ODG – Salle 166
Robert Kluender (Université San Diego)
The Role of Working Memory in Syntactic Representation and Processing (2/4)
This session changes gears to focus on recent research in the visual working memory literature, which in many ways complements and presents intriguing parallels to advances in verbal working memory research. Here there appear to be two rather divergent, seemingly opposing strains of investigation. One, based primarily on fMRI studies of humans, aims to demonstrate the illusory nature of presumed maintenance functions (at least in terms of neural activation) and to buttress the case for the purported working memory system being nothing more than the interaction of the attention and long-term memory systems. The other is based primarily on ERP studies of humans, and appears to assume both a storage function (similar to the verbal working memory ERP literature, based on the presence of sustained potentials during delay periods) and a visual working memory component of some undetermined sort, but redefines capacity limits in terms of the ability of attentional processes to filter out irrelevant information from the purported storage component. Monkey single-unit recording data in visual working memory tasks have cut both ways: while early studies were compatible with an active storage function, more recent studies show neural activation only during encoding and retrieval phases, and none during delay periods.