Does the Internet and its generalized, hypertextual publication space — the Web — fundamentally alter the veracious and/or fictional relationship of language with the world? It is this seemingly simple question that I would like to submit for debate.

Undoubtedly, ethical questions — or simply deontological interrogations, in certain professional contexts — are regularly posed regarding the abuses, excesses (or weaknesses) and new forms of transgression (or constraint) introduced by the Web. Piracy, hacking, hoaxes and fraud of all kinds are generated by the development of global networks, through the medium of electronic mail as well as via the Web. For some observers, these new possibilities for fraud or dissimulation — or simply the difficulty of determining editorial status or even interest — intensify but hardly change in a fundamental way the profoundly destructive tendencies manifested in human relations over the millennia. In an extreme form of such arguments, the widespread development (in more or less disguised form) of paedophilic, violent, racist, revisionist, or simply “uninteresting” sites, is claimed to be primarily a result of the fundamental perversity of human nature. For others, the unprecedented level of textual production is the cause of the problem, but is considered the price we must pay for the power of new technologies, which new regulations must restrain in the future (1). I would like to reflect on various reasons specific to the new medium of the Net, that, though perhaps less extreme and delicate than these other infractions, are by their very ambiguity less visible and more fundamental. In other words, the question is whether it might be necessary to rethink in a more profound way the relationship that this medium introduces and maintains between language and the world.

Since Frege, analytical philosophy has investigated, as a complement to other conceptions of language, the relationship of reference (Bedeutung) and thereby of truth, that language maintains with the world, as distinguished from dimensions of sense (Sinn) — fiction for Frege — according to which language can signify even though verification may not be practical or even possible. Errors, fraud, pretence and even fiction therefore clearly become domains of linguistic activity, without a real engagement of the logical possibilities of language. This aspect of the philosophy of logic and of language has long been clearly established (2). Henceforth, the philosophical debate is open as to whether fiction is a non-referential linguistic form (i.e. without literal denotation). Nelson Goodman contests this claim, preferring to speak of “non-denotational reference” — thereby revealing the possibility of a positive understanding of the power of fiction, without having to contradict Frege.
How does all this relate to the World Wide Web ?
The idea is that the Web, as a result of its geographic reach, its multi-linguistic (and therefore trans-cultural) scope, but also its intrinsic dynamics (a more or less automated hypertextualisation, the interruption of the authorial function, the reappropriation of a whole range of semiotic, iconic, symbolic and auditory worlds), would seem to permit — at the very least — expressive and communicational practices in which the conditions of verification and referencing become, if not impossible, at least difficult or random. However — and this seems, for the moment, to contradict the present argument — we are forced to recognise that attempts to create fiction on the Internet (except for a few, perhaps excessively experimental efforts) have remained “disappointing” in the sense that they contribute nothing more than that which literature or cinema have already provided.
It seems to me that, particularly as a result of its unlimited power of hypertextualisation, the Web is largely “self-referential” (or “connotational”), and that the borders between fictional and “true” or “verifiable” worlds are thus much more difficult to determine.

This is more or less my position; I suspect that that the lack of distinction among the practices mentioned is more intrinsically linked to the nature of the medium itself, to that which distinguishes it qualitatively, and that it is necessary to broadly clarify this specificity using varied and innovative heuristic hypotheses. Numerous symptoms warn us that vigilance is necessary. When journalists (those guardians of reference!) express worry over the potential power (and possible abuses) of Google News as a means of producing automatic press reviews (3), unavoidable questions are posed regarding the fundamental techniques of “page ranking” and the logical elaboration of the system. When the Web becomes a privileged forum for the dissemination of rumours (for example, the recent allegations of conspiracy in the wake of Meyssan’s book on the September 11th terrorist attacks, despite all evidence to the contrary), provoking sensitive and radical responses from the journalistic community (4), the appropriate reaction is not to hide our heads in the sand.

The evidence suggests that a new “pragmatics” of the Web must be taken into account — not as a marginal phenomenon, but as an element which, though perhaps not yet entirely visible, is fundamentally linked to the future development of the Web. By “pragmatics,” I mean that these on-line linguistic and semiotic practices lead to new human behaviours. For example, in this respect, the regulated practice of publishing (in the academic sense of the term) falls apart — with undeniable benefits for the increased circulation of scholarly reflection, but also undoubtedly as a result of the negative way in which traditional rules (the condition of authorship, the legitimacy of reading committees, the functioning of scientific committees, the conditions of reception and readership) have been called into question. For this reason, the work of the Pragmatic School (to which my reflections owe a great deal (5)) should be put to more productive use.

Another heuristic trajectory consists in using memetic theories to take into account the fundamental reproducibility that characterises digital documents. To the extent that these theories attempt to characterise the technological evolution of material duplication in a cultural context, now digital, as processes homothetic to those involved in genetic selection and Darwinian natural selection, we must take seriously the analysis of the memosphere in order understand and extract criteria of fidelity, fruitfulness, and longevity attached to communicational processes. From this point of view, the paths of denotational and fictional reference mentioned above owe a great deal to an economy in which the digital meme, like the gene, has its own logic of competition and success: “memes can take different paths in order to succeed, in the same way that genes use different strategies” (6). As a result of the “replicative performance” intrinsic to digital formats, the value of the copy is superior (in strategic terms) to that of the original — in which case, questions relating to authenticity, to “content,” to truth as proof, and to the adequacy of denotational reference, are somewhat disrupted.

In relation to these hypotheses, according to which digital language is, in its “machination,” beyond that which formerly bore the admirably manual name of “manipulation,” the Web begins to follow a path well known to information specialists. Its mode of operation implies an ever-increasing distinction between language and metalanguage, data and metadata. At issue is not only the distinction, within the machinic deployment of language, between two levels in the modus operandi of the search for and identification of information, but also the claim that procedural metadata (7) should be added to these descriptive metadata, filtering, authorising, profiling, according to prior characterisations of the data. As we know, this is related to future advanced treatment techniques for on-line resources: client machines will “understand” complex requests.

To conclude, these considerations could converge in the following manner: the theory or semantics of possible worlds, as an adequate conceptual space for representing the characteristics of the Internet, could conceivably be juxtaposed with the logical and syntaxic tools that the (semantic) Web will require. The idea is that the transgressive character of the Web, far from being (exclusively) a flaw, could constitute the new system of benchmarks in relation to which web publishing must be put into perspective and evaluated. The veridiction of Web resources becomes a central question, if only because their growth is exponential and uncontrollable in determinist terms. As I have indicated (8), the contemporary foundations of logic (sense, denotation, reference) are quite explicitly necessary. Cognitive questions of error, of fiction, of transgression or falsification of data, and therefore those regarding more generally the ways in which the network’s various manifestations might be regulated, become essential.

Does the extension of an RDF syntax to these modal operators make sense ? (9). From this point of view, we may be able to mobilise the field of modal logic and the semantics of possible worlds, particularly the foundational work of S. Kripke and D. Lewis, provided that we posit that the Web also exists to create possible worlds. The Web, perhaps more than any other medium could (though this is not yet the case), create a “narrative depth,” in which the “real” and the various “symbolic denotations” could ceaselessly reconfigure themselves, not so much in order to dissimulate or to confuse, but to complexify, and thereby enrich, the heuristic possibilities of a reading of the world.

That machines and automata might help us to better structure and differentiate, on the Web, “real” (verifiable) spaces and erroneous or “fictional” spaces, is by no means an insignificant paradox.

(1) This is, to a certain extent, the point of view of Marc Guillaume, see l’Empire des réseaux, for whom the discrepancy is a result of unequal development: rapid for new technologies, slow and uniform for culture.

(2) Jean-Marie Schaeffer, in the Nouveau dictionnaire encyclopédique des sciences du langage, provides a clear presentation of this tradition. See p. 373, “Fiction”

(3) see http://news.google.com/news/gntechnologyleftnav.html and Automates-intelligents n° 36: “Google provides a current example. The question will be posed: in what way is it democratic? Won’t big (notably American) publishers be privileged, as opposed to small ones? What proof do we have that the results produced by the program won’t be quietly altered in order eliminate articles considered politically incorrect? What will happen to journalists and editorialists? Undoubtedly, all kinds of manipulations are possible. However, we don’t see how the proposed system eliminates debate and the liberty of opinion.”

(4) See Le Monde, 26/03/02: “Les journalistes et le livre de Thierry Meyssan”, and “Internet, l’agora de la rumeur”, or the editorial published in Le Monde, 20/03/02: "le Net et la rumeur"

(5) I would like to thank Anne Reboul for her help. See La pragmatique aujourd’hui Points essais Le Seuil, 1998 and Dictionnaire encyclopédique de pragmatique, Seuil, 1994. A. Reboul et J. Moeschler.

(6) Susan Blackmore. The meme machine, Oxford U.P. 1999. Thanks again to A. Reboul for having drawn my attention to these perspectives. See also Denett, d. La conscience expliquée, Odile Jacob 1993, Dawkins, R. The selfish gene. On the convergence between this reflection on reproducibility with those of W. Benjamin in bulletin des bibliothèques de France, 1997: “L’œuvre d’art à l’ère de sa reproduction numérisée” Y. Maignien)

(7) I am referring here to the semantic Web, and more precisely to the Resource Description Framework syntax.

(8) Again, following J. M. Schaeffer et O. Ducrot Nouveau dictionnaire encyclopédique des sciences du langage Essais Points Seuil, Paris 1995.

(9) I think so, obviously, but a working group should be created in collaboration with the W3C in order to test these hypotheses.