ROMANIAN GLOSSONYMS: PROPER NAMES OF LANGUAGES AND (SILENT) CLASSIFIERS

The paper investigates Romanian proper names of languages, referred to throughout as glossonyms, and proposes a tripartite classification of such constructions. The first category of Romanian glossonyms is represented by structures of the type limba română / language.DEF.FEM Romanian.FEM / ‘Romanian language’, for which the paper proposes an analysis in terms of complex proper names (van Riemsdijk 1998; Cornilescu 2007), where limba ‘language’ functions as a classifier. The second category of Romanian glossonyms consists of complex proper names with covert classifiers, of the type LIMBA română / LANGUAGE Romanian / ‘Romanian’. Finally, the third category is represented by -ește glossonyms, where the (nominalizing) suffix is syntactically active, signaling the presence of the same covert classifier. In the latter category, two other points are made in the paper. First, where ambiguity arises between two interpretations of -ește derivatives, i.e. the like X and the language X interpretations, two prepositions serve the purpose of disambiguation: în ‚in’ and pe ‘on’. Second, the same prepositions are used to encode a temporally-bounded interpretation of the speaking event, as opposed to a generic / habitual one in the absence of the prepositions.