Logo image
Language Emergence in Development
Book chapter

Language Emergence in Development

Stewart M McCauley, Padraic Monaghan and Morten H Christiansen
The Handbook of Language Emergence, pp.415-436
John Wiley & Sons, Inc
01/07/2015
DOI: 10.1002/9781118346136.ch19

View Online

Abstract

Computational modeling provides a means to directly test the capacity of simple learning mechanisms to give rise to complex linguistic knowledge. This chapter describes two simple, developmentally motivated computational models of language acquisition which instantiate emergentist principles. The first model, Phonotactics from Utterances Determine Distributional Lexical Elements (PUDDLE) captures children's discovery of linguistic units in the speech stream, while the second model, the Chunk‐Based Learner (CBL) simulates learning to use such units and their distributional properties to comprehend and produce speech. The chapter describes the computational architectures of PUDDLE and CBL, summarizes the outcomes of simulations using corpora of child‐directed speech as input to the models. It also reports the results from the application of both models to a single dense corpus of child‐directed speech. The implications of the findings for emergentist approaches to language are also discussed.
Chunk‐Based Learner (CBL) complex linguistic knowledge computational modeling emergentist principles language acquisition language development PUDDLE simple learning mechanisms

Details

Metrics

16 Record Views
Logo image