Learning from implicit learning literature: Comment on

overlooked one of the most robust conclusions of the experimental studies on implicit learning conducted during the last decade—namely that participants usually learn things that are different from those that the experimenter expected them to learn. We show that the available literature on implicit learning strongly suggests that the improved performance in Shea et al.’s Experiments 1 and 2 (and similar earlier experiments, e.g., Wulf & Schmidt, 1997) was due to the exploitation of regularities in the target pattern different from those on which the postexperimental interview focused. This rules out the conclusions drawn from the failure of this interview to reveal any explicit knowledge about the task structure on the part of the participants. Similarly, because the information about the task structure provided to an instructed group of participants in Shea et al.’s Experiment 2 did not concern the regularities presumably exploited by the standard, so-called implicit, group, Shea et al.’s claim that explicit knowledge may be less effective than implicit knowledge is misleading. Most studies on implicit learning, the process whereby people learn without intent and without being able to clearly articulate what they learn, have focused on a few prototypical laboratory paradigms (for a brief overview: Cleeremans, Destrebecqz, & Boyer, 1998). Need-less to say, beyond the investigation of these prototypical situations, the ultimate goal of the researchers is to gain knowledge about learning in a wide range of situations, including real-world settings. In keeping with this objective, several attempts have been made in the last few years to use the concepts and methods of implicit learning in order to shed light on some of the problems originating in other areas of research, such as the segmentation of language into words (Perruchet & Vinter, 1998), the second language learning issue (Michas & Berry, 1994)