Clear vs. approximate categorization in French and Latvian

This paper aims to put into perspective various issues surrounding categorization and approximation, two frequently opposed yet essentially indissociable operations, by comparing the taxonomic expressions in two languages belonging to different language groups (French and Latvian). We mean to verify in particular whether the expressions exhibit the same kind of semantic multifunctionality in the studied languages or, on the contrary, if we observe specific constraints in each language (lexical and grammatical variations, syntactic marking, etc.). One way to illustrate these issues is to focus on the analysis of the nouns preceding and following the type nouns in the binominal structures (X = N1 type noun Y = N2). It is also important to reinvestigate the terminological imbroglio that closely affects approximation, imprecision, vagueness and categorization. Thus, with reference to psychological literature, we will show that the principle of categorization itself functions by means of relating elements, in a similar way to approximation. We also challenge formal and semantic correlations attached to often juxtaposed interpretative types (clear, approximate, non-prototypical categorization or vagueness). However, we must temper this challenge for Latvian: with its morphological syntax, it marks lexicogrammatical differences in a more clear-cut way than French.