Simulation of 'Sensori-motor Stage V and VI' Language Development: A Connectionist Network Approach

Connectionist networks, or the so-called neural networks, provide a basis for studying child language development in that these networks emphasise learning, either from observations or from being told, and that the design of these networks simplifies questions related to the representation of linguistic and world knowledge through the use of a network of 'simple' nodes and links. We report a connectionist simulation of this phenomenon focusing on the transition from one-word to two-word language. The connectionist simulation comprises the simulation of concept memory, word lexicon, semantic and conceptual relations and word-order. The data used in the simulation has its origins in the longitudinal psycholinguistic study of infants through their various stages of cognitive development, including sensori-motor (stages V and VI) and pre-operational stages. Bloom's (1973) archives of child language data was used in 'training' the connectionist networks. The concept representation scheme for simulating a child's concept memory is semantic feature oriented (Katherine Nelson, 1973) and the semantic relations between concepts are based on Roger Brown's (1973) analysis. The connectionist simulation was carried out using ACCLAIM - A Connectionist Child LAnguage Development and Imitation Model. ACCLAIM is a hybrid connectionist architecture comprising 'supervised' and 'unsupervised' learning connectionist networks, and takes into account the diverse nature of inputs to and outputs from a child learning language.

[1]  Geoffrey E. Hinton,et al.  Learning distributed representations of concepts. , 1989 .

[2]  L. Smith,et al.  A model of perceptual classification in children and adults. , 1989, Psychological review.

[3]  L. Smith.  Young children's understanding of attributes and dimensions: a comparison of conceptual and linguistic measures. , 1984, Child development.

[4]  George A. Miller,et al.  Language and Communication , 1951 .

[5]  Teuvo Kohonen,et al.  Self-Organization and Associative Memory , 1988 .

[6]  Jeffrey Mark Siskind,et al.  Acquiring Core Meanings of Words, Represented as Jackendoff-Style Conceptual Structures, from Correlated Streams of Linguistic and Non-Linguistic Input , 1990, ACL.

[7]  K. Nelson,et al.  Structure and strategy in learning to talk. , 1973 .

[8]  W. Bechtel,et al.  Connectionism and the Mind , 1991 .

[9]  Maureen Caudill,et al.  Understanding Neural Networks; Computer Explorations , 1992 .

[10]  D. Levine.  Introduction to Neural and Cognitive Modeling , 2018 .

[11]  James L. McClelland,et al.  A distributed, developmental model of word recognition and naming. , 1989, Psychological review.

[12]  Thomas R. Shultz,et al.  Simulating Stages of Human Cognitive Development With Connectionist Models , 1991, ML.

[13]  Mallory Selfridge.  A Computer Model of Child Language Learning , 1986, Artif. Intell..

[14]  Jane C. Hill.  A computational model of language acquisition in the two-years-old , 1982 .

[15]  Ellen M. Markman,et al.  Categorization and Naming in Children: Problems of Induction , 1989 .

[16]  M. Anisfeld.  Language Development From Birth To Three , 1984 .

[17]  Lance J. Rips,et al.  Semantic distance and the verification of semantic relations , 1973 .

[18]  Katherine Nelson,et al.  Taxonomic Knowledge: What Kind and When? , 1992 .

[19]  Richard F. Cromer.  Language and Thought in Normal and Handicapped Children , 1991 .

[20]  Roger W. Brown,et al.  A First Language: The Early Stages , 1974 .

[21]  James L. McClelland.  Connectionist models and psychological evidence , 1988 .

[22]  H. L. Owrid.  Language development from birth to three , 1985 .