Introduction

The superior memory of people with autism has been commented on for over thirty years. Kanner, in his first report on autism, mentioned the ‘excellent rote memory’ of autism children (Kanner, 1943). Clinicians who are familiar with these conditions never fail to come across examples of outstanding memory for particular subjects (such as birthdays, the academic names of fish, and so on). However, early studies that used conventional neuropsychological tests reported impairments on various measures of memory in children with autism. So, for example, Boucher and Warrington (1976; see also Boucher 1978, 1981a) found that performance in memory tasks by participants with autism was similar to that of adults with the acquired amnesic syndrome, showing impaired recall from long-term memory (LTM), combined with relatively intact cued recall from LTM as well as intact short-term memory (STM). Thus, there is a discrepancy between the clinical observations and experimental findings.

One possible explanation for this discrepancy is differences between the individuals observed or tested. The participants with autism in early psychological studies were primarily children with varying degrees of intellectual disability (low-functioning autism, or LFA). More recent studies that examined individuals without intellectual disability (high-functioning autism, or HFA) have reported basically unimpaired performance on tasks thought to measure episodic memory (Bennetto, Pennington & Rogers, 1996; Minshew et al., 1992; Minshew & Goldstein, 1993; Renner, Klinger & Klinger, 2000; Summers & Craik, 1994) (but see Bowler, Gardiner & Grice, 2000).