In Praise Ofthe Variant: Critical History of Philology

PARALLAX
•• RE-VISIONS OF CULTURE
AND SOCIETY
Stephen G. Nichols, Gerald Prince, and wendy Steiner

SERIES EDITORS
In Praise ofthe Variant
A Critical History of Philology
T RAN S L ATE D
-
Bernard Cerquiglini
BY Betsy wtng
The Johns Hopkins University Press

Baltimore and London JC)C19
~
1"
Originally published as Eloge de fa variante: Histoire

critique de fa philologie, Copyright Editions du Seui\, 1989
© 1999 The Johns Hopkins University Press
All rights reserved. Published 1999
Printed in the United States ofAmerica on acid-free paper Michel Foucault hoped to see this book,
-
9 8 7 6 5 4 3 21 which is dedicated to his memory.
The Johns Hopkins University Press
2715 Norrh Charles Street
Baltimote, Maryland 21218-4363
www.press.jhu.edu
Library of Congress Cataloging-in-Publication Data will

be found at the end of this book.
A catalog record for this book is available from the
British Library.
ISBN 0-8018-6126-8
~\::l
L;j NaS!Oila\bibHoteKet
!J CNoellr,e ll:\O
'6lfo ICi C~I

)j
f"
.. " ,,'i
·'{~t
,1p
'
I Contents
Translator's Note IX
Introduction Xl
Textuary Modernity 1
2 Mr. Procrustes, Philologist 13
3 The Joyful Excess 33
4 Gaston Paris and the Dinosaurs 46
5 Turn the Page 72
Notes 83
Index 89
Vll
· .1
,
,;
Ii}
'1,.1/
j,f!
~i~\-\
I Translator's Note
There are several words in this carefully conceived text which are far
from common in French, among them textuaire, orature, and
ecranique. I have made little attempt to conceal their unusual, even
jarring, nature in English because each has a meaning that would
otherwise require circuitous phrasing. When new ideas come into a
language they tend to bring their words with them, and it is my
strong belief that a translator's overdomestication of their difference
does neither the writer nor the reader any real favors.
Textuaire, which does not appear in either the Larousse Lexis or the
Petit Robert, is translated here as textuary--a word appearing in the
Shorter OED and both the American Heritage and websters New
World dictionaries, all of which provide a somewhat more limited
definition than textuaire requires. Once or twice phrases such as "of a
textual nature" or "characteristic of texts" could convey its sense eas-
ily enough. But it is clear that Modernite Textuaire-the first chapter
heading-is not sufficiently translated as "Textual Modernity" or
even "Modernity of the Text," which translate French phrases that
the author carefully did not choose. The chapter unfolds to make it
more and more plain that modernity and the notion of text are al-
most one and the same; that text is literally modernity; that moder-
nity belongs to the text; that text formed and forms modernity; that
there is something that legitimately may be called textuary moder-
nity.
Orature is of recent origin and has some currency among both
English- and French-speaking scholars interested in the forms of oral
tradition as they are compared or contrasted with those of literature
IX
Translator's Note
(which word is usually to be found nearby). In this context, though

somewhat unusual, it is perfectly comprehensible.
Ecranique is also a recent coinage, an adjective used to describe
something whose characteristics are determined by a screen-in this
I Introduction
case texts. Screenic seemed the most useful English equivalent.
The manuscript, which has long been pushed to the margins of le-
gitimate reflection and sometimes even obliterated-the abominable
trace of some positivist concern-is now the latest object of analysis.
These days one hears terms (textual criticism, draft, variant) that only
a few unobtrusive specialists of ancient or medieval languages or
members of the old guard had preserved out of conviction. This ab-
sence, however, has not come back as a pendulum swing or a back-
lash, not as some posthumous vengeance on former studies, its ghost
leading the shuffiing cotillion of disenchanted semioticians. This re-
newal of interest is a movement whose magnitude cannot fail to im-
press. Genetic criticism (criticism directed at the genesis of a piece of
writing) is now coming in like a desired and overwhelming tide after
several years of low water, when we had all but gone back to literary
history. The diversity of this phenomenon is also striking. It is the lo-
cus for a rediscovered dialogue among disciplines and professions (li-
brarian and critic, theoretician and editor), a fortunate harmony that
might remind us of the harmony formerly called Humanism. In
shon, methodological attention here goes hand in hand with a con-
cern for practical validation. Once again, criticism makes reading
available in its strongest sense because the final goal of analysis is
publication. None of this is the result of chance or of the past. The
present interest in the handwritten trace, the modalities of expres-
sion, the meticulous attention paid to the preparation of the work, as
well as to the precariousness of writing, reflect a sort of impatient
longing in the presence of the written text. This extremely anxious
scrutiny of the genesis of a piece ofwriting is an attempt to escape the
shackles of the fixed text by destabilizing inscriptions.
x Xl
Introduction Introduction
Such an approach can emerge today, of course, because it is per- a technical gesture, and the gesture of the printing press in its heyday.
fectly inscribed within the critical thought that has been brought to The text, the industrial press, and modernity act as one. It is precisely
bear upon texts for two decades. It extends them, it is their culmina- because of the dissemination of the unity of a single technique that,
tion and their complete realization. The reading that genetic analysis incidentally, the literary text gives meaning to the halo of writing sur-
offers is no longer a coherent construction of one of the work's senses; rounding it.
it is a reading that concretely participates in establishing the letter of Criticism is increasingly fond of whatever is unstable, multiple,
the work. By creating a printed object that makes an act of writing and precarious, whatever temporarily goes beyond the enclosed im-
available to readers, it provides a precise picture of the dynamics of mobility established by the machine. Note that this is happening at
this act. What is usually called the pre-text (all the drafts, manu- the very moment when the computer, the new technology, is pro-
scripts, etc.) constitutes a textual elsewhere, the elsewhere of writing ducing floods of a writing that is mobile, various, and fluctuating.
that is the locus of all determining factors, the space of every tra- Does that mean that the written work is itself at stake? One might
versal: drives, desires, prefabricated ideologies, horizons of expecta- think so because of the extent to which orality now enjoys favor. A
tion. The work exhibits in its attested preliminaries those things that thundering uproar exists in this realm, and it has qualified itselfas the
determine it in various ways. Genetic study clearly meets two re- avant-garde with such self-satisfaction that the attention directed to-
quirements that the literary theories of modernity have voiced, some- ward revolutions in writing could be considered quite negligible and
times in contradictory terms: to return to literature its wealth byas- insignificant. This because it is writing, by and large, that is now ex-
suming the preeminence of writing, and to point out what, on the pected to vanish before long. Although it is possible to distinguish a
order of being or of society, conditions this writing. From this per- bit of good from the worst elements in this chorus as it spreads (the
spective, the unique moment in which the publishable form emerges conventional and empty apologia from the media and the communi-
from the mass of preparatory writing becomes increasingly impor- cations industry), one should hardly spend much time doing so. We
tant. This cesura fixing into a text the plurivocal and multidimen- shall not take a stand here on behalf of the oral, the body, or the
sional machine of writing is crucial. The almost topological fixation voice. The promotion of orality within a great many critical dis-
of literary discipline upon a nodal point where two heterogeneous courses is of course a symptom, but a symptom that is deceptive; re-
and asymmetrical spaces take their bearings, the poipt that makes search, with great fanfare, is entering a blind alley. The technical,
analysis possible, is a very precise indication of its historicity. The vi- hence cultural, mutation that we are part of does not call the written
tal gesture that marks the advent of a text is somewhat reminiscent of word into question once again-quite the opposite. Information
the signature on a final corrected proof. technology is a technology of writing. On the one hand, it is amaz-
The name affixed to the page destined for the printer authorizes. If ingly effective at fulfilling the universally positive and progressive
it orders and allows the multiplication of a single fragment ofwriting functions of the written word (registering, classifYing, relocating, and
by itself, it gives this fragment the status of a text. It equips it with an expanding forms of knowledge). On the other hand, it consumes,
author, that is, with an origin and an entitlement; it endows it with a produces, stocks, and diffuses vast numbers of inscriptions. In short,
canonical form, that is, with a stable conformity. Preparatory writing it introduces new processes of writing and reading. It is within the
of any sort ends only in the gesture that affixes the name. The signa- written word that our intellectual habits are becoming most deeply
ture is the ultimate writing for which the pre-text constitutes the pro- fractured.
tean preproduction. It is at once an institutional and a legal gesture, What has pointed the way to this renewed approach to manu-
xu Xlll
Introduction
scripts has been vaguely perceived by those who idolize the oral, and
it is now demonstrated to us by information technology: our con-
ception of textuality is at stake. How a text is ordinarily thought of, a
conception upon which both the simplest daily practice and theoret-
ical statements are based, is revealed to be thought representing en-
closed conformity and authorized stability. It is our usual notion, In Praise ofthe Variant
-
mental paraphernalia that is so robust and convenient that no one
particularly thinks about it; we forget that it has a history and that
this history is a recent one. At the dawn of the nineteenth century,
extremely diverse phenomena of order, nature, and evolution all
seemed to converge, forming a coherent semantics connected with
the practice and study of texts. Philology is the most significant ex-
pression of this coherence. Its history is the history of our sponta-
neous philosophy of the textual.
XIV
1 I Textuary Modernity
Lucien Febvre, in his inaugural lecture at the College de France, ex-

amined his conscience as the young historian he was and formulated
a litany of textual studies:
In the realm of modern studies, young men formed intellectually

by a culture based only on texts, studies of texts, critical analyses
of texts, moved with no break in their habits from secondary
schools, where it was solely their abilities as students of texts that
ranked them, to the Ecole normale, the Sorbonne, and the other
Faculties where the same work of studying texts and analyzing
texts was assigned to them. l
There are certainly a thousand ways of charac:terizing the nine-

teenth century from which to choose; we intend, for our part, to ex-
amine this religion of the text which so rightly annoyed the recal~i=
.Y~ apprentice; for a historical positivist education is thoroughly
representative of this trust in the factual nature of texts, this confi-
dence in their being established, this desire that they be the basis of )(
critical thought. Starting from reliable texts, or ones that have been i.
made so, in order to construct a superior theory of commentary, is a
procedure that theoreticians of contemporary "textuality" of every
-J ..
IN PRAISE OF THE VARIANT Textuary Modernity
sort would be unable to reject completely. Whatever the point of lures the copyist, whom it catches in its snare and sets up as subject.
view one may take to examine it, the text as a stumbling block seems Whence the other dream: the machine. To turn the scribe into the
one of the moral principles of our modernity. It is the origin of criti- objective mechanics of transcription, achieving this through disci-
cal discourse because the text itself asks the question of origin in a vi- pline and punishment: the untouchable sacred and the supervisor.
tal manner (Who formulates or transmits it; under what conditions Who is the better scribe? The one who does not understand and re-
and to what ends?). And it is the challenge and guarantee of com- produces signs, or the one who understands everything but respect-
mentary because it is materiality itself (published under the supervi- fully adheres to it? The ideal worker is an average person, a terrorized
sion of the author or the philologist, printed, added to the sacred half-scholar.
treasure of the library). The printing press did not fulfill this desire for a machine at the
The phenomenon of the text gets so much credit for many rea- outset. On the contrary, we know that the scriptoria of the late Mid-
sons; for us it is important to show here that technical and legal dle Ages had achieved a quality, if not a reliability, of reproduction
breakthroughs have not been immaterial. The relationship, which that was far superior to the first print shops. The success of the press
probably could be formulated, among these breakthroughs ofvarious is due to other irreversible factors: the lowering of costs, the speed of
sorts is less important than their massive and seemingly inevitable composition, the multiplication of print runs, and so on. The pro- .
convergence at the end of the eighteenth century. gressive conquest of what is ordinary today-the immutable multi-
ple-must be read according to the rising curve of this new technol-
ogy from 1530 to the dawn of the nineteenth century.
The Unique . ..
Literary genetics is thoroughly representive of the modern con-
The notion of a reliable text, or one whose reliability can be en- ception of the text; it explores the polymorphous activity of writing
sured by establishing it, is interdependent with the printing industry's which precedes the final gesture of the hand that has the final power
reaching a state that finally provides satisfaction: what came from the to attest to the conformity of the proof, the hand that gives permis-
press was in keeping with the desire of the author, whatever that sion for reproduction but with no possible intervention. The final
might be, now definitively fixed within the two-dimensional limits of corrected proof is what separates writing from text, writer from au-
the page and infinitely reproducible in its integrity. The text took on thor, freedom from law. The watershed of the literary process, its
supreme importance, accessible to everyone in precisely the form north slope, always deeper and more shadowy, is probed by genetics.
wished by the final writing hand upon delivering it to the press. In If the preparatory instability of signs and forms takes on such great
this state it would act from now on as the reference and authority. value for analysis, it is because the fixation of the literary object seems
There can be no doubt that this represented immense progress both impossible to ignore and natural: the two dimensions of the
and the realization of an old dream, one probably formulated when printed page delineate its inexorable fate. A cruel fate it is, no doubt,
writing was born: the faithful copy. Every copy is alteration; it is nec- which transfigures a production of sense by immobilizing it, but it is
essary, however, t~-;:~~~c~~efully, word for word, certain found- a fate so pleasant that one resigns oneself to it, except in a vague
ing legal or sacred utterances. Transcribing is treacherous because the search, as in genetics, for some earlier dynamics in its preparatory
I human hand finds it very hard to give up the elusive possession of stages. Because this immobility, which we see as being imprinted
, sense. In cases of deliberate revision as well as in careless mistakes, upon the textual phenomenon, referring every written word back to
I, something is at work to restore life to inert inscription; the language mechanical reproduction's objectified immutable, necessarilyaccom-
,I 2
3
II
I'
I, 4 ,'" _
panies finally guaranteed control. It took several centuries for that to hemently addresses some supervisor who is hardly listening. That is
J happen; the text that is standard and definitive in its multiplicity the author-somewhat ridiculous, somewhat out of place, or, at least,
i dates from the very end of the eighteenth century. he has not yet found his place, his status. Reactions from that time
It would be good to write a history of the correction of things that on were individual and in every shape or form. There was the
have been printed. In the very technical and partial studies devoted haughty refusal, at one end of the scale, to go to the print shop or
to this particularly daunting subject, the birth of the modern author participate in any way at all in proofreading (even if this meant slip-
and his work can be read. 2 Anxiety comes first, a fear so great that the ping in a spiteful preface that spoke of "correctorum errata emendare,"
technical progress extinguishing it at the dawn of modernity seems to etc.). At the other end, we see an almost military and somewhat ma-
bring history to a halt. When the first works were entrusted to the niacal occupation of the shop, as by Erasmus, who spent eight
press, the last scribes were not working so badly and people stayed months in Venice in 1508 while Aldus Municius printed his Adages,
with them in the homogeneous and reassuring space of writing by or as by Jacques Peletier du Mans, who promptly moved in with
hand; the print shop, a frenetic microcosm, concentrated the disqui- Michel de Vascosan to oversee the printing of his Oeuvrespotftiques in
eting novelties of the technical hand, of specialization, of the ma- 1547. However, and this is more important, any desire for control was
chine. The agent (author or editor) entrusting the press with the blighted by correction procedures that were awkward and annoyingly
work, if he were attached to the letter of it, had the feeling of turning diverse throughout the course of the book's production. Until the
it over to dispersal and lack, turning himself over, ill equipped, to dawn of the nineteenth century, the supervision of conformity, when
some perverse doubles of himself: typography, which made mistakes, it existed, had little to do with any sovereign gesture or final com-
the impudent proofreader, who invented them. The printed text was mand at the press. Corrections were made during the printing pro-
anything but certain: grievances were commonplace, and old books cess-indeed, even afterward; and the bad-tempered presence of the
\ repeat a long litany of "Errata, si quae occurrent, benevolus Lector," to author in the shop seems of a piece with this constant manipulation
plead not guilty in a rather offhanded manner that represented little of the text. There were corrections during the print run (with the re-!
respect for what was being published. Generally speaking, until the suit that copies from the same edition can be different, which, for us,\
second half of the eighteenth century, there was no coherent and con- negates the primary role of the printing press), and there were cor- \
scious intervention vouching for the letter, other than grievances, in rections at the end of the print run (by the addition of inserts, the
answer to the very relative faithfulness of the text. Specifically, the di- gluing in of strips of paper, or corrections by hand). This last proce-
versity of reactions, situations, and methods did not form any homo- dure was rare but very important: as the author took back his book,
geneous concept of the author defined in relation to his text. These he seemed to triumph over the machine and imposed upon it the
last two ideas come later and are interconnected. sovereignty of writing. A pathetic power, in fact, it was exercised
For that matter, what place does the new technology give the au- everywhere and resided nowhere. Paradoxically, it was his exclusion
thor? A famous engraving by Holbein represents a print shop. In the that gave the author a semblance of status. He finally had a place: at
foreground is the master printer seen from behind, huge, organizing home.
the space as he distributes the work. All around him, bustling and Of course, the notion of the proof as being corrected by the hand
merry, are the workers of Humanism: running the press, washing the of the person who vouches for its letter is very old, and we have evi-
type, drying the page. In the background is a table where a man is dence of it starting in 1530; this proofreading, however, took place at
calmly reading: the proofreader. Finally there is a little fellow who ve- the shop itself, in the space and during the time of production. When
4 5
did proof begin to be read in town? This question probably seemed Words written in preparation, consequently, became important
of little interest to the specialists; however, from their work one can whether considered as, in turn, divine inspiration, creative genius, or
draw the conclusion that it was not until the beginning of the seven- blind processes of writing; above all, they took on value. How long
teenth century that there is any evidence of proof being sent home have writers been saving their drafts of printed works before they or
and that it was in the course of the eighteenth century that this prac- their heirs turn them over either to the market or to some devout am-
tice spread and, indeed, became established. The best presses, such as ateur or to a library? Here too, an obscure but fundamental history ('
Oxford University Press, decided once and for all, as Percy Simpson remains to be written, that of the notebook, the tear-out, the scrap of
puts it so well: "to keep the author off the premises." paper. The "author's manuscript," researched and studied, is appar-
By sending him away and by discharging upon him the responsi- ently a modern idea. Louis Hay notes that the first manuscript col-
bility for the text (what he considered true to his words will be mul- lections date from German Romanticism, when national tradition
tiplied), the printer acknowledged that the author had a power and a was glorified. 3 An enlightened and patriotic group of sponsors began
right. At his table, the very same as the one on which the work was gathering together documents that related to German literature, pri-
developed, the author would transform this work into a text that was marily manuscripts and drafts. Private or public foundations served
immutable and vouched for by him alone. Accompanying the final as depositories and research centers, culminating with the Goethe-
game of proof, the act of passing for press, in which writing culmi- und Schiller-Archiv in the Weimar Republic. We can see that it was
nated and absolutely abolished itself, became widespread at the be- during the nineteenth century that the author's manuscripts acquired
ginning of the nineteenth century. It marked the birth of the modern status and social recognition (either for their commercial value or as
text, for it established the deep and defining break between pre-text a specific institution) and that this followed a rapidly rising curve.
and text. Ofcourse, evolution in this specific realm-the proofs for a Toward the middle of the century there were more than a hundred
preindustrial and then an industrial press-was slow and lethargic, important collections in Germany (four times more than the number
and as a result extremely instructive. We know that, from Voltaire to in France). This Germanic advance implied that philology, that pio-
Stendhal, it was customary not to put the finishing touches on the neer from beyond the Rhine, played a role here. Despite the differ-
manuscript but to amend and correct on successive proofs, "on the ence then assumed to exist between ancient or medieval manuscripts
metal." The proof was very much part of the pre-text; it was the final and modern manuscripts-the former something like a remnant, the
draft. In the secular battle against the author, conducted in the name latter a rough sketch-it was the same taste for the priceless traces left
of rational work and economic interest, the printer once again had by a hand that was attractive in both cases. This was a Romantic
been on the side of the text. Though the restriction of "corrections by taste, of course, for things of the past, glorifying the poet of genius
the author" seemed to destroy a privilege, it put the finishing touches right down to his most minor scribblings. Even so, this interest was
on the distinction between the space of writing and that of mechan- never possible, much less intense, until the day that the machine,
ics, opposing a shifting and polymorphous personal element to a which for about two centuries had been bringing the notion of text
public one that was immutable and multiple. Modernity's corrected into line with the image of the final printed page, increased the value
proof, reduced to a certificate concluded by a signature, is this water- of the writing by hand that went before, whether in the history of hu-
shed that can be crossed with a single step but where the hillside manity or in the history of an individual. The draft became desirable
slopes away in opposite directions and the waters divide. when the text became fixed and its author had rights.
6 7
,,/\!
... and His Property place, securely albeit in a half-light. The emergence of the author in
the period from the sixteenth to the nineteenth century is a complex
This second aspect is well known, and only its historical conver-
but well-known phenomenon that falls into the realm of what might
gence with other notions of the text interests us here. It was at the be-
be called "internal literary history." The importance of the postclassi-
ginning of the nineteenth century that the idea of literary property
cal notion of belles-lettres. which developed as the old rhetoric be-
acquired the force of law. At that time the law stated that every text
came fractured, is part of this history. Having lost one after the other,
was first of all something belonging to the person who conceived it:
pronuntiatio and memoria (to the theater), then inventio and disposi-
origins and paternity. In other words, as Michel Foucault clearly saw,
tio (shifted to logic), and finally reduced to elocutio alone, that is, to
the idea of the author took hold at the center of the notion of text,
a purely ornamental art, rhetoric, which had now become belles-Iet-
which became constituted as the "work of": the modern text is geni-
tres, highlights the unique talent of somebody who knows how to
tive. 4 It is certain that, as usual, the legislature translated into legal
say, like nobody else on earth, what everybody thinks. Obviously, the ""
terms something that people already thought of as a fact, namely, the
emergence of the author is also a matter of external literary history.
complete and apparently definitive emergence of the notion of the
The slow waning of royal patronage; the ups and downs of private
literary author.
sponsorship; the first demands during the 1720S for financial, not to
Evolution was slow, from the autumn of the Middle Ages until the
say professional, autonomy, coming mainly from the humble rank
end of the Enlightenment. It ended with modernity, which it helped
and file of writers-who asked at most to get some profit from the
to found, and it forms a meaningful parallel with the history of print-
sale of their books; the conflicting relations with the booksellers' pro-
ing, which we mentioned earlier. The invention of the printing press
fession; and finally (though not before the last quarter of the cen-
was far more a psychological than a technical revolution, that is, a
tury), the obtaining of rights over print runs and later editions-all
stuttering and inconvenient one, certainly, but one that, in reality,
of these need to be mentioned.
implied a different relationship with the text. After all, as Elizabeth
We prefer to emphasize the paratextual aspects of the phenome-
Eisenstein remarks, Erasmus could try to put things to rights and ob-
non, and once again the history of the book is very instructive'? Our
tain a correct edition by publishing errata, but Saint Jerome and Al-
perfectly ordinary (but highly standardized) title page, which Lucien
cuin could not do so.5 The fact is that the idea never dawned on
Febvre described as the "registry" of printed works, was slow to be-
them. Gutenberg's desire, for all its novelry, was not fully realized un-
come established. The printed text of the first books began, as did
til the beginning of the nineteenth century. Things went along as if
manuscripts, on the recto of the first page, introduced by a few brief
I these techniques of the written word were the bearers of radically new
I iii
words (no doubt the former incipit) indicating the subject of the
ideas, and for a long time people saw them as offensive. The same
work and occasionally--only occasionally-the name of the author.
had been true for the invention of the codex, and it is true today with
There was a very material reason for the origins of the title page: be-
computer word processing.
cause the recto of the first page was particularly likely to get soiled, it X
The author is not a medieval concept. We shall come back to that,
became usual to begin printing on the verso of this page. The un-
and, although the emergence of the figure and practice of the writer
printed recto, therefore, could have a number of things on it: an il-
can be shown starting in the fourteenth century, what looks like a
lustration, the printer's stamp, the address of the bookseller, the title
functional anachronism is attached to the expression medieval au-
-more and more comprehensive (there was plenry of room) if not,
thor. 6 As with the printing press, the Renaissance set new ideas in
indeed, an outline of the work-and finally, in establishments run by
8 9
Humanist printers, the author's name. One cannot follow the prog- Lakanal proposed a "declaration of the rights of the genius," the de-
ress made by this last item during the sixteenth century simply in cree of 21 July 1793 granted authors (and their heirs) the exclusive
terms of its absence or presence; the reason for mentioning it was also right to sell, cause to be sold, and distribute their works or to surren-
important. The decorative function of this first page turned the sur- der, wholly or partially, ownership of them. Thus a noble idea, seeing
name into what was, apparently, an element of illustration demon- in the "fruits of thought" the most sacred and most personal of prop-
strating the skill of the engraver in a complex game of masking and erties (very belatedly acknowledged and honored, it is true), gave the
light (framing it voluminously, setting it in infinite regression, etc.). author a "sovereign power," an indefeasible moral right and a patri-
The title page, with the author's name as a component, purely typo- mony. Obviously, what was granted to the author exceeded anything
graphical and informational in nature, seems to have become estab- that could be expected from a reasonable and fair settlement of the
lished fact (if one excludes baroque editions) in the first third of the dispute with booksellers, on the one hand, and counterfeiters, on the
seventeenth century. 8 It would be nice to have the same sort of infor- other. We are face to face with a change in the way people thought,
mation about bindings (at what date does the piece of leather glued which went deeper and moved faster because of the revolutionary up-
to the back of a book bear not only the title of the work but also the surge. This can be verified in other ways: the 1793 decree of the revo-
name of its author?), and about libraries (at what date do books be- lutionary assembly is still the basis of the French legal system as far as
gin to be organized upright so that the spine is showing? when does literary property is concerned; revisions and extension that came later
cataloguing according to surname appear?). Our way of doing these (the Berne Convention, inspired by France in 1886; the law of 19 May
things apparently goes back no earlier than the eighteenth century. 1925, etc.) were executed in relation to the principles and sometimes
We can agree, therefore, that displaying and defining the printed in the terms of the decree of 1793. 9 A memorable law, it should be in-
work under the authority of the name of the person to whom the im- scribed on the frontispiece of every textbook of literature: it an-
mutability of the letter is attributed is a rather recent custom. nounced to the world the birth of the modern author.
In short, this genesis of the concept of author, as it forms de facto Clearly, many notions that, either in theoretical vigilance or in the
our notion of the text, had a legal aspect. This all came about in the semiconfusion of practice, we connect to the idea of text result from
final years of the eighteenth century, principally through the reduc- a technical, mental, and legal upheaval that took shape at the dawn
tion of the privileges of the booksellers' corporation. In 1777, the year of the nineteenth century-that is, if we employ a broad synchrony
in which, following Beaumarchais's wishes, the Society of Dramatic and bear in mind that there are multicentury evolutions, with their
Authors was founded, the privileges granted booksellers were limited inherent rhythm and intensity, coming to an end. All the same, we
by the decrees of 30 August, inspired by the anticorporation politics can pinpoint a convergence that is sufficiently marked for it to con-
ofTurgot. Specifically, publication rights became legally transferable stitute one of the definitive instruments of our thought. Or one of
to the author, who wanted to exploit his own works. Legal proce- those presumed to be so, for computer word processing, which we see
dures set in place by the Revolution completed this development. now spreading far and wide, demands our urgent attention. It poses
The law of 2 March 1791 extended the abolition of privileges into the thorny legal problems, and this time it does not seem possible to re-
economic domain, and hence the rights of booksellers as well as the solve them in the spirit of 1793. Word-processing technique, despite
ownership of authors lost all legal basis. Published works fell prey to an inert relatedness by vocabulary, bears no resemblance at all to the
counterfeiters in a period when the market for printed material expe- printing press: the screen represents vast numbers of pages; a print-
rienced enormous growth. Following a preliminary report in which Out is not a proof.
10 II
~ I
IN PRAISE OF THE VARIANT
In the nineteenth century the idea of the text as it was thus con-
stituted came into its own and gave birth to a particular and funda-
mental science: philology. It is a modern science to our minds, even
Mr. Procrustes, Philologist
though, since the 1950S, a young, impatient, and fractious body of
critics has attempted to throw off the yoke. A modern science of the
modern text, philology is applied to ancient objects. We shall exam-
ine how it relates to medieval works, where the historicity of its prin-
ciples and, put simply, its very peculiar anachronism are most clearly
apparent.
There are legitimate reasons for examining nineteenth-century philo-

logical practice in its application to medieval manuscripts written in
the vernacular rather than in its relationship with the ancient works
or sacred texts for which it had been created. Displaced in this man-
ner, or, to risk a metaphor, uncalled for, this practice plainly reveals
its dynamics. Moreover, in a new translatio studii, medieval culture
now takes precedence over Greek and Latin culture, formerly the ma-
trix of references, identifications, and illusions, as the locus of our in-
tense examination. There are, to be sure, many reasons for keen in-
terest in the Middle Ages. The media have worked some minor
miracles, which no doubt merged with the excellence of a historical
school as well as with anxieties born of industrial society and its cri-
sis: this is, in short, the vast legacy of Marc Bloch and bucolic nostal-
gias. Nonetheless, interest that is focused equally on food and the fine
arts, on feudal ideology and on trivial everyday acts, is a sign that this
examination holds a great deal of meaning for a world in which some
founding otherness is acknowledged.
Study of the first texts written in French, between the eleventh
and thirteenth centuries, examining their most material aspects ac-
cording to the specific features of this writing activity, is part of an in-
12 13
IN PRAISE OF THE VARIANT Mr. Procrustes, Philologist
quiry that reveals discontinuities at the heart of something that has sloughing off the old rhetoric, made progress in the twelfth century,
been accepted as an almost ontological continuum. It thus represents enlisting logic, or dialectics, on the side of modern thought and
an archeology of our own mental structures-all the more true be- abounding in neologisms: scholastic Latin, swaggering before it ossi-
cause we are in a pioneering field that serves as a model. fied.
The written word made progress that was decisive, with no going
back. For, though men of letters (able to read, perhaps even write,
The Stakes ofthe Written WOrd and knowing Latin-one implied the other) remained in the minor-
The page that was blank from the sixth century to the millennium ity (unless the overestimation of medieval illiteracy is, once again, a
would be first written on again in Gallo-Roman territory. The fate of fantasy of the nineteenth century), they were the bearers of positive
scholarly culture and written works-that is, Latin-is well known; and legitimate values that constituted a system of reference. This was
for several centuries it flickered as the final flame of ancient culture, a critical. The change that took place between A.D. 1000 and IIOO was,
night light kept burning where people worshipped. Precarious sup- on the one hand, in the mere quantity of inscriptions of every sort
port for it came from the church, thanks to several schools, and above produced by medieval society (measured by the consumption of
all thanks to the liturgy, which reinforced the specialized, ritual aspect manuscripts and ink as well as sealing wax). On the other hand, the
of the Latin language. The language of religion, Latin was also an ex- greatest change was that the written object was beginning to be insti-
tremely useful medium of communication for the very few scholars, tutionalized. In England, for example, the least baronet would soon
as well as the vehicle for the production, or rather, reproduction of be required to read and sign (or to possess a seal), as would more than
knowledge. It cut itself off from the rapidly developing vernaculars one of his newly rich serfs. 2
rooted in it, in a break reinforced by the Carolingian renaissance and One can grasp the start of a new mentality that accompanied and
especially by the Latin pronunciation proposed by Alcuin, which fur- facilitated (not to say, permitted) the great renewal of the twelfth cen-
ther separated the regenerated Latin that was taught from maternal, tury. It constituted a deep, rich humus of craving for the written
everyday languages. What little writing took place was in Latin; word, new layers of visual capacities, and a conception of knowledge
nothing expressed in the vernacular was worthy of being inscribed that would be the ground for the culture of print after its break with
and certainly not worth slaughtering a few sheep over. the practice of writing by hand. The appearance of this new "tech-
. None of the conditions existed which would allow the develop- nology of the intellect," as Jack Goody calls it, among all the techni-
II ment of a literature in a vernacular language, the first such condition cal inventions of the millennium, had widespread and diverse conse-
being an audience. There were circles ofscholars and their pupils, but quences. It constituted a set of mental tools very well described by
they did not constitute this cultivated audience required by literature. Brian Stock. 3 There was a new sense of temporality together with a
Erich Auerbach, a faithful disciple of Curtius, noted its absence and different perception of space and its organization (the structural links
saw in it long centuries of drastic impoverishment of the ancient her- between scholastic thought and gothic architecture are, moreover,
itage. l From this point of view, the revival around the millennium, a well known).4 With this came the notion of exchange at a distance
joyful and rapid expansion (masked by the "millenarian terrors" of and communication, in short, an intellectualism enhanced by reflec-
Romantic historiography), was stupendous. Development of cities tion and knowledge. The written word, even ifit eluded most people,
and commerce, safer roads, land clearing, the first court societies, and was the new figure of authority; it was the basis of both commitment
bourgeois money all opened the way for a cultural revival. Latin, and debate, constituting the inalienable reference and permitting
14 15
positive criticism. Abelard's Sic et Non, bringing apparently contra- ing as techniques became associated now, as we have mentioned, with
dictory fragments of sacred texts together to link them logically, is an the development of trade. In addition, futures were opening up and
admirable example of the progress in reasoning made possible by new social groups emerging (among which great importance would
writing. Even more, reflection on the sign, whether by grammarians be accorded the scholars, who held all this knowledge and its tools).
or philosophers, was promoted by the written trace through its role The oral or, to use Jack Goody's terms, the contextualized use of lan-'
as intermediary between concept and reality, as well as by the setting guage was suitable to a regional and particularist society that was
down ofdiscourse on the page, showing relations as a whole that could more concerned with inherited forms of status than with the dynam- ,
be mastered. The written word, though it might be precarious and ics ofsocial roles. 5 Finally, it is a simple fact that oral language is con- \
scarce, was a basic value; acknowledged as such, it stood surety for servative, wherever or whenever it exists. Written language, by con-
the exchange of ideas just as money, whose parallel development was trast, as the appropriation of a decontextualized knowledge, can be a
spectacular, made possible commerce in merchandise and property. factor (one to be exploited, as Jules Ferry and the republicans did) in
The symbolic stakes in this passage to writing (or toward writing) progress and freedom. And though the view that writing runs through'
which affected medieval society are understood but still need to be all of medieval society is modern and anachronistic, it is inappropri- I
correctly evaluated. As it happens, the estimates ofwidespread illiter- ate to wax nostalgic over a medieval orality that was sincere, free, and I
acy in the Middle Ages and the dubious consequences drawn from of the people, as the victim of clerical and scriptural oppression. '
this have misled research somewhat less than the simplistic manipu- The appearance of writing in the vernacular provides the best
lation of the opposition between oral and written, no matter which proof of how vigorously this change in habits and references took
pole, moreover, was favored. Between the seventh and eleventh cen- place. The exclusive bond between Latin and writing became looser,
! turies, the vigor and significance of oral practices were the concomi- opening up--briefly at first, but the move was decisive-to the ver-
tant side of the tenuous preservation of writing. Nor should one nacular. This innovation, though usually reduced to an archeological
bring into consideration here the very short-lived opposition between curiosity, mere beginnings, scattered bits, should be granted its full
scholarly and popular cultures. Some writing was protected, techni- importance. It has been our usual habit to trivialize this inscription of
cal, and professional, but beyond that, human conduct as a whole, French, which, even if it is indirect and incomplete, represents a
from the most modest to the most official, as well as the transmission tremendous conquest: the vernacular language had to conquer a le-!
of social knowledge, had the spoken word for its basis, its reference gitimacy, and this conquest had to include the instruments of that le-
and its vector. gitimacy. (For the scribe this was no mean, grammatical feat: new
After the millennium, oral expression was certainly no less wide- sounds had to be transcribed into the Latin alphabet, the statement
spread, but orality began to lose its authoritative character. In other had to be broken into segments, etc.). The historical linguistics of
words, the written word robbed the spoken word of its authority but French attaches little value to this writing process, which it consid-
did so in a mode more typical of cultural revolutions, through shift- ered important only in that it finally permitted the establishment of
ing and replacing values rather than through breaks and upheavals. reconstructions on some attested language. Ignoring the fact that eth-
This was a decisive change, moving medieval society from an oral nologists have widely and confidently proven that the passage of ver-
and preliterate status to a mixture of oral and written (according to a nacular languages into writing has many consequences, historical lin-
• subtle typology of illiteracy, near literacy, and literacy). And it was an guistics puts few questions to the attestations it collects. Following an
irreversible change because it represented progress. Writing and read- extremely scholarly philology, its criticism implies that they are seen
16 17
as clumsy and scarcely reliable documents of language as it was spo- the need to Christianize the servants and staff in monasteries and
ken. This point of view forgets, on the one hand, that writing, which places of worship had some bearing upon the very first texts in the
decontextualizes language and makes it audible to more than one vernacular, and the great movement of pilgrim masses was accompa-
person, is always in common usage and very quickly judged to be nied by the development of a literature, as the manifestation of a cul-
"good," because it surrounds itselfwith prestige: these few traces pre- tured audience (think of the many scenes of familiar reading offered
served in folios are the first monuments of literary language, provid- us, like an infinite regression, by twelfth-century romances). Im-
ing evidence of the earliest workings of writing. On the other hand, proved economic and cultural conditions and the sorts of experi-
such a point ofview underrates the extent to which the written word mentation that this promoted deserve, nonetheless, some general
turns around and influences language itself. Basically, no one has mention; they explain why French was the first of the vernaculars to
studied this influence; the idea of doing so is beyond historical lin- embark on the adventure of the written word. For at least two cen-
guistics (for which writing is merely evidence and proof); it is un- turies it was in the northern Gallo-Roman lands that everything was
thinkable for those who believe that the essence of medieval cultures happening. Think of the real melting pot surrounding Saint-Denis
lies in their orality; it goes against good sense (reminding us how rare for forty leagues or so between 1130 and 1270. That was where West-
this written vernacular is). Studies far and wide have confirmed, how- ern philosophy and architecture were forged. The first texts com-
ever, that such an influence exists, and French provides such a good posed in written French constitute the laboratory of writing in the
1111" illustration of the phenomenon that it deserves to figure in textbooks medieval vernacular; it is not simplistic, therefore, to choose them for
of ethnology-an illustration so overwhelmingly obvious that no one our subject. Finally, on the symbolic level, which is what is most at
has noticed it. stake here, the spectacular rise of the written word, written Latin,
No matter what their orientation, treatises on the subject ofFrench must be singled out by itself as it swept the vernacular along with it.
phonetics are in agreement in dating the major phenomena that give It has frequently been remarked that French inscription took its first
French its specific phonic color (the dropping of unstressed vowels, steps in the shadow projected by Latin writing. The Serments de Stras-
the dipthongization ofstressed syllables, the sounding then dropping bourg (842) were cited in Nithard's Latin Histoire; and the next works
of intervocalic consonants, palatizations, etc.) between the third and (Sequence de sainte Eulalie at the end of the ninth century, the Passion
the tenth centuries. In the tables, diagrams, and synopses included in of Clermont and the Vie de saint Leger in the tenth century) were in-
these treatises, there is nothing afterward other than phenomena that serted into Latin manuscripts, where they sometimes filled a lower
are clearly only reductions and simplifications. In other words, the portion of the folio, which had been left blank. We should not, how-
phonetic evolution of French became stable and sorted itself out dur- ever, make too much of this subordinate position. These unfinished
ing the same period as it acquired a written form. This, we consider, folios, their value well known, were filled in not with just any old
is no coincidence, but rather the proof, disturbing perhaps, of the ef- Latin poetry but with an aesthetically worded native language. This
fects of the first written documents on proto-French. We need, there- wording must, therefore, have been considered precious, translating a
fore, to submit these effects to an inquiry, which we know will be si- successful endeavor-the incredible emergence of the mother tongue.
multaneously difficult, far reaching, and disconcerting. It was a phenomenon of the greatest importance, humble, no
Numerous reasons have been advanced to explain the develop- doubt, and obscure, but immensely powerful and venerable. One is
ment all at once of a written vernacular. Certain very specific cir- face to face with the written word, no longer just the words of me-
cumstances should not be overlooked. Thus, according to Auerbach, dieval Latin, that beautiful, cultivated, and professional language cus-
18 19
I
II
1"'
1
;
,
.f
'"
~!;fr'
tomarily used for presenting the Scriptures. But now the words were for example) at the heart of manuscript transmission; in dictation to
in the language used for how one approaches existence and comes to a scribe (not to mention the "interior dictation" of reading some-
understand its meaning, the language that was the basis of interde- thing-probably aloud, moreover-and writing it down slowly); and
pendence, the one used to express most deeply one's desires and suf- in the convivial and almost professional reading of nearly all of this
ferings. The mother tongue, for the first time, confronted all the risks literature (as a playwright nowadays writes words to be spoken). But
and possibilities of everything that literature specifically is. What ba- all these interventions by the oral were fragmented, kept at a dis-
sis exists for the use of this term, however precarious its object, lies in tance, and confirmed in their heterogeneity precisely by the activity
the conflictual contact between meaning linked to the constituent of writing. Whether or not· the work was composed and transmitted
language as it is practiced and the constraints and possibilities of its orally, whether or not it was later interpreted, it was linked, for a mo-
written formulation; in the conviction that this will result in a spe- ment, to the gesture of the hand, it yielded to the linearity of the
cific aesthetic experience that is new in process as well as in effect; in graphic signifier, and it confronted the decontextualization and vir-
the desire to preserve this experience, varying and going back over its tualization of meaning, its variation and deferral, its diffirance. Here
effects. When the mother tongue learned writing (in its inseparable we are putting our finger on the specificity of the first literature in the
technical and aesthetic senses), it involved a double break on its part. French language. This was a literature not yet forced into the shack-
First, it broke with Latin and very rapidly emerged from its les of established forms of the written word (the author as the tute-
shadow. It may be noted that, even in its first expressions, French lit- lary origin, textual stability, etc.), which were very late in coming, as
erary language demonstrated very few Latinisms, and its stylistic we have seen, and it shows us in an exemplary manner the euphoric
forms (the decasyllabic and octosyllabic lines used at first; then the appropriation by the mother tongue of the gesture that transcends it.
prose that appeared at the end of the twelfth century) owed nothing This appropriation found expression in an essential variance,
to Latin forms. More generally, we can see that its models were not which philology, modern thinking about the text, took to be merely
taken from Latin literature (ancient or medieval); French literature a childhood disease, a guilty offhandedness or an early deficiency of
invented its genres, from the epic poem to a form destined for some scribal culture, whereas the variance was, quite simply, joyful excess.
success, the romance.
Second, it broke with orality. Though a symbolic break, it matters
The Alterity ofthe Manuscript
to us. Of course, one can surmise that orature in the vernacular lan-
guage, though losing some of its legitimacy, remained very vigorous Everything about medieval literary inscription seems to elude the
and produced an abundance of works: for a long time to come (for- modern conception of the text, of textual thought. Still, if one con-
ever, probably, in poetic practice), aesthetic experience in the mother siders this inscription to be part of out patrimony, informing us
tongue remained physically articulated. Actually, human speech in- about an aesthetics, or continuing to provide enjoyment, we need to
tervened in various ways at the very heart of this attempt at literature make it available to readers by publishing it, that is, by editing it. The
proper with which we are concerned. And it did so in very different nuance is important, because we shall take care to refuse the always
ways; this literature was traversed by multiple voices: in the oral com- latent temptation (from Romantic typography to the love of first
position of certain chansons de geste (to the extent that this genre did drafts demonstrated by certain literary geneticists) of the facsimile.
not mimic in writing the formulas and processes of orature); in a In the often stated and sincere desire not to reduce what is be-
stage of oral transmission which occurred for certain texts ifabliaux, lieved to be the immanent truth of the object but to deliver it in
20 21
abundance, the alluring facsimile, which represents an abdicated re- vice and practices of the workshop (which certainly should not be
sponsibility for thought and, in this case, a dreadful illusion, is turned minimized) as their only basis, scribes had to devise a segmentation.
over to the machine. There is no end to the boundless desire for a This was the first and minimal act required for putting French into
"restoration" and an imaginary possession (from the reproduction of writing, for the accompanying notion of regular recurrence (not to
colored writing to the vellum format). The perfect facsimile is some- say norm) and for the empirical reflection on the part of grammari-
thing our of Borges-it is the original work itself. Editing, on the ans implicit here. The results were extremely disparate and long con-
other hand, is choice: one must cut and know the reasons for this tinued to be so; certain perceptible tendencies, however, are worthy
challenging gesture. In other words, given the difference between our of note if we watch out for them and have some confidence in the
established attitude toward the written word and that of the Middle scribe. The two manuscripts of Marie de France's Lai de Lanval thus
Ages, it is important to keep whatever in the medieval disposition demonstrate tendencies that are divergent, rather regular, and quite
makes sense and ought to be saved at a possible cost to everything considerable.
else. The editor chooses what he considers to be the specificity of the Manuscript H Manuscript C Translation
work, what is for him its truth, and makes it understood. That being Anglo-Norman Anglo-Norman
mid-13rh cemury end of 13rh cenrury
the case, it is clear that every edition is based on a theory--often im-
plicit-about the work. The methodology of editing, no matter what net ensouvient ne lensouvient did not remember him
net enpeseit ne lenpeseit did nor rrouble him
work it is applied to, always puts a literary theory into practice. From nel apele ne lapele does not summon him
ne1 empleinsist ne len plensist did nor piry him
this point of view, medievalist philology, which is based on textual nel alasse ne lalasse would nor go
thought, has opted for maximum reduction of the manuscript to a del amur de lamur of rhe/wirh love
del esrencele de sa estencele wirh rhe/wirh irs spark
contemporary textual object and to its accompanying notion ofliter- un kil ot servi un que lout servi one who had served him
de ces kil our avilee de ce que lour avilee because he had oUffaged her
ature. sil apela se lapela and called ro hiln/her
When the fantasy of absolute reproduction or the solely docu-
mentary project of fidelity to the manuscript (which is not a reading) Pronouns and articles are considered to be clitic in both manu-
is abandoned, from that moment on, editing becomes an activity scripts; they are almost always enclitic in H, proclitic in C. Whereas
with increasing choices and privileges. Thus simply transferring the a study of the syntax may discover something to think abour here,
inscription on the manuscript into the type characters of our Roman this sort of layout obscures any other way of looking at it. Such an
alphabet ("diplomatic" editing), necessarily entailed some graphic arrangement, dividing the forms in a graphic series, so that this is
representation of the wording (breaking it into words), and punctu- how the reader first perceives them, is more strange and disconcert-
ation to be legible. Words had been cut apart in Latin manuscripts ing than what makes each of the old forms and the modern ones dif-
starting in the eighth century (scriptio continuo being hard to deci- ferent from each other (peseit for pesait, apele for appel/e, etc.). Con-
pher), which no doubt represented some progress in silent reading in sequently, this needs to be scaled down. The editor, Jean Rychner, did
private. More profoundly, this segmentation was quite certainly con- so by joining elements that the scribe differentiated:
nected with the Carolingian revival of the teaching of Latin, which prefix and root: mes parla
we know consisted mainly of morphology. In the absence of any intensifier and intensified: i tant
grammatical instruction in the vernacular language, and with the ad- composite preposition: de dans
22 23
.'
'116'"
~~f
Mr. Procrustes, Philologist
and differentiated what had been joined: Edited lIersion

C'est la vertuz que Ii dons de sapience plante ou cuer contre
determiner and substantive: laventure [the adventure, l' outrage de gloutenie, quar sapience enseigne sobriete, si comme
the event]
dit Salemeons. Sobrietez est uns arbres moult precieus, quar il
samur [his / her love]
garde la sante de l' arme et dou cors, si comme dit Ii Escripture. Et
preposition and substantive: despine [of the thorn] de gloutenie et d'outrage de boivre et de meingier vienent souvant
This editing, which imposes the canonical segmentation of con- moult de granz maladies, et souvant la mort soudeinnement, si
temporary written French upon medieval discontinuity, which is comme l' en prant Ie poisson al' ame<;:on, c'est adire Ie morsel en la
bouche.
taken to be random, recalls somewhat, all things being equal, the dif-
ficulties encountered in editing a tape recording of spoken French, a Translation
signifYing series that we perceive as opaque. 6 If the act of editing im- [It is the virtue that the gift of wisdom places in the heart against
poses the modern definition of the word, it does the same thing on a the sin of gluttony, for wisdom teaches sobriety, as Salomon said.
higher level with the definition of the sentence (which, in the final Sobriety is a very precious tree for, as the Scriptures say, it protects
analysis, is merely what one prints beginning with a capital letter and the health of the soul and the body. Great illnesses and often sud-
ending with a period). What is disconcerting to modern eyes is not den death frequently result as much from gluttony as from excess
of drink just as one catches a fish with a hook, that is, the morsel
so much that words have not been separated as that they have been
still in its mouth.]
separated differently. In the same way, medieval manuscripts seg-
mented the broad units of French wording but according to a logic
We can see that the sign * corresponds sometimes to our period
displaying its difference right from the start. We could compare, for
(ending a sentence), sometimes to our comma (ending a proposition
example, this (semi-) diplomatic transcription (with the words sepa-
or a long constituent), and sometimes breaks something that seems
rated) of an extract from the Somme Ie roi by Brother Laurent (1279),
to us a syntactic connection (plante * ou cuer). Rhythmic punctua-
which Jacques Stiennon provided in his manual of paleography, with
tion, it separates parts of the statement, (syntagm, proposition, or
one possible way of editing this extract. The manuscript under con-
phrase), cutting apart the "elements of reading" with a logic that can
sideration makes use of several capital letters and a period (tran-
be examined only after having rid oneself of the presuppositions and
scribed here as *) half-way up the characters.
prerequisites of phrasal syntax.7 Reading, however, accustomed to the
Diplomatic Transcription structure provided by commas and periods, becomes disoriented and
*C'est la vertuz que Ii dons de sapience plante* ou cuer contre out of breath because it cannot relocate its usual pattern. Conse-
l'outrage de gloutenie* quar sapience enseigne sobriete si comme quently, the editor needs to punctuate, though indeed with a light
dit Salemeons* Sobrietez est uns arbres moult precieus* quar il touch, thoughtfully and sparingly. It is clear that editing obscures the
garde la sante de l' arme et dou cors* si comme dit Ii escripture* et segmentation of medieval literary exposition, at the level of words as
de gloutenie* et d'outrage de boivre et de meingier vienent sou-
well as at that of larger elements, whether it is hesitant or innovative
vant moult de granz maladies* et souvant la mort soudeinnement*
or follows a model. Thus, for certain forms of research, innumerable
si camme l' en prant Ie poisson aI' ame<;:on* c'est a dire Ie morsel en
la bouche. (Frere Laurent, La Somme Ie roi. Paris: BN, fonds facts are lost, not to say the object itself. This cannot be made avail-
fran<;:ais 938, fo1. 145) able to readers (except at the expense of making the published text
24 25
obscure for the benefit of a few) and can hardly be put into a note and in a manner deserving of interest. This being the case, what is
(except by adding to the impenetrable jungle of the hyperscholarly disconcerting to the modern eye (just as it was in the way words were
edition), yet one simply ought to be able to regain it. separated and punctuated) is not the absence of a paratext. It is,
The transfer, the manipulating and ordering of a medieval work rather, that its paratextual practice differs even more markedly from
that is represented by modern publication, can be seen more gener- ours than does that of Latin manuscripts but gives no signs of doing
ally when a scribal inscription becomes a book. The materialization so. From that point on, editing very clearly and concretely must re-
of a text intended for readers-which for us, out of cultural necessity, duce what is before it to the model of the modern paratext, that is, to
amounts to making a printed book-obeys rules that bring into play textual thought. An extremely simple example of this may be seen by
a finite set of pertinent elements (from format to author's name, from going back to the Lai de Lanval and the very remarkable edition of it
epigraph to subheadings) whose typology and functioning were de- that Jean Rychner provided. 9 It consists of a small book made up of
,i/ scribed very well by Gerard Genette in Seuils. A synchronic and this one text, which gives it its title. On the cover, in bold letters
structural analysis of the "paratext," it opens the way for a historical (called quite appropriately titulaires), the name of an author, Marie
survey of moments of genesis and rupture. In this history the situa- de France, whose work consequently goes on the shelf between Mal-
tion of French medieval manuscripts is exemplary from more than raux and Maupassant, and who takes her place in the homogeneous
one viewpoint. Initially, through their slow conquest ofliterary status and reassuring paradigm of French authors.
for the vernacular, they were less quick to innovate and stayed pre- The Lai de Lanval took the material form of a unique, entitled, at-
modern, so to speak, longer than Latin manuscripts, which, tran- tributed, and closed text occupying edge to edge the space of a book
scribing legal, philosophical, or theological texts, were closely linked (called, to be exact, a volume). Its medieval form was very different.
to the transmission of knowledge and, hence, to its appropriation. The Lai de Lanval, a work that we think of as a unit, appeared in four
For the latter, one can follow the appearance of a critical peritext in manuscripts with important and numerous variants. That is the first
keeping with the technological progress of the written word, progress difference-its essential plurality-to which we shall later return at
that the printing press would prolong: the alphabetical layout of the greater length. Next, the manuscript privileged by Jean Rychner, and
Elementarium doctrinae Erudimentum ofPapias (ca. 1053), the classi- upon which he based his edition, is a codex preserved in the British
fication by symbols, the separation of text and commentary, and Museum (Harley 978). Along with Lanval (fo1. I54v-I59v), it con-
then, during the twelfth century, the invention of a functional layout: tains the twelve lays attributed to Marie de France (fo1. I39-I8I-ex-
running heads, subtitles in red, initials alternating blue and red, or ceptional in that it is the one "complete" manuscript from our point
decreasing in size, and so on. 8 of view); but it contains also a treatise on religious music (in Latin),
The genesis of this medieval paratext is a good illustration of the a Roman calendar (in Latin), the Sept signes de mort, the Synonymes
tabular and type foundations of "graphic reason": the written word is des plantes, recipes for medicine, a letter from Hippocrates to Caesar,
not simply a deposit of knowledge; it is above all an incomparable the Isopets of Marie de France, twenty-two Latin texts, the Doctrinal
means of classifying and retrieving it. In this sense computer word sauvage, the Bestourne of Richard, the Mariage du pere et de la mere de
processing, as a tool for information storage and search, is linked very saint Thomas Becket, and a treatise on falconry. ) f
basically to the written word. French manuscripts, whose task was Like most medieval codices in which French literary works can be
less to preserve information and references than to bring into being a read, the Harley 978 collects texts in a manner combining ill-assort-
language, participate obliquely in this development, more tentatively ment and composition following a logic that is, at the very least, hard
26 27
to detect. Until the arrival of the printing press, the medieval manu- [At the conclusion of this writing
script in form, and probably in function, was an anthology, a collec- That I have said and produced in popular language
tion. Always a higher unit than the work (whether this was a poetic I will name myself for remembrance's sake:
piece or one of the longest romances, which we bind into large vol- Marie is my name, and I am from France]
umes), the codex was the open space of a confrontation, a gesture As well as another passage from Espurgatore saint Patrice, which is also
that brought together. We know that the paratextual practice of the attributed to her:
codex, when it existed, was something entirely different, serving pri-
Jo, Marie, ai mis en memoire
marily to distribute and give rhythm to this space of writing. Thus
Le livre de l'Espurgatoire
the three other manuscripts in which the Lai de Lanval appears pre- En romanz, qu'il seit entendables
sent this rubric: A laie gent e covenables
(Ed. Warnke,~ 2297-23°0)
P c'est de Lanval [it is about Lanval]
S c'est Ie lay de Lanval [it is the lay of Lanval] [I, Marie, have put into memory
The book of Purgatory
C lei comence Ie lay de Lanval [Here begins the lay of Lanval]
In popular language so that it will be understandable
which we can certainly take as the ancestor of our title. Its function, And suitable for lay people]
however, as incipit clearly distinguishes it from the thematic (or rhe- "It is generally believed that these three Maries are one and the
matic) finality of the modern title overhanging a work installed in its same," Jean Rychner wrote in his edition (85), "a woman who lived in
absolute singularity by being made into a book. Finally, there is noth- England during the time of Henri IL" We are obliged to observe that
ing in folios I54V and I59V of the Harley 978 manuscript that attributes these three Maries exist only as a few brief textual effects. From these
the Lai de Lanval to an author named Marie de France; nor is there three separate marks of an intradiegetic-heterodiegetic narrator (to
anything formally attributing to her the collection oflays found there. use Gerard Genette's terms), an author has been created and en-
No title, no rubric; one just reads at the beginning of the first lay: dowed with the beautiful name Marie de France (constructed, we
must note, from start to finish on the basis of "Marie ai num, si sui
Oez, seignurs, ke dit Marie
Ki en sun tens pas ne s'oblie (Guigemar, v. 3-4) de France,") and turned over to the inexhaustible hordes of conjec-
tures and speculations.
[Hear, lords, what Marie has to say
Clearly, this edition, the creation of a canonical text, has taken,
Who in her own time does not neglect her talent]
enclosed within the space of a book, and assigned to a memorable in-
Critics connect this textual (and not paratextual) note to the epilogue dividual a segment of the written word, borrowed from an uninter-
of the Fables attributed to the same author: rupted column, running through several folios of a codex manu-
script. This has important consequences for us concerning the very
AI finement de cest escrit composition of the utterance as well as its language.
Qu'en romanz ai traitie edit,
In Rychner's edition, the Lai de Lanval begins:
Me numerai pur remembrance:
Marie ai num, si sui de France L'aventure d'un autre lai,
(Ed. Warnke, v. 1-14) Cum de avint, vus cunterai.
28 29
11];'","
" 'thf'C,
"\1,,
I 't
Faiz fu d'un mult gentil vassal: de Lanval, though they are not necessarily continuous. However, in
4 En bretanz l'apelent Lanval. fact, they are read continuously in folio 154 (we reproduce here the
[I The adventure of another lay, diplomatic edition provided by K. Warnke):
As it occurred, I shall relate to you.
lt was made about a very noble vassal: Laventure ke auez oie
4 In Breton they call it Lanval.] Veraie fu nen dutez mie
De bisclaueret fu fet Ii lais
Pur remembrance a tut dis mais
The construction of the sentence occupying the first rwo lines is Laventure dun autre lai
worth noting: the nominal group that is the verb's object is placed Cum ele auient v 9 cunt'ai.
first, followed by a circumstantial subordinate (with a pronominal re-
[The adventure that you have heard
peat); then comes the verbal group. There is a widespread belief that Was true don't doubt it at all
in Old French, word order was free; but the reasons advanced for this About the werewolf the lay was made
freedom, though sometimes those of grammar, are groundless (the For the sake of remembrance from now on
supposed declension ofsubstantives indicating in the earlier language The adventure of another lay
the syntactic functions that modern French distinguishes through a As it occurred I shall relate to you.]
fixed order in the wording), sometimes metaphors of childhood (the
genesis of the language) or of disorder (a mosaic of dialects, variance The passage gains a striking unity and becomes possible to analyze
among the manuscripts). It is our thinking that, on the contrary, the grammatically. The construction detaching and anteposing the nom-
order of the words in Old French was, basically, already that of mod- inal object group ("Laventure d'un autre lai / Cum ele avint") is def-
ern French. (What is different about medieval syntax is generally not initely a thematizing articulation, which picks up again the word
located where people thought it was.) In the subordinate, the an- aventure used earlier, where it functions, moreover, as an anteposed
teposition of the subject is constant; it is in the principal proposition subject ("Laventure ke auez oie / Veraie fu"). Adventure and Lay be-
that the anteposition of other elements, words detached and empha- come the highlighted theme of the opening statement of the Lai de
sized, is encountered. These various thematizing processes, which Lanval; there is no solution of continuity:
gradually were eliminated from the classical written sentence, are dis-
Laventure ke auez oie ...
cursive phenomena that are all the more visible in medieval French De bisclaveretfit Jet Ii lais ...
because it is a literary language. Therefore they are always meaning- Laventure d'un autre lai,
ful and have to be explained. In the passage just quoted, the nominal Cum ele avint, vus cunterai.
object group is, on the one hand, anteposed and, on the other, de-
[The adventure that you have heard .
tached by the pronominal anaphora of line 2, which often is the About the werewolf the lay was made .
equivalent of discursive articulation; if one adds the expression "autre The adventure of another lay,
lai" (other lay) from line I, one sees that this introduction is particu- As it occurred, I shall relate to you.]
larly anaphoric, turned towards something said earlier but which
eludes us. Now, in the Harley 978 manuscript, upon which this edi- What we see here is a unique and multipropositional discourse so-
tion was based, an "other lay" (called de Bisclaveret) precedes the Lai lidified by a dense nerwork of anaphora. For this discourse the edi-
30 31
!""'".
~ I
tion substitutes one autonomous and closed sentence, with its

anaphorical elements and functions disconnected. The Joyful Excess
We need to stress one other major aspect of the work by means of
which it concretely eludes our methods of approach and clouds
analysis whenever it is reduced to the printed form, though it can it-
self provide us with a wealth of material and constitutes something
that must be given greater importance and visibility because it is the
basis of this writing: that is, its essential variance. lO
In the Middle Ages the literary work was a variable. The effect of the
vernacular's joyful appropriation of the signifYing nature suited to the
written word was the widespread and abundant enjoyment of the
privilege of writing. Occasionally, the fact that one hand was the first
was probably less important than this continual rewriting of a work
that belonged to whoever prepated it and gave it form once again.
This constant and multifaceted activity turned medieval literature
into a writing workshop. Meaning was to be found everywhere, and
its origin was nowhere. Usually an anonymous literature, its ony-
mous state is a modern fantasy (we saw that the name Marie de
France was an invention of editors) or else an admirable medieval
strategy: the name Jean Renart was an expression of cunning, Chre-
tien de Troyes (the Troy of an ancient culture revealed to faith) ex-
pressed the cultural ambition of the Middle Ages, namely translatio
studii. In this way it is a literature that is in conflict with the authen-
ticity and uniqueness that textuary thought connects with aesthetic
production. Of course, genetic literature has somewhat undermined
the closed stability of the modern text as well as the simple and tele-
ological vectorization of its production; nonetheless, though it is use-
ful to associate a space of scribal elements (advance projects, drafts,
33
32
The Joyful Excess
revisions, etc.) with the text, this space takes its bearings and stitches that, starting in the eleventh century (i.e., at the moment when the
itself together at one point: the completed version, ready for the written word attained full development), textus was used more and
press, authenticated and authorized. It is the break in writing activity more exclusively to designate the codex Evangiliorum. Attested
through which the text happens and for which genetic literature is re- around 1120, the French word tiste, changed then to texte (a scholarly
ally a regret. The medieval situation is a fine example of the premod- word), means "the book of the gospels." This text was the Bible, the
ern. Consequently, it is disorienting to a philology that originated at immutable word of God that may, of course, be annotated, but not
the beginning of the nineteenth century when the text gained its al- rewritten. An utterance that is stable and finite, a closed structure:
most perfect and immutable reproduction, attested contents, and le- textus (the past participle of texere) was something woven, braided,
gal paternity. In the generalized authenticity of the medieval work, all interlaced, constructed; it was a weft, a framework. The past perfect
that philology could see was a lost authenticity. Medieval philology is form of the verb tisser, textus possessed a connotation of fixity, of
the mourning for a text, the patient labor of this mourning. It is the structural completeness, which textuary thought would then provide
quest for an anterior perfection that is always bygone, that unique with full semantic power, that is, as denotation. Medieval writing, on
moment in which the presumed voice of the author was linked to the the other hand, was a reprise; it put things together, constantly and
hand of the first scribe, dictating the authentic, first, and original ver- again and again wove works, worked endlessly, striving like those
sion, which will disintegrate in the hands of all the numerous, care- three hundred princesses who were prisoners and weavers, the riddle
less individuals copying a literature in the vernacular. It is the desire encountered by Yvain in a castle:
to reduce the troubling image of the other to a primordial sameness
which is endlessly afforded by the writing of variance through its ex- qui diverses oevres feisoient:
treme instability of detail. de iii d'or et de soie ovroient
That instability of medieval works in the vernacular is a clear il- chascune au mialz qu'de savoit.
(Chretien de Troyes, Ie Chevalier au lion,
lustration of what is particular to both the written manuscript and,
ed. Roques, v. 5189-91)
more generally, scribal culture. As Gerald Bruns points out, in such a
culture notions of originality, imitation, and plagiarism would have a [who were working at various tasks
they were working with thread of gold and of silk
completely different meaning from the one they would have in the
each one doing her best.]
culture of the printed word (besides which, we should note, they
make no sense at all in an oral culture, in which the work is ab- A writing workshop. It was later understood, a trait particular to
solutely its enunciation).! The work copied by hand, manipulated, scribal culture, that originality for such an aesthetic lay more in the
always open ~;d~ g;--;;d as unfinished, invited intervention, annota- form of the narrative than in what was narrated. 2 It was a formal lit-
tion, and commentary. Confronted with an earlier piece of writing, erature throughout, a writing that grew out of itself; and that was its
it constructed itself and sustained itself simply with the distance it as- greatness and its joy, inventing its forms and playing with them on
sumed in relation to the urterance that was its basis. The scribal work the basis of some earlier utterance. Everything had always already
was commentary, paraphrase, supplementary meaning, supplemen- been said. By the ancients: translatio studii, textual translation, and
tary language, brought to bear upon a letter that was essentially un- the proud progress ofwriting in the vernacular-through some pred-
finished. We can see that the term text is hardly applicable to these ecessor whom one went back to, continued, pretended to finish. The
works. There was only one text in the Middle Ages. Du Cange notes medieval corpus is composed of innumerable Continuations (Contin-
34 35
IN PRAISE OF THE VARIANT The JoyfUl Excess
uations de Perceval, Roman de la Rose, etc.), explained by critics by memory, whose abilities were exercised in a way that is now incon-
means of biographical fantasies (one premature death after another), ceivable to us; oral practices to which we have become deaf but
without seeing that continuation was the first principle of this litera- whose effects moved into the written word, which amplified and
ture. Was not the predecessor, in fact, a necessary fiction for scribal made use of them; the power of rhetoric, a mental structure, the
writing? Finally we understand the astonishing repetitiousness of this habitus shared by the huge collection ofliterate, half-literate, or para-
writing. It has been thoroughly studied in the chanson de geste, which literate people who saw topical repetition as a creative act. In an aes-
is constructed according to an algorithm of stereotypical expressions thetics of return, where pleasure lay in variance, writing made minute
(formulas and motifs), bringing into play or staging the techniques of shifts in what was already known, and the acts of reading and listen-
oral composition and memorization. However, a formulaic style may ing lent themselves to the vicissitudes of recognition and surprise.
also be easily found at the other extreme, in eighteenth-century This variation is longitudinal, so to speak, running throughout
prose, though clearly this is a matter of the written word, one no the work, and though it bothers modern critics and troubles the edi-
doubt inducing a solitary, even silent reading: a closed system of ar- tor (terribly tempted to correct it), it does so far less than do the pro-
ticulation for the sentences of the narrative, a strict and calculable liferating lateral variants. Here the philological mind becomes seri-
grammar for the integration of discourse} What we perceive as the ously alarmed. Philology, created for editing the ancient and sacred
heavy-handed repetition of this prose participated in the aesthetic of Latin and Greek works that were reproduced especially during the
return which lay at the basis of vernacular writing. Middle Ages, is a measured and patient practice of comparison; it
No matter what genre is considered (from the epic to the fabliau, compares manuscripts separated only-this is axiomatic-by the
from verse narrative to prose fiction), if one pays close attention, rid- changes specific to the act of copying. When tradition (i.e., all of the
ding oneself of the modern scorn for needless repetition, a whole col- manuscripts that have come down to us) presents different readings
lection of processes, figures, and motifs whose sequence is meaning- (i.e., lessons from lectio: what one reads) at a certain point in the text,
ful becomes apparent. A humble formula for spring, an expression there is a variant (philology sometimes calls it innovation, as a re-
for riding a horse, a figure for surprise, speaking one's grief, and so proach), and one needs to makelsure which is the good text (for the !..-
on, all repeat and vary so that we should neither be surprised by them "good reading," etc.). When this practice turns to medieval manu-
nor have to work our some outrageous justification for them. The scripts in the vernacular, the philological automaton gets carried away
question we must ask of every work in the vernacular concerns the and panics. There is so much variability in number, extent, and na-
specific forms and the reasons for their occurrence. In other words: ture of the readings that the work is immense and success illusory;
"Tell me how you repeat yourself and make use of your repetition." the whole, vast undertaking seems maddening and humiliating. 4 A
Writing that was paraphrastic in essence, it thus presents the syntac- poor, rather disenchanted relative, blind to the positive effects of
tician with a wealth of material as soon as he takes into account that writing in action and thus ignorant of its treasures, Romance philol-
he is still dealing with literary works and their particular qualities. ogy attempts to salvage and promote some of the beauties of a litera-
This language that is worlds apart from us comes alive (its rules and ture that, basically, it considers to be a bit childish and carefree. In
flow) ifwe take it at face value. Such an aesthetics of return, the plea- good company, that is, among Bedierists, one cannot really escape
sure in sameness and difference, was the result of a number of phe- this scepticism except by taking a radically different perspective.
nomena: the slight influence of Latin models and, correlatively, the Variance is the main characteristic of a work in the medieval ver-
scrupulous and constant establishment of vernacular forms; human nacular; a concrete difference at the very basis of this object, it is
36 37
"r,·,',
IN PRAISE OF THE VARIANT The ]oyjitl Excess

something that publication should, as a matter of urgency, make vis- glo-Saxon audience is when confronted with the editions of Shake-
ible. This variance is so widespread and constitutive that, mixing to- speare provided by the cheerful proponents of the New Bibliography.
gether all the texts among which philology so painstakingly distin- These critics, on the one hand, demonstrate that the various edi-
guishes, one could say that every manuscript is a revision, a version. tions of Shakespeare published during his lifetime, the ones he could
Setting forth what seems to be a conventionally unorthodox prin- be said to have been in charge of, present very important variants
ciple does not, of course, keep one from later proposing a typology of (which philological edition, starting in the nineteenth century,
variants, if only to examine the degree and nature of variability spe- carefully erased). On the other hand, they explain that these variants
cific to each work, or each type of work. One can distinguish among reflect the mobility of a text constantly revised by its author (or the
the genres; thus, the epic, because it is closely or distantly rooted in person who published in his name) following performances (idiosyn-
oral composition, seems to carry with it the maximum number of cratic improvements, some good idea an actor had liked, etc.). Then
variants (the epic genre "exists only in variants," said Ramon Menen- finally they prove that to reduce this plurality into one unique and
dez Pidal). The seven complete manuscripts of the Chanson de supposedly established text loses something that is there in Shake-
Roland represent as many versions, realizations of this epic, and the spearian works (images, vocabulary, language characteristics).
genetic branchings that stem from an original archetype with which This is a critical notion, forcing one to think, for example, that the
philological editions are adorned are touching fictions. quarto (1608) and folio (1623) editions of King Lear are not two cor-
One can analyze the genre itself; thus, in his study of the fabliaux, rupted states of a perfect (and unfortunately lost) original, but two
Jean Rychner carefully brought to light the impressive number of in- versions of the same play, each one authoritative. 6 There are, there-
terventions (slight variations, innovative copies, revisions, etc.) to be fore, two King Lears (that need to be laid out in a parallel edition),
found in this corpus.5 The meaning and value (whether the effect is and one of these distinguished iconoclasts authenticates five Mac-
truly literary or rubbish, etc.) of these can be gauged: "in these oper- beths. Shakespearian writing is no longer presented as a closed, origi-
ations everything-style, versification, narrative organization and co- nal, and seminal utterance; it is constant and multiple production.
herence, intention and spirit, can all be changed" (132). When our literary presuppositions have become sufficiently un-
Finally, one can examine a single work in itself Such a study is si- hinged that Shakespeare is affected, it is not hard to conceive of the
multaneously one of the most frequent (it accompanies the working disturbance gradually spreading to most of premodern writings.
out of a critical text) and extremely rare. A work is usually examined There are certainly two Percevals, at least. This romance, it should
from every angle to retain its "good readings" and reconstitute an ar- be noted, could provide a good counterexample to our assertion. It
chetype rather than to observe the semantic and textual drift affect- was very famous and its influence was far-reaching as well as long-
ing it. Doing this would break up the unity of the work, the unique- lasting; it is reasonable to think that it was copied with great care
ness of its composition, the perfection of the masterpiece and would (which, in a sense, is true). However, it does fall within the province
traumatize our notion of the text. It is hard enough to get used to the of this writing of variance which we are discussing. Because variance,
idea that there might be more than one Chanson de Roland, all of until the end of the thirteenth century, was the basis of the medieval
them authentic. But does one have to put up with, for example, sev- literary aesthetic, this aesthetic is the perfect antithesis of modern aes-
eral true Percevals by Chretien de Troyes-the most famous romance thetics of the text.
of the European Middle Ages? Think how appalled a cultivated, An- The fifteen manuscripts that have come down to us constitute a
38 39
\
IN PRAISE OF THE VARIANT The JoyfUl Excess
"particularly confusing and obscure" tradition, as Felix Lecoy de- Manuscript T

scribes it in his edition (99), leading us to understand that scribal ac- Atant clui autre vallet vindrent
tivity was particularly heavy (in philological terms: crossings, bor- Qui candeliers en lor mains tindrent
rowings, contaminations, exchanges, rewritings, and alterations). De fin or, ovrez a neel.
None, of course, is autographic; all (and this remains true until the Li vallet estoient molt bel
Qui les chandeliers aponoient.
fourteenth century) came later than the hypothetical date of what is,
En chascun chandelier ardoient
supposedly, the first writing. This break obviously emphasizes the ap-
Dis chandeilles a tot Ie mains.
propriative process of copying. Like all literary works in the vernacu- Un graal entre ses deus mains
lar, Perceval was rewritten prolifically. From our point of view, this Une damoisele tenoit,
rewriting is not basically different from the numerous continuations Qui avec les valles venoit
to which it was subjected (Chretien's work was unfinished, which, for Bele et gente et bien acesmee.
that matter, resulted more from an ethics of writing than from a sud- Quant ele fu laiens entree
den death); nor is it any different from the prose versions or the cycli- Atot Ie graai qu'ele tint,
Une si grans dartez i vint
cal revisions affiicting it in the next century. All make the same ges-
Qu' ausi perdirent les chandoiles
ture of completing the sense.
Lor darte come les estoiles
The intrinsic plurality of Perceval is demonstrated extremely well Font quant solaus lieve ou la lune.
by the availability of two commercial editions of it. The one provided (Roach, ed., V. 3213-29)
by William Roach (Geneva: Droz, 1956) follows manuscript fro 12576
Manuscript A
in the Bibliotheque nationale (MS T); Felix Lecoy's edition (Paris: Et lors dui autre vaslet vindrent,
Champion, 1972-75) conforms to manuscript fro 784 of the BN (MS qui chandeliers en lor mains tindrent,
A), copied at Provins around 1230 by Guiot the scribe, beloved son of de fin or, ovrez a neel.
the philologists (because, really, he was the first one of them). When Li vaslet estoient molt bel
one studies this manuscript, it becomes clear that it is two hundred cil qui les chandeliers portoient.
and seventy four lines shorter than manuscript T and has at least five An chascun chandelier ardoient
hundred variants that are worth noting, that is, one change every X. chandoiles a tot Ie mains;
un graal antre ses .11. mains
eighteen lines.? This is not very many (Jean Renart's Lai de I'Ombre
une dameisele tenoit
has a variant every three words), and no doubt we are at the limit of et avoec les vaslez venoit,
variant writing-but still well within this limit. One becomes con- bele et jointe et bien acesmee.
vinced of this after comparing the most intense, almost sacred, pas- Quant ele fu leanz antree
sage of the work in both manuscripts. Because it is the passage that a tot Ie graal qu'ele tint,
sets one to dreaming and whose heritage is immense, it is the passage une si granz dartez an vint,
in which one would expect the copyists to exert their maximum ac- ausi perdirent les chandoiles
curacy: the first vision of the Grail granted to young Perceval. Ignor- lor darte come les estoiles
qant li solauz lieve, et la lune.
ing graphical variations, the following shows in italics the variants (Lecoy, ed., V. 3201-17)
that are strictly textual:
4° 41
The Joyful Excess
Translation [A young woman held,
[Then two other young men came Who was coming with the young men]
Who held candelabra in their hands
Of pure gold, wrought with black enamel. In the places where manuscript A detaches the antecedent and em-
The youths were very courtly phasizes it by repeating it again with a demonstrative:
Who were bringing in the candelabra.
Li vaslet estoient molt bel,
In each candelabrum were burning
cil qui les chandeliers porroient. (A)
Ten candles at the very least.
A grail between her two hands [The young men were very courtly,
A young woman held, Those who were carrying the candelabra.]
Who was coming with the young men,
Courtly and noble and beautifully adorned. or else coordinates:
When she had entered inside une dameisele tenoit
With the grail that she held, et avoec les vaslez venoit, (A)
Such a brilliance came in there
That the candles thus lost [a young woman held,
Their own brilliance just as the stars and she was coming with the young men]
Do when the sun rises or the moon.]
ManuscriptT expresses consecutive relation with the conjunction que:
The translation is of MS T. The variants in MS A will be translated be-
low as they are cited. Trans. une si grans dartez i vint
Qu' ausi perdirent les chandoiles (T)
There is, of course, no fundamental modification (along the lines
of those listed by Felix Lecoy in the appendix to his edition), to be [Such a brilliance came in there
That the candles thus lost]
gleaned from this seventeen-line section, but there is enough to
destabilize somewhat the textual sequence and permit us to sketch where A uses a simple paratax:
out a hypothesis covering specific instances and their recurrence.
une si granz darrez an vint
Thus it appears that manuscript T strongly articulates the sentence, ausi perdirent les chandoiles (A)
employs more explicit and regular syntax than manuscript A (Atant,
MS T; Et (ors; MS A; atant being the most dearly beloved initial word [such a great brilliance came from it
thus the candles lost]
in thirteenth-century prose), and makes strict use of relative con-
struction: In T the construction of comparison is explicit, making use of the
Li vallet estoient molt bel substitute verb faire (to do):
Qui les chandeliers aporroient. (T) Lor darre come les estoiles
[The youths were very courtly Font quant solaus lieve ou la lune (T)
Who were bringing in the candelabra.] [Their own brilliance just as the stars
Vne damoisele tenoit, Do when the sun rises or the moon.]
Qui avec les valles venoit (T)
whereas in A the expression is less heavy-handed:
-- 1 '
2
4 • ',; 43
IN PRAISE OF THE VARIANT The Joyful Excess
Lor darte come les estoiles this we can reply, with about the same level of presupposition (and,
qant Ii solauz lieve, et la lune. (A) ultimately, less anachronism) that the great writer of medieval ro-
mances is the one who plays subtly with the canons, the models, and
[Their brilliance, just as the stars
when the sun rises, and the moon.] the topoi that he has inherited. Clearly, this debate is absolutely
pointless. Because the question lies elsewhere.
(Perhaps the expression "ou fa lune" in manuscript T may be seen as The one concrete fact about literary objects in the vernacular is
additional evidence for construing it as the "canonical" text. It is cer- their excessive number of variant manuscript forms. Any study needs
tainly more what one would expect than the turn of phrase "et la to be prefaced by a question somewhat like the one any linguist must
lune" in manuscript A.) But deciding which of the two syntaxes il- ask when confronted with the heterogeneity that makes up a lan-
lustrated by manuscripts T and A reflects the supposedly original guage: "How can one manage all this?" Whether one is a grammarian
writing is both impossible and irrelevant. Siniilarly, there is no sense or has some reason to be interested in the medieval esthetic use of the
in wondering if this first version conformed with the canonical por- vernacular, a sceptical and careful attitude concerning every edition
trait of feminine beauty ("Bele et gente" [courtly and noble], MS T) should be adopted, inasmuch as the editor has always "managed it
or the more innovative one (bele et jointe" [courtly and ?], MS A): all," reducing the excess ofwhat he was copying to the unity of a text.
analysis must be comparative, not archeological. This is all the more Taking something and putting it into a book cuts into a continuous
true because archeology reduces something that derives its meaning flow, as we saw earlier, and in this case, it is a multiple operation and
from difference into something that is just one. Thus it is important more Draconian in other ways. What is eliminated from the read-
to keep the distinction between the supernatural light suddenly en- able, not to say thinkable, is the surplus of text, language, and mean-
veloping the scene and the Grail: ing. One can well imagine how interesting methodological debate in
this area might be. 8
Atot Ie graal qu'ele tim,
Vne si grans dartez i vim (T)
[With the grail that she held,
Such a brilliance came in there]
and the light welling up from the sacred vase:
a tot Ie graal qu'ele tim,

une si granz dartez an vim (A)
[with the grail that she held,
such a great brilliance came from it]
In any literary or mythographic analysis this difference is important.

Critics, one imagines, saw in the ease, the light hand, and the orig-
inality of manuscript A (Guiot's copy) proof that it was good evi-
dence of the ease, the light hand, and the originality that belonged,
by definition, to the great writer of romances, Chretien de Troyes. To
,
« j ~l
45 •
4I
Gaston Paris and the Dinosaurs
polygraph who, at the age of twenty-six, living on a stipend provided

Gaston Paris and by Guizat, whom he had won over, got hold of the Chanson de
Roland in the Bodleian Library at Oxford (Francisque Michel, Chan-
the Dinosaurs son de Roland ou de Roncevaux, 1837). This period saw the first insti-
tutional involvement (Francis Guessard launched the collection "An-
ciens Poetes de la France" in 1856, underwritten by the imperial decree
of 12 February). Editing, however, was a matter of taste, the respect of
novices, and the scrupulousness of students formed at the Ecole des
chartes (the recently founded Parisian school from which students of
medievalism-a new discipline-now graduated; Latin and Greek
philologists had a more traditional education). Editing was, in shoft,
a matter of an eminent Romanticism. The language, which me-
dievalists were not sure they understood, was hardly ever corrected; a
manuscript they were so happy to have was hardly one to fiddle with,
and in any case, its freshness represented their literary sense of the
The history of methods that have been proposed for editing medieval Middle Ages. Whenever their sources differed, they stuck with their
texts in the vernacular is strangely full of sound and fury. It is symp- favorite manuscript, even if this required notes or appendixes con-
tomatic of the stakes in laying the ground for what will be the subject taining a multitude of variants (in 1835, Polycarpe Chabaille pub-
matter of grammatical and literary discourse, making it available to lished Supplement, variantes et corrections to Meon's edition of the Ro-
readers and informing this reading. But these activities are also se- man de Renart). Or else they would provide several versions (as when
cretly formed by this subject matter: textual criticism, a dry and Eugene Hucher published the texts of the Saint-Graal in prose in
rigourous discipline, is the unspoken praxis of literary theory. And 1875), inviting the reader to form his own judgment in an edition
philology here is a force to be reckoned with: behind its mask of that looked more like a file. These works were respectable, and more
naivete it manipulates the facts. than one of them can still be used.
The Empirical The Positive

The first distinguishable period (1830-1860) in this recent history That remark would have brought a smile to philologists' faces in
does not deserve to have been plunged into a state of nothingness by the third quarter of the century because of a deep split marking off a
the generation of savants. It was a time, of course, of empirical edi- second period (1860-1913) in the history of medieval philology. There
tors, but these were editors who, when confronted with the alterity of were many different factors behind this break. Systematic and scien-
the manuscript, were not content merely to reproduce editions tific perusal oflibraries turned up vast numbers of manuscripts, mul-
printed in the sixteenth century, as had been done in the preceding tiplying the sources of each work and forcefully raising the issue of
century. They were pioneers who frenetically published a literature how to edit them. A medievalist establishment developed: chairs were
discovered by Romanticism, ones like Francisque Michel, a brilliant created (College de France, 1852; Sorbonne, 1883) as well as revues
'I;
46
tl ~
--- J. _ _ _ _ _ _ _ _- - 0 1 , /
IN PRAISE OF THE VARIANT Gaston Paris and the Dinosaurs
(Revue critique d'histoire et de litterature, 1866; Romania, 1872), and studies, made use of comparatist methodology, the desire for recon-
societies (Societe des anciens textes franlfais, 1875), not to say institu- struction, but also the feeling of decadence. ("I walk in a field of
tions (Ecole pratique des hautes etudes, 1868). Finally, though late by ruins," said Schleicher, and Lachmann could have said it.) Lach-
at least a generation, French medieval philology, having been highly mannian philology, a mechanical archeology of the lapsus, began by
motivated after 1870 by a fierce desire to avenge Sedan, caught up automating the scribe, who was denied any positive and conscious in-
with German philology. The epistemological unity of these years is tervention. This philology could not, without collapsing, think that
striking; while linguists finally had a sure method of classifYing Indo- a scribe, when confronted with an uncertain reading, for example,
European languages and, like paleontologists, took on the task of re- might be able to improve it or indeed even rediscover the "original"
constructing the primordial and perfect Ursprache, philologists reading. Otherwise, from that point on, the linear filiation demon-
adopted a rational method of classifYing manuscripts which pro- strated by constant deterioration would become unclear. The scribe
moted the reconstruction of the archetype Urtext. It was a visionary was a machine, and this machine had to function poorly in order for
reconstruction created by this half-century, in the shape of a tree the multiplicity and the excess of variants to fall into place, showing
fraught with lopped off boughs where improbable monsters sheltered the slippery slopes of adulteration and delineating the genealogical
their wretched, disparate nature. branches of the manuscript family. Philology is a bourgeois, pater-
The major figure in this field was Karl Lachmann, the mythic fig- nalist, and hygienist system of thought about the family; it cherishes
ure of a great, political, and worldly professor, an impeccable and ex- filiation, tracks down adulterers, and is afraid of contamination. It is
tremely arrogant Prussian scholar. By his influence as much as by his thought based on what is wrong (the variant being a form of deviant
work, synthesizing the methods used by Hellenist editors, Lachmann behavior), and it is the basis for a positive methodology.
(who died in 1851) spread a method to which he gave his name, thus Lachmann started from the principle that copyists do not sepa-
establishing the practice of "scientific" editing of old texts,! The rately commit the same error at the same place (though any minor
method consisted of reducing the editor's subjectivity as a factor in schoolteacher who has regularly required exercises in dictation would
the choice of the good readings and using an almost mechanical have been able to tell this prestigious professor at the University of
method for reconstructing the original. Advances ~ere achieved Berlin that certain linguistic sequences lend themselves to th~ same
through the importance and precedence accorded, in the words of mistake). The results were twofold. First of all, the "shared wrong"
the Humanists, to the recensio (comparative study of several manu- could only be an inherited defect, and consequently it even repre-
scripts) over the emendatio (one-to-one contact between a copy and a sented a form of filiation; in other words, the manuscripts that show
philologist): "recensere ... sine interpretatione et possumus et debe- a shared wrong were copied, and only the thing copied, once distin-
mus." Lachmann thus provided scholarly and renowned editions of guished from its copyists, was worthy of interest, because it allowed
the Greek text of the New Testament, of Lucretius's De Rerum one to come closer to the original. Second, the agreement among sev-
Natura, and so on-ancient and venerated texts that the scribes of eral manuscripts against one other was a sure indication that they
late antiquity and the Middle Ages had copied with respect. From represented the good reading, because the mistake was isolated, or
that point on, Lachmann assumed that these copyists were guilty rather, isolation was the mark of the mistake.
only of mistakes due to incomprehension, inadvertence, and fatigue Lachmann's method thus claimed to have the means of locating
and that these errors represented degradation. Every copy represented dependent relations among manuscripts and to show these relations
decline. This philology, in common with the first Indo-European by a genealogical tree (stemma codicum). Thanks to this tree model,
48 49
showing simultaneously an origin (the archetype) and the hierarchy Gaston Paris at the end of the 1860s into the bosom of French me-
of relative positions occupied by the various preserved manuscripts in dievalism. A daring innovation, it was doubly imported, first from a
relation to the archetype, one could, by setting "two (at least) against methodology thought up by the German enemy, whom Gaston Paris
one," reconstruct word for word (that is, drawing by turns on one or knew well (Paulin, his father, had sniffed "a little breeze from Ger-
another manuscript) the archetype, the original figure. many in the air," and had sent him to study with Friedrich Diez in
That was the critique of readings, and it then remained to the cri- Bonn).3 Then, also, it was a technique imported from a methodology
tique of forms to make the archetype homogeneous. Composed of conceived of for other ends and other texts. We have mentioned the
bits and pieces borrowed from manuscripts ofvarious periods-a de- intrinsic variance of the medieval vernacular work; even from Gaston
formed diplodocus that didn't hang together, designed by some mad Paris's perspective, which saw a lost authenticity, the scribe acted with
scholar-it had to be made closer to the supposed original. To do so a freedom that made him more a reviser than a copyist; and what
the language had to be unified according to what was generally un- might be called conscious intervention (supposed to be infinitesimal
derstood about the history of Greek and Latin: that is, specifically, it in old texts) played a huge and absolutely incalculable part here.
had to be rewritten in the language assumed to have been that of the From our perspective, which sees authenticity as generalized, the pro-
original (just as Schleicher wrote a fable in the Early Indo-European duction of a surplus of text and meaning is constitutive of medieval
which he had reconstructed). But any error that could not be rewriting in the vernacular. What Paris did flatly negated this speci-
paired by comparison would still be inaccurate in the homogeneous ficity: every variant was considered an error, and the positive method
perfection of the archetype thus reconstructed. The primordial mis- was applicable.
take committed by the first scribe, exemplary of the essential degra- Yet applying it required a forcible takeover. Bedier was the first to
dation represented by the copy and spread henceforth throughout spread the word that Gaston Paris, won over to the German method,
the entire manuscript tradition, remained to be found. set out to extend it to French literature of the Middle Ages. The real-
Here the philologist showed his mettle. Using his knowledge and ity was more complex and in another way more meaningful. Paris did
subtlety, he worked with the cunning of a seasoned sleuth, if we are not like Lachmann and made it clear in the text purportedly showing
to judge by Louis Havet's masterpiece, which at the same time repre- his adherence to Lachmannism. The text was a review of the (very
sented an insane amount of erudition, a treatise on criminology and Lachmannian) edition by Karl Bartsch of the Nibelungenlied. 4
a manual on the psychopathology of the scribe. 2 This positive philol- Let us skip the barbed remarks aimed at the person of Karl Lach-
ogy seemed to play the role of a very refined older sister (but one who mann (stubborn, disdainful, haughty, "no one dared contradict him,"
could be effective when necessary: it was as a philologist that Louis etc.). They only demonstrate through their Prussophobia that the
Havet defended Captain Dreyfus) in relation to the young and young Gaston experienced some difficulties with father figures. Re-
rapidly developing science, criminal anthropology, whether the latter member that he was the son of Paulin Paris, one of the founders of
was more influenced by psychology (Cesare Lombroso) or sociology French medievalism, and that he had the painful fortune of succeed-
(Gabriel Tarde) or had a more technical orientation (Alphonse ing him at the College de France. (This probably explains Gaston's
Bertillon). Moreover, we know that in the realm of fiction, as in desire for positive science, Paulin Paris having been the perfect exam-
Merimee's Lokis, the figure of the philologist could stand in for that ple of the proper Romanist, all pleasant intuitions and Romanticism,
of the detective. who had published, in 1830, a well-chosen date, a little Apologie de
That was Lachmann's method, introduced with great fanfare by lecole romantique.)
51
5°
IN P R A 1 S EO F THE V A RIA N T Gaston Paris and the Dinosaurs
It is important to note that the compliments Gaston Paris ad- for the original poem of the Nibelungen is an event of the greatest im-
dressed to Bartsch's edition were based on a sharp criticism of the edi- portance for science" (186).
tion of the Nibelungenlied published by Lachmann himself In fact, Clearly, apart from slightly undermining Wolfs hypotheses, Paris's
on the one hand, the Prussian scholar did not limit himself to demonstration proved that all the manuscripts of the Nibelungenlied
philologia sacra but also worked on the German Middle Ages, editing derived from a single original, which came first, and that Lachmann's
some of its most important works: the Nibelungenlied, 1826; the po- method was perfectly applicable to them. If, moreover, such a posi-
etry of Walther von der Vogelweide, 1827; Iwein by Hartmann von tive philology could be practiced in this limited realm of the epic, a
der Aue, 1843. Yet, on the other hand, Lachmann as a medievalist was fortiori it could be brought to bear on every other genre (lives of the
far from Lachmannian. s The complexity of manuscript traditions in saints, fabliaux, romances, etc.) from the Middle Ages. Therefore it
this realm and the endless variability of the works had made him was now a matter of vigorously Lachmannizing French medieval lit-
sceptical and circumspect; he thus contented himself with providing erature and demonstrating to the Prussians (and to his father) that
an image of the original, usually following a single manuscript the French could do it better.
(Bedierism before the term was invented?) and intervening quite Allow us a digression. It has been generally agreed since Bedier
freely and intuitively in it (the return of a Humanist judicium?). This (God save us from good disciples!) that it was Gaston Paris who in-
had been his approach to the Nibelungenlied. Conforming, on the troduced to France the Romantic theses attributing a popular origin
one hand, to manuscript ~ he espoused, on the other, the theories to epics (cantilenas); Joseph Bedier, for his part, in Ugendes epiques.
of Friedrich-August Wolf about the popular origins of the Homeric Recherches sur la formation des chansons de geste (I908-I3), strongly de-
epic. Convinced that the Nibelungenlied was only a collection of epic nounced this thesis and suggested that the French epic (especially the
popular songs, he used the typography of his edition to distinguish Chanson de Roland) be seen as a literary composition of the eleventh
the twenty "authentic" (and independent) songs that he had redis- century. The antagonism typical of the clash between Paris and
covered at the heart of this "rhapsody." Bedier is perfectly exemplified here; it cuts medieval studies as a
The edition by Bartsch (who was moreover a good Romanist and whole right down the middle.
a friend of Gaston Paris) was based on an entirely different postulate: Bedier, who succeeded Paris in his chair at the College de France,
that there was a single original writing that reunited in one single brought a new scientific paradigm into play and founded a new
poem, into one work, various different traditions or popular songs. school. Consider, for instance, Lucien Fouler's thesis, dedicated to
This original writing was supposedly composed at one time (around Bedier, in which he demolished the notion that the Roman de Renart
1140) and in one place (Austria). Above all, it was supposedly the had its origin in folklore and buried it, rather, in the intertextuality of
work of one individual (Bartsch suggested a doubtful Konrad de the clerical culture of the twelfth and thirteenth centuries). But what-
Kiirenberg). Gaston Paris was in complete agreement. For him the ever meaningful conflict may have existed, it was not really of that
epic was not something "entirely primitive"; though based on earlier order.
"heroic songs," it transformed and combined them into an organic Gaston Paris's ideas were different, as we have seen, and they did
whole, he claimed, which assimilated these elements and was quite not change from the time of his thesis (Histoire poitique de Charle-
different from a chance collection of pieces patched together (187)· magne, 1865) all the way through his mature teachings (Litterature
Consequently, "the demonstration of the existence of a single author franraise du Moyen Age, 1888). Two points summarize these ideas:
52 53
First, there is the fear of emptiness. Between what happened at Ron- vellously logical and harmonious poems were in flower, later set upon
cevaux (778) and the Chanson de Roland available to us (dated by by stupid men who rewrote them?" (Ugendes epiques, I: 332). Second,
Paris in the eleventh century), there must have been a persistent lit- Bedier then showed the greatness that epics attained as literature (es-
erature or rather, orature. Rejecting for political reasons any Ger- pecially Roland), with their solid composition, their sure and beauti-
manic influence on the primitive epic in the vernacular, Paris fullanguage, and their very elevated poetry, sometimes extremely cul-
nonetheless borrowed the idea of popular heroic songs, cantilenas tured in its inspiration (with Virgilian influences); these were the
sung by warriors. These had not been attested, of course, and they work not of mediocre jongleurs but oflearned scholars who, like Tur-
came along with all the pleasant fantasies of theories of oral trans- old in the case of the Chanson de Roland, could be considered real au-
mission, but they came at just the right moment to fill in a gap of thors. (In the beginning was the poet.) Paris and Bedier were in agree-
several centuries. For, and this is the second point, the theory ad- ment, therefore, on what was essential-the epic's unifYing textuality.
vanced by Gaston Paris was based on a notion of decadence. Here it What separated them was secondary-and generational. Gaston
clearly contradicted Wolf and Lachmarin: for the French Romanist, Paris's theory, with its leaning toward "fierce warriors," echoed Indo-
the period of cantilenas and heroic songs of a national poetry was Europeanism and reconstructivist positivism; it was extremely close
over by the end of the tenth century; the genre lost momentum; it to the lyrical pages devoted by Renan to the primitive epic. 6 Joseph
was collected by professionals, the jongleurs, who, with more or less Bedier represented a second stage and was a more "modern" figure at
talent assembled a work (the epic) out of songs that had come to have the heart of textual criticism. We shall find the same opposition in
no idea or link in common, which "had lost their interest, often even matters concerning the editing of medieval texts.
their sense" (Histoire poetique, 12). This work was doubtless less Thus it certainly was a forcible takeover when Gaston Paris made
grandiose, but it was unique; in other words, the Chanson de Roland all of French philology (until Bedier) adhere to a strictly Lachmann-
that we have was the result of a synthesis brought about by one jon- ian (or rather, super-Lachmannian) conception of medieval textual
gleur, a primitive composition entrusted to the written word, then criticism. Aware of scientific progress as both a premise and an
subjected to the workings of the manuscript tradition; the manu- epiphany, this young man was one of the small number of directeurs
scripts of Roland "seem to derive from a manuscript, and not from di- d'etudes appointed in November 1869 to the Ecole des hautes etudes,
verse oral traditions independently entrusted to writing" (Litterature just a few months after its establishment (31 July 1868).
franraise, 57). It was thus possible to apply Lachmann's method- The extent to which this new branch from the trunk ofthe old Sor-
somewhat in spite of Lachmann. bonne would surpass the hopes invested in it is well known. Open to
Bedier's criticism, in fact, focused on these two points. First of all, everyone, with no academic prerequisites, the ecole's mission was to put
~
he saw no need to invent cantilenas to fill the void of centuries. The into practice the new sciences: "exercizing young people in the prac-
memory of Roland was enough to give birth to an epic literature in tice of the labors of erudition, in critical and scientific methods," do-
il'l'
the eleventh century: a few legends, a sword and a horn they would 't ing this by means of a rejuvenated pedagogy that abandoned the au-
show you, preserved in the churches along the road to Roncevaux. thoritative lecture for conferencespratiques in small groups.? As Gaston
(In the beginning was the road.) Nor, consequently, was there any Paris remarked in the introduction to his edition of the Saint Alexis,
need at all to lament these cantilenas. "If chansons de geste developed, these small lectures "alone are able to propagate the methods effec-
did this necessarily represent a movement of order toward disorder? tively and create what we are most lacking, a scientific tradition" (vii).
Is one required to imagine a golden age in the beginning when mar- This insistence on method, on practice, and on the small group
54 55
was an obvious reference to the benefits and superiority of the Ger- made an ephemeral figure of an archetype appear from the jumble of
man Seminar, of which France had just become aware. It was in this lines, puzzled out word by word in the patient and industrious con-
context of intellectual, scientific, and pedagogical revival that Gaston viviality of the conftrencepratique. By definition, this archetype is un-
Paris listed his first course (January-June 1869), which he devoted to like the manuscripts that philology had at hand, the bits of parch-
the Saint Alexis, a saint's life from the eleventh century, an edition of ment to be looked at, the scattered remains where hands search.
which he and his students were preparing. Delayed by the war of Consequently, the fourth line of the poem (which expresses a lauda-
1870 (during which Gaston Paris, like several of his colleagues, gave a tio temporis acti that is, moreover, very pleasing to the philological
patriotic course on the Chanson de Roland), the project would appear mind: the world is no longer what it used to be) can be read in the
in 1872 in the "Bibliotheque des hautes etudes."8 Gaston Paris edited following fashion in all four manuscripts (L being for Gaston Paris
the eleventh-century version, and Leopold Pannier, "a student of the the closest to the original):l0
Ecole," worked on the revisions of the twelfth, thirteenth, and four-
teenth centuries. 9 L (12th century, England)
This majestic edition, whether one reveres or mocks it, is the ba- tut est muez perdut ad sa colur
sis of French medieval philology. Gaston Paris was conscious of this [everything is changed, has lost its color]
and presented at length (138 pages) the method he followed as if it A (12th century, England)
were a manifesto. He did so, moreover without even once mention-
tut est muez perdu ad sa culur
ing the name of Karl Lachmann. This "attempt at an integral restora-
[everything is changed, has lost its color]
tion of the poem" (vii) fitted into the mission of textual criticism,
whether aimed at productions of the Middle Ages or those of antiq- S (end of the 13th century, France)
uity. This comprehensive philology "has as its goal to rediscover, as si est muez perdue a sa valour
much as possible, the form that the work to which it is applied had [so is changed, has lost its valor]
when it came from the author's hands" (8); based on a transhistorical
P (13th century, France)
notion of the author-the origin of his work, whose meaning he
tot est muezperdu a sa color
controls-it was the response to a transhistorical demand for recon-
struction. For this, medieval manuscripts presented conditions that [everything is changed, has lost its color]
were, compared to works from antiquity, at the same time better Gaston Paris edited:
(usually being closer to time period of the author) and worse (because
of their numerous inaccuracies and "changes"). Nonetheless, when Tot est mudez, perdude at sa color
the shared errors and readings were compared, it became possible to [everything is changed, has lost its color]
set up a classification of the French medieval manuscript tradition,
a reconstitution that he further pressed home (even including a pro-
"delicate, complicated and painstaking work, but indispensable, one
nunciation) in the 1903 edition and to which he returned again in the
upon which all scientific criticism rests"(u).
inaugural series of the Classiques franrais du Moyen Age (19U, vol. 9):
In fact, with this treelike classification, the criticism of readings
was able to operate and, thanks to processes that are "virtually math- Toz est mudez, perdude at sa colour
ematical" (23), able to reconstitute the letter of the original poem. It [everything is changed, has lost its color]
56 57
The text sent to the printers was even further removed from the the old language is the marker of Romanism at the height of its de-
information in the manuscripts, because the archetype had suffered velopment. There are many reasons overdetermining the amount of
the consequences of the second stage of the Lachmannian method: attention focused on the question of morphology. Historical gram-
after the critique of readings, which reconstructed the archetype, mar found here its topic of choice. The existence of medieval declen-
came the critique of forms, which reconstituted the language. Gaston sion clearly demonstrated where the language belonged (remember
Paris devoted any number of pages to this second archeology, and it the arguments over the Celtic origins of French). Declensions were
is important to understand what was at stake here. the visible trace left by Latin structure in French. In a manner that
In order to establish itself as a science in the nineteenth century, was simple and calculable, it showed the diachronic evolution (from
linguistics had to provide itself, through a process of reduction, with six cases in Latin, to two in Old French, to none in modern French)
an object that was stable, simple, regular, and homogeneous. From and easily represented the alterity and the specificity of Old French,
this point of view Old French, to the eyes of the positivist scholars an inflected language, in relation to modern French. Moreover, since,
who set out to describe it, seemed to have a pronounced and consti- in historical grammar, morphology was an extension of phonetics (it
tutive heterogeneity. This was because, in addition to the endless tex- was phonic evolution that changed the forms so drastically) and pho-
tual variance offered by the manuscripts, there was an infinite varia- netics was the perfect homogeneous subject as well as one that could
tion in linguistic forms. The malleability of the earlier language set be formulated, declension made it possible to think about the ques-
grammarians to thinking, as soon as they understood that this was tion of syntax-always a stretch for historical grammar-through a
something not always explainable by the absence of a normalized, or field that was certain (phonetics-morphology). Word order was free
merely regular, spelling. The multitude of suffixes and radical vowel because the syntactic functions were substantially differentiated.13
gradations showed a teeming and expansive morphology; the multi- The idea of declension, then, in a very concrete manner allowed one
ple constructions as well as the obvious sequential freedom made it to put order in the disconcerting multiplicity of forms. Thus the fi-
clear that syntax was also a variety of different forms competing
freely. As Sonia Branca demonstrates, the response ofscience was not fl,
I" nal s curiously adorning certain singular substantives is explained not
by a spontaneous graphic anarchy but by the declension of the first
to come up with a system for this heterogeneity but, rather, to dis- i '~ class of substantives:
solve or reduce it. ll The first attempts were to dissolve it and make it j'; singular plural
go away: Guessard, (an empirical editor who was as recalcitrant with subject murs rour
Lachmann as he was deaf to Diez) dismissed Old French as a primi- I, object

[wall]
rour
[walls]
mues
tive babel ("the age-old and well-known story of the language of [wall] [walls]
childhood, or a language in its childhood, which amounts to about Thus the double formations (sire / seigneur [lord], ber / baron
the same thing"). 12 Next they reduced it, according to geography (the [baron], suer / sorour [sister], etc.) which we find in certain words is
multiplicity of forms going back to the diversity of dialects) or ac- made clear by constructing a second paradigm of inflection, showing
cording to grammatical order. a shift of accent:
In this regard, the discovery or the invention of declension in Old singular plural
French constituted the most important act of the linguistic reflection. subject sire seigneur
[lord] [lords]
From Raynouard's first intuitions to the splendid, inflected tables of
object seigneur seigneurs
fin de siecle historical grammars, the study of the two case system in (lord] [lords]
58 59
This theory of medieval declension, in the end, was identical with charters. This, after all, was a fantasy: the idea the editor had formed
the notion of decadence. The manuscripts display such numerous of primitive Old French, a language neither attested nor described, a
"errors of declension" that analysis (and teaching) of Old French in pure object of speculation, the specular image that philologists and
this respect and many others is a discourse rotten to the core. The el- grammarians batted back and forth forever. And the most important
egant and practical system of inflection had, in fact, sunk; and characteristic of this language was its regularity and perfection.
though there is no agreement about the date of the shipwreck, one The theory that the copy represents degeneration presupposes a
can contend that it was earlier than any of the manuscripts we have, flawless original; the author has no right to any lapsus. Similarly, the
that is, before all written traces, documents, or bits of information. idea that language becomes degraded implies an impeccable origin:
The grammarian has flotsam and jetsam to contemplate. And dreams the author has no right to bad language either, or to dreadful puns,
on: the perfect beauty of (proto)-medieval declension derives from or, indeed, to the diversity of his way of speaking. By subscribing to
the impossibility of attesting it. The most effective means of reducing this, philology surreptitiously annexed a literary theory-the theory
the heterogeneity sullying the old language is the fantasy of bygone of the genius. 14 By magnifYing a transcendent author it tied an au-
perfection. thoritarian theory of the subject (the master of the sense as well as of
Here the philologist came to the rescue of the grammarian. The the signifier expressing it) to the notions of origin and textual stabil-
critique of forms provided the simple purity of original homespun to ity. The author, great by definition, and unique, the most pre- of pre-
the archetype and its proscenium arch. Of course, as Gaston Paris reproduction by the unity of his conception, the opaciry of his work
marked, such reconstruction is more tricky in the case of medieval (the argument of the lectio difficilior), and the quality of his language,
French texts than it is for Latin manuscripts, which were written in a stood in sharp contrast to scribal diversity, ignorant and purposeless,
language that virtually never ceased to be taught and commented which pluralized the work, trivialized its expression, and impover-
upon. But reconstruction was possible, nonetheless, and should be ished its language. A truly metaphorical solution, by displacement.
attempted so that the editing of medieval texts could be established Medieval philology, representative of the earliest thought about
as a science and the unification of general philology could be brought the text, attributed to a primordial subject the textual stability that
to completion. it sought in the midst of variants. As Michel Foucault noted, "privi-
The editor, whose basic mistrust led him to turn away from the leging the author," which we view as the basis of positive philology,
multiplicity of forms before him, latched onto an elsewhere he did after Mallarme, would in the main be transferred to the text and
not have. He had three things at his disposal to make this undertak- writing itsel£ This, then, would pave the way for the final textual
ing successful. First, several concrete arguments essentially derived epiphany of literary theories since New Criticism. Philology plots
from versification; it became established as a principle (largely unver- the curve of this evolution at the heart of modernity as it relates to
ified, moreover) that renewals and rejuvenations never, or hardly ever, the text.
touched either rhyme or rhythm. Second, he could turn to paltry For Gaston Paris the language of the original Saint Alexis was an-
subterfuge: Natalis de Wailly, publishing l'Histoire de saint Louis in cient, regular, and homogeneous, and it was in this spirit that he cre-
18 74, held that the scribe to whom Joinville dictated his narrative ated the reconstructed archetype. An example is the declension of
around 1306 must have been a member of the chancellery. This substantives, which the editor did not doubt had been altered in the
scribe, after closely analyzing the charters written in Champagne dur- manuscripts and which, correlatively, in the original was perfectly
ing the same period, rewrote the narrative in the language of the correct. "Thus, the irregularities of our manuscript as regards spelling
60 61
Gaston Paris and the Dinosaurs
All manuscripts are in agreement: Folques OA, Forques B, Foukes

must be attributed to the influence ofAnglo-Norman habits, and one
CDE.
is justified in eliminating them and in reestablishing French declen-
This is a phenomenon deserving our attention because it seems to
sion in the form that it existed during the period in which the poem
go against the defective nature of the proper noun as regards inflec-
was composed throughout" (106). However, it is just as true today as
tion, as well as against our linguistic understanding (at least the offi-
in 1872 that there are plenty of areas within the standard theory of in-
cial one): a subject marker has been added to an attribute of the ob-
flection where grammatical description is not very certain. In mor-
ject. It gives the impression that analysis of the ancient language is
phology, for instance, there are proper nouns that seem reluctant to
different. It extends the class ofattributive verbs through an operation
accept the final s required by inflection in the nominative singular.
in which marks of inflection playa part. There are other examples:
For example, in manuscript L, considered to be extremely close to the
original, the proper noun euftmien never appears as euftmiens when Li empereres sefait e balz e liez
it functions as a subject: (Chanson de Roland, ed. Moignet, v. 96)
[The emperor becomes both joyful and glad]
tut sui sen est eufemien turnet
[entirely alone Eufemien turned around] (L, v. 6ge) and the phenomenon, moreover is not unknown in modern French
(Elles ont l'air heureuses [They(f., pI.) have a happy airs(f., pl.)]). Still,
Manuscripts P and A vary in their practice; in S, which was later,
the important thing is that the medieval artifact must be made avail-
the word is always given as euftmiens in this case. To the eyes of the
able to the reader. In this passage from the Saint Alexis, the three
grammarian, this is a paradoxical and extremely interesting situation,
main manuscripts agree on a reading in which an occurrence of the
demonstrating a pronounced use of noun inflection at work in S
expression avoir nom is included:
(correct usage and hypercorrection). Gaston did not really worry
about this very long; he corrected manuscript L (the one he was fol- Ii uns acharies Ii altre anories ot num (L, v. 62b)
lowing for the line in question) and published the following critical Ii uns akaries Ii altre honorie out num (P. v. 62b)
text (where all the attributive characteristics of euftmien have been
Ii uns achaires Ii altres oneries out num (A, shifted to v. 72b)
corrected):
[One had the name Arcadie, the other Honorie]
Toz sols len est Eufemiens tornez (Saint Alexis, ed. Paris, v. 6g e)
Except for the form honorie in P, there is complete agreement,
[entirely alone Eufemien turned around]
therefore, upon a "good reading" in the nominative case (acharies,
There is another example of an equally doubtful and paradoxical anories, etc.). Gaston Paris, usually so quick to distribute an inflex-
use of syntax. In Old French, the expression avoir nom, avoir a nom ional s in order to restore faulty nominatives, here acted with equal
(to be named) is often followed by a proper name in the nomina- swiftness, but this time in the opposite direction. The stability of a
tive-if not in texts, at least in the critical apparatus (where, luckily, grammatical model organized according to classical canons is such
the material prior to its correction can be read). Thus, in this phrase that it can be applied to primitive organization following notions of
from Villehardouin: what is natural. The editor removed from these proper nouns "an s to
un saint home en France qui ot nom Folques de Nuilli which they have no right, because when it comes after avoir nom (to
(La Conquete de Constantinople, ed. Faral, § r) be named) the proper noun is naturally put in the objective case"
(107); in his edition it reads:
[a holy man in France, who was named Folques of Nuilli]
63
62
IN PRAISE OF THE VARIANT 'l Gaston Paris and the Dinosaurs
Li uns Arcadie, Ii altre Honorie out nom came acknowledged and put into practice, they no longer provoked
(Saint Alexis, ed. Paris, v. 62b) any debate but, rather, initiated a sort of end to philological history
which lasted more than half a century; the steady editorial activity
The problem vanished, and the grammars that carne after, relying on was extremely peaceful, its sleepy thought sustained by a revealed truth.
this famous edition, were able to assert that the object's predicate in Now, we should note that this truth is basically dubious and that
Old French, is "naturally" in the objective case. Then other editors, it is far from being general. Bedier was no theoretician, and we must
reconstructing other texts, took their authority from these grammars. take into account that he was a rather Germanophobe individual
"I think it will be readily acknowledged that the text of this poem, who mistrusted the mathematization of a literature he knew marvel-
as I present it, offers an acceptable example of good French as it must lously well, a man who was always extremely clever, endowed with a
have been spoken and written in the middle of the eleventh century" keen intelligence and with a taste for being provocative. Did he not
(135). And as it is nowhere attested in the manuscripts of the Saint choose the Lai de I'Ombre, one of the rare medieval texts where, if
Alexis, which, like all the other works in the vernacular, date from the one takes the traditional point of view, one can suppose that its au-
twelfth century at the earliest. Neglecting the linguistic abundance thor revised a later copy of it? Bedier, analyzing at length five readings
that was attested but "late" and muddled things somewhat, the shared by manuscripts D, E, and F of the lay, concluded that they are
philologist dreamed of an ancient language, pure and reedy, stripped the work of a man who was "extremely intelligent, who spoke the
bare and white, like a skeleton pulled from the desert, like an im- same language as Jean Renart, wrote in the same style, used the same
maculate Romanesque church, to the taste of the Second Empire: "I technique for rhyming and whom the Muses had endowed with pre-
have tried here to do for the French language what an architect would cisely the same gifts"(65)'
do who wanted to reconstruct on paper Saint-Germain-des-Pres as Of course, the application of medieval stylistic models, the con-
the eleventh century admired it." AB Gianfranco Contini amusingly stant interaction within coded forms of writings can bring forth a less
remarked: "ll Paris ha voluto essere, insomma, piu che il Cuvier, il simple and individual explanation. For Bedier, however, any such
Viollet-Ie-Duc del francese del Mille."15 high praise for the reviser would lift him over the basic threshold sep-
arating copyists from author: endowed with such talent, this scribe
could be none other than Jean Renart himself, and that, indeed, was
Doubt
what Bedier suggested. In this case, however, if the person who pro-
It was Joseph Bedier himself who inaugurated the third period, in vided the perfect form and sense in some primordial time had van-
a startling manner, republishing in 1913 at the Societe des anciens textes ished, then made the revision; if there was some later intervention by
franrais his earlier, youthful edition of a short verse narrative, the Lai the author, then the origin splits in two, the stemma codicum spins on
de I'Ombre by Jean Renart (an edition strictly adhering to the Pariso- its axis, the beautifully organized arborescence turns into a disorderly
Lachmannian method).16 The new edition, however, expressed and flaccid rhiwme. Genealogical classification becomes impossible,
doubts about this method. Amplified and justified, these doubts were and the reproduction of the archetype (now multiple) becomes im-
in the form of a long article published in 1928, which, in magnificent practicable. Such an argument, even though based on a case, is slen-
language and with the grace ofa fine mind, laid Gaston Paris to rest. I? der, and lateral.
Somewhat regrettably, the ideas of Joseph Bedier became utterly Similarly, Bedier did not make a frontal attack on Lachmannian
successful in France and in Anglo-Saxon countries. Once they be- method. He said nothing about the critique of forms (devoting him-
64 65
r' Gaston Paris and the Dinosaurs
self to the critique of readings) and shared, as we have seen, the nos- tained a perfect binarism (the original division into two families),
talgic and anachronistic ideology concerning authorial origin: "Is that is, until it could go no further. There was also another and
there any sure way of reconstituting a sufficient approximation of the deeper reason driving it until the system was blocked. "Obscure
contents of the first manuscript, presumably unsullied by errors, the forces, confined within the depths of the subconscious, exerted their
one that Jean Renart, on the eve of publishing his poem, must have influence" (15). (It is perfectly apparent what theories this worthy aca-
written in his own hand and then, to be certain the text was right, demician was echoing here, in 1928, two years after the French trans-
must have read and reread with his own eyes?"(2). Bedierism is basi- lation of the Traumdeutung). This was desiring man somehow taking
cally the regret over not being able to give a positive response to that revenge on the machine.
question. Rather than attacking the basic principles of the critical In fact, with classification in the shape ofa tree, the archetype was
method imported by Gaston Paris, Bedier brought to light an odd reconstructed, as Gaston Paris said, mathematically; the rule of "two
imperfection linked to the practice of this method. In 1913 he had, in against one" set in motion an automaton that, step by step, engen-
fact, discovered a "surprising law": among all the editions he had con- dered the text. Now, though the classification is binary on the most
sulted (among them, we should note in passing, the Saint Alexis general level, in every case in which two readings are irreconcilably
edited by Gaston Paris), 95 percent of those that present the classifi- opposed, the automaton "holds out to the editor ... the offering of
cation of their manuscripts in schematic form do so with a stemma its two arms, which, though immobilized, are laden with variants"
that is split in two, with only two branches (that is, two supposed (14). And this editor is free to choose, with an entirely clear con-
copies) deriving from the original (see diagram). science, the reading that suits his taste and intuitions, not to say the
venerable judicium. And, since almost all the classifications set forth
o are two-branched, Bedier lets it be understood that Lachmann's
/~ method never had, in fact, been applied to French medieval works

but had always been circumvented by philological cunning, which
had the skill to weaken the automaton, the skill, indeed, to paralyze
a
p it completely. If, therefore, "scientific" method was secretly under-
mined by desire, why persist in creating heteroclite monsters; why
••• • •• • ••
M
•• • • •• • ••
''A tree that is split into two branches is in no way bizarre, but a
not go back to a "Humanist" conception of editing, conscious of
one's limits and one's choices? The best idea was to accept a good
manuscript (one in a favorable position within the stemma) and to
stick with it, reducing and explaining the necessary corrections:
"Consequently, the most commendable method ofediting is perhaps,
grove of two-branched trees, a forest? Silva portentosa" (12).18 in the final analysis, one governed by a spirit of self-doubt, a spirit of
Bedier provided two explanations for this surprising phenome- prudence and extreme 'conservatism,' a vigorous will-almost a bias
non, each a matter for a different stage of philological consciousness. in favor of scribes, giving them the most credit and not touching the
The first was inscribed within the logic of the system: the Lach- text being published except when there is an extreme and almost ob-
mannian method, fastidious and virtually fanatical in its indefatigat- vious necessity" (71).
ible comparatism, was a force that worked by dichotomy until it ob- Rejecting the Lachmannian edition of the Lai de I'Ombre which
66
67
·':
,.
r· Gaston Paris and the Dinosaurs

he had delivered in 1890 (adorned with a magnificent two-branched this phenomenon by the thesis of universal catastrophe that suppos-
stemma), Joseph Bedier published, in 1913, manuscript A ("a good edly destroyed all species then alive, so that we no longer have any
manuscript, published almost without alteration and accompanied knowledge of them except through scattered fragments, which,
by notes that mark a return to the technique of the old Humanists" through patient application of comparative anatomy, can be pieced
[17]) and then, in 1928, another good manuscript, E, which "offers a together by the naturalist. 21
lush and well-formed text of our poem" (66). Gaston Paris's philology was based on much the same thesis: the
This is the Bedierism, in almost caricatural contrast to the basic copy is a catastrophe that displaces and destroys the original work so
assumptions circulating since Gaston Paris, which was adopted with that all that remains to us now are (Paris could have repeated Cuvier's
almost no debate by French and Anglo-Saxon philology. Some words) "isolated bones thrown helter-skelter, almost always broken
adopted it out of laziness and many out of a general conservatism, and reduced to fragments" (58). We are right, therefore, to recon-
but above all, this came about as an evolution in keeping with the struct them by means of the meticulous application of comparative
spirit of the times. 19 Despite what Bedier himself said, it was no re- philology. Lamarck offered a solution to the problem raised by the
turn to square one, and the clash with Gaston Paris was not as it has fossils, but it went unheard until after Cuvier: fossils belong to still
usually been described but, rather, represented a remarkable aggior- existing species, but they have since changed and ceded to the species
namento of thought concerning the text. that are currently living. 22 There is no need at all to suppose that
Bedier's criticism, as we have mentioned, picks up once more the there was some universal catastrophe that came to topple everything;
postulates of nineteenth-century philology, and the lateral argumen- "nature works slowly and by successive degrees" (16). Against Cuvier's
tation that he developed shows that he had little room to manoeuvre. creationism, the correlative to his notion of the catastrophe, Lamarck
What was probably newest and, for us, most important about this opposed a description of life forms which was simultaneously taxo-
theory was its committed bias in favor of the scribes, which gave nomic and diachronic: "Vegetable and animal types are not, for the
them the most credit. Bedier turned his pragmatic attention to the most part, primitive, but they are derived, through successive trans-
data of philology, those manuscripts it was important to bear in formations, from other forms of which there are living or fossil rep-
mind, going through them with the eyes of an editor. Each is a seg- resentatives" (20).
ment of language that the scribe's interventionism quite specifically Though something like a faint echo of Freud (or Charcot) runs
homogenized or attempted to homogenize. 20 It had been realized through Bedier's discourse, it is most unlikely that this traditionalist
with a situation of effective literary communication in mind. It was Catholic would have been interested in Darwin (translated in 1872 in
something real. A living creature, in short, not flotsam, not jetsam, France), who was commonly considered a profoundly antireligious
and not a fossil, but one realization among others in the midst of an thinker. By contrast, French thought, strictly biological (and not in-
evolving process. The three "good" manuscripts-A, E, and F-of scribed within any philosophy of general evolution), constituted by
the Lai de I'Ombre, were described by Bedier, in fact, as three "forms the neo-Lamarckian transformism that was the rule starting in 1880,
of the text" ("differently, but almost equally coherent and harmo- the period in which Bedier had been a student, could hold his atten-
nious" [68]). tion and appeal to him. In his study on the manuscript tradition of
The fossils found by Cuvier in Montmartre's gypsum did not for the Lai de I'Ombre, Joseph Bedier, in short, introduced neo-Lamarck-
him correspond to any contemporary living animal. He explained ian thought into philology: Manuscripts evolved like species, inher-
_________1
68
~ 7
'n ~. :
"C,
' c·'".'.P.'.'.
ited acquired modifications (readings in common), differentiated Bedierism provides no picture of the intrinsic variance that is me-
themselves by successive speciations (manuscript families), and dieval writing, only snapshots, which are, of course, preferable to il-
showed variations among themselves that were more adaptations lusory reconstruction. Like the latter, however, they leave this surplus
than degradations. Though, of course, he might regret the primitive oflanguage and meaning in the place allotted to it by textuary think-
species (created by the author, meaning God), the task of the philol- ing-in the margins.
ogist, like that of the naturalist, was not to reconstruct this primor-
dial unity but to compare all the different available species that are,
of course, imperfect but alive.
From this point of view, Bedier's theory represented an important
stage in philological thought, which was thus brought in line with
contemporary science. We need to be careful not to simplifY the pic-
ture and find reassuring contrasts: only one generation, as we men-
tioned, separated Gaston Paris and Joseph Bedier, but it was one that
saw a break take place in the nineteenth century as it drew to a close.
Philology was still well within the tradition of thought as it was
formed by and concerned with text, with Bedier representing its final
stage as far as the treatment of medieval texts was concerned; the
good-natured, academic comfort in which Bedierism has dozed along
ever since is one sign among others. Bedier's antimethod, as much as
any other, reduced medieval works to the stable, closed, authorized
texts of modernity. Of course, very fortunately this did not consist of
relying on a ghost-the primordial subject, the guarantor ofstability
-for the process of reduction, pretending that there was a unity to
its conception. But the notion of text was reduced to the manuscript,
which was real, of course, but it was still considered unique, with
complete disregard for variational space. It was as if one had had the
same edition in its essentials ever since the Middle Ages, so the
philologist merely used elaborate care in correcting any possible blun-
ders. One can better understand, therefore, the particular fondness of
Bedierists for the copy made by the scribe Guiot of the works of
Chretien de Troyes, which is the copy they faithfully reproduced. The
language here is smooth and beautiful, its quality pleasant and dis-
tinguished, nothing sticks out, and nothing is surprising. It is a good
manuscript; it is a good edition. Gaston Paris was suspicious of vari-
ants; Bedier loved them, to the point ofediting them exclusively. But
7° 1 7' I
~ I
Turn the Page
us? Surplus worthy of note has been carefully placed in the critical
Turn the Page apparatus. This deposit, nevertheless, though not a reject (it is
arranged in order), by its very disposition takes on a prisonlike air.
The patient and recalcitrant reader, dissatisfied with the unity pro-
vided him, can gain from it a few fragments, splinters, and scraps,
but not the other of the text. The secret function of the critical appa-
ratus is to dissipate this in silence.
The result is the following passage from paragraph 70, accompa-
nied here by the corresponding extract from the critical apparatus,
printed at the bottom of the page:
kill Alexis si m prist son frere l' empereorn , si o Ii traist les iaulz de la
teste et seP fist empereor'l en tel traison con vos avez o'i r.
[This same Alexis so took his brother the emperor, so pulled his eyes
from his head and made himself emperor in such a betrayal as you
The question could not be more concrete. Philology, a practical sci- have heard.]
ence, can be judged by its works: whatever the perspective adopted,
(I) Et cil B, Cil CDE. (m) Alexis si manque dans CDE. (n) tempereor
whatever the method used, whatever the object chosen, it always
manque dans B. (0) et CDE. (p) Manque dans B. (q) empereor de soi B. (r)
comes down to manipulating pieces of writing, arranging them ac- come oes CDE.
cording to one's convictions, then making their calm, self-assured or-
(I) Et cit B, Cil CDE. (m) Alexis si is lacking in CDE. (n) l'empereor is
der available to readers. Medieval philology's affiliation with textuary lacking in B. (0) et CDE. (p) is lacking in B. (q) empereor de soi B. (r)
thought is clearly seen in the works with which it has filled libraries,
in its unveiling of these austere treasure troves. But in that case we
~ com oes CDE.
need to answer, in a practical manner, the question in turn entailed, Despite the superscript letters that dance their disconcerting bal-
by our critical study: What is your philology? let of occasional otherness along the line, this is a very simple exam-
In 1938, Edmond Faral provided a fine edition of La Conquete de ple. The sentence is perfectly clear. (This same Alexis so took his
Constantinople as told by Geoffroy de Villehardouin.! This Bedieriste brother, so pulled his eyes from his head and made himself emperor
edition followed manuscript a (Oxford, Bodleian Land. mix. 587), in such a betrayal as you have heard.) The variants, which are ele-
correcting very explicitly (introd., L) the only readings singled out by mentary and far from numerous, at first sight contrast manuscript B
general agreement in the manuscript tradition. Thus one reads a on the one hand with the group CDE on the other; nevertheless, it is
good copy, a "document," as Faral would say, "which is of greater in- obvious that a second, even dual view is necessary for a curious
terest than an attempt at conjectural and partial reproduction." What reader. How can such an arrangement, indeed, be read? And where
about the other manuscripts, however, whose quality the editor has does one begin? Let us stick with group CDE. One notices as the text
demonstrated and which are part of the textual production of this progresses and as we come across them in the notes: a divergence in
the demonstrative (icil "reinforced" in 0; cil the "simple" demonstra-
1
narrative of the crusade which the Middle Ages has left behind for
7 73
Turn the Page
tive in CDE); the absence of a segment of text (Alexis si is lacking in The first proposition has as its subject a proper noun, preceded by
CDE: a proper noun plus its coordinate; this segment, circumscribed an intensive demonstrative, and separated out to the beginning of the
in this manner, has no syntactic reality); another coordination (si in sentence by the linking element si, which introduces the predicate.
0; et in CDE); a change in tense (com vos avez of/ com oes). A bit of This element introduces the second predicate as a break (unlike the
everything (grammatical elements and an agrammatical fragment), coordinating et); this second predicate is coordinated to the third. In
nothing that concerns the drive or the unfolding of the text, nothing short, the temporal location comes before the situation of enuncia-
that has to do with meaning. Slim pickings for the meekest of gram- tion (con vos avez or). Before our eyes two homogeneous utterances
marian's analyses, which is the only one invited to authenticate its with coherent variation materialize; thus the alternation:
wealth, and which at the very most and in the most unremarkable
manner can fill its trivial files under entries such as "demonstrative," cit prist +--+ ici/ AkxiJ si prist
"coordinate," "sequences of verbs." No syntax at all: the variants that pronoun verb intensive proper separating verb
demonstrative noun link
have been noted in this sentence have been defined and classified ac-
cording to a classical morphology, blind to the movements of the text. altered by variants I and m. These utterances translate two perspec-
If, however, one takes some distance (taking a piece of paper and tives, one thematizing and the other nonthematizing. One need not
a pencil, too) and regroups the variants in CDE, a structure appears attempt to discover which is closer to the "original" (a philologist's re-
-and its meaning: flex), or even which is the older (grammatical reflex); one must as-
Cilprist son frere l'empereor et Ii traist les iaulz de fa teste et sefist em- sume their equivalence and grasp medieval language as it swings back
pereor en tel trai'son com ols. and forth between the two. Thus literary writing of the Middle Ages
produced a constant torrent of paraphrastic utterances that are the
[This one took his brother the emperor and pulled his eyes from
precise data required by syntactic inquiry. The most precious thing
his head and made himself emperor in just such a betrayal as you
are hearing.] that the contemporary linguist can get from a native speaker of a liv-
ing language, that is, the dynamic of the language, its play of form
There are three simple propositions (the subject of the first is the and meaning, is all there to be seen in medieval manuscripts, but
pronoun cil) which are coordinated by et; the three verb predicates, then it is dissipated when they are edited.
thus balanced, are eventually linked on the level ofenunciation (pres- How can one visualize and anchor the variants more precisely?
ent indicative). Which suggests that we consider once again the read- The reply given by philology is not just a technical answer; it is
ing (which we shall accept as being a complete utterance) of manu- linked to the major choices made by historical grammar and reveals
script 0; then we can detect the operations of thematization and what they are. In the second half of the nineteenth century, linguis-
rupture that it consists of: tics, in order to acquire the status of a science, carved out for itself an
object that could be formalized and calculated by reducing language
leilAlexis si prist son frere l'empereor, si Ii traist les iaulz de fa teste et simply to its phonic layer-as far as possible from the subject of
se fist empereor en tel trai'son con vos avez oi:
meaning. There are immense consequences to such a decision: it is
[This same Alexis so took his brother the emperor, so pulled his the basis for a science, and furthermore, it legitimated the experi-
eyes from his head and made himself emperor in such a betrayal as mental role of this science among the disciplines concerned with hu-
you have heard.] manity.
74 75
],
Turn the Page
Rejecting the traditional commentary on good authors and also out the study of the language, unifYing it. One can follow its trace
the wider speculations about language and its origins, linguists took from phonetics, the hard nucleus with its pioneering research, to his-
real languages and their history as their object. It was an object that torical and comparative linguistics, and from there to the ancillary
could be enunciated (languages are bodies of sounds), represented science whose task it is to establish the data-philology. It is under-
(sounds evolve according to laws that are inescapable and blind to standable that such an approach to language, based on the study of
meaning), and calculated. This solved the age-old question of where individual sounds, would scarcely venture beyond the lexicon (the
our languages came from. study of individual words) and morphology (the study of their indi-
Progressing into the relatively homogeneous realm ofIndo-Euro- vidual forms). Historical linguistics, a consideration of the element
pean languages, linguistics brought positivist method into play with and not the system, has neither the means nor the desire to take syn-
its factual thought and genealogical classification-at the cost, as we tax as its object. The latter remained in the margins of research, often
have mentioned, of an extreme reduction: language became reduced in appendixes (the layout of historical grammars is revealing); it rarely
to its phonetics and, more specifically, since the phonic layer was not goes beyond an impressionistic summary of curious facts, and it is
conceived of as a structure, to the individual nature of the sounds. equally given over to old prejudices and unverified hypotheses. The
This was, of course, an impressive gesture, enunciating laws at the theories of Gustave Guillaume, which were deployed with some suc-
very heart of what seems most personal in language {pronuncia- cess among comparatists (Guillaume was one of Meillet's students),
tion)-which also, among the thousand characteristics of intonation thus expressed the ambition to think syntax through by resorting to
which delineate a personality, showed what belonged to a group. completely unfalsifiable hypotheses deriving from Bergsonian philos-
Phonetic law, however, has nothing to say about structure or sys- ophy.
tem; it is about fate. Historical phonetics, the basis of historical and It is understandable that, throughout the field, there is a recogniz-
comparative grammar, details the evolution of each of the phonic el- able way of working, one in which the fastidious study of the note-
ements that can be heard in the language. Because relationship with worthy bit is given most importance; philology's great fondness for
another sound is not systematic but accidental (conditioned trans- notes, short papers, and critical analyses is well known. In the bibli-
formation), the classification adopted by the description is of little ography of great philologists, besides editions, how many small notes
importance-the first work done by Diez on vernacular languages devoted to a dialectal characteristic, how many brief remarks on three
listed the sounds in alphabetical order. Moreover, a good phonetics lines or on a variant are listed! And finally, one understands what is
manual is only a collection oflaws, and it is strictly the linguist's busi- the p~rsistent, steady basis of this critical apparatus, from Karl Lach-
ness to discover them; articulatory factors, for instance, are simply at- mann to Joseph Bedier and his present successors. It is a critical ap-
tempts at explanation at a later date. Nineteenth-century linguistics, paratus both predisposed and functioning to link the way historical
a science that was inherently historical, adopted the behavior suitable linguistics thinks (the variant is a word, an isolated form) with the
to the triumphant history of its era. Phonetics is a purely descriptive
history of sounds; it establishes, documents, and arranges in chrono-
j; textuary will to provide readers a stable and unique text, by explod-
ing and scattering the textuality surrounding it (the variant is part of
logical order the accidents that have happened to noteworthy indi- the nontextual). The variant, a tiny bit, a brief note, the hysteria of
viduals. detail, is simultaneously the sieve that retains nothing of the syntax
From that point on, a conception of the individual, the element, except a lexicon and the curse directed at the adjacent, threatening,
ofwhat is noteworthy and guaranteed and fragmented, runs through- \~~i
textual element, breaking and tearing it apart. Now, medieval writing
76 77
IN PRAISE OF THE VARIANT Turn the Page
does not produce variants; it is variance. The endless rewriting to plete form all the intact data, which will become the wealth of the
which medieval textuality is subjected, the joyful appropriation of reader. Because a loyal magnanimity was their only choice, they for-
which it is the object, invites us to make a powerful hypothesis: the got that every edition is a theory: though one must show what is
variant is never punctual. Paraphrastic activity works on the utterance there, one must above all make it understood. Also, there is no way
itself, like dough; variance is not to be grasped through the word; this to leave the two-dimensional space of the printed page in these edi-
must be done, rather, at least at the level of the sentence if not, in- tions, where medieval writing is set before one's eyes but not set in
deed, at the very heart of the complete utterance, of the segment of motion. The solution lies elsewhere.
discourse. Remember that the variation cil/icil in the earlier example, The computer, a valuable aid and one worth considering, provides
a local trifle marked by the critical apparatus, played a part in an the obvious solution. Recently progress has been made in the way the
overall and differently conveyed orientation of the utterance. Vari- computer visualizes and manipulates texrual data on screen and the
ance is the construction of a sense, of a sequence of writing. It is a way it manages this data. Progress has been such that our most ordi-
syntax, the "building," as Ramus called it, of a collection oflanguage nary use of the computer brings us face to face with a form of writ-
phenomena that take on meaning only through the link uniting ten matter that can no longer be described as bookish; we shall call it
them. It is in this way that variance is to be grasped and appreciated, a screenic presentation. Always in the process of further development
that it is important to have it understood. and refinement by technicians, the screen is simultaneously dialogic
Because the variance of a medieval work is its primary characteris- (it offers a constant interaction between the user and'the screen) and
tic, the concrete otherness of discursive mobility, the figure of a pre- multidimensional {through the use of "windows,"it allows one to
modern written word, editions must give it priority, following it bring together and consult information belonging to separate enti-
closely. We cannot, in fact, examine each critical apparatus or each ties}. Utilizing these two qualities, one can conceive of the type of
manuscript from every angle and work out a restored textuality as we edition that would result from this assembling of separate entities
did with group CDE of Villehardouin's manuscripts. Luckily, we represented by codices. In such an edition medieval works would no
have a good edition of manuscript B and can make a line-by-line longer be subjected to the two-dimensional and closed stucture of the
tit
comparison. 2 With patience one could also do the same with the printed page because a diskette accommodates varied textual masses,
editions following different manuscripts for Chretien de Troyes's which the reader consults by making them appear in different ways
Perceval, or the Roman de Renart. That was basically what Bedier on a computer screen. Certainly, it is possible to irp.prove on the edi-
wished to do by publishing first A and then E of the Lai de l'Ombre. tions in parallel columns diffracting a book which We mentioned ear-
Is it necessary to multiply books in this manner? Or indeed to defract lier because the computer executes a diffraction and proliferation of
and proliferate the book, as did Jean Rychner with the Lai de Lanval textual spaces: in each of the "windows" of the screen there appears
and the fabliaux, Peter Dembowski with the Vie de Marie l'Egypti- an inscription that is saved in the computer's memory. This inscrip-
enne, Willem Noomen and Nico Van den Boogaard in their Nouveau tion can be arranged or shifted, acquiring visibility and importance
Recueil complet des fabliaux, publishing the different versions of a {linking, rooming in, etc.}. As the articulation of a vast memory, the
work side by side in parallel columns. 3 These were all good and very computer, with its instant management and multiple visualization, is
useful attempts, answering a need but unsatisfYing in the end. the answer to the editorial requirements that we have noted. The use
Tempted by the diplomatic copy, these editions have been drawn into of the computer, because the effects and constraints of the printing
the fantasy of the facsimile, of honestly providing in the most com- press are irrelevant to it, thus allows the reader to see and consult not
78 79
IN PRAISE OF THE VARIANT Turn the Page
only the totality of the manuscripts of a particular medieval work but mensional screen, simulates the endless and joyful mobility of me-
also the editions (empirical, Lachmannian, Bedierist, etc.) which !' dieval writing as it restores to its reader the astounding faculty of
took these manuscripts as their objects. Moreover, it can provide a memory-the memory that defines its aesthetic reception and is ba-
great many minor bits of information, which should remain virtual sic to the pleasure taken by a reader.
so they will not get in the way of reading but which one needs to be This simulation, however, suggests that the written word we have
able to locate: makers of concordances, frequency lists, tables of designated as screenic might very well escape from textuary thought.
rhymes, every sort of calculation, codicological and paleographic Mechanical writing is extremely flexible because of the way text is
data, and so on~verything that a printed edition usually abandons presented; reading on screen calls up, brings together, and arranges
or from which it makes a painful choice, everything that the hyper- segments of writing taken from the vast, obscure, and inconceivable
scholarly edition hypostasizes to the point of unreadability. It is not reserves of electronic memory. What the screen provides the reader is
simply a question, however, of making what is an honest and upright an always instantaneous grasp, the ephemeral visualization of differ-
fidelity even more magnanimous; edition must lead to a reading: it is ent and mobile textual spaces, of texts that are no longer that palpa-
the theory of the work. ble reality of ink on a page but are only the immateriality of a few
Though information technology satisfies our need to be exhaus- electrical pulses. The two-dimensional stability of the text, attributed
tive, it also upholds the intelligence of the text, and this is where its to an originating and controlling subject, no longer seems the basis
most important and newest technical contribution can be seen: the for this new technology of the written word. Computer inscription is
computer is able to help us detect the dynamics of the text by mak- vanance.
ing visible the connections prepared and suggested by the editor. It is At first, people see a technology that brings new ideas as threaten-
less a question, therefore, of providing data than of making the reader ing. In the case of the written word, this phenomenon has taken
grasp this interaction of redundancy and recurrence, repetition and place before, toward the end of the fifteenth century. The printing
change, which medieval writing consists of-and to do so according press had just been invented, and books that came off the press were
to the two axes that we have brought to light. Vertically, along the considered as, at the very most, excellent manuscripts. Though we
thread that leads through the work, it can bring back all the things may think of the computer of our everyday intellectual existence as
that each noteworthy utterance constantly echoes but which modern merely the most desirable of typewriters, there is something, perhaps,
memory no longer hears; the screen unrolls the infinity of memorable silently stirring in our conception of the text. There is some ques-
context. Horizontally, it can compare the utterances within a perti- tioning of our spontaneous philosophy of the textual, which literary
nent and chosen range ofvariants that are paraphrases of one another criticism, fascinated by the parts of writing that precede and exceed
from one manuscript to the next, even indicating by some symbol or the reification of the completed work, may echo. There is some vague
note what the characteristics of this relationship are. We would do portent that textuary modernity-this convergence of progress in in-
well to hang the changing constellation of the medieval written word dustry, in law, and in letters that is the mainstay of philology, the
in the boundless space that technology offers inscription today. That modern science of the modern text-might be coming to an end be-
truly would be publishing on a grand scale, editing on a scale never neath this new surge of technology. The outlines of a post-textuary
before realized, yet indispensable, and only the information technol- philology are appearing on the computer screen; it is a tempting op-
ogy of today can provide us with the means, probably even the idea, portunity to use the post-text instrument to provide some image of
of doing so. Because the computer, through its dialogic and multidi- what was there before.
80 81
I
So the page would be truly turned. Electronic writing, through its

mobility, reproduces the medieval work even in its variance. Because Notes
information technology, beyond modernity, is rediscovering the path
of an ancient literature whose trace the printing press had erased: and
this is something to think about.
Chapter I: Textuary Modernity

I. Lucien Febvre, "De 1892 a 1933, Examen de conscience d'une his-
lOire et d'un historien," inaugural lecture at the College de France (13
Dec. 1933), reprinted in Combats pour l'histoire (Paris: Colin, 1953), 5-6.
2. In particular, Percy Simpson, ProofReading in the Sixteenth, Sev-
enteenth and Eighteenth Centuries (London: Oxford University Press,
1935); and Hans Widmann, "Die Lekttire unendlicher Korectturen,"
ArchivfUr Geschichte des Buchwesem 5, 1963-64: 778-826.
3. Louis Hay, "La Critique genetique: Origines et perspectives," in L.
Aragon et al., Essais de critique genetique (Paris: Flammarion, 1979),
227-36.
4. Michel Foucault, "Qu'est-ce qu'un auteur?" Bulletin de la Societe
franraise de philosophie 64 (1969): 73-104.
5. Elizabeth Eisenstein, The Printing Press as an Agent of Change,
Communicatiom and Cultural Transformatiom in Early-Modern Europe
(Cambridge: Cambridge University Press, 1979), 80.
6. Jacqueline Cerquiglini, "Un Engin si soutil. Guillaume de Machaut
et l'ecriture au Xl17e siecle (Paris: Champion, 1985).
7. Gerard Genette, Seuils (Paris: Editions du Seuil, 1987).
8. Lucien Febvre and Henri-Jean Martin, L'Apparition du livre (Paris:
Albin Michel, 1971 [first published 1958]), 122-28.
9. Henri Desbois, Le Droit d'auteur en France (Paris: Colin, 1953), 5.
82 83
Notes to Pages I4-32 Notes to Pages 34-SI
Chapter 2: Mr. Procrustes, Philologist tre et la voix. De la "litterature" midievale [Paris: Editions du Seuil, 19 87],
"Intervocalite et mouvance, " 160 ff.).
1. Erich Auerbach, Literary Language and Its Public in Late Latin An-
tiquity and in the Middle Ages (Princeton: Princeton University Press,
1965), 248-62. Chapter 3: The joyful Excess
2. M. T. Clanchy, From Memory to Written Records: England
I066--I307 (London: E. Arnold, 1979), pt.!. 1. Gerald L. Bruns, "The Originality of Texts in a Manuscript Cul-
3. Brian Stock, The Implications ofLiteracy. Written Language and ture," Comparative Literature 32 (1980): U3-29.
Models ofInterpretation in the Eleventh and Twelfth Centuries (Princeton: 2. Henry John Chaytor, From Script to Print (Cambridge: Cambridge
Princeton University Press, 1983). University Press, 1945), 140.

4. Erwin Panofsky, Architecture gothique et pensee scolastique, trans. 3· Jean Rychner, L'Articulation des phrases narratives dans la "Mort
Pierre Bourdieu (Paris: Editions de Minuit, 1967). Artu"(Faculte des Lettres de Neuchatel; et Geneve: Droz, 1970); Bernard
5. Jack Goody, La Raison graphique. La Domestication de la pensee Cerquiglini, La Parole midievale (Paris: Editions de Minuit, 19 81 ).
sauvage (Paris: Editions de Minuit, 1979); and Goody, La Logique de 4· One expert recently remarked "In fact, without touching on the
l'ecriture. Aux origines des societes humaines (Paris: Colin, 1986). question of principles, and without a doubt, no scholar has ever suc-
6. Claire Blanche-Benveniste and Colette Jeanjean, Le Franfais parle. ceeded in classifying beyond dispute the manuscripts of a French me-
Tr-anscription et idition (Paris: Institut national de la langue fran~aise, Di- dieval work, whenever the manuscripts are the least bit numerous" (Felix
dier Erudition, 1987). Lecoy, ed., jean Renart. Le Lai de I'Ombre [Paris: Champion, 1979], vii).
7. Christiane Marchello-Nizia, "Ponctuation et 'unites de lecture' 5· Jean Rychner, Contribution a l'etude des Fabliaux. Variantes, re-
dans les manuscrits medievaux," Langue ftanfaise 40 (1978): 32-44; maniements, degradations (Faculte des Lettres de Neuchatel; et Geneve:
Henri-Jean Martin, "Pour une histoire de la lecture," Revue ftanfaise Droz, 196o).
d'histoire du livre 16 (1977) 583-6°9. 6. Michael J. Warren, "Quarto and Folio King Lear and the Inter-
8. Richard H. Rouse and Mary A. Rouse, "Statim invenire: Schools, pretations of Albany and Edgar," in D. Bevington and]. L. Halio, eds.,
Preachers and New Attitudes to the Page," in R. L. Benson and G. Con- Shakespeare: Pattern ofExcelling Nature (Newark, Del., 1978), 97.
stable, eds., Renaissance and Renewal in the Twelfth Century (Cambridge: 7· Alexandre Micha, La Tr-adition manuscrite des romans de Chretien
Harvard University Press, 1982), 201-25; "La Naissance des index," in de Tr-oyes (Paris: Champion, 1939), 292-93.
H.-J. Martin and R. Chartier, eds., Histoire de l'idition ftanfaise, t. I 8. Alfred Foulet and Mary B. Speer, On Editing Old French Texts
(Paris: Promodis, 1982), 77-85. (Lawrence: Regents Press of Kansas, 1979), 1-39; Mary B. Speer,
9. Jean Rychner, ed., Lai de Lanval, by Marie de France, Textes lit- "Wrestling with Change: Old French Textual Criticism and Mouvance,"
teraires fran~ais (Geneva: Droz, 1958). Olifant 7 (1980): 3U-26.
10. This expression is to be distinguished from the excellent term
mouvance created by Paul Zumthor to describe the "incessant vibration" Chapter 4: Gaston Paris and the Dinosaurs
and the "fundamental instability" of medieval texts (Essai de poetique
midievale [Paris: Editions du Seuil, I972 ], S07). In effect, this notion, I. Sebastiano Timpanaro, La Genesi del metodo del Lachmann (Flo-
which followed the ever more "oralist" evolution of Zumthorian reflec- rence: Le Monnier, 1963), 36-42, 56-68.
tions, came to designate the effects of the nomadism of the voice, of the 2. Louis Havet, Manuel de critique verbale appliquee aux textes latins
concrete and originating voice, on texts whose writing is no longer (Paris: Hachette, 19U).
thought of as anything but secondary and reductive. Under orders to 3· Hans Ulrich Gumbrecht, "Un Souffie d'Allemagne ayant passe":
make an original voice heard, the written word is put in doubt, in a Friedrich Diez, Gaston Paris and the Genesis of National Philologies,"
manner somewhat recalling the gesture that founded philology (La Let- Romance Philology 40, no. I (1986): 1-37.
84 85
Notes to Pages 5I-6I Notes to Pages 64-68
4. Gaston Paris, review of Karl Bartsch, ed., Das Nibelungenlied, Re- Genius. The Kane-Donaldson Piers Plowman in Historical Perspective,"
vue critique d'histoire et de litterature 38 (1866): 183-89. in Jerome J. McGann, ed., Textual Criticism and Literary Interpretation
5. Peter F. Ganz, "Lachmann as an Editor of Middle High German (Chicago: University of Chicago Press, 1985), 55-91.
Texts," in Peter F. Ganz and Werner Schroder, eds., Probleme Mittelal- 15· Gianfranco Contini, "Rapporti fra la filologia (come critica tes-
terlicher Oberlie/erung und Textkritik (Berlin, 1968), 12-30. tuale) e la linguistica romanza," Actes du XIIe Congres international de
6. Ernest Renan, L'Avenir de la science. Pensees de I848, in H. Psichari, linguistique et de philologie romanes (Bucharest: Ed. Acad., 1970), I: 63.
ed., Oeuvres completes (Paris: Calmann-Levy, 1950), 3: 883-84. 16. Even with regard to the progressive Lachmannianism of Dom
7. Leon Renier, "Avertissement," in the first volume of the Biblio- Henri Quentin (Essais de critique textuelle [Ecdotiquel [Paris: Picard,
theque de 1'Ecole des hautes etudes series (Paris: Franck, 1869). 1926]), which abandoned the troublesome notion of error in favor of
8. Gaston Paris, "La Chanson de Roland et la nationalite fran<raise," in that of the variant; classification was unbiased, working statistically,
La Polsie du Moyen Age. Lerons et lectures (Paris: Hachette, 1886), 87-II8 . putting all divergences on the same level.
See Howard Bloch, "The 'Oaths of Strasbourg.' The First document and 17· Joseph Bedier. "La Tradition manuscrite du Lai de I'Ombre;
the Birth of Medieval Studies," in D. Hollier, ed., The Harvard History reflexions sur 1'art d'editer les anciens textes," Romania 54 (1928): 161-96,
o/French Literature (Cambridge: Harvard University Press, 1989). 321-56. Quoted from the later, one-volume edition (Paris: Champion,
9. Gaston Paris and Leopold Pannier, La Vie de saint Alexis, powe du 1970).
XIe siecle, et renouvellement! des XIIe, XIIle et.xJl.-e siecles, with prefaces, 18. A neo-Lachmannian response attempted to prove that the pre-
variants, notes, and glossary, Bibliotheque de 1'Ecole des hautes etudes, dominance of two-branched stemmas was a statistically predictable result
sciences philologiques et historiques (Paris: Franck, 1872), pt. 7· of using the manuscript as a means of transmission (Jean Fourquet, "Le
10. Wendolin Foerster and Eduard Koschwitz, AltftanzOsisches Ue- Paradoxe de Bedier," Melanges I945, vol. 2, Etudes des lettres, Publications
bungsbuch (Leipzig: Reisland, 1902), 98-163. de la faculte des Lettres de 1'universite de Strasbourg [Paris: Klincksieck,
II. Sonia Branca, "Les Debats sur la variation au milieu du XIXe sie- 1946], 1-16; Arrigo Castellani, Bedier avait-il raison? La Methode de Lach-
cle," GARS (Universite de Provence), Recherches sur Ie franrais parle 5 mann dans les editions de textes du Moyen Age [Fribourg: Faculte des Let-
(1983): 263-9°. tres, 1957]). But this did not diminish the power of the argument set
12. Fran<;:ois Guessard, "Examen critique de l' Histoire de la formation forth by Bedier.
de la langue franraise par Ampere," Bibliotheque de l1icole des chartes 2 19. Italian neo-Lachmannism is probably explained by the strong
(1840-41); quoted by Sonia Branca. sense of continuity felt by intellectuals in Italy with their Middle Ages
13. Earlier we said that it is nothing of the sort. Medieval inflection, and by the almost internal way in which they perceive it (an example of
to the extent that it exists, has no syntactic pertinence at all. Rather, it this would be Umberto Eco's Nome della rosa). Their more immediate
plays a rhetorical role as ornamentation, marking "proper usage" of the comprehension of the language, which has gone hand in hand with an
written language, as, for example, its connection to the quality and the uninterrupted series of commentaries and editions, provides them with
beauty of the copy demonstrates. Consequently, we have to adopt an en- an assurance that their French colleagues do not have. In France there
tirely different perspective toward it, specifically one that is diachronic: was a dramatic break introduced by Classicism in the second half of the
certain late (and luxurious: Froissart) manuscripts present, in fact, lan- seventeenth century, fixing the language, instituting the belles-lettres,
guage that is extremely correctly "declined" (Bernard Cerquiglini, "Flir dumping everything that had gone before into alterity and suspicion.
ein neues Paradigma der historischen Linguistik: am Beispiel des Alt- But it should be added that the neo-Lachmannian medievalist editors
franzosischen," in B. Cerquiglini and H. U. Gumbrecht, eds., Der (Maria Corti, D'Arco S. Avalle, Cesare Segre, et al.) were, moreover,
Diskurs der Literatur und Sprachhistorie [Francfort: Suhrkamp, 1983], masters of Italian literary semiotics and well known to a wide audience
for their works on Pavese, Joyce, and Pound. A connection can be seen
440-6 3.
14. Leen Patterson, "The Logic ofTextual Criticism and the Way of between their editorial interventionism (which seems very much like a
86 87
Notes to Pages 68-78
return to positivism) and their well-founded conviction that literary texts

are answerable to theory as well as to a strict methodology of reading and
analysis. But the French editors of medieval texts hardly ever step outside
their domain, put forth few general ideas, and are distrustful, to say the
least, of contemporary literary theories. In everything they adopt the ex-
I Index
treme "conservatism" advocated by Bedier-but without quotation

marks.
20. For Cesare Segre, the manuscript represents the contact (as one
would use the term about languages) between the linguistic system of the
text and that of the copyist: it is a diasystem (Cesare Segre, "Critique Abelard,16 Bloch, Marc, 13
textuelle, theorie des ensembles et diasystemes," Academie royale de Bel- Aesthetic of return, 36-37 Boogaard, Nico Van den, 78
gique, Bulletin de fa classe des fettres et des sciences morales et politiques 62 Alcuin, 8, 14 Branca, Sonia, 58
[1976]: 279-92). Alterity Bruns, Gerald, 34
21. Georges Cuvier, Discours sur les revolutions de fa surface du globe, of manuscript, 21-32
et sur les changements qu'elles ontproduits dans Ie regne animal (Paris: Pas- of Old French, 59 Carolingian revival, 14, 22
sard, 1864 [first published 1815]), 71 f£ Auerbach, Erich, 14, 18-19 Catasrrophe as explanation, 69
22. Jean-Baptiste Lamarck, Philosophie zoologique (Paris: Schleicher, Author Chabaille, Polycarpe, 47
in Bedier's view, 70 Chanson de geste, 20, 36, 54
1809), 15.
emergence of, 8-II Chanson de Roland, 38, 47, 53, 54, 55,
privileging of, 61 56,63
and proofreading, 4-6 Charcot, Jean-Martin, 69
Chapter 5: Turn the Page
signature of, xii-xiii, 6 Chevalier au lion, Le (Chretien de
1. Edmond Faral, ed., Villehardouin. La Conquete de Constantinople text as property of, 8, II Troyes),35
(Paris: Les Belles Lettres, 1938). Author's manuscript, 7 Chretien de Troyes, 44-45, 70, 78
2. Josfroi de Vileharduyn, La conqueste de Costentinoble, editee par la Le Chevalier au lion, 35
Bartsch, Karl, 51-52 meaning of name of, 33
section de traitement automatique des textes d'ancien franc;ais du CRAL,
Beaumarchais, P.-A.c. de, 10 Percevah,38-45,78
Nancy, 1978.
Bedier, Joseph, 53, 54-55, 64-68 Comparative linguistics, 77
3. For Rychner see above, chap. 2, n. 10 and chap. 3, n. 5. Peter Dem- "conservatism" of, 88n.19 Computers, xiii, 8, II, 26, 79-82
bowski, ed., Vie de Marie l'Egyptienne, Publications romanes et franc;aises and critical apparatus, 77 Conquete de Constantinople, Le, 62,
(Geneve: Droz, 1977); Willem Noomen and Nico Van den Boogaard, and Lai de I'Ombre, 78 72-75
eds., Nouveau Recueil complet des fabliaux, vols. 1-3 (Assen, Pays-Bas: Van and Gaston Paris, 51, 53, 64, 66, Comini, Gianfranco, 64
Gorcum, 1983-1986). 68, 70 Continuation, in medieval literature,
Bedierism, 52, 66, 68, 70 35-3 6,40
Belles-lettres, 9 Copying, 2-3
Berne Convention, II as degeneration, 61
Bertillon, Alphonse, 50 facsimile, 21, 78-79
Bible, as text, 35 Criticism, xiii
Binary classification, of manuscripts, genetic, xi-xii
66-67 Cuvier, Georges, 68-69
Bindings, 10
88 89
Index Index
Darwin,· Charles, 69 French language Information technology, xiii. See also binary classification of, 66-67
Declension, in Old French, 58-60, classicism as break in, 87 Computers Marie de France, 23, 27, 28-29, 33.
61-62 Old French, 30, 58-60, 61-64 Italian neo-Lachmannism, 87n .19 See also Lai de Lanval
Dembowski, Peter, 78 and segmentation, 23-24, 25 Iwein (Hartmann von der Aue), 52 Medieval author, 8
Diett, Friedrich, 51, 58, 76 and writing, 17-21 Medieval culture. See Medieval
Du Cange, Charles, 35 French literature and texts, 13-14, 20 society and culture
Jean Renart, 33, 40, 64, 65, 66
du Mans, Jacques Peletier, 5 French medieval manuscripts, 26-27. Medieval inflection, 86n.13
Jerome, Saint, 8
See also specific manuscripts Medieval literary inscription, 21
Joinville, Jean de, 60
French medieval philology, 48. See Medieval literary writing,
Ecole des chartes, 47
also Medieval philology paraphrastic utterances in, 75
Ecole des hautes etudes, 55 Kiirenberg, Konrad de, 52
Freud, Sigmund, 69 Medieval literature
Ecranique, x
continuation or repetition in,
Editing, 22, 24
Genealogical tree, in Lachmann's Lachmann, Karl, 48-52,56,65, 35-37
and computer, 80
67,77 as variable, 33-34
"diplomatic," 22 method, 49-50
Lai de Lanval (Marie de France), 23, variance in, 37-38
in empirical period, 47 Genetic criticism, xi-xii
27, 29-32, 78 See also specific works
humanist conception of, 67 Genetics, literary, 3, 33-34
Lai de I'Ombre (Jean Renart), 40, 64, Medieval manuscript, as anthology, 28
and punctuation, 24-25 Genetre, Gerard, 26, 29
65, 67-68, 69, 78 Medieval philology, 13, J4, 61, 72
and reconstruction, 60-61 Goethe-und Schiller-Archiv, 7
Lakanal, Joseph, 10-11 empirical period of, 46-47
scientific, 48 Goody, Jack, 15, 17
Lamarck, ].-B.-P.-A., 69 and methodology of editing, 22
and segmentation, 21-24, 25 Graphic reason, 26
Language, study of, 76-77 positive period of, 47-64
self-doubt in, 67 Guessard, Francis, 47, 58
Latin language, 14-15, 17, 19-20 third period of doubt for, 64-70
and variant forms, 45 Guillaume, Gustave, 77
and French declensions, 59 Medieval society and culture
Education, and texts, I Guiot the scribe, 40, 44, 70
and manuscript reconstruction, 60 development of, 14
Eisenstein, Elizabeth, 8 Guizot, F.-P.-G., 47
Latin manuscripts, 26 and orality, 16-17
Elementarium doctrinae Erudimentum
Laurent, Brother, 24 and philology, 13
(Papias),26
Hartmann von der Aue, 52 and rise of written French, 17-21
Erasmus, 5, 8 Lecoy, Felix, 40, 42
Haver, Louis, 50 and technology of the intellect,
Espurgatore saint Patrice (Marie de Lexicon, 77
Hay, Louis, 7 Libraries, 10 15-16
France),29
Historical linguistics, 17-18, 77 Medieval syntax, 30
Evolution, and Bedier's philology, Linguistics, 58, 75-76
Holbein, Hans, engraving by, 4 Medieval texts, fundamental
69-70 comparative, 77
Hucher, Eugene, 47 instability of, 84n.1O
historical, 17-18, 77
Humanism, xi Medieval writing
nineteenth-century, 76
Fables (Marie de France), 28-29 and Bedier's Lai de I'Ombre and computer processing, 79
Literary genetics, 3, 33-34
Facsimile, 21, 78-79 manuscript, 68 variance of, 77-78
Literary property, 8, II
Faithful copy, 2-3 in editing, 67 Meillet, Antoine, 77
Lombroso, Cesare, 50
Faral, Edmond, 72 and Holbein engraving, 4 Menendez Pidal, Ramon, 38
Lucretius, 48
Febvre, Lucien, I, 9 Michel, Francisque, 46-47
Ferry, Jules, 17 Middle Ages. See at Medieval
Inflection
First page, 9-10 Mallarme, Stephane, 61 Millennium, 14-15
medieval, 86n.13
Fossils, and catastrophe theory, 69 Manuscript, xi Modernity
in Old French, 59-60, 63
Foucault, Michel, 8, 61 alterity of, 21-32 and authorship, 8
uncertainty of, 62
Foulet, Lucien, 53 author's, 7 and corrected proof, 6
9° 91
Index Index
Modernity (cont.) and textual criticism, 46 vs. printing press, 4 Textual studies, I
and text, ix, 2 Phonetics, 76-77 and segmentation, 23 Textuary modernity, ix-x, 81
textuary, 81 and morphology, 59 Segmentation, 22-24, 25 Tide page, 9-10
Morphology, 59, 62, 77 Pre-text, xii, 6 Sequence de sainte EulaIie, 19 Turgot, A.-R.-J., 10
Municius, Aldus, 5 author's manuscript, 7 Serments de Strasbourg, 19 Turold, 55
proof as, 6 Shakespearean works, variants in, 39 Two-branched classification, of
Neo-Lachmannian medievalists, Printing "Shared wrongs," 49 manuscripts, 66-67
87 n .19 and manuscripts, 81 Signature, of author, xii-xiii, 6
New Criticism, 61 and proofreading, 4-6 Simpson, Percy, 6 Variance
Nibelungenlied, 51, 52 and text, 2-4, 8 Society of Dramatic Authors, 10 in medieval literature, 37-38
Nithard,19 and value of pre-text, 7 Somme Ie roi, La (Frere Laurent), 24 of medieval writing, 77-78
Noomen, Willem, 78 and word processing, II Stiennon, Jacques, 24 Variant, 37, 77, 78
Nouveau Recueil complet des fabliaux, Proofreading, 4-6 Stock, Brian, 15 Vascosan, Michel de, 5
78 Punctuation, 24-25 Vie de Marie l'Eyptienne, 78
Tarde, Gabriel, 50 Vie de saint Leger, 19
Orality Ramus, Petrus, 78 Text, xii Villehardouin, Geoffroy de, 62, 72, 78
and French language, 20-21 Raynouard, E-J.-M., 58 in Bedierism, 70 Vogelweide, Walther von der, 52
in medieval society, 16-17 Renan, Ernest, 55 and computer, 81
promotion of, xiii Renart, Jean. See Jean Renart emergence of idea of, I I - a Wailly, Natalis de, 60
Orature, ix-x, 20 Repetition, in medieval literature, "fundamental instability" of, Wolf, Friedrich-August, 52, 54
Oxford University Press, 6 36-37 84 n.1O Written word
Rhetoric, 9, 37 and literary genetics, 3 freedom and progress through, 17
Pannier, Leopold, 56 Roach, William, 40, 41 as literary property, 8, II and French language, 17-21
Paris, Gaston, 51-58, 69 Romance philology, 37 and medieval literature, 34-35 as graphic reason, 26
and Bedier, 51, 53, 64, 66, 68, 70 Roman de la Rose, Continuations of, and modernity, ix, 2 medieval development of, 15-16
and binary classification, 67 35-3 6 and printing, 2-4, 8 and medieval literature, 33
and inflection, 63 Roman de Renart, 47, 53, 78 Textual criticism, 46 screenic presentation of, 79, 81
on reconstruction of French Romanticism, German, 7 Textuality, xiv as threatening, 81
texts, 6o Rychner, Jean, 23, 27, 29, 38, 78
and Saint Alexis, 56, 61-62
Paris, Paulin, 51 Saint Alexis, 56, 61-64, 66
Percevais (Chretien de Troyes), Saint-Graal47
38-45,78 Schleicher, August, 49, 50
Continuations of, 35-36 Screenic presentation, x, 79, 81
Philology, xiv, 12, 37, 72, 77 Scribal culture, 34-36
and evolution, 69-70 Scribes
and family comparison, 49 and Bedierism, 68
history of, xiv Guiot, 40, 44, 70
medieval, 13, 34, 61, 72 (see also and inscription of French, 17
Medieval philology) and Lachmannian philology, 49
post-textuary, 81 as machines, 3
romance, 37 and Gaston Paris, 51
92 93
Library of Congress Cataloging-in-Publicarion Dara
Cerquiglini, Bernard.
[Eloge de la varianre. English]
In praise of rhe varianr: a critical history of philology / by Bernard
Cerquiglini; translated by Bersy Wing.
p. cm. - (Parallax)
Includes bibliographical references and index.
ISBN 0-8018-6126-8 (alk. paper)
1. Criticism. Textual-History. 1. Tide. II. Series: Parallax
(Baltimore. Md.)
P47·C47 13 1999 9 8-45777
801'.959-dc21 elP

In Praise Ofthe Variant: Critical History of Philology

Uploaded by

Copyright:

Available Formats

In Praise Ofthe Variant: Critical History of Philology

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

In Praise Ofthe Variant: Critical History of Philology

Uploaded by

Copyright:

Available Formats

PARALLAX

Stephen G. Nichols, Gerald Prince, and wendy Steiner

In Praise ofthe Variant

A Critical History of Philology

The Johns Hopkins University Press

Originally published as Eloge de fa variante: Histoire

Library of Congress Cataloging-in-Publication Data will

'6lfo ICi C~I

2 Mr. Procrustes, Philologist 13

3 The Joyful Excess 33

4 Gaston Paris and the Dinosaurs 46

5 Turn the Page 72

(which word is usually to be found nearby). In this context, though

case texts. Screenic seemed the most useful English equivalent.

Lucien Febvre, in his inaugural lecture at the College de France, ex-

In the realm of modern studies, young men formed intellectually

There are certainly a thousand ways of charac:terizing the nine-

There are legitimate reasons for examining nineteenth-century philo-

IN PRAISE OF THE VARIANT Mr. Procrustes, Philologist

Mr. Procrustes, Philologist

and differentiated what had been joined: Edited lIersion

tion substitutes one autonomous and closed sentence, with its

IN PRAISE OF THE VARIANT The ]oyjitl Excess

"particularly confusing and obscure" tradition, as Felix Lecoy de- Manuscript T

and the light welling up from the sacred vase:

a tot Ie graal qu'ele tim,

In any literary or mythographic analysis this difference is important.

polygraph who, at the age of twenty-six, living on a stipend provided

The Empirical The Positive

Lachmann as he was deaf to Diez) dismissed Old French as a primi- I, object

All manuscripts are in agreement: Folques OA, Forques B, Foukes

/~ method never had, in fact, been applied to French medieval works

r· Gaston Paris and the Dinosaurs

IN PRAISE OF THE VARIANT Gaston Paris and the Dinosaurs

So the page would be truly turned. Electronic writing, through its

Chapter I: Textuary Modernity

Princeton University Press, 1983). University Press, 1945), 140.

return to positivism) and their well-founded conviction that literary texts

treme "conservatism" advocated by Bedier-but without quotation

You might also like