Multiple Systems of Perceptual Category Learning: Theory and Cognitive Tests

Distinguishing the contributions of implicit and explicit processes to performance of the weather prediction task

“…These findings converge with previous studies of WPT performance that have demonstrated normal WPT performance despite significant delays in feedback delivery (up to 6 sec; Aron et al, 2004;Heffernan & Newell, 2008;Poldrack et al, 2001). Neurobiological theories of category learning (e.g., Ashby & Valentin, 2005;Shohamy et al, 2008) argue that implicit learning depends on reinforcement-mediated dopamine signals that alter the synaptic connections associated with the immediately preceding stimulus and associated response. If reinforcement is delayed, the dopamine release will modify inappropriate synapses, thereby disrupting implicit learning.…”

“…Considerable evidence points to the necessity of striatal activity in both explicit (e.g., Ell, Marchant, & Ivry, 2006;Eslinger & Grattan, 1993;Nomura et al, 2007) and implicit (e.g., Nomura et al, 2007) category learning. Thus, one cannot interpret striatal activation during a category-learning task as evidence that the task specifically demands implicit, or explicit, processes (Ashby & Valentin, 2005). Given accumulating evidence pointing to a necessary role for declarative-learning processes in WPT performance, it is reasonable to conclude that striatal activity during WPT performance reflects a reliance on explicit, rule-based learning.…”

“…One of the most successful multiple-system models of category learning is the competition between verbal and implicit systems (COVIS) model (Ashby et al, 1998;Ashby & Valentin, 2005;. Central to the COVIS model is the idea that the implicit and explicit systems operate competitively and that responding is governed by the system producing the most strongly weighted response.…”

“…More specifically, Experiment 1 manipulated the delay between categorization response and feedback delivery, a modification previously shown to selectively impact implicit category-learning processes (Maddox, Ashby, & Bohil, 2003;Maddox & Ing, 2005). Neurobiological models of category learning (Ashby et al, 1998;Ashby & Valentin, 2005;Shohamy et al, 2008) argued that nondeclarative learning depends on the immediate delivery of feedback to strengthen cue-outcome associations. The regardless of the probabilistically determined feedback provided on that particular trial.…”

Idealness and similarity in goal-derived categories: A computational examination

“…The finding that a similarity-based model cannot account for the typicality gradient in goal-derived categories supports the hypothesis that the human conceptual system is flexible and can accommodate different types of representations and different systems of categorization (e.g., Ashby & Valentin, 2005;Blair & Homa, 2003;Minda & Smith, 2001;Poldrack & Foerde, 2008;Voorspoels, Storms, & Vanpaemel, 2011). Next to similarity-based representations such as exemplar and prototype representations, idealness, too, seems to be an organizing principle of category representations.…”