Autism is a neurodevelopmental disorder characterized by the early (before 3 years of age, if not earlier) presentation of deficits in social abilities (and in all abilities that depend upon social abilities) and language (delays and/or inappropriate use) and by repetitive behaviours or apparent obsessions. Approximately 70% of individuals with autism are mentally retarded, and nearly 50% of cases never develop useful speech. These disturbances are lifelong, although they may be modified by education, by circumstances, and perhaps by maturation. Autism is surprisingly common, with an incidence of approximately 1/1000. Although first described in 1943 by Kanner (1943) and in 1944 by Asperger (1944), it came under far more intense scrutiny and saw greater public awareness beginning in the 1970s. Milder forms have been recognized, other conditions (such as the general categories of ‘developmental language delays’ and ‘mental retardation’) are now being appreciated as frequently harbouring the diagnosis of autism, and individuals with autism are being more publicly visible, and even in some cases speaking out on their own behalf (e.g. Grandin & Scariano, 1996).

Nevertheless, autism is a confusing condition to many health care professionals. The term ‘autism’ is confusing partly because its characteristic deficits, in social abilities, communication and language use, and in the flexibility and spontaneity of behaviour, are all domains that are often difficult to assess without a detailed history from good observers, and in which a wide range of normalcy (in development tempo or degree of achievement) is generally allowed. The diagnosis of autism is also confusing to many because the term does not really apply to a single condition or even to a spectrum of severity along a single dimension of disease features. Instead, it describes a set of multidimensional clinical entities that differ in both their specific pattern of features and in the severity of each feature. The term can be used to characterize an award-winning mathematician(Baron-Cohen et al., 1999) as well as a mute, severely retarded child who spins in a corner by himself all day long. Whether such cases have a unifying neurobiologic basis is not yet known. Therefore, the diagnosis of autism is still based on an imperfect phenomenology, not on neurobiology.