10.2002b FYI France Essay: "French libraries online -- the view from the outside"

by Jack Kessler, kessler@well.sf.ca.us

Originally published in French in Bulletin des bibliothèques de France t. 47 n. 5 2002, pp. 10-31, ISSN: 0006-2006
-- online in French as, "Les bibliothèques françaises en ligne : Une vue de l'extérieur", at

All this appears here as a feature of the FYI France ejournal, ISSN 1071-5916, which since 1992 has been distributed for free via email every month except August. Ejournal subscriptions may be obtained via email request to: kessler@well.sf.ca.us.

Here this file is one of a number made available -- hopefully attractively, all in one place, and relevant to libraries and online digital information work in France and Europe -- as part of FYI France (sm)(tm), an online service to which anyone can subscribe by postal mailing a check for US $45 payable to Jack Kessler, to PO Box 460668, San Francisco, California, USA 94146 (site licenses also are available): please write your email address on the front of your check. Please email suggestions for improvements to me at kessler@well.sf.ca.us .



French libraries online :
the view from the outside

by Jack Kessler, kessler@well.sf.ca.us -- April 1, 2002

One of the greatest pleasures of Internet use is the discovery of remote resources. In my personal case this first happened one evening in 1988, at the University of California -- a fellow graduate student showed me how I could look up bibliographic records in the libraries at Oxford, all the way across the Atlantic Ocean and over in the UK, from a basement laboratory in Berkeley. That was magical, to me, back then.

Since then, many of us in the library community have come to take these things for granted. "Well, of course you can see bibliographic records online," we say now, "so now what about the texts, and the pictures, and associated sound files and multimedia?" -- and how to use all these interactively for distance education and lifelong learning, and how to integrate them with medical and military and onsite scientific research and other strange new applications? -- and in what format do the records appear, and using what encoding, and is their metadata proper? -- and do they show their accented characters correctly? Too often we forget, amid all this detail and our excitement over recent developments, about the sheer magic of Internet access.

To many outside the field -- and outside the few nations, and the few truly-"wired" cities within those nations, which currently are leaders in digital information and communication -- this Internet access magic still has not arrived, or if it has it still is magic only. To vast populations of the globe -- in Africa and Asia and Latin America, but also in villages in the Dordogne and in Kansas, and in certain apartments in Limoges and in Tulsa, and in certain significant pockets of resistance still holding out within our most advanced offices and libraries, and among the poor and the illiterate nearly everywhere -- Internet access still is magic only, even if it is something known about at all. With all our advances online since 1988, we still are the few speaking to the few.

The specialty which I personally developed, back then, was one which emphasized this "international" library Internet access. The UK, however, seemed too familiar: exciting though Oxford access appeared to be, the language used nearly in common by the Americans and the British made the consideration of Internet access between those two countries seem a little too obvious and tame. France was among the non-anglophone leaders in Internet development -- it still is -- and it had superb libraries, interesting innovations of its own such as the Minitel, and very similar questions to ask of the new digital techniques.

And there was nothing obvious or tame about Internet usage in France, neither in the explosive popularity of the new techniques there, nor in the great difficulties encountered in trying to do all of it in two languages. For me, then, France seemed the fascinating place to study, both for its sophistication and for the fact that it so much better represented all the other countries of the entire non-anglophone globe, which soon would be facing the Internet onslaught, than any nation which already used English. And the fascination continues, as the Chinese, and the Japanese, and the Indonesians now take up so many of the same issues which originally faced the French.

So what I propose here is a consideration of what it means to be a "foreigner", in using these digital techniques. Here, specifically, I am a US-based foreigner trying to find and use Internet library resources which happen to be located physically in France. But the procedure could be reversed: I could be a French user trying to handle an opac offered online by some library in Kansas. Or I could be someone in Japan trying to use either, or someone in the US or France trying to figure out what the National Library of China's excellent online site is trying to say. In each situation, much has been improved since 1988, but also many of the old issues remain, and are as distressingly unresolved and even unacknowledged as they were back then.


I. Finding France -- the view from the outside

How does someone who is outside of the country -- a foreigner -- find and use, online on the Internet, a library which happens to be in France? Is this process easy? Is it useful? How can it be improved?

The "search engines" are the first thought -- the Internet "portals" -- Yahoo, Altavista , Nomade, Voila, Wanadoo, Google... The great challenge for any library -- in France or elsewhere -- wanting to be "found" easily by users who rely on Internet search engines is to get listed, and stay listed, with the public search engine services.

The largest of these, run by firms which rely primarily on advertising and sales revenue for their incomes, include a website automatically using "knowbots" and "spiders" -- software routines which periodically scour major portions of the Internet and index the sites find. Tricks for making a site more visible to these people include the use of appropriate metadata (see below), minimizing the layers in a site because some services look through only the first two or three, and including the right keywords frequently -- such as "library" and "digital" and "France" -- on higher level pages in prominent places such as titles.

Other very large search services, which rely for revenue on selling their mailing lists to others, or which are restricted for some other reason, require registration of a site before it will be shown online and searched -- Yahoo, for example, does this in part, employing teams of evaluators who list a site only if it conforms to certain policies of the firm, although Yahoo supplements this by offering a general Web search feature as well. Growing numbers of professional and otherwise registricted search engines require registration. A site wishing to be listed must file an online application, and with some search engines renewing these applications has to become a regular task -- search engine registration services exist, online, which will perform this periodic renewal function across many different search engines for a single site registration.

Finally there are, inevitably, generally-smaller commercial search engine sites online -- again in increasing numbers -- which will register a site for a fee. For whatever added value the specific search engine service might bring -- appeal to a particular profession or commercial sector, features not offered by other search engines, attractive "portal" interface design, and so on -- payment of such a fee may or may not be worthwhile for a library.

The point, however, is that little of this is automatic. Increasingly it will be insufficient merely to set up a Website and wait for users to come to it: as the Internet continues to grow, any site which fails at least to pay attention to its own "search engine presence" -- how it looks to users who try to find it on the search engines -- runs the risk of being ignored, and having the increasingly-considerable investment involved in designing and maintaining Internet access lead only to a few mentions of their site begin lost somewhere deep in the pile of "3,830 retrievals" on Google, or "152,725 results" on Altavista. And per the oldest adage of le marketing: "the tree in the forest that falls unheard does not fall".

There is a fundamental and rapidly growing problem, however, with the "search engines. All of them exist, reputedly, to help users find things on the Internet. But do they do this? Do they find libraries, in this French example just from a search on "bibliothèque"? A random check: made in this case on March 16, 2002 at 6am San Francisco time... these things change rapidly, and constantly...

WWW.Wanadoo.fr"343 réponses"
altavista.com"152,725 results"
fr.Altavista.com"77,055 pages France"
"152,725 pages tout le Web"
nomade"200 sites francophones"
"145,000 pages web mondial"
voila.fr"86 web francophone"
"625 web mondial"
yahoo.fr"44 categories and 431 sites"
yahoo.com"33 sites"
google.com"3,830 retrievals"

What is the meaning of these numbers, and what is their meaning to a user? To a librarian, bibliometrician, or Internet systems retrieval analyst, such statistics are difficult enough to interpret. To a library user they are meaningless. And yet this is how most foreign users currently make their first encounter with French libraries online: through the "information overloaded" Internet.

There are, in addition, a number of conceptual gaps which any user confronted with this overload and its immense variety then must cross:

Until Unicode solves all such problems -- and gets fully implemented, worldwide, and all legacy systems and data not using it then get replaced -- foreign users and systems will have trouble with a word like "bibliothèque".

None of the various formulations of the Internet's and digitization's generally anglo - centric / americo-centric systems, for solving the problems of international spelling -- not "ASCII", nor "extended ASCII", nor "Microsoft ASCII" or "ISO Latin 1 / ISO 8859-1, 2, 3, 4, 5, 6, 7, 8, 9 (Latin 5), 10 (Latin 6)" or any other standardization attempt -- yet has gained universal commercial acceptance, much less solved the intractable and perhaps permanent problem of "legacy systems and data and users".

Still, and perhaps forever now that so many digital systems already are in place, users will come up with new or old methods of entering "h" into their terminals in such a way that current systems cannot recognize them. Whether this is some strange combination of computer keyboard key strokes, or a mouse click invoking same, or a "copy and paste" function from some strange encoding made long ago or very eccentrically, sooner or later the user always finds some way to obtain the dreaded null set retrieval, or "HTTP Error 403 / 404 page forbidden / page not found" error.

And if "h" sounds like a simple-enough problem, containing mathematically so few possibilities that server systems should be able to anticipate them all, consider that the French language is being used here simply as the tip of the iceberg: how many other possibilities are there -- for user eccentricity and legacy problems -- among all the other human languages of the world all now busily getting digitized and being mounted online? The Unicode effort --


-- is the best bet for making a beginning at overcoming this for new systems, but user eccentricity and legacy issues are library user problems far beyond Unicode's scope and abilities.

Systems people will deny this. But librarians know the truth and importance of it well. In library service "the customer is king", and library "customers" always are a little eccentric, and sometimes difficult: that is why they have come in to the library, after all -- the "easy answer" was not out there. So one initial problem, for finding "bibliothèques" using Internet search engines, is that the users won't all spell things the same way, and that the information systems and their designers -- so far certainly, and perhaps forever -- cannot anticipate this user variety in advance.

The logic of "search and retrieval" creates a second initial barrier for users seeking libraries now -- perhaps the most well-known of the many problems deemed "intuitive", by logicians and mathematicians, but frustratingly non-obvious to normal users. Intuitive though it might be, the concept of adding "Limoges" to a search on "bibliothèque" -- in order to "limit the search using a boolean 'and'" -- seems counter-intuitive or at least very frustrating, to many users.

User studies, and the testimony of nearly any experienced reference desk librarian, plus the presence of detailed and much-ignored printed "searching instructions" placards next to all opac terminals, all will attest to the difficulties which many digital information seekers have in extending search criteria beyond an initial term and default parameters: "bibliothèque" in "simplesearch" too often is as far as they go -- and as far as they get -- faced with all those thousands and now even hundreds of thousands of retrievals, they do what library users have done for centuries and simply pick the one on the top of the pile to look at, if that.

Again, systems people have waged brave battles to extend the sophistication of the user's search: multiple criteria, date and other limitations, nested retrieval and stored search possibilities -- the very imaginative list goes on and on. The most successful strategy has been to bury the logic within the system itself: complex search algorithms, building in all sorts of assumptions -- some valid, a number very often not -- about what the user "really" will deem to be "relevant" govern the most successful search and retrieval procedures, such as that of Google.

The systems find and rank things by relevance for the user, rather than forcing the user to do it -- so that when the user picks something "off the top" it has a better chance of being useful. The most successful commercial systems, such as Yahoo's and Amazon's, let the user decide which of a few criteria to use in this ranking, in ascending or descending order -- for example, Amazon's criteria for search and retrieval of its printed and digital books for sale,

or Yahoo's for stock investment research,

-- the user is able to sort a limited retrieval, using such criteria, enabling the famous "what-if scenario" which proved so useful to establishing the personal computer's role in modern life in the first place. Good library opacs can do the same. But the general point is that the most successful search engine systems -- Google, currently -- do the searching and retrieving behind the scenes, instead of forcing the user to do it.

So, pre-coordinate and post-coordinate indexing for search and retrieval... such great tools, if only the users would use them... a modern gloss on the very old problem that, "you can lead a horse to water but you can't make it drink". In library terms, exemplified by that most famous of library reference questions, "do you have a book about horses?": so few reference questions ever have come in to the library which have not needed at least some filtering and refinement -- even the famous "known item" search, where the user is so certain of the exact title and the spelling of the author's name, and the precise publication date... only to find that the librarian uncovers variant subtitles, erroneous spellings, and multiple editions...

Libraries in France and elsewhere have experimented with great varieties of techniques, for refining user inquiries prior to their submission to the search engine, and for sorting them once they are retrieved. Both forms of indexing need constant study and renewal, because of the very rapid development of both the users' own hardware and software and the users' own skills

Rapidly increasing personal computer capacities -- typical processing speeds in the gigahertz ranges now, RAM in the 100s of megabytes, vast amounts of memory -- and sophisticated and inexpensive new "off the shelf" software for information retrieval and databases, make this a different world from that for which most current library online public access catalogs were designed. Users now have more data manipulation capacity in their laptops than many libraries had in their main systems, back when user access first was being thought about. So some re-thinking is needed now -- and will be again, as users' capacities increase further in the future -- the users can do far more with the data than simply read it on a screen, now.

The users themselves, furthermore, are far more sophisticated, and they expect far more from the data and from the library. A student who just has spent three hours downloading and uploading vast numbers of complex music clips to the Internet -- having run all of these through her local personal computer sound software, modulating things and changing formats to suit her MP3 player and resorting her data endlessly into playlists of various types -- understands far more about the library opac's underlying structure and capacities than is suggested by the simplistic screens and command line ASCII output now offered to her by most of them. Give her the data in sufficient amounts, and she can work wonders with it herself, now; but like any library user throughout history, first she needs the data -- she now has the intelligence, and the tools, but too many libraries still provide only the barest minimum.

And then the "bibliothèque" found might turn out to be located not in France but in Canada...

The problem of "false drops" plagues the user seemingly at every step. Anticipating the variety of indexation is the Great Game of information retrieval: figuring out, in advance, where the user will look -- whether something called "industrial secret" will be searched for in "copyright" or "patent" or "intellectual property" -- always has been the greatest search - and - retrieval challenge.

So much depends upon the characteristics of the individual user. Is this a little 7 year old wanting to learn to ride a horse, or a captain of industry wishing to drop $100,000 into a personal stock market investment just before lunch, or a professional attorney looking for a "hook" on which to hang a lawsuit, or a starving author looking for an "exit" through which to escape that same lawsuit? The librarian's oldest and most valuable professional expertise is not in designing systems or supervising the search and retrieval within them, but in characterizing the users, and anticipating and representing them in the design process.

So when a user comes to the reference desk and asks for "French libraries", the professional librarian should know to respond by asking, at least, "libraries in France or libraries which contain French materials?". But a simplesearch single - term information system -- the type which users use, if not the type which they are offered -- returns vast numbers of "bibliothèque" entries for excellent Canadian resources, useful if that is what you were looking for but not so useful if you wanted libraries which are in France.

The "false drops" problem then goes further. The multiplicity of semantic and syntactic vagaries which may be obtained on a modern search engine boolean exercise is formidable. Currently, a search on "bibliothèques" in even the best of them -- Google -- finds, inter much alia,

among many Canadian and other "irrelevant" entries, all highly-ranked for "relevance" by the search algorithm. Even the "boolean" search for "bibliothèques (and) Limoges" currently retrieves, inter - nearly - as - much - alia,

as the very first and presumably most "relevant" entry... little wonder that the users get frustrated...

A library concerned about the "false drops" problem -- the "relevance calculation", another of the oldest problems of librarianship -- might do well to back away from recent "digital" and "hitech" and "artificial intelligence" solutions which have been proposed, and examine some of the older thinking.

Now that the users have so much more digital information sophistication than they had before -- and such better hardware and software and systems tools for their personal use -- much of the burden of performing the "relevance calculation", of deciding exactly which among the vast retrievals obtained by the search engine's algorithms are "most" relevant and "less" relevant to her inquiry, might in fact be shifted to the user. The "what-if scenario" was the chief contribution of personal computing in its earliest business application years: the ability, provided through "spreadsheet" software, of any user to sort, and re-sort, vast amounts of data to see how each alternative might or might not affect a given result -- "what if we did it this way...", "what if we did it that way...".

There may be no one answer to a "relevance" problem, in other words, there may be several -- a "wicked" epistemology problem, notoriously difficult for automated systems to perform, but everyday "judgment" work for any human mind. So if the library systems begin giving the newly-sophisticated users ranges of alternatives containing sufficient "tags", the users on their own systems can sort, and re-sort, those tags to obtain results of greater relevance to their personal inquiry.

This might extend even to Internet search engine retrieval of a "bibliothèque in France" address: library web page, and metadata tags (see below), perhaps should be designed more with the data manipulation capacities of the new generation of users equipped with their far more capable personal information systems, than they have been in the past.

And then there is the great Internet promise and problem of "layering": the multitudes of hierarchical levels which have grown up, separating the seeker from the thing sought -- in this communications medium which was supposed to have flattened all the hierarchies, promoted direct access and democracy, and made information "hypertextual" and "free". (see "change", below).

In this information retrieval case, the layers are the lists within lists, which sometimes are lists within lists themselves: retrievals of hundreds of thousands of Internet resources, many of which in fact are nothing more than lists of thousands of Internet resources, many of which are... ad infinitum.

The path even can lead the user back to the beginning -- a Yahoo search can reference a website which references an email discussion which references the original Yahoo search -- a wonderful circuitry for spreading rumor, erroneous information, outdated addresses, and so on.

No better illustration of the Borgesian "bibliothèque de Babel" nature of the Internet exists than its "lists within list" characteristic.

A library does well to focus its own Internet presence and web page design on conveying its own authoritative presence to the user. This emphatically extends to the use of an appropriate "domain name" which the user will see on obtaining a retrieval: increasingly, among the entries which she will find listed online, will be great numbers of secondary and tertiary etc. resources which lead to anything but the item she sought -- to use an example from the highly-commercialized hotel industry, currently a Google search for the "Bayerischer Hof hotel in Munich" obtains the hotel's website,

but that is buried, one among very many, well down in a very long list of things like,

-- all of these being commercial hotel reservation and travel agent services, and city guides, and tour operators, and municipal websites, and many other things, rather than the hotel itself -- secondary resources, some of which lead to each other and few of which in fact lead to the "known item" actually being sought in the search.

This is one of the oldest of information search and retrieval problems, well-known by any indexer in the library or documentation fields, but libraries must address it better than they have -- by more frequent use and better placement of keyword terms in Web page design, by establishing and constantly maintaining Internet search engine registrations, and so on, if users are to be able to find them online amid the growing cacaphony. At the very least, a library "URL" -- still the thing most likely to be scribbled down by hand on a piece of paper, or perhaps "copied and pasted", by any of the new generation of more sophisticated Internet users -- should read as something better, more authoritative-sounding, than

-- much attention needs to be devoted to the online appearance of that address.

The Protean nature of the Internet frustrates a user's search for French libraries on it even more than all these other problems do, moreover. Even making allowances for this being digital information's Age of Incunabula -- a time when practices such as title pages and colophons and pagination, even standardized spelling and punctuation, all are in flux or yet - to - be - invented -- there seems to be an impermanence built into the very structure of the information systems now offered to users for finding things.

Not only do hardware and software change very rapidly now -- in the latter case faster than every 6 months now, according to admittedly self-serving statements by leading Silicon Valley software vendors -- so that both the user and the library must race to keep up, but even within the fundamental telecommunications structure of the Internet change appears also to be endemic.

Two leading examples, both of them formidable obstacles to the effective use by users -- or anyone -- in finding online information, are the URL and the "path".

The ease with which institutions change a valuable URL / Uniform Resource Locator (now also called URI / Uniform Resource Identifier, in the endless propagation of Internet acronyms) --

-- may illustrate the extent to which said institutions get out of step with their own users.

The transfer of a database or service from something pedestrian like "frmop11.ac.fr" to something more exciting such as "Oulipo.fr" may make great superficial marketing sense to some, in an organization. But at least two problems are involved:

i) if the underlying IP / Internet Protocol number is changed, the users and the rest of the Internet as well -- all those links -- will not find it. A "forwarding" address page,

must be inserted, and maintained -- for several years, in many cases, as Internet links from others to the site may not get reviewed and updated more often than that. So often -- too often -- systems which make such a change find that their old host, who had promised to maintain the "forwarding" function, has reneged on the promise, or gone out of business, or that something else has happened, and suddenly an entire swathe of users gets cut off.

ii) And even if the underlying IP number has been maintained, moreover, with only the domain names having changed -- still "", for example, only now no longer "frmop11.ac.fr" but "Oulipo.fr" -- systems operators will assure you that this will "take care of your problem, don't worry"... -- there still is the "legacy user" difficulty, which is that although some users "cut and paste" IP number-based addresses into their online links and other texts, most users still write, with a pen or pencil, "frmop11.ac.fr" down onto lots of pieces of paper, all of which suddenly are rendered useless for information search and retrieval purposes by the "safe" domain name upgrade. The number of systems which have tried this... most of them enormous old university and public agency systems, very advanced in other respects, who are trying to graduate from old-style engineering - and - engineer - driven labeling to "le marketing"...

Even worse than a change of URL / URI, if only because more insidious and far more difficult to control internally in any institution, is a change of "path". High level policy makers normally become involved nowadays -- are permitted to become involved -- in something considered so fundamental to institutional image as a change in domain name: if the change from "frmop11.ac.fr" to "Oulipo.fr" sounds logical and not particularly controversial, consider what a change from "bnf.fr" to "Oulipo.fr" might entail...

But within the information system -- in-house, protected by layer upon layer of professional expertise and administrative structure and "turf wars" considerations, the way in which the information department structures its memory banks is guarded, very jealously, by the computer engineers: non-engineers, and particularly high-level decision makers, enter very much at their own peril, and no-one else simply is allowed in. So if a new system is purchased, or even an old one reconfigured, suddenly there are a couple of new directory layers added -- named, for example, "www" and "htdocs" -- and the old URL / URI,

becomes instead,

What this means to the computer engineers -- sysops, system administrators, and so on -- may be far greater convenience, many hours of expensive labor saved, perhaps merely conformity to a new vendor spec but at least that in itself is something.

But what it means to the library user tentatively searching online for "bibliothèque" is,

-- the greatly-dreaded "404 Not Found now you are lost" message... (there was no "Room 404" at CERN, when Tim Berners-Lee was there... ) -- the typical Internet user instantly will give up and "click" elsewhere...

Changing and, particularly, deepening the "layers" of the address also can place a site out of reach of search engines simply because so many of them scan only the home pages and top - level pages of a site: so that the address change, and two new levels of hierarchy, may be sufficient to knock a site "beneath the radar" of the entire Internet, suddenly -- and something which is "beneath the radar" does not "show up on the screen"...

And, again, the systems department's own internal effort to control this difficulty -- through automatic "forwarding" tricks of their own -- addresses only a part of the problem. The users still will have written the old path, sans the "/www/htdocs" addition, down somewhere on a piece of paper -- or, worse, have included it in a formal printed text, or much worse might even have published it in print somewhere -- they, or some reader or reviewing committee member, will have hand-typed the entire string into their computer and received the above "404", and nobody will thank you or your institution for your increased internal efficiency.

So change in and of itself is bad -- in this most changeable of all media, ironically -- pace all the Internet industry and research efforts to formulate "permanent" or "persistent" URLs, until they can guarantee that users will not write an Internet address down on a paper lunch bag, using a pencil, any change in a URL or URI or IP number, even and perhaps above all a "path", needs to be stubbornly and effectively detected and opposed, by any librarian or other professional representing users' interests in information work.

Any library, anywhere, needs an internal and very formal "review and update" procedure for its online Internet presence: someone / some group simply must look at and try to use -- "standing in the shoes of the user" / as though he or she is a "typical" user -- literally everything offered online by the library, on a very regular calendared basis.

Once every year, once every three months, once every month -- every Tuesday -- all this needs to be reviewed and updated. The hitech industries themselves boast that they now change everything which they are doing more often than once every six months -- if change is that rapid for them, a system can count on even more rapid change among the users and the various resources to which it is trying to link online.

Some decision-maker at the Ministry of Culture will decide to shift something from "livre et lecture" to "multimidia" -- for political reasons, or any other -- and suddenly several of the links on the library's own pages no longer work... Or -- even more likely -- the library's own well-intentioned Web designer will make a "tiny little change" which to her makes no real difference but suddenly means that users' default fonts become too small for older people to read (she is 23 years old, herself)... Or -- even more likely than that -- M. Gates and M. Case and M. McNeely will wage yet another iteration of their ongoing market domination war, and suddenly all of the oldest versions of some browser or another no longer will work with something...

The library's "review and update" committee needs to check, regularly, everything which the users might be using: every browser version, every cybercafi and home and Nokia flip-phone connection, every font size possibility and default setting, certainly every means in use of depicting accented French characters as these change so often, and all the latest plus the older means of manipulating library-generated data which might be downloaded to the users' own systems -- not so difficult a task, if performed regularly by people who enjoy doing it, and one without which the library systems simply cannot function, in an atmosphere of such radical and continual change.

The caveat regarding change extends, unfortunately, even to art --

French literature, perhaps more than any except possibly the Chinese, places great value on the indirect, the double entendre, the mere suggestion and the "thing left out", in the naming of things -- "Lettres Anglaises", "Surveiller et punir", "Apologie de Raimond Sebond" -- so often, French text titles contain no hint whatsoever, or sometimes even a misleading one, as to their contents. What then happens, to the digital information system's search and retrieval function, when such a deliberately obscure and even evasive "title" is offered as a sole access point?

A search on "surveiller et punir" obtains a fairly precise retrieval, online -- turning up mostly references only to Foucault's work, at least in the higher relevance rankings.

But the terms' translation into English, as "discipline and punish", finds more: Foucault references do show up at the top, but very quickly also a user has to face,

Proximity connectors and quotation limiters and other search and retrieval tricks would help, of course: "discipline within 5 punish", or "discipline and punish", or "case - sensitive Discipline and Punish". But how many users standing at a public terminal or dialing in from home have the knowledge or perhaps more the patience to employ these? Instead, the few top retrievals are picked, and the search goes elsewhere.

And "lettres anglaises" retrieves, inter much alia as well,

Again, boolean searching and proximity connectors and all sorts of sophisticated pre - coordinate and post-coordinate and limiting tricks and devices come to mind -- but so does the tendency of users simply to "click" away to the next search...

The artful naming of a text -- "Lettres Anglaises", "Surveiller et punir", "Apologie de Raimond Sebond" -- so as to stand out on a bookshelf among other titles, or so as not to -- so as to be obvious but just as often to be obscure -- runs a real risk of undergoing profound transformation, now, under the simplifications, and new sophistries, of digitally-based information search and retrieval.

If the search engines are looking for a thing in the titles, include that and it will be found -- the term "Internet", for example, or "digital library", to cite only one other current buzzword -- and if the search engines are trying to avoid something else, such as "sex", avoid including that in the title. There is a search engine tyranny at work here, perhaps -- perhaps nothing more than the old "le marketing" at work, perhaps something new. The rebels -- poetry magazines named "ZYZZYVA" (it exists! -- see ISSN: 8756-5633), for example -- will become notorious, or will be ignored.

Libraries may need to compromise, here. The development of metadata (see below) and other indexing tricks can do only so much to alleviate this "artful naming" problem.

But even if internal systems solutions like these are developed and adopted, the "legacy systems, data, user problem" remains -- the irritating but inevitable tendency of many users to download, printout, or even simply scrawl in pencil on a piece of paper, text as it appears to them on a screen, thus escaping all of the internal programming sophistications designed to help them. This is one reason why the full URL itself needs to be shown in any important link, instead for example of the currently-ubiquitous "click here", which means nothing to the user once such text is downloaded, printed out, or scrawled on paper.

A library wishing to call itself "For the children", online -- or perhaps "L'Heure Joyeuse" -- has to consider whether an Internet audience inclined to download and print out and scrawl on paper, or simply to scan through enormous search engine retrievals without thinking too deeply, will be able to figure out that such a name in fact indicates a "library". For some audiences it might, for some it might not. Search engine tyranny online may dictate the wisdom of naming something "bibliothèque" or perhaps "médiathèque" instead, in the title, and only including "For the children" or "L'Heure Joyeuse" in a subtitle somewhere later on.

One innovative way around all of this, under a development so rapid as to have become a race -- between information retrieval specialists determined to channel the flood, and information providers, hackers, and various other groups just as determined to circumvent the controls -- is the metadata movement.

The outstanding libraryland example is the Dublin Core Metadata Initiative --


-- nickname, "The Dublin Core".

The basic metadata idea is to "embed" -- write in so as to be unreadable by users but reachable, and usable, to any online search engine and database software -- identification tags for "subject", "title", "author", "publisher", format", and so on, very similar to those used in bibliographic data and other meta-description, using a tag set agreed-upon by a group of objective and international developers. Search engines then can find the tags, more easily than they can scan entire documents and websites, and tag descriptors can be made to conform, hopefully, to pre-determined classification schemes which mean something more to information professionals than "lettres (and) anglaises" .

The effort has come very far, enough so as to have become terrifically complex. But so long as it gets fully implemented -- and it hasn't, yet -- it offers the best promise of the information professions that their expertise, and experience, will be able to ease the search and retrieval task of users at some future date.

So an author will be able to entitle a text whatever she wants, and perhaps even a user will be able to search for it using a multitude of search terms unlikely to be found now in the limited "Author" "Title" and "Subject" classifications currently available -- all sorts of meta-indexing can be done, linking likely user descriptors with standardized descriptive data, once the embedded metadata are in place on webpages.

But this has not happened yet. Already there is an enormous and rapidly growing corpus of non-standard -- in Dublin Core terms -- web pages, candidates for the "restrospective conversion" projects of the future. As with Unicode and "persistent URLs" and other international standardization efforts, unless and until Dublin Core is finalized and fully implemented, that corpus will continue to grow at the Internet's phenomenal expansion rates. The information overload flood is not over yet.

French libraries are active in the development of Dublin Core and other metadata efforts. But they need to become more active, as do the libraries of any culture which wants to see its language, and conceptual patterns, and cultural practices represented adequately in the rapidly-globalizing information search - and - retrieval structures being assembled now.

The ability to "un-do" any of this later on, if it becomes too biased in favor of one culture or another, or if it simply becomes too rigid, will be very much out of the hands of current standards developers: as we have seen, at least because of the "legacy systems and data and users" problem -- the tendency of users to download and print out and scrawl things in pencil on paper, thus "freezing" them into practice and acceptance -- if not for any other reason. Today's tentative, first-draft, temporary - subject - to - later - revision effort has a tendency, now that the Internet has exceeded critical mass in a number of respects, to become permanent in spite of its many caveats and qualifications. People wishing to influence developments must be "in at the beginning".

And, lest we forget, a great many people are not online yet -- including library resources, and libraries. In France alone -- using this particular country only as a leading and "best - practices" example, as we are throughout here -- great numbers of libraries still have no W3 address, and not even any email or telnet or Minitel access which is readily available to users -- and where these exist they very often are difficult to learn about, little effort having been made to publicize their existence beyond the local user area if that.

More expansion, then -- far more resources than currently already are choking users' information retrieval results on a search for something as simple, and as specialized, as "bibliothèques". When all libraries in France truly are online, and are listed, plus given the great quantity and sophistication of the online offering of a typical French library once the decision has been made, money found, conversion accomplished, and connection launched -- all of which only can increase -- the current flood will seem a trickle in retrospect. Imagine all of the numbers given earlier here,

WWW.Wanadoo.fr"343 réponses"
altavista.com"152,725 results"
fr.Altavista.com"377,055 pages 'France'"
"152,725 pages 'tout le Web'"

-- multiplied by factors of 100, or of 1,000, or perhaps more: with some energy devoted to the task by French libraries and librarians, including all of those still currently "unlisted" -- recon projects, digitization exhibits, interactive multimedia offerings mounted by all of those "médiathèques" -- today's information retrieval inundation (above) could become a true digital nightmare. Nobody ever will be able to find anything -- many of the users may not even bother.

One of the greatest contributions made so far by French libraries, to Internet development, has been in their role of reminding developers that the entire world does not in fact speak English.

The simple problem of representing "accents aigus" properly, in the original ASCII which excluded them, now has been extended to the depiction of other character sets for other languages. Other technical standards development -- MARC formats, telecommunications procedures such as OSI and tcp/ip, domain name assignments, intellectual property attitudes -- has benefitted greatly from at least the reminder, offered by la diffirence frangaise, that there is more than one way of doing a thing.

But differences in conceptual pattern can be cultural -- and certainly differences such as industrial development and antitrust and competitive and social welfare attitudes, which underlie the entire development of the hitech industry and its Internet, are cultural and political in every sense.

So the presence of a friendly critic is invaluable -- there are plenty of others, not yet really reached by digital techniques in many cases, who are not going to be so friendly. Both within the societies which already have adopted them, and outside -- among the many nations which still do not have enough to eat, much less the inclination or even ability to consider them -- digital information techniques can benefit from an open discussion of different approaches to solving the problem of the "information poor".

Social welfare is viewed very differently, these days in France, than it is viewed in the US, for example, as are questions of providing multilingual access for immigrants, and as are problems of assisting in the development of other nations overseas -- libraries in France, by continuing their efforts to show others on the Internet how they approach the problems of "the unlisted", greatly broaden and deepen the debate, among all of us, as to how such problems might be solved.


"Filters" has been the response of the library profession, and of many others: some device, some rules, some system, some person or group or authoritative agency -- to winnow out the digital harvest, separate the information "wheat" (bits - with - relevance) from the data "chaff" (bits - without - relevance), and serve us the former while consigning the latter to the functional oblivion of digital data storage.

No less a personnage than Umberto Eco has described the development and implementation of information filtering, for dealing with information overload:

And yet it was Umberto Eco who also has been terribly concerned about book preservation: acid paper has troubled him --

-- but so have the realities of book preservation,

So now comes "filtering", which may cut down on the information overload problem, but perhaps threatens even more to dictate "which books to retain", for Eco as for all of us. The promise is, as Eco points out, that better filtering will,

but the problem is, as Eco also points out,

What gets left off of the list, and why, and -- most important of all -- who decides?

There are plenty of candidates for the last group: Platonic philosopher-kings -- Eco himself seems to favor this, but then he would be a candidate, as Plato would have been... -- or how about librarians, or a committee of leading scientists, or generals, or leading corporate business giants, or whatever politician happens to be in power at the moment? All have been self-suggested.

We put our trust in those whom we have trusted before, perhaps: best in our local library, where increasingly information "filters" -- of precisely the type for which Eco and others have called -- are appearing online. Local librarians, in France as in the US and elsewhere, carefully and deliberately developing excellent little "recommended" lists -- what in name - conscious France are called by some, appropriately if cutely, "Sitothèques" (for some years, some in France have called a "books plus" library a "médiathèque", after all) -- little lists, which have a way of growing very large very fast, of "good Internet sites".

The implication is, as with similar "good books" lists assembled by generations of librarians in the past, that the Internet links thus supplied, while they may not have formal approval of anyone and might not have passed some particular set of structured criteria and certainly have not been "censored", nevertheless have been "vetted", as the English say -- reviewed by someone competent, who has weeded out the worst of the bunch and at least tried to find the best, following "worst" and "best" criteria which may reflect nothing more than things like middle class morality but might at least avoid the worst Internet excesses.

Such is filtering, in its current incarnation. It is not such a bad solution, really. It means that the local librarian -- with whatever intelligence and good faith and biases and education and class and religious and political background and professionalism s/he can bring to bear personally on the project -- has assembled a list of W3 sites deemed relevant to the searches of that local library's users.

In the case of the "sitothèque of Montfleury - sur - Loire" the result can be quaint or cute or curious -- a similar effort mounted the Bibliothèque nationale de France (see below) can be formidable, involving substantial amounts of effort and intellectual and other energy -- either way, the "sitothèque" of a local library is an information filter, one assembled by a human being who has kept the needs and capacities, of other human beings in mind. One starting point, for paring down the current and growing information overload of the online search engines, are the little lists of links offered nowadays by growing numbers of libraries on their W3 sites: as in the case of "do you have any books about horses", the users once again can trust the same people to answer "do you know any relevant websites".

Perhaps... there always are abuses... As both users and librarians in France know all too well, even libraries are not totally immune to the vagaries of politics and budgetary and other pressures. A few local politicians in France recently have made news by pressuring local librarians to conform their practices and collections to their particular political platform -- the scandals at Vitrolles, where the Front National mayor has leaned so heavily on the librarians, have outraged many in France.

Such pressures are exerted in the US as well, on occasion: the post-September 11 hysteria last year produced a Children's Internet Protection Act which threatens to censor far more than just children, in US libraries, also outraging the library community in the US.

These problems and abuses existed, however, long before the invention of the Internet and digital information. The problem of filtering -- for information overload purposes, at least, if not for its extensions to "pornography" and "Nazi memorabilia" and "terrorism" and other issues, as advocated by various groups on both sides of the Atlantic -- is not one easily answered by automated processes. "Spelling" problems are difficult and become nationalistic; "boolean" searching becomes unworkable; "false drops" become so numerous that it appears as though every one of the world's "bibliothèques" must be located in Québec; we all are waiting for "Unicode" and "metadata" developments, but these haven't fully developed yet...

The best thing to do -- as with "democracy", not the perfect solution but the best -- may be to trust our local librarian. A user searching for links to a "bibliothèque" in France could do worse than scan through the W3 links offered online by a local library, either in France or elsewhere -- given the information overload realities of the Internet search engines now, the user probably could not do much better with those.

A second source of "filters" for information search and retrieval is government. France has an excellent example, in the outstanding online offerings of its Ministry of Culture:


which includes, for specifically "bibliothèques"


Since nearly the beginning of the digital revolution, "culture.fr" has provided a leading and nearly comprehensive portal into all things Internet under way in France. This may be a matter of money: the Ministry has a lot of it, after all -- at least more of it than do regional or local government, most corporations, and all libraries -- so that the staff and equipment and time necessary to develop such enormous and complex online resources are relatively more available there.

It also is a matter of national pride, however, which in France counts for so much in cultural affairs: French internauts point to government information resources online with a pride very unlike the "minimal tolerance ranging toward barely disguised contempt" which citizens in other countries so often exhibit in speaking of their own national government Internet efforts.

It also is a matter of national difference, as well, though -- la différence française in this regard, perhaps -- that "minimal tolerance ranging toward barely disguised contempt" has been well-earned, in the case of many other national governments. Even in the US case, for example, where national government is trusted far more, and far more broadly, than in most, such trust never would be blind, least of all in the area of information "filtering". In the US, in fact, national government is perhaps the last entity which would be so-entrusted, a long national civil rights tradition of adversarial if healthy relations between "government" and other institutions getting firmly in the way. Nor would "filtering" be relegated to "the Church" as a trusted intermediary, in the US, as it might still in other countries -- in the US, in fact, competing private commercial firms might be considered the most trusted information "filtering" intermediary, as they never would be in France.

But national government does perform the "filtering" purpose in the French case well, as the Ministry of Culture's efforts illustrate. At the local level too, many mairies now have city and town W3 sites which offer the best first stop for reliable local online information -- including where to find the local "bibliothèque", online. At least it simply is good to remember that, just as the US "competing commercial firms" solution might not be the best for France, so the French "national government resource" approach might not be most suitable for the US -- and that neither solution might work in Mongolia, or in Kashmir, or in Mozambique...

Within most countries outside of the US, and even within the somewhat unique case of the US as well, the "national library" offers a third alternative resource -- in addition to library websites and government websites, and better than the information-overloaded online search engines -- for finding things online which have been "filtered". Most national libraries worldwide now have websites:

and most of these offer "sitothèques" now, among them the outstanding service of the BnF,

So, a list of places to find "bibliothèque" online, "vetted" and arranged and presented in a user-friendly format by professionals -- librarians -- trained in anticipating the interests and needs of users...


An outstanding problem of using any online "search engine" or "directory" for reference work, however -- a problem in any sort of reference work, for that matter, even in the print world which preceded the digital -- is providing intellectual access in addition to mere access. It is one of the fundamental corollaries of illiteracy work -- that it never is enough just to be able to read, but one also must be able to understand. Understanding, knowledge, wisdom -- all qualities so much more important than mere contact; the ability really to use the information found, more than just seeing it and noting its existence -- the dream of every teacher, and the oldest of problems that, "you can lead the horse to water but you can't make it drink".

To supplement the online reference guidance currently provided -- to do better, in some cases -- the Internet offers econferences / listservs: the online moderated discussion groups, sometimes numbering international memberships in the many thousands, to which a user can take a quest for "bibliothèque" and have it answered directly, fleshed out with analysis and comment and suggestions, discussed and debated, provided with context and history and bibliography and other resources -- an enormous online seminar, in the best cases, to which any question within the group's purview might be taken.

As with "do you have a book about horses?", most users' research questions need a little tailoring before they may be answered. Even the famous "known item searches" normally require some: which edition is desired? need it be in translation? are all of the spellings in the citation correct? if the volume is not in the collection how best might it be obtained elsewhere?... There always is a little room for reference work.

How inadequately performed, moreover, by the current crop of online search engines and directories: as though the complex psychology of the "reference interview" might be rendered by delivering a simple one-line answer -- any more than by shoving 152,725 books at the user and telling her to "take her pick" -- even having sorted through them perfunctorily and telling her that you think that "the ones at the top of the pile might be best"... no wonder she simply grabs the one at the top, and then runs...

Econferences provide one very good answer to the "online reference" problem. If the single known-item search and any other does not succeed, with the online search engines and the directories -- and even if it does -- a good econference, in the subject area of the inquiry, or of information specialists who might know how to find things, is an excellent place to pose a question online.

The format encourages multiple participation -- there are French history econferences with several thousand participants, and librarian econferences with over 10,000 -- better than even the single reference librarian no longer provided by so much in online searching, an online econference not only provides more than one, it can reach many more than that, and even encourages discussion and debate which can be very beneficial both to the participants and the user.

Such a service has not developed yet really. Most online econferences are not yet provided as a service to inquirers but instead, still, simply as a discussion forum among those who would answer such an inquiry -- historians getting together to discuss how they do their work, book collectors gathering to discuss theirs, librarians doing the same -- in such a setting an inquiry from an outsider, no matter how innocent and earnest, is considered an intrusion, like asking a doctor at a cocktail party for a diagnosis.

But the opportunity is there. Within econferences plenty of collegial information exchange takes place already -- among the historians, book collectors, librarians -- on the model of the physical professional conference, where "back room" and "corridor" and "social occasion" chat so often is more important than the formal conference sessions.

Already, too, an informal protocol enabling users to ask "personal" questions exists for many econferences: although addressing "the entire list" might be deemed inappropriate, a personal email sent solely to one subscriber who appears to be an expert in an area can be not only appropriate but highly productive -- "I saw on XXX that you are familiar with libraries in France, would you mind telling me how I can find out the email address of one in Limoges? And which one in Limoges might be best for answering the following question...."

The discussion which can result -- the interchange, give - and - take -- can be a valuable part of the provision of true "intellectual access" to the user. Just as student questions and observations can be turned over and over in classroom discussion -- pulled apart and reassembled -- to the benefit of the student asking as well as for the others present -- so the give - and - take of online econference discussion, with individual members or with the general list, can add substantially to the user's understanding of his own question.

There are thousands of online econferences now -- tens of thousands -- not all of them good, not all of them well-moderated, but some of them outstanding and ranging from very good to invaluable for their participants. It would be a small step from their current state of cooperative consensual talkshop to online reference information work: some may settle for existing informal arrangements -- others may insist on some more formal structuring, to insulate subscriber participants from an onslaught of outsider inquiries -- either way, the econference represents an outstanding opportunity for online reference service.

For my own interests in French librarianship, I myself use the excellent biblio-fr,

plus H-France and DigiLib and PACS-L and Exlibris,

-- for other interests, some related to France and some not, although various members in each of these latter places happen to be extremely knowledgeable about France as well.

In each case the general discussion, plus personal correspondence with individual subscribers, supplies precisely the sort of information overload "filtering" called for above by Umberto Eco -- conveniently via email, personally-tailored, a personalized "give - and - take" discussion far more productive in fact than the old one-shot individual "reference interview" with an harassed reference desk librarian ever was in the past.

Improvements could be made, to the use of econferences as a filtering device --

-- but even as they stand now, online Internet econferences are a leading candidate for the active digital information overload "filters" demanded by Eco and other users.

Libraries in France have been leaders in econference development so far. If this leadership is to continue, and if econferences are to expand their traditional "talkshop" role to encompass online reference service and other functions, French libraries need to be present and make their voices and sometimes-unique approaches heard in the development discussions -- without them, econference development will be poorer, and most likely restricted to the needs and perceptions of anglophone users, to the detriment of users non-anglophone and anglophone as well.


At least one outstanding problem of intellectual access remains, however, even if a user -- one seeking "bibliothèques", or whatever -- is able to reach search engines, or directories, or econferences. Multilingual access extends to far more than just language. Cultural differences represented by linguistic differences extend to underlying concepts, perceptions, biases both hidden and acknowledged, pervasive attitude differences. The latter all deserve attention -- and sometimes get it, in an adequate personal "reference interview", and perhaps even will obtain some attention in a good econference discussion -- but usually, so far, they don't.

A start would be made, as already mentioned, if the foreign language spellings and character depictions were correct, and even here things are only just beginning. The tortuous problems of language character sets -- the alphabet soup of ASCII, extended ASCII, microsoft ASCII, ISO 8859, Unicode, Big 5 Chinese and the rest, mentioned earlier here -- creates an initial intellectual access challenge which merely suggests challenges to come.

Two other problems related to multilingual access exist:

i) Translation is an imperfect artform at best, but acknowledged as such it can be an invaluable aid to intellectual access and understanding. The process need need be perfect -- cannot be -- the biography of the most famous English translator of classical Chinese poetry was suitably entitled "Madly Singing In the Mountains", his translation of an ancient line, in terms which only could have been uttered by an Edwardian Bloomsbury gentleman.

But translation has to be tried, at least, and it is done so rarely on the Internet still. The arrogant assumption that everyone else in fact speaks one's own language covers up, too often, insecurites about the appeal of one's own sites to members of other cultures, and perhaps sheer laziness as to the effort involved in mounting something multilingual.

It does require effort. But libraries in France -- many of which provide excellent multilingual sites online now -- can serve as demonstrations that the effort both can be integrated with the main website and is worthwhile.

ii) Any number of techniques exist now for accomplishing a good translation of a site. This can be the achievement of language students at the host site, in the case of a school or university. On-the-fly translation, such as that offered by much personal computer software -- and online at the google.com and babelfish.com sites already mentioned -- can provide occasionally-laughable but nevertheless very usable translated versions, ideal for an econference but useful even for an active website, at least as a home page utility offered to users in the latter case.

Most importantly, however, any site with any pretension to offering an accurately - translated version needs to rely on native speakers, ultimately -- at least to edit a final text, if not to undertake the entire translation process. The nuances of foreign language are too great, to use an imperfect process -- automated on the fly translation or even a student who is really good at it -- in what essentially is neither perfectible nor a process but an art form. Human expression has too many ranges of meaning: and the associations change with the daily news -- references to "Watergate" and "Red Square" need superintending by a native speaker who keeps up with the newspapers and with the possible sensitivites of users. In these days of the global reach of the Internet it should not be too difficult for a website to find some native speaker, somewhere, who can cast at least a casual eye over the translations on a website.


Another outstanding intellectual access problem, well known to librarians but not so much to others, is format access. The variety of MARC / MAchine Readable Cataloging formats in use, plus varieties of local practice in their implementation, have long trained library catalogers in the necessities of format integration, and international standards in bibliographic description.

Not so the users, for whom most of this has been invisible. But until very recently, user bibliographies have consisted of handwritten or typewritten entries, manually transcribed from paper catalog cards or library computer output -- the immensely complex standards and formatting efforts underlying this data was masked, the user seeing the output only.

Since the early 1990s, however, a two-pronged revolution in user access to bibliographic records has taken place, at both the user end and at the library end. "Bibliographic software", a consumer product tailored to the end-user's requirements and capacities, began appearing back then in the general commercial market. This enabled users both to build enormous personal databases of bibliographic records, and also to download these records directly -- in massive quantities now -- from library catalog databases. At the same time, libraries themselves have vastly improved the search and retrieval and output options of users unmediated interactions with their catalogs.

There has been a snowballing effect. Initially a user might have retrieved just the opac bibliographic records which fill a single computer screen, laboriously transcribing these by hand or keyboard from there onto paper, and then "paging down" to the next screen. Then email output was added, so that first individual citations could be sent from the library source to the user's personal email account, then several citations at once, then larger bundles and finally hundreds.

To date, this downloading has increased to thousands: users of bibliographic software easily can collect bibliographic records sufficient to constitute personal digital library catalogs which once major institutions would have been proud to have assembled, using email and ftp and Z39.50 and various other telecommunications channels.

The key to all this has been the formats -- the library's "Title" appearing where the user's bibliographic software expects it to be.

By the end of the decade, this user downloading activity had progressed to images and sound files and multimedia presentations -- all of it still very much dependent upon countless hours of format work "behind the scenes", on standards such as JPEG and MPEG and MP3 and many others. And now in the early years of the new millenium, ebooks have been added to the mix, so that now users search and retrieve and download entire "books" -- from commercial vendors like amazon.com and from university seminars and other online resources -- expecting and practically taking for granted that the product that they acquire thereby will come complete with the illustrations and various textual enhancements which they had with the printed book. The ebooks are inexpensive -- users can assemble enormous libraries of them now, as they once could assemble enormous library catalogs of downloaded bibliographic records, to fill up the cavernous data storage disks with which personal computers nowadays come equipped.

Format integration and standardization has been one of the greatest success stories of the digital revolution so far. It has not been an easy process. Early battles between OSI and tcp/ip, and Microsoft's DOS and its competitors, and among the various MARC formats, all were extensive and sometimes bitter -- but the result so far has been the explosive and revolutionary flood of online information which has brought the library's catalog, and the photographer's images, and the recording studio's music, and now even the publisher's "books" -- title page and illustration and layout and colophon and pagination and all -- to the user's personal desktop. Next come the cinema industry's "movies" -- not so far away, as already the "news" end of that now-unified industry purveys its product in streaming video clips on that same desktop -- all of this unimagined and unimagineable to most of us a decade ago when it all began.

As with econference development, format development has benefitted greatly in the past from French library participation. If it is to expand in the future -- further to images and sound and multimedia, and even to cinema and virtual reality and whatever else might come along -- international participation will be crucial. The myopia of any culture would be a handicap, in this sort of development: at least because of the delay and resistance it would encounter in implementation, and certainly if implementation might have deleterious results which could have been avoided by a little consultation and cooperation in the development. So it is very much to be hoped that, when "Infotainment" and whatever else really hits all of us, online, that it will have had French as well as other international and trans-national input.


More intractable than the problems of formats and standards, however, have been the difficulties of obtaining the ultimate product. One of the greatest frustrations of online library catalog access, in the early years, was the user's inability to get the books: from San Francisco, all those bibliographic citations in the Oxford libraries were nothing more than a tempting frustration -- invaluable for specialized bibliographers, but merely an indication of local inadequacies for a user who merely wanted to read the strange books thereby found.

The online connection to France only made things worse: once the enormous treasure of the Bibliothèque nationale was revealed so easily on the screen by Opale, local inadequacies came to seem a gaping hole -- if an anglophone country's library collections were inadequate in their provision of English imprints, how much moreso they lacked in publications from France.

And, following the general argument here, France was an indicator of even more exciting, and far more frustrating, possibilities: if the BN had so much, what more might reside in collections in Germany, and Italy, and Spain, and Russia -- what might a scholar interested in Japan, and China, unearth once these records were online and so easily available.

Inter-Library Lending has tried bravely to "scale up" to this demand for "final product", accelerated so violently since the early 1990s by online access to cataloging records, as has Document Delivery. Both information services, constructed originally for an era in which catalog records were not even easily accessible by users in print, much less online -- in the earliest years of the British Library Document Supply Centre at Boston Spa, the early 1960s --


-- library users more typically would put inquiries through a librarian than ever wade into massive printed bibliographic reference works or even the typed-out collection catalog cards themselves, both of which then still were part of the mysteries of the professional librarian's trade.

Now, thanks to online digital information and much brave effort in format integration and standardization, all this has landed directly in the user's lap, with no user-friendly librarian present to interpret. So both ILL and physical document delivery services have experienced a nearly unanswerable surge in demand for their services -- requests from California, for "your book about horses", to little local libraries in / near Limoges brave enough to have mounted their catalog records online.

A partial response to the problem has been offered by remote "on-demand printing" solutions, such as that enabled by the Xerox firm's Docutech machines: they receive text electronically, and print it as needed, so that a text resource located Limoges may be used for the printing of texts in Shanghai -- as needed and only as needed in the latter location, without all the intervening trouble and expense of producing and shipping entire stocks of texts.

This is further improved and publisher's solutions, such as that of the McGraw-Hill firm's Primis series, which permit the publication of separate chapters of individual texts -- for use in education and in business, where assembled "readers" consisting of articles and chapters from a variety of sources so often are more useful than entire texts -- "Now you can have EXACTLY what you want for your students", the firm asserts.

And an even more practical response to this "end resource" problem has been provided by online digital document delivery. Better than packaging up the book and shipping it, and allocating labor power to photocopying and faxing for answering the waves of overseas inquiries, digitization of the collection's holdings gives libraries the ability to provide "content" to the users directly and easily -- to permit the users to retrieve that content themselves.

Libraries and publishers in France have held somewhat aloof from the fundamental changes which have been sweeping book distribution in the international industry. Unique French characteristics such as the loi Lang, and the general treatment of text as a cultural rather than a commercial commodity, have prevented or removed the incentive for the free-form restructuring of these industries which recently has taken place -- French attitudes toward antitrust competition and intellectual property are different, from those current in the UK and US and many other places now, and they have sheltered industries in France from the changes to some extent, to some extent exposed them.

As so they should be, perhaps -- different, that is. Perhaps there is enough room here for differences in approach. Perhaps there is a need, elsewhere in the international world than in the US and UK -- or perhaps more appropriately, elsewhere in the world than in the few Global Cities which really so far are the primary beneficiaries of online and other digital information , for something other than the high volume mass distribution model currently under implementation in the few major urban areas really using the new Amazon and Barnesandnoble.com and La FNAC online services. The attractions of carrying slim inventory are severe: Amazon now advertises that it is able to generate its phenomenal revenues from a working inventory representing only 5% of its previous 12 months' sales, a figure astounding enough to set most commercial mouths watering in France or anywhere -- librarians who in some cases carry hundreds of years of inventory by comparison to their usage statistics can only wonder at such efficiency.

But great inventory turnover, and efficiency, are not everything -- as more and more commercial firms get seduced by the Amazon model, perhaps there is a need for others, in commercial book publishing and distribution, and perhaps there is a need for models altogether non-commerical, in the publishing and distribution of digital texts.

Consideration in this case of the French difference might cause, if not necessarily the acceptance of the current French model for use in either France or other cases, at least the acceptance of the idea that various models might work in various situations -- that there is no "one size" which "fits all, in book publishing and distribution, inter-library lending, document delivery, intellectual property, cultural protectionism and propagation, any more than there is in shoes.


Virtual resources are a partial answer, but they pose a difficult problem for librarianship: that is, will the activity of providing online distance reference service and virtual materials throw a wedge into reference service itself, dividing the virtual from the real.

As with non-digital "documents" so with digital, to some extent: publication, distribution, intellectual property and payments considerations, even cultural property, storage, and search - and - retrieval all can be very similar, as between a "book" or "printed text" and something digital -- digital "document delivery" resembles the older printed "document delivery" in many ways.

From the traditional library point of view, the greatest difference between the older "book and document delivery" and the newer "digital" -- at least the greatest difference to have emerged clearly, so far -- is ease of access for the user. Even costs may shown to be similar -- information infrastructure and retraining / rehiring are considerable expenses, although rarely attributed directly any longer to information services provision, but new computers and hitech personnel do cost a lot, at least as much as books.

The information access provided by the Internet once the infrastructure and training are established, however, unquestionably are greater than any older availability of information to local library users -- people living nearby can reach and use information much more easily online now, and people in Kansas trying to reach Limoges, or the reverse, could do so with only the greatest difficulty and expense before. Virtual resources certainly outdo physical resources for ease of access.

Perhaps the greatest question facing libraries in France or anywhere else institutionally, however, stems precisely from this enormous difference in ease of access. To what extent, one wonders, is this great difference -- between the great ease of access to online digital information, and now the relatively much greater difficulty in reaching and using the older printed materials -- driving a wedge between the one and the other, the old and the new, into the heart of librarianship separating those most concerned with preserving the old materials before they decay from those more concerned with purveying access to the new digital things as widely and inexpensively as possible.

If such a split occurs it would do so very much at the expense of both: older printed materials need dissemination as much as newer digital ones do, and the new digital formats turn out to have a preservation problem in fact much worse than any of the acid paper and other depredations which plague the older printed collections -- each side does continue to need the other. But that a split already in fact is occurring is indicated by several alarming developments: among these the atrophy of professional library schools resulting in their demise or rechristening as schools of computer science, shrinkages in the pool of library profession applicants , and bitter controversies resulting too often in a veritable dialogue de sourds over the digital preservation of printed materials and their mistaken destruction in the process of "saving" them.

Print is print and digital is digital, it seems, in the minds of too many information advocates nowadays, sight being lost of the "information" contained in all of the excitement over the container.

The French term "support" is instructive, for these debates: the idea, translated into English, that communications media actually "support" the texts which they bear conveys exactly the notion of sustenance and enhancement and creation of a unique experience of appreciating that particular text which the far more neutral term "media", used in English, fails to represent -- words on a screen and the same words on a Pleiade edition's page can be the same, or they can be different, depending primarily on the intentions of the users in choosing the one medium or the other -- in a perfect world, both media would be supplied, and the user allowed to choose.

Libraries in France lead, in both preservation and access. They also take attitudes very different, from those employed by other leaders in for example the US and UK, towards questions of the purposes of preservation and of access, as they do towards cultural patrimony, and the art of reliure, and so many other things -- including basic terminology, such as the concept of "support".

Their presence online -- the encouragement to "think different", offered by such resources as the BnF's "Gallica" and many other projects -- is a constant challenge and inspiration to those outside of France who believe, or may fear, that the one true answer to all questions regarding information sources and their use may already have been find. La différence française -- surrogate, in fact, for differences everywhere.


From the commercial world, a lesson: customer service is an important thing -- commercial customers are less interested, ultimately, in the product being sold to them than they are in the way that it is sold, and in the service provided during and after the sale, as they can go elsewhere to get the product after all so it is the service really which keeps them coming back.

From the point of view of the commercial firm, this emphasis on service makes a great deal of sense. You can charge a great deal more for it, for one thing: the operating margin in services -- the difference between the price charged and the cost to the firm of providing it -- is far greater than the margin in products. This is a measure, perhaps, of the customer's need -- they value the service more. It also may be a measure of competition -- they can get the product elsewhere, perhaps, but not the service. But service also is an intangible -- composed of many difficult - to - define factors, subject to endless variation -- so there is room within customer service for designing something unique, which in a commercial setting can command a unique price.

In the commercial globalization process now under way, nearly all firms are becoming "service" companies as quickly as they can: realizing that the global economy now has made this transition from "products" to "services" -- and realizing that they can charge more money per unit for the latter and still sell as many units, thanks to the new global information technology, which is perhaps more to the commercial corporate point. From IBM, which makes more margin on its service contract than it does on its computers, to the corner grocery store which

The analogy to the commercial firm should not be taken too far with libraries. But the idea of customer service -- being both something which users want and something which the institution can offer uniquely and efficiently -- is one which libraries, with their long traditions of library service and now information service, do well to examine closely and implement to its fullest possible extent.

Much thought, in economics and business administration, has been given to customer service distinctions which might be very useful to libraries now. The basic difference, for example, between the product and the service -- in the digital library case between the book and the information in it perhaps, although an historic preservation book library might make a different distinction -- can demonstrate useful new avenues for the library to pursue.

If the library user is the important point of focus, different priorities and strategies take precedence than for situations in which preservation of the collection is the outstanding concern; even in library website design, as determined by surveys, or polls, or the users themselves -- just as Yahoo! permits users to design their own MyYahooPage, so a library might offer its users a MyLibraryPage, online at the library's own site -- a true customer service orientation might yield very different results than other "prestige" and "design" and "card catalog imitation"-oriented libary websites found online today.


From the commercial world as well, then, another lesson, as an extended example of thinking through a "customer service" commercial analogy for libraries: follow-up is 9/10ths of a sale -- if a commercial salesperson sells a single product, to a single customer, and then fails to follow up on that sale with some sort of program designed to attract that customer back in, to buy again and again in the future, the firm has lost money on that one sale and that salesperson in all likelihood has lost her job. This is particularly true nowadays in modern low-margin, mass market retail selling industries, in which the profit obtained from any one sale to any one customer may be counted at most in pennies -- unless an economy of scale of many sales of that product to many "repeat" customers, customers who come back in to buy again and again, may be realized those pennies never will add up to enough money to pay the firm's bills.

So commercial firms follow-up with their customers: literally pursuing them, using the address and other very often intrusive personal information mandatorily gathered at the initial point of purchase -- flyers to the home, piles of email "spam", increasing numbers of automated and "survey" telephone calls which can drive a customer to distraction and ruin a peaceful "evening at home", trading mailing lists back and forth with competitors for great sums of money -- plus generalized advertising on billboards, in magazine ads and inserts which drop out in the customer's lap and onto the floor, on the radio, on television, now even on the side of a city bus or the name of a sports stadium.

The point is not to imitate the practice -- already there is legislation under way in several countries to restrict it, after all, and severe reactions against it by the customers concerned. The point is to realize the immense value of "customer follow-up", apparently, in the commercial setting, and to ask whether or not such "customer follow-up" might not have some value in the library setting as well. Might not a user, and a library, derive some benefit from the library's "following-up" the inquiries and library usage: suggesting additional relevant material, for example, and notifying the user when new things which might be of interest to her either come into the collection or pop up on the Internet?

Might a library use information gathered from the users to organize discussion groups, and reading sessions, and lectures, concerts, poetry slams and other events, even econferences and targeted mailings of various types. Many libraries do such things now -- a dedicated local "chess club" librarian may contact chess-players about upcoming tournaments.

But this normally is a modest matter of a few dozen recycled library cards, arduously assembled and maintained by the librarian in his few precious spare time moments at work: the suggestion is that automation and information technology -- compiling library user profiles, sorting these, using them to generate email and other distributions which would target individual users for various purposes -- the way any commercial firm nowadays would / must follow-up with its customers -- is a neglected and perhaps important information technology avenue for libraries in France and elsewhere to follow now.


But are libraries in the "commercial" world? And if they are not, why are they not? Once again the French case can be instructive, if only because it is different -- as instructive for those outside of France who think that they possess the single answer which might improve French librarianship, as for any in France and elsewhere who believe that there might be a single answer for all librarianship -- even if the French ever might conform, in Cambodia they will do it differently.

The allure of commercialization is significant, now, in too many fields. "Globalization", which ought to be of such universal significance, is seen far too often now solely in its commercial sense, having nothing to do with anything social or cultural or political or governmental. Searching for some of these latter things, all of us interested in globalization, or involved in it whether or not we want to be, need some non-commercial models: at least to consider, if only to reject -- we won't know, unless we consider them.

So if we take a few lessons from commercial models for libraries -- institutions marked historically by a significant lack of commercialism -- we need also to consider some non - commercial aspects of these models as well, and libraries perhaps can help us with this. The lesson can work both ways.

Likewise it can work internationally: just as French libraries, commercial and otherwise, can benefit from considering and perhaps even adopting some of the current "customer service" and "customer follow-up" practices of US institutions -- among them perhaps a few US libraries -- just so US libraries might be interested to consider French practices in their field, even less inclined to commercialism historically than are their US counterparts, particularly insofar as all libraries nowadays, like any other cultural institution, are subject to such pressure under commercial globalization trends to perform on their own financially. And if they US and France can learn from each other, so can both of these and any other country: each can learn from the other -- at least how not to do a thing -- there is no one way only of doing anything.


All of which poses a final question for libraries: do libraries need online users, as much and perhaps more nowadays than online users even need libraries?

There have been closures, of library schools, and downsizings of libraries and their budgets. Of library schools which remain, nearly all have recast their names, at least, in some form so as to be associated better with the "information" revolution -- in most cases altering the curriculum in some manner, in some cases transforming both school and name into an institution concerned more with computer science, or something else, than anything recognizable as traditional librarianship or library users.

And the users do not help things. What if they, rather than search for "bibliothèque" -- as has been the primary example used here -- search instead simply for a "known item" or a "subject", on Google and the rest -- what if they, deliberately or in their innocence, cut out the "library" middleman and "go direct"? What becomes of the library then?

Librarians can sit, complacently -- simply assuming that if the users think libraries are not needed now they will be back, later, once they have faced the wall of Internet "information overload" a couple of times. Or librarians can "take arms against a sea of troubles and by opposing end them": go out to get the users, online and elsewhere, instead of sitting and waiting for the users to walk back into the library.

The oldest commercial retail marketing adage in history is that to sell a thing you cannot wait for the customer to come into the store, you have to go out and seek the customer. Like a number of the suggestions which already have been made here, this might involve libraries in commercial practices which would have to be carefully weighed and suited to essentially non - commercial library situations.

Holding a "Fureur de Lire" event, and managing personalized "recommended reading" lists for users, and offering online reference service -- all of which might boost sagging library attendance and support -- may be lessons from the commercial sector which need not be commercial, however.

It does not have to be marketing, necessarily, with all of the private sector, commercial and capitalist connotations with which that term normally is associated, for cultures like that of France in which the private sector, commercialism and capitalism are not always comfortably associated with librarianship: call, it, instead, "aggressive librarianship" -- active, that is, with respect to recruiting and assisting users, as opposed to passive in just sitting and waiting for the user to walk into the library, or to click on the library's website.

Some libraries feel they need this, others feel they don't. In some places, libraries are greatly overcrowded, and can handle no more users than they currently have, and the library profession is over-subscribed, the schools teeming with new candidates eager to get out and begin offering information service.

But in other places, libraries when crowded are filled simply with "teenagers playing video games" or "students using the building as a study hall", the collections sitting dormant on the shelves while people read at most from photocopied course "readers" and current periodicals collections, library schools find it hard to attract applicants, and vast swathes of the population never see the inside of a library -- crowded or not -- or its websites.

All of this would benefit from discussion and exchange of ideas, and conceivably even cooperation and resource-sharing -- in both / all directions -- for all of which, adequate access from the outside world to France and from France to the outside world is a must.


II. France found -- the view from the inside

Stating the general problem addressed here the other way around, however -- we asked originally, "How does a foreigner find a library in France on the Internet?" -- we need also to ask , "Does that library in France really want to be found?"

The cost -- in time, in materials, in money, in distraction from other perhaps more important activities -- of mounting and then publicizing and then even "maintaining" and "following-up" a library presence on the Internet, is considerable. Typically, library administrators want to know -- need to know -- what the return is for the library, what is being achieved, and whether it is in the interests of the library to achieve it. The view from inside, in other words: what is the mission of the library, in France or in the US or in China or anywhere else, in the age of the Internet?

One approach to this seemingly-intractable problem might be to consider the users, once again, but this time in groupings according to how far these users are located physically from the library in question: not that physical location ought to be relevant in this "age of the Internet", but it is -- at least in terms of the remaining non-virtual collections and perhaps more in terms of the ability to pay fees and to cast votes in elections of importance to the public-supported library --

A subscriber base, where one exists, generally is the "closest" to a library, physically and financially and in terms of interest and other connections -- people who have taken the trouble and perhaps expense to subscribe "care" more. There are exceptions to proximity, in that subscribers can be international or at least physically located much further away than any local user base, but in the case of libraries these tend to be exceptions: most subscribers are local as well -- they are, in any event, one of the groups the library normally least wants to offend.

The local user base, then -- a campus, a city neighborhood, a prison, a laboratory -- includes more than just current library users, unlike the subscribers group. Consideration of the local users raises interesting questions of whether some library decision -- to mount a Website, for example, or to go a further expensive step to do so in English as well as in French -- detracts or at least distracts from the efforts to reach and maintain the subscribers and the local user base. In a public library case, for example, potential users in the locality may be more important to the institution than they might for an academic library -- the latter being relatively more concerned with inter-institutional connections and resource-sharing than a local public library would.

At the regional level, two new considerations at least are added: resource-sharing possibilities, and funding possibilities. Regional library consortia can provide great economies not only in physical collections but also in Internet development and expense. As numerous examples show, just as several library systems in a region can divide up physical collection acquisition policies -- physical proximity of users, and of librarians and ILL services, enabling such resource sharing -- so Internet development can be enhanced, through pooling bibliographic records in a single cooperative regional opac, or pooling Web development and maintenance expense in a single regional Website, both just as any main city or campus library does for its branches.

As nearly as many examples also illustrate, such regional consortia are notoriously difficult to maintain over time -- like any administrative umbrella -- as initial enthusiasms wane and initial enthusiasts retire or simply tire and get replaced and regional differences get rediscovered. But regional consortia are one proven path for pooling talents and expenses -- of great increased interest, and worthy of much additional study, now that trans-national solutions to resource-sharing are more possible than they ever have been before.

Of greatly increased interest, then, at the regional level, is the question of whether and to what extent remote users -- of the type likely to be "dialing in from abroad" via the Internet -- can and should be accommodated, even at the expense of the "local" and "subscriber" user bases. The more fluid and novel construction, which usually typifies the regional consortium -- local libraries tend to have their long-established loyalties and missions, while a regional consortium very often can be an entirely new project designed specifically to experiment with new ideas -- may be able to make greater room more easily for the innovation needed to mount and maintain a service to accommodate a new flood of remote, non-fee-paying and non-voting, users.

Whether it ought to do so, however, is less clear: cooperative cataloging efforts and other "regional" schemes normally are designed to save money, not spend it on new ideas, and prestige or simply glamor to be garnered from some new Internet scheme may be seen to be the claim of the member institution and not of the region. Likewise, the expertise -- rare book librarians, "original" catalogers, local fonds experts, the children's room -- may reside at the local level, so that the regional becomes merely a forwarding address for a user inquiry flood which lands on the local institution anyway.

In France particularly, the national control over library developments -- through grant - funding but also through a national administrative centralization and a sense of national pride which are stronger than they are in many other cases -- augurs well for library Internet projects which can be "national" in scope. Such a project can be designed to reach "all the nation's citizens" and, as such, to constitute excellent advertising for national prestige and values with respect to any international users who might try, via the Internet, to reach and use it -- not incidentally, thus qualifying the project for national government support and finance. The lists of "national" cultural Internet projects offered by the Ministry online at,


-- illustrate one national government's selection of resources which they believe deserve such "national" status and support. Other national governments might make other selections. In the current context, however, the point is that these sites -- and efforts similar to them -- seem to be prime candidates for sites to be made accessible to foreign users, even at the expense if necessary of access of these same sites and other institutional resources to regional and local and even subscriber users of the institution. The trade-off -- national pride and prestige versus regional or local or subscriber information service, which might have seemed inappropriate earlier, may be more suitable now for a resource of "national" interest.

Do library W3 Internet sites fit in here? Is there enough "national" interest, in a particular library and its Internet offering, to merit the expense and distraction of providing it online to a national much less an international user base? And if so, who is to pay the expense?

The Bibliothèque national de France? -- well, clearly. The online efforts of a major bibliothèque municipale or an outstanding bibliothèque universitaire? -- clearly as well. But what of a bibliothèque universitaire which already is having to struggle to provide access to its own local user base -- and what about the bibliothèque municipale of a tiny country town, which arguably might have used the funds and time and distraction on providing better children's services to the local population, or hosting a concert for a local subscriber's group -- what are the arguments in these latter, harder, cases for a library effort which will reach users in Kansas and Mozambique?

For now the issue is an easier one: the digital information techniques are new and exciting, and they are relatively simple -- any library, no matter how small or how financially impacted, very justifiably can argue that a library Website project will repay itself many times over in its introduction of library personnel and users to the Internet and the new technologies -- the Internet works both ways, after all, providing access to foreigners but also foreign access to French users and librarians, and techniques have been simple in these earliest stages, so that many imaginative projects have been created by librarians online in spare time at work or at home.

As the Internet and its use by local users both grow, however -- and as the information overload trickle from the outside world becomes a flood, of router-choking hits on the sites and databases which local users need to see as well, of strange inquiries from strange foreigners in strange languages, of viruses, and advertising, and knowbot and spyder searches, local and even regional institutions may begin to ask whether continuing maintenance of valuable local projects is worth their effort.

"Information wants to be free", runs a famous phrase from the earliest days of digital information -- but it isn't, as so many idealistic dotcom people discovered of the Internet just last year -- sooner or later someone must pay, perhaps local users and subscribers of a too-Internet-minded library, and then the difficult discussions and decisions must be undertaken. If not now, then in few years, once the magnificent projects now being mounted attract and keep their enormous and international Internet publics -- better to begin the thought process behind justifying them now, than to wait and see them simply unplugged then by some budget committee.

In the case of France, Internet development like so many other things has a potentially very rosy and optimistic future, thanks to "Europe". Just as "regional" consortia offer a safety valve, for local efforts wondering whether they have the time and the energy to devote to something not entirely within their traditional local mission statement and user base, so national and really any effort in France nowadays has the safety valve -- if traditional enthusiasms are lacking or begin to flag -- of becoming an effort of pan-European interest, deserving of "European" support and "European" money.

As the largest international trading block in the world, one steadily becoming a fully - integrated and operative political and governmental system, "Europe" offers a French Internet project support possibilities which can be among the best -- most broadly-based, best-funded -- of any online.

If it can be made more efficient, that is. Already, the EU's DGXIII / DG SI / Direction Générale de la Société de l'Information


-- has been active for years in promoting and funding Internet development projects. As "Europe" goes forward, presumably this will continue and increase, so that library or other Internet project in "France" will have one additional and very powerful protection against any local institutional questions about suitability -- if the project is well thought out enough, and of broad enough appeal, "Europe" may pay.

US and other foreign librarians need not look on in wonder at this: the same possibility exists, increasingly, for information workers in any number of smaller countries -- in Europe for the citizens of other nations besides France, in Latin America andAsia, where regional economic and political groupings although not anywhere as advanced as the European Union nevertheless are under way and promising something better than the current fragmented world order which they may in fact replace.

It would be nice for a Cambodian, someday, to be able to receive support for an Internet project from some sort of South East Asian regional international entity, just as someone in France can from the EU today. In the meantime France, once again, is an interesting subject of study -- for this among so many other reasons.

Library cooperation on an international basis in these things is more problematic. The reason is their reliance -- of the libraries as well as of the developments -- on hitech and digital information which now have become so much the focus of national pride. In the US, France, and elsewhere, information technology and information policy have become so newsworthy that national politicians -- and bureaucracies -- have taken up the issues as a cause, injecting their own particular slant or even ideology into their treatments of the general issue.

Thus, one leading US politician has taken credit for having "invented the Internet", to the derision of information workers in his own country and abroad. Thus too, concerns in France about Nazi memorabilia have been expanded into a broad-based suspicion of digital techniques generally, as once also occurred with "Minitel Rose" publicity, when such issues simply are isolated applications of a technology which does so many other things as well.

Now, post-"September 11", and information technology effort which garners national attention must run a gauntlet of national security and anti-terrorism and official secrets challenges, not to speak of the censorship and extremism and immigration concerns which in the US at least have become jumbled together with the others.

The problem is the publicity -- the media, not the message -- all the national level focus has made information technology a difficult subject for any really practical international application. A modest library project wishing to link Internet resources in one country to those in another -- formally, through official channels -- runs a great risk now of seeing hordes of national officials at both ends attempt to climb on board, weighing down and even swamping the effort with extraneous material, irrelevant issues, unnecessary procedural difficulties, and in the end not enough money to justify the trouble -- all for the sake of the publicity involved.

There is a relatively new and very interesting possibility for cooperation at a level higher and broader than the local, regional or national, and even specifically European or even international, however -- one which possibly can avoid the chest-thumping and posturing which currently occurs at the national level in so many cases.

International relations writing has been making the distinction between "international" and "trans-national" for some time: the former referring to relations between national governments, the latter to relations among smaller groups sharing common interests but which happen to be located in different nations. For "international" affairs work can be assisted by, but just as often impeded by, the formal structure of nation-states, with their national governments and bureaucracies and regulations and politicians and so on.

The "trans-national" arena, however -- which is very much the product of increased global transportation and communications possibilities in the modern era -- a group in France sharing an interest with similar groups in the US and Japan, for example, can combine with the others easily, to exchange views and cooperate on projects and even pool finances and personnel and other resources. Such an effort is far easier now -- even simple, using the Internet -- than it ever was in previous eras, distinguished more by difficult transportation and communication, and formal relations between nation-states, when things international all had to pass through national government channels.

The NGOs are the leading example, of interest to international relations theorists -- "Non - Governmental Organizations" -- of which the many "global" ecology movements currently are the most colorful example, although others go back far earlier to organizations spawned by the United Nations, League of Nations, and even completely private groups such as the Institute of Pacific Relations of the YMCA.

Thus, a José Bové can work with, and learn profitably from, activists in his subject area who happen to be in Kansas or Brazil.

Thus also, a lone woman working from a farmhouse in Vermont can, using her Internet connection, summon the world's conscience to do what the world's nation-states do not have the ability or the courage to undertake in banning landmines, winning a Nobel Peace Prize in the process.

My suggestion is then, that -- not instead of but in addition to local and regional and national efforts, also pan-European and "international" where appropriate and possible, libraries today have "trans-national" possibilities which they can use to further cooperation in resource-sharing and document delivery and even collection development and finance and Web development and many other areas of common interest. A medical library in Limoges and one in Nashville have much in common, and they duplicate the efforts of each other in many respects particularly now in online information development. In previous eras, contact between the one and the other might have had to be "official" -- communications through "the national organizations" or, even worse, "Washington D.C." and "Paris" -- now, all it takes is an email.

It requires a relatively peaceful world to make trans-national cooperation possible -- nation-states were invented, by the Treaty of Westphalia, for the waging of war -- but now that we have relative peace, it does seem incumbent upon us to reap its fruits, and one of the best of these might well be trans-national cooperation.


The barriers to communication -- the world is not yet so rosy that merely sending an email can solve a problem, even now that trans-national cooperation is so possible --

The greatest traditional barrier can be the greatest incentive as well. Information, technology on both sides of the Atlantic and elsewhere, currently is in its first great phase of consolidation. From the earliest days of the public Internet -- the early 1990s, really 1993 when public access and commercial use first became available -- growth was continuous and explosive, so much as to have formed an enormous commercial speculative bubble -- which burst, in the year 2000, like the "South Sea bubble" and the "Tulip Mania" and so many other speculative financial binges before it.

Now, as the "dotcom bubble" settles, a new phase of explosive information technology development is gathering steam, this time to take place not in tiny, entrepreneurial, "startup" firms with no prior experience in business, but in major industrial organizations all of which know business operations very well, and all of which now realize at last that they will not be able to survive competitively, going foreward, unless they adopt the new information techniques as well. Many billions of dollars -- far more than fueled the "dotcom bubble" -- are being spent already to engineer this re-tooling of mainstream global industry to use the new information technologies.

So much of the current pessimism about the dotcom revolution's demise is short-sighted: there will be money out there -- far more than ever before -- as mainstream industry, now, retools for its own information revolution. How these funds will trickle down may vary, in each national situation : filtered by government, or granted directly from corporations to university and other researchers -- biased more towards wealthy users, or reserved for "public information space" and "the information poor" -- most recently in the US, "defense" (aka "War on Terrorism")-related or "other" -- but there is money there.

Government and social policy, however, does not stand still -- nor do trends in library policy, and bias. As the world saw in the post-September 11 reaction, still going on, nationalism is very much alive in many countries, and very likely to raise its head in any crisis, now and in the future -- so that the priorities of and interference by national government, in any sort of information work, are not to be taken for granted.

Politics, likewise -- for reasons mentioned already, of the visibility of and publicity accorded to the Internet and digital information work nowadays -- can intrude on seemingly innocent and a-political information projects, now. So that an attempt at trans-national cooperation among health libraries, or to share "land mines" information, or even to promote children's librarianship, quickly can find local and even national politicians weighing in with grandstanding on extraneous issues, simply for sake of the immense publicity nowadays associated with the Internet.

Library policies, as well, do not stand still -- and they react, just as any politics does, not always in the long-term interests of the people involved. So, for example, any time a librarian "shoos away the teenagers" from the overcrowded Internet terminals -- never thinking, for a moment, that this really is the first time he has had anything in his library which really attracts teenagers, his future public, and that more Internet terminals and teenagers is what he really needs -- that librarian is shooting himself and his profession in the foot, crippling the future for the sake of a few of today's short-term realities.

There is fear, at work, in all of this. Fear of the unknown, largely: of teenagers and of the Internet, of too rapid political and social change, of terrorism and globalization and novelty in international relations. But there always has been fear. This is not the new thing. The new thing is the hope that digital information technology will contribute to the amelioration, if not necessarily the total resolution, of some of these age-old uncertainties.

The response to such fear, in some quarters, is censorship -- also cutbacks and downsizing and restrictions of various types. In the US, after September 11, the government and the general society have embarked upon a broad-ranging crusade of such measures -- censoring things, imprisoning some people, "questioning" others, restricting liberties, tightening up. It is a predictable and even an understandable reaction to uncertainty, even a necessary one in some instances. But even where necessary it must be temporary: in the long run, "The only thing we have to fear is fear itself", one of the greatest US presidents said. Exchanges of views and information with our friends in France can greatly ease these transitions -- from confidence to fear, and back to confidence again -- as things change, and we respond, input from the outside world can be a very useful thing.



Libraries in France now are found easily, on the Internet -- almost too easily, considering the "information overload" which characterizes the most difficult of looming problems at both ends of information retrieval. Just as users already are swamped with too much data , so libraries in France and elsewhere increasingly feel the pinch of providing information service to the entire world over the Internet.

And yet there still are so many libraries which are not even online, and so many users who have no Internet connections -- the unlisted, and the information poor.

There are many remedies for this, some of them drawn from the ongoing experiments of the commercial world with digital information, some from reactions against that. Libraries may be able to use both.

The most crucial task in this for libraries is to clarify the institutional mission. Channels exist now for that mission to be expanded -- expanded far beyond the information distribution dreams of any previous generation. To prepare for this task, and to anticipate both its predictable demands and its surprises, a library needs to decide what user publics it is most interested in serving, and what rewards and sacrifices will occur from its embracing something new. The greatest possibilities now exist, but it is best not to embrace them blindly.

With a little more clarity than we have now, the revelatory joy which most librarians feel when they find a new resource online -- particularly something really remote, as libraries in France were to me, from California back in 1988 -- can be extended to many others. Without it, there can be confusion, and information overload, and reaction. But with broad-ranging thought, and long-term planning, libraries may be able to use the Internet to bring information service to users in ways never even imaginable before.

-- Beckett, "En attendant Godot" (1952)



From here you can link / jump up to,

The FYI France Home Page ,

or you can link / jump over to:

1.00 FYI France: Print Libraries in France
2.00 FYI France: Digital Libraries in France
3.00 FYI France: E-Newsletter and Archive
4.00 FYI France: Publishers in France
5.00 FYI France: Book-Dealers in France
6.00 FYI France: Calendar
7.00 FYI France: Discussion and Debate
8.00 FYI France: La Francophonie
9.00 FYI France: Internet Training & Consulting
10.00 FYI France: Essai
11.00 FYI France: Translation Services
FYI France: Bibliographies / Resource Lists
or you can,
Return to the top of this page .

M. Eiffel

Copyright © 1992- by Jack Kessler, all rights reserved.
W3 site maintained at http://www.fyifrance.com by Jack Kessler.
Document maintained by: Jack Kessler, kessler@well.sf.ca.us
Last update: November 29, 2007