Languages, Books, and Reading from the

Printed Word to the Digital Text

Roger Chartier

Translated by Teresa Lavender Fagan

I would like to begin this reflection on languages, books, and reading in

the age of electronic textuality with two fables, as the author calls them. The
first relates an enduring nostalgia in the face of a loss of linguistic unity; the
second presents the disturbing reality of its utopian restoration. Borges’s
“The Congress” was published in 1975 in The Book of Sand. In this fable,
Alejandro Ferri, who, like Borges himself, wrote an essay on the analytical
idiom of John Wilkins, is assigned to identify a language that should be used
by the participants of the “Congress of the World,” one that would “rep-
resent all men of all nations.”1 In order to find such a language of universal
communication, the initiators of the project send Ferri to London. He de-
scribes his stay:
I roomed at a modest boarding house behind the British Museum,
where mornings and afternoons I studied in the library in search of a
language worthy of the Congress of the World. I did not overlook uni-
versal languages, investigating both Esperanto—which Lugones quali-
fies as “impartial, simple, and economical”—and Volapük, which,
declining verbs and conjugating nouns, attempts to work out all lin-
guistic possibilities. I also weighed the arguments in favor of and against
the revival of Latin, a nostalgia for which has endured down through
the centuries. I even dwelled on an examination of John Wilkins’ ana-

1. Jorge Luis Borges, “El Congreso,” El libro de arena (Madrid, 1995), p. 35; hereafter abbreviated
“EC”; trans. under the title “The Congress,” The Book of Sand, by Norman Thomas di Giovanni
(New York, 1977), p. 33; hereafter abbreviated “C” (“un Congreso del Mundo que representarı́a a
todos los hombres de todas las naciones”).

Critical Inquiry 31 (Autumn 2004)

䉷 2004 by The University of Chicago. 0093-1896/04/3101-0003$10.00. All rights reserved.

133
134 Roger Chartier / Languages, Books, and Reading
lytical language, in which the definition of each word is to be found in
the letters that spell it out. [“C,” p. 42]2
Ferri examines the three types of languages that are capable of going beyond
the infinite diversity of vernacular languages. First of all, he looks at the
artificial languages invented during the nineteenth and twentieth centuries,
such as Esperanto or Volapük that were supposed to assure understanding
and concord between all peoples.3 Second, he considers a possible return
to a language that was, throughout the history of the West, a universal ve-
hicle of communication: Latin. And, third, Ferri examines the formal lan-
guages that promise, as the philosophical language of Wilkins, a perfect
correspondence between words in which each letter is a signifier and the
categories, elements, and the individual things that the words describe. If
we look at what Borges himself wrote on this language that was invented in
England in 1668, we can see how the perfect language functions: “de means
element; deb, the first of the elements, fire; deba, a portion of the element
of fire, a flame.”4 In this analytical, perfect language, each word is defined
by itself, and the language becomes a classification of the universe.
In the end, Ferri’s research at the British Library proves useless. To con-
vene a Congress of the World was ultimately an absurd notion, as Don Ale-
jandro, the man who initiated the project, admits:
“It has taken me four years to understand what I am about to say,” don
Alejandro began. “My friends, the undertaking we have set for ourselves
2.
Me hospedé en une módica pensión a espaldas del Museo Británico, a cuya biblioteca concurrı́a
de mañana y de tarde, en busca de un idioma que fuera digno del Congreso del Mundo. No
descuidé las lenguas universales; me asomé al esperanto—que el Lunario sentimental califica de
“equitativo, simple y económico”—y al Volapük, que quiere explorer todas las posibilidades
lingüı́sticas, declinando los verbos y conjugando los sustantivos. Consideré los argumentos en
pro y en contra de resucitar el latı́n, cuya nostalgia no ha cesado de perdurar al cabo de los siglos.
Me demoré asimismo en el examen del idioma analı́tico de John Wilkins, donde la definición de
cada palabra está en las letras que forman. [“EC,” pp. 45–46]
3. See Anne Rasmussen, “A la recherche d’une langue internationale de la science 1880–1914,” in
Sciences et langues en Europe, ed. Roger Chartier and Pietro Corsi (Paris, 1996), pp. 139–55.
4. Borges, “El idioma analı́tico de John Wilkins,” Otras inquisiciones (Madrid, 1976); trans.
under the title “The Analytical Language of John Wilkins,” Other Inquisitions 1937–1952, by Ruth
L. C. Simms (Austin, Tx., 1964), p. 102.

R o g e r C h a r t i e r is directeur d’études at the École des Hautes Études en

Sciences Sociales, Paris, and Annenberg Visiting Professor in History at the
University of Pennsylvania. His work is devoted to the history of the book,
publishing, and reading in early modern Europe. Te re s a L a v e n d e r F a g a n
is a freelance translator based in Chicago. She has published over a dozen book-
length translations and has translated numerous essays and articles.
 Critical Inquiry / Autumn 2004 135
is so vast that it embraces—I now see—the whole world. Our Congress
cannot be a group of charlatans deafening each other in the sheds of an
out-of-the-way ranch. The Congress of the World began with the first
moment of the world and it will go on when we are dust. There’s no
place on earth where it does not exist.” [“C,” p. 47]5
Therefore, the search for a universal language was as futile an idea as that
of convening a Congress of the World because the world is already there,
made up of an insurmountable diversity of places, things, individuals, and
languages. To attempt to erase that multiplicity is to design a disturbing
future.
In another fable also published in The Book of Sand, “Utopia of a Tired
Man,” the world of the future, in which the narrator is lost, has returned
to linguistic unity.6 The visitor to the future, Eudoro Acevedo, who, like the
author, is a professor of English and American literature and a writer of
imaginative tales and who, like Borges when he was director of the Bibli-
oteca Nacional in Buenos Aires, has an office on Mexico Street, does not
know how to communicate with the very tall man he encounters:
I tried out various languages, and we did not understand each other.
When at last he spoke, he did so in Latin. I dusted off what I remem-
bered from my now distant school days, readying myself for conversa-
tion.
“By your clothes, I see you come from another century,” he said. “A
diversity of tongues favored a diversity of peoples and even of wars. The
world has fallen back on Latin. There are those who fear it may degener-
ate again into French, Lemosi, or Papiamento [a language that sounds
like birdsongs, spoken by certain island peoples of Central America],
but that is not an immediate risk. Be that as it may, neither the past nor
the future interests me.” [“U,” p. 90]7
5. “Cuatro años he tardado en comprender lo que les digo ahora. La empresa que hemos
acometido es tan vasta que abarca—ahora lo sé—el mundo entero. No es unos cuantos
charlatanes que aturden en los galpones de una estancia perdida. El Congreso del Mundo
comenzó con el primer instante del mundo y proseguirá cuando seamos polvo. No hay un lugar
en que no esté” (“EC,” p. 51).
6. See Borges, “Utopı́a de un hombre que está cansado,” El libro de arena, pp. 96–106; hereafter
abbreviated “UH”; trans. under the title “Utopia of a Tired Man,” The Book of Sand, pp. 89–104;
hereafter abbreviated “U.”
7.
Ensayé diversos idomas y no nos entendimos. Cuando él habló lo hizo en latı́n. Junté mis ya
lejanas memorias de bachiller y me preparé para el diálogo.
“Por la ropa . . . veo que llegas de otro siglo. La diversidad de las lenguas favorecı́a la
diversidad de los pueblos y aun de las guerras; la tierra ha regresado al latı́n. Hay quienes
temen que vuelva a degenerar en francés, en lemosı́n o en papiamento, pero el riesgo no es
inmediato.” [“UH,” p. 97]
136 Roger Chartier / Languages, Books, and Reading
The linguistic unity that has been recovered thanks to a return to Latin at
the same time signifies the loss of history, identities, and names: “‘You said
your name is Eudoro. I can’t tell you my name, because I’m simply called
Someone’” (“U,” p. 91).8 And, what is worse, this return to a world without
memory, without museums, without libraries, leads to an accepted destruc-
tion. Leaving the house with its inhabitants, Eudoro Acevedo notices a dis-
turbing building: “In the distance I made out a kind of tower, crowned with
a cupola. ‘It’s the crematory,’ someone said. ‘Inside it is the lethal chamber.
It’s said to have been invented by a philanthropist whose name, I think, was
Adolf Hitler’” (“U,” p. 96).9 The utopia of a world without differences, with-
out inequality, without a past ultimately resembles death. Commenting in
the epilogue of The Book of Sand on the different tales he had assembled,
Borges states that “Utopia of a Tired Man” is “the most honest and mel-
ancholy piece in the collection.”10 Melancholy, no doubt, because every-
thing that in classic utopias appears to promise a better future, withoutwars,
without poverty or wealth, without governments here leads to the loss of
that which defines the very humanity of human beings: memory, names,
and differences.

These various lessons Borges teaches us are perhaps not irrelevant for a
better understanding of the present. Indeed, what is the language of this
new “World Congress” that is being constructed by electronic communi-
cation? Such a language recalls the three characteristics of the universal lan-
guages Ferri encountered in the British Library. The first, the most
immediate, the most obvious characteristic is related to a domination by a
specific language, English, as the universally accepted language of com-
munication, within and beyond the electronic medium, both for scientific
publications and for informal exchanges. Such domination reflects the con-
trol exerted by the most powerful—that is, American—multimedia com-
panies over the digital database market, the websites, and the production
and dissemination of information. As in the frightening utopia imagined
by Borges, such an imposition of a single language and of the cultural model
that it conveys can only lead to the mutilating destruction of diversity.
But this new way of raising the “questione de la lengua”—as the Italians
of the Renaissance, from Pietro Bembo to Baldassare Castiglione, once
8. “Me has dicho que te llamas Eudoro; yo no puedo decirte cómo me llamo, porque me dicen
alguien” (“UH,” p. 99).
9. “Divisé una suerte de torre, coronada por una cúpula.—Es el crematorio—dijo alguien—.
Adentro está la cámara letal. Dicen que la inventó un filántropo cuyo nombre, creo, era Adolfo
Hitler” (“UH,” pp. 105–6).
10. Borges, afterword to El libro de arena, p. 124; trans. under the title afterword to The Book of
Sand, p. 124 (“La pieza más honesta y melancólica de la serie”).
 Critical Inquiry / Autumn 2004 137
said—which is connected to the domination of English, must not obscure
two innovations of electronic textuality. On the one hand, the electronic
text reintroduces into writing an element of the formal languages that, be-
ginning in the eighteenth century, or even the seventeenth, sought a sym-
bolic language capable of adequately representing different processes and
registers of thought. Thus Condorcet in his Esquisse d’un tableau historique
des progrès de l’esprit humain, written in prison during the French Revo-
lution, stresses the need for a universal language capable of formalizing cog-
nitive operations and logical reasoning and that could be translated into
different vernacular languages. Such a universal language could be written
with conventional signs—symbols, charts, tables—which he calls technical
methods, that enable the formal transcription of relationships between ob-
jects to be known and operations for knowing them. The universal language
that Condorcet imagines is made possible by the invention and the dissem-
ination of the printing press that allows the fixing and standardization of
graphic conventions.11 In the contemporary world a new formal language
is being created in relation to the electronic text, a language that proposes
pictographic signs that are immediately decipherable by everyone regardless
of the idiom that they speak and write. Emoticons are symbols that use
different keyboard characters (parentheses, comma, semicolon, colon) to
indicate the emotional significance the writer wishes to attribute to what
he or she is writing: joy :-) sadness :-( irony ;-) anger :-@ and so on. Emo-
ticons reflect the search, within the language written on the screen, for a
nonverbal language that, for this very reason, enables the communication
of emotions and can unambiguously establish the semantic register on
which the discourse should be understood.
On the other hand, it is possible to say that the English of electronic
communication is at least as much an artificial language with a unique vo-
cabulary and syntax as it is a particular idiom raised to the level of universal
language. In a certainly less obvious way than what can be said of the lan-
guages invented in the nineteenth century, the English transformed into a
lingua franca of the electronic world is nevertheless a sort of new language
that reduces its lexicon, simplifies its grammar, invents new words, and
multiplies abbreviations. The ambiguity of a universal language that is
rooted in a specific language but which nonetheless utilizes original con-
ventions has several important consequences.
First, there is a reinforcement in the United States of the belief in an
unshared hegemony of the English language and thus the implication that

11. See Chartier, Culture écrite et société: L’Ordre des livres (XIVe–XVIII siècle) (Paris, 1996), pp.
20–24.
138 Roger Chartier / Languages, Books, and Reading
it is unnecessary to learn other languages. A long time before the devel-
opment of the digitalized world, a former governor of Texas, “Ma” Fergu-
son, declared: “If English was good enough for Jesus, it’s good enough for
the children of Texas.”12 In our times, according to statistics published by
the New York Times in April 2001, only 8 percent of American high school
and college students study a foreign language.13 Second, electronic English
to a certain extent implies a specific training that is not necessarily obtained
through the study of the English language in its classic form. As Geoffrey
Numberg points out: “The English that one finds on the web is in some
ways more difficult than that which is required to be able to write formal
texts.”14
A third consequence is the graphic imperialism of English, a language
that has neither tildes nor accents, and which often forces languages such
as French, Italian, Portuguese, or Spanish to eliminate them on the com-
puter screen. As Emilia Ferreiro notes, linguistic imperialism is thus accom-
panied by a graphic imperialism that subjects other languages to its laws
when they are written or read through the medium of electronic commu-
nication.15
It is possible to qualify these statements because the gap between the
English-language community and others in the electronic world is narrow-
ing. Nevertheless the data published by the electronic site Global Internet
Statistics show that an overwhelming majority of electronic addresses are
located in English-speaking countries—45 percent, as compared to only 4.5
percent for Spanish, 3.7 percent for French, and 2.5 percent for Portuguese.16
Such an imbalance is a clear reflection not of the respective demographic
weight of the different linguistic communities but of their unequal level of
development, whether economic, social, or cultural.
The second reality is that progress in the teaching and learning of lan-
guages, particularly in Europe and Latin America, if not in the United States,
has enabled the possibility of communications in which each person uses
his or her own language while being capable of understanding that of some-
one else. It is from this perspective that we might define a modern poly-
glotism such as that which Umberto Eco proposes in his book on the search
for the perfect language:

12. The quote can be found at [Link]/issues/dispatch/2002–06–14/

[Link]
13. See New York Times, 16 Apr. 2001, pp. A1, A10.
14. Geoffrey Numberg, “La Langue des sciences dans le discourse électronique,” in Sciences et
langues en Europe, p. 254.
15. See Emilia Ferreiro, Pasado y presente de los verbos leer y escribir (Mexico City, 2002), pp.
55–56.
16. See Global Internet Statistics, [Link]
 Critical Inquiry / Autumn 2004 139
Generalized polyglottism is certainly not the solution to Europe’s cul-
tural problems, like Funes ‘el memorioso’ in the story by Borges, a global
polyglot would have his or her mind constantly filled by too many im-
ages. The solution for the future is more likely to be in a community of
peoples with an increased ability to receive the spirit, to taste or savour
the aroma of different dialects.17
Here we have a fundamental challenge concerning linguistic studies that
should enable individuals not to speak or write every language, but at the
very least to understand a rather large number of them so that multilingual
communication will become possible. This also implies a pedagogical pro-
ject that has consequences for the teaching of languages and a civic project
that alone can prevent the domination of a single language, whatever it
might be.
Monolingual or polyglot, the world of electronic communication is a
world of textual overabundance in which the written texts that are offered
go far beyond the reader’s ability to take advantage of them. Often, literature
has denounced the uselessness of accumulated books, the excessive number
of texts, the overload of information.18 Such a diagnosis expresses a strong
concern over a growing textual world that has become uncontrollable. In
the utopian world of “Utopia of a Tired Man,” the dialogue between Ac-
evedo and the nameless man he has encountered in the future describes this
anxiety in its own way. Acevedo leafs through a 1518 edition of Thomas
More’s Utopia and declares:
“It’s a printed book. At home, I had over two thousand of them, though
they were neither as old nor as valuable as this one.” I read the title
aloud.
The man laughed. “No one can read two thousand books. In the four
centuries I have lived, I haven’t read more than half a dozen. Besides, re-
reading, not reading, is what counts. Printing—which is now abolished,
since it tended to multiply unnecessary texts to the point of dizziness—
was one of man’s worst evils.” [“U,” p. 92]19

17. Umberto Eco, La ricerca della lingua perfetta nella cultura europa (Rome, 1993); trans. under
the title The Search for the Perfect Language, by James Fentress (Oxford, 1995), pp. 350–51.
18. See Ann Blair, “Reading Strategies for Coping with Information Overload ca. 1550–1700,”
Journal of the History of Ideas 64 (Jan. 2003): 11–28.
19.
“Es un libro impreso. En casa habrá más de dos mil, aunque no tan antiguos ni tan preciosos.”
Leı́ en voz alta el tı́tulo.
El otro se rió.
140 Roger Chartier / Languages, Books, and Reading
More than three centuries earlier, the dialogue that Lope de Vega invents
in his comedia Fuente Ovejuna between Barrildo the peasant and Leonelo,
a student who has just come back from Salamanca, illustrates the same lack
of trust in the multiplication of books brought about by the invention of
the printing press—a recent invention in 1476, the date of the historic events
recounted in Lope’s comedia. Barrildo praises the effects of the printing
press (“So many books are being printed now, / There’s not a soul but boasts
he is a sage!”), and Leonelo responds: “It seems to me they know less than
before, / Because the great excess of books creates / Confusion in the minds
of readers now, / And avid readers are the most confused / By all the titles.”
Surprised by this remark, Barrildo asserts: “But printing is important, all
the same,” of which the “licenciado” is hardly convinced: “The world got
on without it very well / For many centuries, and in this one / There is no
St. Jerome or Augustine!”20 The increasing number of books is a source of
confusion rather than of knowledge, and the printing press that generated
such an excessive number of books did not produce any new geniuses.
This leads us to a question regarding the present: how are we to view
reading when we are confronted with textual offerings that are increasing
even more rapidly through electronic technology than they ever did with
the invention of the printing press? In 1725 Adrien Baillet wrote in a work
entitled Jugemens des savants sur les principaux ouvrages des auteurs: “We
have reason to fear that the multitude of books that is increasing every day
in a prodigious manner will put the centuries to come into as difficult a
state as that in which barbarity had put the earlier ones after the fall of the
Roman Empire.”21 Was Baillet right, have we fallen into a textual barbarity
similar to that which followed the fall of the Roman Empire and do we share
the same anxiety vis-à-vis the excessive number of texts, information over-
load, and uselessness of accumulated discourse? To respond to this ques-
tion, we must carefully distinguish between the different levels of changes
that characterize the revolution of digital texts. The first changes relate to

“Nadie puede leer dos mil libros. En los cuatro siglos que vivo no habré pasado de una
media docena. Además no importa leer, sino releer. La imprenta, ahora abolida, ha sido uno
de los peores males del hombre, ya que tendió a multiplicar hasta el vértigo textos
innecesarios.” [“UH,” p. 100]
20. Lope de Vega, Fuente Ovejuna, trans. William E. Colford (Woodbury, N.Y., 1969), pp. 63, 65.
Spanish text: “Barrildo: Después que vemos tanto libro impreso, / no hay nadie que de sabio no
presuma. Leonelo: Antes que ignoran más siento por eso, / por no se reducir a breve suma; /
porque la confusión, con el exceso, / los intentos resuelve en vana espuma; / y aquel que de leer
tiene más uso, / de ver leteros sólo está confuso” (ibid., p. 62).
21. Adrien Baillet, “Advertissement au lecteur,” Jugemens de savants sur les principaux ouvrages
des auteurs (Amsterdam, 1725). I owe this reference to Ann Blair.
 Critical Inquiry / Autumn 2004 141
the order of discourse, the second to that of reason, and the third to that of
property.
In the order of discourse, we are confronted with what is perhaps the most
fundamental rupture of all. Indeed, in the written culture as we know it, such
an order is established from the relationship between objects (the letter, the
book, the newspaper, the magazine, the poster, the form, and so on), cate-
gories of texts, and different uses of the written word. This connection, which
links types of objects, categories of texts, and forms of reading, is the result
of a historical layering of three fundamental innovations. The first took place
in the early centuries of the Christian era when the codex as we still know it,
that is, a book made up of leaves and pages gathered into the same binding
or covering, replaced the scroll or volumen, that book of a completelydifferent
structure which belonged to Greek and Roman readers.22
The second rupture occurred in the fourteenth and fifteenth centuries,
before Gutenberg’s invention, with the appearance of the libro unitario, as
Armando Petrucci called it, the book that, for texts in the vernacular, as-
sembled in the same binding all the works of a single author, or even just
one sole work.23 Although this type of book was already the rule for juridical
collections, canonical works of the Christian tradition, or the classics of
antiquity, the same was not true for texts in the vernacular that, in general,
were assembled into miscellanies made up of works of different dates, gen-
res, or languages. It was around figures such as Petrarch or Boccaccio, Chris-
tine de Pisan or René of Anjou that, for “modern” writers, the unitary book
was born, that is, a book in which the connections between the material
object, the work (in the sense of a specific work or of a series of works), and
the author were finally established.
The third legacy is, of course, the invention of the printing press in the
mid-fifteenth century. From that moment on, without it having caused the
disappearance of scribal publication—far from it—the printing press be-
came the most utilized technology for the reproduction of the written word
and for the production of books.24

22. See Colin H. Roberts and T. C. Skeat, The Birth of the Codex (London, 1987); Les Débuts du
codex: Actes de la journée d’étude, ed. Alain Blanchard (Turnhout, 1989); and Guglielmo Cavallo,
“Testo, libro, lettura,” in Lo spazio letterario di Roma antica, ed. Cavallo, Paolo Fedeli, and Andrea
Giardino, 5 vols. (Rome, 1989–91), 2:307–41 and “Libro e cultura scritta,” in Storia di Roma, ed.
Aldo Schiavone, 4 vols. (Torino, 1988–93), 4:693–734.
23. See Armando Petrucci, “From the Unitary Book to Miscellany,” Writers and Readers in
Medieval Italy: Studies in the History of Written Culture, trans. Charles M. Radding (New Haven,
Conn., 1995), pp. 1–18.
24. See Elizabeth Eisenstein, The Printing Press as an Agent of Change: Communications and
Cultural Transformations in Early Modern Europe (Cambridge, 1979) and The Printing Revolution
in Early Modern Europe (Cambridge, 1983), and Adrian Johns, The Nature of the Book: Print and
Knowledge in the Making (Chicago, 1998).
142 Roger Chartier / Languages, Books, and Reading
We are the heirs of these three historical developments, as much for the
definition of the book that is for us both an object different from other
objects of written culture and an intellectual or esthetic work endowed with
an identity and a coherence assigned to its author, as for a perception of
written culture based on immediate, material distinctions between objects
that are associated with different textual genres and usages.
It is this order of discourse that electronic textuality calls into question.
Indeed, it is the same apparatus, in this case the computer screen, that en-
ables different types of texts to appear in front of the reader, texts that, in
the world of the scribal and a fortiori printed cultures, were distributed
among distinct objects. In the digital word, all texts, whatever their genre,
are produced or received through the same medium and in very similar
forms, usually decided on by the reader him- or herself. Thus a textual con-
tinuity is created that no longer differentiates discourses on the basis of their
materiality. Hence the anxiety or the confusion of readers who must con-
front or overcome the disappearance of the most strongly internalized cri-
teria that enabled them to distinguish, to classify, and to categorize different
types of discourse.
Because of this, the perception of individual works as works becomes
more difficult. Reading in front of the computer screen is generally a dis-
continuous reading process that seeks, using keywords or thematic head-
ings, the fragment that the reader wishes to find: an article in an electronic
periodical, a passage in a book, or some information in a website. This is
done without the identity or coherence of the entire text from which the
fragment is extracted necessarily being known. In a certain sense, one might
say that in the digital world all textual entities are like databases that offer
fragments, the reading of which in no way implies a perception of the work
or the body of works from which they come.
Regarding the order of discourse, the electronic world thus creates a tri-
ple rupture: it provides a new technique for inscribing and disseminating
the written word, it inspires a new relationship with texts, and it imposes a
new form of organization on texts. The originality and the importance of
the digital revolution must therefore not be underestimated insofar as it
forces the contemporary reader to abandon—consciously or not—the vari-
ous legacies that formed it. This new form of textuality no longer uses print-
ing (at least in its typographic sense), it has nothing to do with the libro
unitario, and it is foreign to the material nature of the codex. It is therefore
a revolution that in the same period in time, and for the first time in history,
combines a revolution in the technical means for reproducing the written
word (as did the invention of the printing press), a revolution in the me-
dium of the written word (like the revolution of the codex), and a revolution
 Critical Inquiry / Autumn 2004 143
in the use of and the perception of texts (as in the various revolutions in
reading). This no doubt explains the confusion of the contemporary reader
who must transform not only the intellectual categories he or she has em-
ployed to describe, structure, and classify the world of books and of other
written materials but also his or her most immediate perceptions, habits,
and gestures.
The second change concerns the order of reasoning, if by that we mean
the way an argument is organized and the criteria that a reader might em-
ploy to agree or to disagree with it. For the author, electronic textuality en-
ables the development of demonstrations and arguments following a logic
that is no longer necessarily linear or deductive, as is the logic imposed by
the inscription, however it is done, of a text onto a page. It enables an open,
fragmented, relational articulation of the reasoning, made possible by hy-
pertextual connections.25 For the reader, the validation or the refutation of
an argument can henceforth occur by consulting texts (but also images,
recorded speech, or musical compositions) that are the very object of the
study, provided, or course, that they are accessible in digital form. If that is
the case, the reader is no longer constrained to trust the author; he or she
can in turn carry out all or part of the author’s research. Here we have a
fundamental epistemological mutation that profoundly transforms the
techniques of proof and the modalities of the construction and validation
of the discourse of knowledge.26
Let us take an example: In the world of the printed word, the history
book assumes a pact of trust between the historian and the reader. Notes
refer to documents that the general reader will not be able to obtain. Biblio-
graphical references mention books that the reader, most often, will only
be able to read in a library. Citations are fragments that the historian alone
chooses to extract, without the reader being able to know the complete text
out of which they came. These three classic means for a proof (notes, ref-
erences, and citations)27 are profoundly modified in the world of digital
25. For the new possibilities in argumentation offered by the electronic text, see David Kolb,
“Socrates in the Labyrinth,” in Hyper/Text/Theory, ed. George P. Landow (Baltimore, 1994), pp.
323–44, and Jane Yellowlees Douglas, “Will the Most Reflexive Relativist Please Stand Up:
Hypertext, Argument, and Relativism,” in Page to Screen: Taking Literacy into the Electronic Era, ed.
Ilana Snyder (London, 1988), pp. 144–61.
26. For the definitions of hypertext and hyperreading, see J. David Bolter, Writing Space: The
Computer, Hypertext, and the History of Writing (Hillsdale, N.J., 1991); Landow, Hypertext: The
Convergence of Contemporary Critical Theory and Technology (Baltimore, 1992) and Hypertext 2.0
(Baltimore, 1997); Snyder, Hypertext: The Electronic Labyrinth (Melbourne, 1996); and Nicholas C.
Burbules, “Rhetorics of the Web: Hyperreading and Critical Literacy,” in Page to Screen, pp. 102–
22.
27. On the citation, compare Michel de Certeau, L’Écriture de l’histoire (Paris, 1975), pp. 109–13;
on the note, compare Anthony Grafton, The Footnote: A Curious History (Cambridge, Mass.,
1998).
144 Roger Chartier / Languages, Books, and Reading
textuality from the moment when the reader is in a position to be able to
read a book read by the historian and to directly consult the analyzed doc-
uments him- or herself. The first uses of these new modalities in the pro-
duction, organization, and accreditation of scholarly discourse show the
importance of the transformation of cognitive operations implied by the
use of electronic texts.28
A third level of change is linked to the order of property, meaning prop-
erty both in a juridical sense, that of literary property or of copyright, and
in a textual sense, that of the characteristics proper to each written text or
genre. The electronic text as we know it or have known it is a moving, mal-
leable, open text. The reader can intervene not only in its margins, but in
its very content, by removing, reducing, adding, or reworking textual units.
Unlike the manuscript or printed objects in which the reader can add his
or her writing only within the blank spaces of the handwriting or typo-
graphical composition, with the digital world the reader can intervene
within the text itself. The consequences are important. They lead to the
disappearance of the name and the presence of the author because the text
is constantly modified by a multiple and collective writing. One might think
that this possibility offers writing the new form Michel Foucault has dreamt
of many times while imagining an order of discourse in which the individual
appropriation of texts would disappear and in which each writer, anony-
mously, would leave his or her mark in the layers of an authorless dis-
course.29
But the mobility of the open and malleable text seriously challenges the
criteria and the categories that, at least since the eighteenth century, estab-
lished the author’s ownership of his or her works and consequently the pub-
lisher’s ownership of the works he or she has acquired. The recognition of
copyright (the word appeared in 1704 in the registers of the Stationers Com-
pany) implies that the work can be firmly identified in its uniqueness and
originality. Thus in the eighteenth century Blackstone, one of the lawyers
involved in the trials that have contributed to the birth of the notion of

28. For an example of the possible connections between historical demonstration and
documentary sources, see the two forms, printed and electronic, of the article by Robert
Darnton, “Presidential Address: An Early Information Society: News and the Media in
Eighteenth-Century Paris,” American Historical Review 105 (Feb. 2000): 1–35 and AHR webpage:
[Link]/ahr/. See, as examples, for theoretical physics, Josette F. de la Vega,
La Communication scientifique à l’épreuve de l’Internet: L’Émergence d’un nouveau modèle
(Villeurbanne, 2000), in particular pp. 181–231; for philology, Filologı́a e Informática: Nuevas
tecnologı́as en los estudios filológicos, ed. José Manuel Blecua (Barcelona, 1999); and for history,
Rolando Minuti, Internet et le métier d’historien: Réflexions sur les incertitudes d’une mutation
(Paris, 2002).
29. See Michel Foucault, “Qu’est-ce qu’un auteur?” Dits et écrits, 1954–1988, ed. Daniel Defert,
François Ewald, and Jacques Lagrange, 4 vols. (Paris, 1994), 1:789–821.
 Critical Inquiry / Autumn 2004 145
literary property, justified the ownership of the author by stating that a work
is always the same if, beyond the variations in its material forms, what he
described as sentiment, style, or language can be recognized. A close con-
nection is thus established between the unique identity of the texts, always
detectable and perpetuated, and the juridical and aesthetic regime that at-
tributes ownership to their authors.30 This is the foundation of the notion
of copyright that protects a work that is assumed always to be the same
regardless of the forms of its publication. It is obvious that the palimpsestic
and polyphonic texts of digital textuality challenge the very possibility of
recognizing such a fundamental identity.31
This challenge has led to a reflection that has begun in the last few years
regarding whether or not it is possible in digital textuality to reconstitute a
perpetuated and perceptible identity for texts or, at least, for certain texts.
This has also led to the suggestion for a reorganization of the digital world
so that authors’ rights, as well as those of their publishers, can be protected.
This reorganization could lead to a stronger distinction (even if it is made
difficult by the medium, which is a single machine transmitting different
sorts of texts) between, on the one hand, electronic communication as we
know it, which makes it possible to send or receive open, mobile, free texts,
and on the other, electronic publishing, which is the result of editorial work
and implies that the text is fixed, delimited, and closed so that its ownership
is clearly defined and, by that fact, so are the rights of the author and the
income of the publisher. This discussion has truly crystallized around the
advent of the e-book because this new type of computer product does not
allow the transmission, copying, modification, or even printing of texts
published in an electronic form and placed on the market. Electronic pub-
lishing, which implies the same operations as conventional publishing (the
preparation of texts, the creation of a catalogue, copyediting) would thus
be defined in contrast to the free and spontaneous communication on the
digital network.32
The tension between the free communication of ideas and research and
electronic publishing that fixes and closes texts is a major issue in the con-
flicts between scientific communities and publishers. In the last few years,
a very heated controversy has opposed scientific journals, which have in-
creased the number of electronic editions protected by passwords that pro-

30. See Mark Rose, Authors and Owners: The Invention of Copyright (Cambridge, Mass., 1993),
pp. 89–90.
31. See Jane C. Ginsburg, “Copyright without Walls? Speculations on Literary Property in the
Library of the Future,” Representations, no. 42 (Spring 1993): 53–73.
32. See Darnton, “The New Age of the Book,” New York Review of Books, 18 Mar. 1999, pp. 5–7.
146 Roger Chartier / Languages, Books, and Reading
hibit the copying or printing of articles in order to maintain a captive market
for journals, subscriptions to which can cost up to $10,000 or $12,000, and
researchers, in particular those in the biological and cognitive sciences, who
demand free access to the advances in research. Two different logics are at
issue here: the logic of free communication, which is associated with the
ideal of the Enlightenment that upheld the sharing of knowledge, and the
logic of publishing based on the notions of an author’s rights and of com-
mercial gain. For reaching a compromise, certain journals, such as Molec-
ular Biology of the Cell or Science, have agreed to allow their articles to be
freely consulted after a few months or a year of restricted access.33
The example of journals illustrates the profound difference that exists
between readings of the “same” text when it is moved from a printed me-
dium to an electronic form. The case of newspapers is particularly illumi-
nating. In the printed newspaper, the meaning the reader gives to each
article depends on the presence, on the same page or in the same issue, of
other articles or other elements (photographs, cartoons, advertisements,
and so on). The reader constructs the meaning of any article by relating it,
even unconsciously, to what precedes it, accompanies it, or follows it, and
from his or her perception of the editorial intent and of the intellectual or
political design that governs the publication. In an electronic form, a read-
ing of the “same” article is organized out of the logical architecture that
structures the domains, the themes, the headings, and the keywords. Such
a reading proceeds in the same way as the analytical language of Wilkins,
that is, from an encyclopedic organization of knowledge that proposes texts
to the reader that have no other context than that attributed by their be-
longing to a certain theme. This difference must be pointed out at a time
when, in all the libraries of the world, people are discussing the need to
create digital collections, in particular of newspapers and journals. Digiti-
zation projects that will enable long-distance communication are absolutely
essential. But they must never lead to the abandonment or, worse, the de-
struction of printed works in their original form.
The heated debate opened in the United States over novelist Nicholson
Baker’s Double Fold: Libraries and the Assault on Paper, devoted to the de-
plorable effects of microfilming collections of books and newspapers, shows
that the fear of more destruction, through digitization this time, is not with-
out precedent.34 Since the 1960s, the Council on Library Resources has up-
held a policy of reproducing millions of volumes and periodicals on
microfilm, citing two justifications: the need to empty out library storage

33. See Libération, 14–15 Apr. 2001, pp. 16–17.

34. See Nicholson Baker, Double Fold: Libraries and the Assault on Paper (New York, 2001).
 Critical Inquiry / Autumn 2004 147
areas to make room for new acquisitions and the importance of preserving
texts by transferring them onto a new medium. This policy was carried to
the extreme in England in 1999 when the British Library decided to sell or
destroy all its collections of American newspapers dating after 1850 after
microfilming them. The consequences were disastrous on both sides of the
Atlantic, with the disappearance of entire collections, destroyed during the
work of microfilming itself or broken up in order to sell individual issues.
The scandal was such that in England and the United States a step was taken
backward in the assault on paper and the great massacre of newspapers and
books was halted.35 But the losses are immense and irreparable. That lesson
must not be forgotten today, when the possibilities offered by digitization
increase the number of collections accessible from a distance but also re-
inforce the idea that there is an equivalence between media and that a text
is still the same regardless of its form: printed, microfilmed, or digital. That
notion is fundamentally wrong since the processes through which a reader
attributes meaning to a text depend, consciously or not, not only on the
semantic content of the text but also on the material forms through which
the text was published, distributed, and received.36 Therefore, it is essential
that the ability to consult texts in their successive forms be preserved and
that the process of digitization—completely necessary, by the way—never
lead to the destruction of the objects that transmitted written works to the
readers of the past and even of the present.
As D. F. McKenzie strongly emphasizes, “forms effect sense.”37 A text is
always conveyed by a specific materiality: the written object upon which it
is copied or printed; the voice that reads, recites, or otherwise utters it; the
performance that allows it to be heard. Each of these forms of publication
is organized in its own unique fashion, and each form, in different ways,
influences how meaning is produced. Thus, looking only at the printed text,
the format of the book, the layout, the divisions of the text, typographic
conventions, punctuation, all are invested with an expressive function. As
a result of different intentions and interventions (those of the author, pub-
lisher, printer, typesetters, or copyeditors), these elements are used to give
a status to the text, to constrain its reception, to control the reader’s un-
derstanding of it. Guiding the reader’s—or the listener’s—unconscious,
they govern, at least in part, the process of interpreting and appropriating
the written word.
Such a perspective forces one to reject all critical approaches according

35. See Darnton, “The Great Book Massacre,” New York Review of Books, 26 Apr. 2001, pp. 16–
19.
36. See D. F. McKenzie, Bibliography and the Sociology of Texts (London, 1986).
37. Ibid., p. 9.
148 Roger Chartier / Languages, Books, and Reading
to which meaning is produced only out of the automatic and impersonal
functioning of language. This position, which completely detaches the text
from its materiality, is the one held by structuralist critics or by the New
Criticism. It is based on several postulates: the reduction of the text to its
verbal structure alone, the insignificance of the author whose intention is
without any importance, and the separation between the single or several
meanings of the work and the historical modalities of its transmission, re-
ception, and interpretation.
In a 1977 essay McKenzie proposed an approach that was completely dif-
ferent from the one that viewed a text as having no materiality, no author,
and no readers.38 The analysis of the innovations introduced into the 1710
publication of the works of William Congreve, by the playwright himself;
his publisher, Jacob Tonson; or the master printer, John Watts shows how
material and formal changes, apparently devoid of textual significance (for
example, the change from the quarto to the octavo format, the numbering
of scenes, the presence of an ornament between each scene, listing at the
beginning of each scene the names of the characters to be present at the same
time on the stage, stage directions) have an important effect on the status
and the understanding of the plays.
On the one hand, such changes enable a reader of the plays to experience
something of the dramatic movement, how the actors move, how the scenes
unfold. On the other hand, the typographic layout used in the publication,
which imitates editions of French plays, gives a new status to Congreve’s
works, granting them a legitimacy that led their author to eliminate certain
licentious passages in order to make the plays more appropriate to the dig-
nity that their new printed form granted them.
We can derive many lessons from this pioneering study: in contrast to
an abstraction of the text, which reduces it to its semantic structure, it dem-
onstrates that the status and the interpretations of a work depend upon its
successive forms; in contrast to the death of the author, according to Roland
Barthes’s expression, it emphasizes the role that the author can play, along
with others (the publisher, the printer, the typesetters, the editors) in the
always collective process that gives texts their materiality; in contrast to an
absence of readers, it reminds us that the meaning given to a text is a his-
torical production, located at the crossroads of the abilities or expectations
of the readers and of the designs, both graphic and discursive, that organize
the objects being read.

38. See McKenzie, “Typography and Meaning: The Case of William Congreve,” Making
Meaning: “Printers of the Mind” and Other Essays, ed. Peter McDonald and Michael F. Suarez
(Amherst, Mass., 2002), pp. 198–236.
 Critical Inquiry / Autumn 2004 149
In stating that “new readers . . . make new texts, and that their new mean-
ings are a function of their new forms,” McKenzie leads us to consider the
relationship that connects the varied forms in which written works are pre-
sented, the definition of the audience of their potential readers, and the
meaning that those readers attribute to the texts they appropriate.39 Thus
we can understand how the status and the reading of theatrical works are
changed by the transformations in the way they are published.40 Multiple
are the effects of such transformations—for example, the theological con-
sequences, very strongly emphasized by the concerns of John Locke, of the
dividing of the Bible into verses; or further, the many instances of the play
between typographical layout and textual meanings Joyce introduced in the
original 1922 edition of Ulysses—and which have been lost in all subsequent
editions that have changed the layout of the text. The novel, at least from
the time of Richardson, is another example of the effects produced on the
text by the changes in its materiality, either because they alter its literality,
as is the case with abridged versions or anthologies of remarkable passages41
or because, without altering the work, they present it in forms that antici-
pate very diverse expectations and reading habits: the serial in a magazine
or publication in installments, both of which present the work over time,
the book intended for the circulating libraries that fragments the work into
several volumes, cheap editions, fiction anthologies, the “works” of a single
author, and so on. In each case the very form of the publication of the
“same” text demands quite different modalities of categorization, classifi-
cation, or of the understanding of the “same” work.
Such a position leads to two essential conclusions. On the one hand, it
is clear that, far from being mutually exclusive, a morphological analysis of
the materiality of texts and a social and cultural analysis of readers and read-
ing must necessarily be combined. It is in function of the assumed abilities
and expectations of the targeted readers that the author, publisher, or
printer decide on the forms that texts will be given. But those forms, in fact,
have their own dynamics that can, or may not, create a new audience (for
example, one that is wider and less elite) and encourage a new appropriation
of texts that previously circulated in other ways and among other readers.
An emphasis on the very materiality of texts enables one to challenge a too
narrow sociological approach that implicitly states that prior, crystallized

39. McKenzie, Bibliography and the Sociology of Texts, p. 20.

40. Compare David Scott Kastan, Shakespeare and the Book (Cambridge, 2001), and Margreta
De Grazia and Peter Stallybrass, “The Materiality of the Shakespearean Text,” Shakespeare
Quarterly 44 (Autumn 1993): 255–83.
41. See Leah Price, The Anthology and the Rise of the Novel: From Richardson to George Eliot
(Cambridge, 2000), pp. 13–42.
150 Roger Chartier / Languages, Books, and Reading
social divisions necessarily govern the unequal circulation of texts. On the
contrary, it points out how, thanks to the diversity of their forms, texts are
able to be revisited and reinterpreted by the different audiences that they
reach or create.
The great variety of forms and thus of meanings of the same text estab-
lishes the decisive role that libraries must play at present and in the future.
Granted, the electronic revolution has seemed to signal their end. The abil-
ity to transmit electronic texts anywhere in the world makes their universal
availability conceivable, if not immediately possible, while at the same time
the library is no longer believed to be the obligatory place for the conser-
vation and communication of written culture. Every reader, regardless of
where he or she reads, will be able to receive on the screen any text from
the library without walls, where ideally all the books of humanity will exist
in digital form.
The dream is no doubt seductive. But it should not mislead us. Indeed,
more than ever before one of the essential tasks of libraries is to gather,
protect, catalogue, and make accessible the physical objects that have trans-
mitted the written works of the past. If these works were exlusively com-
municated or, worse, if they were conserved only in an electronic form,
there would be a great risk of losing the intelligibility of a textual culture
that is inseparable from the objects that have transmitted them. To maintain
the communication of the texts in the various forms that they have, si-
multaneously or successively, received is essential so that we may under-
stand the practices and readings of their previous readers. The electronic
conversion of old texts and their hypertextual publication is no doubt a
precious innovation that makes, paradoxically, the diversity of the forms of
the “same” work more immediately obvious than does the printed word.42
Nevertheless, it cannot be considered comparable to the intelligibility that
comes from the analysis of the very objects that previous readers have held
in their hands. This conclusion is valid for the most canonical of works, but
is equally true for the more humble and recent products of print culture
that have been and still are the first victims of the illusion that texts are only
linguistic structures without material existence.
In 1978 Borges asserted: “People speak of the disappearance of the book;
I believe that is impossible.”43 He was not entirely right; in his country, for
two years, books were burned or destroyed and authors and publishers dis-
appeared, assassinated. But obviously his diagnosis expressed something

42. On the electronic publishing of the plays of Shakespeare, see Kastan, Shakespeare and the
Book, pp. 125–36; on that of the works of Nietzsche, see HyperNietzsche, ed. Paolo D’Iorio (Paris,
2000).
43. Borges, “El libro,” Borges oral (Madrid, 1998), pp. 21–22.
 Critical Inquiry / Autumn 2004 151
else: a confidence in the survival of the book in the face of new means of
communication such as films, television, and recordings. Can we continue
to express the same certainty today? The question is recurrent, but perhaps
it is badly put insofar as the reality of our present is characterized above all
by the appearance of a new technology and modality of inscription, of dis-
tribution, and of the appropriation of texts. The screens of the present are
not screens of images that are to be contrasted to the culture of the written
word. They are in fact screens of the written word. Granted, they convey
images, both fixed and moving, sounds, spoken words, and music, but
above all they transmit, multiply, perhaps in an uncontrollable excess, the
written culture.
And yet we do not know how this new medium offered to readers trans-
forms how they read. We know, for example, that reading the volumen in
antiquity implied a continuous reading, involving the entire body because
the reader had to hold the scroll with two hands, and this prevented the
reader from writing while he or she read. We know that the codex, first
handwritten and then printed, enabled practices previously [Link]
reader could leaf through the book, which was henceforth organized by
quires, leaves, and pages. The book could be paginated and indexed, which
enabled the reader to cite precisely and to easily find a given passage. Thus
the form of reading encouraged by the codex is discontinuous, but it is a
discontinuous reading in which the overall perception of the work, imposed
by the very materiality of the book, is always present.44 How might we char-
acterize the reading of an electronic text?
We may advance two observations borrowed from Antonio Rodrı́guez
de las Heras,45 which distance us from our inherited habits or our sponta-
neous practices. First of all, we must not consider the screen as a page, but
as a three-dimensional space, possessing width, height, and depth, as if texts
arrived on the surface of the screen from deep within the monitor. Con-
sequently, in digital space, it is not an object that is folded, as in the case of
the printed page, but the text itself. Reading therefore consists of unfolding
this moving and infinite textuality. Such a reading brings ephemeral, mul-
tiple, and unique textual units onto the screen, units that are created fol-
lowing the will of the reader, and they are in no respect pages set down once
and for all. The image that has become so familiar, that of surfing the web,
clearly indicates the characteristics of a new way of reading: segmented,
fragmented, discontinuous. If such reading is suited to encyclopedic texts,

44. See Stallybrass, “Books and Scrolls: Navigating the Bible,” in Books and Readers in Early
Modern England: Material Studies, ed. Jennifer Andersen and Elizabeth Sauer (Philadelphia,
2002), pp. 42–79.
45. See Antonio Rodrı́guez de las Heras, Navegar por la información (Madrid, 1991), pp. 81–164.
152 Roger Chartier / Languages, Books, and Reading
whose fragmented structure corresponds to that type of reading, it is dis-
turbed or disoriented by genres the appreciation of which implies famil-
iarity with the work in its entirety and a perception of the text as an original
and coherent creation. The success of electronic encyclopedias, the Ency-
clopaedia Britannica or the Encyclopedia Universalis, for example, as well as
the disappointments of the pioneers in electronic publishing of mono-
graphs or novels, clearly attest to the connections that exist between certain
ways of reading and certain literary genres and also to the greater or lesser
ability of the electronic text to satisfy or to transform these inherited prac-
tices. One of the great questions of the future is whether or not digital tex-
tuality will be able to overcome the tendency toward fragmentation that
characterizes both the structure of texts and the modes of reading that it
proposes.
Another challenge relates to the discordance, particularly strong in the
case of the youngest generation of readers who (at least among those with
sufficient means) have entered into the written culture in front of a com-
puter screen, between, on the one hand, the categories that have been es-
tablished throughout the centuries to define works by their coherence and
totality and, on the other, a practice that very immediately and very spon-
taneously fragments all types of texts. The potential consequences of such
a discrepancy are not insignificant because they will lead either to the in-
troduction of conceptual and technical devices into digital textuality that
would be capable of perpetuating the classic criteria for identifying written
works or to the abandonment of those very criteria in favor of a new way
of perceiving and of conceiving of written texts considered as discourse that
the reader feels quite free to cut up and reconstruct.
Will electronic textuality be a new and horrifying book of sand, whose
number of pages was infinite, which no one could read, and which had to
be buried in the storerooms of the Argentine National Library on Mexico
Street?46 Or, with the promise it offers, will it lead to an enrichment of the
dialogue that each book undertakes with its reader?47 The answer is uncer-
tain, and no one knows it. But every day, as readers, without necessarily
knowing so, we are fashioning an answer.

46. See Borges, “El libro de arena,” El libro de arena, pp. 130–37; trans. under the title “The Book
of Sand,” The Book of Sand, pp. 117–22.
47. See Borges, “Nota sobre (hacia) Bernard Shaw,” Otras inquisiciones, pp. 157–60; trans. under
the title “For Bernard Shaw,” Other Inquisitions, pp. 163–66.