2. From perfection to absent interpretation
The normative character of hard-boiled cognitivism is expressed in the fact that it claims and
maintains without empirical backing, much as beliefs are held, and we should ponder on the origins
of such a methodology. It is well known that the wish for universality is a constant feature of
Western metaphysics: it went hand in hand with the political and ideological conquest of the world.
In linguistics, it results in the programs of universal grammars. Chomsky estimates that his universal
grammar is one of the "hypothetical components of our genetic inheritance" (1984, p.21).
The Leibnizian enterprise to seek out a universal characteristic, resumed by Frege's
Begriffsschrift, was to produce a universal and perfect ideography, free from all the irregularities of natural languages. The cognivist feat will have been to fit
this perfect language into the very operation of the human brain, as the means of all possible
Here is not the place to make fun of such a serious subject: Hebrew was long held to be the
perfect language, the language spoken by Adam and Eve. The calculation of first-level predicates,
boosted by modal operators,1 or propositional
calculation, was thought, up to very recently, to provide a format to mental representations as well
as to the language of thought. After several failed attempts at establishing the perfect language, are
we not ready now to discover it within our genetic inheritance?
The refusal of hermeneutics has undoubtedly led the symbolic paradigm to elect the format
of a semiotics characterized by interpretational suspension. It certainly had the merit of underscoring
the role of semiotics in human cognition, but its computationalism has pigeonholed it into a single
formal version. That resulted in the absolute primacy of syntax over semantics, that only symbols
could convincingly implement:
the brain is first and foremost a syntactic machine, which can advantageously be regarded
as emulating a semantic machine, but one where significations themselves never have precedence,
never dominate and have no influence, ever so little, over the raw mechanical or syntactic flow of
local causality in the nervous system (Dennett 1992, p.31).
Besides, certain authors, such as Stich, explicitly refuse to provide a semantic interpretation
to their theory.
That did not prevent, quite to the contrary, rather hubristic ambitions, even totalitarian ones,
which show through in claims such as the one made by Johnson-Laird: "language and society ultimately depend on the mental ability to compute recursively defined linguistic structures"
(1983, p.450). As cognitivism links its naturalization of meaning agenda and its stated purpose of
eschewing the hermeneutic circle,2 it has located the
origin of meaning in a syntactic machinery, thereby suppressing the problem of interpretation.
III. Semiotics and interpretation
After the image of linguistic theories, among which some are universal theories of language,
in line with traditional philosophy, and others are general theories of the languages (as in comparative
linguistics), two kinds of semiotics can be identified: universal semiotics and federative semiotics
(Rastier 1993). Universal semiotics are quick to develop into a science of the sciences, or to give the
label of signs to everything (see Peirce).
However, no concept makes it possible to summarize the whole of the semiotic level,
quite simply because its inside is where conceptualizations thrive: Meaning, as soon as it becomes
hypostasy, is no longer within reach, because it falls within the realm of transcendental thought, and
not within the scope of the sciences. It appears necessary to take a step back, in scepticism or as a
tactical manoeuvre, and to keep the door open for some philosophy of Meaning to visit, before the
semantics of natural languages, along with the semantics that study other sign systems, can be
developed further. In short, three roads can be taken.
1. This computation articulates the logical patterning of the
semantic interpretation of sentences in the Chomskyan theory of the Eighties; or then again the
predicative format which, according to van Dijk and Kintsch, would have made it possible to encode
the signification of all sentences in all texts. [RETURN]
2. Dan Sperber wrote, under the title of The Knowing
of Knowing ("Connaître l'acte de connaître"): "There is no thought without signification. Is
this to say that signification should also receive a Darwinistic explanation? Can signification "be
naturalized"? Here is undoubtedly the Grail of cognitive philosophy. If we manage one day to explain
the signification of a speech or the contents of a thought without bringing back them to other
significations, to other contents, if, in other words, we are able to leave the "hermeneutic circle",
then, indeed, there will be a cognitive revolution. The gap between the natural science and the social
sciences will have been filled" (1993, p.32). According to us, the gap only appears to be one to uncritical
objectivists, whose turn of mind is a late revival of last-century scientism. It will be bridged once
the hermeneutic dimension of "natural" sciences is established, as authors like J. Stewart are trying
to do in biology or G. Cohen-Tannoudji in physics (Rastier et al., forthcoming). [RETURN]