Rastier, François: "On Signs and Texts"


c) Symbols are the signs of the language of thought, independently of natural language. The cognitivists' originality is that they have maintained that this language is a symbolic one, in the image of formal languages, and that it plays for the mind the same role machine code plays for the computer. This speculative thesis is obviously a means by which dogmatic rationalism runs the cognitive science show. And as Reason is nothing but the secularized form of the Soul, animals are mysteriously deprived of mental language.
Only assertions and arguments of authority substantiate the idea that thought is a symbolic system: "mental representation [sic] are symbols: they have both formal and semantic properties" (Fodor 1981, p.26). Let us not expostulate on the atomization of mental life, since symbols are discrete entities, nor on the resilience of the representational paradigm (Rorty 1990). Mental contents are reduced to representations and beliefs, to which propositional cores and propositional attitudes respectively correspond. This distinction is actually based on the ontological opposition between categorematic and syncategorematic elements. Symbols have semantic properties, but do not have an interpreter to establish them (contra Edelman 1992); and the hermeneutic issue is once again sidestepped.
However the difficulty of the perceptive, and more generally biological, background of the symbolic level arises. If symbols ground cognition, what kind of metalanguage are they grounded on themselves? As a formal language cannot be its own metalanguage, as the ineluctability of the hermeneutic circle is rejected, and as the semiotic is not regarded as autonomous, the problem is addressed by a naive brand of nativism: we are born thus equipped.
Because of its sheer simplicity, the semiotics of the symbol is integrative. In short:
(i) if the symbol has no signified, or at least if its interpretation can be suspended, it does not have a signifier either. At least, according to Fodor and so many others, the symbol is amodal, and is as such independent of any perceptive modality.
(ii) It has no contextual variations.
(iii) It is not governed by a textuality of any sort, but by compositionality.
This simplicity makes it possible for it to be placed everywhere. The purely syntactic definition of the symbol does not prevent, quite the opposite, that traditional cognitivism turns it into a mediator between neuronal states and states of affairs: "mental symbols are neuronic configurations with physical, chemical and biological properties (studied by the neurosciences) as well as with formal or syntactic properties. In addition, being representational, symbols also have contents, and semantic or intentional properties: they represent aspects of the environment" (Andler et al. 1992, p.12).
One of the lesser accomplishments of symbols is to account for all of cognition __ world, mind and brain. They indeed mediate between the environment and the neurons. One can however have doubts and be somewhat dissatisfied, in spite of the lofty intents in some sub-branches of the neurosciences, as to the assertion that mental symbols are neuronic configurations. It is worth wondering about the ideological function of neurons: is it not to anchor cognitivism in nature?
Cognitivism undoubtedly is the most evolved and systematic among the theories that originated from the logico-grammatical branch. As such, it is unlikely that another paradigm may compete with it on its own terrain (see below, V).

Page - 1     Page + 1

AS/SA Nº5, Article 4 : Page 12 / 27

© 1998, AS/SA

E-mail to the editors
Pour écrire à la rédaction