The vast majority of the research done at LEAD revolves around learning and is part, to a greater or lesser extent, of the work done in linguistics, music, graphomotor activities, modeling, etc. A common feature of most of this work, one that is specific to our approach to research, is that we are less interested in analyzing explicit, intentional learning processes, such as those one might encounter in a school setting, than we are in understanding how behavioral modifications subsequent to interactions by the learner with a structured environment. These behavior changes often occur without (verbalizable) awareness on the part of the learner that they have occurred, hence the use of the term “implicit” learning applied to this work. The expression “statistical learning” has come to be used with increasing frequency to designate this type of learning, the emphasis being on the fact that behavioral modifications track the statistical regularities in the material to be learned.

Along with the research in specific, well-established paradigms, such as spoken or written language, some of LEAD research makes use of artificial, laboratory-created situations to study learning. While the generalization of the results from this latter approach to natural behavior in real-world situations is more tenuous, this highly constrained laboratory approach has the advantage of allowing fundamental theoretical issues to be resolved with a high degree of certainty. On of the major theoretical results from this approach was to show that, even though regularities in sequences of stimuli might have been generated by formal (arbitrary) rules, the rules themselves are not learned by participants. Rather changes in the learner’s behavior are observed that show that he or she is sensitive to the statistical regularities that result from using rules to derive the sequence of stimuli.