Learning of rules that have high-frequency exceptions: New empirical data and a hybrid connectionist model

Theorists of human learning, in domains as various as category learning and language acquisition, have grappled with the issue of whether learners induce rules or remember exemplars, or both. In this article we present new data that resect both rule induction and exemplar encoding, and we present a new connectionist model that speciÞes one way in which rule-based and exemplar-based mechanisms might interact. Our empirical study was motivated by analogy to past tense acquisition, and speciÞcally by the previous work of Palermo and Howe (1970). Human subjects learned to categorize items, most of which could be classiÞed by a simple rule, except for a few frequently recurring exceptions. The modeling was motivated by the idea of combining an exemplar-based module (ALCOVE, Kruschke, 1992) and a rule-based module in a connectionist architecture, and allowing the system to learn which module should be responsible for which instances, using the competitive gating mechanism introduced by Jacobs, Jordan, Nowlan, and Hinton (1991). We report quantitative Þts of the model to

[1]  D. Palermo.  On Learning to Talk: Are Principles Derived from the Learning Laboratory Applicable?. , 1966 .

[2]  John K. Kruschke,et al.  Three principles for models of category learning , 1993 .

[3]  James L. McClelland,et al.  On learning the past-tenses of English verbs: implicit rules or parallel distributed processing , 1986 .

[4]  V. Marchman,et al.  U-shaped learning and frequency effects in a multi-layered perception: Implications for child language acquisition , 1991, Cognition.

[5]  R. Nosofsky.  Attention, similarity, and the identification-categorization relationship. , 1986, Journal of experimental psychology. General.

[6]  B. MacWhinney,et al.  Implementations are not conceptualizations: Revising the verb learning model , 1991, Cognition.

[7]  Steven Pinker,et al.  Generalisation of regular and irregular morphological patterns , 1993 .

[8]  L R Brooks,et al.  Perceptual manifestations of an analytic structure: the priority of holistic individuation. , 1993, Journal of experimental psychology. General.

[9]  R M Nosofsky,et al.  Recognition memory for exceptions to the category rule. , 1995, Journal of experimental psychology. Learning, memory, and cognition.

[10]  David S. Palermo,et al.  An experimental analogy to the learning of past tense inflection rules , 1970 .

[11]  S. Pinker,et al.  On language and connectionism: Analysis of a parallel distributed processing model of language acquisition , 1988, Cognition.

[12]  David S. Palermo,et al.  On the learning of morphological rules: An experimental analogy , 1968 .

[13]  John K. Kruschke,et al.  Human Category Learning: Implications for Backpropagation Models , 1993 .

[14]  Roman Taraban,et al.  Categorization by humans and machines , 1993 .

[15]  R. Nosofsky.  Attention, similarity, and the identification-categorization relationship. , 1986 .

[16]  C. Ling,et al.  Answering the connectionist challenge: a symbolic model of learning the past tenses of English verbs , 1993, Cognition.

[17]  R. Nosofsky,et al.  Rules and exemplars in categorization, identification, and recognition. , 1989, Journal of experimental psychology. Learning, memory, and cognition.

[18]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[19]  S Pinker,et al.  Rules of language. , 1991, Science.

[20]  Eric H. Lenneberg,et al.  New directions in the study of language , 1964 .

[21]  S Pinker,et al.  Overregularization in language acquisition. , 1992, Monographs of the Society for Research in Child Development.

[22]  D. Shanks,et al.  Characteristics of dissociable human learning systems , 1994, Behavioral and Brain Sciences.

[23]  Thomas J. Palmeri,et al.  Recognition memory for exceptions to the category rule. , 1995, Journal of experimental psychology. Learning, memory, and cognition.

[24]  Geoffrey E. Hinton,et al.  Adaptive Mixtures of Local Experts , 1991, Neural Computation.

[25]  J. Kruschke,et al.  ALCOVE: an exemplar-based connectionist model of category learning. , 1992, Psychological review.