What's the connection between implicit learning and autism? Individuals with autism spectrum condition (ASC) are typically characterized by social, communicative, and motor impairments. Since implicit learning is an important mechanism for acquiring social, communicative, and motor skills, it is reasonable to ask: Do the social, communicative, and motor impairments evidenced in those with autism spectrum condition arise, in part, from a general deficit in implicit learning?
Studies have explored implicit learning patterns in ASC. Some studies report impairments, while others report intact performance on a range of implicit learning tasks. There is some reason to question the findings of these earlier studies, however.
First, there is reason to believe that the so-called implicit learning tasks that were administered were actually tapping heavily into explicit thought processes. This is an important point given that the same studies that reported an ASC-deficit did not stringently match ASC and control groups for IQ, potentially disadvantaging those with ASC. We do know that explicit, in contrast to implicit, processes correlate strongly with IQ. We also know that the use of explicit strategies usually change performance on implicit learning tasks . Therefore, differences between diagnostic groups on the so-called implicit learning tasks may be attributable to differences in the explicit rather than the implicit component of the task.
To shed new light on the implicit learning-autism link and overcome the potential limitations of earlier studies, I conducted a study with a team of researchers led by Jamie Brown at the Laboratory for Research into Autism at the University of Cambridge [read full article in The Quarterly Journal of Experimental Psychology here].
We compared the performance of 26 children with high-functioning autism spectrum condition to 26 typically developing children on four implicit learning tasks, each one designed to tap into a different domain of implicit learning. Importantly, we matched the two groups on IQ. And to better identify whether the differences largely rely on explicit or implicit learning processes, we used implicit learning procedures specifically adapted for use with ASC children and that avoid the use of explicit strategies. Our tasks were intentionally constructed so that the learning was probabilistic, which would make it rather difficult to use explicit strategies to figure out the underlying rules. Prior studies used deterministic implicit learning tasks that could be more easily solved by explicit thought processes.
In one task, which has a social element to it, we had participants implicitly sort cartoon faces and characters ('potato heads') into various categories. Another task involved motor coordination, another involved perceptual processing of context, and another task involved implicitly learning an artificial grammar, a task that it has been argued relates to language learning.
We also included a measure of explicit learning (paired- associates learning) to compare explicit learning performance to implicit learning performance. We also had participants complete the Social Communication Questionnaire (SCQ), a reliable index of autistic symptomatology and related that to everyday real-world implicit learning performance. This way we could test whether implicit learning relates to real-life impairments.
What did we find?
We found intact implicit learning in individuals with autism spectrum condition. The equivalence between the ASC group and the typically developing control group was not a consequence of compensation by explicit learning ability or IQ. In fact, there was no equivalence between the two groups in our measure of explicit learning even though there was on our measures of implicit learning. We also found no relationship between implicit learning performance and real-world ASC symptomatology.
We think these findings undermine the argument that a deficit in general implicit learning processes play a key role in the social, communicative, or motor impairment found in those with ASC. Our findings also converge with some recent reports on intact implicit learning in ASC [3, 4, 5, 6] and reports of intact performance on related incidental procedures such as implicit memory and priming [7, 8, 9].
There are reports of impairments in implicit learning in ASC [e.g., 10, 11] but we think that the impairments observed in those studies can be accounted for by the tasks they employed, which may have relied more heavily on explicit processes and therefore disadvantaged the ASC groups that were not matched for IQ.
Based on our results and other recent reports, we propose that processes other than implicit learning disrupt the operations of otherwise intact implicit learning mechanisms of individuals with ASC, impacting negatively on the development of these skills.
What are some of these potential processes? We can think of four possibilities.
One possibility is that real-world implicit learning impairments may result from the propensity for ASC individuals to use explicit strategies to learn about their world. This overuse of explicit strategies by ASC individuals may interfere with the capacity to learn language, social, and motor skills implicitly. Indeed, there is evidence that for implicit acquisition to proceed normally, the learning must not be obstructed by explicit strategies and people with ASC do tend to have an over-reliance on explicit learning strategies.
Another possibility is that the "implicit learning mechanisms" of people with autism are intact but are simply not directed at the same things that typical individuals are. This could occur as a result of well-established differences in ASC attention. For example, from a very early age, autistic individuals tend to attend to mouths more than eyes. This kind of attentional difference could certainly cause a difference in what is learnt implicitly, even when fundamental mechanisms are equivalent. In line with this explanation, autistic individuals have been found to possess considerable skill at distinguishing different pictures mouths but relative deficits in distinguishing different pictures of eyes. In one study, Grossman and Tager-Flusberg  found enhanced performance in people with ASC on a task involving mouth expertise, an area of the face to which ASC individuals typically allocate an unusual amount of attention.
A third possible explanation is that the implicit learning mechanisms are intact in people with ASC but the knowledge derived from implicit learning is not applied successfully. This could explain why in the laboratory people with ASC may display intact implicit learning but not be able to apply their learning in real-world situations. This possibility would be consistent with other research on the dissociation between ability and application in ASC and would be in line with "a recent shift toward understanding ASC in the context of dysfunctions in introspection or self-referential processing" .
Lastly, it is possible that there may be impairments in the long-term consolidation of the implicitly learned skills. Studies have shown the importance of off-line learning for further improvement after implicit learning. We also know that sleep is important for the development of insight from implicit learning episodes. Since ASC is highly associated with sleep difficulties, differences in the consolidation of implicitly learned information may account for some of the ASC decrements in real-life skills associated with implicit acquisition.
Our data combined with data from a number of other researchers point to the conclusion that individuals with autistic spectrum condition can in fact learn implicitly about the world around them and that it is unlikely that these processes are the cause of their real-world impairments in language, social, and motor skills. We look forward to more research on this important topic, with a wider range of individuals on the autistic spectrum and with a wider age range. We believe such research will be important for understanding the cognitive deficits of those with autism and in developing programs to increase their real-world functioning.
© 2010 by Scott Barry Kaufman
 Brown, J., Aczel, B., Jiménez, L., Kaufman, S.B., & Grant, K.P. (2010). Intact implicit learning in autism spectrum conditions. The Quarterly Journal of Experimental Psychology. doi: 10.1080/17470210903536910. [pdf]
 Gebauer, G. F., & Mackintosh, N. J. (2007). Psychometric intelligence dissociates implicit and explicit learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33, 34-54.
 Barnes, K. A., Howard, J. H., Jr., Howard, D. V., Gilotty, L., Kenworthy, L., Gaillard, W. D., et al. (2008). Intact implicit learning of spatial context and temporal sequences in childhood autism spectrum disorder. Neuropsychology, 22, 563-570.
 Kourkoulou, A., Findlay, J. M., & Leekam, S. R. (2009). Local processing of visual context facilitates implicit learning in autism spectrum disorder. Manuscript submitted for publication.
 Muller, R.-A., Cauich, C., Rubio, M. A., Mizuno, A., & Courchesne, E. (2004). Abnormal activity patterns in premotor cortex during sequence learning in autistic patients. Biological Psychiatry, 56, 323-332.
 Travers, B. G., Klinger, M. R., Klinger, L. G., & Mussey, J. L. (2008). Implicit sequence learning in persons with ASD. Poster presented at the International Meeting For Autism Research, London, May 15-17.
 Bowler, D. M., Matthews, N. J., & Gardiner, J. M. (1997). Asperger's syndrome and memory: Similarity to autism but not amnesia. Neuropsychologia, 35, 65-70.
 Gardiner, J. M., Bowler, D. M., & Grice, S. J. (2003). Further evidence of preserved priming and impaired recall in adults with Asperger’s syndrome. Journal of Autism and Developmental Disorders, 33, 259–269.
 Renner, P., Klinger, L. G., & Klinger, M. R. (2000). Implicit and explicit memory in autism: Is autism an amnesic disorder? Journal of Autism and Developmental Disorders, 30, 3–14.
 Gordon, B., & Stark, S. (2007). Procedural learning of a visual sequence in individuals with autism. Focus on Autism & Other Developmental Disabilities, 22, 14-22.
 Grossman, R. B., & Tager-Flusberg, H. (2008). Reading faces for information about words and emotions in adolescents with autism. Manuscript submitted for publication.
 Chiu, P. H., Kayali, M. A., Kishida, K. T., Tomlin, D., Klinger, L. G., Klinger, M. R., et al. (2008). Self responses along cingulate cortex reveal quantitative neural phenotype for high-functioning autism. Neuron, 57, 463–473.