Implicit learning: Below the subjective threshold

Abstract

and DIANNE BERRY University ojReading, Reading, EnglandIn this review, we consider three possible criteria by which knowledge might be regarded as implicit or inaccessible: It might be implicit only in the sense that it is difficult to articulate freely, or it might be implicit according to either an objective threshold or a subjective threshold. Weevaluate evidence for these criteria in relation to artificial grammar learning, the control of complex systems, and sequence learning, respectively. Weargue that the convincing evidence is not yet in, but construing the implicit nature of implicit learning in terms of a subjective threshold is most likely to prove fruitful for future research. Furthermore, the subjective threshold criterion may demarcate qualitatively different types of knowledge. We argue that (1) implicit, rather than explicit, knowledge is often relatively inflexible in transfer to different domains, (2) implicit, rather than explicit, learning occurs when attention is focused on specific items and not underlying rules, and (3) implicit learning and the resulting knowledge are often relatively robust.

Positive transfer and negative transfer/antilearning of problem-solving skills.

“…Good control performance is unaccompanied by declarative knowledge of the task environment (Berry & Broadbent, 1984;Dienes & Berry, 1997;Dienes & Fahey, 1995), and unaccompanied by self-insight into the processes used to control it (Berry & Broadbent, 1984). In addition, practice can lead to further improvements in the controllability of the system, but not to similar increases in declarative knowledge (Berry & Broadbent, 1988).…”

“…In addition, practice can lead to further improvements in the controllability of the system, but not to similar increases in declarative knowledge (Berry & Broadbent, 1988). Implicit learning theorists (Broadbent, Fitzgerald, & Broadbent, 1986;Dienes & Berry, 1997;Dienes & Fahey, 1995, 1998) also claim that dissociations are found because the two systems operate over different types of knowledge. Typically, in CDCTs, the input-output relations are non-salient and therefore difficult to acquire.…”

“…The former approach suggests that the learning processes underlying skill acquisition are bottom-up and unavailable for conscious inspection (e.g., Berry & Broadbent, 1988;Dienes & Berry, 1997). The latter approach claims that top-down processes (e.g., hypothesis testing and evaluative processes) influence the success of control behaviors (e.g., Burns & Vollmeyer, 2002;Sweller, 1988).…”

Lost in dissociation: The main paradigms in unconscious cognition

“…Dienes & Berry, 1997). These findings have been reported in tasks using sequence learning, artificial grammars, and complex simulated systems in the context of learning, and, now in the context of memory, in tasks of priming and skill performance (for reviews, see, e.g., Augusto, 2010;Cleeremans, 1997;Schacter, 1992;Shanks, 2005).…”

“…(a) Subjects are acquiring and/or applying information about stimuli or features thereof despite conceiving no conscious intention and/or no conscious strategy (control) to do so (the intentionality criterion; e.g., Graf & Schacter, 1985;Schacter, 1987; see also Jacoby, 1991;Jacoby et al, 1992); (b) Subjects are acquiring and/or applying knowledge of stimuli or features thereof without being aware of that fact (the metaknowledge criterion; e.g., Dienes & Berry, 1997;Dienes & Perner, 2002); (b.1) Subjects display an above-chance performance while claiming that they are merely guessing (the guessing criterion: introduced by Cheesman & Merikle, 1984; labeled by Dienes et al, 1995); (b.2) Subjects exhibit a mismatch between accuracy in the performance of a task and confidence in their own performance (the zero-correlation criterion: introduced by Chan, 1991; labeled by Dienes et al, 1995). These criteria in turn require the establishing of thresholds of awareness (Cheesman & Merikle, 1984;.…”