Memory has been studied in autism (this term is used broadly in this review to refer to all autism spectrum disorders) for more that 50 years, with important and sometimes contradictory findings. Memory in Autism: Theory and Evidence is the first book on this topic, and does an admirable job of presenting a variety of findings that have emerged since the resurgence of interest in memory and autism in the 1990s. It also puts current findings in the context of earlier reports. Most importantly to this reader, it elucidates themes and integrates information across studies. It captures the considerable strengths typically observed in the memories of people with autism, including rote learning, perceptual memory, mnemonic abilities, priming, cued recall, and recognition memory. It describes circumscribed weaknesses which include: free recall of episodic memories, encoding of contextual or relational information, social memory and, in some cases, encoding of semantic relationships. It makes the all important point that, despite early reports of an amnestic syndrome in autism, recent investigations find intact memory in most high functioning individuals with autism on many tasks. Memory in Autism also elucidates the importance of the process by which individuals with autism learn, and how that influences what they remember and retrieve. Finally, this book makes a point that is important not only for how we understand memory, but other aspects of cognition in autism as well: that findings in high functioning individuals with autism are sometimes better understood as differences in cognitive style, or preferred processing modes, than as deficits. What follows is a brief summary of each of the chapters in the book.
Chapter 1, by Gardiner, provides a useful review of the common terminology used in memory research to describe basic memory systems, and encoding and retrieval processes. It also has a helpful discussion of how these different theoretical approaches to memory converge. Gardiner closes the chapter with a reminder that the importance of memory for the individual, and the crux of the import of memory in autism, lies in “the decisions and actions that are taken on the basis of memory, most especially on the basis of conscious experiences of memory, and with the personal and social uses to which memory is put” (p. 17).
Chapters 2 through 6 comprise Part II, which addresses the neurobiology of memory in autism. Chapter 2 by Bachevalier provides an excellent review of: the medial temporal lobe and its memory functions; what we have learned from monkey lesion studies; and how those data relate to neuroanatomical and neurofunctional findings in autism. She summarizes her own work investigating the impact of neonatal lesions in the monkey, which is more relevant than adult lesion data for any investigator studying memory in the context of a neurodevelopmental disorder such as autism. She outlines the role of amygdala lesions in producing anxiety and of the hippocampus in relational memory, both difficulties seen in autism. She makes a compelling argument that damaging the medial temporal region in neonates alters the development of the prefrontal-striatal dopamine system, which generates interesting hypotheses to explain the combination of both social and non-social impairments in autism. She also hypothesizes that the range of severity observed in autism could relate to the size of medial temporal abnormality.
In Chapter 3, Mayes and Boucher review the literature on human adult acquired memory disorders as a route to understanding the parcellation of memory functions in the brain with a special focus on anmestic syndrome. The chapter is valuable for the hypotheses it generates in terms of memory function in autism, particularly that low functioning autism may be associated with extensive medial temporal abnormality and that in high functioning autism (HFA) there is an impairment in “binding” information related to abnormalities in systems used to acquire semantic information. Chapter 4 explores the neuropathology of memory profiles in three conditions: autism, developmental amnesia, and prematurity. It reports evidence for hippocampal abnormalities and selective weaknesses in episodic memory in all three groups, but notes that autism is associated with the mildest form of episodic memory impairment and much greater difficulty with social memory. Indeed, it is difficult to disentangle social from episodic memory difficulties in their investigation. Chapter 5 addresses parallels between memory and emotion processing in autism from a neuropsychological perspective. The authors hypothesize parallel processes in memory and emotion in which low functioning individuals with autism produce “raw’” representations (perceptual memories or physiological emotions) only and high functioning individuals add a logical fact based interpretation to these “raw” data. Neither group converts or integrates these representations into mental states. Disrupted conversion of limbic (hippocampus and amygdala) system representations to medial frontal representations is blamed. Chapter 6 hypothesizes a central role for hippocampal dysfunction in autism. It emphasizes the role of the hippocampus in “multidimensional combinatorial processes” (p. 116) and the creation of deep semantic associations. It links this cognitive hypothesis with abnormalities in serotonin, and proposes a specific gene as the source of these difficulties.
The third section of Memory in Autism focuses on psychological findings. Chapters 7 through 10 discuss three specific areas of memory impairment in high functioning individuals with autism: memory for complex information, memory for episodic contextual information that requires recollection, and memory for social information. In Chapter 7, Williams, Minshew, and Goldstein take a different approach to memory, stating that there is no declarative memory deficit in HFA, but there is a problem with encoding complex information because individuals with HFA do not use organized schemas, do not plan, and are not efficient learners, all of which can affect retrieval as well. They summarize several of their important investigations which have made central contributions to our understanding of this topic. Toichi’s chapter on episodic memory, semantic memory, and self awareness in HFA integrates the author’s own provocative research with previous findings to make a particularly compelling argument that captures strengths in autism as well as dissociating specific areas of weakness. A key finding is: intact organization of semantic knowledge, but an impaired relation between semantic and episodic memory in HFA. Lind and Bowler’s chapter gives an elegant description of impairments in self awareness, representational abilities, and temporal cognition, and the impact they have on episodic memory, which requires integration of complex information, autonoetic awareness, and the ability to reason about time. Webb’s chapter addresses social memory weaknesses. It reviews somewhat equivocal behavioral evidence for weaknesses in social content (eyes, faces) and more generic processing abilities that support social memory (configurational processing). It also presents evidence from neuroimaging regarding differential activation of neural circuitry in the context of social stimuli.
In Chapter 11, Pring reviews savant skills in autism. One of her key points is that although savants tend to develop expertise within a specific domain of experience, many of them do use the data’s inherent structure and organization, and not just rote learning. In this way, savants develop a flexible knowledge of their topic that is akin to the knowledge of an expert. The working memory chapter contains a helpful review of strengths and weaknesses in memory in autism, and introduces the important role of executive functions in memory in autism, an otherwise neglected topic in this volume. Regarding working memory, the authors identify that spatial memory tasks are primarily impaired, and note that they are also inherently more complex in terms of visual attention demands than other working memory tasks. They conclude that attention to task support at the point of memory retrieval, task difficulty, and information-processing complexity are key variables to consider when attempting to unravel apparently contradictory findings in working and short term memory. The chapter on rehearsal and directed forgetting in Asperger syndrome describes deficits in elaborative, not maintenance, rehearsal. It relates that to the episodic memory weaknesses frequently observed in autism, and a lack of spontaneous engagement of the most effective encoding strategies, which the authors link to organizational deficits in autism. Chapter 15 deals with memory, language, and intelligence in low functioning autism. It links impaired language and intelligence to a pervasive declarative memory deficit in nonverbal individuals with autism, while arguing for spared procedural and immediate memory in low functioning but verbal individuals with autism.
Memory in Autism closes with an overview section which contains a chapter on the practical implications of memory in autism. It provides a useful integration of many specific findings described throughout the book. There is also a provocative chapter by Mottron, Dawson, and Soulieres that cautions against the pursuit of deficits in the memory of people with autism. They rightfully point out how much memory performance is entirely intact in autism, and that earlier findings of deficits have in many cases been replaced with an understanding that people with autism may learn information differently, but not worse than typical individuals. The book closes with an excellent chapter by Bowler and Gaigg that integrates several themes from the book. It describes the tendency of people with autism to organize information by perceptual stimulus parameters as opposed to conceptually. It summarizes findings that people with autism tend not to make semantic links or episodically defined clusters that associate data in memory. This issue of the absence of, or differences in, linkages at both the neuroanatomical and neurofunctional levels in autism constitutes an important theme of this book. Some examples include: impaired limbic prefrontal connections; a tendency not to link episodic memories with their contexts or semantic knowledge; and difficulties “binding” and organizing information.
Memory in Autism should stimulate thinking and research. Many of its chapters make hypotheses about possible processes in memory that are provocative and testable. It also is friendly to investigators who are not memory experts. The provision of a chapter on memory terminology, and extensive cross referencing between chapters make the book user friendly. Many chapters also contain helpful summaries. As the editors point out in the preface, the volume is not meant to be read cover to cover, as it can become somewhat repetitive, although even this is informative at times. Many seminal memory papers are cited again and again, but often with a somewhat different interpretation. As a resource for the investigator of memory in autism, or even cognition more generally, Memory in Autism is an indispensible tool.
