FYI France: the B.de France at Berkeley -- part 3 of 4

[Jack Kessler <kessler@WELL.SF.CA.US>]

The promise of the conference's subtitle, "...and the Future of the
Library", was realized in the third session, on Sunday afternoon.

In his superbly-titled talk -- "Les Non-bibliothe`ques", or "the Non-
libraries" -- Goe`ry Delaco^te made a superb presentation. Delaco^te's
background is fascinating. He directs the Exploratorium in San
Francisco, is a physicist by training, is a graduate of the Ecole
Normale Supe'rieure, was a member of the group which assembled the
Me'diathe`que de la Villette in Paris, and was the architect of the
CNRS' INIST: the Institut de l'Information Scientifique et Technique.
He is a leading figure of the science establishment in France.

Delaco^te described his primary effort at INIST as having been to
establish a "built-in process of adaptation and transformation":
without these elements, in balance, Delaco^te insisted, any institution
like a museum or a library will atrophy. INIST, Delaco^te teased, was a
librarian's dream: "no visitors, no users, only the documents".
(Another, perhaps teasing, suggestion was that at Nancy windows had
been inserted into the INIST building design, "so that the towers might
be usable later" once they might no longer be needed for documents,
according to Delaco^te: the realists and the cynics in the audience
immediately thought of the B.de France's much-vaunted and much-
criticized "book" towers -- Delaco^te's face gave no clue that he was
alluding to those.) The INIST project had been to establish document
delivery and database services for France's giant CNRS science research
operation. From the first, this was conceived as a non-library project,
more like the US' National Library of Medicine or even the anti-library
approach of UK's Boston Spa center, than like a conventional book-and-
journal library.  Throughout, the emphasis at INIST was on the delivery
of information to users, rather than on its storage and preservation.

The driving concepts of INIST became 1) production and 2) marketing:
both ideas foreign to the traditional library.  France is reorganizing
all its culture, not just its sciences and its libraries, Delaco^te
observed. The key to this for any institution, he said, is integration,
and a strong research and development component: otherwise one might
establish a fantastic flagship, but a flagship only, and one unable to
adapt to change. The Exploratorium has, he said, 1) exhibitions, 2)
teaching and learning programs, and 3) media and communications work,
to provide their R&D component and to give their institution
flexibility. Speaking directly to his friends from Paris he advised,
even warned, that without similar approaches they would find their new
library institution to be lacking.

Even more important, asserted Delaco^te, was the development of a
"marketing attitude". This sounded strange to the US audience, but
perhaps was familiar to the French. Commercial marketing ideas -- much
of it taken from US sources -- are the rage now in French education and
government circles. French business schools -- unthinkable as a concept
in the 1960's -- are over-subscribed and multiplying in number now.
But Delaco^te's institution at Nancy is built, the B.de France people
are building theirs, and the institutions of the US sceptics in the
audience, who were shocked by his "marketing" ideas, are on the ropes
or are closing. The problem of education in the US and in France is a
worldwide problem, he asserted, demanding a revolution in approach. His
suggestion is that only some new mix of cultural, educational and
professional networks will be able to serve the new realities. Ideas
like marketing and built-in processes of adaptation and transformation
will be keys to the success of any new institutions, including museums
and libraries, Delaco^te concluded.

Geoffrey Nunberg, of Xerox PARC and Stanford, then gave an intriguing
talk, fascinating his French and American audiences alike. He was
concerned, he said, to discuss "the place of books" in a new
information age, without indulging in either of the two extremes, as he
described them, of "bibliophilia" or "cyberphilia". He is not a
determinist, he asserted, and believes that we must understand that the
future will be every bit as subtle and as complex as is the present.

Discussions of libraries, said Nunberg, too often "mistake the cultural
content for the artifacts". He said that "information overload" always
has been with us, and in ways which we often haven't recognized: the
printed output of the Traveller's Insurance company, he pointed out,
would fill the Bibliothe`que Nationale's shelving within six weeks.
Cultural forms are at work constantly, he suggested, sorting and
organizing our information.

Nunberg then spoke about some of the artifacts. He talked about the
opposition between print media and the screen, saying that he thinks
books will be with us for some time, for some things: "paper documents
are a superior medium for sustained reading," he believes.  From the
"electronic samizdat" of the Internet, he suggested, old media forms
are perhaps being reborn, as well as new forms created.  The pamphlet,
for example -- the short format, immediate news source of the printed
text revolution -- may be experiencing a rebirth, in the bulletin board
and e-conferencing formats of the Internet.

Electronic publication, said Nunberg, resembles the 14th century
scriptorium more than it does the recently-modern printing house.
Printing's distinctions among different printed forms -- journal
articles, monographs, books -- are blurred in electronic media as they
were in the scriptoria. Once again, as then, it is hard to "locate"
texts. Perhaps in a bow to the presence of Le Roy Ladurie in the
audience, Nunberg praised what he called the "shimmering
inter-textuality of annales": the fascinating suggestion that the
electronic media may contain the very inter-disciplinary possibilities
which have so attracted the annales historians, and other thinkers who
have crossed traditional disciplinary boundaries. He has little trouble
with the so-called problems of electronic journal publication, he
said:  for some time, in science, publication has confirmed rather than
created -- there is room for both, he thinks, both the publication
which confirms discovery after the fact and for publication so
immediate that it takes part in the act of discovery itself.

One interesting corollary to the "library without walls" idea, said
Nunberg, is that "the permeability goes both ways". Just as the
information can get out, the users can get in: copies can be made with
great ease, one can't "burn" electronic texts, he said.  Electronic
texts, so easily reached, furthermore can become a locus for "imagined
communities": any active e-conference participant can attest to the
truth of this assertion -- tele-commuting and tele-communities are only
two of the phenomenal social by-products of the technology.  The
_Tatler_ and the _Spectator_ are the two older parallels closest to
modern electronic publishing, Nunberg asserted.  Electronic publishing
supplies the modern ideal of community in anonymity, he said. Despite
the gains of electronic media, though, he continues to believe there
will be a role for paper publishing, if only as a ceremonial practice,
to endow with legitimacy efforts made using other media as well.

Alain Giffard's presence on this panel was curious. He is no doubt as
concerned with the library of the future as are the others, but his
charge of actually implementing his own and others' dreams at the B.
de France made his talk that much more interesting to his audience.  To
accomplish this, said Giffard, the B.de France has divided the task of
implementation among six domains:  1) administration -- he reminded us
that his library will have 2000 employees; 2) technical management; 3)
collection management -- selection, acquisition, conservation, and so
on, all the traditional activities of the Bibliothe`que Nationale,
following much the same system; 4) public information -- accreditation
of readers, reception, document delivery, circulation -- his hope is to
achieve a single sequence of information processing which will serve
all these needs; 5) electronic management of documents; and 6)
management of audio-visual spaces -- there will be many, for
conferences, meetings, films and videos.

The bidding process for all of this has been completed, Giffard said,
and the French firm Cap-Gemini Sogeti, in association with the Canadian
firm GEAC, won the bid. He expects the process to cost about US $ll
million, and employ about 100 people, and he bravely asserted that it
"would be complete": as a concept by the end of 1992, as an operational
reality by the end of 1993, and as a working system by the library's
opening in 1995. It would seem that Giffard's ability to complete his
particular task will be the key to preventing the B.de France's being
merely a badly-overstocked book warehouse by the date of its opening.

Giffard then turned to perhaps the most exciting, "gee-whiz" feature of
the B.de France project, in the minds of many in his audience:  the
digitization of texts. By 1995, he said, they now hope to have between
150,000 and 200,000 titles digitized. (This represents a retreat from
previously-announced estimates of 300,000 and 415,000; but those were
not official, and one sensed that the more modest figure might in fact
be realised.) These are to be the "grand texts of reference, those
which will be the 'classics' of the 21st century, and 'rare and
precious' works". The librarians in his audience wondered who might be
on the selection committee; as Umberto Eco has warned in this context,
"even Plato and Dante have known their periods of disgrace." (Nouvel
Observateur no 1406, 17-23 Oct., 1991, p.16)

Two techniques are being used, said Giffard. Fragile, ancient works are
being microfilmed. (nb.There is a project at Yale devoted to the
question of digitizing microfilmed texts.) More recent works will be
scanned in.  (He mentioned nothing about optical character recognition
or compression algorithms, or the resolution levels at which this
scanning will be performed and stored: one can only hope that the ocr
and compression will work, and that resolution levels will be high
enough for later reproduction technologies to use well.) Already 9,000
works have been digitized, Giffard said, some using robot page-turners
to speed the process. They are determined to establish a production
rate, this year, of 40-50,000 works to be digitized per year.

Giffard then mentioned the need to develop a single workstation which
would enable a user to assemble documents, make notes, make marginal
notes, and compose a final document: they are dedicated, he said, to
the idea that their workstation will combine such functions. One key
idea, though, is to make their approach modular, so that as any
particular aspect of the technology pulls ahead of the others, its
component might be replaced, without having to replace the entire
workstation system. Experiments are being run already, he said, using a
digitized file of the works of Henri Bergson, and using Sun, Apple and
Next equipment. For this sort of development the B.de France is relying
upon liaisons with Grenoble, the CNRS, the Ecole des Hautes Etudes, the
University of Toulouse, MIT, Project Xanadu, and Cornell.

Giffard is concerned to keep pace with the pc's coming transformation
into a vehicle for multimedia. He sees, however, great continuity
between what they are undertaking at the B.de France and information
work of the past. "Digitized text is the paper of the computer", he
declared, "the word processor is the successor of the printing press".
There no doubt might be revolutions to come in the meaning of text,
particularly as to its relations with its context and with its
"linearity", Giffard said. But hypertext and cybertext are old ideas,
he reminded us, it's just that some new electronic techniques at last
perhaps are making them realizable.

The last speaker of this second day, John Gage of Sun Microsystems,
told us of Donald Davidson's idea that language is only a theory, a
"way of knowing your way around the world". He then reminded us that
theories change: he brought in Richard Rorty's idea of language and
culture, "as a process of change of metaphor": "when you wish to change
the world you change the metaphor" -- that's the old power of magic,
Gage said, "to break our common ideas". The B.de France's broader
purpose, then, provides just such a metaphor -- that of the world
library -- and, like other powerful metaphors it has within itself the
power to change our perceptions.

One example we'd just heard, Gage continued. That was Giffard's
references to "reaching in" and "reaching out" at his new library: a
metaphor, asserted Gage, with powerful suggestions as to the type of
institution which Giffard's may become. Are libraries "containers of
precious objects", Gage asked, or are they "conversations among
people"; are they "windows into the treasures of France", or are they
"windows onto the world"? Unless these ideas are addressed, said Gage,
as they appear to be at the B.de France, we are in some danger of being
drowned in a "babble of languages of passing theories".

Gage suggested that although the metaphors might be new, the ideas
themselves very often were old. Much of the new today derives from the
ideas of Bell in 1876, or of Otlet in 1934 or Bush (and others) in
1945. Gage mentioned Sutherland, Engelbart and Furness, all of whom
worked in the 1960s. Sakamura, Kahn and Cerf, and the groups at Bell
Labs, Harvard and SLAC during the 1970s were mentioned.  From all this
we now have a New Information Technology, Gage says, but what seems to
be lacking is a New Intellectual Technology to go with it. We need, he
concluded, to extend our previous oral civilization beyond our current
writing civilization, into some sort of "new object civilization" if we
truly are to take advantage of our new information tools.

The day ended with a panel discussion, ably chaired by Gage, consisting
of Giffard, John Garrett of the Coalition for Network Research
Initiatives, Jean-Pierre Sakoun of GEAC, a firm which has entered the
library information market very successfully in Europe, and William T.
Crocca of Xerox, who personally has been working on well-known
experiments in digital preservation at Cornell and Harvard.

Crocca told us that the Cornell project now has 1000 books online, and
he projects digitization costs of no more than $4 per book. He showed
us a very handsome volume, reproduced using their techniques, which did
not at all have the "feel" of being a "reproduced" book. Sakoun then
lamented the lack of financial support in the US for public libraries:
not the case in France, where currently public libraries are getting
plenty of support, he said, although there academic libraries still are
lacking resources.  He mentioned the good efforts of the EC's DGXIII,
with their Projects of Cooperation for Inter-connection, and other
European efforts, such as those of the Deutschebibliotek, the British
Library, the Swiss Union Catalog, and PICA efforts in the Netherlands.
Gage suggested that the problems no longer are technical: with advances
in Z39.50 and public key encryption, most of the remaining barriers to
shared information over the networks are political and social.

John Garrett then waded into one outstanding social and political
problem which politely had lain quiet until then. "I challenge the
French", he said, looking directly at the B.de France visitors from
Paris in the front row, "to join the rest of the networking world, and
join the Internet!" The Internet is present now in 78 countries, he
said, linking over 500,000 computers, handling over 2 million
simultaneous transactions, sustaining a 15% per month message growth
rate, and serving untold millions of individual users. The French are
faced, said Garrett, with two choices: either they can build pretty
buildings which contain lots of books, or they can join a world
information system which already is thinking through its technology
decisions in the framework of a world network concept. Giffard bravely
tried to answer. Much work has been done, he said, to advance
information networking within France. One barrier to its success
perhaps has been, he suggested, the immense success of Minitel, which
has robbed French networking of some of the incentive to take its
development forward. Giffard might also have mentioned the great amount
of work which has been done to improve European X.25 e-mail techniques,
including the development of the ASTRA query system which the Library
of Congress itself had said it was adopting for its own European
projects at the conference only that morning. Ultimately, though, all
this was a passing reference to the "protocol wars" which have
smoldered for a while between US and European efforts in networking. It
provided the conference's most delicate drama, but no resolutions:
more on this in a later posting, as there are some new developments.

Next: Final analysis -- the Grand Historians, and "will it be built?"