Representing artificial grammars: Transfer across stimulus forms and modalities
Louis Manza, Arthur S. Reber · Oxford University Press eBooks · 1997
This chapter addresses the character of the mental representation that underlies implicit learning. Is the representation best viewed as a distributive (exemplar) or fragmentary system based on characteristics of the physical stimulus display, or is it abstractive in nature and based on patterns of covariation among types of stimulus elements? To answer the question, the discussion presents six experiments using the transfer paradigm in artificial grammar learning. Taken together, the findings support the characterization of artificial-grammar-based knowledge as being represented in a complex multifaceted form. In nearly all experiments, there is evidence to support the fragmentary, distributive, and abstractionist views. The chapter argues that there is no default mode of representation; the learning contexts carry with them particular patterns of constraints and invitations that encourage one or another representation system to be established.