The lexicon-encyclopaedia interface as a measure of meaning complexity

The English verb see is one of the most general in the conceptual domain of vision, namely it allows a great deal of co-textual and contextual modulation and can foreground various conceptual dimensions, and correlated meanings, presupposed by the domain. The work analyses some of the most relevant factors responsible for the modulation of the meaning of this verb and the corresponding variation of translations into Italian. Relevant data were mainly gathered from the British National Corpus (through the sketch-engine, ). The data discussed here are part of a prototype for a lexicographic file for the verb see as developed at Pisa University within a research project on lexical complexity. The lexicographic file is a prototype for a multilingual dictionary in electronic format, which might become a useful tool in translation. The innovativeness of the file rests on the attempt to overcome the limits of traditional dictionaries (e.g. circularity) by developing a lexicographic instrument which, ideally, can account for word meanings in a way close to the hypothetical representations in our minds. Rather than having an alphabetical organisation, word meaning representation has an encyclopaedic scope and follows a criterion of lexical complexity based on the theory of complex dynamic systems (cf. Bertuccelli Papi and Lenci, 2007). My work has the general goal of providing parameters (cf. conceptual dimensions) for measuring the complexity of the lexical item at issue so as to account for meaning variation as systematically as possible. Moreover, the specific claim I make is that, whenever involved, encyclopaedic knowledge can indeed be viewed as a complexity factor in the modulation of the meaning of the verb. The key question leading my research has thus been ‘which factors are responsible for the emergence of a given meaning of the verb?’ Indeed, the work shows the dependence of different emergent meanings on the following: - co-textual factors such as different kinds of lexicalisation of subjects and objects, and the presence of adverb/ials of space and time, - the temporal-aspectual morphology of the verb, cf. tense and -ing form, - inferential extensions of various types via recourse to relevant sections of encyclopaedic knowledge. A progression is then proposed, starting from the more basic meanings of the verb (cf. the dimension of Perception and Cognition) to more and more ‘distant’ meaning dimensions (viz. Experiential, Action, Action + Affect). The latter are selected via inferential work and correlated metaphorical and metonymical extensions hinging upon varying sections of encyclopaedic knowledge, and tend to correspond to less predictable translations into Italian. Of course, the selection of examples in the file does not exhaust the meaning multidimensionality allowed by see but is especially designed to illustrate the rationale behind our proposal. In fact, research on other relevant factors of variation in more extended contexts is needed. Nevertheless, the proposal appears to provide promising guidelines for investigating the syntax-pragmatics interface of lexical items