“Professor Eco, you’re a man of letters, a writer, philosopher, a
historian. On the desk beside you is a computer. Is modern computer
technology actually functional for you as an author and literary
researcher?”
Eco glances over at the computer, smiles, then nods thoughtfully:
Yes, but sometimes the computer can also give paralyzing results. I will
give you an example: I was invited by Jerusalem University to a
symposium whose theme was the image of Jerusalem and the temple as an
image through the centuries. I did not know what to do on this
particular topic. Then I said to myself, well OK, I have worked with
stuff from the beginning of the Middle Ages; my dissertation was on
Thomas Aquinas.
He points to the rows of well-filled bookshelves on my left...
Here I have all the works of Thomas Aquinas with a reasonably good
index, and I looked there to see how many times he quoted Jerusalem and
tried to say what use he made of the image of Jerusalem. Now, if I only
had these books - well, that index is a reasonable index which focuses
only on the larger, more intensive treatments of the word ‘Jerusalem’ -
I would have found say 10 or 15 tokens of ‘Jerusalem’ which I would have
been able to examine. Unfortunately I now have the Aquinas hypertext...
He glances again at the computer in the corner... and there I found,
that there were - well I don’t remember the exact number - but there
were round 11,000 or so tokens... Working with 11,000 references is just
impossible. That’s far too many.
“So the system you use doesn’t ‘filter’ well enough in other words?”
I cannot manage to scan as many as 11,000 tokens. Now, if I had only my
old traditional limitations then I would probably have done something
more or less reasonable on that particular topic.
“That’s because the human person who is searching does it in a kind of
sensible, intuitive way, whereas the computer just does it in a very
mechanical way and just picks out every single example?”
My theory is that there is no difference between the Sunday New York
Times and the Pravda of the old days. The Sunday New York Times that can
have 600 or 700 pages altogether really just contains old news fit to
print. But one week is not enough to read a number of the Sunday New
York Times. So therefore, the fact that the news items are there is
irrelevant, or immaterial, because you cannot retrieve them. So what
then is the difference between the Pravda, which didn’t give any news,
and the New York Times which gives too much? Once upon a time, if I
needed a bibliography on Norway and semiotics, I went to a library and
probably found four items. I took notes and found other bibliographical
references. Now with the Internet I can have 10,000 items. At this point
I become paralyzed. I simply have to choose another topic.
“So information overload and this extreme, non-intuitive selection of
information is the main problem?
Yes, we have an excessive retrievability of information. It is neither
ironical nor paradoxical, I think, what has happened with Xerox copies.
Eco picks up a pile of papers from the desk in front of him and waves
them.
Once I used to go to the library and take notes. I would work a lot, but
at the end of my work I had, say, 30 files on a certain subject. Now,
when I go into the library - this has happened frequently to me in
American libraries - I find a lot of things that I xerox and xerox and
xerox in order to have them. When I come home with them all, and I never
read them. I never read them at all!
“You never seem to have the time, do you? Once you know that it is
there, you feel reassured, and so you don’t read it.”
Exactly...
"Xeroxing then can paralyse your reading activity? That’s another risk”
Sure...That’s another risk which is sometimes very real.
“Yes, well then, what do you think about the idea of these personal
information filters. This idea that you can kind of make a personal
profile, and the system will search Internet on the basis of this?”
This is what I call the art of decimation...
“Decimation?”
Decimation. You kill only one person out of ten...
He gestures towards the well-filled bookshelves again. The number of
books that only concern my specific domain, not to speak of the other
ones that I receive weekly certainly, exaggeratedly, overwhelms my
reading...
“Your capability, capacity?”
...my capability, my time. If you have a certain experience you are able
to... well, you can make a very random decimation. On this or that
subject for instance, there may be no more than ten possible new ideas.
It is rare that that is the case.
“And the working hypotheses you make are based on these?”
So .. if I read only one out of ten books, probably there will be an
idea in there I can find, and if it is not there, then it will be in the
next group of ten books that I pick up. But this is a very random thing.
“But it is also very much based on your past experience, obviously?”
Oh, sure, it is random, but based upon past experience.
He reaches for a book from his desk and begins to thumb through it. OK,
now I am able to open this at the first page, to look at the summary, to
see the bibliography and to understand if the fellow is reliable or not;
if there is something new there or not. And since I have long
experience, my decimation is oriented. I sense it is better that I read
this, and not that etceteras.
“So you are able in a way to recognize newness, or innovation?”
In a way, in a way. I can commit mistakes of course, but if I make a
mistake today, I probably won’t do that tomorrow. Possibly I may choose
to disregard some book or other and that may be a mistake, but the next
week I will come across yet another book, and so on. But a student of 20
years old, or even of 30 does not have this kind of filtering ability.
We have to invent a practice, a theory. A practice or training in
decimation.
“Well now, how do we do that?”
Eco leans forward eagerly in his chair. Well, it still has to be
invented. There must be some rules. There are some very elementary rules
such as: control the dates of the bibliography for instance. If you are
working on a certain subject you may find many references from 1993 and
1994. But in relation to other subjects finding only references from
1993 and 1994 might be negative, you ought to be finding some older
dates. So if you read a book on Kant and you have only a bibliography
from the nineteen-nineties then this is suspect. The author is working
from secondary sources. If you are reading a book on hypertext and you
find an old bibliography then this is suspect, because every day there
is something new about this particular field. So there may be some
first, elementary rules you can use in order to isolate certain things
immediately.
(to be continued)
