Psycho-computational modelling of the mental lexicon

Over the last decades, a growing body of evidence on the mechanismsgoverning lexical storage, access, acquisition and processing has questionedtraditional models of language architecture and word usage based on the hypothesisof a direct correspondence between modular components of grammarcompetence (lexicon vs. rules), processing correlates (memory vs. computation)and neuro-anatomical localizations (prefrontal vs. temporo-parietal perisylvianareas of the left hemisphere). In the present chapter, we explore the empiricaland theoretical consequences of a distributed, integrative model of the mentallexicon, whereby words are seen as emergent properties of the functional interactionbetween basic, language-independent processing principles and the language-specific nature and organization of the input. From this perspective,language learning appears to be inextricably related to the way language isprocessed and internalized by the speakers, and key to an interdisciplinary understandingof such a way, in line with Tomaso Poggio's suggestion that the developmentof a cognitive skill is causally and ontogenetically prior to itsexecution (and sits "on top of it"). In particular, we discuss conditions, potentialand prospects of the epistemological continuity between psycholinguistic andcomputational modelling of word learning, and illustrate the yet largely untappedpotential of their integration. We use David Marr's hierarchy to clarify the complementarityof the two viewpoints. Psycholinguistic models are informative abouthow speakers learn to use language (interfacing Marr's levels 1 and 2). When wemove from the psycholinguistic analysis of the functional operations involved inlanguage learning to an algorithmic description of how they are computed, computersimulations can help us explore the relation between speakers' behavior andgeneral learning principles in more detail. In the end, psycho-computational models can be instrumental to bridge Marr's levels 2 and 3, bringing us closer tounderstanding the nature of word knowledge in the brain.