The semic matrices of meaning

The possibility of a quantitative approach to the problem of sense outlined in the following article, is one which has been suggested innu merable times in those sciences which are based on a model of the sign (e.g., linguistics and logic). The controversy between partisans of the quanti tative or statistical method and those who accept the qualitative or semantic one remains unresolved.

If the 1930's, influenced by logic, favored the first method, recent years have evinced a preference for the second. Nevertheless, the problem remains. Generative grammar eliminates the question when it proposes calculations of syntactical transformations and mathematically organizes the generation of " signified " structures as if they were "signifying " structures without ever formulating the problem of sense.

Semiotics is an heir to this situation, again seriously raising the dilemma, to organize as signifiers, the signified in systems which it studies. The complementarity of the two approaches (quantitative, qualitative) is essential. One cannot overemphasize the fundamental importance of the first method, more concretely and more specifically in the realm of semiotics, that is, the importance of mathematical and logical proce dures for articulating the signifying of the corresponding signified material.

M. Guiraud's text, while strictly in the realm of linguistics, is of considerable interest to semiotics because of the methodological and theo retical problems which it raises : 1) The homology of the signifier/signified. 2) The acceptance of the notion of " sense " a) as a function of a statistical factor (word frequency), b) as semic content on the one hand, and sense, in its fullest meaning, on the other, and c) as an " orientation of the word toward words, as a relationship with the semic properties which it does not possess ". Semiotics, still in its formative stages, can render more flexible the qualitative models which it has hitherto used, by these and other propo sitions of quantitative linguistics. With this article, we initiate the regular publication of studies involving the application of quantitative models, thereby working towards the rigorous axiomatization of semiotic systems. (Editor's note.)