Roberto Casati

The so-called "virtual world" is often described with the help of metaphors derived from ordinary discourse on perception and action. This should not be surprising, since virtual objects were partly conceived on the basis of these metaphors. Yet, it is not a given that these metaphors are appropriate; one might need to begin using different concepts and eventually to invent new ones, more appropriate to the phenomena they describe. It might even happen, as I shall show, that these new concepts will themselves be used in situations described by the discourses on perception, action and social behaviour in a way that so far seems perfectly natural, but that, in turn, might reveal itself to be entirely inadequate. One could use these new metaphors, born as they are of new practices and new usages, for the reinterpretation of the non virtual world. The subject of this piece is the metaphysics of the book. I shall look at the way in which the Web frees it from our inadequate conception of it. This emancipation, strangely, seems to involve an economic liberation.

Let us look at the "recycle bin" icon on the computer desktop. What this icon represents is not a recycle bin. In actual fact, it is not even an icon: it only represents the bin indirectly. Instead, this is the image of a sensitive field which activates a particular object. The activated object itself obviously is not a bin, and, if one looks carefully, it does not look much like one.

A recycle bin contains documents in the form of bits of paper. There is no order in there. The first bit to land in the bin ends up more or less at the bottom. A bin may contain things of a completely different kind than a document, and one never even thinks of reorganizing it.

The virtual "bin" stores data. It behaves intelligently, keeps a record of what has been thrown away, has an unlimited capacity for content, allows for the limited (re)use of the objects it contains, and so on. One can reorganize the bin.

Actually, one may begin to think that this is not at all a recycle bin, but a proper archive. Or a limbo. And the metaphor of a limbo, with all its redemptive ramifications, is very different from the metaphor of a bin. In a philosophical investigation, one can have rather discordant intuitions even about rather common objects, and even about rather basic concepts. With slightly more complex objects (and with much more complex objects, such as books) one is in great difficulty as soon as one tries to describe them with the precision that is necessary in certain contexts such as the legal and economic ones. The philosopher often does nothing else than weigh up these intuitions. What is a book? An ambiguity is inherent in this question:

My book weighs a kilo.

My book is inspired by yours.

Is the book that weighs a kilo the same thing as the book that is inspired by yours?

A. It is the same thing. There is only one book on the shelf.

B. It is not the same thing. I can destroy the book that weighs a kilo without destroying the book that is inspired by yours. The book that is inspired by yours could remain in my mind; it is a mental-book, a text. It could be printed on lighter paper and give birth to an object-book that weighs half a kilo.

This might look like an exercise in metaphysics. But it so happens that the issue of understanding what kind of thing a book is has given food for thought to the experts of on line publication, who have tried to define the ontology of the numerical heir of the book, the eBook. In 2000, the Open eBook Forum (OEBF, a consortium which includes editors such as McGraw Hill and Random House, producers of software such as Adobe and Microsoft, computer constructers such as IBM and new electronic publishing houses like iBooks) published a programmatic document open to the discussion on the ontology of the eBook.

In the preceding paragraph my use of the word "eBook" was ambiguous; this intrinsic ambiguity points to the nature of the problem in which the OEBF is interested. It is an ontological ambiguity. An "eBook" can signify:

The machine that receives the text,

and the text sent onto the machine.

And of course the same ambiguity holds good for "book". In the case of the eBook, the ambiguity is more vivid still. The eBook-machine can receive an infinity of eBook-books. The uncertainty with regard to the two notions prevails in the OEBF document, and this might pose difficulties in the project of establishing an ontology, instead of helping define a standard for it. We shall see that this indecision does necessarily need to be resolved.

When I say that I have read a book or remember it by heart, I am talking about the immaterial content. But if I say that I burned it, it is the physical support that I am referring to. If, on the other hand, I say that I have sold the book, I leave open the two possibilities. Now, we can take the question about the nature of the book further precisely by asking ourselves what place it occupies in human transactions, and especially in commercial transactions. What is it that is one selling when one sells a book? In an article published in 1785, Immanuel Kant discussed the issue of the copyright. His idea is simple. Some ways of copying a text are a form of piracy, while others are not.

My publisher prints many copies of my book and sells them, and he is not a pirate.

My neighbour prints many copies of my book and sells them, and he is a pirate.

What is the difference? According to Kant, books generate a right to stop anyone else from copying them or reading them in public, and this right can be ceded to a publisher against payment. But where does this right come from? For Kant, it arises out of the very nature of the book, which is not a thing like other things. To produce a book is not like producing a chair - unless the chair is a signed object. Kant was concerned with finding a good justification for copyright because he noticed that remuneration for words and ideas was not self-evident. If words and ideas are commodities, they are strange ones indeed. Let us stay for a moment with this comparison between ideas (and the words that express them) and chairs.

1. If I build a chair and give it to you, the chair leaves with you. For this reason it is easy to sell and buy chairs: the transfer of the object can recapitulate the transfer of the rights I have with regard to the object. 2. When you buy my chair, you can do whatever you want with it. In particular, you can resell it.

1. If I sing a song to you, neither text nor music leave me when they reach your ears. For this reason it is difficult to understand how one can sell a song. 2. When I sell my book to you, you may give it to a friend or burn it, you can even sell it to a second-hand bookstore, but you cannot make copies of it and sell it in your turn. The invention of the copyright has turned the sale of an abstract product into that of a concrete product - a record, a printed volume, a photography. The trick here is that the transfer of rights is limited.

There is an assumption in this book-protecting mechanism that cultural contents are a commodity - albeit a commodity of a strange kind. We know, however, that this does not have to be the case. If no one were willing to pay for songs there would be no sense in trying to sell them or in inventing ways of selling them. There have always been cultural contents that were never sold: songs improvised during a walk, drawings given away, books printed at the writer's own cost, love letters, anecdotes told during a conversation. Why pay for songs and books? Maybe because they have an intrinsic value? No: if we were all to become illiterate, books would no longer have much worth. A putative value depends on external circumstances. How can one know the value of a cultural content? Again, a comparison will help.

"Concrete" products, such as chairs, are subject to laws of the market which allow one to value them in relation to other concrete products. If I compare the price of a chair to the price of a match, the resulting information concerns the relative value of the chair and of the match, within a specific context. If, for example, the chair costs ten pounds and the match costs a hundred thousand pounds, I might conclude that there is a penury of matches and an excess of chairs. Pricing systems function with regard to physical objects just as they do with regard to pools of information. (We know, however, that the system can be twisted, especially in situations where the market is not free. In a planned economy pricing systems do not tell us much, since the match can cost a hundred thousand pounds and the chair ten pounds even if there is an excess of matches and a penury of chairs.)

Cultural products have never been really subject to the free market. For the market of cultural products, distortion is the norm. In effect, the mode of their selection, the marketing which promotes them to the public, the presence of innumerable mediators between author and reader, not least the bookseller who chooses what to display, the pressure of institutions, whether academic, ecclesiastic (think of the imprimatur) or governmental (think of books by famous politicians published with the tax-payer's money and distributed in schools), the shortage or partiality of reviewers - all these elements prejudice any attempt at obtaining reliable information about the value of cultural products from the observation of their commercial life. For example, academic publishers tend to request subsidies for publication, or the guarantee that the author's students will "adopt" the book; they may also ask the author to buy a certain number of copies. Their publishing guidelines are dictated by the traffic of favours between members of reading committees, or more simply, by cultural fashions. Cultural contents are not like chairs because their economy has no connection to the free market, and it cannot reflect their value. For a counter-example, think of your bookseller's reaction if you were one day to return a book to him, demanding reimbursement because its contents had disappointed you.

At this point, we can bring in the Web, and turn to a series of examples that, I hope, should help explain what I mean when I suggest we redefine objects and functions belonging to the pre-Web world according to what we can learn from the Web-world. Let us take the case of a producer of cultural contents, a research scientist.

I could want to publish a text with a publisher, but I would have to wait for months, await the scrutiny of a reading committee which might possibly be hostile to my research, for the most varied reasons. Moreover, I can be sure that the text will be read only by those who have been charged for access to the book, which will eventually be sold out and out of print.

When I publish a specialized research article on my Web site, I hope for a non-restricted access to it, indeed for the widest possible access, free of charge. I wish for my article to be read and for a response on the part of the readers.

Given such a choice, I should not hesitate to publish on the Web. Authors of scientific papers are tired of private and institutional filters to their work, and so they inevitably tend to publish straight on the Web. Here, on the Web, it is possible to assess the work in a real sense; and indeed, a constant assessment is what takes place, not by private or institutional mediators but by consumers, who effect it in a similar, but also significantly different way from the price system. I would like to propose three models for the assessment of on-line contents.

First model: visits (salon.com)

Consumers judge "concrete" products by buying them.

On the Web, cultural contents are voted for by visits to the pages most liked. The number of hits gives the measure of the extent to which the page is appreciated.

There have been cases in which on-line magazines (such as salon.com) fired those journalists whose articles did not get enough hits. What is interesting about the on-line world is that it is possible to obtain visitor figures that are more precise than those of any other media. Television channels, which use highly sophisticated recording systems, could not even hope to achieve the kind of precision yielded by visitor counters. These are capable of identifying individual computers with extremely simple programs that are available to anyone who builds a Web-site. This precision is the on-line writer's bogeyman. Until now, his or her article was judged only by editors; now, its readers give their direct vote.

Two sorts of problems emerge here. One, is it desirable to entrust assessment directly to readers? And two, how reliable is an assessment based on visitor figures?

Is it desirable to entrust assessment directly to readers?

There is always someone deciding what is good and what is bad to read; editorial teams make choices. The test case is that of an excellent paper, with excellent analyses, but read by no one. Should one keep it alive only in order to defend the quality of the press? To want to do so can lead to cultural paternalism and to the invention of ideal readers who should be interested in certain things. The point is that one does not really know what 'excellent' means; nor does one know what are the criteria according to which ideal readers are defined. Counting visits to a page obviates this situation: visits are the currency which allows demand to be measured in the realm of cultural products, where until now such quantitative evaluation has been traditionally resisted and therefore hard to apply. There is even the new possibility of a comparative evaluation, which might generate some surprises: an article signed by a famous writer could turn out to be less liked than the one written by a young unknown.

How reliable is an assessment based on visitor counts?

A reader can very well have visited a page without having read an article; he or she could have read it distractedly, or read it and not liked it at all. But even if what the visits measure exactly is unclear, one has to suppose that readers have an idea of what they want to read and that they visited a particular page for the sake of satisfying their own desire to read. This gives a rough estimate of a text's value. But is there a way of correcting the distortions of this rating system? Clearly, one cannot ask readers to give a summary of what they read in order to check that their judgement is informed rather than plucked out of thin air; but one could ask them to vote for pieces by expressing an opinion ("Did you like this? Yes/No"). The argument, however, might not impress a publisher. Indeed, it is possible that readers will loathe a piece, while the publisher's main goal is simply to make sure his paper is read. A special button could be set up to enable readers to answer a more precise question, such as "Were we right to publish this piece?" if not "What is the value of this piece in the context of the magazine's other contents?" In this case a piece on the dying art of lacing could be rehabilitated in the context of an issue which contains another - probably more visited - piece on the summer adventures of a celebrity performer. One could also imagine a system of reviews (which would themselves be reviewed -  and indeed the visitors have the possibility of rating the reviews on amazon.com). But there is a deeper problem. The vote-visitor mechanism is potentially irresponsible insofar as there is nothing at stake for readers when they vote-visit. In a future (that many hope, and others fear is not far off), where there will be a way of charging for on-line contents, a subscription might simply amount to the purchase by the reader of a power to vote for or against pieces. Journalists would then be paid directly by the readers: here the readers' action would be responsible, since it would have a price.

There is, however, a second way of obtaining information about a content's quality on the basis of visit counters.

Second model: links (google.com)

Consumers judge "concrete" products by buying them.

On the Web, cultural contents are voted for. How? By the creation of a link to the page one has liked.

Let us return to the case of the academic researcher who has to choose between publishing his own text on line or sending it to a journal. Journal reading committees are obsolete as soon as the texts published on the Web are judged by the readers, who create links to the pages they like. The system tends to generate a virtuous circle.

If I create a link to pages judged good by readers of my page, my page will be given a positive assessment by these readers and will in turn receive many links.

If, instead, I publicize mediocre sites on my page, no one will vote in its favour.

Honesty and competence are recompensed. A page that receives positive votes acquires an authority it transmits to its own links. I am not calling here for a new way of analysing cultural phenomena or of evaluating cultural products: I am only describing the democratic reality of the Web, one which exists already. A search engine like Google exploits the information contained within the link structure. Links from site A to site B are interpreted as votes for site B. This seems reasonable: if one wants visitors to one's site A to know that the best site on football (or on gamma waves) is B, one will vote for B by creating a link to it. If site A contains many useful links, other sites, at some point, will vote for it, and the authoritativeness it so acquires will be passed on to its links. Google collects the results of such votes. Why should one trust (enough) in the result? For the same reason as that for which we trust (enough) price tags in a shop. Behind those Web pages, there are individuals who exert their judgement, just as behind the price tags there are consumers who acquire the products. We are all mini-experts. Google envisions the Web as one great system of votes. This system is similar, in the end, to the pricing system, which informs us as to the relative value of products. The main difference between publishing cultural contents on the Web rather than through other means (books or television, say) is that the Web contains a huge amount of immediately accessible information about the value of the published products (a bit like feedback).

SCENARIO: But this means the death of cultural contents transmitted outside the realm of the Web. One would immediately assume these to be valueless. If publishers do not take on the risk of making the texts of their authors available on the Web, free of cost and unabridged, they will end up in a marginal economic niche.

I would welcome a research program on the link system which could study its informational potential, its possible distortions and the corrections to these distortions. It seems to me that much too little attention has been dedicated so far to this aspect of the Web's structure.

Third model: the expert's opinion (about.com)

There is an alternative, of course: the expert opinion-filter. Various experts, including Umberto Eco, have been defending the expert's role as a guide through the mass of information on the Web. Eco has dealt in particular with the issue of access to the Web in schools, although his argument can be applied to many other types of cultural transmission. This argument seems solid: "There are no instruments available to teach how to select information. One does not know how to distinguish between serious and crazy sites. The site of a nazi criminal like Eichmann could end up becoming the same thing as that of Mother Theresa of Calcutta." Filters are useful: there is too much material on the Web. "Until now the role of the Church and of scientific institutions was to filter and reorganize knowledge and information. Although these intermediaries restrict my intellectual freedom, they ensure that the community has filtered what is essential. ... I insist on the need for a filter external to the Web, whether it applies to schools, books or newspapers" (Libération, 2000/07/01). Again, filters are useful: almost everything is available on the Web, but without a good guide, one gets lost.

A highly important distinction might not be noticed here. Eco speaks of two types of filters.

Filters of the first type - let us call them negative filters - forbid access to objectionable pages.

Filters of the second type - let us call them positive filters - select only pages considered commendable.

But there is no reason why one filter should do both things. Eco seems to suggest that the need (in the end quite understandable) for negative filters could help make the case for positive filters. A positive filter, according to Eco, is an expert, or an institution external to the Web. The expert compiles -  for example - a page that contains links to other sites worthy of a visit. (A site like about.com uses exactly this principle.) The site gains credibility because the expert who compiles it is authoritative.

A problem arises here. How does one get to a credible site? I trust Eco, but does Eco know all the sites? It could be that a site unknown to Eco delivers more authoritative information than that delivered by a site Eco knows. And how do I get to Eco? Of course the state can set up educational portals. But if the prospect of negative filters should make us think about the possibility of forbidding access, it should not for that matter encourage us to select the contents that will be made available. Other scholars, like Omar Calabrese, have pronounced themselves in favour of officially certified sites. Does this not seem a strange idea? Just think of what the governmental approval of officially certified publishers would mean in the context of the book trade.

Google is not an expert - in fact it is ignorant and it acts blindly. But it comes close to the perfect librarian described by Musil in The Man Without Qualities: in order to do his job well, he should know next to nothing. Furthermore, one could simply argue that there are no experts on the information available on the Web.

But why prefer link-votes to experts? The link-vote system can be distorted just like the price system. In fact, within the link system everyone is considered a mini-expert. Why prefer so many mini-experts to one great expert-filter? Would it not be better to trust such an expert? No. If I may use a metaphor, expert-filters are to the vote-link system what planned economies are to the market. The fate awaiting expert-filters on the Web is that of planned economies. In five years all new knowledge will pass through the Web. And if they are of no use on the Web, where else do we need experts?

Here one may tie together the threads of the discussion on these three models. The Web and the link-economy make the nature of cultural contents explicit. On the Web, contents become what they are, abstract entities that are difficult to fit into the copyright system. This derives from the specific nature of judgements based on feedback.

Let us return to the book and to the nature of the book. We have seen that cultural contents are evaluated as soon as they become immaterial within the Web, and this throws light on something that the paper book tended to mask. But what can we expect, and how profound will be the changes affecting the channels of culture? The next example will tell us.

In 2000, Stephen King sold Riding the Bullet for two and a half dollars. You can purchase it (only on the Web), download it onto your computer or eBook reading device, and send a chapter to a friend by e-mail. Gripped from the very first lines on, that friend will then not hesitate to click on the button that will enable her to have the complete version, in turn paying two dollars and a half.

SCENARIO: The paper book is intoning its swan song. Unless it turns out that the paper book is irreplaceable.

Let us see. The boring objections to the possibility of transferring print production to eBooks usually appeal to a number of facts:

One can tear away a page from a book and send it to a friend. If a book falls it is not damaged. Books do not risk going dead in the middle of chapter three. The "functioning" of a book only depends on the reader, no maintenance is required. A book is ergonomically perfect, it is an object that never ages.

The eBook cannot be taken apart without there incurring significant damage to its functionality. Since it is high-tech, it is liable to various risks of breakage or malfunctioning. There is no agreement yet on what the standard for eBooks should be. The hardware of eBooks and the text formatting can be changed very quickly (think of how much computers and documents.doc have changed over the past ten years).

These details are interesting, but mere smoke when compared to the much more serious problems that are, in my view, too frequently overlooked. Just as with paper books, we can shift the discussion now to questions around the functional role of the eBook within the chain of social relations, for example to the ways in which the author (and his or her publisher) can be protected from copyright violations. But we can go further and try to redefine the social ties created by the circulation and electronic sale of cultural contents.

I can give a book as a present. No one ever gives as a present an introductory chapter of a book in order to get the friend to read the whole thing. I can sell the book to a second-hand shop. I can lend the book to whomever I want without depriving myself of the pleasure of reading a thousand other books in my library.

I would find it terribly rude for a friend to send me by e-mail the first chapter of a text by King or whoever else along with a button to click to acquire it. I cannot lend an eBook without depriving myself of the hardware. The idea of a loan is hard to carry out in the world of eBooks. And the idea of a book that I cannot lend (or resell) after reading it does not make me feel much sympathy for the author.

The idea that the author defends himself with software seems to me to signify that he thinks of me as a potential pirate - not exactly the image of the reader I thought he had in mind when writing his text.

One can see that, once broadened in this way, the discussion on the nature of eBooks forces us to think again about their eventual use. The field has been electrified by the sale of a few bestsellers, but it still has no precise place in the world of transmission of contents. Despite its impact on the media, electronic content that is sold rather than free has had difficulty in establishing itself. There might be an issue of cost here. Contentville.com sells for twelve dollars the electronic version of a novel which costs seventeen dollars in its bound version. The saving this represents does not seem very significant. It is sometimes thought that the difficulty of selling electronic content is due to the lack of appropriate gadgets on which to read it (it is true that the eBook reading devices on sale nowadays are a bit primitive), but this does not seem quite accurate. One American family out of three has access to the Internet and can download a book onto the house computer without having to acquire a new gadget.

To summarize:

Tested social norms and rules have crystallized around the book over the centuries and have protected it. This is not an expression of nostalgia but a fact, inherent in the function of the book, which is to ensure the circulation of ideas at a low cost and in a format whose advantages include not only the ease with which it is consulted physically but also the possibility of transmitting it to others or copying it, of using it repeatedly, of giving it away: the book is an object of exchange and communication. One does not yet know how to replace these social practices surrounding the book. The eBook is actually a hybrid product, somewhere in between book and portable computer; it was born when someone looked at a book and wondered: 'how can I make this book electronic'? But one looked into the past and not at the nature of new technologies, which create machines that are completely integrated and capable of treating information of all kinds. It is necessary to understand this and to invent something new; otherwise the eBook will be a gadget amongst a thousand others. Paper books, then, will resist as long as it they serve certain functions: culture is a highly complex phenomenon, connected not only to means of transmission but also to social practices. And to try to tie electronic content down to the book metaphor means not to take advantage of the various possibilities it opens up.

Let us accept the double nature of the book. Numerical format frees the content, but the paper book is a perfect object. How can one resolve this antinomy? It is not so certain that one has to resolve it. The market could stabilize with a dual object.

Authors will make their unabridged texts available on the Web, and for free. Will you read the book on line? Perhaps yes, perhaps no.

But books still make great presents. By clicking on a button, and by paying two and a half dollars at the on-demand printer-distributor, your friend will receive a much appreciated paper-bound volume.

SCENARIO

Try this thought experiment and imagine this scene. Millions of people publish books each year at their own cost (fattening up a particular category of pay-for-yourself publishers). Millions of people simply want to publish, or to make public their own cultural products. Nowadays they can do exactly that on the Web, and for free; hundreds of million of other people can read these books for free, without paying two dollars and a half. Of course, Stephen King is Stephen King, and an amateur scribbler is an amateur scribbler. But big numbers are big numbers. The choice between two and a half dollars for the super-famous Stephen King and zero dollars for a beginning author (who might have been drawn attention to by an aficionado, maybe in a small, but - thanks to a links page - accessible market niche), repeated millions of times in one day, simply becomes a choice between two and a half dollars and zero dollars. Free is free, especially if what I receive is a screen page and not a volume endowed with weight.

Another problem:

Payable contents should be protected

Free contents cannot be pirated.

If the market of cultural products tilts towards electronic contents, it will have to literally take account of the fact that, in the electronic world, free contents do not have competitors. Which means that contents must be free if they are not to disappear. Which means that paying contents disappear, and with them, publishers, agents and authors who live off royalties. Moreover, the figure of the author living off his royalties is a recent phenomenon and there is no reason to believe that it is an everlasting institution. This scenario can give rise to others. In order to avert the disappearance of cultural commerce, one could forbid the publication of free content by buying all providers; or impose, monopolistically, the purchase of a hundred books along with the purchase of a computer, as does a certain kind of software.

Copyright? It will continue to exist, in a "lighter" version, because an author may well be glad that people are reading his or her book for free, but not that someone else is profiting from it on the author's back. (This view of authorial behaviour may seem optimistic, but the alternative would be that everyone believes they are Stephen King, capable of making a fortune with their books, which is completely unrealistic.)

Again, the Web makes the nature of cultural contents explicit. Here contents become what they are, abstract entities, difficult to frame within the notion of copyright.

A reader's ethic

I now would like to look at the reader's perspective and to tell a story with a political slant which demonstrates just how much the old conception of books will have to be revised in the light of a profound reflection on cultural contents.

In 2000, three hundred francophone writers signed a petition addressed by a number of publishing houses and corporations to the French ministry of culture. The signatories were well-known French figures - Bernard-Henri Levy, Tahar Ben Jelloun, André Comte-Sponville, Jean Ziegler, and others - whose cause was rather odd: it was to attack public libraries. Libraries, they said, were acting as pirates by allowing their readers to read books for free. A reader who borrowed a book from a library did not buy it and so was reading at the expense of other people, without rewarding the author for writing the work and the publisher for taking the risk of publishing it. The petition called for the State to find a way of making sure a fee was paid to authors and their publishers every time a book found its way off the library shelf.

There are clear figures here. 300 million books are bought in French bookshops every year; 150 million are lent by libraries. Divide this by the six and a half million public library members and you get an average of twenty three books a year per user (or one book every two weeks). Now request from each user ca. € 15 per year, and divide the proceeds, leaving 30% to the authors and the remaining amount to the publishers.

There are objections to this reasoning. The figures are approximate: how many of the 150 million books lent every year generate royalties for the author? Authors would surely lose less money if publishers settled their royalties with monthly rather than annual payments. And how can one count all the times a library book has been signed out? In order to simplify what is a technically complicated exaction, one could, it has been suggested, remunerate all authors whose books can be found on library shelves, without taking account of loan frequency. This suggestion, however, seems to invalidate one of the very points of the petition. Why should an author read by no one receive as much as an author whose book is taken out often? The injustice suffered by the second author cannot be put right, and it is hard to understand what justice is being done by the first. One is overlooking, too, the fact that having a library at one's disposal allows for the correction of another injustice, this one committed against the reader-client. Why should one buy books blindfolded? Libraries are the only way we have of reading a book, finding it unsatisfactory, and returning it, at no cost.

But beyond the economic aspects of the issue, lies a moral problem which bears discussion. The moral addressees of the petition signed by the three hundred authors are not libraries or ministers of culture, but readers. This represents the unilateral breach of a subtle, almost invisible trust which connects authors and readers in the republic of letters. In fact, in their view, all readers who enter a library become criminals. And all loans end up as criminal acts - including the simple loan of a book to a friend.

If I consider legally liable all those who download bits of my music from the Web at no charge, I would logically have to call the police each time someone played that piece to his friends, or when two people exchanged a CD (without copying it). But of course this is absurd. It would certainly be unacceptable for a pirate record company to gain income from contents without having asked the author permission to do so. But there is a big difference between pirated contents, exchanged contents and free contents, and one has to make sure that regulations aimed at stopping the circulation of pirate contents do not harm the other two.

When social pacts are broken, so do the behaviours associated with them. Some authors have asked libraries to make clear that their books were not to be loaned out. I cannot imagine anything more strange for an author - a red tag telling the reader: do not read me.

In conclusion: the liberated book and its dual nature

The Web allows us to understand what are contents, it has liberated them economically and metaphysically. It makes explicit their criteria of identity precisely in virtue of the character of the circulation and sale of contents in electronic form. This holds for all contents, and not only those one finds on the Web, because the most profound feature of the nature of contents is their amenability to transmission.

There is a wide grey area, however, and one will have to establish strict norms for the regulation of competition, necessary to stop a concentration of providers. One has to prevent the purchase of many providers by one giant publisher which would get hold of its clients' Web pages and forbid them to publish contents free of charge. This would be an encroachment on the right to free speech. The Internet allows users to go straight to the contents of interest to them; if there is a risk here of creating a "culture on demand", the risk is minimal, given the huge amount of data one has access to. The user's autonomy is what enables one to dodge the filtering of information, which in theory should only improve the quality of our knowledge but which actually is the retrograde defense of a caste of intellectuals who try to impose themselves as the only possible point of access to knowledge.

In this text, I have argued that no problem will arise out of the book's transformation, because the two conceptual poles made out of the physical book, on the one hand, and of the immaterial book, on the other (and its evanescent digital incarnation), can coexist perfectly well within a dual product. We will pay for the physical aspects of the book, but we will also make sure that its immaterial aspects circulate.

Roberto Casati
CNRS-Institut Nicod
Paris

Translated by Noga Arikha.

Copyright © 2001 by Bibliothèque publique d'information - Centre Pompidou