![]() ![]() It is surely one of the most closely analyzed and criticized documents in cognitive science, starting with three long critiques in an issue of the journal Cognition, the most influential being Pinker and Prince's ( 1988). The chapter generated an enormous amount of controversy, acting as ground zero for critiques of the PDP approach. Moreover, the model did not simply memorize the patterns on which it had been trained knowledge was represented in a form that supported the generation of past tense forms of novel verbs. The “uniform procedure” by which the model generated past tenses obviated the distinction between rule-governed forms and exceptions. The model performed a version of the WUG task (Berko, 1958): It took the phonological form of a verb's present tense as input and generated its phonologically specified past tense as output. Rumelhart and McClelland ( 1986 hereafter RM) described a neural network model that presented an alternative to this linguistic orthodoxy. Irregular forms exist outside this system of core linguistic knowledge and are learned and generated by other mechanisms such as memorization and association. Phenomena such as tense on verbs and number on nouns have been taken as simple, decisive demonstrations that grammatical rules are an essential component of linguistic knowledge (Pinker, 1999). Third, it is quasiregular (Seidenberg & McClelland, 1989): There is a main pattern but also irregular forms that deviate from it in differing degrees (e.g., keep-kept, run-ran, go-went). Second, it is productive: People can readily generate past tenses for novel forms such as nust-nusted or wug-wugged. First, it is systematic: Most past tenses are formed by adding the morpheme that is spelled - ed and pronounced as in the examples baked, baited, and bared. Why the past tense? Because it exhibits three interesting characteristics. The chapter described the application of some of the tenets of the PDP approach to a sliver of English grammar: the inflection of verbs for the past tense. The title, “On learning the past tenses of verbs in English,” was one that, like “The information available in brief visual presentations” (Sperling, 1960) or “Studies of interference in serial verbal reactions” (Stroop, 1935) gave little hint as to its significance. Chapter 18 in the second volume of the Parallel Distributed Processing (PDP) books created a fuss. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |