Vision and Language in Cross-Linguistic Research on Sentence Production
چکیده
A classic assumption in the cognitive sciences has been the existence of an invariant cognitive architecture underlying language processing. Perhaps for this reason, the early days of modern psycholinguistics saw little engagement with cross-language data. Indeed, despite the vitality of the discipline in the 1960s and 1970s, it is striking that the language in which experiments were conducted were seldom explicitly considered as an important potential variable of interest. As Cutler (1985) observes, during this period, experiments undertaken in one language could be supported or refuted by experiments in another language, without consideration of whether the particular language could have affected the processes under study. Similarly, the major psycholinguistic textbooks of the day did not make reference to cross-linguistic data or comparative argumentation (see, e.g., Fodor et al. 1974; Glucksberg and Danks 1975).1 Beginning in the 1980s, this picture began to change, as researchers in the field of speech perception began to consider cross-linguistic differences more closely. In the process, the blanket universalist assumption was challenged as it was revealed that certain critical cross-linguistic differences can, in fact, affect language processing (Cutler et al. 1983, 1986, 1989; Mehler et al. 1993, 1996). Mehler et al. (1981) found, for example, that French speakers responded faster in