In traditional generative linguistic theories, idiomatic constructions are seen as a sort of “anomaly”, and dismissed as non-decomposable items of non-literal language, uninteresting and “peripheral”. Contrary to this view, in the last decades psycholinguistic and corpus-linguistic studies have shown that idioms can often undergo structural modification and display different variation patterns, according to their specific formal and semantic properties. In virtue of these findings, the present study aims to investigate the levels of stability and variation in Italian idioms from a socio-cognitive point of view, in a two-step fashion.
In the first stage, a set of 150 idiomatic constructions will be selected from a dictionary (Sorge 2010) and, taking the categorization proposed by Langlotz (2006a) as a starting point, a cognitively motivated typology of Italian idiomatic constructions will be drawn. Langlotz's parameters and categories will be used to classify Italian idioms into a structured taxonomy based on a set of notions which are generally accepted and employed by proponents of functionally-oriented approaches to language; these notions will be applied taking the Italian cultural context into consideration, in order to avoid (potentially hasty) claims about their supposed universality. Then, the mutual relationship between different idioms on the one hand and between idiomatic and non-idiomatic constructions on the other hand will be addressed and accounted for in the light of a constructionist perspective on language.
In the second part of my study, a sample of occurrences of a subset of 50 idiomatic constructions will be downloaded from a large Italian corpus, in order to observe their variational behavior in the context of actual interactions in a contemporary setting. Particular attention will be paid to the potential correlation between the category an idiom was allocated to in the previous stage and the variation patterns observed in its occurrences, with the specific aim to understand if a causal connection can be established between the idiom category and the (quantitative and qualitative) level of variation observed in real language data.
The two phases of the study will be treated as deeply interconnected, and a dynamic-systems approach will be adopted to highlight the several links between the two stages. An integrated model of the mechanisms which regulate the “life-dynamics” of idiomatic constructions will be provided, taking distinct dimensions, time-scales, and levels of granularity into account. Finally, the results of the study will be scrutinized in order to assess the adequacy of a dynamic-systems perspective to accurately explore and describe the self-organizing nature of linguistic constructions and their relationship with other aspects of human cognition and interactivity.