English Riddles
15. Ib id ., p. 248. of the Exeter Book, p. 102.
16. Charles Malamoud, Le ju m e a u so la ir e , p. 131. On writing and speech in clas­ 33. The solution would appear to be radrod, “well sweep,” a tool for drawing
sical India more generally, see ib id ., pp. 127-4 9 ; Charles Malamoud, Femtnitedela water. See Williamson, The Old English Riddles of the Exeter Book, p. 312; cf. John D.
parole: Etudes sur I 'ln d e cla ssigu e, pp. 6 1-9 1. Niles, Old English Enigmatic Poems and the Play of the Texts, pp. 89-92.

17. G o p a tha -B rah m a n a, 1.31. Cited in Malamoud, I.e ju m e a u so la ire, p. 142. 34. Patrick J. Murphy, Unriddling the Exeter Riddles, p. 65. On the use of runes

18. Kumarila, T a n tra v a rtik k a ad J a i m i m - S u d r a , 1.3.7. Cited in Malamoud, le in the Exeter riddles, see also Dieter Bitterli, Say What / Am Called: The Old English

jum eau solaire, p. 142. Riddles of the Exeter Book and the Anglo-Latin Riddle Tradition, esp. pp. 84-97.

19. Malamoud, F em in ite d e la p a r o le , p. 63. 35. Ludwig Traube, Nomina sacra: Versuch einer Geschichte der christlichen

20. Strabo 15.1.53; English in T h e G e o g r a p h y o f S tra b o , vol. 7, pp. 87-89. Kurzung.

21. See the analytic commentary on Megasthenes apud Strabo in Harry Falk, 36. Ibid., p. 24.
37. W. M. Lindsay, “ Traube’s Nomina Sacra and Posthumous Works,” p. 133.
Schrift im a lten In d ie n : E in F o rsch u n g sberic h t m il A n m e rk u n g c n , pp. 291-93.
38. Traube, Nomina sacra, p. 33. The manuscript is Primasius, Oxford Douce
history of writing in ancient India more generally, see Gerard Fussman, “ Histoire
140.
du monde indien: Les premiers systemes d ’ ecriture en Inde." On the status of

201
2 0 0
 DARK T O N G U E S NOTES

39. Ibid., p. 284. n . Ibid., p. 51.

40. Antoine Meillet, “ Quelques hypotheses sur des interdictions de vocabu- 12. Ibid., p. 52.
laire dans les langues indo-europeennes,” in Linguistique historique et linguistique 13. Ferdinand de Saussure, Ms. 3962.7.20 v, reproduced in Prosdocimi and
generale, p. 281. Marinetti, “ Saussure e il saturnio,” p. 56.
41. Ibid. 14. For Saussure’s self-diagnosed “graphophobia,” see Paola Villani, “Docu-
42. Emile Benveniste, “ La blasphemie et Teuphemie,” in Problemes de linguisique menti saussuriani conservati a Lipsia e a Berlino,” p. 26. Elsewhere, Saussure
generale, vol. 2, p. 257. The article first appeared in 1969. Cf. his “ Euphemismes evokes his “ horror of writing” (ibid., p. 24).
anciens et modernes," in Problemes de linguistique generate, vol. 1, pp. 308-14. 15. The Course now exists in several forms. For a critical edition of the 1916
43. Benveniste, “ La blasphemie et l ’euphemie,” p. 257. text, see Ferdinand de Saussure, Cours de linguistique generate. On the textual
44. H.L. Mencken, “American Profanity,” p. 243. The examples (ibid., pp. composition of the Course, see Robert Godel, Les sources manuscrites du Cours de
242-43) are drawn from M .L. Walter’s Dictionary o f Profanity and Its Substitutes, linguistique generate de F. de Saussure. Cf. the critical edition by Robert Engler,
which Mencken consulted in a special collection o f the Princeton University Saussure, Edition critique du Cours de linguistique generate de Ferdinand de Saussure,
Library. with the supplementary view provided by Eisuke Komatsu and George Wolf (eds.),
in Saussure, Saussure’s First Course of Lectures in General Linguistics; Saussure’s Second
CHAPTER NINE: A N A P H O N ES Course of Lectures in General Linguistics; and Saussure’s Third Course of Lectures in
1. Ferdinand de Saussure, Letter of January 10, 1906, to Antoine Meillet, in General Linguistics.
Emile Benveniste, “ Lettres de Ferdinand de Saussure a Antoine Meillet,” p. 105. 16. Jean Starobinski, Les mots sous les mots: Les anagrammes de Ferdinand Saussure.
2. Ferdinand de Saussure, Letter of January 23, 1906, to Antoine Meillet in Robert Godel was the first to evoke 99 anagram notebooks: see his “Inventaire des
Benveniste, “ Lettres de Ferdinand de Saussure,” p. 106. manuscrits de F. de Saussure remis a la Bibliotheque publique et universitaire de
3. For more recent work on the Saturnian, see Thomas Cole, “ The Saturnian Geneve,” p. 11. Michel Dupuis, in his account of the Geneva manuscripts, puts the
Verse”; and Jed Parsons, “A New Approach to the Saturnian Verse and Its Relation number of anagram notebooks at 117: Dupuis, "A propos des anagrammes saussuri-
to Latin Prosody.” ennes,” pp. 11-12. W. Terrence Gordon and Henry G. Schogt count 134notebooks:

4. Horace, Epistles 2.1.157. “ Ferdinand de Saussure: The Anagrams and the Cours,” p. 140.

5. For a summary of the debate, see Frank Frost Abbott, “ The Accent in Vulgar 17. William Camden, Remaines concerning Britaine: But especially England, and

and Formal Latin.” the Inhabitants thereof p. 168.

6. Saussure, Letter of September 23, 1907 to M eillet, in Benveniste, "Lettres 18. See Christine Luz, Technopaignia: Formspiele in der griechischen Dichtung,

de Ferdinand de Saussure,” p. 109. pp. 147-78-

19. Ibid., p. 150.
7. Peter Wunderli, “ Ferdinand de Saussure, ter cahier a lire preliminairement:
20. Ferdinand de Saussure, Letter of July 14, 1906, in Starobinski, Les mots
Ein Basistext seiner Anagrammstudien,” p. 204.
sous les mots, p. 21.
8. Ferdinand de Saussure, Letter of July 5, 1906, in Aldo L. Prosdocimi and
21. Ibid.
Anna Marinetti, “ Saussure e il saturnio: Tra scienza, biografia e storiografia,’ p.
22. Ibid. p. 22.
43- 23. Ferdinand de Saussure, Lettter of September 23, 1907, in Benveniste,
9. Ibid.
“ Lettres de Ferdinand de Saussure,” p. 110.
10. Ferdinand de Saussure, Letter of July 17 ,19 0 6 , in Prosdocimi and Marinen-
24. Ibid. Livius Andronicus 18.1-2.
etti, “ Saussure e il saturnio,” pp. 4 5-4 6 .

203
2 0 2
 d a r k t o n g u e s
NOTES

25. Saussure, Letter of September 23, 1907, in Benveniste, “ Lettres de Ferdi­ 47. S taro b in sk i, L e s m o t s so u s le s m ots, pp. 69 and 78; see also the manu­
nand de Saussure,” pp. 110-11. script ( m s .fr . 39 6 4 ) cited in W underli, F e r d in a n d de Saussure und die Anagramme,

26. Saussure, Letter of July 14,1906, in Starobinski, Les mots sous les mots, p. 21. p. 5 1. G andon ( D e d a n g e r e u x e d ific e s , p. 16) holds that “cryogram” is synonymous
27. Saussure, Letter of September 23, 1907, in Benveniste, “ Lettres de Ferdi­ w ith “ anagram .”
nand de Saussure,” p. 111. 48. Ferdinand de Saussure, Letter of August 7,19 06 , “ Correspoodance Bally-
28. Saussure, Letter of July 17, 1906, in Prosdocimi and Marinetti, “Saussure Saussure,” C ahiers Ferdinand de Saussure 48 (1994), p. 115.
e il saturnio,” pp. 52-53. 4 9. Gandon, D e d a n g e r e u x e d ific e s , p. 116.
29. Starobinski, Les mots sous les mots, pp. 134 and 133. 50 . Starobinski, L e s m o ts so u s le s m ots, pp. 100-107.
30. Ibid., p. 36. 51. W underli, “ Ferdinand de Saussure, l e r c a h ie r a lir e p r e lim in a ir e m e n t ," p. 212.
31. Ferdinand de Saussure, Letter to Meillet of September 27, 1907, in Ben­ 52. Starobinski, L e s m o ts so u s le s m ots, p. 60.
veniste, “Lettres de Ferdinand de Saussure,” p. 114. 53. I b id ., p. 125.
32. On metathesis, see Peter Wunderli, Ferdinand de Saussure und die Ana- 54. Wunderli, F e r d i n a n d d e S a u ss u re u n d d ie A nagram m e, p. 15.
gramme: Linguistik und Literatur, p. 38. 55. Starobinski, L e s m o ts so u s le s m ots, p. 59.
33. Starobinski, Les mots sous les mots, p. 47. 56 . O n the distinction betw een the synchronic (or “ idiosynchronic”) and the
34. Ibid., p. 47. “ diach ron ic,” see Saussure, C o r n s d e lin g u is t iq u e g e n e ra le , pp. 114-40.
35. This diagram is drawn from Peter Wunderli, “ Saussure’s Anagrams and the 57. Starobinski, L e s m o ts so u s le s m ots, p. 133.
Analysis of Literary Texts,” p. 177. 58. On the status of the sign in the C ou rse, see, among others, Jean-Claude
36. Saussure, Letter of September 23, 1907, in Benveniste, “ Lettres de Ferdi­ Milner, L e p e r i p l e s t r u c t u r a l: F ig u r e s et p a ra d ig m e , pp. 15-44; Franfoise Gadet, Sau­
nand de Saussure,” p. 110. ss u re : U n e s c ie n c e d e la la n g u e , pp. 29-48; Simon Bouquet, Introduction a la lecture

37. See Wunderli, Ferdinand de Saussure und die Anagramme, pp. 20-21. de F e r d in a n d d e S a u ssu re , pp. 27 9 -9 1. Cf. the materials collected in Ferdinand de
38. Starobinski, Les mots sous les mots, p. 50. Saussure, E c r it s d e lin g u i s t i q u e g e n e r a le and Saussure, F erd in a n d de Saussure: Linguis­
39. See ibid., pp. 50-55. t ik u n d S e m io lo g ie : N o t iz e n a u s d em N a c h la f , Texte B riefe und Dokumente.

40. Ibid., p. 51. 59. Jean -C lau d e M ilner, L ’a m o u r d e la la n g u e , pp. 85-97.

41. Saussure, Letter of September 23, 1907, in Benveniste, “ Lettres de Ferdi­ 60. I b id ., p. 89.
nand de Saussure,” p. 113. 61. O n Saussure’s principle of linearity, see Wunderli, Ferdinand de Saussure und
42. Starobinski, Les mots sous les mots, p. 37. On Saussure’s notebooks on Vedic d ie A n a g r a m m e , pp. 17 8 -8 5 ; and W underli, “ Zur Geltung des Linearitatsprinzips
hymns, see David Shepheard, “ Saussure’s Vedic Anagrams,” pp. 513-23. bei Saussure.” T he principle has been questioned. For a critical appraisal, see Jean-
43. See Starobinski, Les mots sous les mots, pp. 79-10 0 ; on Saussure’s work on Claude M iln er, I n t r o d u c t io n a u n e sc ien ce du la n g a g e , pp. 385-95.
Lucretius, see Francis Gandon, De dangereux edifices: Saussure lecteur de Lucrece, Les 62. Saussure, C o r n s d e lin g u is t iq u e g e n e ra le , p. 103.
cahiers d ’anagrammes consacres au De rerum natura. 63. Ib id ., p. 170.
44. Starobinski, Les mots sous les mots, p. 27. 64. Starobinski, L e s m o ts so u s le s m ots, p. 47.
65. Saussure, L etter o f September 23, 1907, in Benveniste, “ Lettres de Ferdi­
45. On the hypogram, see Wunderli, Ferdinand de Saussure u n d d ie Anagramme,
nand de Saussure,” p. 113; Letter to Bally, July 17,190 6 , in Prosdocimi and Mari­
pp. 46-51.
netti, “ Saussure e il saturnio,” p. 53.
46. For Saussure’s terms, see Wunderli, Ferdinand de Saussure und die A n a ­
66. Ferdinand de Saussure Letter to Meillet, January 8, 1908, in Benveniste,
gramme, pp. 42-54; cf. Gandon’s glossary in De sangereux edifices, pp. 381-93.

205
204
 DARK T O N G U E S
NOTE*

“ Lettres de Ferdinand de Saussure,” p. 118. 114-28; and Joshua T. Katz, “ Wordplay” and “ Saussure* Anaphonir."
&7. See Gandon, De dangereux edifices, p. 18. 84. Jean Starobinski, “ Lettres et syllables mobiles: Complement a la tenure
hS. For the “ furor of phonic play,” see Saussure, Letter to Meillet, January 8, des C a h ie r s d’anagrammes de Ferdinand de Saussure,” p. 8.
mo$, in Benveniste, “ Lettres de Ferdinand de Saussure,” p. 118. 85. Starobinski, L e s m o ts so u s le s m ots, p. 135.
(>9. See Starobinski, Les mots sous les mots, pp. 14 6 -4 8 . Saussure mistakenly 86. Ibid.
believed Thompson lived within a generation of his own time. 87. See ibid., p. 133.
70. Milner, L’amour de la langue, p. 88. 88. Ibid., p. 135.
71. See Starobinski, Les mots sous les mots, pp. 115-20 .
72. Ibid., p. 154. For the correspondence with Pascoli, see also Giuseppe Nava, CHAPTER TEN; PATTERNS
“Lettres de Ferdinand de Saussure a Giovanni Pascoli.” 1. Robert Godel, “ Inventaire des manuscrits de F. de Saussure remis a la Bib
73. Starobinski, Les mots sous les mots, p. 151. liotheque publique et universitaire de Geneve,” p. n.
74. Ibid., pp. 133 and 128. 2. Jean Starobinski, “ Les anagrammes de Ferdinand de Saussure.”
75. Saussure alludes at several points to a calculus of probabilities: see Staro­ 3. Jean Starobinski, “ Les mots sous les mots.”
binski, Les mots sous les mots, pp. 5 0 ,12 8 -32 ,133 ; and Saussure, Letter of September 4. Roman Jakobson, “ La premiere Lettre de Ferdinand de Saussure sur les
23,1907, in Benveniste, “ Lettres de Ferdinand de Saussure,” p. 112. anagrammes.”
76. “The Two Saussures” was the title of a special issue of Recherches devoted 5. Ibid., p. 21.
to the anagram notebooks in 1975, edited by Sylvere Lotringer. See his introduc­ 6. Roman Jakobson, “ Retrospect,” in S e le c te d W ritings, vol. 4, p. 685.
tion, “Flagrant delire.” 7. Jakobson, “ La premiere Lettre de Ferdinand de Saussure,” p. 21.
77. Starobinski, Les mots sous les mots, p. 154. 8 . I b id .

78. For “madness,” see Michel Deguy, “ La folie de Saussure”; for “ dementia,” 9. Ibid. Jakobson’s reference is to Starobinski, “Les anagrammes de Ferdinand
see Jean-Jacques Lecercle, The Violence of Language, p. 30; for “schizophrenia,” see de Saussure,” p. 256.

Luce Iragaray, “ Le schizophrene et la question du signe.” Recently, Gandon (De 10. Jakobson, “ La premiere Lettre de Ferdinand de Saussure,” p. 22.

dangereux edifices, p. 69) has evoked “obsession.” 11. See Starobinski, “ Le mots sous les mots,” p. 1907.

79. For an annotated bibliography, see Ivan Callus, “A Chronological and 12. Roman Jakobson, “ Retrospect,” in S e le c te d W ritings, vol. 5, p. 570.

Annotated Bibliography of Works Referring to Ferdinand de Saussures Anagram 13. See Jakobson’s account in his retrospective “Vers une science de l’art poe-

Notebooks,” Cahiers Ferdinand de Saussure 55 (2002), pp. 269-95. tique,” in Selected Writings, vol. 5, pp. 241-44.

80. John E. Joseph, Saussure, p. 488. For a different account of the importance 14. Roman Jakobson, “What is Poetry?,” in S elected Writings, vol. 3, p. 750.
15. Roman Jakobson, “ Linguistics and Poetics,” first published in Thomas A.
of the anagram research in the development of Saussures linguistic theory, see
Sebeok (ed.), S ty le a n d L a n g u a g e ; it appears in S elected W ritings, vol. 3, pp. 18-51.
Bouquet, Introduction a une lecture de Ferdinand de Saussure, pp. 79-80 n. 5.
16. See i b i d . , pp. 21-22.
81. Roman Jakobson and Linda Waugh, assisted by Martha Taylor, The Sound
17. Ibid., p. 24.
Shape of Language, p. 221.
18. Ibid., p. 25.
82. Elizarenkova, Language and Style o f the Vedic R$is, p. 124.
19. Ibid., p. 27.
83. See in particular, Vladimir Toporov, “ Die indoeuropaische Poetik und
20. See Ferdinand de Saussure, C ou rs de lin g u istiq u e g en era le, pt. 2, ch. 5, pp.
ihre Urspriinge,” pp. 189-251; Fran^oise Bader, Anagrammes et alliterations; Ranko
170-76.
Matasovicf, A Theory o f Textual Reconstruction in Indo-European Linguistics, pp.

207
206
 DARK TO N G U E S
NOTES

21. Ibid. p. 171.

41. Jak obson , “ L in gu istics and Poetics,” in Selected Writings, vol. 3, p. 43. Cf.
22. Jakobson, “ Linguistics and Poetics,” in Selected Writings, vol. 3, p. 27.
the m ore exten d ed discussion in Jakobson, “Language in Operation,” ibid., pp.
23. Ibid.
7-17.
24. Ibid.
42. Poe, “ The Philosophy of Composition,” Essays and Reviews, p. 18.
25. Edgar Allan Poe, “ The Rationale of Verse,” in Essays and Reviews, p. 33.
43. Roman Jakobson, “ Subliminal Verbal Patterning in Poetry,” in Selected
26. Jakobson, “ Linguistics and Poetics,” in Selected Writings, vol. 3, p. 39.
Writings, vol. 3, p. 138.
27. Gerard Manley Hopkins, Journals and Papers, p. 85, cited by Jakobson in
44. Roman See Jakobson, “ Une Microscopie du dernier ‘Spleen’ dans Lesjleurs
“ Linguistics and Poetics,” Selected Writings, vol. 3, p. 39. du Mai,” in Selected Writings, vol. 3, pp. 465-81.
28. Letter of July 14 ,19 0 6 , in Jean Starobinski, Les mots sous les mots, p. 21. 45. Roman Jakobson, “ Shakespeare’s Verbal Art,” in Selected Writings, vol. 3,
29. Roman Jakobson and Linda Waugh, The Sound Shape o f Language, p. 221, p.300.
The reference is to Starobinski, Les mots sous les mots, pp. 21-23 and 33. 46. See, for example, Jakobson and Levi-Strauss’s suggestion that the unvoiced
30. Perhaps for this reason, Jakobson could go so far as to suggest that arith­ fricative /ch/ plays a remarkable role in its near absence from Baudelaire’s son­
metical regularities could, in themselves, play a functional role in verse. See, for net: “Les Chats de Charles Baudelaire,” L’Homme 2.1 (1962), pp. 5-21, in Selected
example, Roman Jakobson, “ Shakespeare’s Verbal Art in T h ’ Expense of Spirit,” in
Writings, vol. 3, pp. 4 4 7-6 4 . Cf. Jakobson, “ ‘Si Nostre Vie: Observations sur la
Selected Writings, vol. 3, pp. 284-303. O f course, if one reasons in modern terras,
composition structure de motz dans un sonnet de Joachim du Bellay,” in Selected
incommensurable quantities may also be considered to fall within the domain of
Writings, vol. 3, pp. 239-74, esp. pp. 266-68.
arithmetic. Yet Jakobson appears not to have considered the possibility that quanti­
47. Roman Jakobson, with S. Rudy, “Yeats’ ‘Sorrow for Love’ through the
ties of this kind could play a role in verse.
Years,” in Selected Writings, vol. 3, pp. 613-14.
31. Starobinski, Les mots sous les mots, p. 47.
48. See Roman Jakobson, “ The Grammatical Texture of a Sonnet from Sir
32. Roman Jakobson, Six Lectures on Sound and .Meaning, in Selected Writings,
Philip Sidney’s Arcadia,” in Selected Writings, vol. 3, pp. 275-83, esp. p. 278.
vol. 8 , pp. 378-79- 49. Roman Jakobson, “ Poetry of Grammar and Grammar of Poetry,” in
33. Saussure, Corns de linguistique generale, pt. 2, ch. 5, p. 170. Selected Writings, vol. 3, p. 92.
34. See, Jakobson, Six Lectures on Sound and Meaning, in Selected Writings, vol. 50. I b id ., pp. 92-93.
8, pp. 378-79; for a later presentation, see The Sound Shape o f Language, pp. 122-76. 51. “ La grammaire, l ’aride grammaire elle-meme, devient quelque chose
35. In this sense, Jakobson’s theory o f the phoneme constitutes an example comme une sorcellerie evocatoire; les mots ressuscitent revetus de chair et d’os,
of the structuralist procedure. It may be even its clearest illustration. See Jean- le substantif, dans sa majeste substantielle, l’adjectif, vetement transparent qui
Claude Milner, Le periple structural: Figures et paradigme. l’habille et le colore comme un glacis, et le verbe, ange du mouvement, qui donne
36. Jakobson, Six Lectures on Sound and Meaning, in Selected Writings, vol. 8, le branle a la phrase.” Charles Baudelaire, CEuvres completes, vol. 1, p. 431; cited in
p.380. Roman Jakobson, “ Retrospect,” in Selected Writings, vol. 3, p. 769.
37. For a summary presentation, see John J. McCarthy, “ Nonlinear Phonology. 52. Poe, “ The Rationale of Verse,” Essays and Reviews, p. 33.
38. Jakobson, “ Language in Operation,” in Selected Writings, vol. 3, p. 11. Jako­ 53. Jakobson, “ Subliminal Verbal Patterning in Poetry,” in Selected Writings,
bson refers to Roland G. Kent, “Assimilation and Dissimilation." vol. 3, p. 147.
39. Jakobson, “ Retrospect,” in Selected Writings, vol. 3, p. 775. 54. Roman Jakobson, “ On the So-Called Vowel Alliteration in Germanic
40. Edgar Allan Poe, “ The Philosophy of Composition,” in Essays and Reviews, Verse,” in Selected Writings, vol. 5, p. 195. Jakobson cites E. H. Sturtevant: “we may
take it for granted that an obligatory feature of versification must be in some way

2 0 8
209
 DARK T O N G U E S
NOTES

audible.” See E.H. Sturtevant, “ The Doctrine o f Caesura, a Philological Ghost,” 8. Jean Dufournet, “ Tzara et les anagrammes de Villon"; cf. the liter version
P- of the essay by the same title that appears in Dufournet, Nouvellet recbercba sur
55. Paul Kiparsky has suggested that linguistics must go further, recognizing Villon, pp. 249-73.
“that even linguistic features which must remain latent play a role in verse.” Kip­ 9. Tristan Tzara, “ La signification des anagrammes.”
arsky, “ Roman Jakobson and the Grammar of Poetry,” p. 34. 10. Behar, “A mots decouverts,” p. 96.
56. Jakobson, “ Retrospect,” in Selected Writings, vol. 3, p. 769. 11. Tristan Tzara, “ L’Actualite de Villon,” in Oeuvres completes, vol. 5, pp.
57. Jakobson, “ Subliminal Verbal Patterning in Poetry,” in Selected Writings, 115-24, p. 119.
vol. 3, p. 142. 12. Ibid.
58. See, among others, Tzvetan Todorov, “ Roman Jakobson poeticien”; Leo 13. Ibid.
Bersani, “ Is There a Science of Literature?” ; Jam es A. Boon, From Symbolism to 14. Tristan Tzara, Le Secret de Villon, in Oeuvres completes, vol. 6, p. 9.
Structuralism: Levi-Strauss in a Literary Tradition, pp. 3 8 -6 1; Jonathan Culler, Struc­ 15. Ibid. The Middle French text cited here appears in the format given by
turalist Poetics: Structuralism, Linguistics, and the Study o f Literature, pp. 55-74; Paul Tzara. For a more recent edition of Villon’s text, see Francois Villon, Poesies com­
Werth, “ Roman Jakobson’s Verbal Analysis o f Poetry” ; Linda R. Waugh, “The pletes, p. 107.
Poetic Function in the Theory of Roman Jakobson” ; and Jean-Claude Milner, 16. Francois Villon, The Testament, in The Poems of Francois Villon, p. 38.
“A Roman Jakobson, ou le bonheur par la sym etrie,” in Le periple structural, pp. 17. Tzara, Le secret de Villon, in CEuvres completes, vol. 6, p. jo.
131-40. 18. See, most recently, Claude Thiry’s remarks in Villon, Poesies completes, p.
59. Culler, Structuralist Poetics, p. 57. 106.
60. Ibid., p. 58. 19. See Villon, CEuvres de Villon, p. 158. Before being noted by Foulet, they
61. Michel Riffaterre, “ Describing Poetic Structures: Two Approaches to were already observed by Jean Acher in a contribution that Foulet most likely read:
Baudelaire’s Les Chats,” p. 202. Acher, review of Lognon, p. 22. See Behar’s notes, CEuvres completes, vol. VI, p. 539.
62. Kiparsky, “ Roman Jakobson and the Grammar o f Poetry,” p. 33. The passage appears as folio 124.
63. Riffaterre, “ Describing Poetic Structures,” p. 215. 20. Francois Villon, Lais 81 (stanza XI), in Poesies completes, p. 67; Testament
64. The term “ linguistic and poetic actualizations” is Riffaterre’s; see ibid. 970 (stanza XCIV), p. 169.
p. 213. 21. Tzara, CEuvres completes, vol. 6, p. 539.
65. Jakobson, “ Linguistics and Poetics,” in Selected Writings, vol. 3, p. 50. 22. Tzara, Le secret de Villon, in CEuvres completes, vol. 6, p. 10.

66. Ibid. 23. Le Gentil, Villon, p. 22.

24. Tzara, Le secret de Villon, in CEuvres completes, vol. 6, pp. 10-11.
25. Stephane Mallarme, “ Crise de vers,” in CEuvres complete, p. 368.
CHAPTER ELEVEN: SECRETS OF T R IS T A N TZARA
26. Tzara, Le secret de Villon, in CEuvres completes, vol. 6, p. 11.
1. Jean Couvreur, “ Du nouveau sur Francois Villon?,” p. 9.
27. Ibid., p. 12.
2. Ibid.
28. See Dufournet, Nouvelles recherches sur Villon, p. 251.
3. C h a r le s D o b z y n s k i , “ L e S e c r e t d e F r a n £ o i s V i l l o n , ” p p . 1 - 4 *
29. Tzara refers (Le secret de Villon, in CEeuvres completes, vol. 6, pp. 292-95)
4. Ibid., p. 1. to Louis Thuasne’s “ La rime chez Villon,” in Thuasne, Villon et Rabelais: Notes et
5. Ibid. commentaires, pp. 369-418.
6. Henri Behar, “A mots decouverts,” p. 96. 30. See Tzara, Le secret de Villon, in CEuvres completes, vol. 6, pp. 13-16. Cf.
7. Pierre Le Gentil, Villon, pp. 21-30 .
211

2 1 0
 d a r k t o n g u e s NOTES

Dufburnet, N'ouvclks recherches sur Villon, p. 251. 58. Ibid., p. 220.

31. Tzara, le secret de Villon, in (Euvres completes, vol. 6, p. 16. 59. See Harrett Ann Watts, Chance: A Perspective on Dada.

32. Ibid,, p. u . 60. Tristan Tzara, “ Dada manifeste sur l'amour faible et lamour amer," in

$3. Louis Aragon, “ Tristan Tzara decouvre une oeuvre nouvelle de Rabelais,” (Euvres completes, vol. 1, p. 382.

p. 7 . 61. See Richter, Dada, Kunst und Anti-Kunst, pp. 51-65:; and Watt, Chance, pp.

34. Tzara, Le secret de Villon, in (Euvres completes, vol. 6, p. n . 5 1-13 6 .

35. Villon, Testament 9 6 6 -9 6 9 (stanza C X III), in V illon , Poesies completes, 62. Richter, Dada, Kunst und Anti-Kunst, p. 54.

p. 167; for Tzara’s analysis, see Le secret de Villon, in (Euvres completes, vol. 6, pp. 63. Tristan Tzara, (Euvres completes, vol. 6, p. 525.

278-82. 64. Jean-Loup Debardie, Candide, July 26,1961.

36. Kinnell, The Poems of Francois Villon, p. 87. 65. M. Puissegur, “ Rabelais, Dada, et les probability,” p. 10.

37. For the name Charles d’Orleans, see Tzara, Le secret de Villon, in duvres 6 6 . Ibid., p. 23.

completes, vol. 6, p. 254; for the “ ironic” names, see ibid., pp. 2 78 -79 . 67. Tristan Tzara, (Euvres completes, vol. 6, p. 526.

38. See Tzara, Le secret de Villon, in (Euvres completes, vol. 6, pp. 280-81. 68. Ibid., p. 525.

39. Ibid., p. 282. 69. Ibid.

40. For Tzara’s “ simultaneous poems,” see Tzara, (Euvres completes, vol. 1, pp. jo . Ibid., p. 530.

492-500. 71. Lynn D. Stults, “A Study of Tristan Tzara’s Theory Concerning the Poetry

41. See the statement accompanying the poem, ibid., pp. 4 9 2-9 3. of Villon,” p. 433.

42. Hugo Ball, cited in Hans Richter, Dada, Kunst und Anti-Kunst: Der Beitrag 72. Ibid., p. 4 5 3 .

Dadas zur Kust des 20. Jahrhunderts, p. 28. 73. Ibid., pp. 4 5 3 - 5 4 .

43. Tzara uses the term “ superposition” at various points. See, for example, Le 7 4 . Ibid., pp. 445 -46.
secret de Villon, in (Euvres completes, vol. 6, p. 18. 75. Ibid., p. 451.

44. Ibid., p. 17. 76 . Ibid., p. 4 5 4 .

45. Ibid., p. 519. 77- Ibid., p. 45 5.

46. Ibid., p. 91. 78. Ibid., pp. 4 5 7 - 5 8 .

47. Ibid., p. 91. 79. Ibid., p. 45 8.

48. Ibid., p. 16. 80. Michel Bernard, “ Le Secret de Villon a 1’epreuve de l’ordinateur: Tzara et

49. Ibid., p. 89. les anagrammes.”

50. Ibid., p. 17. 81. Ibid., p. 245.

82. Ibid.
51. For a discussion, see Dufournet, Nouvelles recberches sur Villon, p. 264.
83. Ibid., pp. 2 4 5 - 4 6 .
52. Tzara, Le secret de Villon, in (Euvres completes, vol. 6, p. 215.
84. See ibid., p. 249.
53. Aragon, “ Tristan Tzara decouvre une oeuvre nouvelle de Rabelais,” p. 7 -
85. Ibid., p. 248.
54. Tristan Tzara, “ Manifeste Dada,” in (Euvres completes, vol. 1, p. 366.
86. Tzara, “ Dada manifeste sur l’amour faible et l’amour amer,” (Euvres com­
55. Ibid., 364-67.
pletes, vol. 1, p. 381.
56. Tristan Tzara, “ Conference sur Dada,” in (Euvres completes, vol. 1, p. 4 21,
87. Henri Behar, in Tzara, (Euvres completes, vol. 6, p. 550 n. 4.
57. Tzara, Le secret de Villon, in (Euvres completes, vol. 6, p. 219.

213
212
 , „ cre( J e K ite , in <£“" « c°mPlk e S ' TOl- 6 >P ' '7 -
$8. Tz^ra' *•e

S,‘ 'W" S!v°.uac .Andre Breton n’ecrira plus," Le Journal du Teapie, April ^
,J e e p i * ^ Breton, (Eurres s p ik e s , vol.., p. tt,5. Bibliography

Abbott, Frank Frost, "The Accent in Vulgar and Formal Latin,” Clamcel PkiU„0
2.4(1907), pp. 4 4 4 -6 0 .

Acher, Jean, review of Lognon, ( E u v r e s d e V illo n , Z e it s c h r ift fu r f r a z d s is c h e Sprache

u n d L it e r a tu r 38.2 (1911), pp. 18-25-.

Acker, Paul, and Carolyne Larrington (eds.), T h e P o e tic E d d a : Essays on O ld Norse

M y t h o lo g y (New York: Routledge, 2002).

Akehurst, F. R. P., and Judith M. Davis, A H a n d b o o k o f th e Troubadours (Berkeley:
University of California Press, 1995).
Amory, Frederic, “ Towards A Grammatical Classification of Kennings as Com­
pounds,” A r c h iv fo r n o r d is k filo lo g i 97 (1982), pp. 67-80.
Anglade, Joseph (ed.), L e s l e y s d ’a m o u r s , 4 vols. (Toulouse: E. Privat, 1919-1920).
Anonymous, T h e B o o k o f V a g a b o n d s a n d B e g g a r s : W ith a V ocabulary o f T heir L a n ­

guage, ed. Martin Luther, trans. James Camden Hotten (London: James Cam­
den Hotten, i860).
Aragon, Louis, “ Tristan Tzara decouvre une oeuvre nouvelle de Rabelais,” femes
fr a n ip a is e s 1000, October 24, 1963, p. 7.
Ashley, Leonard R. N., “ Rhyme and Reason: The Methods and Meanings of Cock

ney Rhyming Slang, N am es 25.3 (i9 7 7 )> PP-124 ~54-

Ave-Lallement, Friedrich Christian Benedict, D a s d e u tsc h e G a u n e r

sozialpolitischen, literarischen und linguistischen Ausbildung zu semem eutige

B esta n d e (Leipzig: Brodhaus, 1872). University Press,

Ayto, John, O x f o r d D i c t i o n a r y o f R h y m i n g S la n g (Oxfor

Bader, Fran5oise, L a l a n g u e d e s d i e u x o n 1 h e rm e tis m e d e s p '

Giardini, 1989).

21S
 DARK T O N G U E S
B I BLI OGRAPHY

, Anuqrumnkw (t alliterations (Louvain: Peeters, 1993).

Boutiere Jean, and Alexander Herman Schulz, Biographies des troubadours hues
Bartholonueis. V. IV , “ Du role et des origines de la tornade,” Annales du Midi 19 provengaux des Xllle et XlVe siecles (Paris: Nizet, 1964).
lw o 7V.pp. Breton, Andre, (Euvres completes, ed. Marguerite Bonnet et al., 4 vols. (Pan*;
Baudelaire, Charles, (Euvres completes, ed. Claude Pichois, 2 vols. (Paris: Galli- Gallimard-La Pleiade, 1988-2008).
mard'La Pleiade, 1975). Brik, Osip Maksimovich, Two Essays on Poetic Language, preface by Roman Jakobv»n
Behar, Henri, “A mots decouverts,” Europe 53-555-556 (1975)» PP- 95-H 2. (Ann Arbor: University of Michigan Press, 1964).
Bellows, Henrv Adams (ed. and trans.), The Poetic Edda (N ew York: American- Brodeur, Arthur G., “ The Meaning of Snorri’s Catgeories," University of Californio
Scandinavian Foundation, 1926). Publications in Modern Philology 36.4 (1952), pp. 1-13.
Beltrami, P.G., Rimario trobadorico provenzale, 2 vols. (Pisa: Pacini, 1988). Calder, George (ed.), Auraicept na N-Eces: The Scholar’s Primer, Being the Texts ofthe
Benveniste, Emile, “ Une differenciation de vocabulaire dans l ’Avesta,” Studia Indo- Ogham Tractfrom the Book of Ballymote and the Yellow BookofLecan, and the Text
lranica: Ehrengabefur Wilhelm Geiger (Leipzig: Harrassowitz, 1931), pp. 219-26. of the Trefhoculfrom the Book of Leinster (Edinburgh: J. Grant, 1917).
----- , “ Lettres de Ferdinand de Saussure a Antoine M eillet,” Cahiers Ferdinand de Callus, Ivan, “A Chronological and Annotated Bibliography of Works Referring to
Saussure 21 (1964), pp. 93-130. Ferdinand de Saussure’s Anagram Notebooks,” Cahiers Ferdinand de Saussure
----- ,Problemes de linguisique generale, 2 vols. (Paris: Gallim ard, 1966-1974). 55 (2002), pp. 269-95.
Bernard, Michel, uLe Secret de Villon a l’ epreuve de l ’ordinateur: Tzara et les ana- Camden, William, Remaines concerning Britaine: But especially England, and the
grammes,” Romania 113 (1992-1995), pp. 242-52.
Inhabitants thereof (London: Simon Waterson, 1614).
Bernart de Ventadorn, Bernart de Ventadorn: Seine Lieder mit Einleitung und Glossar,
Camporesi, Piero, II libro dei vagabondi, preface by Franco Cardini (Milan: Gar-
ed. Carl Appel (Halle: Niemeyer, 1915).
zani, 2003).
Bersani, Leo, “ Is There a Science of Literature?,” Partisan Review 39 (1972), pp.
Cerquiglini-Toulet, Jacqueline, “ Marot et Villon,” in Villon et ses lecteurs: Actes
535 - 63- du colloque international des 13-14 decembre 2000 (Paris: Champion, 2005), pp.
Bertran de Born, The Poems of the Troubadour Bertran de Born, ed. William Paden,
19-31-
Tilde Sankovitch, and Patricia H. Stablein (Berkeley: University of California Cervants, Miguel de, Saavedra, La Galatea, ed. Juan Bautista Avalle-Arce (Madrid:
Press, 1986).
Espasa-Calpe, 1985).
Bitterli, Dieter, Say What I Am Called: The Old English Riddles o f the Exeter Book and
Chambers, F. M., Proper blames in the Lyrics of the Troubadours (Chapel Hill: Univer­
the Anglo-Latin Riddle Tradition (Toronto: University of Toronto Press, 2009).
sity of North Carolina Press, 1971).
Blacking, John, “ The Social Value of Venda Riddles,” African Studies 20.1 (1961). Clay, Jenny, “ The Planktai and Moly: Divine Naming and Knowing in Homer,”
pp.1-32. Hermes 100.2 (1972), pp. 127-31.
Bode, Wilhelm, Die Kenningar in der angelsachsischen Dichtung (Darmstadt: Cleasby, Richard, An Icelandic-English Dictionary, subsequently revised, enlarged,
Zemin, 1886). and completed by Gudbrand Vigfusson, 2nd ed. (Oxford: Oxford University
Boon, James A., From Symbolism to Structuralism: Levi-Strauss in a Literary Tradition Press, 1957).
(New York: Blackwell, 1972). Coelho, Francisco Adolpho, Os ciganhos de Portugal, com um estudo sobre 0 calao
Borges, Jorge Luis, “ Las Kenningar,” in Historia de la eternidad (Buenos Aires: Viau (Lisbon: Imprensa nacional, 1892).
y Zona, 1936), pp. 35-56. Cohen, Marcel, “ Note sur l’argot,” Bulletin de la Societe Linguistique de Paris 21.67
Bouquet, Simon, Introduction a la lecture de Ferdinand de Saussure (Paris: Payot, (1919), pp. 132 - 4 7 -

1997 )- Cole, Thomas, “ The Saturnian Verse,” in Studies in Latin Poetry, ed. Thomas Cole

2 l6 217
 DARK T O N G U E S BI BLI OGRAPHY

and Christopher Dawson (Cambridge: Cambridge U niversity Press, 1969), D o b z y n s k i , C h a r l e s , “ L e S e c r e t d e F r a n c o is V illo n / ' Lettresfran$aites 8o j, Derem

PP- 3- 7 4 - ber 17 - 2 3 , 1 9 5 9 . PP- i - 4-

Coleman, Julie, A History of Cant and Slang Dictionaries, 4 vols. (Oxford University Dronke, Ursula, T he P oetic E d d a , 3 vols. (Oxford University Press, 1969-2011).
Press: 2004-2010). D u b o i s , M i c h e l , “ S u r u n p a s s a g e o b s c u r du Je u de saint Nicolas,” Romania 40 (1929)-

Cook, Eleanor, Enigmas and Riddles in Literature (Cambridge: Cambridge Univer­ pp. 256-58.
sity Press, 2006). Dufournet, Jean, “ Tzara et les anagrammes de Villon,” Europe 555-56 (1975), pp.
Couvreur, Jean, “ Du nouveau sur Francois Villon?” Le Monde, December 22,1959, 113-34 .
p.9. ------ , “ Tzara et les anagrammes de Villon,” Nouvelles recherches sur Villon (Paris:
Cropp, Glynnis M., Le vocabulaire des troubadours de Vepoque classique (Geneva: Champion, 1980), pp. 249-73.
Droz, 1975). Dumezil, Georges, “ La tradition druidique et 1’ecriture,” Revue de Tbistoire des

Culler, Jonathan, Structuralist Poetics: Structuralism, Linguistics, and the Study of re lig io n s 122 (1940), pp. 125-33.
Literature (Ithaca: Cornell University Press, 1975). Dupuis, Michel, “A propos des anagrammes saussuriennes,” Cahiers d'analyse tex-
Dante Alighieri, Opere minori, vol. 3, pt. 2: De vulgari eloquentia e Monorchia, ed. tu e lle 19 (1977). PP- 7-24-
Pier Vincenzo Mengaldo and Bruno Nardi (Milan: Ricciardi Editore, 1999). E i s m a n n , W o l f g a n g , a n d P e t e r G r z y b e k (e d s.), Semiotische Studien zum Rdtsel:

Davies, Anna Morpurgo, “ The Greek Notion of D ialect,” Verbum 10 (1987), pp. Sim p le Form s R eco n sid ered I I ( B o c h u m : N. B ro ck m ey e r, 1987).
7-28. Elizarenkova, Tat’jana Jakovlevna, Language and Style o f the Vedic R$is, ed. with an
Debardie, Jean-Loup, Candide, July 26, 1961. intro. Wendy Doniger (Albany: State University of New York Press, 1995).
Deguy, Michel, “ La folie de Saussure,” Critique 25 (1969), pp. 20-26. Estienne, Henri, A p o lo g ie p ou r Herodote, ed. Paul Ristelhuber, 2 vols. (Paris: Isidore
Delaplace, Denis, LeJargon des Coquillars a Dijon au milieu du XVe siecle selon Marcel Liseux, 1879).
Schwob (1892) (Paris: Classiques Gamier, 2011). Elliott, R. W. V., R u n es (Manchester: Manchester University Press, 1959).
Demsky, Aaron, “A Proto-Canaanite Abecedary Dating from the Period of the Evans, David, “ Riddling and the Structure of Context,” Journal of American Folklore
Judges and Its Implications for the History of the Alphabet,” Tel Aviv 4 (1977), 89.352 (1976), pp. 166-88.
pp.14-27. Eyvindr Finnsson Skaldaspillir, D ie Dichtung: Edition-Kommentar-Unteisuchungen,
Derolez, Rene, Runica manuscripta: The English Tradition (Brugge: De Tempel, ed. Arnulf Krause (Leverkusen: Literaturverlag Norden Mark Reinhardt,
1954 )- 1990).
Detienne, Marcel, and Gilbert Hamonic (eds.), with the participation of Georges Falk, Harry, S c h rift im a lten In d ien : Ein Forschungsbericbt mit Anmerkungen (Tubin­
Charachidze, Marcel Detienne, Gilbert Hamonic, Charles Malamoud, and gen: Gunter Narr, 1993).
Carlo Severi, La deesse parole: Quatrefigures de la langue des dieux (Paris: Flam- Felices, Jose Caldereon, “ Lengua de los dioses/ Lengua de los hombres,'" Faventia
marion, 1995). 4.1 (1982), pp. 5-33-
Diels, Hermann (ed.), Die Fragmente der Vorsokratiker: Griechisch und Deutsch, 6th Fidjestol, Bjarne, “ The Kenning System: An Attempt at a Linguistic Analysis,"
ed., edited by Walther Kranz, 3 vols. (Berlin: Weidmann, 1951). in S elected Papers, ed. Odd Einar Haugen and Else Mundal, trans. Peter Foote
DiNapoli, Robert, “ Odd Characters: Runes in Old English Poetry,” in Verbal (Odense University press, 1997), pp. 16-67.
Encounters: Anglo-Saxon and Old Norse Studiesfo r Roberta Frank, ed. Antonina Formisano, Luciano, La liric a (Bologna: II Mulino, 1990).
Harbus and Russell Poole (Toronto: University o f Toronto Press, 2005), pp. Frank, Istvan, R ep erto ire m etrique de la poesie des troubadours, 2 vols. (Paris: Cham­

146-61. pion, 1966).

219
218
 BIBLIOGRAPHY
DARK T O N G U E S

5 -11.
Frank, Roberta, Old Norse Court Poetry: The Drottkvcett Stanza (Ithaca: Cornell
Gonda, Jan, N o te s on N am es a n d the Name o f God in Ancient India (Am*terdam:
University Press, 1978).
North Holland, 1970).
----- , s.v. “ kenning," in Dictionary of the Middle Ages, ed. Joseph Strayer, 13 vols.
Gordon, E. V., A n In tro d u c tio n to O ld Norse (2nd rev. ed., A.R. Taylor (Oxford:
(New York: Scribner, 1982-1989).
Oxford University Press,1957).
Frege, Gottlob, Funktion, B e g n ff, B e d e u tu n g , ed. Gunther Patzig, 6th ed. (Gottin­
gen: Vandenhoeck und Ruprecht, 1986). Gordon, W. Terrence, and Henry G. Schogt, “Ferdinand de Saussure: The Ana­

Friedrich, Johannes, “ Gottersprache und Menschensprache im hethitischen grams and the C o u rs,” in The Emergence o f the Modern Language Sciences: Studies

Schrifttum,” in Sprachgeschichten u n d W o rth ild u n g : F e s t s c h r f t A . D ebru nn er on the T ran sitio n f r o m H istorical-C om parative to Structural Linguistics in Honour

(Bern: Francke, 1954), pp. 135-39- o f E . F. K . K o e rn e r, vol. 1, H istoriograph ical Perspectives, ed. Sheila Embleton,

Fussman, Gerard, “ Histoire du monde indien: Les premiers systemes d’ ecriture en John E. Joseph, and Hans-Joseph Niederehe (Philadelphia: John Benjamins,

Inde,” Annuaire du C ollege de France 89 (1988-1989), pp. 5 0 7 -14 . 1 9 9 9 ). PP- 139-50-

Gade, Kari Ellen, The Structure o f Old Norse Drottkvcett Poetry (Ithaca: Cornell Gourian, Gerard, L ’am our et la guerre: L ’oeuvre de Bertran de Born (Aix-en-Provence:

University Press, 1995). Universite de Provence, 1985).

----- , Poetryfrom the Kings’ Sagas, vol. 2, From c. 103s to 1300 (Turnhout: Brepols, Grimm, Jakob, D eu tsch e M yth ologie, 4th ed., 3 vols. (Berlin: F. Diimmler, 1875).

2009). Guiraud, Pierre, L ’a rg o t (Paris: Presses Universitaires de France, 1956).

Gadet, Fran^oise, Saussure: Une science de la langue (Paris: Presses universitaires — — , Le ja r g o n de V illon , ou Le g a i savoir de la Coquille (Paris, Gallimard, 1968).

de France, 1987). Giintert, Hermann, U ber die ahurischen und daevischen Ausdriicke im Awesta: Fine

Gandon, Francis, De dangereux edifices: Saussure lectern de Lucrece, Les cahiers sem asiologische S tu d ie (Heidelberg: C. Winter, 1914).

d ’anagrammes consacres au De rerum natura (Louvain: Peeters, 2002). , Von der Sp rach e d er G otter und Geister: Bedeutungsgeschichtliche Untersuchungen

Gardner, Thomas, “ The Old English Kenning: A Characteristic Feature of Ger­ z u r hom erischen u n d eddischen Gottersprache (Halle: Niemeyer, 1921).

manic Poetical Diction?,” Modern Philology 67.2 (1969), pp. 10 9 -17. Halle, Morris, “ Phonology in Generative Grammar,” Word 18 (1962), pp. 54-72.

Gatien-Arnoult, Adolphe Felix (ed.), Laflors del gay saber, estier dichas Las leys Halliday, M. A. K., “Anti-Languages,” in Language as a Social Semiotic: The Social

d’Amors, 3 vols. (Toulouse: J.-B. Paya, 1841). In te rp re ta tio n o f L a n g u a g e a n d M eaning (Baltimore: University Park Press,

Georges, Robert A., and Alan Dundes, “ Towards a Structural Definition of the 1978), pp. 164-82.
Harman, Thomas, C a veat or W arening f o r Common Cursetors (London: William
Riddle,” Journal of American Folklore 76.300 (1963), pp. 111- 18 .
Griffith, 1567).
Gerbstman, Aleksandr Iosivic, “ O zvukovom stroenii naradnoj zagadki,” Kusskij
Harries, Lyndon, “ The Riddle in Africa,” Jou rn a l o f American Folklore 84.334(1971),
folklor 11 (1968), pp. 185-97.
Gesenius, Wilhelm, Hebrdisches und chaldaisches Handworterbuch iiber das Ahe Testa­ PP- 377- 93-

Hesiod, T h eo g o n y, with prolegomena and commentary by M.L. West (Oxford:

ment, 2nd ed. (Leipzig: F. W. C Vogel, 1823).
Clarendon Press, 1966).
Godefroy, Frederic, Dictionnaire de 1’ancienne languefrangaise et tous ses dialectes du
Heubeck, Alfred, “ Die homerische Gottersprache,” Wiirzburger Jahrbiicherfiir die
IXe au XVesiecle, 10 vols. (Paris: F. Viewig, 1881-1902).
A ltertu m sw issen sch a tf 4.2 (1949-1950), pp. 197-219.
Godel, Robert Les sources manuscrites du Cours de linguistique generale de F. de Sau­
Heusler, Andreas, A ltgerm an isch e D ichtung, 2nd ed. (1923; Darmstadt: Hermann
ssure (Geneva: Droz, 1957).
Gentner, 1943).
----- , “Inventaire des manuscrits de F. de Saussure remis a la Bibliotheque pub-
Hoepffner, Ernest, “ Le Castiat du troubadour Peire Vidal,” in Melanges de philologie
lique et universitaire de Geneve,” Cahiers Ferdinand de Saussure 17 (i960), pp.

2 2 0
 BIBLIOGRAPHY
PARK lONUUI*

H d 'ln ^ r r tjftrts a M Antoine Thomas (Paris: Champion, >927), pp. 211-20. trans. W illiam Alexander McDevitte and W.S. Bohn(NewYork: Harpen,
Holland. via»\, “ keimings, Metaphors, and Semantic Formulae in Norse drott- 1870).

tnrtt,'* tdh* (or nonirvk filologi 120 (2005), pp. 123-47. Kalischer, Alfred Christlieb, Observationes in poesim romanensem proviciahbui in

iMUnder, lee M. ^trans.V The Poetic Edda (Austin: University of Texas Press, primis respectis (Berlin: Duemmler, 1866).
Katz, Joshua T., “ Saussure’s Anaphonie,” in The Other Senses: Antiquity beyond
Hvvmer. Hud, trans. A.T. Murray (Cambridge, MA: Harvard University Press, the Visual Paradigm, ed. Shane Butler and Alex Purves (Durham: Acumen,
14:4V forthcoming).

, iXfvssev, trans. A.T. Murray (Cambridge, MA: Harvard University Press, ------, “ Wordplay,” in Proceedings of the 20th Annual UCLA Indo-European Confer­
14:4V ence, ed. Stephanie W. Jamison, H. Craig Melchert, and Brent Vine (Bremen:
Hopkins, Gerard Manley, TheJournals and Papers, ed. Humphrey House, completed Hempen Verlag, 2009), pp. 7 9 -114 .
b> Graham Storey (Oxford: Oxford University Press, 1959). Kent, Roland G ., “Assimilation and Dissimilation,” Language 12.4 (1936), pp.
Houben, Jan E.M., “ Ritual Pragmatics of a Vedic Hymn: The ‘ Riddle Hymn’ and 245- 58.

the Pravargva Ritual,” Journal of the American Oriental Society 120.4 (2000), Kiparsky, Paul, “ Roman Jakobson and the Grammar of Poetry,” in A Tribute to
pp. 4 9 9 - 3 t>. Roman Jakobson, 1896-1982 (Berlin: Mouton, 1983), pp. 27-39.
Hultgart, Anders, “ The Wisdom Contest in the Vafprudnismal,” Analecta Septen- Kretschmer, Paul, review of Herann Giintert, Von der Sprache der Gotter und Geister,
tnonalia 65 (2009), pp. 531-39. Glotta 13 (1924), pp. 266-67.
Iragarav, Luce, “ Le schizophrene et la question du signe,” Recherches 16 (1975), Kuhn, Hans, Das Drottkvcett (Heidelberg: Winter, 1983).
pp. 31-42. Lacan, Jacques, L’Ethique de la psychanalyse, 1959-1960 (Paris: Seuil, 1986).
Jacobsohn, H., “Aeolische Doppelconsonanz: Zur Sprache und Verstechnik des Laroche, E., “ Hattie Deities and their Epithets,” Journal of Cuneiform Studies 1.3
Homerischen Epos,” Hermes 45.1 (1910), pp. 67-124 . (1947), pp. 187-216 .
Jakobson, Roman, “ La premiere Lettre de Ferdinand de Saussure sur les ana- Larrington, Carolyne, The Poetic Edda (Oxford: Oxford University Press, 1996).
grammes,” L’Homme 11.2 (1971), pp. 15-24. ------, “ Vafjjrudnismal and Grimnismal: Cosmic History, Cosmic Geography,” in
----- , Selected Writings, 8 vols. (The Hague: Mouton, 1962-1988). Acker and Larrington, (eds.), The Poetic Edda: Essays on Old Norse Mythology
Jakobson, Roman and Linda Waugh, assisted by Martha Taylor, The Sound Shape of (New York: Routledge, 2002).
Language (Bloomington: Indiana University Press, 1979). Lattimore, Richmond, “ Herodotus and the Names of Egyptian Gods,” Classical
Jauss, Hans-Robert, DieAlteritat und Modernitat der mittelalterichen Literatur: Gesa- Philology 34 .4 (1939). PP- 357-65-
mmelte Aufsdtze 1956-1976 (Munich: W. Fink, 1977). Law, Vivian, Wisdom, Authority and Grammar in the Seventh Century (Cambridge:
Jeanroy, Alfred, La poesie lyrique des troubadours, 2. vols. (Paris: Didier, 1934) Cambridge University Press, 1995).
Jonsson, Finnur, Degamle Eddadigte (Copenhagen: G. E. C. Gads Forlag, 1932). Lausberg, Heinrich, Handbuch der literarischen Rhetorik: Eine Grundlagung der Lit-
Joseph, John E., Limiting the Arbitrary: Linguistic Naturalism and Its Opposites in eraturwissenschaft, 2nd ed. (Munich: Max Hiiber, 1973).
Plato’s Cratylus and Modern Theories of Language (Amsterdam: John Benjamins, Lazzeroni, Romano, “ Lingua degli dei e lingua degli uomini,” Annali della Scuola
2000). Normale di Pisa 26 (1957), pp. 1-25.
------, Saussure (Oxford University Press, 2012). Lecercle, Jean-Jacques, The Violence o f Language (London: Routledge, 1990).
Julius Caesar, Commentaries on the Gallic and Civil Wars, with the Supplementary Leeuwen, J. van, “ Homerica XVI: De lingua deorum,” Mnemosyne n.s. 20 (1892),
Books Attributed to Hirtius, Including the Alexandrian, African and Spanish Wars, pp. 127-40.

222 223
 DARK TONGUES BIBLIOGRAPHY

Lefkow iti, Natalie J., “ Talking Backwards in French,” The French Review 63.2 ------ , review o f Hermann Giintert, Von der Sprache der Cotter und CeitUr, Bulletin

pp. 3*2-22. de la Societe linguistqiue de Paris 23.2 (1922), pp. 56-57.

l e dentil, Fierro, Villon (Paris: Hatier, 1967). Meissner, Rudolf, Die Kenningar der Skalden: Ein Beitrag z.ur skaldischen Poetik

Leighton, Albert C.. “ Secret Communication among the G reeks and Rom ans,” (Leipzig: Schroeder, 1921).

Technology and Culture 10.2 (1969), pp. *3 9 - 5 4 ~ ------ , “ Die Sprache der gotter, riesend und zwerge in den Alvissmal," Zeitschnft

lindow, John, “ Riddles, Kennings, and the Com plexity o f Skaldic Poetry,” Scan­ fu r deutsches Altertum und deutsche Literatur 61.2-3 (19 24 ). PP-128-40.

dinavian Studies 47,3 (197s), pp- 311 —27. Mencken, H. L., “American Profanity,” American Speech 19.4 (1944). pp. 241-49.

Lindsay, W .M ., “ Traube’s Nomina Sacra and Posthumous W orks,” Classical Quar­ Messenger, John C . , J r . , “ A n a n g P r o v e r b - R i d d l e s , ” Journal o f American Folklore

terly 3.2 (1909), pp. 132-36. 73,289 (i960), pp. 225-35.

Lohmann, Johannes, Musike und Logos: Aufsatze zur griechischen Philosophic und Michel, Francisque, Etudes de philologie comparee sur I’argot (Paris: Firmin Didot,

Musiktheorie, ed. Anastasios Giannaras (Stuttgart: M usikwissenschaftliche 1856).

Verlagsgesellschaft, 1970). Milicic, Vladimir, “ Subliminal Structures in Folklore: Anagrammatized Answers

Lotringer, Sylvere, “ Flagrant delire,” Recherches 16 (1975), pp. 7 _ 1 4 - to Riddles,” Slavic and East European Journal 26.1 (1982), pp. 63-76.
Luz, Christine, Technopaignia: Formspiele in der griechischen Dichtung (Leiden: E .J. Milner, Jean-Claude, L’amour de la langue (Paris: Seuil, 1978).

Brill, 2010). ------ , Introduction a une science du langage (Paris: Seuil, 1986).
Malamoud, Charles, Cuire le monde: Rite et pensee dans l ’lnde ancienne (Paris: La ------ , Le periple structural: Figures et paradigme (Paris: Seuil, 2002).
Decouverte, 1989). Mohnan, Karuvannur Puthanveettil, “ Lexical Phonology,” Ph.D. dissertation,
------ , Feminite de la parole: Etudes sur l ’lnde classique (Paris: Albin M ichel, 2005). Massachusetts Institute of Technology, 1981.
------ , Lejumeau solaire (Paris: Editions du Seuil, 2002). M olk, U lrich, Trobar clus, Tiobar leu: Studien zur Dichtungstheorie der Trobadors
Mallarme, Stephane, CEuvres completes, ed. Henri Mondor (Paris: Gallim ard, 1945). (Munich: Fink, 1968).
Marie de France, The Fables, trans. Mary Lou M artin (Birm ingham , A L: Summa Moberg, Lennart, “ The Languages of Alvissmal,” Saga-Book 18.4 (1973), PP-
Publications, 1984). 299-323.
Marold, Edith, Kenningkunst: Ein Beitrag zu einer Poetik der Skaldendichtung (Berlin: Moret, Andre, Anthologie du Minnesang (Paris: Aubier, 1949).
De Gruyter, 1983). Murphy, Patrick J., Unriddling the Exeter Riddles (Philadelphia: Pennsylvania Uni­
Marot, Clement, CEuvres poetiques, ed. Gerard Defaux, 2 vols. (Paris: Classiques versity Press, 2011).
Gamier, 1990-1993). Nava, Giuseppe, “ Lettres de Ferdinand de Saussure a Giovanni Pascoli,” Cahiers
Matasovic, Ranko, A Theory o f Textual Reconstruction in Indo-European Linguistics Ferdinand de Saussure 24 (1968), pp. 73-81.
(Fankfurt am Main: Peter Lang, 1996). Niceforo, Alfredo, IIgergo nei normali, nei degenerati, e nei crimianli (Turin, Fratelli
Maurer, D. W „ and Sidney J. Baker, “Australian Rhyming Argot in the American Bocca, 1897).
Underworld,” American Speech (1944), pp. 183-9 5. Nielsen, Karl Martin, “ Runen und Magie: Ein forschungsgeschichtliche Uber-
McCarthy, John J., s.v. “nonlinear phonology,” in International Encyclopedia o f the blick,” Fruhmittelalterliche Studien 19 (1985), pp. 75-97.
Social and Behavioral Sciences, ed. Neil J. Smelser and Paul B. Baltes (Oxford: Niles, John D., Old English Enigmatic Poems and the Play of the Texts (Turnhout:
Pergamon, 2001), pp. 11392-95. Brepols, 2006).
Meillet, Antoine, Linguistique historique et linguistique generale (Paris: Champion, Nykl, Alois Richard, Hispanio-Arabic Poetry and Its Relations with the Old Provencal
1921). Troubadours (Baltimore: n.p., 1946).

224 225
 BIBLIOGRAPHY
DARK T O N Q U k S

Niemeyer, 1925), pp. 646-60.

OtvKmL \nd\, TV Elder FdJiK A Book i f Viking Lore (London: Penguin Books,
Prosdocimi, Aldo L., and Anna Marinetti, “ Saussure e il saturnio: Tra uwn/t,
: vm»V
biografia e storiografia,” Cahiers Ferdinand de Saussure 44 (1990), pp. {7-71.
t>v(KW English Dictionary* 2nd ed., 20 vols. (Oxford: Oxford University Press,
Puissegur, M., “ Rabelais, Dada, et les probability,” Bulletin de TAssoaation des
professeurs de mathematigues de Yenseignement public 277 (1971), pp. 9-2}.
Pj^e. R .L. “ Anglo-Saxon Runes and Magic,” in Runes and Runic Inscriptions: Col­
Queneau, Raymond, Excercises de style (Paris: Gallimard, 1947).
lected Essays on Anglo-Saxon and Voking Runes, ed. David Parsons (Rochester:
------ Exercises in Style, trans. Barbara Wright (New York: New Directions, 1981).
The Boydell Press, 1995), pp. 105-25.
Rankin, James Walter, “A Study of the Kennings in Anglo-Saxon Poetry,” Journal
. An Introduction to English Runes, 2nd ed. (Suffolk: Boydell & Brewer, 1999).
Pasjis. Dan, ‘il sod hatum: Le-toldot ha-hidah ha-‘Ivrit be-Italyah uve-Holand (Jeru­ o f Germanic Philology 8.3 (1909), pp. 357-422 andJournal of Germanic Philology

salem: Y.L. Magnes, 1986). 9.1 (1910), pp. 49-84-

----- , "Towards a Theory of the Literary Riddle,” in Untying the Knot: On Riddles Renou, Ernest, and Lilian Silburn, “ Sur la notion de brahman,” Journal asiatigue

and Other Enigmatic Modes, ed. Galit Hasan-Rokem and David Shulman (New 237 ( 1 9 4 9 ), PP- 83-166.

York: Oxford University Press, 1996), pp. 81-108. Richard li Biaus, ed. Wendelin Foerster (Vienna: Alfred Holder, 1874).

Parsons, Jed, “A New Approach to the Saturnian Verse and Its Relation to Latin Richter, Hans, Dada, Kunst und Anti-Kunst: Der Beitrag Dadas zur Kust des 20. Jah-

Prosody,” Transactions of the American Philological Association 129 (1999), pp. rhunderts (Cologne: DuMont Schauberg, 1964).
Riffaterre, Michel, “ Describing Poetic Structures: Two Approaches to Baudelaire’s
H7 - 37 -
Pasero, Nicolo, ed., Poesie di Guglielmo IX (Modena: S.T.E.M., 1973). Les Chats,” Yale French Studies 36-37 (1966), pp. 200-42.

Pepicello, W.J., and Thomas A. Green, The Language o f Riddles: New Perspectives Rig Veda, The, ed. and trans. Wendy Doniger (Newr York: Penguin, 1981).

(Columbus: Ohio State University Press, 1984). Riquer, Martin de, Los trovadores: Historia literariay textos, 3 vols. (Barcelona:

Pillet, Alfred, “ Grundlagen, Aufgaben und Leistungen der Troubadours-Forsch- Planeta, 1975).

ung,” Zeitschriftfiir romanische Philologie 47 (1927), pp. 316 -38 , rpt. in Der R iq u e r , Isa b e l d e , “ Linhaure: C e n t a n s d ’ e t u d e s su r u n senhal,” Revue deslangues
provenzalische Minnesang: Ein Querschnitt durch die neuere Forschungsdiskussion romanes 96.1 (1992), pp. 41-67.

(Darmstadt: Wissenschaftliche Buchgesellschaft, 1967), pp. 20-65. R o c h e t t e , B r u n o , “ G r e c s e t l a t in s fa c e a u x la n g u e s e tran g e res: Contribution a

Plenat, Marc, “ Le javanais: Concurrence et haplologie,” Langages 25.101 (1991), l ’ e t u d e d e l a d i v e r s i t e l i n g u i s t i q u e d a n s l ’a n tiq u ite classiqu e,” Revue beige de

pp. 96-117. philologie et d ’histoire 73.1 (1995), pp. 5-16.

------, “Morphologie d’un ‘langage secret’: Le javanais de Queneau. Description R o n c a g l i a , A u r e l i o , “ T r o b a r c l u s : D is c u s s io n e a p e rt a ,” Cultura neolatina 39.1-2

dans un cadre autosegmental,” Cahiers de grammaire 6 (1983), pp. 150 -9 4 . (1969), pp. 5-55.

------, “ Morphologie du largonji des loucherbems,” Langages 20.78 (1985), pp. Ross, Margaret Clunines, Skaldskaparmdl: Snorri Sturluson’s ars poetica and Medieval

73- 9 5 -
Theories o f Language (Odense: Odense University Press, 1987).

Poe, Edgar Allan, Essays and Reviews, ed. Gary Richard Thompson (New York: Rotolo, Vincenzo, “ La communicazione linguistica fra alloglotti nell’antichita

Library of America, 1984). classica,” in Studi classici in onore di Quintino Cataudella, vol. 1 (Catania: Uni-

Pollock, Sheldon, The Language of the Gods in the World o f Men: Sanskrit, Culture versita di Catania, 1972), pp. 395-414.

and Power in Premodern India (Berkeley: University of California Press, 2006). Ruby, Pechon de, La vive genereuse des Mercelots, Gueuz et Boesmiens, contenans leur

Porzig, Walter, “ Das Ratsel im Rigveda: Ein Beitrag zum Kapitel Sondersprache,” fagon de vivre, subtilitez Scgergon, ed. Denis Delaplace (Paris: Champion, 2007).

in Gemanica: Eduard Sievers zum 75. Geburtstage 25. November 1925 (Halle: Max Sachau, Edward C., Alberuni’s India: An Account of the Religion, Philosophy, Literature,

227
226
 1

BIBLIOGRAPHY
OA**\ T O N B U f S

t 'htonoh'gv, Cmwomv, Customs, laws and Astrology o f India about Shepheard, David, “ Saussure’s Vedic Anagrams,” Modern language KtfUw jj y

{ p. ,v,\\ ; sols, omlon: Kogan Paul, Trench, Triihner & c o „ 1910). (1 9 8 2 ), pp. 513-23.

Sainean, l arare. Us sources de I'argot ancien, 2 vols (Paris: Champion, 1912). Shklovsky, Victor, “Art as Technique,” in Russian Formalist Criticism Lout Lssap,

Nantini, Ciiovanna, Rimario dei trovatori (Rome: Nuova Cultura, 2010). trans. and ed. Lee T. Lemon and Marion J. Reis (Lincoln: Univmitv

Saussure, Ferdinand de, Fcrits de linguistique generale, ed. Simon Bouquet and Nebraska Press, 1965), pp. 3-24.

Rudolf Fngler (Paris: Gallimard, 2002). ------ , “ Isskustvo, kak priyom,” Sborniki 2 (1917), pp. 3-14.
. Cours de linguistique generate, ed. Tullio de Mauro (Paris: Payot, 1974). Starobinski, Jean, “ Les anagrammes de Ferdinand de Saussure,” Mercure de lrante
, Fdition critique du cours de linguistique generale de Ferdinand de Saussure, ed. (1 9 6 4 ), pp. 2 4 3 - 6 2 .
Robert Fngler, 3 vols. (Wiesbaden: Harrassowitz, 19 6 7-19 74 ). ------ , “ Lettres et syllabes mobiles: Complement a la lecture des Coiners
-, Unguistik und Semiologie: Notizen aus dem Nachlafi, Texte Briefe und Doku- d ’anagrammes de Ferdinand de Saussure,” Litterature 99 (1995), pp. 7-18.
mente, ed. Johannes Fehr (Frankfurt am Main: Suhrkamp, 1997). ------ , Les mots sous les mots: Les anagrammes de Ferdinand Saussure (Paris: Gallimard,
----- , Saussure’s First Course of Lectures in General Linguistics, ed. Eisuke Komatsu 1971).
and George Wolf (London: Pergamon, 1996). ------ , “ Les mots sous les mots,” in To Honor Roman Jakobson: Essays on the Occa-
------, Saussure's Second Course o f Lectures in General Linguistics, ed. Eisuke and sion cfhis Seventieth Birthday, vol. 3 (The Hague: Mouton, 1967), pp. 1906-17,
George Wolf (Oxford: Pergamon, 1997). Steiner, Richard C ., “ The Two Sons of Neriah and the Two Editions of Jeremiah
------, Saussure’s Third Course o f Lectures in General Linguistics, ed. Eisuke Komatsu in the Light of Two Atbash Code-Words for Babylon,” Vetus Testamentum 46.1
and George Wolf (Oxford: Pergamon, 1993). (1 9 9 6 ), pp. 7 4 - 8 4 .
Savinel, Pierre, Poesies de Francois Villon, preface by Tristan Tzara (Paris: Audin, Sternbach, Ludw ik, Indian Riddles: A Forgotten Chapter in the History of Sanskrit
1 9 4 9 )- Literature (Hoshiapur: Vishveshvaranand Vedic Research Institute, 1975).
Scott, Charles T., Persian and Arabic Riddles: A Language-Centered Approach to Genre Strabo, Geography, trans. Horace Leonard Jones, 9 vols. (Cambridge, MA: Harvard
Definition (Bloomington: Indiana University Press, 1965). University Press, 1932).
Schone, Lucien, Le jargon et jobelin de Maistre Frangois Villon, suivi de Jargon au Stronski, Stanislaw (ed.), Le troubadour Folquet de Marseille (Cracow: Academie des
theatre (Paris: Alphonse Lemerre, 1888). Sciences, 1910).
Schultz-Gora, Oskar (ed.), Die provenzalische Dichterinnen: Biographien und Texte Stults, Lynn D., “ A Study of Tristan Tzara’s Theory Concerning the Poetry of Vil­
(Leipzig: Gustav Fock, 1888). lon,” Romania 96 (1975), pp. 433-58.
Schwob, Marcel, Etudes sur I ’argotfrangais (Paris: Editions Allia, 2004). Sturluson, Snorri, Edda, ed. and trans. Anthony Faulkes (London: Dent, 1987).
------ , Frangois Villon (Paris: Editions Allia, 2008). ------ , Edda: Hdtttal, ed. Anthony Faulkes (London: Viking Society for Northern
------ , (Eurres, 2 vols. (Paris: Mercure de France, 1921). Research, University College London, 2007).
Sebeok, Tomas A., Style in Language (Cambridge, M A: Technology Press o f the ------ , Edda: Skaldskaparmal, ed. Anthony Faulkes, 2 vols. (London: Viking Society
Massachusetts Institute of Technology, i960). for Northern Research, University College London, 1998).
See, Klaus von, Beatrice La Farge, Eve Picard, Ilona Priebe, and Katja Schulz, ------ , The Prose Edda, trans. Arthur Gilchrist Brodeur (New York: The American-
multiple volumes, Kommentar zu den Liedern der Edda (Heidelberg: C. Winter, Scandinavian Foundation, 1923).
1 9 9 7 -)- Sturtevant, E. H., “ The Doctrine of Caesura, a Philological Ghost,” AmericanJour­
Shapiro, Marianne, “ The Trobairitz and the Lim its o f C ourtly Love,” Signs 3.3 nal o f Philology 45 (1924), pp. 329-50.
(»978), pp. 560 -71. Suetonius, C. Tranquillus, The Lives of the Twelve Caesars, trans. Alexander

2 2 8 229
 BIBLIOQRAPH V
DARK TO N G U ES

Vallet, Eduardo, “ II senhal nella lirica trobadorica,” Studi testuali 5 (2003), pp.
Thomson, revised and corrected hv T. Forester (Cambridge, M A: Harvard
111- 5 9 .
Unhersitv Press,
------ , “ II senhal nella lirica trobadorica (con alcune note su Bel/ Bon Esper in Gau
, l \ ritae caeserum, ed. Maximiliam lhm (Leipzig: Teubner, 1908).
celm Faidit),” Rivista di studi testuali 6-7 (2004-2005), pp. 281-325.
Suter, Ann, “ l anguage of Cods and Language of Men: The Case of Paris / Alexan­
Villani, Paola, “ Documenti saussuriani conservati a Lipsia e a Berlino," Cahien
d r a ,“ lexis 7-8 ^1991). PP-13—25-
Ferdinand de Saussure 4 4 (1990), pp. 3-33.
Tacitus, Cornelius, The Complete Works o f Tacitus, trans. Alfred John Church and
Villon, Francois, CEuvres, ed. A. Lognon, revised by Lucien Foulet (Paris: G. Lres,
William Jackson Brodribb (New York: Modern Library, 1942).
1919).
----- . De Germania, ed. Henrv Furneaux (Oxford: Oxford University Press, 1894).
------ , The Poems o f Frangois Villon, trans. Galway Kinnell (Hanover: Houghton
Tavlor, Archer, English Riddlesfrom Oral Tradition (Berkeley: University of C ali­
M ifflin, 1965).
fornia Press, 1951).
------ , Poesies completes, ed. Claude Thiry (Paris: Librairie Generale fra^aise,
—— , “ The Riddle,” California Folklore Quarterly 2.2 (1943), pp. 12 9 -4 7.
1991).
Thompson, George, “ The Brahmodya and Vedic Discourse,” Journal o f the Ameri­
Virgilius Maro Grammaticus, Opera omnia, ed. Bengt Lofstedt (Leipzig: Teubner,
can Oriental Society 117.1 (1997), pp. 13-37.
2003).
Thuasne, Louis, Villon et Rabelais: Notes et commentaires (Paris: Fischbacher, 1911).
Tobler, Adolf, and Erhard Lommatzsch, continued by Hans Helmut Christmann, Vries, Jan de, Altnordisches etymologisches Worterbuch (Leiden: Brill, 1961).

Altfranzdsisches Worterbuch, 11 vols. (Wiesbaden: Franz Steiner, i960). Watkins, C alvert, How to K ill a Dragon: Aspects of Indo-European Poetics (Oxford:

Todorov, Tzvetan, Les genres du discours (Paris: Editions du Seuil, 1978). O xford University Press, 1995).

— —, “ Roman Jakobson poeticien,” Poetique 7 (1971), pp. 275-8 6. ------ , “ Language of Gods and Language of Men: Remarks on Some Indo-European

Tolkien, Christopher (trans.), Saga Heidreks konungs ins vitra: The Saga o f King M etalinguistic Traditions,” in Myth and Law Among the Indo-Europeans: Studies

Heidrek the Wise (London: Thomas Nelsons & Sons, i960). in Indo-European Comparative Mythology, ed. Jaan Puhvel (Berkeley: University

Toporov, Vladimir N., “ Die indoeuropaische Poetik und ihre U rspriinge,” Poetica of California Press, 1970), pp. 1-17.

13 (1981), pp. 181-251. Watts, H arriett Ann, Chance: A Perspective on Dada (Ann Arbor: UMI Research

Tory, Geoffroy, Champjleury, ou L’art et science de la proportions des lettres (Paris: Press, 1980).

Charles Bosse, 1931). Waugh, Linda R ., “ The Poetic Function in the Theory of Roman Jakobson,” Poetics

Traube, Ludwig, Nomina sacra: Versuch einer Geschichte der christlichen Kiirzung Today 2.1a (1980), pp. 57-82.

(Munich: C.H . Beck’sche Verlagsbuchhandlung, Oskar Beck, 1907). W einrich, U riel, Modern English-Yiddish, Yiddish-English Dictionary (New York:

Troubridge, St. Vincent, “ Some Notes on Rhyming Argot,” American Speech 21.1 Y 1VO Books, 1968).

(1946), pp. 45-47. Werner, Jurgen, “ Kenntnis und Bewertung fremder Sprachen bei den antiken

Trumper, John, “ Slang and Jargons,” in The Cambridge History o f the Romance Griechen I. Griechen und ‘ Barbaren’: Zum SprachbewuBtein und zum eth-

Languages, vol. 1, Structures, ed. Martin Maiden and John Charles Smith (Cam ­ nischen BewuBtsein im friihgriechischen Epos,” Philologus 133.2 (1989), pp.
16 9 -7 6 .
bridge, MA: Cambridge University Press, 2011), pp. 6 6 0 -8 1.
------ , “ Nichtgriechische Sprachen im BewuBtsein der antiken Griechen,” in Fest­
Tzara, Tristan, CEuvres co m pletes, ed. Henri Behar, 6 vols. (Paris: Flam m arion,
schriftfu r Robert Muth, ed. Paul Handel and Wolfgang Meid (Innsbruck: Inns-
1975-1991).
brucker Beitrage zur Kultuwissenschaft, 1983), pp. 583-95.
------ , “ La signification des anagrammes,” Europe 53 .5 55-55 6 (1975), pp. 86 -9 5.
Werth, Paul, “ Roman Jakobson’s Verbal Analysis of Poetry,” Journal ofLinguistics
Valery, Paul, CEuvres, ed. Jean Hytier, 2 vols. (Paris: Gallim ard, i960).

231
230
 THE FIFTH HAMMER

2012047271
12.1 (1976), pp. 21-73..
West, M. L., Indo-European Poetry and Myth (Oxford: Oxford University Press, 2007)..
Williamson, Craig (ed.), The Old English Riddles of the Exeter Book (Chapel Hill:
University of North Carolina Press, 1977).
Wunderli, Peter, Ferdinand de Saussure und die Anagramme: Linguistik und Literatur
(Tubingen: Niemeyer, 1972).
, "Ferdinand de Saussure, ler cahier a lire preliminairement: Ein
Basistext seiner Anagrammstudien," Zeitschr!ft fur franzosische Sprache und
Literatur 82.3 (1972), pp. 193-216.
___ , "Zur Geltung des Linearitatsprinzips bei Saussure," Vex romanica 31 (1972).
pp. 225-52.
, "Saussure's Anagrams and the Analysis of Literary Texts," in The Cambridge
Companion to Saussure, ed. Carol Sanders (Cambridge: Cambridge
University Press, 2004), pp. 174-8S.
Zumthor, Paul, Langue et techniques poetiques a 1'epoque romane (Xle–XIIIe
siecles) (Paris: Klincksieck, 1963).

232
 Index

ABBREVIATION, 23-34, 102-105, 140. Blasphemy, 104-107.

Alliteration, S6, 111, 115-16. Bodel, Jean, 32.
Alphabets, 95-96, 158-59. Book of Vagabonds and Beggars, The, 33•
Anagrams, 81, 113-16, 152, 154, 156-6o, Bo r ge s , J o rge Lu is , 6 3.
161, 162-8o, 181, 182-83; compared Born, Bertran de, 51.
with anaphones, 119-20, 162. See also Bornelh, Guiraut, 51.
Anaphones. Bragi, 57-58, 64. See also Eddas, Skalds.
Anapho nes, 10 9-3 0 , 1 39, 11 3-3 0 ; Brahmansa, 97.
variant terms for, 119-2o; and synchrony,
CAESAR, JULIUS, 95, 100-101.
122-24, 140-41.
Camden, William, 113, 129•
Andronicus, Livius, 114-15.
Cant, 31-43, 187 n.1. See also
Anti-languages, 37-38, 42.
Jargon. Canting Academy, The, 30, 35.
Apocopation, 39-41.
Castelloza, Na, 54.
Aragon, Louis, 16o, 169. Caveat or Warening for Common. Cursetors,
Argot, 25, 31, 35-36, 38-39, 41, 182.
34.
See also Cant; Jargon.
Aristotle, 9, 11-12, 68. Cercamon, 48.
Arnaut Daniel, 47, 51, 190 n.12. Cervantes Saavedra, Miguel de, 80.

Arp, Hans, 171. Chance, 125-30, 145, 170-83.

Atbash, 95-96, 199 n.3. Clay, Jenny, 86.
Ave-Lallement, Friedrich Christian Cohen, Marcel, 37-38, 188 n.25.
Benedict, 36. Company of the Shell. See Coquillars.
Copyists. See Scribes.
BADER, FRANCOISE, 8 4- 8 5, 91. Coquillars, 20-29, 33, 186 n.2; Francois
Bandits. See Rogues. Villon and, 180.
Barbarians," 11. Course in General Linguistics, 113,
Bartholomaeis, V. de, 52. Behar, 122, 124, 1 3 1 -3 2 , 1 3 6 , 1 39, 203
Henri, 154, 180. Bellman n.15. Cratylus, 11-12, 86.
ofLondon, The, 34. Benveniste, Cryptograms. See Anaphones.
Emile, 106. Culler, Jonathan, 146, 147.
Bernard, Michel, 1 7 6 -77, 1 7 8, 179. Cutups, 170-71.
Bernart de Ventadorn, 50, S3•
Bible, The, 12-13, 95. D A D A , 1 5 4, 1 6 5 -6 6 , 1 6 9 -
al-Biruini, too. 7 0 . Dante, 16, 47.
Blacking, John, 66-67. De return natura, 119, 120-21.

233
 DARK TONGUES

Dekker, Thomas, 34-35. Harman, Thomas, 34.

Democritus, 12. Head, Richard, 30, 3S.
DiNapoli, Robert, 102. Herodotus, 12, 32.
Diodorus, 12. Hensler, Andreas, 6o-62, 71.
Divine language, 84-94, 98-99. Hittite, 90-91.
Dobzynski, Charles, 153-54. Hoepffner, Ernest, 51.
Donats proensals, 32. Homer, 11, 84-89, 116.
Druids, 100-101. Homonymous substitution, 23, 41.
Dufournet, Jean, 154. Homophony, 41, 79-81.
Hopkins, Gerard Manley, 138.
EDDAS, 55-57, 62, 71. Horace, 110.
Elizarenkova, Tatyana, 81, 128. Hypograms. See Anaphones.
Enigmas. See Riddles.
Encomium, 129, JACOBSOHN, H., 86.
Epitomae, 96. Jakobson, Roman, 127-28, 131-50, 199
Etienne, Henri, 35. n.43, 208 n.30.
Euphemism, 105-107. Jargon, 20-28, 31, 37-38, 94; compared
Eustathius of Thessaloniki, 86. with cant, 32.
Exeter Book, 102, Jargon among the Normal, the Degenerate,
and the Criminal, 36.
FABLIAUX, 33. Javanais, 40.
Finnsson Skaldaspillir, Eyvindr, 6o. Jeanroy, Alfred, S2.
Jeu de saint Nicolas, Le, 32.
Fouler, Lucien, 157-58, 162, 211 n.19.
Joseph, John E., 127, 2o6 n.80.
France, Marie de, 24.
Journals and Papers, 138.
Frank, Roberta, 59-60, 63.
Frege, Gottlob, 68,
KALISCHER, ALFRED, 48.
Friedrich, Johannes, 90. Kandragupta, King, 99.
Kenning, 57-64, 193 n. 29.
GALLIC WARS, 100.
Garnier, Joseph, 19. Khlebnikov, Velimir, 143.
Genesis. See The Bible. Kiparsky, Paul, 147, 210 n.55.
Genre, 4S, 80, 129. Kretschmer, Paul, 86.
Gentil, Pierre Le, 154, 158.
Georges, Robert A., 67-68. LANGUAGES, DIVERSITY OF, 12-15.
Gerbstman, Alexander, 81. Largonji de loucherbeme, 4o.
Gergons, 24, 32. Laws of Love, The, 49.
Germanic Mythology, 87. Lazzeroni, R omano, 87.
Godel, Robert, 131. Leewen, J. van, 8 6.
Gonda, Jan, 97. Letters, 95-99, 101-105, 119, 139,
Gordon, E.V, S6, 62-63, 64. 1S9-6o. See also Anagrams.
Grimm, Jakob, 87. Lever, Pierre, 26.
Lindow, John, 61, 71.
Guilhem de Petieus. See William of
Aquitaine. Lindsay, W. M., 104.
Guiraud, Pierre, 41, 186 n.2. Logos, 9-13, 68.
Guieyesse, Georges, 38, 39• Lucian, 16.
Giintert, Hermann, 87-89. Lucretius, 119-21, 120.
Luther, Martin, 33-34•
HANDBOOK OF THE LEARNED, THE. Lycophron, 113.
See Scholar's Primer, The.

234
 INDEX

MALAMOUD, CHARLES, 97-99. Prostitutes, 36, 40, 42.

Mallarme, 10, 158-59.
Proto-Indo-European, 13.
Maroc, Clement, 27.
Meillet, Antoine, 87, 105-106. Puissegur, M., 172-74, 176.
Meissner, Rudolf, 60-61. Puns, 41.
Memoires de la Societe Linguistigue
de Paris, 19, 25. QUENEAU, RAYMOND, 40, 189 n.33.
Memory, 99-102. Quintilian, 16.
Mencken, H. L., 106.
Messenger, John, 66.
RABELAIS, FRANCOIS, 173-74.
Metaphor, 24, 39-40, 61, 64, 67, 70-71.
Rabustel, Jean, 21, 22, 24, 26.
Metathesis, 117. Raimbaut d'Aurenga, so, 51-52.
Metonymy, 23-24, 39, 70-71 78-79, 143. Raimbaut de Vacquerias, 50, 51.
Michel, Francisque, 19. Remains concerning Britaine, 113.
Milicic, Vladimir, 81. Renou, Ernest, 75.
Milner, Jean-Claude, 39, 123, 125. Rhyming slang, 41-42, 189 n.39
Moret, Andre, 47. Richard h Maus, 25, 32.
Richter, Hans, 171.
NERVAL, GERARD DE, 177. Riddles, 65-82, 94, 102-103, 196 n.52;
Niceforo, Alfredo, 36. Precedent and Sequent in, 66-69, 76,
Noms de metier, 22-23. 79-81.
Nykl, A. R., 52. "Riddle Hymn," 69-70, 89.
Riffaterre, Michel, 147-48.
ON THE LANGUAGE OF THE GODS AND Riquer, Martin de, 52.
THE LANGUAGE OF SPIRITS, 87-89. Riquier, Isabel de, 51.
OPOJAZ, 134. Ritual, 75-76, 89-90. See also
d'Orleans, Charles, 24, 175-77. Sacrifice. Rogues, 19-28, 31-38.
Rousseau, jean-Jacques, 16.
PAGIS, DAN, 80. Runes, 101-103
Pantomime, 16. .
Parallelism, 115, 141-48. Pascoli, SACRIFICE, 74-75, 89-90, 100.
Giovanni, 126. Saturnian, 109-18, 121-23, 125-26, 138-39.
Peire Vidal, 51. Saussure, Ferdinand de, 109-50.
People of the Shell. See Coquillars. Savinel, Pierre, 155.
Philology, 13, 28, 33, 35, 51, 104; Scholar's Primer, The, 90-91.
Divine speech and, 86-94. Schone, Lucien, 27.
Phonology, 40, 79, 81, 114-15, 140-41. Schwob, Marcel, 19-20, 25-29, 38, 39.
Phonetics, 16, 84, 110-11. "Scramblings," 95-96.
Pho nic i mit at io n. See Ana p ho nes. Scribes, 102, 104-105, 166-67.
Pipet, Alfred, 46. Secret of Villon, The, 153-83, 158-82.
Pini of Urbino, Teseo, 33. Selected Writings, Poetry of Grammar
Plato, 11, 67, 86. Grammar of Poetry, 148.
Poe, Edgar Allan, 138, 141-42, 144. Shklovsky, Viktor, 42.
Po e t ic fu n c t i o n, 1 2 1 , 1 3 4 - 5 1 . Six Lectures on Sound and Meaning, 139-40.
Poetic theology, 93-94. Simultaneous poem, 165-66.
Poeticity. See Poetic function. Politics, 9. Signifiers, 123-24, 139-40.
Po rz i g, W a lt e r, 6 9 -7 0 , 7 8, 9 4. Skeggjason, Markus, 59..
Probability, 172-79. Skalds, 55-64, 71-72.

235
 DARK TONGUES

Slang, 31-32, 38-41. West, M. L., 89, 90.

Sordello, 50. William of Aquitaine, 48.
Sound Shape of Language, The, 138-39. Words upon Words, 113, 126-29, 131-32.
Special languages, 22, 37-42, 69-70, Wright, Barbara, 40.
77-78; compared with divine jargon,
98. ZACUTO, MOSHE, 80.
Speculum cerratanorum, 33. Zumthor, Paul, 45.
Speech events. See Poetic function.
Sphinx, The, 68, 78.
Starobinski, Jean, 113, 126-32,
146, 203 n.16.
Stints, Lynn D., 174-77.
Sturluson, Snorri, 56, 57-58, 62, 63, 64,
192 n.10.
Suetonius, 95.
Syllabograms. See Anaphones.
Synonymy, 12, 38, 93; riddles and, 69,
81-82. See also Riddles, Precedent
and Sequent in.
Syntagmatic relation, 136-37, 143-44,
193 n.29.

TABOO, 87, 105-107. See also blasphemy.

Tacitus, 105-107.
Taylor, Archer, 67.
Thompson, Sir Edward Maunde,
103, Pordarson, Sturla, 6o.
Thuasne, Louis, 159•
Tory, Geoffrey, 27, 187 n.23.
Todorov, Tzvetan, 81.
Toporov, Vladimir, 81.
Traube, Ludwig, 103-105.
Troubadours, 45-54; Domnas and,
46-48; Senhals and, 49-50, 51-52.
Tzars, Christophe, 154.
Tzara, Tristan, 152, 153-83.

VALERY, PAUL, 42.

Vallet, Eduardo, 48.
Vedas, 69-70, 75, 78, 88-89, 99-100, 128.
Verlan, 40, M.
Vidocq, E. F., 35.
Vilgram, 175-77.
Villon, Francois, 25-29, 35, 152, 153-83.
Virgilius Mare, 96.

WAUGH, LINDA, 138-39.

Watkins, Calvert, 90-93.

216