martedì 11 maggio 2010
Wiki@Home is pleased to present an interview with Umberto Eco. Professor Eco received wikimedian Aubrey in his home in Milan for a chat about Wikipedia, the Internet, collaboration, and of course books. The interview took place in Milan on April 24, 2010.
Eco: I am a compulsive user of Wikipedia, also for arthritic reasons: the more my back hurts, the more it costs me to get up and go to check the Treccani, so if I may find someone's birthday on Wikipedia it's all the better.
I am a car user and could not live without them, but this does not prevent me from stating all the defects and troubles of cars.
I once made a distinction between things good for the poor and things good for the rich, where rich and poor have no immediate connotation in terms of money, but in terms, say, of cultural evolution ... A graduate is rich, an illiterate is poor. There can obviously be a big entrepreneur who is poor and a little clerk who is rich.
Television thus is good for the poor and bad for the rich: it taught the poor to speak Italian, it is good for old women who sit alone in the house. And it harms the rich because it prevents him from going out and seeing things more beautiful at the cinema; it restricts his ideas.
The computer in general, and the Internet in particular, is good for the rich and bad for the poor. That is, Wikipedia is good for me, because I am able to find the information I need; I do not trust it, because everyone knows that as Wikipedia grows, the errors also grow. I found steep follies written about me, and if no-one had pointed me to them, they would be there still.
The rich are grown-up people, they can compare the information. I look at the Italian Wikipedia; I'm not sure that the news is correct, so I go to check the English version, then yet another source, and if all three tell me that this gentleman died in 371 AD, then I begin to believe it.
The poor picks the first piece of data he gets, and that's all folks. So Wikipedia, like the whole Internet, has the problem of filtering the news. It keeps both false and real news; but the rich know filtering techniques at least for the areas they know how to check. If I have to do a search on Plato, I have no problem immediately identifying the sites written by madmen, but if I am researching stem cells it's not certain that I can identify the wrong sites.
So there's this huge problem of filtering. Collective filtering is useless, since it could be subject to fluctuations. I noticed that in a certain period of Berlusconi's triumph people went looking for information about me in right-wing books and placed it in Wikipedia: as propriety prevents me from changing it directly, I left it. But obviously it was an entry made by the winners of the moment.
Collective control is therefore useful up to a certain point: it is conceivable that if one gives a false length of the equator, sooner or later someone will come along and fix it, but correction of more subtle and difficult issues is more complicated.
And it seems to me that the internal control is minimal, that is, it cannot control the millions of new changes flowing in. At most, it can check if a madman wrote that Napoleon is a racehorse, but there's not too much it can do.
Eco: There's assistance in case of insults, right. But those are the big things.
The principle of Wikipedia, in a certain sense, is that the more people there are, the more they are interested, the better the work is done. This is a bit of a paradox. There has been some research on it, the last I remember was in February 2007, from HP Labs in Palo Alto. It was purely quantitative and statistical, based on the English Wikipedia; it found out that the pages with more changes are on average those with the highest quality. The more people there are, the better it is. Then there's the actual problem of the long tail. There are so many pages just fairly important or problematic or debatable. The page about you, for example, may fall into this set of pages, and besides it's a biography. Biographies of living persons are the most problematic, because of recentism (adding useless tidbits which happened just now), identification of the sources, etc. All biographies are generally a problem, although in the case of historical figures there is more agreement. I find it interesting to look at the discussion pages which in theory should be the most problematic, on topics such as creationism, or intelligent design. In the English-language Wikipedia they are abysmally long, because people often quarrel not just over whole paragraphs, but over individual words, or the starting sentence. It's better to have more eyes, like the "wisdom of crowds" theory by Surowiecki, that says that when there are 4 parameters (independence, diversity of opinion, aggregation, decentralization), on average, judging by a crowd is better than that of the experts.
Eco: I don't quite agree with this. I am a disciple of Peirce, who argues that scientific truths are, ultimately, approved by the community. The slow work of the community, through revisions and errors, as he put it in the nineteenth century, carries out "the torch of truth". The problem is the definition of truth.
If I were forced to replace "truth" with "crowd", I would not agree. If you make a statistical analysis of the 6 billion inhabitants of the globe, the majority believes that the Sun revolves around the Earth, there's nothing you can do. The crowd would be prepared to endorse the wrong answer. This also happens in a democracy: we are noticing it these days, the crowd votes for Bossi. To carry on his coup d'état, Napoleon III broadened out the electorate and included the peasants, because the crowd of the countryside was more reactionary than the crowd of the cities.
We must therefore find another criterion, which I think is the motivated crowds. People who work on Wikipedia are not just an aristocracy, just professors, but they are not the indiscriminate crowd either: they are the part of the crowd who feels motivated to work with Wikipedia. Here it is: I'd replace the theory of the "wisdom of the crowd" with the theory of the "wisdom of the motivated crowds." The general crowd says we should not pay taxes; the motivated crowd says that it's fair to pay them. In fact, it's not the ditch diggers or illiterates who contribute to Wikipedia, but people who already belong to a cultural crowd for the very fact they're using a computer.
Eco: Peirce was thinking of the scientific community, of course; especially in his time, it was definitely separated from the common crowd.
Eco: Wikipedia has two unrelated functions in my opinion. The first one is to allow quick searches for information, and it is just an extension of Garzantine [a popular Italian series of compact encyclopedias], period. The other, and this is what we are talking about now, is whether the control from below can be many times more successful than the control from above. Since the world is full of expert idiots, certainly it can be.
Just an example: some days ago I was correcting an essay about Benedetto Croce. Croce, building on his authority, spread false ideas for 50 years in Italy, and everyone in Italy had accepted them, without considering that he knew nothing about art. He was the aesthetics master for two or three generations without having understood anything about art. So you see, sometimes the authority... Responses of artists, children, students would have been really more useful. This control by the mass can, produce a development in the long run, as Peirce said.
But I keep saying that I am increasingly exposed to the risk of my inability to filter the news. Lately I started writing down some false information, some errors that one can find in Wikipedia. In the same article, for example, there were two contradictory reports, a sign that there had been an amalgam.
Eco: Not in that case. I don't edit pages, except for the page about me when I found it written that I married the daughter of my publisher, since as a matter of fact I didn't. Poor soul, she ran such a risk! [laughs] Another time I was described as the eldest of 13 brothers.
Eco: Yes. If the error was made by another person, I don't see why I should waste my time to correct it. I am not the Red Cross. [laughs, N.d.R]
Thus, I actually noticed that there was a contradiction, within the same article. The problem is that I'm good, I can notice the error, because that's my job; another person, less competent, could read just half of it and take the first version.
Eco: Of course, it's a matter of time. When I write, I consult Wikipedia 30–40 times a day, because it is really helpful. When I write, I don't remember if someone was born in the 6th century or the 7th; or maybe how many n's are in "Goldmann"... Just a few years ago, for this kind of thing you could waste a lot of time. Nowadays, with Wikipedia and Babylon, which checks the spelling, you can save a lot.
Furthermore, this often implies that the text itself is free, meaning released with a free license. This mechanism is obviously related to several issues: the issue of filtering, the issue of bottom up vs. top-down process, as well as the issue of having a community of peers with different values and motivations than a scholar community.
In your opinion, is this experience exportable to other writing mechanisms not aimed to collective knowledge production, as Wikipedia is?Wikipedia, in fact, was born as an encyclopaedia developed within a wiki, a specific software, and has been a great success, against all odds. People have tried several times to build similar projects: the Los Angeles Times, once, tried to aggregate collective editorials, it was a failure. Thus, it seems that some projects can be collaboratively developed, meanwhile some others can't.
Eco: You are now talking about collective collaboration. Well, there are a few things that the Internet provides: the first are mere data, as the train schedule that no one can correct. Another is encyclopaedic information, which can always be corrected, because the author could be wrong or simply has not said everything yet. The third ones are texts: should I edit other's texts? Moreover, there is the whole universe of blogs and Facebook; but it doesn't matter right now, they are people talking to each other, conversing.
In these very days I had to debate on Hypatia: I looked for some information on the Internet, and I found interesting and less interesting texts. But they are texts. The Internet provides us classical and contemporary texts, but if they are wrong or I do not agree with them I surely don't edit them. I cannot say "Your opinion on Aristotle is wrong".
Eco: More, it is signed. In fact I found many interesting documents that are not signed, I never understood why.
Eco: Do you mean bibliographies?
Eco: I stumbled upon some of them. Actually regarding Hypatia, I found a project where different scholars collaborate to translate a text from the 10th century.
Eco: This is yet another topic. Congresses were made by textual critics to investigate this topic. These are truly auto-controlled communities.
Eco: Yes, but where we know that a single scholar belongs to a single university, we know where he comes from. In this case, it happens something similar to when people used to collaborate in writing a book, and they needed to take the train once a week to meet and discuss. It is collaborative team work that is controlled by someone. It is not the wisdom of the crowds. It is simply the scaling and the simplification of the collective research work that once required filthy travels and nowadays it can be done online daily... I'd rather call them uncontrollable and controlled communities.
For example, I was thinking of a project for the Italian studies community, which could be granted from institutions and still let the community free to auto-control itself.In your opinion, is this auto-organization also possible in these scholarly communities?
Eco: I recall a conference in Bologna, about textual criticism studies, that was dedicated mainly to digital humanities projects and text research environments and functionalities. Evidently, this was a leaderless community, auto-controlled and leaderless. But "leaderless" is a phrase: because in scientific communities which self-legitimate there's always someone who gains more authority: if an important philologist propose an interpretation, the others will follow.
Therefore [online collaboration within scholarly projects] it is not the same thing of Wikipedia.
Eco: Let's take as an example the magazine Nature. In the scientific world, if a paper appears on Nature - where a peer review has been carried out and there's a wide control - it's taken seriously. It's anyway possible that Nature can make a mistake and reject a brilliant paper: nonetheless Nature is considered a center of reliability, with fringed boundaries. Because an error, or a small academic revenge, can always happen...
Now take me as an example: with my age and my body overweight I entered the high-glycaemia phase of a type II diabetes. Once, the limit for defining glycaemia "high" was 140, today it's 110: we all know that this new limit has been set by the pharmaceutical companies for selling their products. So, 140 is risky, maybe 110 is too low, one can get along with, say, 120. Maybe in a decade the limit will be adjusted to 120, or they'll decide that 110 is good in terms of preventive medicine [laughs]. We realized that swine flu was partly rubbish, spiced up by the vaccine manufacturers. We realized it too late, after billions had been spent; we realized that far fewer people than expected died, that they maybe overstated it.
In one way or another things fall into place: these are the controlled communities, not anarchical, but with a fringed authority. That has nothing to do with Wikipedia, where the anarchy is bigger.
Eco: There is an adjustment. Galileo, Tycho Brahe and Kepler in the end agreed that Kepler was right. Infinitesimal calculus has been discovered both by Newton and Leibniz but in the end everybody agreed with Leibniz. [he laughs]
They might have been the wrong choices, but they've been made that way.
There wasn't any authority, the emperor, who decided it. It's been a collection of habits and applications.
Eco: Sure. The Accademia del Cimento began first! And without the Internet. [he laughs]
Eco: First there was a few of them in Florence, then a few more at the Royal Society; now it's a crowd.
Back to the previous point, in Wikipedia also we can notice a cultural difference between the articles about technology, science, maths, physics and the articles about humanistic topics. Humanistic articles are much less (philosophy, history, literature).This in Wikipedia. Within the academic communities, in a similar way, there's a different impulse to the collaboration. In the "soft" sciences, the authorship, the authoritativeness and even the interpretation, matter more.
Eco: For what are soft sciences, there is absolutely less impulse to collaboration. There is much more interest to be the main character of an idea, than being just a "water carrier".
That's for sure. A scientist in these cases is used to not being mentioned and to know that however is carrying forward a fundamental research. In soft sciences, this happens only to the exploited student who is sent to gather data that the professor will sign and profit by.
That's an old story, there's no escape from that...
Eco: I don't believe so. Think about ancient Greece. Plato and Aristotle, one being the other's disciple, developed two opposite philosophies. On the other hand, Euclid came and it is still discussed, his fifth postulate survived for two thousand years.
Eco: Science is cumulative-destructive, it stores what it needs and throw away what it doesn't require. Humanities are totally cumulative, they don't throw away anything: in fact, there is always a return to the past.
On the other hand, they are totally destructive in the way, as Maritain stated regarding to Descartes, "a philosopher is a novice in the Absolute". For Descartes, everything that philosophy stated before him was false. If a mathematician did that, it would be the end of mathematics.
In volunteer-based projects like Wikipedia, the problem matters less, but given that the scientific world is moving toward more and more intense collaboration (and the humanistic world as well, although more slowly) we have to face the fundamental question of copyright.In Wikipedia it's been solved by adopting free licenses, and the culture of the nicknames - or no names - helps; in the academic and scientific world the culture of the name, related to important things such as a personal career, leads us to a complex problem of recognition of the intellectual property.
Eco: This is certainly coming out, also in the world of books; I think that in 50 years we'll have a very deep mutation. We'll probably have a cultural situation similar to the one in the Middle Ages, where comments and comments were produced, and the authoriality was lost. Then, from the Romanticism on, the authoriality became excessive.
But I cannot say up to which point we can reach a total anonymity. Although it can look democratic, total anonymity gives the idea that just one and only one truth exists. Can we have a moment in a future where Wikipedia itself, on certain articles (not the one about the multiplication table, of course), can open sections called "Conflicts" where - signed - different theses can appear in opposition?
In spite of the always present denying madman, we're certain that Napoleon died in Saint Helena. That Pius XII did the right thing during the Holocaust, it's an open debate. What does Wikipedia do? It says that Pius XII did not do enough (irritating millions of catholics)? It says he did (irritating millions of non-believers)? Or does it open an appendix in which a series of authors, each assuming responsibility for their words, expose in twenty lines the conflict of interpretation?
Eco: I never went to see it.
Eco: Of course, everything can change tomorrow.
Eco: Yes, that's why it remains and is not modifiable, because that is the article and is not reprinted. They create an appendix, of course. The destiny of the Treccani is to wikipedize itself.
Eco: With the current speed of renewal of the culture, if an encyclopedia doesn't go online for being updated month by month, it is doomed forever. Even when it talks about Parmenides, because even tomorrow a book casting a new light on him can be published... but never mind of Parmenides. Take "Aeroplane", as example: who knows what the article about the "Concorde" said before the Concorde crashed.
Eco: It's a proposal for the Dizionario degli Italiani [a collection of biographies], but it's being withdrawn. Since writing articles is too expensive, they asked the readers to "donate" some, not considering that revising those donations takes such a group of editors that the costs are higher than simply paying for the articles.
Eco: I'm not sure I understood well what you said, but if I did, all of this depends on the fact that - apart of botanic and zoologic taxonomies - a global classification does not exist, but only a local one does. In my last book "From the tree to the labyrinth", I wrote a 100-page essay exactly about the history of the classification, from Porphirius' tree to what we today dumbly call "ontologies".
The problem here is that centuries have been spent in trying to make a total classification, but it's impossible, it's always local and in perspective. Consequently, it can be authorial and not collective. It's a goal attainable in certain fields only, for example animals and trees, as they are universes somehow finite. And it leaves anyway big problems in the classification of the insects. And there's the famous example of the ornithorhynchus, for which it took them eighty years, but they found an agreement, all together.
Animals then are finite and - one way or another - can be categorized. In those cases where elements are more disparate, instead, total and collective categorization is impossible.
Eco: Yes, although probably there's more than one book for each person. Yes, I do. But this question is like "Why did you care about the Middle Ages?", that it's like asking "Why did you marry that one and not another?" [he laughs] If you're interested, I made the translation of that book and talked about it in a collection of essays about literature... but this has nothing to do with the question.
Eco: I am very empirical. I make my living on the gains of intellectual property, but every time I've been an object of piracy I got off cheaply. It happened that my American publisher sued a university for having made thirty copies of a book of mine, and I protested. It's fine for me like this, at least 3 or 4 of my books can be downloaded through eMule... Why am I so careless about this? Considering that I live with that, I should be worried. One answer might be that I'm earning enough this way, the other is that I am a good democrat.
Let me make an example. When the newspaper La Repubblica decided to distribute books with the paper, they began with Il nome della Rosa, giving me a modest flat sum. And then they sold two million copies that day. I decided not to mind, I didn't earn anything but it was all right. Six months later I checked the reports of my publisher and the sale of the paperback hadn't changed at all. That is, those two million people were people that would never have bought my book in a bookstore. I didn't lose a sale. This means that the "space" is so big that [the piracy] doesn't look like a tragedy to me. It's the author that sells a thousand copies that gets angry if a hundred of them are bootlegged.
Up to the 17th and 18th centuries, a writer made his living from a benefactor's will. Maybe we'll return there, we won't be paid by the audience, but by a patron. Ariosto got off well, why shouldn't I? [he laughs]
They got off even before. Then the 18th century revolution - when the storyteller went around selling his own books - gave birth to the rights. In a sense, this democratize that work, because the writer and the philosopher did not have to lick the benefactor's ass any longer.
Well, nothing changed that much between the way that Ariosto licked the Estensi's ass and the way a lot of people lick everybody's ass. [he laughs] Ariosto doesn't interest us less because he writes two ottava rima to thank the Estensi.
Eco: I don't understand all these protests against Google Books. Honestly, I get angry because I can see two pages and I cannot buy the book. The publishers should be enthusiastic, I don't understand. It's like the pedestrian areas: when you close a road to the cars, all the shopkeepers protest, although it's scientifically demonstrated that such an action increases sales.
Eco: Each writer lives a conflict about this: on one hand he's happy that his book is read, on the other he's sorry that his grandchildren won't earn anything from the rights. Now, my publisher said he'll give the rights of Il nome della rosa for making an eBook for the Kindle, I think. The percentage is much lower than for normal books, but it's all right. I personally don't believe in it, I think that people still want paper for reading a book, but I have no problems, it's correct that people asking for an electronic version might have it. It doesn't look complicated, they pay for the rights, although less because the eBook is cheaper. Either it'll be a smash and you'll sell millions, or you'll sell few copies, and it'll be all right anyway.
I think that everybody is overreacting, just like the publishers against Google. Google Books is for selling books, not for selling less books. It plays the same role a bookshop does, when you go to browse the books. You can buy them, or just read a couple of pages, or read the index. Just as with Google.
And there's the trend to give more and more things for free. I cannot stand Adobe, that every year asks me to pay for reading the PDFs. In a few minutes I found programs that do the same for free. I don't understand where's the return for the developers...
Eco: There's also OOorg, that substitutes Word. It's very good and it works very well.