Speech: A dynamic process

by Nick McCullough, Pierre Divenyi, Martin Minogue

Overview

Speech: A dynamic process takes readers on a rigorous exploratory journey to expose them to the inherently dynamic nature of speech. The book addresses an intriguing question: Based only on physical principles alone, can the exploitation of a simple acoustic tube evolve into an optimal speech production system comparable to the one we possess? In the work presented, the tube is deformed step by step with the sole criterion of expending minimum effort to obtain maximum acoustic variations. At the end of this process, the tube is found divided into distinctive regions and an acoustic space emerges capable of generating speech sounds. Attaching this tube to a model, an inherently dynamic and efficient system is created. In the resulting system, optimal primitive trajectories are seen to naturally exist in the acoustic space and the regions defined in the tube correspond to the main places of articulation for oral vowels and plosive consonants. All this implies that these speech sounds are inherent properties of not only the modeled acoustic tube but also of the human speech production system. This book stands as a valuable resource for accomplished and aspiring speech scientists as well as for other interested persons in search for an introduction to speech acoustics that takes an unconventional path.

Product Details

ISBN-13:	9781501510601
Publisher:	De Gruyter
Publication date:	04/24/2017
Pages:	241
Product dimensions:	6.10(w) x 9.06(h) x (d)
Age Range:	18 Years

About the Author

Rene Carre, Universite Lyon-2; Pierre Divenyi, Stanford University; Mohamed Mryati, GIZ.

Table of Contents

Preface v

Introduction 3

1 Speech: results, theories, models 10

1.1 Background 10

1.2 Speech production 15

1.2.1 Speech production theories 15

1.2.2 Speech production data 19

1.2.3 Speech production modeling 23

1.3 Speech perception 26

1.3.1 A short history of the study of speech perception 26

1.3.2 Models of speech perception 30

1.4 Conclusions 33

2 Perturbation and sensitivity functions 35

2.1 Introduction 35

2.2 Sensitivity functions for process modeling 35

2.3 Resonances and energy in an acoustic tube 36

2.4 From perturbation theory to sensitivity function 37

2.5 Approximating the tube by N sections 44

2.6 Example on a relation between A(x), SF_j(x), and F_i 45

2.7 An example of a non-uniform tube 46

2.8 Conclusions 48

3 An efficient acoustic production system 49

3.1 Algorithm for efficient change of tube shape 51

3.2 Computational results 52

3.2.1 Exploring the F₁-F₂ acoustic plane 52

3.2.2 Behavior of an initially uniform tube constrained only by area limits 58

3.2.3 Toward a more realistic sound-producing tube 66

3.2.4 Starting from a non-uniform tube 72

3.3 Conclusions 72

4 The Distinctive Region Model (DRM) 73

4.1 Introduction 73

4.2 The model 74

4.2.1 The two-region model for the closed open case 75

4.2.2 The four-region model for the closed-open case 77

4.2.3 The eight-region-model for the closed open case 79

4.2.4 The closed-closed tube model 88

4.3 Use of the model to discover primitive trajectories 90

4.3.1 The model with simplified controls 90

4.3.2 From a closed-open to a closed-closed model 92

4.3.3 Deduction of seven primitive trajectories 93

4.4 Efficiency and optimality of the model 96

4.5 Summary and conclusions 97

5 Speech production and the model 100

5.1 The articulatory level 104

5.2 The DRM and vowel production 107

5.3 Vowel dynamics 108

5.4 Consonant production and the model 112

5.4.1 CV syllabic co-production and the model 112

5.4.2 VCV production 116

5.5 Discussion and conclusions 117

6 Vowel systems as predicted by the model 120

6.1 Problems of vowel system predictions 120

6.2 Prediction of vowel systems from DRM trajectories 122

6.3 The phonology-phonetics relation 126

6.4 Conclusions 127

7 Speech dynamics and the model 128

7.1 Characteristics of speech gestures as dynamic phonological primitives 133

7.1.1 The issue of duration range 133

7.1.2 The issue of kinetics 136

7.1.3 The issue of gesture synchrony 137

7.2 Formant transition rates 142

7.2.1 Vowel-to-vowel transition dynamics 142

7.2.2 Consonant-to-neutral vowel transition rates 148

7.3 Discussion and conclusions 153

8 Speech perception viewed from the model 156

8.1 Properties of the auditory system, in a nutshell 156

8.2 Auditory pattern recognition 162

8.3 Multiple auditory objects: information and streams 164

8.4 The dynamics of the auditory system 167

8.5 Is auditory perception optimal for speech? 175

9 Epistemological considerations 182

9.1 The integrated approach 183

9.1.1 Deductive method 183

9.1.2 Iterative process 184

9.1.3 Modeling 185

9.2 Findings 186

9.2.1 A dynamic approach to speech 186

9.2.2 Coding and acoustic phonology 187

9.2.3 The importance of time 189

9.2.4 Complexity 190

9.3 Conclusions 190

10 Conclusions and perspectives 192

10.1 Summary of findings and conclusions 192

10.1.1 Deductive-iterative modeling process 192

10.1.2 Efficiency and optimality 192

10.1.3 Acoustic phonology 193

10.1.4 Dynamic process 194

10.1.5 Explanatory process 195

10.2 Perspectives 195

10.2.1 Extension of the model to speech sounds other than oral vowels and plosive consonants 195

10.2.2 Co-production-deconvolution of gestures 197

10.2.3 Transition slopes and time normalization 198

10.2.4 From iterative to evolutionary process? 198

10.2.5 Practical applications 201

Bibliography 205

Index of terms 224

Author Index 227