Linguistic evolution through language acquisition: Formal and computational models

This is a study of how children acquire language and how this affects language change over generations. Written by an international team of experts, the volume proceeds from the basis that we can not only address the language faculty per se within the framework of evolutionary theory, but also the origins and subsequent development of languages themselves; languages evolve via cultural rather than biological transmission on a historical rather than genetic timescale. The book is distinctive in utilizing computational simulation and modelling to help ensure the theories constructed are complete and precise. Drawing on a wide range of examples, the book covers the why and how of specific syntactic universals; the nature of syntactic change; the language-learning mechanisms required to acquire an existing linguistic system accurately and to impose further structure on an emerging system; and the evolution of language(s) in relation to this learning mechanism. Half-title......Page 3 Title......Page 5 Copyright......Page 6 Contents......Page 7 Contributors......Page 9 1.1 Linguistic theory and evolutionary theory......Page 11 1.2.1 Generative linguistics......Page 13 1.2.2 Language agents......Page 14 1.2.3 Languages as dynamical systems......Page 15 1.2.4 Languages as adaptive systems......Page 18 1.2.5 Languages as complex adaptive systems......Page 21 1.2.6 Genetic assimilation......Page 23 1.3 Methodological issues......Page 25 1.4 What next?......Page 28 References......Page 29 2.1 Features of human language......Page 33 2.2 Signaling systems, innateness and symbolic reference......Page 34 2.3.1 Innate non-symbolic systems......Page 35 2.3.2 Innate symbolic systems......Page 36 2.3.3 Learned non-symbolic systems......Page 38 2.3.4 Learned symbolic systems......Page 39 2.4 Transmission of learned communication systems......Page 40 2.4.2 Signaling systems and communicative accuracy......Page 42 2.4.3 The learning model......Page 43 2.4.4 Willshaw networks......Page 48 2.4.5 Cumulative-Association networks......Page 49 2.4.6 Hebbian networks......Page 52 Scalability......Page 54 2.4.7 Summary of simulation results......Page 56 2.5 The problem of observing meaning......Page 57 2.6 Discussion......Page 58 References......Page 59 3.1 Introduction......Page 63 3.2 The Talking Heads experiment......Page 66 The guessing game......Page 67 The conceptualization module......Page 68 Verbalization module......Page 69 Examples......Page 70 3.3 Tendencies in natural language......Page 72 Semiotic landscapes......Page 74 3.5 Global evolution of coherence......Page 76 3.6 Damping synonymy and ambiguity......Page 78 3.7 Conclusions......Page 81 References......Page 82 4.1 Introduction......Page 85 4.2 Words as memes......Page 86 4.3 Language learning by generalization......Page 87 4.3.1 Feature structures and unification based grammars......Page 88 4.3.2 The principle of generalization-based learning......Page 89 4.3.3 Learning the meaning of a word......Page 91 4.4 Learning a sound–meaning association......Page 92 4.4.1 Learning syntax......Page 95 4.4.2 Evolutionary origins of language learning......Page 99 4.4.3 The analogy with chemistry and DNA......Page 100 4.5 Language change by evolution of word feature structures......Page 101 4.6.1 Language universals......Page 103 4.6.2 Regularity and irregularity......Page 110 4.7.1 The speed of language evolution......Page 113 4.7.2 Constraining the theory......Page 116 4.8 Conclusions......Page 118 References......Page 119 5.1 Introduction......Page 121 5.2 Negotiated communication systems......Page 122 5.2.1 Systematic communication......Page 123 5.2.2 Conventional analyses of structural mappings......Page 124 5.2.3 Exemplars......Page 126 5.2.4 The computational model......Page 127 5.3 Internal representations......Page 128 5.3.1 Formula sets......Page 129 5.3.2 Renaming arguments......Page 130 5.3.3 Combining formula sets......Page 131 5.3.4 Formula sets as meanings......Page 132 5.4.2 Complex phrases......Page 133 5.4.3 Exemplars......Page 134 5.4.4 Combining and modifying exemplar phrases......Page 135 5.4.5 Searching a set of exemplars......Page 136 5.4.6 The origins of phrases......Page 138 5.5.1 Acquiring exemplars......Page 139 Constructing a complex phrase......Page 140 Renaming an exemplar’s variable......Page 142 Creating an empty token......Page 143 5.5.2 Reinforcement and discouragement......Page 144 Discouraging inconsistent exemplars......Page 145 Competition......Page 147 5.6 Computational simulations......Page 148 5.6.2 An example simulation......Page 149 5.6.3 The emergence of agreement......Page 152 5.7 Negotiated systems......Page 154 5.7.1 Introducing formulae......Page 155 5.7.2 Partitioning......Page 156 5.7.3 Empty tokens as delimiters......Page 158 5.7.4 Marking reflexive predicates......Page 162 5.7.6 Overruling the default mapping......Page 164 5.7.7 Equivalent empty tokens......Page 166 Partial regularity......Page 169 Causes of partitioning......Page 170 Strict ordering versus empty tokens......Page 173 Promoting versus inverting systems......Page 174 5.8 Discussion......Page 175 5.8.1 Meanings and learnability......Page 176 5.8.2 Controlling recursion......Page 177 5.8.4 Deficiencies of the model......Page 178 5.8.5 Conclusion......Page 179 References......Page 180 6.1.1 Evolution without natural selection......Page 183 6.1.2 A computational approach......Page 184 6.2 A working model of linguistic transmission......Page 185 6.2.1 Learning......Page 187 Grammatical representation......Page 188 Rule subsumption......Page 189 6.2.2 Invention......Page 191 6.2.3 Summary of simulation cycle......Page 193 6.3.1 Degree-0 compositionality......Page 194 6.3.2 Infinite language, finite means......Page 199 6.4 Bottlenecks and universal grammar......Page 201 6.5 Conclusion......Page 208 References......Page 209 Appendix: Details of rule subsumption......Page 211 7.1.1 The problem of language change......Page 215 7.1.2 The Cavalli-Sforza and Feldman theory of cultural transmission and change......Page 217 7.2.1 One parameter models......Page 219 7.2.2 An alternative approach......Page 221 7.2.3 Transforming NB models into the CF framework......Page 223 7.3.1 TLA and its evolution......Page 224 7.3.2 A historical example......Page 228 A Final Note......Page 234 7.4 A generalized NB model for neighborhood effects......Page 236 7.4.1 A specific choice of neighborhood mapping......Page 238 7.5 Conclusions......Page 240 Appendix: language learning......Page 241 8.1 Introduction......Page 245 8.2 Evolution as hillclimbing......Page 246 8.3 Learning guided evolution......Page 247 8.4 The 'no external reality' thesis......Page 249 8.5 A simulation......Page 250 8.6 Results......Page 254 8.7 Discussion......Page 259 Acknowledgements......Page 262 References......Page 263 9.1 Introduction......Page 265 9.1.2 Linguistic selection......Page 266 9.2 The language acquisition device......Page 267 9.2.1 The (universal) grammar......Page 268 9.2.2 The parser......Page 272 9.2.3 Parameter setting......Page 273 9.2.4 Implementation......Page 277 9.3 Feasible and effective grammatical acquisition......Page 281 9.4 Populations of language agents......Page 287 9.5 Linguistic selection experiments......Page 288 9.6 Coevolution of the LAD and of language......Page 292 9.7 Creolization......Page 296 9.8 Conclusions......Page 304 References......Page 307 10.1 Introduction......Page 311 10.1.1 Characteristics of expression/induction models......Page 312 Learned vocabulary transmission without bottlenecks......Page 317 Vocabulary transmission with only a production bottleneck......Page 318 Vocabulary transmission with only a semantic bottleneck......Page 319 Frequency effects......Page 320 10.2.1 Does true syntax emerge in E/I models?......Page 321 10.2.2 Phases in the emergence of syntax......Page 323 10.2.3 Agents' representations of syntax......Page 324 10.2.4 Representation and mapping of meanings......Page 330 10.2.5 Invention and production algorithms......Page 335 10.2.6 Induction algorithms......Page 339 10.2.7 Bottlenecks......Page 342 10.2.8 Population dynamics......Page 345 10.3 Methodology: What IS a language?......Page 349 10.4 Conclusions......Page 350 References......Page 352 Index......Page 355 Publisher Description (unedited Publisher Data) This Groundbreaking Study Of How Children Acquire Language And The Effects On Language Change Over The Generations Draws On A Wide Range Of Examples. The Book Covers Specific Syntactic Universals And The Nature Of Syntactic Change. It Reviews The Language-learning Mechanisms Required To Acquire An Existing Linguistic System (accurately And To Impose Further Structure On An Emerging System) And The Evolution Of Language(s) In Relation To This Learning Mechanism. Introduction / Ted Briscoe -- Learned Systems Of Arbitrary Reference: The Foundation Of Human Linguistic Uniqueness / Michael Oliphant -- Bootstrapping Grounded Word Semantics / Luc Steels, Frederic Kaplan -- Linguistic Structure And The Evolution Of Words / Robert Worden -- Negotiation And Acquisition Of Recursive Grammars As A Result Of Competition Among Exemplars / John Batali -- Learning, Bottlenecks And The Evolution Of Recursive Syntax / Simon Kirby -- Theories Of Cultural Evolution And Their Application To Language Change / Partha Niyogi -- Learning Guided Evolution Of Natural Language / William J. Turkel -- Grammatical Acquisition And Linguistic Selection / Ted Briscoe -- Expression/induction Models Of Language Evolution: Dimensions And Issues / James R. Hurford. Edited By Ted Briscoe. Includes Bibliographical References And Index. This collection of papers examines how children acquire language and how this affects language change over the generations. It proceeds from the basis that it is important to address not only the language faculty per se within the framework of evolutionary theory, but also the origins and subsequent development of languages themselves, suggesting that languages evolve via cultural rather than biological transmission on a historical rather than genetic timescale. Ten papers include: (1) "Introduction" (Ted Briscoe); (2) "Learned Systems of Arbitrary Reference: The Foundation of Human Linguistic Uniqueness" (Michael Oliphant); (3) "Bootstrapping Grounded Word Semantics" (Luc Steels and Frederic Kaplan); (4) "Linguistic Structure and the Evolution of Words" (Robert Worden); (5) "The Negotiation and Acquisition of Recursive Grammars as a Result of Competition among Exemplars" (John Batali); (6) "Learning, Bottlenecks and the Evolution of Recursive Syntax" (Simon Kirby); (7) "Theories of Cultural Evolution and their Application to Language Change" (Partha Niyogi); (8) "The Learning Guided Evolution of Natural Language" (William J. Turkel); (9) "Grammatical Acquisition and Linguistic Selection" (Ted Briscoe); and (10) "Expression/Induction Models of Language Evolution: Dimensions and Issues" (James R. Hurford). (Papers contain references.) (SM) This is a groundbreaking study of how children acquire language and how this affects language change over the generations. The book is written by an international team of experts and is distinctive in utilizing computational simulation and modelling to help ensure the theories constructed are complete and precise.