Abstract
The English verbs cut and break, and their counterparts in other languages, often figure in discussions of verb semantics, constructional meaning, and the relationship between the two. Ironically, the core meanings of these verbs are rarely addressed: the critical portion of the meaning is typically rendered as a constant (“in need of further exploration”), as in break: ‘x cause [y become BROKEN]’. So what does it mean to ‘cut’ or ‘break’ something? And how are elements of meaning distributed across the predicate in different languages (e.g., simple verb, verb-particle or verb-verb constructions)? This study investigates how languages partition the semantic domain of ‘cutting and breaking’, and how children learn the system of their target language. In the first phase of the research, descriptions of a set of videotaped ‘cutting and
breaking’ events were elicited from adult speakers of 28 diverse languages. Multivariate
statistics show that there is considerable agreement on the ‘semantic space’ underlying
this domain – e.g., all the languages recognize a dimension to do with the agent’s degree
of control over the point of separation in the patient’s material integrity. But there is
variation in how many categories languages recognize, where they place the boundaries between them, and how semantic elements are apportioned to different parts of the predicate.
In the second, developmental phase of the project, a new ‘child-friendly’ set of videotaped ‘cutting and breaking’ events was used to elicit descriptions from child and adult speakers of DUTCH, TAMIL, and MANDARIN CHINESE (3 age groups per language, 10
subjects per group: 4 yrs., 6 yrs., adult). Multivariate statistical and qualitative analyses
reveal that although children identify the relevant construction patterns and approximate the semantic categories of the target language by as early as age 4, they are still shifting and fine-tuning category boundaries even as late as age 6. Discussion will highlight not
only substantive findings, including the ongoing process of semantic category construction, but also the methodological innovations of this study: new techniques for
exploring semantic structure in large crosslinguistic data sets.