5hlqiruflqj Yv Lqirupdwlyh Ihhgedfn Zkloh Frqwuroolqj D G\qdplf V\vwhp

Dealing with the issue of implicit learning in dynamic system control, more particularly with the so called Sugar Factory task, we developed (Fum & Stocco, 2003; in press) a model that explains the performance improvement normally shown by participants by relying exclusively on the ACT-R subsymbolic procedural learning mechanism. Our model contrasts sharply with more popular accounts of the phenomenon that assume that people store in memory, and successively exploit, concrete, declarative knowledge of the single interaction episodes (instances) they deal with.