Cleeremans, A.
Université Libre de Bruxelles
There is wide agreement around the idea that implicit learning and statistical learning share common mechanisms: Both are instantiations of the fact that people can become sensitive to the associations that may exist between events in the environment through mere incidental exposure to such events. Different kinds of elementary computational mechanisms are capable of capturing such contingencies, starting with the simple Hebb rule. However, and despite this common ground, the relevant literatures have remained largely disconnected from each other. Further, the role that consciousness plays in such phenomena has received surprisingly different treatments in the three domains. Thus, while there is continuing debate about the extent to which implicit learning genuinely involves unconscious learning mechanisms, there is wide agreement around the idea that people?s sensitivity to complex contingencies outpaces their ability to report on their knowledge. In contrast, there is a very little convincing evidence that simple associative learning or conditioning can take place without awareness in humans. Finally, and somewhat surprisingly, the literature dedicated to statistical learning has remained almost mute about the role that awareness plays in such paradigms. Here, I will overview these issues, drawing both on established findings and on recent evidence. I will suggest that associative learning, statistical learning, and implicit learning lie on a continuum defined by the complexity of the relevant contingencies. Noting that awareness cannot simply be turned off in normal participants, I will then speculate that this is the main cause of the observed differences in the availability of the learned knowledge to conscious awareness.