Are all sciences not cognitive?
Two major problems, mainly investigated by language scientists, dominate the Western
epistemological tradition. They correspond to a pair of underlying preconceptions about language
__ whether it is seen as a means of representation or as a means
of communication. In short, the former posits meaning as a relationship between a subject
and an object, and the latter as a relationship between subjects.
The dominant of these views, meaning the former, which follows a logical and grammatical
tradition, stresses signs and syntax in its approach to language, relating them to the laws of reason.
It is centred on cognition, and cognitivism represents its current state of development.
The second school of thought, belonging to a rhetorical and hermeneutical tradition, examines
text and discourse in terms of their generative and interpretative processes, and can be considered to
focus on communication 1 through pragmatics, which
borrows some topics from rhetoric, and offers a restrictive outline of communicative processes; this
approach is largely based on the logical positivism of Morris and Carnap, a kindred spirit to the
discipline. For the sake of concision, we will classify the former approach into a problematic
of signs, and the latter into a problematic of texts.
Recognizing a distinction dating back at least as far as Dumarsais, let us adopt the convention
that signification refers to a property of signs, and meaning to a property of texts. The transitional
concept of context may help distinguish between the two problems. If we look further into the
distinction between meaning and signification, a sign, at least when in
isolation, does not have meaning, and a text does not have signification.2 Signification results indeed from decontextualisation, a
dispelling of context, as one may see in lexical semantics and terminology. In this respect
signification is a stake of high relevance for ontological questioning, since traditionally Being has
been characterised as identity with oneself. On the other hand, meaning implies maximum
contextualisation, in language (context is all of the text) as well as through situational criteria
(inclusive of the co-dependent variables of history and culture, beyond the sole here and now
considered by pragmatics). Therefore, whereas signification is traditionally posited as a relationship,
meaning can be represented as a process.
By privileging the study of meaning, interpretative semantics (Rastier 1987)
focuses on the text, rather than the sign, and defines meaning as interpretation. It is based
on text disciplines (law, theology, literary criticism, among others) and can be ancillary to two
kinds of theories: philosophical hermeneutics and philological hermeneutics. Since it aims to describe
a great diversity of texts, it is naturally closer to the latter, because while the former seeks the a
priori conditions of any interpretation, the latter seeks on the contrary to specify the incidence
of social practices, and leads to a typology of texts.
If of course the study of signs and that of texts supplement one another, logico-grammatical
problems and rhetorical-hermeneutic problems differ significantly. The first one enjoys widespread
recognition and has strong unity, for until very recently grammar and logic developed in parallel and
around the same categories (such as the very concepts of category, predicate, categoreme and
syncategoreme, etc). The second one hardly has unity at all, and apparently rhetoric and hermeneutics
till their own fields: the spoken and the written, enunciating and interpreting, Reform and Counter-Reform, persuasion and grace, Latinicity and Germanicity, etc.
* Research Director (INaLF-CNRS), 57, rue de Paris. 94340 Joinville-le-Pont. E-mail: firstname.lastname@example.org
Reprinted from Intellectica, 1996, vol.2, nº23, p.11-52. Translated by the Editors, including quotations from the French. We have chosen to distinguish between "natural language" (une langue) and unspecified "language" (un langage), inclusively of several kinds of semiotic systems such as animal communication, artificial languages, and codes. [Translators' note]
1. Considering the failings of communication theories, we
would rather adopt the term transmission (Rastier 1995b), insofar as cultural transmission,
semiotic inheritance, can be included in it. [RETURN]
2. Three paradigms of signification centered on
the sign, can be said to dominate the history of Western linguistic ideas:
reference, inference, and difference.
a) The paradigm of reference, in the aristotelian tradition, defines signification
as a mental representation, precisely a concept. It is variously continued today by vericonditional
semantics and cognitive semantics.
b) The paradigm of inference, in the rhetorical and Augustinian tradition, defines
signification as an intentional action of the mind, connecting two signs or two objects. It is developed
today by pragmatics.
c) The paradigm of difference, of sophistic origin, and developed by the
synonymists of the Enlightenment, then by the so-called structural semantics, defines signification
as the result of a contrastive categorization.
The synthesis of which we have just offered the principle consists in determining inference
and reference through difference, then to place these problems of signification under the authority
of semantics, by assuming that the global (the text) plays a determining role over the local (signs).