Abstract

The syndrome of autism was first systematically identified in the 1940's, and has been the focus of a broad range of work since that time. Its symptomatology is seemingly diverse, and involves a rough division between 'personal' and 'nonpersonal' tendencies. In the personal category are difficulties in understanding and interacting with other persons, socialisation, empathy and communication. In the non-personal category are difficulties in adaptability, occasional special abilities, and a wide range of peculiarities in learning, generalisation, pursuit of narrow interests, and so on. Some tendencies, such as peculiarities in the use of language, seem to span both categories. A central question in the theory of autism, therefore, is whether these two categories of impairment share a common pattern or character. It is true that received sets of diagnostic criteria, drawn from empirical observation, give the impression of a split syndrome in which non-personal and personal tendencies occur together but are different in nature. However, we cannot simply trust the language games which inform the presentation of these observations: as Wittgenstein repeatedly argued, forms of words and 'analogies in language' can have the effect of obscuring both similarities and differences between things. The main thesis of the present paper is that there does exist a pattern common to these two areas of impairment, in that both involve dilogical structures