Connectionist Language Generator

Connectionism has been gaining ground as a psychological modelling technique annd shows great promise as a way to build fast, robust systems to perform intelligent tasks. Connectionism contrasts with the older tradition, that of explaining intelligent behaviour in terms of the manipulation of complex symbol structures. Without expecting either style of research to be exclusively correct, it is still important to seek out the relative strenghts and weaknesses of each. The partisans of the two camps have taken-up the challenge; the controversy has been fierce, as befits the prize-acceptance as the best technique for understanding people and building intelligent systems. Language, the quintessential human activity, is often used as a touchstone for the two approaches. The traditional approach views language as composed of symbols and language use as the manipulatoin of symbol structures. Connectionists have largely conceded to this, accepting the idea of structure as an essential component of language. This text calls into question this common assumption, that "structure is necessary for language modelling", by presenting a generator which produces appropriate natural language utterances without building structures along the way. It backs up this demonstration by an analysis of the generation task, which leads to the conclusion that massively parallel computation and numeric combination of evidence are, in fact intrinsically necessary for generation, and that, conversely, structure building is computationally awkward.

Table of Contents:
Preface Introduction 1 Motivations for this Work 2 Preview of the Model 3 Overview of the Book 2 Design Issues 1 Characteristics of the Generation Task 2 Why Previous Research has Missed the Point 3 Design Principles 4 The Principles in FIG 5 On Decisions, Algorithms, and Modules 3 Le,dcal Knowledge and Word Choice 1 Meanings of Words 2 Inference 3 Syntactic Properties of Words 4 Other Properties of Words 5 Word Choice at Run-Time 6 Summary 4 Syntactic Knowledge and Its Use 1 Motivation 2 Basics of Syntax 3 Two Details  4 An Example  5 Synergy and Competition 6 Issues and Non-issues 7 What Remains to be Done? 8 History of Syntax in FIG  9 Summary 5 Representing and Using Relational Information 1 The Problems with Case Grammars 2 Proposal 3 Participatory Profiles in FIG 4 Implications for Parsing 5 Open Issues 6 Summary 6 FIG's Grammars 1 Some Details of Syntax in FIG 2 English  3 Japanese 7 Details of FIG 1 Building the Network 2 The Input 3 Activation Flow 4 Special Processes 5 Getting the Correct Overall Behavior 6 Summary of Node and Link Types 7 Size and Speed 8 Miscellany Regarding Connectionism 1 Strengths and Weaknesses of Connectionism  2 Why Structured Connectionism? 3 Artificial Intelligence as an Experimental Science 4 Past Connectionist Generation Research 9 Human Language Production 1 Introspection 2 Pauses 3 Priming Effects 4 Errors 5 Traditional Cognitive Models 6 FIG as a Cognitive Model 10 A Model for Natural Translation 1 The Need for Natural Translation 2 Strategies for Machine Translation Research 3 Present Technologies for Machine Translation 4 Proposal 5 Design Implications 6 Philosophical and Software Engineering Issues 7 Prospects 11 In Conclusion 1 How FIG Measures Up 2 Directions for Future Work 3 What Has Been Learned Appendix References Author Index Subject Index

About the Author :
After completing a degree in English at Oxford University, Nigel Ward trained as a theatre director, working professionally in fringe theatre as well as spending a year as an assistant director at the Royal Shakespeare Company. He has taught drama in both the university sector and in conservatoire settings. He was the Course Leader for the M.A. Performance Studies at Central School of Speech and Drama and is currently a Senior Lecturer at Anglia Ruskin University in Cambridge where he lectures in Drama and is Course Leader for Performing Arts.