Transfer in artificial grammar learning: A reevaluation.

Redington, Martin; Chater, Nick

doi:10.1037/0096-3445.125.2.123

Cited by 175 publications

(207 citation statements)

References 40 publications

(160 reference statements)

Supporting

Mentioning

199

Contrasting

Order By: Relevance

“…The seemingly straightforward conclusion that can be drawn from this finding is that participants must have learned something about the deep structure of the training strings, independent of their specific physical appearance. However, Redington and Chater (1996) showed that transfer to a new letter set can be explained solely on the assumption that participants have stored fragments of the training strings and find analogies between these stored fragments and the fragments ofeach test string. To summarize, both designs described in this section are inappropriate to show that rule knowledge is acquired in AGL.…”

Section: Experimental Designs For Investigating Rule-based Processingmentioning

confidence: 99%

Learning artificial grammars: No evidence for the acquisition of rules

Kinder

Assmann

2000

Memory & Cognition

View full text Add to dashboard Cite

Two experiments investigated whether there is evidence for acquisition of rules in implicit artificial grammar learning (AGL). Two different methods were used in meeting this goal, multiple regression analysis and analysis of receiver-operating characteristics (ROCs). By means of multiple regression analysis, several types of knowledge were identified that were used injudgments of grammaticality, for example, about single letters and about larger stimulus fragments. There was no evidence for the contribution of rule knowledge. The ROCswere in accord with a similarity-based account of AGLand thus did not support the notion that rule knowledge is acquired in AGLeither. Simulations with a connectionist model corroborated the conclusion that the results were in accord with a similarity-based, associative account.An important question in implicit learning research has been whether implicit knowledge is stored in terms of the surface features ofthe stimulus environment or in terms of abstract, rule-like descriptions. A paradigm widely used to investigate this question is artificial grammar learning (AGL). In the experiments reported in this article, we applied two different methods to find out whether the knowledge acquired in AGL can be described solely in terms of surface stimulus features or whether an (additional) rulebased process has to be assumed. The first method, multiple regression analysis (e.g., Johnstone & Shanks, 1999), is an excellent tool to investigate the impact of rule adherence as well as several types ofsurface information on performance in AGL experiments. The second method, the analysis of receiver-operating characteristics (ROCs; e.g., Yonelinas, 1997), was applied to investigate the processes underlying performance in AGL.In AGL, strings are presented that were generated according to an artificial grammar (such as the one depicted in Figure I). During training, a subset of all grammatical strings (i.e., the strings that can be generated by means of a particular grammar) are presented. Usually, participants are told that they are taking part in a simple shortterm memory 'experiment and are instructed to memorize the strings. That way, incidental training conditions are provided. Only after the training stage is over are participants informed about the existence of a set of complex rules constraining letter order. Then they are presented with new strings that are either grammatical or nongrammatical.The authors thank David Shanks and two anonymous reviewers for their helpful comments on a previous version of this article. Stimulus materials, simulation details, and simulation parameters used in this article will be made available on e-mail request. Correspondence should be addressed to A, Kinder, Fachbereich Psycho logie, Gutenbergstr. 18, Philipps-Universitat, 0-35032 Marburg, Germany (e-mail: kinder@ mailer.uni-rnarburg.de).Nongrammatical strings violate at least one ofthe rules of the grammar. When participants are asked to categorize these strings, their performance is typically well above c...

show abstract

Section: Experimental Designs For Investigating Rule-based Processingmentioning

confidence: 99%

Learning artificial grammars: No evidence for the acquisition of rules

Kinder

Assmann

2000

Memory & Cognition

View full text Add to dashboard Cite

show abstract

“…Edges are usually considered uninteresting confounds to be controlled for (e.g., Redington & Chater, 1996, but see, e.g., Dienes, Broadbent, & Berry, 1991;Johnstone & Shanks, 1999;Shanks, Johnstone, & Staggs, 1997, who used them as an explanation for generalization in other artificial-grammar-learning experiments). However, as we argue below, edge-based mechanisms are the psychological basis of surprisingly complex generalizations.…”

Section: Edges In Artificial and Natural Grammarsmentioning

confidence: 99%

Primitive computations in speech processing

Endress

Mehler

2009

Quarterly Journal of Experimental Psychology

View full text Add to dashboard Cite

This is the accepted version of the paper.This version of the publication may differ from the final published version. Previous research suggests that artificial-language learners exposed to quasi-continuous speech can learn that the first and the last syllables of words have to belong to distinct classes (e.g., Endress & Bonatti, 2007;Peña, Bonatti, Nespor, & Mehler, 2002). The mechanisms of these generalizations, however, are debated. Here we show that participants learn such generalizations only when the crucial syllables are in edge positions (i.e., the first and the last), but not when they are in medial positions (i.e., the second and the fourth in pentasyllabic items). In contrast to the generalizations, participants readily perform statistical analyses also in word middles. In analogy to sequential memory, we suggest that participants extract the generalizations using a simple but specific mechanism that encodes the positions of syllables that occur in edges. Simultaneously, they use another mechanism to track the syllable distribution in the speech streams. In contrast to previous accounts, this model explains why the generalizations are faster than the statistical computations, require additional cues, and break down under different conditions, and why they can be performed at all. We also show that that similar edge-based mechanisms may explain many results in artificial-grammar learning and also various linguistic observations. Permanent repository link

show abstract

“…Because of the inherent structure of grammatical strings, larger chunks can be used to process them at test compared to nongrammatical strings, resulting in them having higher familiarity and more likely to be classified as grammatical. Other frameworks and models have been proposed to account for AGL that also fundamentally rely on superficial knowledge of chunks rather than rule abstraction (e.g., Dienes et al, 1991;Jamieson & Mewhort, 2009a;Knowlton & Squire, 1994;Redington & Chater, 1996; although see Higham, 1997a). …”

Section: Introductionmentioning

confidence: 99%

Implicit learning of conjunctive rule sets: An alternative to artificial grammars

Neil

Higham

2012

Consciousness and Cognition

View full text Add to dashboard Cite

A single experiment is reported that investigated implicit learning using a conjunctive rule set applied to natural words. Participants memorized a training list consisting of words that were either rare-concrete and common-abstract or common-concrete and rare-abstract. At test, they were told of the rule set, but not told what it was. Instead, they were shown all four word types and asked to classify words as rule-consistent words or not. Participants classified the items above chance, but were unable to verbalize the rules, even when shown a list that included the categories that made up the conjunctive rule and asked to select them. Most participants identified familiarity as the reason for classifying the items as they did. An analysis of the materials demonstrated that conscious micro rules (i.e., chunk knowledge) could not have driven performance. We propose that such materials offer an alternative to artificial grammar for studies of implicit learning.

show abstract

Transfer in artificial grammar learning: A reevaluation.

Cited by 175 publications

References 40 publications

Learning artificial grammars: No evidence for the acquisition of rules

Learning artificial grammars: No evidence for the acquisition of rules

Primitive computations in speech processing

Implicit learning of conjunctive rule sets: An alternative to artificial grammars

Contact Info

Product

Resources

About