Summoning the indignation to which only the French language can give pure expression, Jean-Noël Jeanneney, the director of the French national library, touched off a national debate not long ago when he said that Google's plans to digitize the collections of major American and British libraries raise ``the risk of a crushing American domination in defining the idea of the world that will be held by future generations.''
Not that Google's virtual library won't be a boon to humanity, Jeanneney said. But its selection criteria will be ``strongly marked by an Anglo-Saxon point of view.'' It would be ``deleterious and detestable,'' he added, if someone surfing the Web for information about the French Revolution could find nothing but British and American accounts that depict ``valiant British aristocrats triumphing over bloodthirsty Jacobins as the guillotine eclipses the Rights of Man.''
In the wake of Jeanneney's remarks, commentators warned that French culture was at risk of being marginalized by the process of ``omnigooglization.'' A few weeks ago, President Jacques Chirac responded to Jeanneney's appeal for a French ``counterattack'' by asking the national library to draw up plans to accelerate the digitization of its collections, and to work with other European nations to put their libraries online.
Martial metaphors aside, that's welcome news for all devotees of French culture and literature, who will have online access not just to Dickens' and Carlyle's versions of the French Revolution, but also to the French side of the story, as told by the historian Jules Michelet.
But the fact that the decision came as a response to the ``American challenge'' demonstrates how the Web has aggravated concerns about the linguistic and cultural domination of English -- and not just among the French. At a conference in Geneva, Switzerland, last year, Kofi Annan, the secretary-general of the United Nations, warned that the predominance of English on the Web was ``crowding out voices and views.''
What's curious is that those concerns seem to be mounting even as the Web becomes less and less of an English lake. The Internet was basically a North American development, and the vast majority of its early users were drawn from the United States and the rest of the English-speaking world. But native English-speakers now constitute a minority of people using the Web. The proportion of non-English content also is increasing sharply as American Internet penetration approaches saturation and as the rest of the world comes online.
In 1997, the linguist Hinrich Schütze and I calculated that about 85 percent of Web pages were in English. Five years later, a study by the Online Computer Library Center came up with a figure of 72 percent. Those estimates are rough, but it's certain that English will account for less than half of Web content within a few years.
True, not all the English on the Web comes from the English-speaking nations. The international character of the Web has encouraged many sites in other nations to post their sites in English so they can be understood by English-speaking foreigners. But the proportion of such sites is dropping as foreign Internet use spreads to individuals and small businesses.
A university in Barcelona or the Swedish Trade Council may find it advantageous to post some content in English. But when bloggers, movie theaters and real-estate agencies come online in those places, they're more likely to post in Catalan or Swedish. Then, too, the technology is now up to representing many non-Western writing systems whose users were forced to post in English before now, though there is still a way to go in this area.
By all rights, that ought to allay some of the concerns of the French and other non-English-speaking nations. Paradoxically, though, the impression seems to be growing that English is ``crowding out'' other languages, as Annan put it, even though that metaphor makes no literal sense in a forum where space is an unlimited resource.
Blame ``omnigooglization.'' The real crowding out is in search-engine results, which show the totality of sites from all over the world and often give higher rankings to the more popular English-language pages that are widely linked to. Do an unrestricted Google search on the name of the late French writer Roland Barthes, for example, and 44 of the first 50 sites that come up are in English, with four in French and one each in Spanish and German.
The prevalence of English pages among those results may be disconcerting to a Parisian who is accustomed to browsing the reassuringly francophone shelves of the bookstores in the Rue des Écoles. As it happens, though, the ratio is roughly consistent with the numerical proportions of English- and French-speaking Internet users -- all it shows is that Barthes is held in equal esteem in both communities. And a search on Barthes restricted to French-language pages still turns up 75,000 hits, enough to keep anyone busy for a while.
Rather than decrying ``omnigooglization,'' the French and other language communities should welcome it as a sign that distance is no longer an impediment to diffusing information. In the world of print or broadcast, for example, only American news media can achieve anything like worldwide distribution. An American in Strasbourg can easily find CNN or Time or the International Herald Tribune, but a French-speaker has to dig hard to find similar French media in Milwaukee.
And for speakers of smaller national languages like Greek or Hungarian, the circulation of information has pretty much stopped at the national borders. Now, however, speakers of those languages can have access to news and opinion in their own languages wherever they happen to be -- already a significant factor in communities with large international diasporas that have been historically shut out of their home countries' political discussions.
Even Yiddish has experienced a modest boom in the discussion groups of the Net. One linguist who up to now has spoken Yiddish only with her parents and their friends tells me that for the first time, she can have Yiddish conversations that don't involve the merits of denture creams and early-bird specials.
In the near future, the leveling effect of the Internet will extend further. Even if a new film by French director Olivier Assayas doesn't get extensive international distribution, cinephiles in Cleveland and Caracas will be able to order or download it directly. And the continued digitization of library collections means that those who speak French, German, Japanese or Hungarian will have access to the cultural patrimonies of those languages, wherever they happen to be.
But if the Internet levels the playing field between English and the other languages of the developed world, it also increases the inequalities between those languages and ones spoken in poor and developing nations, where online access is restricted to a small elite and where even basic information is not available online in the local languages. The digital divide we should all be worrying about is the one between north and south, not the one that runs under the English Channel.
GEOFFREY NUNBERG, a Stanford University linguist, is the author of ``Going Nucular.'' He wrote this article for Perspective.
