Resumen de Classifier learning and modality in a polyglot 'savant'
Gary Morgan, Neil Smith, Ianthi Tsimpli, Bencie Woll
Christopher (C) is a mildly autistic, severely apraxic, savant who learns languages with surprising ease. Surprising because outside language-learning he has severe difficulties in ‘non-verbal’ cognitive domains, particularly where his visuo-spatial abilities are involved. C's unique language learning skill has been reported on in several previous studies, but his verbal abilities have previously been measured mainly in his learning of different spoken and written languages, that is in the modality of oral language. In earlier work we have also described some aspects of C's learning of British Sign Language (BSL), a language produced and perceived in the visual modality. The present investigation gives further details of C's learning of BSL, by focusing on his use of sign language classifiers, and compares his uneven learning with that of a group of university language honours students. Specifically, although after exposure to an introduction to sign language course, C performed within normal limits on a variety of BSL constructions, he had significant difficulty with the production and comprehension of classifiers, a category that crucially involves the interaction of grammatical and spatial cognition. Here we present two new sets of results of tests of C's understanding and use of classifiers: his comprehension of sentences with differing degrees of spatial and grammatical encoding, and his performance on a placement task involving classifiers. In the first task C's poor performance was worst on those sentences with the most complex demands on spatial cognition. In the task placing objects on a table C performed well with real object placement but could not use the classifier system to represent object placement in sign space. Performance on these tasks as well as C's uneven profile of BSL acquisition is interpreted within an established model of the mind but with modifications to accommodate the interface between language processing and non-verbal, visuo-spatial cognition.
