THE OF ENGLISH THE PHONEMES OF ENGLISH

A PHONEMIC STUDY OF THE AND CONSONANTS OF STANDARD ENGLISH.

by

ANTONIE COHEN

THIRD PRINTING

MARTINUS NIJHOFF• - THE HAGUE - 1971 First printing 1952 Second printing 1965 Third printing 1971

ISBN-13: 978-90-247-0639-6 e-ISBN: 978-94-010-2969-8 001: 10.1007/978-94-010-2969-8

©1971 by Martinus Nijhoff, The Hague, Netherlands All rights reserved, including the right to translate or to reproduce this book or parts thereof in any form CONTENTS

Page

Preface VIII

Preface to the second printing x

Chapter I. INTRODUCTION 1

C hap t e r II.

PROBLEMS CONNECTED WITH PHONEMIC ANALYSIS 14 I. Fundamental Points 14 a. Linguistics as an autonomous science 14 b. Phonetics and Phonemics 16 c. Diachronic v. synchronic method 17 d. Structural relationship between sounds 18 2. Significant Function of Sounds 18 a. The word as a basic unit 18 b. Phonemes 19 c. Commutation Test 20 d. The significant function of place. 22 e. The problem of [hi and [ui . 23 f. "Phoneticism": mono- or biphonematic interpretation. 24 3. Phonemes and their Interrelations 27 a. Identification of the word form . 27 b. Pattern phenomena 28 c. Are phonemes only opposing entities? . 29 d. Martinet's conception of distinctive qualities. 30 e. Neutralization, archiphoneme . 35 4. The Positive factor in the Habitus of the 36 5. Consonant and Phonemes 37 a. How to distinghuish them. 37 b. How to classify them 40 C hap t e rIll.

Page THE CONSONANT PHONEMES OF ENGLISH. 42

A. INVENTORY 1. 'Plosives' 42 2. '' . 43 a. Is [d3} one or two phonemes? . 43 h. Is [3} a phoneme of English? 45 3. 'Nasals' and 'Laterals' . . 46 a. The special nature of [V} 46 h. Clear and dark [I} 48 4. 'Fricatives' . 48 a. The special nature of [r} 49 h. The special nature of [hI 49 5. 'Semi·Vowels' . 51

B. CLASSIFICATION 1. Synchronic v. Diachronic Method 52 2. The Problem of Foreign Words 54 3. Initial Clusters .. . 56 a. The problem of iii . . . . 56 h. The interpretation of voiceless [fl} 58 c. The interpretation of voiceless [m} • . • • . 58 d. Consonants appearing as first aiul second members 59 e. Consonants appearing as first members only • 60 f. C~nsonants appearing as second members only • 60 g. Triphonematic clusters • • • • 60 4. Final Clusten ...... • 61 a. Is. the - a phonemic element in English? 62 h. Biphonematic clusters . 65 c. Triphonematic clusters • 69 d. Quadriphonematic clusters 69 5. Medial Clusten 69 6. Corollary 70 Position Table 72 Table of Consonant Combinations: Initial clusters 73 Final clusters.. • 74 Chapter IV.

Page THE VOWEL PHONEMES OF ENGLISH. 75 1. The so·called 'Short' of 'Checked' Vowels 75 a. Is [a] a phoneme? 75 h. Is [u} a phoneme? 78 2. Long versus Short 78 a. Chroneme theory • 79 h. Are differences relevant? 80 c. Are 'long' and 'short' correlated? • 84 d. 'Silbenschnitt' • 84 e. Other approaches . 85 3. 'Long' Vowels 87 4. 89 a. Some earlier interpretations 89 h. lei} and foul 91 c. [ail and [au} 95 d. [ai} • 97 e. {ju:} • , • , • 99 5. 'Centring Diphthongs' 100 a. One or two phonemes? • 100 h. Phonemic transcription . 101 c. The problem 0/ [a} and [r} 102 d. Triphthongs 104 6. Corollary 105

Appendix. Publications consulted 108 Index 0/ authors . . , • 114 Word index . , . . .. . • 115 Subject index • • • • • • 121 Postscriptum . 129 PREFACE

I gladly take this opportunity to convey my heartfelt thanks to those who have guided me on my way as an undergraduate and who have enabled me through their teachings and friendly advice to proceed to preparing for this doctorate thesis. I should like first of all to thank Prof. C. L. Wrenn, M. A., now of Pembroke College, Oxford, who has always been extremely helpful to me and who was generous enough to admit me to the Honours English Course at King's College, University of London. After moving to Oxford he still found time to show interest in my progress and on more than one occasion helped me with his wise counsels. I am also extremely grateful to his successor at King's College, Prof. G .. Bullough, M. A., who likewhise helped me whenever he could. I feel greatly indebted to Prof. D. Jones, M. A., Dr. Phil., who at the time was Professor of Phonetics at University College, London, and from whose lectures and methods of expression I greatly benefited. I am particularly thankful for the kindness shown to me by the staffs of the English department of King's College and of the Phonetics department of University College for the excellent tuition I received from them and for making me feel completely at home among my English fellow students. I am happy to acknowledge the generosity with which Prof. Dr. P. N. U. Harting as head of the English department of this University received me on my return from England and the helpfulness extended to me, enabling me to complete my course within the shortest possible time. I am grateful to Prof. A. G. van Kranendonk for the confidence placed in me and for his readiness to give advice whenever requested. I consider it a great privilege to be able publicly to express my sincere gratitude to my supervisor, Prof. Dr. A. J. B. N. Reichling, who has been such an inspiring mentor to me throughout the preparation of this thesis. He has patiently helped me with numerous stimulating remarks in what must have seemed to him almost interminable discussions, but which were to me exhilarating hours of profit and encouragement. I have tried to do justice wherever I could in the following text and -notes to the debt I owe him. I am greatly indebted to Mr. A. C. Gimson and Mr. J. D. O'Connor, both of University College, London, for many elucidating remarks in the course of a long correspondence which must have taken up a great deal of their valuable time. I am very grateful to the English department of this University headed by Prof. Harting and assisted by Dr. J. Swart and Mr. R. Vleeskruyer, B. Litt., for having shown such a benevolent interest in my work ahd their helpful suggestions. Finally I wish to thank Mr. J. W. M. Verhaar for his very kind help in indexing the book, and my wife who has been a constant source of encourage• ment and a most able and efficient secretary.

Voorburg, September, 1952. PREFACE TO THE SECOND PRINTING

At the publishers' request a reprint of this study which originally ap• peared in 1952 as a thesis under the supervision of Professor A. J. B. N. Reichling (University of Amsterdam), is now offered. The time elapsed since this request first reached me is considerably longer than the time it had taken to write the thesis in the past, which should be some insIication of the qualms I felt and had to overcome before complying with the request. Nobody with any pretention to a scientific frame of mind would im• mediately agree to have his work reprinted unaltered after such a con• siderable lapse of time. The reason why I have been prevailed upon to take the admittedly unusual way, which was by no means that of least resistance, is that some users of this book have given me to understand that they believe it has an historical interest in giving a reasoned account of phonemic theory up to 1952. The book has met some severe adverse criticism and prospective readers may wish to turn to it first. One American reviewer, G. L. Trager in Language 29, 1953, 564-566, complains that the arguments put forward in Chapters I and II are philosophical rather than scientific and begs to differ on many points about my analysis of the English phonemes as shown at the end of the review where his phonemic listing of American English is compared with mine. A witty, but none the less serious, cri.ticism was raised by A. G. Juilland in Word 10, 1954, 106-109. He reached the conclusion that I had, unwittingly, made out a case for some kind of historical importance in carving out a niche for myself by offering the only attempt at establishing a phonemic analysis with due regard to meaningful distinctions but without taking phonetic substance into account, a. type of approach "nobody had thought or cared to put into practice ..... " (107). A final but no less stinging reference in this masochistic catalogue should be a critical note contained in an article by D. Abercrombie as recently as 1963, in Zeitschrift fiir Phonetik, Sprachw. und Kommuni• kationsf. 16, 1963, 9-12, entitled Pseudoprocedures in Linguistics. In this I was taken to task for having indulged in thought experiments without having taken the trouble of carrying them out. Since over the last five years I have largely occupied myself with thinking out experiments as well as testing them, I feel I can afford to agree that there is much truth in this objection. Yet, I believe that linguistic studies have advanced far enough to allow for a division between theoretical and experimental linguistics and in this respect the 1952 thesis should be considered a contribution to theoretical phonemics. It is obvious that, given this division, it would be highly desirable, if not necessary, for the outcome of the one to be confronted with that of the other ap• proach. In general it should be understood that experimental science hardly ever takes its cues directly from theoreticians. The experimenter on the whole is quite happy in hunting out his own problems and does pot seem to stand in need of getting'his work meted out to him by others; any problem that he tackles usually throws up a whole host of new and unexpected ones and he is therefore never at a loss for work. In the field of phonemic studies I believe that an interpenetration of theory and experiment is the only healthy approach. In effect, phonetics cannot be anything but phonemic, or less paradoxically, studying the sounds of speech is a worthwhile pursuit only in so far as it is made within a framework of phonemic knowledge; on the other hand, phonemics should be clljarly embedded in phonetic knowledge for fear of arriving otherwise at solutions fm: the sake of convenience, economy, or congruity that need not coincide at all with what can be checked as operating in actual language usage. Put differently: the outcome of a purely linguistic, phonemic analysis should be regarded as a hypothesis that stands in need of ex• perimental ve~fication. The choice between the "God's truth" and "hocus pocus" approaches can then be seen for what it is worth: a sham dilemma which need not be taken seriously. However "pure" the methods handled by the investigator in his analysis of the phonemic data may be, he can never be sure of having arrived at any "truth", unless he can verify the outcome of his investigation by experimental proof. The same applies to the systematiser, needlingly called the hocus pocus man, who can take no comfort from the purity of the system he has evolved, unless it can be proved to adhere in some way to the language users whose behaviour he ultimately wishes to account for. But enough of controversy for the moment. I believe that we, that is all persons who take phonemic work to heart, may agree that before 1928, i.e. the year of the first international congress of linguists, phoneticians took a fiendish delight in pointing out that no two speech sounds were ever quite the same. In crude terms: all speech sounds were different. It was to the credit of the Prague phonologists to have stressed the point that some were more different than others, thanks to the concept of the phoneme. From a functional point of view it then seemed that all phonemes were more or less alike. It is largely the Bloomfieldian contribution which has modified this view in pointing up distributional classes. It would seem that now we have reached a stage where it has to be admitted that some phonemes are less equal than others. We should allow for a less rigorous way of looking at phonemic data and not necessarily opt for only one solution, purely on the strength of the logic with which it-has been arrived at. In this respect I will submit, following Chomsky, that the heuristic devices manipulated in this book, as indeed in many contributions of the period in which it was written, need not interest us very much at this juncture of linguistic science. The prime assumption underlying this book, and indeed the binding principle of all linguistic studies, should be given the proper emphasis it needs and which it may not have received at the time of its first appearance. IIi explicit terms it is the assumption that it is feasible to the speech continuum into a number of discrete segments. At the time of its inception this was no more than an implicit assumption. Later work in the field of perceptual speech research, of which the reader will find some reflection in the Postscriptum, have convinced me that this constitutes a very serious and highly interesting problem. Finally it should be observed that the specification of the phonic contents of phonemic units has gone all the way from an original formu• lation in terms of articulation, via the acoustic frame of reference based on spectrographic analysis, to the perceptual domain to which it has always belonged and which supplied even the earliest inventors of alpha• bets with a way of applying the phonemic principle avant la lettre.

Attgust 1964. A.C.