Formal Phonology

by András Kornai

Overview

This work, first published in 1995, is primarily addressed to phonologists interested in speech and to speech engineers interested in phonology, two groups of people with very different expectations about what constitutes a convincing, rigorous study. The subject matter, the application of autosegmental theory for Markov modeling, is technical, but not really esoteric – autosegmental theory is at the core of contemporary phonology and Markov models are the main tool of speech recognition. Therefore, it is hoped that anyone interested in at least one of these two fields will be able to follow the presentation.

Product Details

ISBN-13:	9781138321151
Publisher:	Taylor & Francis
Publication date:	04/20/2020
Series:	Routledge Library Editions: Phonetics and Phonology
Pages:	232
Product dimensions:	6.12(w) x 9.19(h) x (d)

About the Author

Andras Kornai is a mathematical linguist. Besides phonology, he studied the formal theory of syntax, particularly finite automata and X-bar theory, and formal semantics, with emphasis on lexical semantics. His new textbook, Semantics, is forthcoming with Springer, where his previous monograph, Mathematical Linguistics, was published in 2007. Kornai is a professor at the Department of Algebra, Budapest University of Technology and Economics, and a member of Academia Europaea. His homepage is at www.kornai.com.

Table of Contents

Preface xi

Introduction xiii

0.1 The problem xiii

0.2 The results xv

0.3 The method xix

0.4 References xxiv

Acknowledgments xxvii

1 Autosegmental representations 3

1.1 Subsegmental structure 4

1.2 Tiers 6

1.3 Association 8

1.4 Linearization 12

1.4.1 The mathematical code 13

1.4.2 The optimal code 14

1.4.3 The scanning code 17

1.4.4 The triple code 21

1.4.5 Subsequent work 24

1.5 Hierarchical structure 25

1.6 Appendix 28

1.7 References 36

2 Rules 39

2.1 Data and control 40

2.2 The rules of autosegmental phonology 44

2.2.1 Association and delinking 46

2.2.2 Insertion and deletion 47

2.2.3 Concatenation and tier conflation 51

2.2.4 Templatic and Prosodic Morphology 54

2.2.5 Domain marking 58

2.3 Automata 60

2.3.1 Biautomata 61

2.3.2 Tone Mapping 63

2.3.3 Vowel projections 64

2.3.4 A reduplicative template 65

2.3.5 The role of automata 67

2.4 Multi-tiered representations 68

2.4.1 Tuple notation 69

2.4.2 Autosegmentalized tuples 69

2.4.3 Tier ordering 70

2.4.4 A distinguished timing tier 72

2.5 Appendix 73

2.5.1 Finite-stateness and regularity 74

2.5.2 The characterization of regular bilanguages 77

2.5.3 Implications for phonology 84

2.5.4 Subsequent work 88

2.6 References 91

3 Duration 97

3.1 The classical theory 100

3.2 The gestural theory 106

3.3 A statistical study 107

3.4 Duration in Markov models 114

3.4.1 The cascade model 116

3.4.2 The tridiagonal model 117

3.4.3 The input model 118

3.5 Markov models as model structures 119

3.6 Appendix 121

3.7 References 128

4 Synchronization 133

4.1 What does feature geometry mean? 134

4.2 Interval structures 140

4.3 The interpretation of large-scale structure 150

4.3.1 Underspecification 151

4.3.2 Hidden variables 153

4.3.3 The interpretation function 155

4.4 Appendix 156

4.5 References 161

5 Structured Markov models 167

5.1 Features in Markov models 169

5.2 The segment-based model 172

5.3 The feature-based model 177

5.3.1 Feature models 178

5.3.2 Building structured models 180

5.3.3 Structured Markov Models and Multi-Layer Perceptrons 182

5.3.4 Expressing feature geometry in structured Markov models 184

5.3.5 Acoustic feature geometry 188

5.3.6 Subsequent work 189

5.4 Conclusions 191

5.5 References 193

Index of definitions 198

Index of authors 200