Y a-t-il une relation entre la valence (pleine) et la synonymie?

Abstract

The article is devoted to the analysis of the possible relationships between (full) valency and synonymy. We first present a very short overview of positions on valency, then we proceed to present the position of researchers who see a relationship between valency (full, i.e., not distinguishing arguments from adjuncts and treating them all together as arguments) and synonymy. The article shows that since a more frequent word would appear in more contexts than a less frequent word, the more frequent word would tend to have more meanings, and therefore it will have more synonyms, and being more polysemous it would result in a greater number of full valency frames of that word. It has been shown that the hypothesis has not been sufficiently precise, because it is the word, as a form, that can be considered polysemous, but it cannot itself have synonyms: it is only a particular meaning of this polysemous word that can have them. Therefore, the analyses could not be sufficiently subtle to identify any relationship, if any, between the two phenomena. On the other hand, the results of the analyses from this not sufficiently precise starting point did not demonstrate that there is a significant correlation, let alone dependency, between the two phenomena. Kendall coefficient, which measures the ordinal association, was estimated at 0.18 in the case of the material analysed (for the range –1/+1). It was pointed out at the end that it is not possible to draw from the fact that the differentiation between arguments and adjuncts is often subtle and sometimes difficult to make, the conclusion that there is no difference between them, that the problem in fact does not exist, and to refrain from searching for satisfactory elements and criteria for differentiation of the two categories or to apply in a consequent way those at our disposal, namely, semantic implication.