Lexeme-Based Model vs. Morpheme-Based Model from Psycholinguistic Perspectives

Résumé En

For the past thirty years, research on morphology in psycholinguistics has aimed at determining the way in which complex words are encoded and represented in long-term memory. Several empirical results have shown, in various languages, that on one side an experimented reader is sensible to the structure of words in her/his language, and on the other that morphology plays a major role in the organization of linguistic units in the mental lexicon. More precisely, experimental data from priming paradigms have shown that morphological units are represented in the memory in an explicit way, which is independent of other units describing the linguistic features of words (orthographic, phonological, and semantic). This paper presents a series of experimental results based on French, which took as their basis suffixed and prefixed words, derived and inflected forms, and complex words formed with bound or free stems. The results permit a precise determination of the locus of morphological units within the lexical architecture. The paper thus elaborates a connectionist model of interactive activation, called a supralexical model of morphological parsing. In this model, morphological units of parsing are in fact situated at the interface between the ortho/phonological and semantic level, and act as organizing units for the mental lexicon.