Umberto Eco
Gloria Origgi: When asked, in varying contexts, about the possibilities and risks of Internet, you have often evoked the problem of filtering information.
Umberto Eco: It is the fundamental problem of the Web. The whole of the history of culture has consisted in the establishment of filters. Culture transmits memory, but not all memory: it filters. It filters well, or badly, but if anything what has allowed us to interact socially is the fact that we have had all of these filters. Then of course the scientist or scholar can put in question these filters, but that is another matter. With the Web, everyone is in the situation of having to filter information that is so vast, and so unsustainable, that if it isn't filtered it cannot be absorbed. It is filtered unsystematically, so what is the primary metaphysical risk of this business? That we'll end up with a civilization in which every person has his own system of filters, in other words where every person creates his own encyclopaedia. Now a society with five billion concurrent encyclopaedias is a society in which there is no more communication. Moreover, the filters we resort to result from our having trusted what we call "the community of learning" that, throughout the centuries, through debate and discussion, gives the guarantee, if anything else, that the filtering is reasonable; so imagine what would be an individual filtering performed by anyone, for example by a fourteen-year-old boy. We could end up with competing encyclopaedias, some of them completely wild.
G. O.: But today we have authoritative filtering systems that belong to this very means of communication, that is, search engines.
U. E.: That's not a filtering system. There are already polemics on the fact that search engines "filter" only information that has been paid for. I don't believe in the possibility of automating the filter's function. The only solution is that there appear authorities, external or internal to the Web, that constantly monitor what is found. Here's an example: I recently searched for "Holy Grail" on the Internet. Thirty sites came up. I'm quite informed about the subject, so I could tell that one of these was philologically correct, two were correctly encyclopaedic and all the others were actually the work of delirious occultist crazies. I, so to speak, am an expert on the topic: but what about the poor sod who looks up the theme of the Grail for the first time? How does he manage to filter? He can be the victim of the first quack who built a site. But how can these monitoring groups take shape? How does one monitor the whole of the Web? Even if it had been monitored on Monday, it would be different already on Tuesday. There should be specialized monitoring groups, for example the International Society for Philosophy could continuously monitor all the philosophy sites, you know, a bit like the old lists of the Centro Cattolico Cinematografico in Italy that pronounced, outside the church, on whether a film was "excluded to those under age", or "for adults", or "suitable for all". Catholic parents who trusted the priest knew, if anything, that that film contained dangerous scenes for a child: they had a filter. Now if I trust the International Society for Philosophy, which tells me: "This site on Kant is rubbish", then I won't use that site. But I've already discussed how these filtering groups could emerge, and how they could
manifest themselves: if they do so within the Web, how does the ingenuous visitor know that this site is the monitoring, expert one? If, on the other hand, they do so outside the Web, maybe on a special bulletin, or monthly and so on, it would be printed matter and would be available to a tiny percentage of surfers. We haven't resolved these problems. If I had the answer I'd probably become a billionaire, but I don't.
G. O.: How does one make the difference between the imposition of a filtering authority and something similar to a new form of "censorship"?
U. E.: A filtering authority is not a "censor" but a consultant. Excuse me but if I go to my economic consultant and I ask him to tell me which shares to buy and which ones not to buy, he's no censor, he's a consultant who tells me that it would be advantageous for me to buy these shares and not these other ones because they've caused a lot of trouble. So, what, the whole of culture is censorship because a teacher who instructs her pupils that two and two does not equal five is censoring the ignorant child - but this job of filtering is an aspect of education. To use censorship is to stop the circulation of information, while to use a filter is to judge the information that is circulating. Those are two very different things: I can criticize the primary school history book and be simply acting as a critic; or I can have it seized by the police, and act as a censor.
G. O.: We are here at the seat of the Master's in Publishing that you direct in Bologna. Is a publisher also a "bottleneck" that filters information?
U. E.: A publisher is a filter, and I can be sure that Mondadori might also choose a bad novel, but not below a certain standard because it has a history behind it; this is all the more true if the publisher is Olschki, which prints only scholarly and philological books. The other problem with Internet is the spread of the samizdat. In the USSR, if one wanted to publish one distributed a samizdat, which circulated within a restricted group of people; now anyone who wants it can put his novel on line. This is a great act of freedom: at least it relieves publishing houses of a load of useless manuscripts, and it can allow an intelligent publisher to check what is on the Web and maybe even to discover a new talent. But in this case too, just put yourself in the shoes of a fourteen-year-old kid who goes online and reads all these samizdat - the function of orientation provided by criticism is gone. When I was in high school, I used to buy La Fiera Letteraria, a cultural magazine that existed then, where there was a section for people who'd sent their poems to a critic who analyzed them, saying "This one is good", "This one is bad", and so on. You can say what you will about this critic: but to me he transmitted the appreciation of discrimination. It could be that his taste is outmoded today, but he taught me that some verses were judged more beautiful for certain reasons, that some were junk, common old stuff - the point is, this filter, constructed by a magazine, gave me a taste for poetry.
G. O.: Who filters literary taste on the Web?
U. E.: That is a huge problem. It could revolutionize our literary taste. For example, it could establish an "anything goes" of taste: it isn't true that Homer is better than John Smith. No: everyone takes whatever he likes. That could be such a revolution in taste that we can't even imagine its outcome. From the point of view of our cultural tradition it would be extremely dangerous. But there is another argument: taste filters in literature have only ever interested 0,5 per cent of the population. If today 70 per cent of the population surfs the Web and considers good any poem or story it encounters, we can say that these people had been excluded from the enjoyment of the literary product and that they have finally come in contact with some form of literary expression. Again, this will certainly be a revolution. A revolution that could be mitigated, in the sense that someone who educates himself on the Internet devours anything, but the minute he goes to university or gets a job, inevitably encounters parameters and evaluates his own previous excesses - but all this is pure prophecy.
G. O.: About this "anything goes" you're talking about: the screen itself is a representation of it, it is a single means for everything, newspapers, publicity, texts that traditionally we call "books". What is it that still distinguishes a book from any other form of information?
U. E.: Above all, I think, the psychological mechanisms of attention. Our species has gotten used to a certain sort of attention, which involves turning pages, lingering with attention; the kind of reading effected on screen inevitably is different, faster, one scrolls with greater speed. Here's an example: there is still a huge difference between correcting proofs on a printed page and correcting the text on screen. There will always be more mistakes left on the screen than on the printed page, because there is a different way of fixing one's attention on a line. However, new generations are born with a mechanism for attention that is adapted to the screen. In the end, if we take a medieval manuscript we can't read it, while at the time of Gutenberg people were complaining that they couldn't read a printed page: they read the manuscript more easily! So mechanisms can change, but even while they do these various types of attention, I think, will continue to exist.
G. O.: So the fact that something is a newspaper, as opposed to a portal as opposed to a book, even if the means remains the same, depends on how much attention the user puts into it?
U. E.: Well certainly if I want to know the weather in Amsterdam the computer screen gives me more than enough information, and it is even maybe more clear and convincing than what I get from the newspaper, where I have to look for it on the last page. If, on the other hand, I want to read the Divine Comedy... Here's an example (although I'm still a representative of the pre-computer generation): I have the whole of the Divine Comedy in the computer. If I have to write an essay in which I quote three tercets of the Divine Comedy, I get them from the computer and so avoid the trouble of copying them. If I have to consult the Divine Comedy it is easier for me to take down the volume, since I remember the page layout and so on, but I can't rule out that a next generation will find it more poetic to read the Divine Comedy on the screen. This is their business; for me it would be difficult: I use it on screen but I return to the book. But I would like to make this point: it has never happened in the history of humanity that the introduction of a technological means killed off all the practices of the previous means. Even the wheel did not entirely supplant the sledge; photography did not destroy painting, if anything it gave it a new direction; and all the statistics tell us that where people watch a lot of television there tends to be a higher rate of newspaper use: it isn't true that people who watch too much television don't read the papers. I believe that the increase in information, even if it is on screen, will not have an effect on the use of books; on the contrary it will increase it. The proof is that in the decade of the Internet there is a proliferation of mega-bookshops in all cities, visited mostly by youngsters. The eBook is another issue. For the moment it has not taken off and so we have to wait and see what happens. I am not at all opposed to the fact that an eBook could perfectly well replace the book, even though, by habit, I prefer the book - thus as far as reading mechanisms are concerned. Then there are the emotional, aesthetic, tactile dimensions - so you can leaf through the beautiful paper of a book, although here too it could be that in three generations things might well be different.
G. O.: If you had to describe to the students in your Master's in publishing class how you envision a publishing house, let's say in ten years, what disappears? Does paper disappear physically? What does a publisher have on his desk? Can we imagine a scenario where authors negotiate directly with eBook manufacturers and stop using publishers?
U. E.: I understand the question but I refuse to give you an answer. Just some five or six years ago, when we began to make the CD-Rom of the Encyclomedia, I brought those who were involved (it was an Olivetti initiative) to an important Italian publisher. We showed them this project, asking whether they wanted to take part, and they said first that there was no future for CD-Roms, and second, that the winning formula would be not a CD but something or other that Philips had brought out and they wanted to bet on that. Nowadays this important publisher makes CD-Roms by the bucket. So if five years ago an important publisher, whose interests are, let us say, not cultural but at least commercial, was incapable of foreseeing the development of a new technology, imagine wether we're in a position to make exact predictions about the future. Of course we're teaching a Master's in print and multimedia publishing because publishing houses will increasingly become producers of printed matter as well as of multimedia. Which one of the two will prevail, I cannot tell.
G. O.: Now, about the problem of copyright. There are "types of intellectual property" for which any paid access, given the quantity of information around, seems intolerable.
U. E.: Listen, the issue of copyright is, and I'm being generous, four centuries old. It began when the privilège du Roi started appearing in seventeenth-century books, which was a declaration that, in a sense, defended the rights of the individual book. That did not stop the publication during the whole of the seventeenth and eighteenth centuries of books that appeared maybe in Adelphia or under the names of non-existent cities, and which copied without any worries a book published in Paris or in Amsterdam - they would make another edition of it. Curse and delight for collectors, since at times it is very difficult to differentiate a real first edition from a pirated one: you have to look at the change in a bookmark. In sum: the defense of intellectual property is pretty recent. Just as it didn't exist in the past, it could well not exist in the future, or take on other forms. I am an author, I receive money from copyright, but when I learn that they made a pirate edition of my books in Cuba or even in Germany or China, while my publisher goes mad with rage, I'm not so unhappy, actually I'm very glad that my work circulates. There might be problems if no longer receive even a cent for a book I've written. But as you see, such problems are easily overcome. For example, just as writers used to be sponsored by patrons, so today they could be sponsored by publicity. That would entail a great loss of freedom, because copyright has been a factor of freedom for the writer, who no longer had to give his due to the patron but to the general public, which bought him or did not buy him. So to lose copyright could bring about a dangerous loss of freedom, because if you're no longer paid in royalties you have to be paid by Berlusconi or by the Vatican or by the Social Democrats or by Coca-Cola, and this certainly is a big problem. Here too the solutions should be legal solutions of some sort. I just want to emphasize the point that, while the system could change, nevertheless the system of protection of intellectual property has signified an increase in democracy and freedom - to lose it might put us in real danger.
G. O.: Certainly the protection of intellectual freedom differs according to the type of text. In his paper for text-e, Stevan Harnad argues that there is a fundamental difference between scientific publication, for which the author seeks maximal diffusion but no financial remuneration in terms of royalties, and other types of publication for which access remains and will remain paid for.
U. E.: A scholarly text should be financed either by the university or by research funds, so it should be public. Even the most successful scholarly book doesn't enrich the author. The problem is really one of so-called "literary" property.
G. O.: The two kinds should thus be separate?
U. E.: Not only that. My opinion is that even if the latest Pulitzer prize novel were immediately to become free on the Internet, a part of the public, out of a desire to collect or to read on the train, would continue to favour the print format; so there would be a balancing out. That is to say, I might - well, personally I would never do it - or someone might read four hundred pages on screen, something that seems pretty difficult to me, because it would mean staying seated and damaging the spine for hours on end. Daniele Barbieri says that changes could occur in the form of large viewers in each room with which you could read a novel even while lying in the bathtub. But what would happen then? If I don't like the novel I delete it, and if I love it, what then? If I print it I end up with an object I can't use, which falls all over the floor and so I go out and buy a print copy. That might mean that authors with a large circulation would lose something in terms of royalties, but here too, what could happen is that the print copy will cost € 20 instead of € 10 and that what the author loses through online reading is recuperated through the public of enthusiasts. But you know, this is what is happening today anyway: the large American publisher produces a hardback and within two or three months it produces the trade version. Then within another two or three months the paperback appears, while all along the hardback continues to be published because while the young, or whoever wants to read the book and throw it away after, will buy the paperback, there is still a large part of the public that prefers a good, readable edition and is ready to pay more for it. Look, I am an author: when my book appears in paperback, I get very few royalties. They are all tied in with the hardback copy, and you only get some from the paperback if its circulation is very high. So today already the author is bound to a contract of the type: "If you want millions to read you, you'll have to accept to get few royalties."
G. O.: Another question about the variety of available supports: one technical possibility that was put forward at one point and that today is more controversial is that of print-on-demand machines. Jason Epstein argued in his contribution to text-e that this would be possible for certain kinds of books. What are your views about this prospect?
U. E.: I strongly believe in it. When Geoffrey Numberg, nearly ten years ago now, first told me about the Xerox project at Stanford, I immediately recognized the great possibilities of these machines. At the time it was only a question of money because a machine like that was very expensive and the possibility of receiving material online did not exist yet. What, then, would happen? First, a book "printed on demand" is a real book: the difference is that, instead of coming off the printer's press, it comes out of a computer. Second, it can be very "friendly" because if, for instance, I have eye problems I can ask the machine to print the book in larger or smaller type. Third, it allows for the recovery of out-of-print book; today if one wants a book, even a very good book, which was published thirty years ago, you'll usually hear that it doesn't exist anymore: the publisher hasn't reprinted it because he knows that only about five hundred copies would be sold and that doesn't suit him. Not only that, but there is a crazy law in many countries according to which the warehouse contents are taxed, so that if a publisher has a thousand books stocked there, each priced, say, € 1, he'll be taxed for € 1000 even though in effect he'll have lost, not made this money! So: the print-on-demand machine would free publishers from the warehouse costs, it would give them the courage to take on new or avant-garde works, because they would not need to employ huge capitals to do so... There are many reasons for which print-on-demand could be very positive, culturally and economically. I've said that I don't prophesize, but if I had to make a science-fiction prophecy I would see a future bookshop that displays only the cover of the disk, just like in video stores, so you'd take that to the counter and request your printed copy. Maybe this is a little exaggerated, because the pleasure of the bookshop remains that of perusing the shelves, sniffing the book and so on. But a few pilot copies might suffice for the bookshop, even some printed-on-demand ones - so no printing press is required here either - and within five minutes the client would have the book. My prediction is that Stephen King's latest book would probably be printed because people would need to buy it in a hurry, like sandwiches, while if you wanted to read Thérèse Desqueyroux by François Mauriac, which might be out-of-print, then you'd need to have it printed.
G. O.: In his piece for text-e, Dan Sperber predicts a literate society in which one no longer writes, that is, in which the corporeal capacity for writing has been replaced by the dictation to a machine.
U. E.: It is highly likely, because this is what is happening with multiplication tables: our generation still knows that eight times six is forty eight, but the new generations that use calculators will lose this capacity. There's an excellent, fifty-year old science fiction story by Isaac Asimov which tells how in an extremely tense cold war situation, following a black-out of all computers, the Pentagon finds in Oklahoma the only person who still knows multiplication tables and imprisons him because he is the only calculator they have. In the same way we might lose our capacity to write, which would be the preserve of artists. But in an era of mechanization, there are still many youngsters who learn manual skills. So writing would not be lost, it would just be practised by a small amount of people. At this point, since writing is educational, schools could intervene and promote the taste for writing as "sport": you no longer write because you need to take notes, you do so in order to win the calligraphy or rapid writing competition, just in the same way that one doesn't run to go from one place to another because we have cars, but one does so to exercise. That could be an important task for the schools of the future. A last point: the French annoy me sometimes with their attachment to this principle of good education - which was fundamental in my day - according to which a letter should be written by hand, so that I never understand what they're writing to me. But it could happen that lovers who get tired of text-messaging on their portable phones rediscover the taste for personal hand-written letters as gestures of affection - in the same way that we could end up in a society in which women conceive children through chemical means but where people don't lose the habit of making love because in the end it is quite a lot of fun.
G. O.: I would like to ask you to comment on the future of libraries, which has been discussed at length in text-e. Is there still a need for libraries, of an archive for the memory of a culture?
U. E.: The problem of the future of libraries is two-fold. The shift took place in the middle of the nineteenth century, when one stopped printing on cloth-based paper and began printing on wood-pulp-based paper. Wood-pulp paper has a seventy-year lifespan. But if you look at some fifty-year old French books, they risk crumbling into pieces before you even open them. The first problem is that, since most of the work published after 1850 is destined to perish, one has to find ways of preventing this deterioration, and many international commissions are studying this. The other problem is the scanning of all this material for it to be transferred onto electronic supports, so that if you want to consult the Figaro of May 5th, 1921, you'll probably find it online soon. Some collectors of old papers might try to buy the original copy and put it under glass in order not to touch it too much, for it would probably crumble. There is of course the issue of how to make the electronic support eternal, while renewing it constantly. It is a huge problem but we'll have to get there. As far as old books are concerned, libraries are like art galleries: you do not replace the Uffizi with a good CD-Rom that shows you the Uffizi, however useful that CD-Rom may otherwise be; there persists the desire to go see the original, just as one wants to touch the old book, to look at the binding and so on. In this sense the great historical libraries must remain. On the other hand, what will happen to student libraries, where one can find text books, etc.? That, I do not know. Certainly new texts, once the problem of copyright is resolved, could be online: that would solve the problems of theft, of books already on loan, and so on. It could also increase solitude, because libraries are also meeting places, although university libraries could also become large spaces with screens where people go to read and can talk with others as well. I think that, for now, there will remain the need for paper memory, so that along with libraries for everyday use you'd still have the Library of Congress with its printed books and journals. But there is also the problem of the physical space for archives, for example those of periodicals and journals, which at one time contained four pages and nowadays contain a hundred pages. It could really be the case that annual runs of journals become the preserve of collectors only, who also embody a form of memory and conservation. Two months ago, I bought for quite a lot of money the entire collection of Lacerba, the journal edited by Giovanni Papini, and it is very moving to have it before one's eyes: even if they put it online one day, it is very nice for me to have the historical record. In this sense, I would also be a collaborator to the social conservation of memory.
G. O.: A last question: what happens to the "original" of a written text?
U. E.: Certainly at this point the idea of the original disappears. If I go look for the originals of my texts, I cannot find them anymore, unless I've printed them from the outset, since I'm wont to correcting them every week on the computer. But that is not really the problem. Let's say that you write the philosophical masterpiece of your life; even if you'll have lost the original there will always be the latest edition. It will be hard to evaluate your mental evolution, but if your contribution is interesting, that is all right. The problem is with the alterations that I can make to the texts of others. If I download onto my computer the Critique of Pure Reason, then begin to study it and to write all my comments between the lines, either I am a good philologist and I can tell which are my comments, or after three years I no longer know what is mine and what by Kant. We'll be like those medieval copyists who automatically corrected the text they were copying because to them it seemed right to correct it in that way, so the risk is that the philological spirit end up disappearing altogether. But here too, the risk exists for the young student who does not realize that he has manipulated the text; it isn't there for the scholarly and academic worlds, which would remain the guarantors of this philological guardianship. In other words, there would continue to be new critical editions of Kant in which one will be sure that the text is his. Anyway, originals are not interesting otherwise than emotionally, unless you become an extremely famous writer, in which case some American university is bound to offer to buy all your manuscripts. But let me tell you the story of an Italian writer, no longer alive, nor that famous, but of solid reputation. One day, a good American university asked him if they could have the original manuscript of a novel he had published, for about $ 5000 of that time. But he no longer had the original manuscript. So he had his whole printed book typed up, then with a pen he deleted various lines and wrote over them what had been deleted under the pen mark. Officially, then, this is the original manuscript of the author, although it isn't at all - so you see, even with the Internet there can be serious philological mistakes!
An interview by Gloria Origgi
Translated by Noga Arikha
Copyright © 2001 Bibliothèque publique d´information- Centre Pompidou