Toward a phenomenal model of implicit learning

Chapter 1 argues that Implicit Learning Theories usually adopt one of three levels of description: macro-descriptions, phenomenal descriptions, or task-specific descriptions. It is argued that all three levels of description are necessary and that complete understanding of implicit learning will exist only when explanations from all three levels of description are pieced together seamlessly. Chapter 1 reviews several accounts of implicit learning which adopt different levels of description. Chapter 2 presents a series of task-specific work on the flanker task in which it was shown that cross-dimensional contingency learning could occur in some conditions. It is argued that when dimensional maps have unused processing capacity (i.e., when a given dimension has relatively little variation to capture) "cross-talk" between neighboring dimensional maps is possible. It is further argued that the obtained results refuted the theoretical claim of Cohen and Shoup (in press) that "cross-talk" was not possible. Chapter 3 presents a phenomenal framework capable of explaining the implicit learning effects of four commonly used implicit learning paradigms (i.e., flanker tasks, serial learning tasks, control tasks, and artificial grammar tasks). Several advantages of the model are discussed. ^