The Transformations of the Book
Books & Ideas: I’d like to talk to you about the way in which the book as a physical object is metamorphosing nowadays under the influence of Internet technologies (e-books, print on demand etc.). Can you go over some of the changes the book has undergone since the invention of the codex ?
Roger Chartier: The first problem is: what is a book? This is a question Kant asked in the second part of the Groundwork of the Metaphysics of Morals, and he defined quite clearly what a book is. On the one hand, it’s an object produced by any kind of manufacturing process – manuscript copy, printing or even electronic production – and which belongs to whoever acquires it. At the same time, a book is also a work, a discourse. Kant says it’s a discourse that is addressed to a readership and that always remains the property of whoever composed it and can only be distributed under the mandate he gives to a bookseller or a publisher to put it in the realm of public circulation.
All the problems involved in thinking about it derive from this complex relationship between the book as a material object and the book as an intellectual or aesthetic work, because to this day that relationship has always been established between these two categories, between these two definitions: on the one hand, works that have a certain logic, a coherence, a completeness and, on the other, the material forms of their inscription, which in Antiquity and up to the 1st century AD were confined to the scroll. In this case, the work was very often disseminated by means of several different objects. Beginning with the invention of the codex (that is to say the book as we still know it, with notebooks, leaves and pages), the opposite situation arose: one and the same codex could – and this was actually the rule – contain several books in the sense of different works.
What’s new about the present day and age is that this nexus between types of object and types of discourse has been severed, since there is a textual continuity to be read on a screen, and the material inscription on this unlimited surface no longer corresponds to a specific type of object (the scrolls of Antiquity, manuscript codices or the printed book after Gutenberg). This now gives rise to disputes that may even have legal ramifications regarding copyright and ownership issues. How does one maintain categories of ownership in a work within a technology that no longer delimits the work the way an object like the ancient scroll or the codex used to do? This may also have consequences for the recognition of the status of scholarly authority. In the age of the codex, a hierarchy of objects could more or less indicate a hierarchy in terms of validity of discourse. There was an immediately perceivable difference between an encyclopedia, a book, a newspaper, a journal, a note sheet, a letter and so on, which were provided in material form to be read, regarded, physically handled, and which corresponded to registers of discourse that fell within this plurality of forms.
But today the sole object – and there’s one right here on this desk – is the computer, which holds every type of discourse, whatever it may be, and which renders the continuity between reading and writing absolutely immediate. So we can enter into contemporary considerations, while returning to this duality that we often forget. This brings up the problem of the electronic book, with a rematerialization in an order of objects like the e-book or the laptop computer, which are unique objects for every sort of text. So from here on, the relationship takes on new terms.
Books & Ideas: Michel de Certeau draws a distinction between the written trace, which is fixed and enduring, and reading, which is of a transient nature . But texts on the Internet are constantly changing. To stretch the point a bit, one might say the Internet is a world of “plagiarized plagiarizers.”  Is this a major upheaval in your opinion, or would you say that over the course of history, and especially in the 17th century, the text has never been a stable form?
Roger Chartier: Yes. In his distinction, Michel de Certeau refers to the roving reader who constructs meaning based on constraints while concomitantly constructing meaning based on freedoms, that is to say he “poaches.” If we do poach, it’s because there is a domain that is set, protected and off limits. De Certeau often compares writing to tilling the soil and reading to traveling (or poaching). This is indeed a view that has inspired works on the history of reading or the sociology and anthropology of reading, from the moment that reading was no longer enclosed in the text, but was the product of a dynamic, dialectical relationship between a reader, the scope of his expectations, skills and interests, and the text he is coming to grips with.
But this constructive distinction can also obscure two aspects. The first is that the poaching reader himself is rather strictly determined by collective determinations that are shared by interpretative or reading communities, so this creative freedom, this consumption that is production, has its own limits; it is socially differential. Secondly, as you say, this textual domain is a more mobile terrain than a plot of land to the extent that this mobility actually did exist for any number of reasons. The technological conditions of text reproduction, for example the manuscript copy (a practice that endured till the 18th and 19th centuries), are amenable to this textual mobility from one copy to another. Apart from highly sacred texts, which have to be respected to the letter, all texts are open to interpretation, additions and alterations. In the first period of printing, that is to say between the mid-15th and early 19th century, print runs were still quite limited, for many reasons, to between 1,000 and 1,500 copies. From then on, the success of a work was ensured by multiple republications. And each new edition was a reinterpretation of the text, whether in terms of the letter of the text, which is modifiable, or even in terms of the physical aspects of its presentation, which are another form of variation. Supposing not even a comma is changed in a given text, changes in forms of publication – fonts, the presence or absence of illustrations, how the text is divided up etc. – create a mobility in the possibilities of appropriation.
So we have compelling reasons to assert this textual mobility. There are other, intellectual or aesthetic reasons, too: up to Romanticism, stories belonged to everyone and texts were written based on formulae that were already there. This malleability of stories, this plurality of resources available for writing, creates another form of movement, which is impossible to confine to the letter of a text that remains forever stable. We might even add that copyright only serves to reinforce this fact. Which is paradoxical, of course, since copyright recognizes that a work is always identical to itself. What then does copyright protect? In the 18th and 19th centuries, it protected every possible form of printed publication of a text and, in our day, every possible form of publication period, whether it be a motion picture adaptation, a television program or various editions. So we have a legal principle of unity that covers the indefinite plurality of successive or simultaneous states of the work.
I think we need to resituate contemporary mobility, with the electronic text, that polyphonic textual palimpsest, in a long-term conception of past textual mobilities. What remains of the question is the fact that there are constant attempts to reduce that mobility in the electronic world. This is the precondition for products to be saleable – the “opus mechanicum,” as Kant would have called it – and the precondition for proper nouns to be recognizable both as authors and as beneficiaries of their creative work. Hence the profound contradiction that Robert Darnton developed between this infinite mobility of electronic communication and the effort to enclose the electronic text in mental or intellectual categories, but also in material forms that fix it, define it, transforming it into a plot of land on which the reader might poach – but a plot with sufficiently stable boundaries, limits and contents. Herein lies the great challenge, which is to determine whether the electronic text ought to be subject to inherited concepts and therefore altered in its very materiality, that is permanently fixed with built-in safeguards, or whether, conversely, the potentialities of this anonymity, this multiplicity and endless mobility are going to dominate the uses of writing and reading. I think this is the focus of the contemporary discussion, uncertainties and vacillations.
Books & Ideas: To end this batch of questions about changes in the book as a physical object, I’d also like to ask you about changes in the place where this object is traditionally kept: namely the library. In its Google Books Library Project, Google has scanned the collections of 28 libraries, including those of Harvard, Stanford and Oxford. This project has its (critical) proponents like Darnton and its opponents like Jean-Noël Jeanneney. Do you think Google is going to give rise to a global library that is open to everyone?
Roger Chartier: Here again, underlying this project you’ll find myths or figures of Antiquity, specifically that of a library that contains every book in the world. That was the plan of the Ptolemys in Alexandria. Google has subscribed to this prospect of a library containing every existing book as well as the books that may yet be written. Technically and ideally speaking, there’s no reason to think that all the books that exist in one form or another couldn’t be digitized and included in a universal library.
But one of the primary limitations is that the Google project is handled by a capitalist enterprise. It is governed by an economic rationale, even if that is not readily apparent, and that rationale may also govern the advertisers and media of this gigantic corporation. Furthermore, this is a project which, even if it claims to be all-encompassing, clearly favors the English language. As an ex-governor of Texas once said, “If English was good enough for Jesus, it ought to be good enough for the children of Texas.” No doubt she’d read the Bible only in the King James translation and not the previous versions. The project is not presented along those lines, but still, given that the first five libraries chosen were Anglo-American, the bulk of the collections was inevitably in English.
So what are the possible responses? One suggestion was to reorganize national and European libraries in such a way as to come up with an alternative project. It was alternative in terms of language diversity and also because it was based more on public power and not on private enterprise. But presumably through these portions of universal libraries we could arrive at a universal library, even if it isn’t unified by a modern-day Ptolemy. And there’s no reason to think it couldn’t be accessible in electronic form.
The question that follows from there is not only that of languages and responsibility, but also whether this universal library, which wouldn’t need to be located anywhere insofar as anyone can access one book or another on his computer wherever he happens to be, won’t spell the demise of libraries as we knew them: a place where books are preserved, sorted and consultable. I think the answer is no, it won’t. The digitization process makes an even stronger case for retaining the traditional definition because we always come back to a fundamental point, which is, as Don McKenzie puts it, that forms affect meaning. The great danger in the digitization process is to suggest that a text is the same no matter what media it is carried on. As crucial as access to digitized texts may be, I find that this digitization nevertheless reinforces the importance of preserving our cultural heritage, the successive forms that texts have had for their successive readers. The task of preserving, cataloguing and consulting texts in the forms in which they were originally published becomes an absolutely fundamental requirement, which reinforces the importance of libraries in preserving our cultural heritage.
Any number of examples can be adduced in this regard. In the 19th century, the novel existed in multifarious physical forms: serialized in weekly or daily supplements in the papers, in the form of publication in installments, in the form of books for reading rooms, in the form of anthologies of a single or several authors, in the form of complete works and so on and so forth. Each form of publication involves different possibilities of appropriation, different waiting periods, a different temporal relation to the text in question. The need to reinforce this role in the preservation of written heritage is not only a good thing for scholars seeking to reconstruct the history of texts, but also for the relationship current-day societies have to their own past, that is to say to the successive forms that written culture has taken in the past.
This is the crux of the main debate over projects like Google’s, which were subsequently imitated by library consortiums. When they got wind of the Google Library Project, some librarians drew the conclusion that they’d be able to empty the stacks and assign the reading rooms to a different use. We also see this in the controversy raging in the United States over the destruction of 19th and 20th-century newspapers as soon as they were reproduced on a substitute medium, in this case microfilm; but the risk would be even greater with digitization. The libraries sold their collections, or they were destroyed in the course of the microfilming process. One American novelist, Nicholson Baker, wrote a book to denounce that policy, which was the policy of the Library of Congress and the British Library, and incidentally to try to save this written heritage himself: he actually put together a sort of archives of collections of American dailies from the 1850s to 1950.
What is Reading?
Books & Ideas: Ever since the invention of writing, reading practices have changed continually. We read aloud in family circles in the evening, for example, or alone in silence. Can you talk about the different forms of reading down through history?
Roger Chartier: There are two dimensions: morphological and chronological. We can pinpoint moments in history at which the preconditions for reading underwent massive changes. In a very long-term medieval development, more and more readers were able to read the way we do, that is in silence and with their eyes, although oralized reading was both a normal way of sharing text among the literate and one of the conditions for textual comprehension. The cause and consequence of progress in silent visual reading were a new form of textual inscription, specifically the introduction of separations between words, which did not previously exist in most Latin texts. That was one of the major revolutions in reading.
We have been able to talk about a “new revolution” in reading in the 18th century, but this expression is debated. The objects that were read diversified: this was the age of the widespread circulation of periodicals, of growing quantities of lampoons and pamphlets, of increased book production throughout Europe. On the other hand, reading moved away somewhat from the respectful, obedient, sacred practices that still marked it heavily to become a more casual, critical and mobile practice. In the 18th century – and contemporaries really felt it, by the way – there was like a mania for reading, a reading craze. Another crucial transition occurred in the 19th century, an age of mounting tension between the reading norms imposed by school and the rampant spread of reading in ever-broader social milieux. This 19th-century proliferation of writing could be seen in cities on the walls, the public notices, the posters, in the press, which changed in nature during that period, and from the second half of the century in book collecting by the working class.
So it is possible to identify the changes, some of which were related to the morphology of reading (silent or oral), others to the tension between imposed norms of “how to read properly” and the wide range of unfettered everyday practices. Historians have debated the validity of each of these putative turning-points and whether they can be qualified as “revolutions in reading.” On the other hand, the plurality you mentioned is not just a morphological and chronological plurality: in each of these societies (medieval, Enlightenment, 19th century) we see a differentiation in what we might call “interpretive communities” or “reading communities,” based on the same skills, the same expectations in respect of writing and the same reading conventions. There’s a well-known article by Michel de Certeau on mystical Spanish and French communities of the late 16th and early 17th century united by a rapport with the book, by specific reading practices, by a gradual drift away from prayers. We have also tried to discern what might characterize “working-class reading,” that is reading as practiced by the least literate social strata or those accustomed to narrower textual repertoires. So there are efforts to identify this plurality, which is directly rooted in social and cultural differences. I think the way to parse these diverse ways of reading would consist in fusing these chronological and morphological dimensions with an identification of sociocultural differentiations.
Books & Ideas: There’s a very funny and savvy book by Pierre Bayard, Comment parler des livres que l’on n’a pas lus? [i.e. “How to talk about books you haven’t read”]. After all, it is often true that all we know of books is what the critics say or the filmmakers who adapt them. Would you say we do read books nowadays or that we only become acquainted with their by-products?
Roger Chartier: The question is whether there’s been any innovation around the idea that you can get to know books you haven’t read, in other words through different forms of mediation. This mediated knowledge has been reinforced by the development of places of mediation. But similar forms existed in the past as well. In this regard Don Quixote is probably the first text to usher in the modernity of reading, first of all because its main theme is the projection of the text onto the world and the presence of the world incorporated into the text, but also because a great many readers very soon got to know Don Quixote without having read it. References to Don Quixote circulated quite early on by dint of impersonations of the characters at courtly revels and at carnivals, the spread of iconographic representations of scenes from the novel, its adaptation for theatrical performances, though also the piecemeal reading of passages from the novel, which was made possible by its division into chapters. It cannot be inferred from all those references, however, that readers actually read the whole book, much less both parts when the second volume came out in 1615. So this represents the first matrix of these forms of access to a text through the mediation of either fragmentary readings or manifestations of the text outside the text. In my opinion this is a very important history: how characters and stories emerge from the page – on stage, at festivities, in discourse – to become realities that depend on and yet differ from the writing proper.
One might also consider the dominant technique in Humanism: that is the technique of commonplaces, that is to say the ability to reuse examples, maxims, models that serve in producing new discourse. This is a reading technique that dismembers texts and sometimes makes use of snippets cut out by others, since you can consult collections of commonplaces, where you’ll find rhetorical and stylistic resources for your own productions.
So this idea that one can know texts without having read them, that reading fragments is often an ersatz for reading the whole thing, is not new. What has no doubt amplified this phenomenon in modern-day society even more than the proliferation of printed reproductions of artworks is the multiplicity of audiovisual adaptations, first in the cinema and then on television, engendering a familiarity with literary works that people have never read. So it is the mode of impact that has changed.
Books & Ideas: In a recent article, Robert Darnton says it’s “important to get the feel of a book – the texture of its paper, the quality of its printing, the nature of its binding. […] Books also give off special smells.”  Allow me in closing to ask you a personal question: what is your particular way of loving books? How do you read?
Roger Chartier: “Le moi est haïssable [i.e. egoism is odious],” somebody [i.e. Pascal] once said somewhere. By the way, I think this question is a booby-trap if you bear in mind what Bourdieu says about the biographical illusion. This type of question presumes you will give an answer in which, even if unconsciously, you construct an image of yourself. The main thing, especially in the first part of his observation, is that Darnton is relaying his own work as an historian. In the 18th century, as a matter of fact, as he shows in copious correspondence, many buyers of books were keen on this physicality, the nature of the paper, the ink and so on. All these elements, which inform the nostalgia of those who believe the book is already dead, give pleasure to certain bibliophiles and readers. To my mind, they are not to be taken so much from their affective side, this world of printed pages that we’ve supposedly lost, but from their intellectual side: the forms of inscription of a given text delimit or prescribe its potential for appropriation. This starts with appropriations at the most economic level, since the retail price depends on these physical features. A paperback doesn’t cost as much as a hardback. Above and beyond the conditions of physical and economic appropriation, there are the conditions of the construction of meaning, which involve the choice of format, fonts, the division of the text, the use of illustrations and so on. So Darnton’s remark, which lies on the affective level of an intimate relationship to the object, can be converted into a tool of knowledge.
As to your second question, I think the only answer is the one we mentioned earlier. Nowadays everyone is developing this plurality of relations to the texts they read in accordance with their concerns and their pastimes, their activities and desires. From this point of view, we read intensively and extensively texts that deserve to be regarded as legitimate reading material, and we place the others outside of these categories. Sometimes we hear a diagnosis that says we’re reading less and less these days. That’s dead wrong: no society has ever read as much, or published as many books (even if the print runs are tending down). Never before has so much printed matter been available at kiosks and newsstands, and never before have people read as much – thanks to the omnipresence of screens.
So it’s completely wrong to claim that reading is on the wane. On the other hand, what’s at stake in this sort of observation is the fact that the ones asking and the ones fielding this question often don’t feel the same things are worth reading. Christian Baudelot has put out a book entitled Et pourtant, ils lisent [“And Yet They’re Reading”], in which he stresses the contrast between statements by teenagers, especially boys who don’t want by any means to be pegged as readers (because that carries connotations of “square” bookishness and a conventional mindset, a culture they reject), and their actual behavior: at school, they read; in front of the screen, they read; and those who say they never read actually read a wide range of written matter. We find the same type of analysis in historical studies using interviews of readers born in the early 20th century in working-class and rural milieux.
This points up the tensions between what people say about reading, which always refers to a norm of scholastic and cultural legitimacy, and actual practices, which are boundless, widespread and diverse, and which take up a wide range of printed and written matter over the course of a single day or a lifetime. Defining legitimacy, the connect between what we regard as reading proper and the infinite amount of low-quality practices, which are reading practices all the same, may well be the great challenge facing contemporary society. The multiplicity of widespread practices and appropriations of writing can be viewed as a telltale sign of the fault lines running through society and of the very different resources thanks to which individuals can know themselves better or know others better. This is not a matter of contending that all the texts we read are of equal value, but I don’t exempt myself from this tension between what we read for intellectual purposes or aesthetic pleasure and the countless low-quality texts we read over the course of a day in the press or on the web. This is an answer in which, it seems to me, the actual, individual case can help us think about practices of knowledge, which are what we have in common nowadays.
Interview by Ivan Jablonka
Transcription by Émilie Boutin