segments 9 Combinatorial constraints on phonetic features 10 Segments'in'the speech signal 11 What is a segment?-the speaker focus 13 What is a segment?-the listener focus 14 Phonemes 14 The nature of symbolic representations 16 Phonetic transcription and linguistically relevant information 17 Confusion over what speech sounds 'represent' 18 The domain of speech theory 19 2 Coarticulation: the Phenomenon 21 viii Contents Non-variable targets, variable targets and management 36 The unknown input 38 3 Coarticulation Theory 40 Introduction 40 Characterising the results of coarticulation 41 What must a theory of coarticulation include? 49 The focus of coarticulation theory 53 Theories and models of coarticulation 55 Models of coarticulation-presuppositions 55 Coarticulation-basic theory 57 Symbolic representation and linear context 58 Technical criticism of Wickelgren's approach 59 Lashley's serial ordering approach 60 Ohman's vowel carrier model 61 MacNeilage's three-dimensional spatial targets 61 Holmes' linearly conjoining model 62 Lindblom's coarticulation model 64 Lindblom and Ohman compared 64 Henke's look-ahead model 67 Intrinsic and extrinsic allophones 70 Ladefoged's approach to allophones 71 Classical Phonetics confuses assimilation and coarticulation 72 Kim's 'systematic synthesis' model and coarticulation 74 Daniloff and Hammarberg-coarticulation as feature spreading 75 Keating's model of coarticulation 76 Window setting-target constraints and their domains 82 Coarticulatory resistance and aggression 86 Revisions of the Window Model 87 Coarticulatory resistance 89 Co-production theories 91 Coarticulation vs. co-production 92 Sources of intrinsic variation 95 Inter-segment coarticulatory influence , 96 4 Speech Motor Control 99 Contents xiii Clinical Phonetics-procedures and expert systems: modelling clinical assessment Areas that need further specification Conclusion References Name Index

No abstract

No abstract

No abstract

An experiment in electromyography is described ; data on action potentials obtained from m. orbicularis oris is presented. There is every indication that the EMG signal is statistically insignificantly different in duration and amplitude for initial and final /p/ and /b/ in context with several vowels in monosyllabic words. The data is linked to a provisional theory of production.