| Abstrakt: |
Most researchers agree on the idea that multilinguals have specific advantages over monolinguals in learning more languages and, more generally, develop specific competencies for expanding and managing their multilingual repertoire. Of particular interest, most notably in the context of the European strive for the upward revaluation of smaller, less used languages, is the development of receptive competences in several genetically closely related languages (cf. e.g. the EuroCom-projects). This contribution presents empirical evidence that, on the one hand, seems to support the claim that multilinguals are more efficient in developing receptive competences in new, previously unlearnt languages. The article focuses on the question how multilinguals use their languages in order to guess the meaning of cognates in unlearnt but genealogically close languages. A series of studies is discussed whose aim is to tap into this process of interlingual inferencing. Different measures for phonological and graphematic distances across languages are established and correlated with the rates of successful cognate recognition in the search for a threshold of string similarity beyond which recognition becomes unlikely. The role of different types of the participants' multilingual repertoires is assessed, and other factors influencing good performance in cognate recognition are identified. The process of interlingual inferencing is discussed as a form of abductive reasoning, and quantitative and qualitative data are analyzed to support the idea that this type of abduction is an essential driving force in receptive multilingualism and language comprehension in general. [ABSTRACT FROM AUTHOR] |
