Estudi de la llengua d’Ausiàs March a través de les col•locacions Una aproximacio semiautomatica

Since their first publication in 1593, the complexity and beauty of the verses by Valencian poet Ausiàs March have constituted a challenge for his readers. In this study we present an analysis of the poet’s language using an innovative methodology in the field of literary studies: the analysis of word combinations – collocations – with the support of ICT.

Based on a structurally/morphosyntactically-tagged, lemmatized corpus, we carried out an automatic extraction of the collocations in March’s work by means of a search engine optimized to work with the Old Catalan language. Thus, on the basis of the more frequent characteristic patterns of this language, we were able to obtain complete lists of collocations – in a broad sense, as defined by Sinclair in his first research studies. Subsequently, we used a control corpus in order to contextualize them and describe their usage, which allowed us to establish whether the frequent combinations found belong specifically to the author – as an intrinsic part of his style –, or to the general linguistic usage of his age – as part of established Old Catalan phraseology. Our research showed us the difficulties of studying the phraseology of old languages, as well as the productivity of the collocational analysis of an author’s language.