A Reinterpretation of Some Earlier Evidence for Abstractiveness of Implicitly Acquired Knowledge

Abstract

Reber and Lewis (1977) exposed subjects to a subset of letter strings generated from a synthetic grammar, then asked them to reorder scrambled letters to generate new grammatical strings. The distribution of the frequency of the bigrams composing their solutions correlated better with the frequency of the bigrams composing the full set of strings generated by the grammar than with the frequency of the bigrams composing the subset of strings displayed in the study phase. In his recent overview on implicit learning, Reber (1989) develops this experimental result into one of the main supports for his contention that studying grammatical letter strings gives access to the abstract structure of the grammar. However, this result can be accounted for by a set of biases inherent to the Reber and Lewis procedure. In the present experiment, a group of subjects learned from a list of the bigrams making up the study strings, a condition which precludes the abstraction of any high-level rules. The pattern of correlations outlined above also emerged in this condition, thus lending support to our re-interpretation.

Independent Judgment-Linked and Motor-Linked Forms of Artificial Grammar Learning

“…The results provide further evidence that higher order as well as pairwise knowledge is learned in artificial grammar learning (Dienes, Broadbent, & Berry, 1991;Gomez, 1997;Gomez & Schvaneveldt, 1994;Perruchet et al, 1992) and extends that literature to show that participants are sensitive to both pairwise and higher order dependencies even on a motor-linked measure of learning. Gomez (1997) found no higher order artificial grammar learning on a typing test; however, in her experiment the spatial characteristics of grammar strings were dis-rupted (only one letter appeared at a time, and the entire string was never seen by participants), which could account for her results.…”

“…Judgment performance was above chance for both kinds of nongrammatical strings, indicating that participants learn higher order knowledge. Furthermore, participants who studied only the bigrams present in grammatical strings and then performed grammatical judgments on whole strings performed worse than participants who studied whole strings, indicating that bigram knowledge cannot account for all of participants' ability to classify grammar strings (Gomez & Schvaneveldt, 1994;Perruchet, Gallego, & Pacteau, 1992;Perruchet & Pacteau, 1990). Willingham (1992) argued that learning new mappings between perceptual and motor events takes precedence over learning sequential information.…”

Episodic knowledge and implicit learning

“…Our primary focus will be on evidence derived from categorization tasks, such as artificial grammars. We will not review the evidence taken to support the traditional abstractive account of implicit learning (e.g., incidental learning and transfer to stimuli with dissimilar characteristics), because the limitations of this evidence have been extensively reviewed elsewhere (Brooks & Vokey, 1991;Perruchet, 1994a;Perruchet, Gallego, & Pacteau, 1992;Vokey & Brooks, 1994;Whittlesea & Dorken, 1993a).…”

Implicit versus explicit learning: Some implications for L2 teaching

“…On the basis of these results, Reber argued that subjects had acquired knowledge that was deep, abstract, and representative of the underlying grammar. However, Perruchet, Gallego, and Pacteau (1992) showed how several artifacts could have produced this pattern. For example, the variance of the distribution in the learning stimuli was much smaller than that in the set of all the possible stimuli generated by the grammar for the string lengths used.…”