1. in_ernet_dli_2015_174095_2015_174095_The-Development-Of-Modern-English

The

Development

of

Modern

English

The Development

of

Modern English

BY

STUART ROBERTSON

Temple University

GEORGE G. HARRAP & CO. LTD.

LONDON BOMBAY SYDNEY

First published February 1936

by George G Harrap & Co Ltd.

182 High Holborn, London, W C 1

Copyright. All rights reserved

PRINTED IN THE U. S. A.

Preface

THIS book attempts to present the historical background necessary for an understanding of the English language as it is spoken and written today. The author has endeavored to introduce the general reader and the student who is beginning the systematic study of English to a selected portion of the mass of facts and doctrine that linguistic scholarship, of the last half-century particularly, has made available. In the belief that, for the general reader and the elementary student alike, the field of contemporary language is of paramount interest and value, the emphasis of the book has throughout been placed on present-day English. It has been the author's purpose, however, to deal with the English of our own day in the light of the perspective that is afforded by a survey of the past. He can merely express the hope that his judgment as to the fitting proportion to be observed between past and present has been reasonably sound.

The question of proportion has been present too in the problem of the relative amounts of space to be devoted to the various specialized fields of linguistic study. Naturally the attempt has been to preserve a just balance among the topics that appear as chapter headings, and not to allow the greater attention that is paid to some of them today (phonetics, for example) to dictate that an undue amount of space be given to these. Among the topics, American English is not so much given a

vi

Preface

separate treatment as made the point of emphasis in

almost every discussion of contemporary English. Hand

in hand with the presentation of facts has inevitably

gone indulgence in speculation on their significance.

The author trusts, however, that fact is presented as

fact and theory as theory, so that the reader, accepting

the one, will at the same time be given the basis for

determining whether the accompanying interpretation

or commentary is worthy of consideration. Into the

attitude toward language that the book sets forth it is

impossible to go here, except perhaps to suggest the

writer's belief that a woeful general ignorance of the facts

is chiefly responsible for the curiously wrong-headed

notions about the standards and sanctions of language

that accompany most discussions of that ever fascinating

topic, one's native speech.

The author's indebtedness to others is indeed vast,

but exceedingly difficult either to measure or to acknowl-

edge. He has of course endeavored to give credit for

all ideas that are not common property, by citations in

the text or in footnotes, but he is conscious that such

acknowledgment is sadly inadequate. The footnotes

have been put to other uses too: to supply additional

illustrations of points made in the text, to suggest

supplementary or additional lines of inquiry, and

occasionally to give references that might be helpful

in pursuing these or other topics. References for addi-

tional reading in connection with the general topics of

each chapter have also been added at the end of the

chapter. These lists consist, for the most part, of books

and articles of recent date and, almost entirely, of matter

in English; they do not pretend to form a complete

bibliography of the subject, but are designed to indicate

Preface

vii

supplementary material of a kind that the reader will

be likely to find useful.

To the works of three men, of three different nationalities, the author feels most deeply indebted: Jespersen,

Bradley, and Krapp. The first named, particularly, is in

large measure, though of course quite innocently, responsible for much of the speculation, as well as for many of

the facts about language that the book contains. The

author can only regret the impossibility of singling out

and acknowledging his debt to many another writer and

work. He must, however, make the usual obeisance,

and in no perfunctory fashion, to that wonderful storehouse of information, The Oxford English Dictionary.

There are nearer and more personal obligations that

should also be recorded here. To his colleagues, the

author is grateful for many expressions of interest and

helpfulness; to two of them, Miss Jane D. Shenton

and Mr. Ellis O. Hinsey, both of whom have read and

criticized the entire manuscript, he would particularly

express his thanks.

S. R.

Philadelphia,

November, 1934.

Contents

CHAPTER

I. Introduction: The Nature and Origin of Language. . . . . . . . . . 1

II. The Ancestry of English . . . . . . . . . 14

III. Old and Middle English. . . . . . . . . . 39

IV. Modern English: British and American . 65

V. English Inflections. . . . . . . . . . 98

VI. English Sounds and Their History. . . . . 162

VII. Contemporary Pronunciation . . . . . . 214

VIII. Spelling and Spelling Reform. . . . . . 263

IX. Sources of the Vocabulary . . . . . . . . 308

X. The Making of Words. . . . . . . . . . 358

XI. The Changing Meanings and Values of Words . . . . . . . . . . 422

XII. Syntax and Usage. . . . . . . . . . . . 491

Index of Subjects and Names. . . . . . . 539

Index of Words . . . . . . . . . . . . . . 547

ix

CHAPTER I

Introduction:

The Nature and Origin of Language

NO completely satisfactory definition of language has ever been evolved. Perhaps the following formulation may be admitted to represent not unfairly the terms that go into the usual definition: Language is the agent or medium by which men communicate their thoughts to one another. Passing over for a moment the difficulty presented by the term that is most vague here (agent or medium), let us attempt to see how far the rest of the formula is applicable. By doing this, we may succeed in getting at certain fundamental conceptions of the nature of language.

Does the word men, in the first place, express a necessary and correct limitation? Is language the prerogative of man, to the exclusion of the lower animals? It is probably safe to say that, although as yet comparatively little is known about the meaning of sounds and gestures made by animals, enough has been learned to make one skeptical of the idea that language should be conceived as a human institution only. At any rate, there is clearly a close parallelism between the cries of animals, obviously expressive of emotions like pain, surprise, and joy, and a good deal of the human utterance that is directly concerned with various sorts of feelings. As we shall see presently, the parallelism is so close,

2 The Development of Modern English

indeed, that it has been held in the past, as one of the theories of the origin of language, that all human speech began with involuntary ejaculations much like the cries of animals. It cannot be denied that most of the higher animals possess at least these elements of language, and it is quite conceivable that some of them—the apes, particularly—have a claim to much more than this.

Other terms of our definition, "communicate their thoughts to one another," raise doubts of more than one kind. To begin with, it is quite certain that everyone does not talk in order to express ideas. Professor Jespersen is surely right in feeling that Mme. de Staël expressed a universal truth about language when she praised the French language in these terms (though perhaps her statement is more true of French than of other languages): "It is not only, as it is elsewhere, a means of communicating thoughts, feelings, and doings, but an instrument that one loves to play on, and that exalts the spirits just as music does for some people and strong drink for others."1 Children discover very early in life, and adults—in civilized as well as in primitive communities—never completely forget, that language may be indulged in for the mere pleasure that the exercise of the vocal organs affords. When a person talks at length to a cat or a dog (even a china one or a mantel-piece, as does a character in one of Barrie's plays), it is quite certain that he does not talk in order to be understood by his ostensible interlocutor, or even, necessarily, to set his own thoughts in order. Quite often he talks, as the infant does when he babbles meaningless sounds or the grown-up when he recites nonsense jingles, for

1 Translated from the French quoted by Jespersen in Mankind, Nation, and Individual from a Linguistic Point of View, p. 7.

The Nature and Origin of Language

3

the pure enjoyment of utterance. Even if we limit

ourselves to the idea of communication with another

adult human being, it by no means follows that thoughts

are being conveyed. It may be, as with the cries of

animals just referred to, mere emotion that is being

expressed and communicated. And, what is perhaps

not so readily perceived, it may even happen that neither

thought nor emotion is being conveyed, but that utter-

ance is used (even though the hearer has no glimmering

of the meaning of what is said) to gratify a desire for

sociability on the speaker's part or to disarm a feeling

of hostility that the other might harbor if greeted by

silence. One may speak, not only to one who under-

stands the same language, but also to a savage who

does not, for no other reason than to banish the uncom-

fortable tension that silence creates. What is spoken,

on either occasion, may have little relation indeed to

the conveying of ideas. Is it not, in spite of this,

properly to be designated language?

Let us accept, however, the premise that, with certain

limitations, the formula stated at the outset (men's

communication of thought to one another) does express

an essential truth about language. We may then

proceed to inquire into what is meant by the agent or

medium used in the communication of thought. Evi-

dently, there are three leading possibilities: ideas may

be expressed by mere gesture, by the visible symbol—

drawn, written, or printed—or by the uttered or spoken

sound. To be sure, language is sometimes used meta-

phorically in a way that fits none of these categories;

we hear, for example, of the “language” of flowers,

or the “language” of architecture or of music. These

are clearly, however, figurative extensions of an idea that

4 The Development of Modern English

in any literal or practical application must be limited to the three varieties of media that have been mentioned.

The first, gesture, is an element that has played, and still plays, its part in language. Primitive man doubtless expressed a variety of ideas by movements of the parts of the body, particularly the head and hands. Moreover, civilized man retains such movements to supplement oral speech, to an extent that obviously varies in part with race and individual but that on the whole is more considerable than may at first appear; witness the difficulty experienced by even the less demonstrative in explaining such a term as concertina or circular staircase without the use of supplementary gesture.

Within certain limits, gesture is admirably expressive: the minatory significance of the clenched fist or the forehead contracted into a frown may be as clear as any words could possibly be. It is true that the meaning of such gestures is not everywhere the same; for example, not all men read the same message in the nod of the head or the clapping of the hands. The former signifies, to most of us, “Yes,” while in some parts of the East it stands for “No”; the latter means applause in the Occident, but in the Orient, a summons. Naturally, the best illustration of a language composed exclusively of gestures, and gestures that have a completely conventionalized and widely understood significance, is the system of communication employed by the deaf and dumb. A less familiar example is the intertribal sign language once widely used as a kind of diplomatic code by American Indians of tribes whose dialects were mutually unintelligible.2

2 At the time of his death on April 30, 1934, General H. L. Scott, former Chief of Staff of the United States Army, was preparing a permanent

The Nature and Origin of Language

5

guage, or something like it, represents social intercourse as it might have developed had men not been endowed with physical organs that are adapted to produce and to hear sounds, or had men not learned to use them for these purposes. For it must not be supposed that what we somewhat misleadingly call the "organs of speech" are primarily intended for that use. We refer, for example, to the lungs as furnishing the breath that is the raw material of speech, or to the tongue or the teeth or the palate as modifying the quality of the sound that issues from the mouth and making one sound distinct from another. These parts of the body, nevertheless, have other and more vital functions—respiratory, gustatory, or digestive—than the linguistic one for which they are, after all, somewhat imperfectly fitted; as someone has wisely put it, the organs of speech are no more properly so called than the knees are called the organs of prayer. Before beginning, however, to examine the use of sound in expressing thought, it will be best to deal with the second agent of the three that have been mentioned, the pictorial or the written symbol.

The development of the visual symbol begins, in all probability, with picture writing, in which a series of pictures or characters represents, in a manner not totally unlike that of the contemporary comic strip—though more crudely, at least in one sense, for the pictures must often have been highly conventionalized—a connected chain of narrative. This period of the pictogram or ideogram—the eye-picture—is eventually succeeded by, or blended into, that of the phonogram, or

record of the language. This work, referred to in the Princeton Herald of July 25, 1930, is said to be almost complete. See also Paget, Human Speech, pp. 130-131.

6 The Development of Modern English

ear-picture, in which a symbol is used for a sound-group. Here it is that we observe the amalgamation of two

of our media for the expression of thought, the symbol and the sound. The next step is the evolu:ion of the

alphabetic symbol,3 standing (approximately) for a single vowel or consonant, from the phonogram; in

the course of this process, the original picture presented by the ideogram is so thoroughly obscured in the con-

ventionalized alphabetic symbol as to be, finally, quite lost. Thus, the alef and the beth of the Old Semitic

writing, originally symbolizing the “ox-head” and the “house,” respectively, are completely conventionalized

in the Greek and Hebrew characters that develop from them, so that, in Greek at any rate, the resulting letters

come to stand for nothing more than the initial sounds of the two words. The first use of the visual symbol,

though it evidently takes place far back in the antiquity of the human race, can scarcely be supposed to be so

ancient as the employment of sound for the communica-tion of ideas. Moreover, vital though the development

of the outlined, the written, and the printed symbol has been to the growth of language, it is clear that the

visible symbol has always been secondary in importance to the sound.

The third agent through which thought may be expressed, the uttered or spoken sound, is to be regarded

as first in significance, even though it cannot absolutely be proved to be first in point of time. “Language”

means, above everything else, speech.4 The etymology

3 For a fuller summary of the development of writing, see p. 264.

4 To avoid a vexatious complication, no attempt is made here to differ-entiate language and speech, as is done by Jespersen, op. cit., pp. 11–23 (following Saussure, Bally, and H. E. Palmer), or, quite differently, by

The Nature and Origin of Language

7

of the word itself is suggestive: our English form comes,

through the French, from the Latin lingua (tongue).

It is being more and more recognized that the fruitful

approach to language is through the study of the living

tongue, and that is almost equivalent to saying the

spoken, as opposed to the written, form of the language.

Whether we discuss contemporary linguistic habits,

or speculate, as we shall now do in reviewing theories

of the origin of language, upon the obscure beginnings of

human speech, the spoken language is the aspect of

expression or communication on which we naturally

center our attention.

The first theory, now completely discredited, is that

which finds the origin of language in a divine fiat. Thus,

Plato, in what is perhaps the earliest extant explanation

of the beginnings of speech, insists that “names belong

to things by nature,” and hence “the artisan of words”

must be “only he who keeps in view the name which

belongs by nature to each particular thing.”5 The

implication is that the original perfect language, which

humans must rediscover or re-create, is the work of the

ruler of the universe, the great “law-giver.” Imper-

fections in human language are thus explained as failures

to discover the original “natural” or divine words.

Curiously parallel to this is the view of the origin of

language that was long the orthodox Hebrew (and

Christian) theory, likewise maintaining that language

originated in a divine act. It was supposed that God

A. H. Gardiner in Speech and Language (Oxford, 1932), p. 62. Unfor-

tunately, to distinguish between speech and language in any of these ways

runs counter to the well established practice of using the words synony-

mously or merely limiting speech (as has just been done) to the more

literal implication of the term; that is, to the spoken aspect of language.

5 Cratylus, New York (Putnam), Loeb ed., p. 31.

8 The Development of Modern English

gave to Adam a language fully developed—this was, of

course, believed to be Hebrew—and that the confusion

of tongues at the building of the Tower of Babel accounted

for the variations in human speech. This explanation,

it is probably unnecessary to add, has long since been

given up, by theologians as by linguists. It is now

believed that man was endowed by his Creator not with

speech itself, but rather with the potentialities of speech.

The theory that language is an evolutionary process is

universally accepted, but what were its ultimate begin-

nings is still a matter of speculation.

Turning to the more modern theories, which agree at

least that language is of human rather than divine

origin, we encounter first what is best known by its

nickname, the "bow-wow" theory. This asserts that

primitive language was exclusively "echoic"; that is,

that its words were directly imitative of the sounds of

nature or of animals. All the word-stock is thought to

have originated in a way parallel to the child's calling a

dog "bow-wow" or a duck "quack-quack." The great

objection to this theory is that it has not been demon-

strated that early or primitive languages are composed

exclusively or in great part of onomatopoetic words;

on the contrary, it is clear that the primitive languages

of savage tribes are largely made up of words that are

quite as conventional as those of civilized peoples. At

best, the "bow-wow" theory can explain the origin of

but a part, and not the largest part, of language. Yet

it seems fair to add that the theory has in the past been

somewhat unjustly derided. Words that are imitative

or at least partly so—for there are many gradations

between the purely imitative and the purely conventional

—do form an appreciable part of the vocabulary of most

The Nature and Origin of Language

9

languages. There are many words that we instinctively feel to be symbolic, or semi-echoic.6 Thus, such English words as battle, roar, and thunder have not perhaps a completely imitative quality, certainly not as compared with hiss, whistle, bang, and crash; yet they approach echoism in a way that the conventional words7 of language do not. If, then, the "bow-wow" theory does not solve the riddle of the origin of language, it does at least help to account for the sounds of many words.

Similarly, other discarded theories of the origin of language may, after all, contain an element of truth. The principal ones are the so-called "pooh-pooh" theory, which derives language from instinctive ejaculatory responses to such emotions as pain or joy; and the "ding-dong" theory, which holds that language began with a mystically harmonious response, on the part of man's hitherto silent vocal organs, to a natural stimulus that was fated thus to call forth its perfect expression-"everything that is struck, rings." The obvious criticism of the interjectional theory is the difficulty of bridging the gap between interjections, which, on the whole, are relatively isolated phenomena in speech, and the main body of language; it has been held, indeed, that it is precisely this chasm that separates animal speech, "exclusively exclamatory," from that of men.8 It seems hardly probable that the theory of interjections

6 Cf. p. 362.

7 As an indication of the utterly conventional character of many, or most, of our words, let the reader ask himself if anything would be gained or lost if the words cat and dog were interchanged; or if blue, provided we were equally familiar with this application, would not serve as well as red to designate the color of fire or of blood.

8 Cf. Grandgent, C. H., "The Why and the How of Speech," Getting a Laugh, Cambridge, Mass. (Harvard), 1924, p. 78.

10 The Development of Modern English

accounts for more than the interjections themselves. The other theory is reminiscent of the Greek belief that words exist by nature, rather than by convention, and that there is a necessary and inherent connection between words and the ideas for which they stand. In its eighteenth- and nineteenth-century phases, this theory (once maintained but later rejected by Max Müller) seems no more acceptable as a complete explanation of the origin of language than it does in its Platonic form.

To this account of theories that have been held in the past as to the origin of speech9 may be added, in brief summary, the speculations of two living students of language. Sir Richard Paget, assenting to the general belief that the earliest form of human communication is gesture, presents, as the full explanation, the theory that

. . . human speech arose out of a generalized unconscious pantomimic gesture language—made by the limbs and features as a whole (including the tongue and lips)—which became specialized in gestures of the organs of articulation, owing to the human hands (and eyes) becoming continuously occupied with the use of tools. The gestures of the organs of articulation were recognized by the hearer because the hearer uncon-

9It should be understcod that many others have been propounded—so many, in fact, that the problem has fallen into some disrepute. Perhaps, however, two other attempts to deal with it may be cited here: Noiré's "yo-he-ho" theory, which postulates that the first words were the sounds produced by the vibration of the vocal cords when men, performing some act of strenuous communal labor, expelled their breath forcibly and repeatedly (referred to by Jespersen, Language, p. 415); and Professor F. N. Scott's somewhat similar hypothesis, which holds that the first vocal utterances were the modifications of ordinary breathing due to certain constrictions—these sounds served first as recognition-signs and later as true communications. Scott's theory of "The Genesis of Speech" was the presidential address at the 1907 meeting of the Modern Language Association of America, and may be found in its Publications and also in Scott's Standard of American Speech, New York (Allyn), 1926, pp. 312-345.

The Nature and Origin of Language 11

sciously reproduced in his mind the actual gesture which had produced the sound.10

According to this hypothesis of “oral gesture,” the significant elements in human speech are the postures and gestures, rather than the sounds; the latter merely serve the purpose of indicating the former.

To Professor Jespersen,11 on the other hand, the sounds are of primary importance. His hypothesis, based in part on study of the language of children and of primitive races but chiefly on the history of language, is that emotional songs were the germs of speech. In particular, he believes that the emotion of love called forth the earliest songs, that these songs—and others called forth by different emotions (a chant of victory, for example, or a lament for the dead)—were inevitably accompanied by what were at first meaningless syllables, and that the circumstance that the same sounds were used on similar occasions brought about the first association of sound and meaning.

The conclusion of the whole matter is that the origin of language is still an unsolved enigma. That no certain answer, however, emerges from all the speculations on the question seems no adequate reason for dismissing it, as is frequently done, as something merely vexatious and fruitless. For, often, such speculations are not merely interesting in themselves, but are also suggestive of significant ideas about language. Nor are the theories all mutually destructive. The “bow-wow” and “pooh-pooh” theories, for example, are admitted as partial explanations of the general problem, in recent attempts

10 Human Speech, p. 174. Reprinted by permission.

11 Language, pp. 412-442.

12 The Development of Modern English

to solve the question of the origin of speech—notably that of Jespersen. It may be added that Professors Greenough and Kittredge, in formulating the general line of a possible future theory, which they suppose will be named the “goo-goo,” are able to include in it what they conceive to be the elements of truth in the two theories just mentioned and in the “ding-dong” theory as well.12

If the how of language must still be held to elude us, there can be no doubt as to the why. Language must have arisen through men’s desire and need for communication with each other. Social necessity is evidently the prime factor. Further, the increasingly elaborate character of language goes hand in hand with an increasingly elaborate social scheme of things. Language is useful almost exactly in proportion to its ability to adapt itself to new circumstances, to new needs for expression. To the invention of the radio, for example, are said to be due literally hundreds of new words and phrases. And behind the radio is the airplane, and behind this the automobile—each with its separate progeny of words. Our own language, English, would be much less responsive to the needs of the world it serves, much less suited to the rôle of a world language, were it not able to adapt itself, quickly and effortlessly, to the changing conditions of present-day society.

References for Further Reading

Bloomfield, Leonard, Language, New York (Holt), 1933. Graff, Willem L., Language and Languages, New York (Appleton), 1932.

12 Greenough and Kittredge, Words and Their Ways in English Speech, New York (Macmillan), 1901, pp. 391-392.

The Nature and Origin of Language

13

Jespersen, Otto, Language: Its Nature, Development, and Origin, New York (Holt), 1924.

Mankind, Nation, and Individual from a Linguistic Point of View, Oslo (Aschehoug), 1925.

Paget, Sir Richard, Human Speech, New York (Harcourt, Brace), 1930.

Pedersen, Holger (trans. by J. W. Spargo), Linguistic Science in the Nineteenth Century, Cambridge, Mass. (Harvard), 1931.

Sapir, Edward, Language: An Introduction to the Study of Speech, New York (Harcourt, Brace), 1921.

Vendryes, E. (trans. by P. Radin), Language: A Linguistic Introduction to History, New York (Knopf), 1925.

CHAPTER II

The Ancestry of English

THERE is no proof that human language springs from a single common starting point, or that it does not. It is, however, universally believed by students of language that most of the languages now spoken in Europe and North and South America and some of those spoken in Asia go back to one speech ancestor—the hypothetical Indo-European, otherwise known as Indo-Germanic and also as Aryan.1 Foreign families of languages that have been most directly in contact with Indo-European include the following groups: Finno-Ugrian, comprising certain non-Indo-European languages spoken in Europe, such as Finnish, Lappish, and Hungarian (Magyar); Altaic, including Turkish, Mongolian, and Manchu; Hamitic, made up of certain African languages, among them Egyptian; and Semitic,2 the best-known members of which are Hebrew and Arabic. In addition, there are, of course, many more distant families of languages, such as the Indo-Chinese, spoken in Tibet, Burma, Siam, and China. The total number of families of languages is in the hundreds, but the great majority

1 The last term is best avoided, since it is sometimes used as a synonym for Indo-European, and again as another designation of one of the branches of Indo-European, the Indo-Iranian.

2 The terms Hamitic and Semitic suggest the old belief that from the different languages spoken by the sons of Noah have sprung the great groups of languages later spoken in the world; in line with this theory, the Indo-European was once called the “Japhetic.”

14

of these are primitive dialects, especially in Africa and America, that have not yet been shown to belong to any of the larger groups. The number of native American (Indian) families alone is thus estimated as more than one hundred and twenty. If more evidence were available as to the earlier forms of these languages, it would doubtless be possible to reduce greatly the number of families of languages; it is even conceivable that fuller knowledge of language generally would group together, as branches of a common unit, all the families of languages now recognized. In the present state of knowledge, however, this is impossible. Any consideration of the ancestry of English, therefore, must begin with the conception of an Indo-European starting point.

Before beginning with the Indo-European theory, however, it may be advisable to inquire into the factors that account for the splitting up of a language into subdivisions and new groupings. The first, evidently, is the general tendency of language to change with the lapse of time, and the fact that changes taking place in two physically separated groups of the same language do not proceed at an even pace, or in precisely the same direction. The second factor, then, has already been suggested: migration of some of the speakers of a given tongue to a new home, particularly when the migration has traversed such a barrier as a high mountain range or a broad river. A glance at the map of Europe will illustrate the fact that natural barriers—the Alps or the Pyrenees, for instance—are often also the frontiers of language. On a smaller scale, the separation of Old English into dialects—a separation, to be sure, primarily to be accounted for because three distinct tribes, the Angles, the Saxons, and the Jutes, are represented—is

16 The Development of Modern English

emphasized by the fact that rivers become boundary lines: the region of the Northumbrian dialect was from the Humber to the Forth; of the Mercian, from the Thames to the Humber; and of the West Saxon, the greater part of the country south of the Thames. A third factor is conquest. An alien tongue has often been imposed upon a conquered people, and it tends, of course, to be spoken in a changed way, with some residuum of the original speech of the people. Conquest, moreover, may work in another way: the Normans afford the remarkable instance of a conquering people twice giving up their own language for that of a conquered race; for they first exchanged their Northern language for the Southern tongue of the region they called Normandy, and eventually they substituted for the French dialect that they thus acquired, the Teutonic speech of conquered Britain.

The Indo-European hypothesis, to which we may now return, holds that in the Later Stone Age there lived a people or peoples speaking a tongue that was the common ancestor of the greater number of cultivated languages now existing in the world. As has already been suggested, this does not necessarily imply either that the original Indo-Europeans were racially a unit or that their speech descendants are racially akin. It does, however, assume that the hypothetical language known as Indo-European was the speech of a people inhabiting a comparatively limited geographical area, and that from this central focal point have radiated all the subdivisions of the Indo-European family of languages. During the last half-century, there have been a number of investigations into the problems of the Indo-European home and of the classification of

The Ancestry of English

17

language that the Indo-European theory implies. These investigations, partly anthropological and archeological, but chiefly linguistic, have altered the old belief that the home of the Indo-Europeans was in Asia, and most probably in the fabled region of the Garden of Eden, between the Tigris and the Euphrates. It is now held that the Indo-European home was in central or south-eastern Europe. Some of the methods used in attacking the problem may be of interest.

The evidence of language is fairly clear as to the climate, the animals, and the plants of the region that the original Indo-Europeans inhabited: everything points to the temperate zone, and in all probability not the southern part of that zone. This is argued largely from the presence in many Indo-European languages of the same words (not, of course, words identical in form, but words recognizable as variants of the same roots) for such ideas as snow, winter, and spring; for such animals as the dog, the horse, the cow, the sheep, the bear, but not the camel, the lion, the elephant, or the tiger; for such trees as the beech, the oak, the pine, the willow, but not the palm or the banyan.3 It is urged, further, that the country could not have had access to the sea: there is apparently no common word for “ocean.”4 If it was on the continent of Europe, it is unlikely, for other reasons, that it was near the seacoast:

3 It must be recognized that, singly, many of these words offer but dubious proof; the names of trees, for example, are likely to apply to entirely different species in different times and places. Note the variety of meaning that sycamore has even in the English-speaking world; and it is said that in Greece the word corresponding to English beech means “oak” and sometimes perhaps “chestnut.”

4 This, however, should perhaps not be so confidently asserted as it frequently is; cf. Hirt, H., Indogermanische Grammatik, Heidelberg (Winter), 1927, Vol. I, pp. 77 and 78.

18 The Development of Modern English

the pre-Hellenic inhabitants of Greece, for example, were not Indo-European, nor were the Etruscans in Italy, the pre-Celtic peoples who inhabited Britain, or the Basques in Spain and southern France; and over the greater part of northern (and of eastern) Europe were the Finno-Ugrians. A recent formulation of the problem and a possible solution is this:

The areas that will satisfy the conditions require a land with a temperate climate, remote from the sea and shut off from other areas. . . . The only area which will satisfy the conditions postulated by the languages is the great area in Europe which includes practically the former empire of Austria-Hungary.5

In greater detail, a single argument may be added to illustrate more fully the way in which the problem of the Indo-European home has recently been attacked. Professor Bender6 observes that almost every Indo-European language shares with its cognates a common word for honey or for an intoxicating drink made from honey. The first stem is Indo-European *melit, the second *medhu. The former appears, for example, in Greek μελι (honey) and μελισσα (bee), Latin mel (honey), and Old English milisc (honey-sweet) and mildeaw (mildew—literally, “honey-dew”). The latter appears, among many other places, in Sanskrit madhu (honey, mead), Greek μεθυ (intoxicating drink), Dutch mede (mead), Old English medu (mead), and English mead. The inference is that the original home of the Indo-Europeans was a land where the honey-bee abounded;

5 Giles, Peter, “Indo-Europeans,” Encyclopedia Britannica, 14th ed., 1929, Vol. XII, pp. 262–264.

6 Cf. Bender, H. H., The Home of the Indo-Europeans, Princeton (Princeton University Press), 1922, pp. 19–20.

The Ancestry of English

19

and none of the Asiatic sites that have been seriously considered by linguists falls within the bee-belt. In ways like this, though with full recognition that this or that single argument may be fallacious, the weight of the evidence has been shown to favor the theory of a central or southeastern European site. Professor Bender, incidentally, favors Lithuania as the specific region, since he argues that "Lithuanian . . . has preserved into modern living speech more of the Indo-European past than any other language on earth," and that "the Lithuanian stock has dwelt in its present location for at least five thousand years, the duration of the Indo-European period, so far as it is known."

Accepting, then, the concept of an Indo-European starting point, we may proceed to examine the classification of language that follows from this theory. No single scheme of classification has met with universal assent, but the one most generally favored has divided the Indo-European system of languages into eight great divisions or groups. These eight divisions are sometimes further grouped into the Asiatic and the European sections or, more frequently, according to the treatment of the initial consonant that appears in Latin centum and in Sanskrit satem, the word for "hundred." According to the latter grouping, the Greek, Italic, Celtic, and Teutonic languages are the centum branches, and the Indo-Iranian, Armenian, Balto-Slavonic, and Albanian, the satem. It is unknown, however, on what cause this distinction depends, and

7 This number is increased to nine if we include Tocharian, a recently discovered language of Central Asia, and a centum rather than a satem tongue; or to ten, if we include the Indo-European section of the Hittite languages, also centum rather than satem.

20 The Development of Modern English

it may be doubted, further, whether it really betokens important early differences between the languages,

especially since the Greek, the most noteworthy of the early languages of the centum group, resembles most

closely in its syntax the Sanskrit, the most noteworthy of the early languages of the satem group. To tabulate

the eight groups in a single series, the order to be merely geographical, seems therefore more satisfactory. This

order is as follows:

1. Indo-Iranian. The Indian branch includes these tongues: the ancient language of the Vedas; Sanskrit,

a closely related literary language; Prakit and Pali, and numerous living dialects of India; and the language

of the Gypsies. The Iranian branch comprises Old Persian, Zend, Modern Persian, and related dialects such

as Afghan and Kurdish.

2. Armenian. A literary language, Old Armenian, in which the Christian books of the Armenians were

written, dates from about 400 A. D. It is thought to have affiliations with the ancient Phrygian. Modern

Armenian dialects are its descendants.

3. Hellenic. The Greek dialects, the most important of which are the Ionic, Attic, Doric, and Æolic, make

up this branch, which has literary memorials dating from at least nine centuries before Christ. The common

literary language developed from the Attic in the fifth century B. C. The dialects of modern Greece preserve

some of its features.

4. Albanian. This is the language of ancient Illyri-

cum, and that of modern Albania.

The Ancestry of English

5. Italic. The two branches consist of the Latin and the non-Latin dialects; of the latter, the Oscan and the Umbrian, known only from inscriptions and place-names that antedate the Christian era, are to be distinguished. Latin, from the ancient dialect of Latium, became the literary language of Rome. Among its modern descendants are French, Provençal, Italian, Spanish, Portuguese, and Roumanian.

6. Celtic. This group is usually divided into three branches: Gallic, Gaelic, and Cymric (or Brittanic). Gallic is the little-known speech of the ancient Gauls, whom Cæsar conquered. Gaelic includes the Irish, with a medieval and modern literature, interest in which has been studiously cultivated in recent years; Manx, a dialect still spoken in the Isle of Man; and Scotch-Gaelic, known from about the eleventh century and still in some use in the Highlands of Scotland. Cymric embraces Welsh, known as a literary language from the Middle Ages; Cornish, which became extinct in the nineteenth century; and Breton, or Armorican, a dialect of northwest France.

7. Teutonic (otherwise known as Germanic or Gothonic). The three branches8 of this group are the East Germanic, the North Germanic, and the West Germanic. East Germanic is preserved in the fragmentary translation of the New Testament by Bishop Ulfilas, made about 350 A. D. and hence the oldest text in a Teutonic language. The East Germanic has no modern descend-

8Some scholars prefer a twofold division: East and North Germanic, and West Germanic. For the reasons for this, cf. Pedersen, Linguistic Science in the Nineteenth Century, pp. 30 and 31.

22 The Development of Modern English

ants. North Germanic comprises the Scandinavian languages, represented in both medieval and modern literature; an easterly division includes Danish and Swedish, and a westerly, Icelandic and Norwegian. To the West Germanic branch, from which English stems, we shall return.

8. Balto-Slavonic. The Baltic group consists of Old Prussian, Lettic, and Lithuanian. The Slavonic group is further broken up into two subdivisions: one includes Polish and Czech, or Bohemian; the other, Russian and Bulgarian.

What, then, are the common features of the Indo-European family of languages generally, and what are the special differentia of the Teutonic group? In answer to the former question, it is to be observed, in the first place, that the Indo-European languages are all inflectional (though not all inflectional languages are Indo-European). That is to say, such syntactical distinctions as gender, number, case, mood, tense, and so forth, are usually indicated by varying the form of a single root. Thus, in English inflection, we most frequently use different endings—the -s, for example, to differentiate the plural number of nouns from the singular, and the -ed to differentiate the past tense of verbs from the present—though, as we shall see, initial and internal inflection are also possible. Languages outside the Indo-European system that are not inflectional are usually (1) isolating, like Chinese, in which only monosyllabic and invariable roots are used, and in which the relation between words must be suggested by their position or occasionally by variation in tone; or (2) agglutinative, like Turkish or Hungarian, in which formal affixes are

The Ancestry of English

23

attached to independent and invariable roots in such a

way that root and affix are always distinct. A fourth

type of language, called incorporating, is sometimes

differentiated. In such a language (that of Greenland

is an illustration), a single word may express not only

subject and verb, but also such other concepts as direct

and indirect object; hence, what would be in an isolating

language a sentence of five or six words may become a

single word.9

Further criteria of the Indo-European languages are

similar habits of pronunciation and the possession of a

common word-stock. As to the latter, it is clear even

to the novice that, for example, Greek νύκτος,10 Latin

noctis,10 German Nacht, French nuit, and English night

are variations of a single original root. What is implied,

of course, is that their relation is that of cognates rather

than that one language has borrowed from another.

Further examples of common roots appearing in variant

forms in different Indo-European languages are these:11

English Dutch German Gothic Lithuanian

three drie drei thri tri

seven zeven sieben sibun septyni

me mij mich mik manen

mother moeder mutter — moter

brother broeder bruder brothar brolis

Celtic Latin Greek Persian Sanskrit

tri tres treis thri tri

secht septem hepta hapta sapta

me me me me me

mathair mater meter matar matar

brathair frater phrater — bhratar

9 Cf. Pedersen, op. cit., pp. 99 and 100.

10 The forms given are, of course, those of oblique cases, in order to

show the root.

11 From Whitney, W. D., Language and the Study of Language, New

York (Scribner's), 1867, p. 196.

24 The Development of Modern English

There are, however, special characteristics that the Teutonic languages have in common and that serve to differentiate Teutonic, as a branch, from the other divisions of Indo-European. The principal ones are usually thought of as four in number, as follows:12

1. A simpler conjugation of verbs, including a twofold classification.

2. A twofold declension of adjectives.

3. A fixed, stress accent.

4. A regular shifting of the explosive, or stopped, consonants.

Each of these points requires fuller explanation.

As to the first point, even a slight acquaintance with Latin and Greek makes it apparent that their conjugation of verbs is far more complex than is that of German or English. The terminology sometimes used should not be allowed to obscure this contrast. The fact that the phrase I had loved or ich hatte geliebt may be called a pluperfect tense does not mean that it really parallels, in structure, amaveram: the Teutonic verbal system is largely analytic, expressing the concept of pluperfect time, for example, by the use of pronoun, auxiliary, and participle; the Latin is entirely synthetic, expressing a similar concept by elaborately varying the form of the verbal root. The student of Latin who begins the study of Greek finds that the verbal system of the latter, being

12 It should perhaps be added that the four chief changes from Indo-European to Teutonic give a very incomplete picture of the whole process; cf. Hirt, H., Handbuch des Urgermanischen, Heidelberg (Winter), 1931-

nearer the complexity of the Indo-European verb, has at least one more complication in almost every category: there are not only the two numbers of Latin, singular and plural, but there is also a third, the dual; there are not only the two voices, active and passive, but there is also a third, the middle; there are additional moods, like the optative, and additional tenses, like the aorist.

In sharp contrast to this is the exceedingly simple system of Teutonic, characterized at first, for example, by only two tenses, the one to express past time, the other to express both present and future. The inflected passive voice of Teutonic, likewise, was early lost; it is represented in Old English by only a single form, hātte (was called),13 the passive voice elsewhere being expressed, as in Modern English, by compound tense-forms.

The most distinctive feature of the Teutonic verb, however, is its development of a special type of preterit in many words, and the consequent classification of its verbs into two categories according as they follow, or do not follow, this pattern. The preterit is that with the dental suffix, the -ed of English and the -te of German; the resulting classification of verbs is the division into two types, called “weak” and “strong.” The weak verbs, less accurately called the “regular,” are those with the dental preterit, as walk, walked; the strong or irregular are those with internal vowel change, as sing, sang, sung. The latter group corresponds more nearly to the verbs of other languages, but the former is distinctively Teutonic. We have come to feel, as the designation “regular” suggests, that the conjugation of the weak verbs is not only the simpler but the normal

13 This inflected passive survives in the poetic form hight (was called).

26 The Development of Modern English

pattern; newly created or newly borrowed verbs invariably are conjugated in this way.

To sum up, then, we have in the Teutonic languages a twofold classification of verbs, and in general a verbal conjugation that is unlike that of other languages; and of these two types of verbs, the weak is more distinctively, indeed quite uniquely, Teutonic. The fact that in giving the principal parts of a weak verb in Modern English only two forms14 need be mentioned illustrates the extreme simplicity that the Teutonic verb has attained.

The second distinguishing characteristic of the Teutonic languages is the twofold declension of the adjective. The adjective was declined according to one system when it stood alone before the noun or was used in the predicate relation, and according to another when used substantively or when it was preceded by a defining element, such as an article or a demonstrative. This is not to be illustrated in Modern English, which has lost all declension of the adjective, and thus, necessarily, the distinction between the weak and the strong forms. It was, however, preserved in Old English and is still to be found in German. Thus, in Modern English, we use the identical form good in the expressions good men and these good men. In Old English, however, the corresponding phrases would be gōde menn and þās gōdan menn—as in German the corresponding phrases are gute Männer and diese guten Männer. As with the verbs, it is the weak inflection rather than the strong that is distinctively

14 Compare, for example, the six principal parts of a typical Greek verb. But it is true, of course, that for the strong verbs in Modern English three forms must often be given (and in Old English, four).

The Ancestry of English

27

Teutonic, the strong forms corresponding more or less to adjective declension in other languages. The loss of the distinction in Modern English need not, of course, be mourned; it is difficult to see any gain in expressiveness in the peculiar inflectional complexity. Nevertheless, its former presence in English and its retention in German are significant in marking these languages as unmistakably Teutonic.

The third characteristic of the Teutonic languages is their accent, fixed rather than free or variable, and a matter of stress rather than of pitch. To English-speaking persons it may seem, at first sight, that accent in any language must necessarily be fixed, as the Teutonic accent is. But it is clear that the original Indo-European and the primitive Teutonic accent was variable; for example, it might shift, in the inflection of a word, from the root syllable to a syllable of the inflectional ending.

Greek and Latin partially preserve this, as when the Greek πoûs (foot) is declined πoδós, πoδί, πóδα; the accent, that is to say, shifts in the genitive and dative from the root to the inflectional ending, and comes back, in the accusative, to the root. Latin is again a step simpler, for the cognate word is declined pēs, pédis, pédī, pédem. However, the freer accent of Latin, as compared to the Teutonic languages, is sufficiently obvious; we may take the conjugation of the present indicative of amō, for example, and observe that in ámō, ámās, ámat, amámus, amátis, ámant, we have a similar phenomenon. Or note how in the inflection of gubernator, the accent is shifted from a root syllable in the nominative gubernátor to the syllable of the inflectional ending in the genitive gubernatóris. We should be doing something equivalent in English if we spoke of "the go\uvernor," and then pro-

28 The Development of Modern English

ceeded to speak of "the goveŕnor's house." But this, of course, or anything remotely like it,15 we never do.

The other aspect of the Teutonic accent is even more important: many other Indo-European languages retain something of the original musical, or pitch, accent; in the Teutonic languages, accent has become, to all intents, entirely a matter of stress. The implications of this distinction are considerable. The accentuation of English and German is utterly unlike that of French, which in words of two syllables has more nearly a hovering or distributed accent than a firm stress on one syllable. Such pronunciations as weékénd, bóokcáse, and saúcepán are the exception in English; in most cases, we have come to overemphasize, as compared even with German, the accented syllable and to underemphasize the unaccented.16 Perhaps this is the chief reason why French as pronounced by an English-speaking person, who can scarcely help carrying over his usual habits of accentuation, usually sounds so curiously unreal. The versification of Teutonic languages likewise depends on this distinction between the stressed and the unstressed syllable, and its effect is utterly different from that achieved by the quantitative system of the classical languages and its partial retention in the Romance

15 It may be objected that we have shifting accent, in Modern English, in pairs of words like pérfume (n.), perfúme (v.), or in classical derivatives like phótograph and photógrapher. But these are evidently far removed from an accent which shifts in the inflection of a single part of speech.

16 For this reason, it is frequently said that in the pronunciation of Modern English only those vowels that are protected by the accent have their peculiar quality, all others approximating either [ə] or [ɪ]. The dictionaries do not always indicate this, but observation makes it evident that the unaccented vowel in subject (noun) is [ɪ] rather than [ɛ], and in actor [ə] rather than [ɔ]. Again, observe the difference in the second syllables of weekend and weakened; the former is the exception, the latter the rule.

The Ancestry of English

29

languages. Finally, perhaps the most important effect,

for the development of English, of the heavy Teutonic

stress accent has been that of slurring and frequently

altogether dropping unstressed vowels; because the

accent usually falls on the first syllable of the word and

because our inflection is chiefly in the final syllables,

it has been easy to lose inflectional endings and conse-

quently to simplify the whole system of inflection.

Here is the chief reason, too, for the largely monosyllabic

quality of the native English word-stock.

The fourth distinctive mark of the Teutonic languages

is their almost regular shifting of certain Indo-European

consonants. The regularity of these changes was per-

haps first stated by the Danish linguist, Rasmus Rask,

but their later formulation by Jacob Grimm has given

the principle the designation “Grimm's Law.” A full

statement of the law and its subsequent modifications

would be out of place here. The formula in general,

however, is as follows:

bh dh gh

b d g

p t k

f th h

The three vertical columns are, approximately, the

labials, the dentals, and the palatal-velars. The first

horizontal row consists of the voiced aspirate stops; the

second, of the voiced stops; the third, of the voiceless

stops; and the fourth, of the voiceless continuants.

The formula is to be read in a downward direction:

each consonant (except those in the last row) changes

to the one directly beneath it. Stated briefly, then,

the principle is this:

1. Indo-European voiced aspirate stops lose their

aspirate quality and appear in Teutonic as voiced stops.

30 The Development of Modern English

2. Indo-European voiced stops lose their voiced quality and appear in Teutonic as voiceless stops.

3. Indo-European voiceless stops lose their momentary quality and appear in Teutonic as continuants.

To proceed, then, to illustrations of the working out of the principle. There is the initial difficulty that, whereas Latin or Greek is commonly taken to represent the unshifted or Indo-European consonants, and English to represent the shifted or Teutonic consonants, the former languages do not always give the Indo-European consonants in their original state. Not one of the consonants in the first row—bh, dh, and gh—appears in Latin or Greek.17 Our illustrations, then, if we are to confine them to Latin and English words, must begin with the second row. Even here, since initial b was exceedingly rare in Indo-European, clear-cut examples are wanting for the first pair, b-p. The other five changes may, however, be represented by many examples. In the following pairs, the Latin word18 gives us the Indo-European consonant; the English word, the Teutonic equivalent. In most cases, not only the initial but the medial or final consonants illustrate the shift.

p-f d-t t-th

pedem (foot) duo (two) tu (thou)

piscem (fish) edere (eat) tonare (thun(d)er)

patrem (father) dentem (tooth) tres (three)

17 Bh, for example, has already been shifted in Greek to φ, in Latin (initially) to f; Greek φρατήρ and Latin frater are cognates of English brother, and Greek φέρω and Latin ferō of English bear.

18 Nouns are given in the accusative case, since such forms as pēs do not disclose the root.

19 Excrescent consonants are thus indicated.

The Ancestry of English

31

g-k k-h

genu (knee)20 cordem (heart)

[co]gno[vit] (know)20 canem (houn(d))

agrum (acre) cornu (horn)

Later philologists than Grimm found it necessary to qualify and modify his statement of the consonant-shift.

The second, or High German, shift—important as explaining certain regular differences between the consonants of English and German in cognate words—originally formed part of “Grimm's Law,” but it has been shown21 that Grimm was mistaken in stating the whole formula as one process.

This second shift will therefore be mentioned later, in speaking of the relation of English and German.

The principal qualification in the law explaining the first consonant-shift must, however, be referred to here.

Another Danish linguist, Karl Verner, discovered that many apparent exceptions to Grimm's Law were dependent on a principle of accent.

His postulate was that when the consonant-shift took place, it found the early Teutonic stress on the same syllables as in Indo-European and that it was this stress on many medial or final syllables that accounted for the presence of voiced instead of voiceless consonants.

Thus, if the preceding vowel in the Indo-European and early Teutonic word did not bear the accent, Indo-European p, t, and k became b, d, and g in Teutonic rather than f, th, and h.

To summarize, the primitive Indo-European voiceless stops first became pre-Teutonic voiceless continuants and then, if they followed an unaccented vowel, they became voiced stops.

On the other hand, initial consonants and consonants preceded

20 The initial k in these words was formerly sounded.

21 Cf. Jespersen, Language, p. 44.

32 The Development of Modern English

by an accented vowel underwent only the shifting explained by Grimm. The voicing just referred to took place, of course, before the accent in Teutonic was universally thrown on the first syllable. A complete formulation of Verner's Law is impossible here, but a few illustrations of its application may be given. Teutonic sometimes has d when, according to Grimm's Law, th should be expected, as the shift from Indo-European t; thus, English old (Old English eald) is cognate with Latin altus,22 English yard (Old English geard) with Latin hortus, English sad23 with Latin satis. Likewise, Indo-European p and k, instead of appearing as f and h, sometimes become b and g.

Of the four distinguishing Teutonic characteristics that have been reviewed, it is probably accurate to suggest that the accent contributed most to make Teutonic words distinctive and to give the modern Teutonic languages a unique place among European languages. The stress-shift that took place, as Verner's Law implies, later than the consonant-shift explained by Grimm's Law, was even more important than the other process in changing the character of Teutonic words.24 The result of the stress-shift was that, in general, the Teutonic accent was fixed on the root syllable of the word, except that in nouns and adjectives, and in verbs derived from these, it rested on the first syllable, whether prefix or root. As a result of both consonant-shift and stress-shift, the Teutonic branch

22 Representing an Indo-European *alt-os.

23 Originally meaning "sated"; cf. German satt.

24 For a discussion of the relative significance of these two processes, see Jespersen, Growth and Structure of the English Language (4th ed.), pp. 22-28.

The Ancestry of English

33

emerged as a highly individualized and easily identifiable member of the Indo-European family of languages.

The common Teutonic characteristics having thus been pointed out, it is necessary, in order to place the English language more specifically among its relatives, to return to the divisions within the Teutonic branch.

To the eastern group belongs Gothic; to the northern belong Old Norse and the modern Scandinavian languages; to the western belong German, Dutch, Low German, Frisian, and English. West Germanic,25 therefore, is to be further investigated. Its chief divisions are obvious: the part of it affected by the second consonant-shift is the High German; West Germanic languages unaffected by the shift form the Low German group. The modern representative of High German is the literary language of modern Germany, called simply German; all the other West Germanic dialects and languages that have been mentioned belong to the Low German division. The second consonant-shift began in southern or upper Germany (whence the name High German) and worked its way gradually northward, stopping short, however, of the low regions bordering the North Sea and the Baltic. Unlike the first or great consonant-shift, the history of this process is quite clear, its period covering the sixth, seventh, and eighth centuries. Many German words are identical, or nearly identical, with parallel English words. This is so, in the very great majority of cases, not because English has borrowed26 from German, or German from English,

25 The name, of course, like Indo-European and like Teutonic, signifies a hypothetical language only—one assumed to have existed because of the similarities among derivative languages in which literature is found, not one of which there is any written record.

26 For examples of the relatively small group of borrowings, see p. 332.

34 The Development of Modern English

but because the two vocabularies have a common ancestry in West Germanic, Teutonic, and Indo-European; their relation is therefore that of cognates. Thus, English house, German Haus, English man, German Mann, English finger, German Finger, and English sing, German singen, are evidently pairs of words that are almost identical. In many cases, however, the relationship is less apparent. The first reason for this state of affairs27 is the High German consonant-shift that German underwent and English, as one of the Low German languages, did not. Since the shift began in the south and went northward with diminishing force, and since, in Germany as in England, it was the language of the midland that was the ancestor of the modern literary language, it follows that not all the consonants that might have been shifted were actually shifted. Of the three series—labials, dentals, and palatal-velars—it happened that the dentals underwent the most thorough shifting. The relation, therefore, of English and German dentals is almost perfectly regular, according to this formula:

English d = German t (deep, tief; day, Tag; dear, teuer)

English t = German z, tz, s (ten, zehn; cat, Katze; sit, sitzen; that, das)

English th = German d (this, dies; three, drei; thing, Ding)

Among the labials, p was most frequently shifted, appearing in German as either pf or f (English pepper, German Pfeffer illustrates both possibilities); and

27 It should be remembered also that the tendency of words to shift their meanings will obscure a relationship that otherwise would be obvious; thus, the fact that English deer and German Tier (animal) are cognates is obscured by the narrowing of the meaning of the former to apply to a particular species of animal rather than “animal” in general; and it is somewhat surprising to observe that English dapper and German tapfer (brave) were originally the same word.

The Ancestry of English

35

although f remained unshifted, its voiced equivalent, v, appeared in German as b (note: over, ober; shove, schieben; grave, Grabe). The velars, for the most part, are unshifted in German, except that the explosive k is frequently replaced with the continuant ch; thus, English wake is cognate with German wachen, and week with Woche.

One other characteristic of German, which has served throughout its history to differentiate it from English and other Low German languages, should be pointed out. The fact that High German was affected by the second consonant-shift and Low German was not might seem to indicate that the former was more subject to change, less conservative, in a word, than the latter. In point of fact, however, the contrary is the case. In almost every other manifestation, German has been far more conservative than English. Only the larger aspects of this contrast can be indicated here: (1) the far greater extent of inflectional leveling in English, so that in many respects the inflection of German is more like that of Old English than that of Modern English—as in its full declension of the definite article for gender, number, and case, and its general retention of adjective inflection, even to the extent of differentiating between the weak and the strong position; (2) the thoroughly “mixed” character of the English vocabulary, on the one hand, and the relatively “pure” character of the German, on the other. In the last respect, once more, German is more like Old English than Modern English, for, like Old English, it has always preferred to translate28 borrowed ideas into native terms,

28 The word itself serves to illustrate the contrast: English translate uses the Latin roots for “over” and “carry,” while German übersetzen uses the native ones.

36 The Development of Modern English

while Modern English has more commonly borrowed the

word with the idea, and has seldom hesitated to supersede

a native word with a borrowed one.

The closest affiliations of English, therefore, are not

with High German, but with the Low German languages.

It is interesting to observe that the northern dialects of

Modern German are frequently closer to English in their

vocabulary than to the literary language of Germany.

The following illustrations will serve to indicate this

relationship:

High German.......... als tief Herz dasz ist hatte zwanzig

Low German.......... as deep Hart dat is hadd twintig

English.............. as deep heart that is had twenty

The explanation, of course, is that the dialect, like Eng-

lish, but unlike the literary language, was unaffected by

the second consonant-shift.

A similar relationship exists between Dutch and Eng-

lish. Dutch, too, has shown a tendency to level its

inflections that is more like the radical evolution of

English than the conservative development of German.

In vocabulary also, it is frequently closer to English than

to German, as these parallels will show:

German............. Wasser zwei Herz zwanzig zehn Weib

Dutch.............. water twee hart twintig tien wyf

English............. water two heart twenty ten wife

The nearest relative of English, however, is not Dutch

but Frisian, a language once spoken over a great part of

northwestern Germany, but now restricted to three or

four hundred thousand inhabitants of Holland and of

coastal islands in the North Sea. The modern dialect

called Frisian is so closely akin to English that com-

munication with the people who speak it is said to be no

The Ancestry of English

more difficult for a speaker of standard English than

communication with the peasantry of many parts of

modern England. Old Frisian has so many character-

istics in common with Old English that scholars have

sometimes assumed that they were originally one speech,

a separate division of Low German which they have

named Anglo-Frisian.

The tendency of all this discussion of the ancestry of

English is to emphasize the inevitably Teutonic character

of our language. However radically its forms and its

structure have changed, and however important the

cross-relationships with such languages as Latin and

French have been, it is evident that most important of all

is the Teutonic base. It will be the endeavor of the

following chapters first to set forth the more general

aspects of the development of English and its present

status, and then to examine successively and in greater

detail such phases of the English language as inflections,

sounds, spelling, vocabulary, and syntax.

References for Further Reading

The books by Pedersen, Jespersen, Sapir, Bloomfield, and

Graff enumerated on pages 12 and 13, and, in addition:

Emerson, O. F., History of the English Language, New York

(Macmillan), 1894.

Jespersen, Otto, Growth and Structure of the English Language,

New York (Appleton), 1923.

Krapp, G. P., Modern English, Its Growth and Present Use, New

York (Scribner's), 1909.

Lounsbury, T. R., History of the English Language, New York

(Holt), 1894.

Meillet, A., Introduction à l'étude comparative des langues indo-

européennes, Paris (Hachette), 6th ed., 1924.

38

The Development of Modern English

Schmidt, P. W., Sprachfamilien und Sprachenkreise der Erde,

Heidelberg (Winter), 1926.

For a bird's-eye view of the general distribution of linguistic

groups, see the "Linguistic Map of the World," page 436 of

Graff's book. An elaborate series of linguistic maps appears

in Meillet, A., and Cohen, Marcel, Les langues du monde,

Paris (Champion), 1924.

CHAPTER III

Old and Middle English

THE English language is conventionally divided into three periods: Old, Middle, and Modern English. It is probably unnecessary to suggest that the threefold division does not imply that there was a sudden substitution, at any particular time, of one form of the language for another; the English people did not, for example, stop speaking Old English on the last day of the year 1099 and begin speaking Middle English on the first day of the year 1100. The development of course has been continuous, and the transition from one period to another gradual, in spite of the fact that at certain times in the history of the language changes have taken place more rapidly than at other times. The Old English period, as the term is usually understood, is from the beginning of the invasions of Britain by the Germanic tribes, in 449,1 to 1100; the Middle English period is from 1100 to 1500; and the Modern English period begins with 1500. Some confusion has been caused by the use of "Anglo-Saxon" rather than "Old English" as the designation for the earliest of the three phases of English. It should be understood that the two terms are synonymous, and that it is nothing more important than fashion that dictates that in one generation those who write on the English language shall term the oldest

1 This is the date given by Bede; but it is quite likely that the first invasions were really earlier.

39

40 The Development of Modern English

period “Old English” and that in another writers shall designate it “Anglo-Saxon.”2 To the present writer, the former term seems preferable, particularly because of its suggestion of a continuous development; but both labels should mean precisely the same thing—the language in its almost purely Teutonic aspect, up to or until just after the Norman Conquest.

The history of the English language proper, then, begins with the incursions of the Jutes, Saxons, and Angles3 about the middle of the fifth century. These invaders came from neighboring regions in Schleswig-Holstein; they spoke, not one language, but related Low German dialects. The Jutes came first and occupied the smallest territory, principally Kent and the Isle of Wight. The Saxons occupied practically all of England south of the Thames, with the exception of these Jutish territories and with the further exception of Cornwall, which remained in the possession of the Celts; in addition, they occupied the regions north of the Thames later called Essex and Middlesex. The Angles took for themselves what was left: the greatest part of what is now England, and Lowland Scotland as far as the Firth of Forth, with the exception of the west coast.

2 At the present, fashion is divided. The titles of Krapp and Kennedy's Anglo-Saxon Reader, New York (Holt), 1929, and Flom's Introductory Old English Grammar and Reader (1930) illustrate this diversity. It is curious to note that the designation of one of the older texts in the field, originally An Old English Grammar, has been changed, in recent printings, to An Anglo-Saxon Grammar, though the content of the book is the same and though it continues to include, on the very first page, an argument for the superiority of “Old English” over “Anglo-Saxon.”

3 In addition to these three, a fourth tribe, the Frisians, should perhaps be added. They inhabited in Britain much the same regions as the Jutes, though to some extent they settled in Anglian territory also.

Old and Middle English

41

Here, then, is the explanation of the division of Old English into dialects that have left their trace in the provincial speech of England until this day. The dialectal groupings have been somewhat shifted, but a threefold division is long preserved. The speech of the Jutes became the Kentish dialect of Old English; the principal dialect of the Saxons was known as the West Saxon; and the Anglian tongue split into two dialects, the Mercian dialect of the Midland and the Northumbrian dialect of the north. The later history of the dialects may be briefly suggested: the Kentish and the West Saxon dialects fell together as the southern dialect of Middle English; the Mercian became the Midland, its principal subdivision—the East Midland—being the ancestor of literary Modern English; the Northumbrian dialect became the Northern, the popular tongue on both sides of the Scottish border. Literary supremacy has swung from the south to the north and eventually come to the Midland. In Old English times, the first centers of learning and culture were in Northumbria; the first important literature in English, dating from the late seventh and early eighth centuries, was therefore in the Northumbrian dialect. By the end of the ninth century, another center was established, in the south, in Alfred's kingdom of Wessex; the body of later Old English literature is in this West Saxon dialect. In Middle English times, no one dialect could claim to be supreme until the East Midland finally asserted itself, by the end of the fourteenth century, as the most favored among the claimants. Standard Modern English represents, for the most part, the further development of this dialect, which has not been, however, without important rivalry from the Northern, partic-

42 The Development of Modern English

ularly in the late Middle and early Modern periods of the language.

To outline the general development of the language in its Old and Middle periods requires us to return to pre-Saxon Britain, and to retell a familiar story. When the Germanic invaders came, Britain was inhabited by the Celts, who had dwelt there for centuries, though it is supposed that another race or races had preceded them.4 Celtic Britain had been invaded by the Romans in 55 and 54 B. C., as we are told in the Commentaries of Julius Cæsar, but any attempt to conquer the island was postponed for almost a century. By the end of the first century of the Christian era, however, the conquest was complete, and a Roman colony which embraced the territory as far north as what is now Lowland Scotland was established. This military occupation of Britain lasted until the early fifth century, when the far-flung Roman legions began to be withdrawn from the provinces in order to defend the capital; Britain, the most distant western outpost of the empire, was the first to be abandoned. The Celtic inhabitants, left defenseless against the attacks of the Picts and Scots on the north and the west, after appealing in vain to Rome for help, finally called in the aid of the Germanic sea-rovers who had in the past harried their shores. The call was answered; but the upshot was disastrous to the Celts, for the Germanic tribes who came to help against the Picts and Scots eventually coveted the island for themselves, turned against their Celtic allies, and succeeded in dispossessing them of the greater part of their lands. The Celts were driven to the corners of

4 Cf. Chapter II.

Old and Middle English

43

the island, to Cornwall, to Wales, to the Highlands

of Scotland, and across the Irish Sea. By the middle of

the sixth century, the Teutonic settlement of Britain

was complete.

What were the influences, then, of the earlier lan-

guages of the island—Celtic and Latin—upon the

language of the invaders, the speech that became the

English language? As to the former, the general con-

clusion must be that the Celtic deposit in Old English

is curiously and almost inexplicably small. The words

(apart from place-names) known, with any degree of

certainty, to have been borrowed from the Celts in the

Old English period5—such as bannock, brock (badger),

and down (hill)—do not amount to more than a dozen;

and the tendency of recent scholarship is rather to

decrease than to add to this small group. Place-names

stand on a different footing: hundreds of towns and cities,

like Aberdeen, Carlisle, and Dundee retain their old Celtic

names. This situation, incidentally, exactly parallels

the treatment of the languages of the American Indians

in the United States; for American English, and the

English language in general, has borrowed exceedingly

few common words6 from the Indians, but many places

are still known by their picturesque Indian designation

or some corruption of this, the impulse to substitute a

commonplace European name having, fortunately, often

been thwarted. Since the hypothesis of a wholesale

extermination of aboriginal inhabitants is untenable in

the one case and exceedingly unlikely in the other, the

only plausible way of explaining the small number of

5 This qualification is important. See page 330.

6 See page 337.

44 The Development of Modern English

words borrowed from their languages is that the relation of conquerors and conquered made it unnatural for the former to use the words of the latter. It must be supposed that, where there was contact at all, the relation of Anglo-Saxon and Celt was that of master and slave—on the whole, a different relationship from that which obtained after the Norman Conquest between Norman and Saxon. Only in this way does it seem possible to account for the very different developments in language after the Teutonic settlement and after the Norman Conquest.

The Latin element in Old English, though considerably greater than the Celtic, is likewise relatively small; but the reasons for this are more evident. There is no real mixture of races in question here; and it is further quite clear that the policy of liberal word-borrowing, later firmly established in English, did not exist in Old English times. The attitude of Old English is much more conservative, more like that of German, as has been suggested, than like that of Modern English. When ideas and institutions were borrowed from the Romans by the Anglo-Saxons, the tendency was, wherever possible, to use translated terms for them. The whole number of Latin borrowings in Old English has, it is true, been estimated as totaling about four hundred; but many of these occur but once and evidently did not take root, so that a much smaller figure would more accurately represent the extent of the Latin element in really active use.7 Many of these borrowings were due to the Christianization of the English following the missionary journey of Augustine in 597. Earlier contacts with

7 Cf. Bradley, p. 82, and Krapp, Modern English, p. 216.

Old and Middle English

45

Latin had resulted in the incorporation of a few words

into the vocabulary of the Anglo-Saxons. While still

in their Continental homes, the Germanic tribes had had

dealings with Roman merchants and had borrowed some

of their words, which therefore appear in other Teutonic

languages besides English. Then, there is the possi-

bility of words left by the Romans among the Celts and

later taken over by the Anglo-Saxons, though the meager

Celtic influence on Old English in general makes this, in

all likelihood, a very small group. Here again, however,

it may be said that place-names are the exception; note

the numerous progeny of the Latin castra (camp), adopted

by the Celts and inherited by the Anglo-Saxons.8 But

it seems more plausible, on the whole, to explain the

greater number of Latin borrowings in pre-Christian

times as due to direct contact between Anglo-Saxon

and Roman rather than to the medium of Celtic. It

should be emphasized, at any rate, that at least a few

very common words—mile, wall, street, wine, kitchen, and

cook, for example—antedate the influence of the ecclesias-

tical vocabulary of Rome.

Before turning to the general characteristics of Old

English, one other foreign contact must be mentioned.

The word "foreign" seems less appropriate in this con-

nection, and the effect in general was to strengthen rather

than weaken the Teutonic character of the language; for

the contact was that with a race closely allied to the

Anglo-Saxons, the Danes. The inroads of the Danes

8 The word itself gives Chester, and, with similarly palatalized ch, other

southern names like Rochester and Winchester; the northern names keep

the hard c, as in Doncaster, Lancaster, and Todcaster. To the same

general source belong also place-names like Gloucester, Leicester, Worcester,

and Exeter.

46 The Development of Modern English

began in the late eighth century, and by the time of Alfred the Great these people had extended their territory southward from their original foothold in Northumbria until they commanded all the territory north of the Thames. Their amalgamation with the Anglo-Saxons was finally complete, and their influence upon the language of the latter was certainly very great, though its extent is obscured by the intimate relationship between the two word-stocks. In many cases, indeed, it is impossible to be sure whether the form of a given word is Scandinavian or Northern English. Undoubtedly, many very familiar words have reached us in a Scandinavian rather than a genuinely English dress: sister, for example, is from the Old Norse syster rather than the Old English sweoster; and we should be saying yive and yift, as Chaucer did (rather than give and gift), had it not been for the influence of Old Norse gifa.9 In regard to gift, it is interesting to note, too, that we owe not only the modern form but the modern meaning to the Scandinavian rather than the Old English variant; for the more specialized meaning of the latter was, in the singular, "the price paid by a suitor in consideration of receiving a woman to wife," and, in the plural, "wedding."10 As Jespersen well says, in connection with the last instance, "No subtler linguistic influence can be imagined than this"—nor, one may add, one more difficult to estimate in its entirety.

Consideration of these three loan-elements in Old English—Celtic, Latin, and Danish—has emphasized

9 Since in Old English g was palatalized before e and i; so that giefan (to give) was pronounced with initial y.

10 Jespersen, Growth and Structure of the English Language, p. 69.

Old and Middle English

47

what may well be regarded as its first distinctive quality,

the thoroughly Teutonic character of its word-stock.

Whereas the Modern English vocabulary possesses native

and borrowed constituents that are of approximately

equal value and importance, the Old English vocabulary

was, with on the whole decidedly minor exceptions,

exclusively Anglo-Saxon, or at any rate Teutonic. Had

the Norman Conquest not taken place, and had the

English language persisted in its ultraconservative

attitude toward word-borrowing, our modern speech

would resemble German much more than is actually the

case.

The resemblance between Old English and German

suggests perhaps the readiest way to outline the general

characteristics of the former. Not only, as has been

indicated, in its vocabulary, but quite as strikingly in its

accidence or inflection does Old English present familiar

conceptions to one who has studied German. Modern

English has discarded grammatical gender in favor of

what we are prone to call logical or natural gender;

but Old English, to give a single illustration, could use a

neuter noun, wīf, for “woman,” just as German does in

the cognate form Weib. The noun, further, had an

elaborate system of case endings, which German approxi-

mates much more nearly than Modern English. In other

parts of speech, such as the verb, particularly in its

treatment of the subjunctive, and the adjective, in its

distinctions between the weak and strong positions

as well as its inflection for number, gender, and case, the

resemblance to German is noteworthy. Most striking

of all, perhaps, is the fact that while the Modern English

definite article has a single invariable form, the Old

English article exhibits an elaborate set of distinctions

48 The Development of Modern English

for gender, number, and case that exactly parallel, point for point, the German forms.

In addition to vocabulary and to inflection, a page of Old English seems foreign to the reader of Modern

English in its arrangement of the words in the sentence. There appears to be an awkwardness, a cumbersome

quality, about not only the more complicated forms of individual words but also their very order. The familiar

pattern of Modern English—subject, predicate, object—is frequently departed from. Exactly the same varia-

tions as German exhibits—the inverted order with the verb preceding the subject, and the transposed order

with the verb following the object—are to be observed. Such varied word-order, as we shall see, is compatible

with a higher state of inflection; a stereotyped word-order, like that of Modern English, is a necessary consequence

of inflectional leveling. We are concerned at present, however, with the impression that the former makes upon

the modern reader. It is not surprising that to a reader of our greatly simplified Modern English, German, with

all of its tantalizing resemblances to English, should still appear not merely foreign but also highly involved and

almost ludicrously distorted in the forms of its words and their order in the sentence. And perhaps this is not

an unduly exaggerated description of the impression produced by Old English also.

The readiest way to summarize these first impressions of Old English is to analyze a typical passage of its prose.

Here is a selection from the most characteristic aspect of Old English prose—Early West Saxon, as it was written

by Alfred the Great in the late ninth century. The selection forms part of Alfred’s preface to his translation

of the Cura Pastoralis of Pope Gregory. The king

Old and Middle English

49

expresses his perplexity that so little of the learning that,

before the ravages of the Danes, had flourished in the

land has been handed down in the form of English trans-

lations of worthy books; he concludes that the refusal of

the older generation of scholars to do this work of trans-

lation was deliberate and dictated by their feeling that

learning would increase with the study of languages.

When I then this all remembered, then wondered I exceedingly

Ðā ic þā þis eall gemunde, þā wundrade ic swiðe swiðe

of the good wise men who formerly were throughout England, and

þāra gōdena wiðona þe gīu wēron giond Angelcynn, ond

the books all completely learned had, that they of them then

þā bēc ealla be fullan geliornod hæfdon, þet hīe hiora þā

no part did not wish into their own language [to] turn.

nēnne dēl noldon on hiora āgen geðīode wendan.

But I then soon again myself answered and said: “They not

Ac ic þā sona eft mē selfum andwyrde ond cwæð: “Hīe ne

thought that ever men should so reckless become, and

wēndon þætte ǣfre menn sceolden swē rēccelēase weorðan, ond

the learning so fall away; for the desire (i.e., intentionally)

sīo lār swē oðfeallan; for þāre wilnunga

they it neglected, and wished that here the more wisdom in the

hīe hit forlēton, ond woldon þæt hēr þȳ māra wīsdom on

land should be the we more languages knew.”

lond wāre þȳ wē mā geðīoda cūðon.”

A few remarks, then, on this passage, in the light of

what has been said of the general characteristics of Old

English. Study of its vocabulary will reveal that much

more of it is familiar to us, as part of our Teutonic

inheritance, than appears at first glance. Indeed, per-

haps only a single noun, geðīode (language), and a very

few adverbs and conjunctions, such as swiðe (much),

50 The Development of Modern English

gīu (formerly), þā (then, when), and ac (but), are wholly

strange. Several others, of course, are preserved in

Modern English only as archaisms—eft, cwæð (quoth),

and wēndon (weened)—or as roots the meaning of which

has somewhat changed—giond (cf. yon), cūþon (cf.

un-couth, formerly meaning unknown), and weorþan

(be, become, as in “Woe worth the day,” i.e., “Woe be to

the day”). Notice, too, that in a number of cases, the

form seems nearer to German than to Modern English:

woldon (cf. wollten), weorþan (cf. werden), wāron (cf.

waren), geliarnod (cf. gelernt), wendan (cf. wenden).

Our second point, the far greater complexity of inflec-

tion, also finds abundant illustration here. In this short

passage, the definite article, which in Modern English

appears invariably as the, has five variants: sēo (nom.

fem. sing.), þǣre (gen. fem. sing.), þȳ (instrumental

neut. sing.),11 þā (nom. plur.), and þāra (gen. plur.).

Adjective declension, utterly lost in Modern English,

appears in such forms as rēccelēase (plur.) and gōdena

(gen. plur. and a weak form). Several lost inflectional

endings of the noun may be noticed; for example,

londe (dat. neut. sing.), wiotona (masc. gen. plur.), and

wilnunga (dat. fem. sing.). Other types of plural forma-

tion than the regular -s ending of Modern English are to

be seen: bēc, an unlaut plural12 later regularized into

books, and geþīoda (languages). In the verb, too, a far

higher state of inflection—as in the infinitive endings of

wendan and weorþan and the plural endings of forlēton and

cūþon—is evident.

11 Such a phrase as “the more the merrier” preserves this adverbial

use of the old instrumental case; it parallels the Latin ablative of degree

of difference.

12 Cf. page 115.

Old and Middle English

51

Finally, our selection gives a vivid picture of the more varied word-order that accompanies the greater degree of

inflection in Old English. German might parallel this order: “When I this all remembered, then wondered I

much at the good wise men who formerly were throughout England and the books all completely learned had”—

but Modern English, never! Note, throughout the passage, how infinitives and participles are detached from

their auxiliaries, how a predicate is detached from its subject, and how a pronominal object may precede

rather than follow its predicate. Most striking of all, probably, is the transposed order that is the usual rule in

subordinate clauses. The use of transposition13 is perhaps the most powerful single element in making the

passage appear, to the modern reader, thoroughly un-English.

With this glimpse of the language in its Old English stage, we may pass on to the developments in the Middle

English period. As has been indicated, the transition from the one to the other is to be thought of as gradual

rather than as sudden; further, it is clear that it did not proceed at a uniform rate in all the dialects. In the

north, changes were earlier and more rapid than in the south, so that in the year 1100, which we have taken as

the dividing line, the northern speech had already distinctly assumed its Middle English aspect, while the

southern tongue remained for at least half a century longer essentially what we know as Old English.

13 Though it is to be observed that the substantival pæt clauses are not completely transposed. It seems to be here that the transposed order gave ground most readily to the normal and opened the way for the final establishment of the normal order in all dependent clauses—doubtless aided later by the influence of French.

The great factor in accelerating and making more thorough changes already at work in the language was of course the Norman Conquest.

The danger, however, is much more likely to be that the influence of this Conquest be overestimated than that it be minimized.

To account for all this difference between Old English and Middle English as due to the Conquest is a gross exaggeration.

For example, the transposed word-order and the grammatical gender that have been mentioned as characteristic of earlier Old English had certainly begun to be replaced by normal order and logical gender14 before the Norman Conquest.

In observing the traits of Middle English, therefore, it is well to guard against the easy but erroneous assumption that everything not found in Old English admits of but one explanation.

If, however, we approach the general characteristics of Middle English through the obvious method of comparison with Old English, it is undeniable that in the very first point, and the most conspicuous change of all, a transformed and augmented vocabulary, we encounter overwhelming French influence.

As we have seen, the Old English word-stock is virtually unilingual; the Middle English is clearly bilingual.

The French influence began before the Norman Conquest:15 King Æthelred

14 For word-order, see the preceding footnote. For the other point, the replacing of grammatical by logical gender, see Bradley, pp. 47–50; Classen, Outlines of the History of the English Language, pp. 38–39; and especially, an article by Samuel Moore, “Grammatical and Natural Gender in Middle English,” Publications of the Modern Language Association of America, Vol. XXXVI, No. 1 (March, 1921), pp. 79–103. Whatever the exact steps of the process, it is clear that it was under way long before the end of the Old English period.

15 Castle, clerk, false, mantle, purse, trail, and turn are French borrowings in Old English writing (Kluge, F., Englische Studien, Vol. XXI, pp. 334 and 335).

married in 1002 the daughter of the Duke of Normandy,

and their son, later known as Edward the Confessor,

found many influential posts in the English church and

state for French clergymen and French nobles. It is

to be noted, too, that French influence continued,

though in a new form, after the loss of Normandy in

1204. As early as the twelfth century, the French of

Paris, as learned by English students and priests, had

begun to supplement the earlier influence of the Norman

dialect, and eventually it became a more important

force. Thus, it is not surprising to find that French

words begin to enter the English vocabulary before 1066,

and that they continue to enter, and in greater numbers,

particularly in the later half of the thirteenth and through

the fourteenth century.16

What kinds of words were they? Before answering

this question, it will be necessary to outline the extent

to which French supplanted English and the extent to

which the two languages existed side by side. The policy

of William the Conqueror is believed by recent historians

to have been far more liberal in its attitude toward the

English language than was long supposed. Apparently,

the king made no attempt at all to force his new subjects,

in their relations with one another, to speak French.

But it was inevitable that French should become at

once the language of law, of the church, of civil govern-

ment, and of military organization. It was almost as

inevitable, particularly since the Normans represented

a higher social and literary culture, that their advan-

16 See Jespersen's figures for 1,000 typical French words, Growth and

Structure of the English Language, p. 94. If we could add words used in

conversation to those recorded in literature, the numbers for the earlier

periods would doubtless be considerably increased.

54 The Development of Modern English

tageous position as conquerors should mean the importation of a French vocabulary for the conventions and usages of polite society, for the things of the church, for education, and for literature. The English language was reduced from the status of a literary language to that of a mere spoken dialect; here is a most important reason for the rapidity and thoroughness of the changes that take place, in other aspects of language as well as in vocabulary, in the early Middle English period.

Illustrations of some of the groups of words that have been suggested as representing the fields of French influence may now be given. Most of our legal terminology is to be traced directly to this source: familiar words like suit, plead, plaintiff, judge, jury, and jail17; and more technical ones like larceny, assize, demesne, court of oyer and terminer. Here belong also ecclesiastical terms like cloister, homily, clergy, and service; terms of national government like parliament,18 crown, realm, and sovereign; and military words like fortress, siege, armor, battle, and war itself. We may mention also the terms of feudalism, such as liege, vassal, and the very word feudal; and the terms designating the various grades of the French aristocracy, such as prince, duke, marquis, viscount, and baron,19 as well as those standing for the

17 This "American" (and more phonetic) spelling comes from the form of the word in the dialect of central France; the British gaol, though pronounced identically, perpetuates the Norman variant.

18 The Middle English spelling, parlement, is better both etymologically and phonetically.

19 It might be added that the highest ranks of all, king and queen, retained their Anglo-Saxon titles; the general terms lord and lady likewise survived. Earl, knight, and alderman (alas, how fallen!) also go back to an earlier dispensation. It is curious, however, that the wife of an earl is called by the French title of countess, the lack of a feminine equivalent for earl suggesting a different attitude towards the position of a noble lady in pre-Conquest days.

Old and Middle English

55

virtues supposed to be cultivated at court—including

honor, courtesy, and refinement. Illustrations of French

borrowings like these may possibly suggest this question:

In what classes of objects or ideas has the French influ-

ence not been felt? What groups of words remain purely

Anglo-Saxon? According to Bradley,20 there is just one

class of objects for which the native names have remained

(so far as colloquial use is concerned) without any

French mixture—that consisting of the external parts

of the body. Even here there is one exception: the

French word face found its way into English conversation

comparatively early, evidently because there was no

equally simple Anglo-Saxon noun to contend with it.

Otherwise, this group of words—eye, ear, tongue (French

only in spelling), nose, arm, hand, and so forth—is purely

native. The very fact, however, that a homogeneous

group of words like this is so highly exceptional is most

striking evidence of the omnipresence of French influence

on our vocabulary.

Enough has perhaps been said to suggest the trans-

formation that the vocabulary underwent in Middle

English times. But it should be remembered that the

addition of a French word does not necessarily mean the

loss of an English one. Frequently, the two remain as

synonyms, though more likely than not the earlier,

English term takes on a humbler significance; thus, we

have table as well as board,21 labor as well as work, and

chair as well as stool (formerly as general a term as its

German cognate Stuhl). A striking illustration of this

aspect of the bilingual vocabulary is the often-quoted

20 The Making of English, p. 91.

21 Retained, in a similar sense, in bed and board, boarding house.

observation of Wamba, the jester in Ivanhoe, to the effect that while the Saxon serf calls the animals he tends by one set of words (ox, cow, calf, sheep, swine, and deer), the Norman master knows them, when their flesh appears as food, by another (beef, veal, mutton, pork, bacon, and venison).

One more general comment on the Middle English vocabulary. It may appear to have been assumed, since only the influence of French has been commented on, that this was the sole source of borrowing. Of the three foreign influences that were mentioned in connection with the Old English vocabulary—Celtic, Danish, and Latin—it is true that the first two are of little importance so far as Middle English is concerned. The third, however, has been a potential source of borrowing in all periods of the English language; and the fourteenth century may be mentioned, along with the sixteenth and the nineteenth, as one of the times in which Latin words were introduced in greatest numbers. And it is by no means the least important aspect of the custom of importing French words that this precedent opened the way for a similar treatment of Latin. This is undoubtedly the reason why English contains so much larger a Romanic element than any other Teutonic language. It is not at all true that classical scholarship has been more active in England than in Germany or Holland; but in the history of neither German nor Dutch had the Renaissance impulse, which began to be felt, even in England, as early as the fourteenth century, been preceded by an earlier wholesale borrowing of French words. The earlier importations from Latin as a consequence of the humanistic movement—many of which, to be sure, are difficult or impossible to separate from Latin words borrowed

Old and Middle English

57

through the medium of French—are thus to be reckoned

as an additional, and a very significant, aspect of the

enlarging and developing vocabulary of Middle English.

The more complex vocabulary is accompanied by an

increasingly simpler inflectional system. This will form

part of the subject of a later chapter and needs only to

be summarized here. Two changes are perhaps most

conspicuous, in the general process of leveling to which

the various inflectional endings were subjected: the far-

reaching vowel weakening, by which almost every vowel

other than e, occurring in an inflectional ending, was

changed to e; and the loss of final -n in the inflectional

forms of several parts of speech, especially in the weak

declension of nouns and adjectives and in the infinitives

and certain plural indicatives of verbs. These two

developments alone—the former, at any rate, quite

certainly completed in early Middle English times—

are enough to account for a great part of the simplifica-

tion of Old English inflectional forms. A further step

in doing away with the variety of forms exhibited

in the earliest period of the language, and incidentally

in shortening a very great number of words,22 was

the general loss of final -e. These levelings even-

tually did away with any possibility of inflection in the

adjective and the article, and greatly simplified what

inflection remained in the noun and the verb. Only the

personal pronoun has retained, throughout the Middle

22 Even where the final -e has been retained in the Modern English

spelling—for example, in numerous monosyllables like love, tale, and

name—it is lost in the living, or spoken, form of the word. It might be

added that the final -e of many Modern English words is not a weakened

inflectional ending, for to Old English monosyllables—for example,

wīf (wife), hām (home), and lār (lore)—an uninflectional -e was frequently

added later.

58 The Development of Modern English

English period and in Modern English, anything like the full inflection of Old English.

Greater simplicity was achieved also in the structure of the Middle English sentence. Here, too, the inherent tendency of the language to simplify—as in the replacing of the transposed and the inverted orders with the normal order—was aided by the influence of French. At the same time, and through the same causes, English expression became not only simpler but more varied and more flexible. Just as the influence of the French vocabulary made the word-stock richer and fuller, so also in the structure of the sentence the French example caused the language to become a more supple and more efficient instrument. A single illustration perhaps will serve to indicate the greater variety of expression that was developed in Middle English. In Modern English, as in Middle English, we can still use the Old English, or Teutonic, method of expressing possession; we can say, “the king’s palace.” But we can also use a prepositional phrase, on the French model, for the same purpose: “the palace of the king.” More than that, we can combine the two types of expression, and phrase a similar idea in a third way: “that palace of the king’s.”

It remains for us to sketch in brief outline the gradual victory of English over French in the Middle English period, and the rehabilitation of English as a literary language after its temporary and partial eclipse. Its eclipse, of course, was never total: even in the early Middle English period, the language did not cease to be spoken, and, though to a much smaller extent, written. As early as the twelfth century, there are references which indicate that some, at least, of the Norman nobles and Norman churchmen both spoke and

Old and Middle English

59

understood English.23 The lower orders, naturally, had

never left off speaking English among themselves. For

formal and written purposes, French lasted perhaps

longer than we should have expected; but this is evi-

dently due to the fact that the educational and cultural

supremacy of France throughout the later Middle English

period made it natural that instruction in the schools,

for example, should be given in French long after the

direct influence of the Norman Conquest had ceased.

In the law courts, too, the use of French, established

after the Conquest, was continued, with typical legal

conservatism, as late as the middle fourteenth century;

so far as that goes, a good deal of French terminology

lingers in Anglo-Saxon jurisprudence to this day. As

the language of written documents, and of royal procla-

mations, English reasserted itself by the middle thir-

teenth century, even though, in the former category

especially, it was here in rivalry with Latin as well as

with French. It should be added, however, that this

particular victory was not complete; for the prestige

of French culture in the thirteenth and fourteenth

centuries was great enough to cause the general use of

French, where Latin was not employed, in public

documents until almost the end of the Middle English

period.

A most significant aspect of a written language is its

use for purely literary purposes, and here the eventual

triumph of English over French was heralded as early

as 1200. This is the approximate date of both the

Ormulum and Layamon’s Brut. It should be remem-

23 See Emerson, History of the English Language, pp. 59 and 60. The

same work (pp. 58-83) cites many of the documents illustrating the

replacing of French, for various purposes, with English.

bered, too, that in some sense our literary tradition

has really been continuous, the Anglo-Saxon Chronicle,

in particular, serving to represent the transition from

Old to Middle English literature, and the earliest aspect

of the latter as well. Yet, after the Normans came, the

literary use of the native language was all but discon-

tinued. Norman French took its place as the language

of pure literature, while Latin continued to be, as it

had been before the Conquest, the language of monkish

chroniclers and scholars. From about the middle of

the twelfth century, however, the Anglo-Norman litera-

ture is paralleled by writings in English, until the great

revival of poetry in the fourteenth century saw the

lasting triumph of English as a literary language. This

poetical renaissance, to be sure, was not an unqualified

victory for the Anglo-Saxon tradition; for, while the

old, alliterative measure was used in Piers Plowman,

in the Pearl, and in Gawain and the Green Knight, rhymed

and metrical verse, on the French model, was used by

the court poets, Chaucer and Gower. It need not be

regretted that the practice of the last-named poets24

proved to be the more influential, since one may well

be dubious about the poetic possibilities of the four-

stress alliterative measure.25 Not only in metrical

24 There are probably, however, more echoes of the old Teutonic metre

in Chaucer's verse than has generally been supposed. See my note,

"Old English Verse in Chaucer," Modern Language Notes, Vol. XLIII,

No. 4 (April, 1928), pp. 234–236.

25 Cf. the opinion of Sir A. T. Quiller-Couch. After quoting a fifteenth-

century survival of the old form, he remarks, "Clearly a language to

attain poetry must cure itself of this fashion . . . Historically it

may be valued as a stage in the Teutonic endeavor towards being lively:

as a form of poetry and in comparison with civilized verse it is silly.

For real literature it simply will not do." The Age of Chaucer, London

(Dent), 1926, p. 48.

Old and Middle English

61

form, but in the general conventions of literary technique,

was Chaucer greatly in debt to his French predecessors

and contemporaries. Unlike Gower, however, he chose

to commit his whole literary fortunes solely to the

English language; and ever since he wrote, it has been

unmistakable that the literary language of England is

English.

The fourteenth century not only saw English firmly

and finally established as a literary language, but it

saw the elevation of one dialect into a commanding

position as a literary speech. Hitherto, there had been

rival claimants among the three groups of dialects:

Northern, Midland, and Southern. Beginning with

the fourteenth century, however, a single subdivision

of one of these, the East Midland dialect of the capital,

the court, and the universities, assumes a peculiarly

favored position among its rivals; its position as the

literary standard is never seriously threatened through-

out the rest of the Middle English and the whole of the

Modern English period.26 The causes of this elevation

of East Midland are not far to seek. All things worked

together to bring it about; if England was to be a really

united realm, the speech of London could scarcely fail

to be, in time, the standard for the whole country.

But the chief reason why London English became the

standard dialect just when it did is quite certainly the

influence of Chaucer. Not, by any means, that this

was the sole cause. The happy chance, however, that

26 The nearest approach to a rival literary language and a second

literary center is, of course, to be found in the literary use of Lowland

Scots, and in Edinburgh as a rival of London—particularly in the work

of the Scotch Chaucerians of the fifteenth century, and again in the early

nineteenth century, in the circle of Sir Walter Scott and the contributors

to the Edinburgh Review and to Blackwood's.

62 The Development of Modern English

Chaucer was a Londoner born and bred and wrote for a court that, so far as it was English-speaking at all, spoke the local dialect, helped immeasurably to give the East Midland dialect27 a place apart from the others. It is easy to exaggerate here; and, in the past, phrases like "the father of the English language," "the first finder of our fair language," and "the well of English undefiled" have tended unduly to magnify the contribution of Chaucer to the English language. On the other hand, it is evident that a great part of English poetry, throughout the fifteenth century, is in direct imitation of Chaucer, and hence that its language is overwhelmingly influenced by his. The point perhaps needs no laboring, for the mighty genius of Chaucer—if that is not too pretentious a description of so thoroughly human and companionable a writer—has never been so clearly perceived as in the twentieth century. His twin masterpieces, the "tragedie" of Troilus and Criseyde and the human comedy of the Canterbury Tales are now quite generally admitted to be not only unsurpassed but quite unrivalled in their kind. And this was not altogether unperceived in Chaucer's own day and in the century immediately following. No wonder, then, if his work is to be reckoned the most important cause both of the victory of English as a literary language and of the beginning of its standardization.

The fifteenth century saw the standardization of the literary language carried further, though there are still

27 To be quite accurate, it should be pointed out that the London and the East Midland dialects are not quite synonymous. Chaucer, and London English generally, used occasional Southern forms that are later displaced in the standard speech by East Midland ones.

Old and Middle English

63

occasional references to the fact that dialectal differences are a stumblingblock. Thus, Caxton, who, as Professor Emerson suggests,28 was probably made more conscious of such difference through his long residence abroad, complains that "it is hard to please every man because of diversity and change of language,"29 at the very time that the products of his printing press were helping to remove this cause of complaint. One of the earliest and one of the most famous of these books, Malory's Morte D'Arthur (1385), by common consent the first great prose classic of our literature, is thoroughly representative of the standardization as well as of the simplicity and flexibility that the language had attained in late Middle English times. Considerably less than a century separates Chaucer and Malory; but while readers of Chaucer are impressed by a modern spirit couched in medieval language, exactly the opposite is likely to be the impression created by Malory. However dreamlike and unreal the Arthurian world of Malory may appear, the language is, nevertheless, exceedingly simple, and so nearly akin to Modern English that it presents no important difficulties at all to the present-day reader. As in the relatively barren period just preceding the Norman Conquest, so in this other period of stagnation when English literature awaits a new foreign impulse—in this instance, that of the Italian Renaissance—the language has been simplified with unusual rapidity. Before the end of the fifteenth century, it has become essentially what we know as Modern English.

28 Op. cit., p. 81.

29 Prologue to his version of the Æneid, quoted by Emerson, pp. 81-82.

64 The Development of Modern English

References for Further Reading

The histories of the English language by Jespersen, Emerson, Lounsbury, and Krapp, listed on p. 37 (Chapter II), and, in addition:

Bradley, Henry, The Making of English, New York (Macmillan), 1904.

Classen, E., Outlines of the History of the English Language, London (Macmillan), 1919.

Emerson, O. F., A Middle English Reader, London (Macmillan), rev. ed., 1915.

Flom, G. T., Introductory Old English Grammar and Reader, New York (Heath), 1930.

McKnight, G. H., Modern English in the Making, New York (Appleton), 1928.

Smith, Logan Pearsall, The English Language, New York (Holt), 1912.

CHAPTER IV

Modern English: British and American

We have seen something of the process by which the English language, as we know it, was born; how, after the Germanic dialects had achieved some literary distinction in the eighth and ninth centuries, the literary progress of the language was interrupted for several centuries and an alien tongue introduced in its stead; and finally, how, in the fourteenth and fifteenth centuries, a literary language was again evolved from popular dialects that had meantime undergone rapid changes, and a final and standard stamp set upon it by such men as Chaucer and Wyclif, Caxton and Malory. After this time, there are no violent or revolutionary changes in the language to be recorded. The obvious reason is that never again was England exposed to foreign conquest. Nevertheless, in one sense something like this did happen: in the sixteenth century, the republic of letters was conquered by the humanistic movement, the revival of learning that put the study of Latin on a new basis and introduced the study of Greek. The general effect of the revival of learning in the progress of the English language was twofold: a temporary neglect of the vernacular by those whose classical studies made them almost contemptuous of modern tongues, but a later recognition of the possibility of giving to modern languages something of the grace and something of the sonorous quality that scholars found in the classics. In addition, the

66 The Development of Modern English

developing of nationalistic feeling under the later Tudors gave a new incentive to the literary use of the vernacular. The great tradition of Biblical translation, from Wyclif and Purvey in the fourteenth century, through Tindale and Coverdale in the sixteenth, to the King James version of 1611, is likewise to be mentioned as one of the channels through which literary Modern English took form and exerted a powerful and widespread influence.

What is perhaps most striking in the attitude of scholars and writers toward the English language in the early Modern English period is a tendency to divide into two opposing camps: those who held that English should be "improved" by free importations from without, particularly by borrowings from Latin; and those who believed that the language should rather develop its own resources, and that an admixture of other languages meant not improvement but corruption. These points of view are, of course, signs of the opposite beliefs about vocabulary that have always existed and, to some extent, will always exist; but conservative and radical tendencies in this field have seldom been so consciously and definitely opposed as they were in the late sixteenth century. We shall have occasion later to see just how these two influences affected the development of the vocabulary; but it may be suggested here that the eventual victory was with the radical camp, and that in the Renaissance there was established, once and for all—though there are, to be sure, later qualifications—the principle of liberal word-borrowing as a permanent policy of the language. Not every aspect of this victory, however, is to be regarded as admirable. Free borrowing from Latin was accompanied by an aping of Latin rhetoric, and more than that, a deliberate striving to give English prose certain orna-

Modern English: British and American 67

ments and decorations that would lend it dignity and an

aloofness from ordinary speech comparable to that of

verse. The Euphuism of John Lyly and the Arcadianism

of Sir Philip Sidney are merely two particularly con-

spicuous examples of the general tendency to cultivate

a highly self-conscious and thoroughly artificial language.

The promise of Malory's beautifully simple manner was

set aside; despite noteworthy exceptions (the outstanding

one being the Anglo-Saxon diction of the King James

Bible), the all but universal trend of literary prose is

toward the rhetorical and the ornate until the Restoration

and the early eighteenth century see the establishment of

quite opposite ideals.

Abraham Cowley's Several Discourses, by Way of

Essays, in Prose and Verse and John Dryden's Of

Dramatic Poesy both were published in 1668. These two

works are sometimes thought of as marking the first

appearance of a truly modern prose in English literature

—a language that is to all intents identical, in syntax and

in structure as well as in vocabulary, with the English

of our own day. The long separation between an ele-

vated and ornamental diction for literary purposes

and a plain, terse vocabulary for conversation begins to

draw to a close. Throughout the eighteenth century,

beginning with the work of men like Swift, Steele, and

Addison, the striving is for simplicity and directness.

The vocabulary is no longer looked upon as susceptible of

change and addition. The Elizabethan exuberance, the

joy in language as in a new toy, is gone. Gone too are the

long, rolling periods, the Latinized diction of Milton,

Sir Thomas Browne, and Jeremy Taylor. What takes

their place is conciseness and simplicity—a desire for

clarity and accuracy. The tendency is toward “purifica-

68 The Development of Modern English

tion" and refinement; to make the best use of the

language we have but to suffer no innovations. Swift

protests against the borrowing of foreign words, and, later

in the century, Samuel Johnson believes that his diction-

ary will forever "fix the English language."1 On the

whole, this more conservative attitude toward language is

to be regarded as a salutary reaction from earlier radical-

ism; in language too, it is "our indispensable eighteenth

century."

With the dawn of Romanticism, a freer and a more

liberal atmosphere becomes evident, in language as in

other matters. Sir Walter Scott, for example, takes

delight in reviving obsolete words and in introducing

dialect terms, particularly from the balladry of the

Scottish border but also directly from the living language

of the peasantry. It is, however, not merely the language

of literature that is renovated in the nineteenth century.

The colonial expansion of England and the great advances

in scientific thought are both accompanied by a changing

and freely developing vocabulary. As in the spacious

days of Elizabeth, though to be sure with somewhat more

restraint, original writers like Carlyle strike out

words and phrases for their individual purposes, colo-

nizers and travelers borrow exotic words from the

languages of foreign peoples, and men of science coin the

terms to describe the new ideas and new discoveries

that they have contributed to the world's knowledge.

Just as much, then, in the history of the language as in

literary history is the eighteenth century, with its pre-

cursor, the age of the Restoration, to be regarded as a

1 The phrase is from the Plan of the Dictionary addressed to Lord

Chesterfield. Johnson, however, experienced a change of heart as to

this purpose, as the Preface to the Dictionary indicates.

Modern English: British and American 69

conservative interlude set between two epochs of far greater liberalism. It is to be added, however, that in certain respects the conservative spirit then triumphant established conventions and ideals that have not been overthrown in later periods. Particularly is this true of the standardization of language, in spelling and pronunciation more than in vocabulary, that was largely brought about by the dictionaries, the grammars, and the printing houses of the eighteenth century. The notion, for example, that there is any virtue in uniform spelling is distinctly of recent origin; the earlier attitude, of course, was that spelling offered an opportunity for individual choice. Hence the spelling of a particular word was often due to the caprice of the moment; the variations in the spelling of Shakespeare’s signature are a striking Elizabethan illustration. But when dictionaries, which became an important factor in language in the eighteenth century, began to indicate a single spelling as the only correct and permissible form of each word, the older attitude was necessarily altered. Hand in hand with this influence, and making itself felt in a significant way at about the same time, went that of the modern printing house, with its insistence on uniformity and regularity. As with spelling, so with pronunciation. Occasional literary use of dialectal differences in forms and sounds is to be found in medieval and Renaissance writing; the Yorkshire dialect of the Cambridge students in Chaucer’s Reeve’s Tale is perhaps the earliest instance. But not until the dictionaries of the eighteenth century began to record first the accent of words and then their vowel sounds was there any formal move toward a standardization of English pronunciation. Dictionaries have had so powerful an influence upon Modern English that a

70 The Development of Modern English

sketch of its development is incomplete without some account of the part that they have played in it.

The approach to the modern English dictionaries was by way of Latin-English glossaries, of which many appeared in the Middle Ages and the Renaissance. Just which of their successors deserves the title of first English dictionary is a point on which opinion is divided. John Bullokar's English Expositour (1616) is the earliest claimant;2 but it is soon followed by Minsheu's Ductor in Linguas, or Guide into the Tongues (1617), the first of etymological dictionaries; and by Henry Cockeram's English Dictionarie3 (1623), the first in which the word dictionary is used in the sense in which we now understand it. The full titles of Bullokar's and Cockeram's compilations suggest the early conception of a dictionary as limited to difficult words only : the one reads An English Expositour: Teaching the Interpretation of the hardest Words used in our Language, with sundry Explications, Descriptions and Discourses (incidentally implying also the early affiliation between dictionary and encyclopedia); the other, The English Dictionarie: or, An Interpreter of hard English Words.

Later dictionaries, with greater pretensions to completeness, are that of Milton's nephew, Edward Phillips—New World of Words, or a General English Dictionary (1658)—and that of Nathaniel Bailey, the direct predecessor of Johnson—Universal Etymological English

2 A Table Alphabetical, published in 1604 by Robert Cawdrey perhaps should be mentioned; though, according to its full title, it is intended to be limited to borrowed words, and to importations from Hebrew, Greek, Latin, and French exclusively, it does in fact include archaic native words.

3 A recent reprint of this work is available (Huntington Press, New York, 1930).

Modern English: British and American 71

Dictionary (1721). Professor Weekley4 points out that it

is only just before Bailey's work is published—in the

seventh edition (1720) of Phillips's dictionary, to be

explicit—that the familiar word dog makes its appear-

ance. Bailey's volume seems, on the whole, to be the

best candidate for the position of first complete dictionary.

It was Bailey, too, who began the practice of marking the

accent of words, in which he was followed by Samuel

Johnson in his epoch-making dictionary of 1755. John-

son, however, did not give the full pronunciation,

because, as he observed to Boswell, it was impossible to

model this "after the example of the best company

because they differ so much among themselves." He

had come to believe, in other words, that he could "fix"

the spelling and signification of words but not their

pronunciation.5 Not until after Johnson, therefore,

was the final step taken in indicating the pronunciation of

words: the vowel sounds were indicated, for the first

time in a general dictionary, by William Kenrich in the

New Dictionary which he published in 1773.6 This

precedent was immediately followed by other British

lexicographers of the late eighteenth century, and by

the Americans, Webster and Worcester, in the early

nineteenth.

4 "On Dictionaries," Atlantic Monthly, June, 1924 (Vol. 133, No. 6),

p. 786.

5 For a fuller statement, see the chapter "Johnson's Dictionary" in

McKnight, G. H., Modern English in the Making, pp. 351–376.

6 Two Scotchmen had in the meantime published pronouncing diction-

aries. These were James Buchanan and William Johnston, whose works

appeared respectively in 1757 and 1764. For a recent treatment of

Buchanan's work, see Burt Emsley, "James Buchanan and the

Eighteenth-Century Regulation of English Usage," Publications of the

Modern Language Association of America, Vol. XLVIII, No. 4 (December,

1933), pp. 1154–1166.

72 The Development of Modern English

It has been more than once remarked7 that there is a certain poetic justice in the history of dictionary-making since Johnson, in that the two races he particularly abhorred, the Scots and the Americans, have taken the lead in the production of English dictionaries. Noah Webster's Compendious Dictionary of the English Language (1806), in its various revisions,8 was the leading authority during the greater part of the nineteenth century, and its chief rivals at the close of the century were two other American works, the Century and the Standard. The greatest of all dictionaries, the New English or Oxford, the first volumes of which were published in 1884 and the last in 1928, was begun under the editorial supervision of one Scotsman, Sir James Murray, and finished under that of another, Sir William Craigie.

And what of the authority of the dictionaries, their influence upon the speakers and writers of the language in the three centuries that have elapsed since Bullokar and Cockeram? It is probably correct to say that it is not until the eighteenth century, with the rise of the middle class to social prominence and the development of middle-class anxiety about "correctness" in speech,9 that dictionaries and grammars (which increase in numbers and prestige at about the same time and in a parallel way) take on anything like the character that to many

7 Cf. again Weekley's article, and McKnight, pp. 365 and 366.

8 The elaborate two-volume revision published in 1828 was suggestively called An American Dictionary of the English Language. This work is termed by Professor Krapp "the most significant contribution to English lexicography between Dr. Johnson and the appearance of the first volume of the New English Dictionary," English Language in America, Vol. I, p. 362. Its later forms, particularly the first edition to be designated Unabridged (1864), were successive improvements over the earlier.

9 Cf. Wyld, H. C., A History of Modern Colloquial English, New York (Dutton), 1920, pp. 18-20.

Modern English: British and American 73

minds they now possess: the final arbiters of speech.

The general veneration of "rules" in the early and middle

eighteenth century—implicit, to take a single example,

in the doctrine of Pope's Essay on Criticism that litera-

ture is to be both produced and judged according to a

formula—naturally lends its weight to the treatment of

language as something that must be adjudged, once and

for all, as either correct or incorrect.10 Doctor Johnson's

dictionary purported to give the correct meaning,

spelling, and accent of all words then existing in accepted

usage, and these words Johnson conceived to be all that

would ever be necessary. This eighteenth-century atti-

tude toward language and toward the dictionary as the

final authority on language has been inherited in more

recent times by both the users and the makers of dic-

tionaries (despite the latters' frequent protestations to

the contrary) to what is really an astonishing degree.

Recent dictionaries, it is true, maintain that they do

not profess, like Doctor Johnson's, to tell what the stand-

ard of language should be; they merely record, in tan-

gible form, the standard already set by usage. In

theory, their attitude is much less dogmatic and con-

servative than the eighteenth-century one; in practice,

it is not so very far removed from it.11 New spellings

are recorded with great reluctance, new words sometimes

10 C. C. Fries has an excellent chapter (pp. 1-31), "The Rules as

Measures," in his Teaching of the English Language, on the part that

grammars have played. Cf. also McKnight's Modern English in the

Making.

11 There is even a nearer affinity in temperament between Doctor

Johnson and certain modern lexicographers than is perhaps generally

realized. Note, for example, the magisterial tone of such utterances as

those of Dr. F. H. Vizetelly, editor-in-chief of the Standard: "Keeping

Step with Speech" and "A Matter of Pronunciation," Atlantic Monthly

(August, 1930, and February, 1931, respectively).

74 The Development of Modern English

knock long at the gates for admittance, and the indication of a change in pronunciation lags far behind the actual usage of good speakers.

The attitude of even well-educated persons toward the dictionaries is often curiously naïve. “What does the dictionary say?” is the usual question when any problem regarding the sound or form or meaning of words arises.

The implication, of course, is that there is only one verdict to be found in any dictionary, and that dictionaries, of any kind and any date, are all equally valuable. No thought is given to the possibility of consulting the wrong dictionary,12 or to the discrepancy between dictionaries, or to that between dictionaries and good usage.

No, “the dictionary,” however that authority is conceived, is all-sufficient; to question its omniscience is heresy. In the extremely conservative attitude of the dictionaries and in the docile acceptance of their authority by the great majority of their users we have a most important influence upon the development of the modern language.

It is clear that when dictionaries, grammars, and handbooks of usage are widely circulated and uncritically accepted as the final word, the formerly free development of the speech is constantly kept in check.

Americans seem to inherit the eighteenth-century attitude, or the middle-class veneration of authority, to a greater extent than Englishmen of equal education and culture.

The educated Englishman is more prone to consider that his own observation of what constitutes good practice in speech takes precedence over any

12 Wrong, for example, because of insufficient authority, because of date, or because of national or sectional bias.

Modern English: British and American 75

written guide. Much more than the educated American,

he is likely to think of dictionaries and grammars as

being primarily for those whose background and upbring-

ing have not brought them in contact with the best

tendencies in English. For himself, he quite frequently

scorns their use very much as he would scorn to follow,

in other aspects of social intercourse, the dictates of a

book of etiquette. The following observation of a

traveled and cultured Englishman illustrates this point

of view: "When I came to America nothing struck me

more forcibly than the respect paid to the dictionary,

and the disposition to fly hotfoot to it when any question

arose."13 If we have rejected the eighteenth-century

dogma that dictionaries make usage, we still seem reluc-

tant to accept, in all its implications, the converse of the

proposition.

In discussing Modern English, it has already proved

impossible to treat it entirely as a unit. Reference has

just been made to contrasting attitudes, on parallel

social levels, in England and in the United States. It is

of course evident that, from the seventeenth century on,

the English-speaking world expands with a rapidity and

on a scale paralleled in the history of no other language.

From the point of view of an American, it is quite essen-

tial to treat the evolution of the modern speech, at least

in part, from the angle of American English.

Is that, in the first place, the proper designation? Is

it "American English," or should we rather discuss, as

some would have it, "the American language"? It is

curious that two of the leading books that have been

13 Letter to "Contributors' Column" of Atlantic Monthly, January,

1931, p. 30.

76 The Development of Modern English

produced by the recent attention to American speech indicate in their very titles these opposing points of views: Gilbert M. Tucker's American English and H. L. Mencken's The American Language.14 The divergence of opinion of course is absolute: American speech is either conceived of as a special section of a larger whole or it is thought to have departed so widely from its seventeenth-century point of contact with the English of England as to deserve to be considered an entirely separate entity. Before attempting to argue for either position, it may be useful to summarize the history of the debate. What has been the consensus of American opinion, in the Colonial period and in the successive eras of our national life, on the relation of American speech to the English language?

The usual Colonial attitude was quite naturally one that accepted without cavil the idea that the standard for the English-speaking world was to be found in southern England, and deprecated any such departures from this as doubtless did actually occur in the spoken or written practice of the Colonists. Parenthetically, it may be observed that this provincial and apologetic attitude has not yet been entirely lost. Richard Grant White, arbiter elegantiarum in the middle nineteenth century, is still convinced that American English can be good only in so far as it is identical with British English.15 Even in the twentieth century, particularly in New

14 I find that Sir William A. Craigie has already made this observation ("An American Language," Saturday Review of Literature, Vol. VII, No. 31, Feb. 21, 1931, p. 614); and, very fittingly, he further points out that Professor Krapp avoids definite issue by using as the title of his most elaborate survey of the field, The English Language in America.

15 "In language, everything distinctly American is bad" (Atlantic Monthly, Vol. XLI, p. 495), quoted in Emerson, p. 106.

Modern English: British and American 77

England, this feeling occasionally persists, in scarcely diluted form. "It is a safe rule to require every perceptible departure from English usage to justify itself"—so runs the general advice given in a rhetoric published in 1917 and widely used in American colleges.16 This attitude, of course, scarcely enters into the dispute between American English and the American language, for it amounts almost to a denial that either has a right to exist.

After the Revolution, there was a sharp reaction from the general unquestioning acceptance of British authority, in language as in other matters; indeed, a demand for a "genuine American language" is to be heard. This seems even to have gone to the grotesque length of demanding that an entirely different language—Greek or Hebrew, for choice—be substituted for English.17 Among the more realistically minded, however, such schemes had no place; nevertheless, the conception that the English language as spoken in America would in time be widely different from the English language as spoken in England and might eventually become a separate dialect was not uncommon. Whether such a development should be helped or hindered was, of course, the bone of contention. Several leading men of the new republic—Franklin18 and Jefferson,19 for example —felt that it should be helped.

16 Greenough, C. N., and Hersey, F. W. C., English Composition, New York (Macmillan), 1917, p. 353.

17 Just how serious such suggestions were is a little dubious; cf. Mencken, American Language, 3rd ed., p. 47, and Cambridge History of American Literature, Vol. I, p. vi.

18 Cf. Mencken, p. 48.

19 Cf. Mencken, p. 1, and Krapp, English Language in America, Vol. I, p. 9.

78 The Development of Modern English

The leanings of Noah Webster were undoubtedly, and more particularly in his earlier work, toward an American language. He conceived that the American language of the future would be "as different from the future language of England, as the modern Dutch, Danish, and Swedish are from the German, or from one another."20 "Let us," he said, "seize the present moment and establish a national language as well as a national government."21 But Webster was forced to recede from this extreme position, as the title of his chief work, An American Dictionary of the English Language (1828), significantly indicates. The more conservative and more British character of the work of his great rival, J. E. Worcester, the first form of whose dictionary was published in 1830, is obviously one reason for the moderating of Webster's patriotic ardor. Worcester, indeed, was long preferred by the Tories in language.22

Through the nineteenth century and in the twentieth century, the debate between conservative and radical, on this question as on other linguistic matters, has continued. Men of letters and professed students of language have, for the most part, been more inclined to the conservative tradition; they have felt that it was of more importance to emphasize the likeness of our speech to that of England than its differences. Until quite recently, only works deliberately couched in dialectal speech—Lowell's Bigelow Papers, Holmes's Autocrat, Harris's "Uncle Remus" stories, and so on—have ordinarily striven for a distinctly American idiom.

20Dissertations (1789), p. 22 (quoted by Krapp, p. 9).

21Quoted by Mencken, p. 45.

22"As late as the last decade of the nineteenth century Harvard University, in its English entrance requirements, specified Worcester's dictionary as the authority in spelling," McKnight, op. cit., p. 490.

Modern English: British and American 79

Occasionally, however, it is to be noted that an American writer has protested against the blind following of British tradition; the most famous protest is perhaps that of Lowell in his preface to the second series of Bigelow Papers. William Dean Howells23 and Mark Twain24 have also joined their voices to Lowell's. Walt Whitman too, who has been called by Louis Untermeyer "the father of the American language,"25 is to be remembered for his conviction that America both needed and would have a distinctive speech, and for his experimentation to that end: "The new world, the new times, the new people, the new vistas, need a tongue accordingly—yes, what is more will have such a tongue—will not be satisfied until it is evolved."26 Recently, Mr. Vachel Lindsay has gone on record with the conviction that we need not one but two dialects or planes of utterance for American purposes:

What do shopgirls mean when they say "Talk United States"? We, the citizens of the United States, in order to form a more perfect Union, to provide for the common defense and promote the general welfare, have ordained and established two languages, the United States Language and the American Language, for ourselves and our posterity; the one for informal, the other for formal occasions, one inherited from the Mermaid Tavern, the other from the Globe Theatre, and both of them changed, yet hereditary in style. No shopgirl ever yelled at a floorwalker, "Talk American."27

23 Cf. "The Editor's Study," Harper's Magazine, January, 1886.

24 As in "Concerning the American Language," in the volume Tom Sawyer Abroad, etc.

25 As quoted by Mencken, p. 90.

26 As quoted by Horace Traubel in the preface to An American Primer. For an account of this, see Leon Howard, "Walt Whitman and the American Language," American Speech, Vol. V, No. 6 (August, 1930), pp. 441–451.

27 Review of Charles Cestre, "An Introduction to E. A. Robinson," Modern Language Notes, Vol. XLVI, No. 5 (May, 1931), p. 323.

80 The Development of Modern English

Self-consciousness about our speech seems in the air today, and, as we have seen, it does not necessarily take the form of apologizing for departures from the British standard. The question inevitably arises: Are the differences great, and are they important? Further, of what do the chief differences consist? These questions will be glanced at more fully in later chapters, when such aspects of language as spelling, pronunciation, vocabulary, and syntax are under discussion. Here, only certain observations about the general contrasts between American and British English will be attempted. Incidentally, the warning may be voiced that accounts of American English tend usually to exaggerate the differences and minimize the resemblances. One who investigates the subject is prone to play up the striking but often superficial contrasts in detail, and neglect, as too familiar and too obvious, the identity in more weighty but often less conspicuous matters. Nevertheless, one finds certain contrasting tendencies in speech, when one compares American with British practice, that seem to admit of fairly safe generalization.

It will be quite commonly agreed, in the first place, that American speech is far more uniform than British. We have no dialects, in the sense in which they exist in England; the local speech of any part of America, with all but the rarest exceptions or (according to the point of view) with none at all, is immediately intelligible to a visitor from any other part. Further, we have no geographical standard that stamps a special value upon the speech of a particular city or state. Our ideal is unquestionably national, rather than local or sectional, usage. A third aspect of our linguistic solidarity is this: the gulf that separates the speech of the most highly educated

Modern English: British and American 81

from that of the average citizen is considerably less than the corresponding one in England.

If these generalizations seem, like some of Noah Webster's utterances, put forward with patriotic bias, that objection will scarcely be urged against what seems to the writer a necessary corollary. In the words of Professor Shorey, it is this: “ . . . it is probably true that the educated Englishman expresses himself in sounder idiom and more discriminating choice of words than the American of the same class.”28 If, in other words, the average of American speech is above the average of British speech, it is still conceivable that the best American speech falls below the best British speech. This dictum, it may be argued, will hold for both the spoken and the written tongue.

A familiar misconception of the relation between American speech and English is that which describes our language as relatively nearer the English of Shakespeare's day. A distinguished writer on language29 has expressed this fallacy thus: “Owing also to the curious fact that the colony usually tends to preserve the language of the time of separation, we retain some older words and some meanings of words not now found in England.” That one may find a number of illustrations of archaic words and archaic meanings (by the present British standard) preserved in living use in America does not really establish this “curious fact,” which has been given a surprisingly wide credence. Professor Krapp30 aptly

28 Paul Shorey, "The American Language," Academy Papers, pp. 156 and 157.

29 Emerson, O. F., History of English Language, New York (Macmillan), 1894, pp. 105 and 106.

30 See English Language in America, Vol. I, pp. 49-51, for an extended discussion of the point.

82 The Development of Modern English

points out that the usual illustration of this supposed law, the speech of the mountaineers of Kentucky and Tennessee, is erroneous in that “the speech of these communities is archaic . . . not because it is transplanted speech but because the communities in their general social life have had few social contacts. Thus the speech of Iceland is archaic as compared with that of Norway.” The “arrest of development,” in other words, is due not to the fact that the language was transplanted, but rather to the fact that the regions to which it was transplanted have been curiously inaccessible to outside influence. It is folly to suppose that such a condition applies in most of America, or that American English has not altered as thoroughly as British English from their seventeenth-century meeting place.

To deal briefly, then, with the general aspects of the contrasts between American and British English today. As to vocabulary, it is fair to say that popular belief grotesquely exaggerates the number of words that are peculiar to American English or have a different meaning here. A partial explanation, perhaps, is that in a group of words with which the American visitor to England must come in contact—the vocabulary of travel—the differences are unusually great; it has been suggested that this is so because railroad travel developed after the separation, and the source for railroad terms in America was the vocabulary of the steamboat, while in England it was that of the stagecoach. At any rate, the differences here31 are far from being typical. The proof of the essential identity of the two vocabularies is

31 Some of them, for example, are listed by Mencken, p. 100, and by Greenough and Kittredge, Words and Their Ways in English Speech, op. cit., p. 271.

Modern English: British and American 83

simply the fact that no American has the slightest difficulty in reading a book by an Englishman and printed in England, or an Englishman with a book by an American and printed in America—unless, perchance, the book is unusually full of either British or American slang. (An English edition of O. Henry is furnished with a glossary.) Nor is the fact that an occasional word is spelled differently in the two countries enough of an obstacle to be worthy of the name. American practice has been a trifle more hospitable to spelling reform than British, but the divergence, on the whole is small and unimportant.

It is not to be denied, on the other hand, that an American does sometimes experience difficulty in listening to an English lecturer, and an Englishman in understanding an American speaker. One is tempted to say that the American’s difficulty is the result of ignorance and the Englishman’s the result of affectation; in both cases, more careful and more patient attention would probably reduce any lack of understanding to the vanishing point. What difficulty there is, it should be noticed, comes from somewhat different habits of pronunciation—not, to any significant extent, from different vocabularies.

And what are these different habits of pronunciation? As satisfactory a generalization as any, so far as the less admirable but unfortunately more conspicuous aspects of pronunciation are concerned, is Mr. George Arliss’s recent dictum:32 “The chief fault in speech in America is sloppiness and the outstanding defect in England is snippiness.” It may also very possibly be true, as Americans are often told by Englishmen, that our speech

32 “Mr. Arliss Makes a Speech,” Atlantic Monthly, February, 1931.

84 The Development of Modern English

is harsher and more nasal than theirs; but, at least, it has,

in the opinion of many disinterested European observers,

the merit of superior intelligibility. That which is

sloppy may be less distinguished but at the same time

more understandable than that which is snippy.

If American speech, particularly the spoken rather

than the written variety, is correctly to be designated as

"sloppy" or "slovenly," the explanation may lie, where

Mr. Arliss finds it, in the educated American's fear of

appearing superior—a sensation much less dreaded by

Englishmen of a corresponding social level. Mr. James

Truslow Adams not long ago christened this "the mucker

pose": the affectation, on the part of men of education

and culture, of the language of "uneducated half-wits."33

One can scarcely deny that there is some truth in this;

that one of the less fortunate results of the democratic

dogma of general equality is that men speak more nearly

after the same pattern than in an aristocratic society, and

that bad speech, like bad money, tends to drive out the

good.

But this, as has been suggested, is much more true of

the spoken than of the written language. Strangely

enough, the very American who speaks, as Mr. Adams

has put it, in an idiom reminiscent of the longshoreman's,

is meticulously correct, to the point of being pedantic,

when he puts his thoughts on paper. With all the

license that he permits himself for the informal purposes

of spoken discourse, he still feels an extreme sense of

restraint, an abiding respect for authority and tradition,

when he addresses himself to formal writing. Here it is

33 "The Mucker Pose," Our Business Civilization, New York (Boni), 1929.

that the eighteenth-century feeling for "correctness,"

together with that curious veneration for such guides as

dictionaries and handbooks of speech that we have

already noted, plays its part. The result is likely to be a

diction that is approved by the best authorities, and a

style that is utterly derivative and lifeless. This is

American English in its provincial or parochial aspect.

It will doubtless occur to the reader, however, that this

is not all of the truth. The written language in America

is surely not always to be condemned as colorless and

dull. Side by side with the following of tradition, and

in absolute opposition to it, there is an unmistakable

striving for warmth, for novelty, and for vividness.

It was this purpose, in part, that animated Walt Whit-

man. It is the same purpose, though not always with

happy results, that today animates many of our journal-

ists, our advertisers, and our purveyors of slang and local

color. This phase of American speech it is that has

tempted commentators to apply the term "Elizabethan"

-not with reference to seventeenth-century survivals in

America, but with reference rather to a new delight in

language, an exuberance in playing with words, that is at

least remotely akin to the old. It is usually indulged in

-and here its presence is to be reconciled with the

existence of the conservative feeling for authority and

tradition—with a sense of its appropriateness for less

formal purposes only. The future, it is to be hoped, will

see the coalescence of these two opposite tendencies;

one may, at any rate, look forward to a time when they

will not go their separate ways—when, rather, traditional

correctness will be enlivened by colloquial freedom and

inventiveness, and at the same time unbridled license

will be kept in check by a salutary respect for authority.

86 The Development of Modern English

To return to the question of "American English" as opposed to "the American language." The latter phrase, as has doubtless been apparent, seems to the writer an exaggeration of the real status of our speech.

The differences that have been pointed out between American and British English seem, in the end, far too superficial and trivial to allow one to think of the two brands of English in any other way than as parts of a larger whole.

Moreover, the divergence that does exist seems to be growing less rather than greater. There is every reason why this should be so.

Communication between England and America becomes every year not only quicker but, some would add, more intimate. We have listened, on the radio, to the British Premier both at home and in America; and it is significant that a frequent comment on Mr. MacDonald's American broadcasts was that his speech was so nearly like an American's that the infrequent departures from American English seemed almost an affectation.

The radio is not alone as an instrument of international standardization of English. There are the "talkies," and the increasingly international character of literature in English.

Not that complete standardization is likely ever to be attained. Indeed, one may express the pious hope that this will never happen, for, as has been well remarked, it is the byways rather than the highways of speech that are piquant and interesting; it would be deplorable indeed if there were no national or local deviations from the international standard.

Our question has been succinctly answered by an eminent British lexicographer, who is now engaged on an American dictionary34:

34 Sir William A. Craigie, op. cit., p. 615.

Modern English: British and American 87

A distinct American language, clearly marked off from other forms of English, in the same way as one Germanic or Romanic language is from another, is less likely to arise than seemed possible a century ago . . . The new American Language . . . is . . . impossible, for the reason that it can no longer develop by itself, but must keep in touch with that wider English which concerns the world at large.

Truly of more moment than the relation of American to British English is the subject the last words suggest. Our survey of Modern English may appropriately conclude with a consideration of the topic of “World English.”

One of the most interesting and important aspects of Modern English is its constantly increasing use as an international language. That such a general medium of expression is a real desideratum we have the experience of history to testify. In the Middle Ages, Latin occupied this position, and in modern times French has approximated it, especially as the language of international diplomatic relations. Before examining the present claims of English as a world speech, however, it may be well to glance at what many consider a possible alternative: the use, not of any of the existing vernaculars of the world, but rather of an artificially created universal language, such as Esperanto.

The ideal of an artificial speech, scientifically constructed so as to combine the merits of some of the leading naturally developed languages and at the same time embody none of their defects, is by no means a new thing. Some of the projected artificial languages do not, it is true, fulfil the former of these conditions; that is, they are based not on one or more existing languages, but are purely a priori schemes. None of these,

88 The Development of Modern English

however, is seriously advocated at present35; it is quite generally recognized that a universal language must be founded on one or more of the vernaculars of the world.

To go into the many variations of the project36—such as Volapük, Ido, "Latino sine flexione," Novial—would take us too far afield. A few words must nevertheless be said about Esperanto, the claimant favored by the majority of the advocates of a universal language.

Esperanto has Latin, as the most nearly international and neutral of elements, as the chief basis of its vocabulary; its grammar is exceedingly simple and its spelling is phonetic. That it has won a real, though necessarily limited, measure of successful adoption, is indicated by the support given it by the League of Nations, the International Telegraphic Union (1925), and the Union Internationale de Radiophonie (1927).37 On the other hand, it must be said that the arguments that have been urged against it in the past38 are still largely valid: at most, it now occupies a position analogous to that of a cable code, without any important use in scientific, literary, or social communication; and one advantage that has been urged in its behalf, its strict uniformity all over the world, is partly vitiated by its different pronunciation in various countries. Further, its pretensions as the universal language are obviously damaged by the

35 With the possible exception of Ro. See "Universal Language," Encyclopedia Brittanica (14th ed.), Vol. 22, p. 861.

36 Cf. particularly L. Couturat and L. Leau, Histoire de la langue universelle (Paris, 1903), and Otto Jespersen, An International Language.

37 Encyclopedia Brittanica, article op. cit.

38 Those, for example, of Professor Krapp, Modern English (1909), pp. 40–43.

Modern English: British and American 89

refusal of partisans of the ideal of a universal language39

to give it unanimous support.

If, then, the project of an artificially constructed uni-

versal language has so far met with failure or only a very

limited measure of success, what claims may be advanced

for English as an international language? It seems

more and more certain that if any living speech attains

this position, it will be English.40 In our own day,

English has come to rival French in the field of diplo-

macy, and it is apparently destined to replace French as

the "second" language—the language most useful for

the traveler—all over Europe.41 In other parts of the

world, its prestige is still more commanding. Beach-la-Mar or Sandalwood-English, spoken and understood

all over the Western Pacific, and Pidgin-English, known

in China and to some extent in Japan and in California,

are the most conspicuous examples of corrupt forms of

39 Among them, perhaps the most eminent of living linguists, Professor

Jespersen. It should be added, however, that Jespersen now feels that

there are enough points of similarity among the leading projects looking

toward an international language—including Esperanto, Ido, Nov-

Esperanto, Latin without Flexion, Occidental, and his own creation,

Novial—to justify the hope that a single adequate International Auxiliary

Language will some day emerge. See "Interlinguistics," his contribution

to the symposium International Communication.

40 A middle position between those who support the claims of an arti-

ficial language like Esperanto and those who support the claims of

English is taken by those who advocate an artificially simplified form of

English. This speech is known as "Basic English"; see Ogden, C. K.,

Basic English, London (Kegan Paul), 1930, and The System of Basic

English, New York (Harcourt, Brace), 1934.

41 It may be noted that in recent times French scholars, realizing

the advance of English, have sometimes advocated an Anglo-French

condominium: cf. Bréal, M. J. A., "Le choix d'une langue internationale,"

Revue de Paris, year 8, v. 40, pp. 229–246 (July 15, 1901); and Chappel-

lier, Paul, Une nouvelle solution de la question de la langue universelle,

Paris, 1901.

90 The Development of Modern English

English that have been evolved from the contact with exotic tongues42 and spread over large areas. It is said, incidentally, that the American share in these trade languages is becoming larger than the British.43 But it is not, of course, merely in such corrupt jargons that Modern English has gained a world-wide ascendancy; in its more standard British and American forms, it is, especially since the World War, more and more dominating the civilized, and a large part of the uncivilized, world.

This leadership among the languages of the world has been achieved well within the Modern English period. In the Renaissance, and probably as late as the period of the Restoration, the speakers of English were fewer than the speakers of at least four other European languages—German, French, Spanish, and Italian. Even in the eighteenth century, English was still, for a time, outdistanced in numbers of speakers by four other European languages, since, if Italian had been left behind, a new rival, Russian, had asserted itself. In the nineteenth century, however, English came rapidly to the front, largely as a result of the swift increase in the population of the United States and of the British colonies. Probably by the middle of the century it had outdistanced its competitors. Estimates, in millions

42 For a description of these and similar jargons, see the section “Pidgin and Congeners” in Jespersen's Language (pp. 216-236). Pidgin, as Jespersen explains, is derived from the Chinese distortion of business.

43 Cf. Mencken, p. 393. Mencken also quotes (p. 391) a striking testimony to the use of English as a world language in the account of an English traveler who observed Chinese mandarins speaking English not only in conversation with German ship's officers and Japanese travelers but even with one another. Since they came from various provinces of China, they found English a more convenient medium than their less standardized native tongues.

Modern English: British and American

91

of speakers, made at several times since then, are as

follows:44

English

German

Russian

French

Spanish

1868.............

60

52

45

45

40

1890.............

111

75+

75

51+

42+

1900.............

116(123)

75(80)

70(85)

45(52)

44(58)

1912.............

150

90

106

47

52

1921.............

170

871⁄2

1201⁄2

45

65

The figures, necessarily only approximations, have at

least the merit of coming from a number of different

sources and representing various points of view. Their

testimony to the recent and rapid increase in the number

of those speaking English is perhaps for this very reason

the more impressive. It is evident, of course, that they

do not tell the whole story as to the languages of the

world; but China's 400 millions speak dialects that are

mutually unintelligible45 and that, besides, scarcely

spread beyond her borders; and a similar observation

may be made of those who speak cognate varieties of

Indic—the Indo-European languages of India—per-

haps 230 millions in numbers. Clearly, any rivalry for

the position of a world language must come from the

tongues of Europe; and the figures that have been quoted

are a striking evidence of the way in which English has

44The following sources have been used: for 1868, Brackenbusch, W.,

Is English Destined to Become the Universal Language?, Gottingen, 1868,

and for 1890, Buehler, H. G., A Modern English Grammar, New York,

1900 (both as quoted by Mencken, pp. 383 and 384); for 1900, the mini-

mum and maximum estimates quoted by Jespersen, Growth and Structure

of the English Language, p. 252; for 1912, Hickmann's Geographisch-

statistischer Universal-Atlas, also as quoted by Jespersen, p. 252; and for

1921, L. Tesnière, in Meillet, Langues, as quoted by Bloomfield, Language,

p. 44.

45Though one of them, North Chinese, in all probability has more

native speakers than even English, and another, Cantonese, is not very

far behind.

92 The Development of Modern English

outdistanced its European competitors. In numbers of

native speakers, English is rivalled only by North

Chinese; in numbers of native and foreign speakers,

it is quite unrivalled by any language in the world. The

fact that English has gained the ascendancy is not to be

disputed. There may be difference of opinion, however,

as to why it has done so.

English may conceivably enjoy its present favored

position for either of two reasons: its spread may be

accounted for by external circumstances, or it may have

come about because English is the language best fitted

to be the international tongue. Probably it is nearer the

truth to say that external and internal causes have

worked hand in hand: that English has been so marvel-

ously favored by political, economic, and social forces

that it would have a world-wide vogue even were it

ill-fitted to be an international language; but also that

there is something inherent in the language itself that

has made it particularly useful in such a capacity.

The former point surely needs no special demon-

stration. England's rise as a world power, beginning in

the days of Elizabeth, is accompanied, step by step, by

the ascendancy of English as a world language. The

political union of England and Scotland under James I

in 1603, followed by the formation of the United King-

dom of Great Britain and Ireland, helped to make it

certain that English would be used, and in a form

substantially the same, throughout the islands and in

the newly-founded Colonies. In the New World, the

fall of New Amsterdam in 1664 and that of Quebec in

1759 assured the triumph of English over two important

rivals, Dutch and French; and the expansion of the

United States brought it about that Spanish, the only

Modern English: British and American 93

remaining competitor of English on the North American continent, was reduced to a position of relative insignificance. In the later half of the eighteenth century and throughout the nineteenth century, the English language was established in every corner of the earth through conquest, colonization, and the commercial ascendancy of the English-speaking peoples. It may even be conjectured, as Professor Krapp suggests,46 that the tenacity with which the British and the Americans hold to their own language and their own ways has been a real factor in spreading the English language; for if the English-speaking people will not learn a foreign language, it is evident that foreigners, to trade with them, must learn English. The contrast has been observed, for example, in the far greater readiness of the Germans, as compared both with the English and with the Americans, to learn the Spanish language and adapt themselves to Latin ways, when engaged in commerce with South America. Doubtless, the more complacent Anglo-Saxon attitude is bad for trade, when better linguists, like the Germans, are competitors; but, just as evidently, it has proved in the past to be an appreciable factor in the world-wide use of English.

Having, by the middle-nineteenth century, won first place among the languages of the world, English has been favored by more recent developments so that its primacy is less and less questioned. Inventions that make communication between nations easier have inevitably lent themselves to the spread of English. So we may regard the telephone, the telegraph, the radio, the movies, and the “talkies.” Here too, it may be said, it is often

46 Modern English, p. 39.

94 The Development of Modern English

the American brand of English that is spread abroad, frequently to the chagrin of the British. A Danish visitor to America is thus quoted in the New York Times:47 “An amazing interest in American talkies is evident in all of Scandinavia. English is taught in the schools . . . and now that American talkies are available in so many Danish towns, the whole populace is interested in ‘brushing up’ on English pronunciation.” This observer predicts that American “talkies” will soon make English the universal language of the Continent. Time may prove the prophecy oversanguine, but there can be little doubt that the increasing prestige of English finds fresh illustration here.

So far, we have dealt with external factors in the world-wide use of English. Can as much be said for the intrinsic qualities of the language? Is English, on the whole, well- or ill-fitted for its rôle of world-speech? It is undoubtedly difficult for one whose native tongue is English to see the problem with proper objectivity. More valuable testimony, quite certainly, is that of a foreign observer who has studied English, and other languages, intensively, and who can more dispassionately assess its merits and shortcomings. Professor Jespersen, certainly as competent a foreign observer as could possibly be found, summarizes his impressions thus: “The English language is a methodical, energetic, businesslike, and sober language”;48 and again (with particular reference to its increasing use as a world language),49 “It must be a source of gratification to mankind that the tongue spoken by two of the greatest powers of the world

47 October 15, 1929.

48 Growth and Structure of the English Language, p. 17.

49 Ibid., p. 252.

Modern English: British and American 95

is so noble, so rich, so pliant, so expressive and so interesting. . . . " In still another place,50 he states

his view of English as compared with other languages in these terms: " . . . it seems to me positively and

expressly masculine, it is the language of a grown-up man and has very little childish or feminine about it."

The foregoing terms represent, with entire adequacy, the usual favorable view of Modern English. It is

most frequently praised for its businesslike simplicity—

in sound-system, in grammar, and in at least the native and more frequently used section of its vocabulary.

This simplicity is commonly thought of as that which makes it easy for a foreigner to learn it, and hence, that

which particularly adapts it to use as a world tongue. To modify this judgment, however, we have such an

admission as the following, from the pen of an enthusiastic defender of this very quality of simplicity:51 "The

foreigner essaying it, indeed, finds his chief difficulty, not in mastering its forms, but in grasping its lack of forms."

More subtly, and from a different angle, the apparent simplicity of English has recently been declared to be a

delusion and a snare:

The fact that a beginner in English has not many paradigms to learn gives him a feeling of absence of difficulty, but he

soon learns to his cost that this is only a feeling. . . . The simplicity of English in its formal aspect is . . . really a

pseudo-simplicity or a masked complexity . . . He [the foreigner] may well feel that the apparent simplicity of

English is purchased at the price of a bewildering obscurity.52

50 Ibid., p. 2.

51 Mencken, p. 387.

52 Sapir, Edward, "Wanted: A World Language," American Mercury, Vol. XXII, No. 86 (February, 1931), pp. 202-209. (Reprinted by permission of author and publisher.) These excerpts cannot do justice

96 The Development of Modern English

Granting that there is some truth in these strictures and that the superficial impression of simplicity that English gives is somewhat ambiguous, one may still feel that its forms, its words, and its sounds compare favorably, in the ease with which at least approximate mastery of them can be attained, with those of other languages.

Its vocabulary has the enormous advantage of being compounded almost equally of Teutonic and Romanic elements, so that a good part of it is already familiar to the speakers of many other European languages.

Its grammar is so inconsiderable that the language has been called (of course, inaccurately) " the grammarless tongue." In inflections and word-order, as we have seen, the modern speech has greatly simplified and regularized the practices of Old English.

The one great stumblingblock in the way of the foreigner who would acquire English is, as will be granted on all sides, the spelling—“that pseudo-historical and anti-educational abomination."53 While it is frequently pointed out that the grotesquely illogical spelling of Modern English forms a most difficult barrier to its use as a world language,54 it may, nevertheless, be contended that this one great and obvious disadvantage does

to the keenness of the writer's analysis of the deficiencies of English as an international language. Professor Sapir presents this point of view also in his contribution to International Communication, a symposium by Herbert N. Shenton, Sapir, and Jespersen.

53 Jespersen, Growth and Structure of the English Language, p. 250.

54 This, for example, was the general tenor of the speeches made at the English Language Congress held in connection with the Sesqui-Centennial Exposition in Philadelphia, November 26 and 27, 1926. "The principal topic of the congress was "English as the International Secondary Language." A Swedish project to spread the use of English by improving its spelling advocates the system known as "Anglic." See Zachrison, R. E., Anglic, New York (Stechert), 1930.

Modern English: British and American

97

not overbalance the several important advantages that

have been enumerated. When every just deduction has

been made, it is reasonable to regard the language as

on the whole well fitted by its own positive merits for

that larger usefulness which outward circumstances have

more and more conspired to give it.

References for Further Reading

The histories of the English language referred to on p. 37

and p. 64, and, in addition:

Fries, C. C., The Teaching of the English Language, New York

(Nelson), 1927.

Jesperson, Otto, An International Language, New York

(Norton), 1929.

Krapp, George P., The English Language in America, New

York (Century), 1925.

Mathews, M. M., A Survey of English Dictionaries, London

(Oxford), 1933.

Mencken, H. L., The American Language, New York (Knopf),

3rd ed., 1923.

Shenton, Herbert N., Sapir, Edward, Jespersen, Otto, Inter-

national Communication: A Symposium on the Language

Problem, London (Kegan Paul), 1931.

Society for Pure English, Tract No. XXXIV, London (Oxford),

1930, containing "Interlanguage," by T. C. Macaulay,

and "Artificial Languages," by J. A. Smith.

Tucker, Gilbert M., American English, New York (Knopf),

CHAPTER V

English Inflections

INFLECTION, broadly speaking, is the process of varying the form of the word to differentiate related meanings or uses. As the term is commonly applied, it refers most frequently to such syntactical distinctions as those of gender, number, case, mood, tense, voice, and so forth. In any strict use, inflection should mean variation rather than composition; that is, it should imply that there is a certain stable element in the word that is its nucleus, and that this is given variety of application by internal change or by the addition of certain prefixes or suffixes. These affixes, it is assumed, exist only for the purpose that has been described; they are not separate words or even independent word-elements. In practice, however, it is difficult or perhaps impossible to mark off the limits of true inflection. A point of contention, for example, is whether the -ly ending of such words as quickly and smoothly is to be considered as an inflectional suffix that transforms an adjective into an adverb, and whether the process is hence to be considered as parallel to the addition of the -s ending that transforms most of our singular nouns into plurals. In general, it will be more satisfactory to limit our application of inflection to changes that take place within the part of speech.

It should be added, too, that any strict conception of inflection must exclude analytical formations such as we now have in the future and the perfect tenses of the

English Inflections

99

verb. A synthetic verbal form for the future, like the

Latin habēbō or the French j'aurai, is a strict example of

inflection; the German ich werde haben or the English

I shall have, on the other hand, is inflection only by

courtesy.

Certain large questions suggest themselves when one

considers the general nature of inflection. For example:

If it is true that the history of English, and of many

other languages, includes a gradual simplification of

inflection, how are we to account for the former existence

of an elaborate machinery that has had to be quite largely

eliminated? Again, are we to suppose that, beyond the

comparatively complicated state of Old English inflec-

tion, there is a still greater complexity to be predicated

in West Germanic, in Teutonic, and in Indo-European?

Is language in its infancy characterized by a high degree

of inflection? Finally: What are the origins and the

general course of the development of inflection in lan-

guage as a whole?

The question of the prehistoric development of our

speech necessarily takes us somewhat afield and into

controversial ground. The orthodox theory may be

stated thus, in the words of Professor Jespersen (who,

however, formulates it in order to attack it):1

. . . an originally isolating language, consisting of nothing

but formless roots, passed through an agglutinating stage, in

which formal elements had been developed, although these

and the roots were mutually independent, to the third and

highest stage found in flexional languages, in which formal

1 Language, p. 367. (Reprinted by permission of the publisher,

Henry Holt and Co.) A full discussion of the question of an earlier

noninflectional stage in Indo-European will be found in Hirt, H., Indo-

germanische Grammatik (Heidelberg, 1921, 1927, 1929).

100 The Development of Modern English

elements penetrated the roots and made inseparable unities with them.

Inflection is thus conceived as having been gradually developed, through agglutination as an intermediate step, from a primitive “root stage” of language. An isolating language like Chinese is, therefore, the type of all language in its infancy. According to the generally received hypothesis, then, the process is something like this: Some of the originally separate roots became attached to other roots and lost their independent character, becoming first agglutinative formatives, and eventually inflectional affixes. Or, to view the development from the other angle, we may state the postulate thus: All our inflectional endings were originally independent roots that have gradually coalesced with other words.

Jespersen musters serious objections to the universal application of this theory. Two may be particularly mentioned: the fact that Chinese, the stock illustration of a language in its root-stage, has recently been proved to exhibit traces of earlier inflection, and does not therefore give us the picture of the primeval structure of our own language that it has been usually supposed to give; and the impossibility of proving that most of our inflectional suffixes go back to independent roots. Nevertheless, even Jespersen is willing to admit that the “agglutination theory” contains a good deal of truth, and that it does actually explain the origin of much (but not all) of our inflection. For the provisos and supplements that he feels are to be added to the usual explanation, the reader must be referred to his own treatment of the problem.2

2 “Origin of Grammatical Elements,” Language, pp. 367-395.

English Inflections

101

So far as the English language is concerned, the farther back one goes from the present, the more inflection does one encounter. Back of our slightly inflected Modern English is a higher degree of inflection in Old English, and a still higher degree in the oldest Teutonic language that has been preserved, Gothic; Sanskrit, of much greater antiquity than any Teutonic language, is a still nearer approach to the original complexity of Indo-European inflection. A single illustration of comparative degrees of inflection may be useful: the case distinctions for the noun, pronoun, and adjective. Sanskrit exhibits distinctive endings for eight cases: the nominative, genitive, dative, accusative, instrumental, ablative, locative, and vocative. Gothic, like Greek, has reduced the number to five; certain cases, that is to say, have lost distinguishing endings and so appear in one form. (Latin here is more conservative, since six cases are preserved.) Old English lost the distinctive form for the vocative and so exhibits, at most, five variations; practically, since the dative and instrumental are identical in the very great majority of instances, the number is reduced to four—and often to three, through the falling together of nominative and accusative. Modern English, of course, preserves three cases only in one category—the pronoun; the noun has but two case forms, and the adjective but one.

The general picture, then, that has been suggested is of language beginning with separate and independent roots but gradually developing inflection until it is so overburdened with its variety and complexity that expression is made difficult. Then the reverse process sets in, and words are more and more stripped bare of their inflectional distinctions. Both extremes of the

102 The Development of Modern English

development we can merely speculate about: the extent

of inflection in the remote past, and the limit of its

simplification in the distant future. But the general

direction, so far as English is concerned, is sufficiently

clear; as has frequently been observed, the structure of

English has become less and less like Latin, and more

and more like Chinese. In its larger aspect, and from

the point of view that most interests us, the conjectural

history of inflection may be represented thus.3

It is, of course, evident that the one side of the figure

represents mere hypothesis, however plausible; the other

side represents, almost entirely, a statement of facts.

Before dealing more specifically with the history of

English inflections, it may be convenient to suggest the

three general types of inflection that are possible. Inflec-

tion may occur in the beginning of the word, in the

middle, or at the end; it is a question of prefix, medial

change, or suffix. The terms usually employed to

designate the three types are initial, internal, and final

(or end).

3This diagram should not, of course, be interpreted to mean that

Gothic is derived from Sanskrit or Old English from Gothic.

English Inflections

103

The first category need not detain us, if the question concerns merely the present stage of the language. Modern English contains no example of initial inflection.4 The nearest approach is in such words as manservant and maidservant, and cock sparrow and hen sparrow;5 but these formations are evidently far removed from inflection in the strict meaning of the term. In Latin, a certain type of initial inflection is familiar; thus, the perfect tense cecidī (I have fallen) is differentiated from the present cadō (I fall) chiefly by the "reduplication" of the initial consonant to form an inflectional prefix. The same method is in Greek the typical way of forming the perfect: λúω (I loose), λéλυκα (I have loosed). Early Teutonic also utilized the method of "reduplication," to form certain preterit tenses; but the full forms of Gothic have regularly become, in Old English, "contract" forms, in which the "reduplication" suffix has merged with the radical syllable. For example, *hehald becomes hēold (held). A clearer instance of initial inflection is the frequent use in Old English of the ge prefix with the past participle of verbs, as in German. On the whole, it is fair to say that, even in the earliest stage of English, only a trace remains of true initial inflection.

The other two types, however—internal and final inflection—are abundantly represented in all stages of the English language. We have a familiar illustration of the former in such strong verbs as ring, rang, rung, and in such "irregular" nouns as goose, geese and man, men. These two types of internal inflection, incidentally,

4 Unless to, as the "sign of the infinitive," may be so counted. Cf. Aiken, English, Present and Past, p. 216.

5 Krapp, Modern English, p. 58.

104 The Development of Modern English

are historically quite distinct. Ring, rang, rung represents the process known as gradation or ablaut; that is, the variation of the vowel through a series of changes. Goose, geese and foot, feet, on the other hand, illustrate the process of umlaut or mutation; that is, the single change produced in a vowel by the influence, usually6 of the i-vowel, of the following syllable. Thus, the presence of the i in the second syllable of such Teutonic forms as *gōsiz and *fōtiz eventually brought about the “fronting” of the vowel of the first syllable, and the loss of the vowel that caused the transformation–hence the evolution of the Old English forms gēs and fēt. The same process has operated in other pairs of related words; for example, blood, bleed; doom, deem; proud, pride; foul, (de)file; fall, fell.

The third type of inflection, called final or end, is too common to need particular illustration. Its most familiar illustrations undoubtedly are the -s ending that we have come to feel as the regular plural sign for nouns, and the -ed ending that has a similar place as the indication of the past tense and past participle of verbs. The -s ending has, of course, other functions than the one mentioned; it may also be the indication of the third person singular, present tense, of verbs, or (with the apostrophe) the sign of the possessive case of nouns. Other inflectional endings of Modern English are -ing (of the participle and the progressive forms of verbs), and the two suffixes utilized in the comparison of adjectives, -er and -est. These five endings virtually exhaust the varieties of final inflection that are now in really

6 Always, so far as changes the effects of which are seen in Modern English are concerned.

English Inflections

105

living use—the ones that we apply, without question, to

the words entering the language or to old words used with

a new function or as a new part of speech. Initial or

internal inflection would never be employed in either

of the categories that have just been mentioned. A

noun borrowed from another language, for example,

must make its plural by end inflection7—and usually by

the addition of the -s, whether this ending is felt by

scholars to be appropriate to the noun in question or

not. It is this tendency that accounts for the preference

given (except in certain academic circles) to such plurals

as curriculums, criterions, stadiums, gladioluses, and

campuses over the forms in -a or -i. Even a well estab-

lished mutation plural can scarcely be carried over when

the noun is used in a new compound—hence the difficulty

with the plural of tailor's-goose or mongoose. How strong

is the preference for the usual types of end inflection is

neatly illustrated in the recent history of broadcast.

When this word began to be used as a verb, in connection

with radio, the popular form of the past tense was

broadcasted,8 in spite of the precedent of the simple (and

invariable) verb cast; an elaborate educational campaign

—over the radio itself—has scarcely yet driven it out

altogether.

The fact that only five endings can be enumerated as

the varieties of inflection that are in vital, active use in

Modern English is a striking testimony to the simplifica-

tion that the inflectional system of the language has

undergone. Is the English language, however, unique

7 Unless the foreign plural is also borrowed.

8 A parallel form, employed by certain sports writers, is the past

tense uppercutted. I have also seen the form forecasted: “Forecasted

trouble . . . was lightly brushed aside.” (Philadelphia Public Ledger,

December 9, 1931.)

106 The Development of Modern English

in this respect? The answer must be qualified. It is

evident, as we have already seen, that inflectional

simplification is characteristic of all Indo-European

languages since the establishment of the original elabo-

rate and intricate system of inflection. The Teutonic

group has been especially subject to simplification; even

members of it that have not been particularly exposed to

foreign influence, like Swedish and Dutch, have experi-

enced thorough inflectional leveling. Likewise, certain

languages outside the Teutonic system—French, for

example—exhibit parallel changes in their inflectional

character. The present spelling of French often exhibits

the greater variety of form that once existed but has

now been lost in the living or spoken language. Note,

for example, the appearance of aime, aimes, aime,

aimons, aimez, aiment. The Latin prototypes, of

course, are amō, amās, amat, amāmus, amātis, amant.

Five variations for the present indicative, that is to say,

are preserved in the French spelling—not a great simpli-

fication of the six forms of Latin. The pronunciation,

however, presents but three, since aime, aimes, and

aiment are sounded alike. Thus, the three variations of

form in modern French really represent almost as much

simplification as the two forms of the present indicative

in English.

On the other hand, if it is clear that inflectional leveling

is not peculiar to English nor to Teutonic, it is likewise

evident that the history of English inflection is striking

for both the rapidity and the thoroughness with which

leveling has been carried out. The primary reason,

though it is possible to exaggerate and easy to misstate

this, is the fact that different tongues or different

dialects have been in rivalry, in the relatively restricted

territory of perhaps half the island of Britain, in several

distinct times and ways. The original invaders of

Britain—the Angles, Saxons, Jutes, and Frisians—spoke

somewhat different dialects. The Scandinavian raiders

who conquered and settled in a large part of England at a

later date spoke a language differing from that of the

Anglo-Saxons as a group. The Normans, still later,

established in England and a language still further removed

from the Anglo-Saxon tongue, which was nevertheless, in

some degree, finally incorporated within it. The effect

of this diversity of tongues on the history of English

inflections was inevitably to make their variety smaller.9

When men of different tongues attempt conversation, it is

natural for each speaker to employ both his own and the

other’s language in as simple a form as possible—his

own, for his hearer’s readier understanding, and the

other’s, through his imperfect acquaintance with it.

Inflection is, on the whole, much less important for

conveying one’s approximate meaning than vocabulary.

Some degree of meaning can be conveyed by only the

roots of words. As Bradley remarks,10 “Many English-

Germans, and managed to make themselves fairly well

understood, without ever troubling themselves with the

terminations of adjectives or articles, or the different

ways of forming the plurals in nouns.” Just so, we may

imagine, were the elaborate inflections of Old English

simplified for the mutual understanding of the original

invaders themselves, and later in communication with

the Danes and with the Normans. If a simplified form

9 See Bradley’s treatment of “Simplification of Accidence,” in Making

of English, pp. 17–53, for a fuller exposition of this point.

10 Op. cit., p. 26.

108 The Development of Modern English

of the language was used to make it intelligible to others,

it was natural, in the end, to use the same simplifications

for all purposes. Only the most vital of the inflectional

distinctions eventually were preserved. Not in every

respect, perhaps, but in quite large measure, was it a

case of the survival of the fittest, or at any rate the most

essential.

Inflectional leveling once having begun on a consider-

able scale, it was undoubtedly accelerated and made more

far-reaching by the operation of the great principle of

analogy. This is the extremely important drift in

language through which the exception tends to conform

to the rule. When once a particular declension or

conjugation is established as the normal or typical one,

the rival types of inflection often yield to its influence;

they cease to be distinctive and became more and more

"regular." The most conspicuous illustrations in Eng-

lish are the -s plural ending for nouns and the -ed preterit

suffix for verbs; it is significant that the modern tendency,

however misleading the term may be historically, is to

call these the "regular" types of plural formation and

verbal conjugation. The pull of analogy is, of course,

resisted to some degree and at some points; otherwise,

we should have complete uniformity and only a single

type of inflection in each part of speech. In reviewing

the inflectional history of the various parts of speech, we

shall observe certain counteracting influences that have

served to prevent absolute standardization.

Before turning to the separate parts of speech, how-

ever, the question of the relative rapidity and scope of

inflectional leveling in the various stages of the language

should be considered. A familiar misconception, as

has been indicated, is to attribute the whole process to

English Inflections

109

the influence of the Norman Conquest, and hence to limit the operation of simplification to the early Middle English period. This misunderstanding has doubtless been fostered by the habit of giving to the three periods of English—Old, Middle, and Modern—the additional labels of “period of full inflections,” “period of leveled inflections,” “period of lost inflections.”11 Convenient and useful as the terms are, it may be pointed out that each one is in some measure inaccurate—that Middle English might better be called the period of “leveling” rather than “leveled” inflections, and further that it has no monopoly on even this term; that Old English is not in every respect the period of “full” inflections, certainly not as compared with its Teutonic and Indo-European antecedents; and that Modern English is far from being, in any absolute sense, the period of “lost” inflections.

Inflectional leveling began, of course, before the English language came into being. Its progress in the first stages of English of which a written record is preserved is well marked. Phonetic changes, on the whole with the tendency to obliterate the more elaborate inflectional distinctions of West Germanic, and the diversity of inflectional endings among the Old English dialects themselves—these were the great factors in bringing about a large measure of simplification in grammatical structure, even before what may be called “foreign” influence began to operate. Just what was the effect of contact with other languages may also be misconceived. As Emerson puts it,12

11 These terms were first applied by Sweet and have been commonly used by later historians of the language.

12 History of the English Language, p. 284. (Reprinted by permission of the Macmillan Co.)

110 The Development of Modern English

The effect of foreign tongues, especially of French, was in preventing the establishment of a standard speech, rather than in directly breaking down the inflectional system of the native tongue. For it was only necessary to remove the check of standard usage to bring about all, or most, of the changes taking place in the Middle English period.

We are not to suppose, then, that there was any conscious substitution of the French inflections for English ones or even that (to any great degree) the precedent of the simpler inflections of French was the cause of the simplification of English forms. It is rather that, when the temporary eclipse of English by French took place, and English was thereby reduced in importance and dignity so that it became for a time rather a popular dialect than a literary language, it necessarily changed with greater rapidity and thoroughness than formerly, simply because the conservative checks that usually operate to throttle popular tendencies in language were temporarily in abeyance. The dropping of inflectional distinctions, noticeable from the first in the Old English dialects and becoming more rapid in late Old English, was still more greatly accelerated in Middle English for the reason that has just been mentioned. In Modern English there is much less of it, partly, to be sure, because there is less opportunity for it, but also because of the far greater degree of standardization in the language as a whole. In the later aspects of Modern English, particularly since eighteenth-century ideas of “correctness” have had their weight, the simplification of inflection has been reduced almost to the vanishing point.

The parts of speech in which inflection occurs are now to be treated in order. The noun, with which we may begin, affords a particularly striking illustration of

English Inflections

111

progressive simplification. Just how many declensions ought to be indicated for the Old English noun is to

some extent a matter of opinion, since there is no unanimity among scholars as to the amount of variation neces-

sary to constitute a separate category. Usually, however, four vocalic and at least four consonantal declensions are

listed (the distinction between the two groups depending on whether the original stem ended in a vowel or a

consonant). Nevertheless, a classification like this tends to exaggerate the variety of declension in Old English,

since one of the vocalic declensions included most of the masculine and neuter nouns, and another most of the

feminine, and since the great majority of the nouns of the consonantal declensions were of a single type, the

n-stem. Practically, then, we may say that in the best-known form of Old English, Early West Saxon, there

are only three normal declensions to be distinguished—two varieties of the vocalic declension, the one for

masculine and neuter nouns, and the other for feminine; and one variety, the n-stems, of the consonantal

declension.

For the first group, masculine and neuter nouns with vocalic stems, these paradigms are typical (the second

and third columns illustrating the difference between neuter monosyllables with long and short stems):

Masculine

Neuter

Singular

Nom., Acc.

Gen.

Dat.

stān, stone

stānes

stāne

word, word

wordes

worde

hof, dwelling

hofes

hofe

Plural

Nom., Acc.

Gen.

Dat.

stānas

stāna

stānum

word

worda

wordum

hofu

hofa

hofum

112 The Development of Modern English

The other two groups—feminine nouns with vocalic stems, and nouns of the consonantal declension—may be represented thus:

Singular

Nom. lār, learning (fem.) guma, man (masc.)

Gen. lāre guman

Dat. lāre guman

Acc. lāre guman

Plural

Nom., Acc. lāra guman

Gen. lāra (-ena) gumena

Dat lārum gumum

One minor group must be added, particularly because of its survival in Modern English—the monosyllabic consonantal stems in which inflection may be represented by mutation of the root-vowel only. Such a noun is fōt (foot) (masc.):

Singular Plural

Nom., Acc. fōt fēt

Gen. fōtes fōta

Dat. fēt fōtum

In spite of the variety of inflection that has been indicated, it is evident that even in the early Old English noun, there is a fair degree of uniformity and simplicity. A few general principles may be useful: -um is the invariable ending of the dative plural, and, for the great majority of nouns, -a is the ending of the genitive plural; -es is the sign of the genitive singular in the very great majority of masculine and neuter nouns, as -e is of the dative singular in almost all nouns, including feminines. In general, the endings that are established as normal and typical at a very early stage are these:

Singular Plural

Nom., Acc. — —as

Gen. —es —a

Dat. —e —um

English Inflections

113

It will have been observed that gender in the Old English noun differs from that in Modern English in being grammatical rather than natural or logical: thus, the words for stone and foot were masculine, while that for learning was feminine. Part of the explanation of the substitution of the new system for the old lies, of course, in the dropping or leveling of inflectional endings; the change of the genitive of lār, for example, from lāre or lore to lores, meant the loss of a form that could be recognized as distinctly feminine. The replacing of grammatical by logical gender first began long before the end of the Old English period,13 and by the time that the simplification of noun declension that we think of as typical of Middle English had been accomplished, it was already a completed process.

The typical Middle English declension for the noun may be summarized thus:

Singular Plural

Nom., Acc. — —es

Gen. —es —es

Dat. —(e) —es

Of the six distinctive forms for the typical Old English noun, that is to say, two—the genitive and dative plural —had been lost altogether, and a third—the dative singular—was going. The last mentioned case ending, the -e of the dative singular, is preserved occassionally in Chaucer, after it has presumably been more nearly lost in spoken English, for several reasons: the language of poetry is usually more conservative and hence more archaic; the vowel is useful for its sound value and for rhyme or metre (as in to the rote, rhyming with shoures

13 Cf. p. 52.

114 The Development of Modern English

sote14); also, it tends to survive, as a “petrified” form, in certain familiar formulas, like to bedde, in londe, on fyre. The corresponding ending, it is interesting to observe, is likewise in a disappearing state in the German of our own day. It might be added that the originally distinctive ending of the feminine genitive, i.e., -e instead of -es, also occasionally appears in Chaucer, though to be sure much less frequently; thus, the title of one of the Canterbury Tales is Nonne Preestes Tale (Nun's Priest's Tale) rather than Nonnes Preestes Tale, and we sometimes meet a phrase like in his lady15 grace (in the favor of his lady), in which the feminine genitive is partly preserved, though the vowel has actually been dropped.

There is only a short step from the typical declension of the Middle English to that of the Modern English noun. The latter, of course, is as follows:

Singular Plural

Nom., Acc Dat. — —s

Gen. —’s —s’

The apostrophe in the genitive singular represents the elision of the vowel, which had actually taken place in pronunciation much earlier. But it was still possible in early Modern English for poetical language to use the vowel; thus, Shakespeare’s moonés sphere16 is the same kind of genitive that we find almost regularly in Chaucer

14 Prolog to the Canterbury Tales, ll. 1–2.

15 This particular feminine genitive is preserved even in Modern English in the name of one of the days on the calendar of the English Church: Lady Day (i.e., “Day of Our Lady”).

16 A Midsummer Night's Dream, Act II, scene 1, l. 7.

and occasionally in the popular ballads.17 Likewise,

the plural -s is a syncopation of the Middle English -es,

which in turn represents the “weakening” of the Old

English -as. Modern English has improved on Middle

English by differentiating (through the -s, -’s, and -s’)

forms which had been distinct in Old English, but which

phonetic changes had made indistinguishable in Middle

English. The disadvantage, of course, is that the

modern printer’s device of the apostrophe is useful chiefly

on the printed page; in the spoken language, the leveling

of case endings has been attended by possible ambiguity,

in Modern English as in Middle English.

So far we have been concerned with the normal or

typical declension of the noun. It is of course evident

that not all Middle and Modern English nouns are

declined on this model. The exceptions are usually

to be explained either as survivals of some other Old

English declension than that of the masculine and neuter

a-stems, which tended to absorb the other types, or,

less frequently, as analogical formations. It will be

simplest to approach this question from the angle of

Modern English.

One noteworthy group of survivors or another Old

English declension comprises the mutation or umlaut

plurals. The Modern English plurals that represent

this declension, uncomplicated by any other type of

plural formation, are these: feet, teeth, men, women,

geese, mice, and lice. In Old English, of course, the

group was considerably larger; it included, for example,

such words as bōc (book), āc (oak), and gāt (goat), the

plurals of which have later been regularized, so that

17 Cf.: “the warldis room,” Edward, 1. 45.

116 The Development of Modern English

we now speak of several books, oaks, or goats, rather than

of several beek, eak, or geat (forms that might well have

developed). But how does it happen that the seven

plurals mentioned above have resisted the pull of analogy

and remained in Modern English as a small but useful

body of recalcitrants? There are two partial explana-

tions. In the first place, it is evident that the more

familiar the form, the less likely it is to change; it is

much easier to imagine the irregular plural of a rarely

used word conforming by analogy to the usual pattern

than it is to imagine men being supplanted by mans

or feet by foots. Then, again, euphony has doubtless had

some influence: gooses, mouses, louses, and tooths would

be both longer and uglier than geese, mice, lice, and teeth.

A second group of irregular plurals in Modern English

is made up of the nouns that have the same form in the

plural as in the singular. These unchanged plurals are

either the direct descendants of the Old English neuter

monosyllables with long stems, or analogical forms that

follow the model. Of the first type are swine, sheep, deer,

folk; in these, the vowel of the Old English form was

long, or, if short, was followed by two consonants, making

the stem long, so that a -u was not added to form the

plural. Hors survived in Middle English as an

unchanged plural, as we see in Chaucer's line:

His hors weren gode, but he nas nat gay

(Prolog to Canterbury Tales, 1. 74.)

but it has since conformed, by analogy, to the regular

type of plural. On the other hand, Old English fisc

was a masculine a-stem with plural fiscas, which would

regularly give fishes as the Modern English form;

English Inflections

117

but it has another and commoner plural form, fish,18

in Modern English, apparently because of the analogy of

such other names of animals as those just mentioned.

The unchanged form fish having thus been established

as the usual plural in Modern English, names of various

species of fish—salmon, trout, bass, and so on—make their

plurals accordingly.

A few other unchanged plurals in Modern English may

be mentioned, though they are of different origins.

Pound comes from an Old English long-stemmed mono-

syllable, pund, and the original unchanged plural

survives in such phrases as a five-pound note and The

Twelve-Pound Look. Night does not belong to this

declension historically, being an umlaut plural, nor does

month; but both occasionally use the same form for

plural as for singular—as in fortnight (for fourteen nights)

and twelvemonth—in addition to their ordinary employ-

ment of the -s plural. In vulgar speech, this category

of unchanged plurals for nouns of enumeration or

measurement is more widely extended: “twenty foot

high” and “seven year19 old” are expressions widely

current.

The n-stems, or weak declension, furnish a third group

of exceptional plurals in Modern English. Only a single

word, however—oxen, from Old English oxan—now

18 Sometimes present-day (especially British) usage employs two

plural forms for other names of animals. The distinction is thus

explained by Miss Rose Macaulay: “Animals seen as sport become to

the mind meat, and cease to be reasonable creatures, so that you may

feed fishes, but catch fish, ride elephants but hunt elephant, . . . throw

bread to ducks, but shoot duck, admire moths, but seek to exterminate

moth. . . . ” Staying with Relations, New York (Liveright), 1930.

19 This is the historically correct form, since Old English gēar was

a long-stemmed neuter monosyllable.

118 The Development of Modern English

represents this declension in its purity. Even words in everyday use-those, for example, for sun, moon, eye, and ear-have had their plurals regularized by analogy.20

Here, incidentally, is one of the best evidences of the strength of analogy, that, of the many numerous and important weak nouns in Old English, one alone should survive into Modern English.

In Middle English there were considerably more survivors: toon (toes) (<O. E. tān, tāan), been (bees) (<O. E. bēon), and treen (trees) (<O. E. trēo, trēowu, later made a weak plural by analogy) are Chaucerian forms.

Milton’s shoon, in

The swain

Treads on it daily with his clouted shoon (Comus, l. 635.)

is, of course, a belated occurrence of a weak plural, possible only in the more archaic language of poetry.

Other exceptional plurals in Modern English are explained as combinations of two different types of Old English plural formations.

Here, for example, belong brethren, which joins a weak ending with an umlaut form, the Old English brēðer; and also children, in which the -r ending (cf. Ger. Kinder) was the older sign of the plural, the -en having been added after the other had lost its plural significance.

Another interesting double plural is the poetic form kine.

Here the development that might have been expected is singular cow (<O. E.

20 It is significant of the more conservative development of German that, in a number of instances, the Old English and the Modern German cognates are both of the weak declension, while the Modern English forms have gone over to the typical declension.

Note O. E. ēagan, Ger. Augen, Mn. E. eyes; O. E. cnapen, Ger. Knaben, Mn. E. knaves; O. E. tungan, Ger. Zungen, Mn. E. tongues; O. E. flēogan, Ger. Fliegen, Mn. E. flies; and O. E. naman, Ger. Namen, Mn. E. names.

cū), plural kye (< O. E. cȳ, an umlaut form). In ordinary use, the plural, of course, is the regularized form cows; but kye is preserved in the Scots dialect, and kine, in which a weak ending has been attached to the mutated plural, is a familiar archaic form.

Apart from all the groups of plurals that have so far been mentioned stand the foreign plurals that are occasionally met with in Modern English. With scarcely any exceptions, nouns entering the language before the Modern English period have conformed to the inflection of native words. In Modern English, however, there are many instances of borrowed words retaining their foreign plurals. The present tendency, which the writer believes should be encouraged, is to regularize a number of these forms. Where there is a choice, that is to say, between the native plural ending in -s and a foreign plural, the better principle is to use the former. It must be admitted that the choice does not always exist; thus, axes, bases, crises, and theses (with the second vowel pronounced distinctly as [i:]) are certainly to be preferred to the awkward succession of sibilants that would be present in their regularized forms. The pronunciation indicated is of course particularly necessary for the first two forms in order to make a distinction between the plural of basis and that of base, and between the plural of axis and that of ax. But surely, when two plurals, one native and one foreign, are competing in actual use, it is desirable to use the native. Plurals like banditti, monsignori, prime donne, stadia, gymnasia, and even campi have a certain (usually academic) vogue, but they have also an unpleasant air of something akin to pedantry or affectation. The story is told of a certain celebrated English lady airily remarking, "I don't understand how anyone

120 The Development of Modern English

can ride in those odious omnibi21”—to be met by the

retort, “Ah, there you propound one of the most difficult

of all conundra!”

Sometimes the problem of rival plural forms is resolved

by retaining both, but for somewhat different uses:

the native for general and the foreign for specialized,

frequently scientific, purposes. Thus the mathe-

matician or the chemist may prefer formulæ, but the

man in the street is justified in using formulas; and so it is

with foci and focuses, gladioli and gladioluses. The last

instance is further complicated by varying accentuation;

it seems natural to put the stress, in the foreign plural,

on the second syllable, and in the native plural on the

third. However, the pronunciation last mentioned is

often used with uneasy misgivings as to its correctness;

hence some who have occasion to use the word dodge the

issue altogether and say glads—just as others avoid the

difficulties involved in the plural forms of hippopotamus

and rhinoceros by using the shortenings hippos and

rhinos.

One argument for preferring the native plural where

there is any sanction for it in reputable use is that, after

all, only a small proportion of the English-speaking

world is sufficiently acquainted with Latin, Greek, or

Italian inflection to employ the foreign forms correctly.

Observe, for example, the confusion between alumni

and alumnæ, a confusion which is intensified by the

fact that the preferred English pronunciation of these

21 What makes omnibi particularly grotesque is of course the fact that

the -us with which the singular of the English word ends is not, in the

Latin word from which it is taken, a singular ending at all; the -i plural

is as thoroughly false as it is in ignorami, which the writer has actually

seen in journalistic use.

English Inflections

121

words exactly reverses the sound of the Latin endings. More commonly, however, the difficulty is as to number rather than gender. Thus, it is easily forgotten that phenomena is a plural; it tends then to be used as a singular, and a new English plural, phenomenas, is evolved. And so it is with insignia, data, and perhaps memoranda. It is to be remembered too that even with our native words it is possible for a form etymologically a plural to become a singular in use; wages was once so treated,22 and news is so treated now (present usage exactly reversing that common in Elizabethan English). Foreign plurals have sometimes gone the same way, the meaning of the word changing in the process; thus opera and stamina are both, etymologically, Latin plurals. A recent Latin borrowing, agenda, is on the way to complete Anglicization in an interesting fashion: first the pronunciation with the “hard” g yielded to that with the “soft,” and now the tendency seems to be to treat the word as a singular rather than a plural.23

Such a discussion as the foregoing tends, perhaps, to suggest that the inflection of the noun in Modern English is more complicated than is actually the case. In spite of the attention that has been given to departures, in both native and borrowed words, from the simple pattern of the typical inflection, it remains true that the Modern English noun is inflected slightly and, on the

22 As in “The wages of sin is death.”

23 “Agenda Is Determined on Calendar Reform” (headline in New York Times, June 17, 1931). As for phenomena, the sentence: “. . . Goethe had taken an intelligent interest in all this phenomena of history” appeared in the Times “Book Review” for February 14, 1932; and: “Russia seems to be getting along very well without this phenomena we have been taught we could not do without,” in the Times for November 27, 1933.

122 The Development of Modern English

whole, regularly. This is the case both comparatively

(with other languages and with earlier stages of English)

and absolutely. A similar observation could not be

made about the pronoun. The pronoun, on the contrary,

is the one part of speech that retains in Modern English a

degree of inflection comparable to that which it had in

Old English. The personal pronoun is inflected, in

Modern English as in Old English, for three cases, three

persons, and (for the third person singular) three

genders; in addition, two numbers are distinguished in the

Modern English forms, and (for the first and second

persons) three in the Old English. The evolution of the

first, second, and third personal pronouns will now be

given in some detail.

First Person

Singular Dual Plural

Nom. ic wit wē

Gen. min uncer ūre

Dat., Acc. mē unc ūs

These are the forms commonly used in late Old English

and in early Middle English. In early Old English

separate accusative forms were employed (mec for the

singular, uncit for the dual, and ūsic for the plural),

but eventually—and well before the end of the Old

English period—these forms had been completely super-

seded by the datives, which thus did double service.24

Our modern forms (except of course those for the dual,

all trace of which is lost after the thirteenth century)

have developed, with the regular sound changes, directly

from the ones just listed, except that the consonant in ic

24 In the dual, however, the two forms unc and uncit (the latter a form

of more recent origin) were often used side by side, for both dative and accusative.

English Inflections

123

has been lost and the vowel lengthened, and that the

vowel in ūs has been shortened.

Second Person

Singular

Dual

Plural

Nom.

ðū

git

gē

Gen.

ðīn

incer

ēower

Dat., Acc.

ðē

inc

ēow

These, again, are the typical Old and early Middle

English forms, the distinctive accusative forms—ðec,

incit, and ūsic—having been absorbed by the datives.25

Once more the modern forms—thou, thine, thee, ye, your,

and you—have developed regularly from their Old

English prototypes (though we might have expected

your and you to be pronounced with the diphthong of

our). It is evident, however, that modern usage has

treated the second personal pronouns differently from

those of the first person. A single form, you, does

service, for the most part, in Modern English for both

nominative and accusative, in both singular and plural

numbers. This requires further explanation.

The custom of addressing kings and other great per-

sonages by the plural rather than the singular second

personal pronoun began in the Latin of the Empire and

was spread, through a similar usage in French, through-

out Europe; it exactly corresponds, of course, to the

present royal, and editorial, use of the plural of the first

personal pronoun. In English, the polite use of ye and

you for the singular became common by the end of the

fourteenth century; at the same time thou and thee began

to be relegated to familiar or affectionate use. The

proverb that familiarity breeds contempt is illustrated in

25 Except that incit and inc were frequently used indiscriminately for

both cases, as noted above for uncit and unc.

124 The Development of Modern English

their later history; for a misplaced thou—one used where

the occasion demanded the formal ye—could only be

interpreted as patronizing or derogatory. Thus in early

Modern English thou has much the force of the German

du: it may signify affection, intimacy, or scorn. However,

the more conservative language of poetry and religion

preserved the older thou in certain special categories.

It still persists in the language of prayer and it is kept

likewise in the speech of the Quakers. The latter use,

however, is somewhat different: the form in vogue among

the Quakers is not thou but thee, just as the standard

speech has substituted the formerly accusative you for

the nominative ye; it is joined, moreover, with the verb

not of the second person but of the third (thee goes, for

example, rather than thou goest), perhaps through the

influence of the similar sound of he or she goes.26

The other point in the simplification of the modern

second personal pronoun is that the former distinction

for case, as between nominative and accusative, has also

been lost. In the early seventeenth century, the older

use of ye for nominative and you for accusative is being

replaced by the indiscriminate use of the single form

you. The King James Bible of 1611 preserves the dis-

tinction (as in “Ye in me, and I in you”); but Shake-

speare, less bound by tradition, not only frequently

uses you for nominative but also occasionally uses ye for

accusative. Milton, too, frequently uses the two forms

regardless of their historical distinction: “ . . . you

must then first become that which ye cannot be . . .

And who shall then stick closest to ye . . . ?” (Areo-

pagitica). Modern poetry sometimes retains the ye,

26 Emerson, History of English Language, p. 323.

just as it does the thou and thee ("And O, ye Fountains,

Meadows, Hills," Wordsworth: Immortality Ode); but

save for such poetical, and occasionally also dialectal,

survivals, the invariable form for the second personal

pronoun is you.

Such simplification has not been unattended by

ambiguity. Obviously useful distinctions are made

possible by preserving four forms in the first personal

pronoun—I, me, we, us; and it is clear that the one form

that has to do service for all four categories in the parallel

uses of the second personal pronoun is somewhat over-

worked. The lack of a distinction for case seems less

vital than the lack of distinction for number. Hence

it is that a number of creations—the Southern you all,27

the dialectal you-uns, and the vulgar youse—have been

evolved to make good the deficiency.

The third personal pronoun differs from the other two

in being declined for gender (in the singular), and in

preserving, throughout the Old English period, separate

forms for the dative and accusative singular of all three

genders.

Third Person

SINGULAR

Masculine Feminine Neuter

Nom. hē hēo hit

Gen. his hiere his

Dat. him hiere him

Acc. hine hīe hit

27 Though the point is disputed, the best opinion seems to be that

Southern usage confines you all to the plural, or, more accurately, that

even if it is addressed to one person it still has a plural implication;

that is, that it stands for "you and your family," "you and your friends,"

and so forth, as fits the circumstances.

126 The Development of Modern English

PLURAL

Nom., Acc. hīe, hī

Gen. hiera

Dat. him

These were the forms most frequently used in Old English. The Modern English equivalents are so different that some details about their evolution are necessary.

The third person singular masculine develops regularly into the Modern English paradigm, except that, as with the first and second personal pronouns, the distinctive accusative form is lost and him does service as both dative and accusative. In the feminine, however, a new form developed for the nominative,28 since, if hēo had remained, its later form and sound would have been identical with those of the masculine hē; here too, the original dative form, hīere, became also accusative in use, even though it was identical with the genitive, as modern her still is.

In the neuter, the form hit was nominative as well as accusative, and this may account for its survival (with loss of initial h) in Modern English as the accusative, especially since him had become the masculine accusative as well as the dative, and the use of hit as dative-accusative for the neuter thus brought about a useful differentiation. Nevertheless, what one finds to be most striking, in comparing the modern forms with the old ones, is that in a single instance, the possessive of the neuter, an additional and a totally new form has been evolved—an outstanding exception in the general trend of inflectional leveling. This is its, formed by adding the usual genitive ending to the nominative hit or it; it supersedes the his that once did service as the neuter

28 Perhaps from the feminine of the definite article, sēo, with later palatalization of the initial consonant.

English Inflections

127

as well as the masculine genitive. The substitution of

its for the neuter his comes about in the early seventeenth

century. The English of the King James Bible, again, is

more conservative than the general practice of its day, for

its is never admitted. The old form his is often employed

("If the salt hath lost his savor"), but that it is also felt

to be awkward is shown by the frequent use of para-

phrases such as "of it" or "thereof" ("The earth is the

Lord's and the fulness thereof") where, to us, its would

be the natural word.

In the plural, the forms of Old English have been con-

siderably altered. The general reason for this is evident:

if hīe or hī had been allowed to remain, it would have

been too nearly like hē and hēo, while hiera would have

become identical with hiere,29 of the feminine singular.

What has happened, then, is that the Scandinavian

forms they, their, and them, entering first in the Northern

dialect of Middle English, eventually took the place of

the original hīe, hiera, him. The nominative they was

employed earliest, this being the regular form in Chaucer,

alongside hem as dative and accusative and such forms

as hire and hiere as genitive. In late Middle English

and early Modern English, however, their and them

established themselves firmly.

In listing the personal pronouns, the origin of the

possessive pronouns has also been indicated. Save for

its, the Modern English possessive pronouns developed

from and were once identical with the genitive cases of

the Old English personal pronouns. These forms early

29 That is, both forms would have become hire, here, or hir. This

confusion does, in fact, exist in Chaucer. Compare the similarly ambigu-

ous ihr of Modern German.

128 The Development of Modern English

in their evolution lost almost entirely their genitival

function, and came to be inflected and used like adjec-

tives. The genitives of the first and second personal

pronouns were declined as strong adjectives in Old

English, while the forms his and hiere, indeclinable in

Old English, came to be inflected like adjectives in

Middle English. In Middle English times, too, are to

be observed the beginnings of the modern differentiation

between the modifying and the absolute forms of these

pronouns. The final consonant of mīn and thīn was

dropped before a word beginning with a consonant, and

the forms my and thy thus created; eventually my and

thy were used even before a word beginning with a vowel,

but in the absolute position the old forms mine and

thine are still retained. In the more conservative lan-

guage of poetry, however, the older forms are permissible

in Modern English even in the modifying position,

when used before vowels ("Mine eyes have seen the glory

of the coming of the Lord"). In addition to the pairs

my, mine and thy, thine, the differentiation for these two

uses was accomplished in Middle English for all the

other pronouns except his, which has but one form in

both the modifying and the absolute positions (though

the vulgarism hisn is an understandable analogical

formation, which had indeed a certain literary use in

Middle English, as had also hern, theirn, and yourn).

The final s of his was apparently the pattern for the

Middle English forms in -s for absolute use, which have

become standard: ours, yours, hers, and theirs.

Aside from the personal pronouns, it is the interroga-

tives that have preserved more of the original inflection

than any other group of pronouns. The Old English

forms were these (alike for both numbers):

English Inflections

129

Masc. and Fem. Neuter

Nom. hwā hwæt

Gen. hwæs

Dat hwām

Acc. hwone hwæt

Inst. — hwȳ

Except for the hwone, all of these forms are represented

in Modern English, in the following equivalents: who,

what, whose, whom, and why. The last named is the sole

remnant of the instrumental case in Modern English,

save only for the instrumental the (indistinguishable,

however, in form from other cases of the now indeclinable

the) in nevertheless and in such phrases as the sooner, the

better. But it is evident that usage in present-day

English treats these forms somewhat differently from

what the paradigm might seem to imply: whose is now but

sparingly applied to things, almost always to persons;

and whom in popular usage tends to give way to who.30

Other pronoun groups that were fully inflected in

Old English have for the most part lost this inflection

in Modern English. Such are the simple demonstrative

sē, sēo, þæt, which served also as the definite article, and

the compound demonstrative þēs, þēos, þis, formed by

adding sē to the simple demonstrative to make the

expression more emphatic, just as vulgar speech now

creates such combinations as that there or this here now.

Both demonstratives were fully inflected in Old English,

for three genders and five cases in the singular, and for

30 Especially when the form which is theoretically objective precedes

the verb. Expressions like “Who did you see?” and “Who did you call

on?” are widely current and have won a certain partial sanction. Cf.

Sapir’s opinion of the general “drift” of the language in this respect

(Language, pp. 166–174); his conclusion about whom is this: “It is safe to

prophesy that within a couple of hundred years from today, not even the

most learned jurist will be saying ‘Whom did you see?’” See also

p. 500.

130 The Development of Modern English

one gender and three cases in the plural. Modern English has greatly simplified these forms. The distinction between sē and sēo was lost in Middle English,

the forms of the article becoming thē and thēo, and the latter eventually being lost as grammatical gender fell into disuse. The originally neuter þæt (that) was

retained only as a demonstrative, without distinction for gender. All forms for the definite article were leveled to the. The old plural form þā (M. E. thō) was given an

analogical -s ending in its demonstrative use. This took place, apparently, after that ending had established itself as the normal plural termination of nouns; perhaps

the plural form of the compound demonstrative (O. E. þās) was an additional influence. The plural for that thus came to be those, so that a new plural for this

(which, like that, was at first neuter only) had to be evolved; this form was Middle English thēs, Modern English these. Modern English thus retains of these

elaborate inflections only the following five forms: the, now the indeclinable definite article; that and this, formerly the neuters of the simple and the compound

demonstratives, respectively, but now otherwise differentiated; and the corresponding plurals, those and these, the origin of which has just been indicated. No variation of form is now possible for the article, while the

demonstratives, having lost inflection for case and gender, retain it for number only.

The remaining groups of pronouns show almost no inflectional variation beyond what has already been mentioned. The reflexives, for example, have no separate forms, as they have in more highly inflected languages,

but are made simply by attaching -self or -selves to the corresponding personal pronouns. In Old English the

English Inflections

131

personal pronouns served, without change, as reflexives also, and this custom persisted until early Modern English times; the phrase "Now I lay me down to sleep" illustrates this use. But Old English also sometimes employed the double forms, with -self added, for greater emphasis; and these are the regular reflexive pronouns of Modern English. The indefinite pronouns, being derived from and inflected like other pronouns or adjectives or nouns, likewise possess no separate inflectional forms. One group of indefinites was, in Old English, identical in form with the interrogatives,31 and these are of additional interest as the source of the modern relative pronouns.

Old English possessed no distinctive relative pronoun. The relative function was variously performed (1) by the demonstrative se, sēo, þæt, (2) by the indeclinable relative particle þe, (3) by þe joined with the demonstrative, or (4) by þe joined with the personal pronoun. The sole remnant of all these forms is that, originally the neuter of the demonstrative, but widely used as the relative pronoun for all genders in Middle English, and the only relative in colloquial use today. The other relatives of Modern English—who and which—developed their relative use much later, though both are from Old English forms, the interrogative-indefinites hwā and hwilc, respectively. Which was the earlier to develop a relative function; it became, in early Middle English use, a general and indeclinable relative form, like that. A familiar instance of its earlier application to persons is the opening phrase of the Lord's Prayer, "Our Father,

31 The Modern English indefinite use of who and what is illustrated in Shakespeare's "Who steals my purse steals trash," and such familiar phrases as "What you say is true" (Emerson, p. 340).

132 The Development of Modern English

which art in Heaven." The present use of who as a

relative and the distinction between who for persons and

which for animals and for things were not fully worked

out until the eighteenth century.32 Even yet, as has

been suggested, who and which are used as relatives

less freely than as interrogatives; the man in the street

unconsciously prefers that to who, quite conceivably

because the problem of case distinctions does not arise

in connection with the former.

Much less space need be devoted to the adjective than

to the pronoun, for adjectives in Modern English have

completely lost inflection (unless we consider comparison

as a type of inflection). The Old English adjective,

however, was declined for number, gender, and case,

and in addition, as we have seen, it had two sets of

inflectional endings—one set for when it was used in

the strong position, and another for the weak position.

These forms, ten in number,33 were already leveled in

early Middle English to just two—a form without ending,

and one with final -e. In Chaucer, the only remnant of

the elaborate distinctions of Old English is that a form

without ending is usually found in the singular of the

strong declension; in all other positions, the adjective ends

in -e. With the general loss of final -e in late Middle

English, the last vestige of adjective declension was

eliminated. The only exception in Modern English is

32 The classical illustration is "The Humble Petition of Who and

Which," in the Spectator for May 30, 1711. As Jespersen points out

(Essentials of English Grammar, p. 359), Addison turns historical truth

topsy-turvy by describing the relative that as an upstart that has recently

done injury to who and which.

33 There were of course many duplications, so that the number of

inflectional distinctions is far greater than this.

English Inflections

133

the occasional use of a borrowed inflection, as in the

phrase Knights Templars (from the French).

Closely allied to true inflection, however, is the process

of comparison. Modern English utilizes two methods

of comparison: the addition of the suffixes -er and -est

to form the comparative and superlative degrees, respec-

tively; and the use of the auxiliary adverbs more and

most. The latter method of comparison is comparatively

modern, becoming common only in late Middle English.

The use of the suffixes, however, goes back to primitive

Teutonic. In Old English, most adjectives made their

comparatives in -ra and their superlatives in -ost; from

these endings come the Modern English suffixes -er

and -est. There were, in addition, a smaller number of

adjectives with mutated comparative and superlative

forms (the latter usually ending in -est), and a few adjec-

tives in which mutated comparatives and superlatives

are of a different root from the positive form of parallel

meaning, the original root having been displaced. Of

the former type are such forms as strengra and strengest

as comparative and superlative of strang (strong),

lengra and lengest of lang (long), and ieldra and ieldest

of eald (old). The only survivals in Modern English

are elder and eldest, and even these have been superseded,

for most purposes, by the regularized forms older and

oldest. There are, however, several survivals of the other

type (that which utilizes two separate roots in the process

of comparison); good, better, best and little, less, least are

the clearest illustrations.

Certain Modern English adjectives show in their

comparative and superlative forms a "pyramiding"34

34 The term is Mrs. Aiken's. Cf. English, Present and Past, p. 226.

134 The Development of Modern English

of inflections. There are, for one thing, double forms like inmost, outmost, foremost, and hindmost, in which

an original superlative in -m35 has been reinforced by a second superlative suffix -est. In these words the double

ending -mest has been altered through association with the auxiliary adverb most, since this word was used in

the analytical superlative forms. Parallel to these are double comparatives like nearer, lesser, and worser.

The Old English comparison for the first word was neah, nearra, niehst36—in the modern equivalents, nigh,

near, next. Near (or ner or nere) remained as the comparative in Middle English,37 but was eventually used,

perhaps under the influence of an adverbial near, as the positive. A new comparative, nearer, and a new superla-

tive, nearest, were then built up on the assumption of a positive near. The most extreme example, probably,

of the accumulation of comparative and superlative endings in a single word is nethermost. Though the point

is disputed, the word may possibly be analyzed as fol-

lows: ne “down” + -th (comp.) + -er (comp.) + -m (sup.) + -ost (originally, sup. -est). Only one degree

removed from this quadruply inflected word are such triple forms as innermost, uppermost, furthermost, and

outermost.

As we have seen, inflection is possible in the Modern English adjective only in the limited sense of comparison. Save for this sense, the adverb is uninflected in

35 Cf. Latin optimus and primus, the latter cognate with Old English forma (first).

36 Cf. German nah, naher, nächst.

37 This is regularly so in Chaucer. Note: “And as he dorste, he drough him ner and ner” (“Prioress’ Tale,” l. 1710), which Wordsworth absurdly translates into the meaningless “And as he durst he drew him near and near.”

both Modern English and Old English. Nevertheless,

inflection was used in the formation of the Old English

adverb. The regular method was to add the instru-

mental ending -e of the adjective to form the correspond-

ing adverb; for example, georn (eager), georne (eagerly);

wīd (wide), wīde (widely). Adjectives already ending

in -e became adverbs without change of form; thus

blīðe meant either “joyful” or “joyfully,” and clēne,

either “clean” or “cleanly.” In addition, a second

adverbial termination, -līce, was evolved through the

circumstance that when certain adjectives ending in

-līc were made into adverbs [earmlīce from earmlīc

(wretched) and frēondlīce from frēondlīc (friendly)],

the ending -līce, rather than -e, came eventually to be

felt as the adverbial suffix. It was then added, instead

of -e, to form other adverbs; thus, we have eornost

(earnest) and eornostlīce (earnestly) (rather than eornoste).

The -līce termination, and its modern descendant -ly,

gradually became established as the normal or typical

adverbial ending. This suffix has been added by analogy

to many words, native as well as foreign, to which it did

not originally belong. There is a special reason for this

in the phonetic change that the old adverbs in -e have

regularly undergone.

In the Middle English period, inflectional final -e was

regularly dropped. This meant, of course, that adverbial

forms which employed this suffix could no longer be

differentiated from the corresponding adjectives: Old

English heard and hearde became alike hard; fæst and

fæste became alike fast. The modern adverbs hard and

fast are therefore not properly to be described as adjec-

tives used adverbially; they are true adverbs, though

identical in form with adjectives. Only a few of them,

136 The Development of Modern English

such as hard, fast, first, and wide, remain in both common and "correct" use. The great majority of the adverbs which once possessed the -e suffix and later lost it have conformed by analogy to the -ly pattern. Even in the cases mentioned, hardly, as well as hard, and widely, as well as wide, are possible as adverbs (though with a differentiation in meaning); and firstly, by analogy with the other ordinals secondly, thirdly, and so forth, is knocking lustily on the doors for admission.

It is true that, in popular, and sometimes in archaic, literary use, other adverbs without ending (representing a lost -e suffix) are to be found. But most of these are definitely incorrect according to standard usage. Thus, it does not seem accurate to put slow and fast, as adverbs, on the same basis.38 Fast is the only possible adverbial form for present use, but slowly has, in precise use, long since superseded the adverbial slow; this has happened in spite of the fact that slow was once correct, as in Milton's line

Swinging slow with solemn roar,

and that there is considerable use today of the same form, perhaps because "Go slow" is easily made parallel in form to the phrase of opposite meaning, "Go fast." It might be added, too, that frequently a stereotyped formula will preserve the older form of the adverb for that particular collocation; thus, "to laugh loudly" is correct, but also "to laugh loud and long."

38 As, for example, Professor Krapp apparently does when he remarks, "This type of adverb formation persists in Modern English in adverbs without ending, as in the constructions 'Go slow'; 'He fought hard.'" (Modern English, p. 72.) For further discussion of slow and slowly, see p 523.

English Inflections

137

Further, it is to be pointed out that the difficult task of discriminating between adjectives and adverbs39 is made harder not only by certain ambiguous forms—for example, well and ill—but by the occasional occurrence of the adjective in what is usually the adverbial position, just after the predicate or just after the object. Thus, to use the adjective of resultant state is the more accurate in the phrase “to roll the cigaret tight”; and the same reasoning makes “the sun shines bright” at least as logical as “the sun shines brightly.” To many, it is difficult to see that the verbs of sensation have a copulative force exactly parallel to that of be, seem, and appear, which require a predicative adjective rather than an adverbial modifier. This, of course, is the explanation of “the rose smells sweet,” “the milk tastes sour,” “the ground feels rough,” and the like. Even here, however, there is sometimes another possibility:40 one can justify both “it smells bad” (i.e., “appears, when smelled, to be bad”) and “it smells badly” (i.e., “abominably,” “atrociously”). In a similar way, with a slight difference in emphasis, both “he stands conspicuous” and “he stands conspicuously” are possible.

To return to the use of inflection in the original formation of adverbs. After the instrumental -e, the most common inflectional ending used adverbially in Old English was the usual genitive ending of nouns and adjectives.

39 This has been proposed as the supreme test of the educated person: “He may be a Ph. D. from any place you like; but if he confuses adverbs and adjectives, he is not an educated man.” (Katharine F. Gerould, “What Constitutes an Educated Person Today,” Atlantic Monthly, Vol. 127, p. 745.) Truly a hard saying, if strictly interpreted.

40 For further discussion, see p. 522.

tives, -es. Some of these formations that have survived in

Modern English are nēades (needs) (as in "He needs must

go"; i.e., "He must of necessity go"); nihtes (nights)41;

ānes (once) (where changes in spelling and pronunciation

have obscured the origin, as also in twice and thrice);

sūðeweardes (southwards); and hāmweardes (homewards).

The adverbial genitive his weges (on his way) is preserved

in the Shakespearean "Go your ways" and in the present-

day colloquialism "He went a long ways." Modern

English, in fact, makes frequent use both of the genitive

ending used adverbially—for example, sideways, back-

wards, always, sometimes—and of the corresponding

prepositional phrase—as in of course, of old, of late, of a

truth, and so forth.

One other case ending used adverbially may be men-

tioned, although there are very few survivals of it in

Modern English. This is the dative-instrumental -um,

as in Old English miclum (greatly), wundrum (wonder-

fully), stundum (from time to time), and so on. The

form whīlom, now archaic, comes from hwīlum, originally

meaning "sometimes," "at times," but later "in former

times." Seldom (O. E. seldum) is an analogical creation

in which an earlier form seldan (cognate with German

selten) has been altered to seldum through the influence of

other adverbs with this ending. The fuller suffix

-mǣlum [< mǣl (time), (measure)] was used in a number

of adverbial forms; for example, styccemǣlum (here and

there), and dropmǣlum (drop by drop). This is pre-

served in Chaucer's flokmele (in a flock) and stoundemele

41 In "He goes out nights," the final -s is not plural but genitive singu-

lar in adverbial use; the expression is thus exactly equivalent to the old-

fashioned "of an evening," in which the prepositional phrase replaces

the genitive ending.

English Inflections

139

(hour by hour), and in a single Modern English example,

piecemeal.

Distinguishing features of the inflection of the English

verb have already been mentioned, in the discussion of

the Teutonic verbal system. These may be briefly

recapitulated on the basis of the Old English forms:

two completely conjugated tenses, present and preterit;

two moods, indicative and subjunctive, for these two

tenses, and an imperative for the present only; one voice,

the active; three persons; and two numbers. In addition,

there are the verbals: the infinitive and the gerund, the

latter historically the dative case of the infinitive,

used as a neuter noun, after to;42 and the two participles,

present and past. The general trend of the develop-

ment of the verb in English has been toward simplicity

and regularity. It must be noted, however, that while

inflectional endings have been more and more leveled,

new analytical formations have constantly been replacing

the lost synthetic forms. The two most striking illustra-

tions of this drift, perhaps, are the histories of the future

tense and the subjunctive mood. Early Old English

had no future tense; the original synthetic future of

Teutonic had been lost, and the analytical future of

Modern English was only beginning to be built up, by

the use of the auxiliaries willan (wish, will) and, occa-

sionally, sculan ("must," later "shall"). Meantime,

the present did service for the future also, as it still does

in Modern English in such a phrase as "I leave tomor-

42 The inflected form very early began to drive out the simple or

uninflected form of the infinitive. This accounts for the preservation

of the to as the "sign of the infinitive," after the distinction in form

between infinitive and gerund had been lost in Middle English. The

modern gerund is of course an entirely different thing; it is identical in

form with the present participle but different in use.

140 The Development of Modern English

row" or "The boat sails at midnight." The present

elaborate distinctions between "shall" and "will"

(whether we regard them with admiration or with dis-

may) must surely be considered to illustrate simplicity

replaced by complexity. It is perhaps not generally

recognized that the subjunctive mood reveals a somewhat

similar story, though with the difference that the earlier

synthetic method was not entirely lost before it began to

be replaced by the analytical. In Modern English, it

is of course obvious that little use is made of the once

prevalent inflected subjunctive. That is not, however,

equivalent to saying that the subjunctive has but little

place in Modern English; on the contrary, the subjunc-

tive formed by the auxiliaries may, might, should, would,

could, ought, and so forth (a method exactly parallel to

that used in the compound tenses of the indicative) has

been constantly given a wider and more subtly differ-

entiated use.43 There is danger, then, of assuming

overhastily that the undeniable loss of inflectional end-

ings necessarily means greater simplicity in the Modern

English verb. The two illustrations just cited indicate

that greater complexity, though of a different kind, may

be the eventual result of the simplification of verbal

endings.

The usual classification of the Modern English verb

follows, almost exactly, the distinctive twofold division

of the verb in the Teutonic languages generally; that is,

its separation into two chief groups, strong and weak.

For the most part, the weak verbs of Old English have

become the so-called "regular" verbs of Modern English,

and the strong verbs the "irregular." As has been said,

43 See Curme, Syntax, pp. 390–430.

English Inflections

141

the basis of the distinction is the use of the dental suffix (-ed, -d, or -t) in the one case and of an internal vowel change (ablauts series) in the other. The former group, the weak verbs, has been the larger in all periods of English, and it grows steadily even now, both because newly created or newly borrowed words are invariably conjugated in this way and because verbs that were once strong not infrequently become weak; thus, dived has replaced dove, and climbed has replaced clomb or clumb.

There are a few exceptions to the latter type of change, however: in certain instances a shorter form, made by the analogy of some existing strong verb, has replaced a longer weak past tense; thus strove, stuck, and dug have displaced strived, sticked, and digged. But on the whole the simplifying and regularizing tendency is very manifest; the strong verbs have tended both to conform to the simpler pattern of the weak, and to level the variations of the ablaut series within their own conjugation. Both tendencies have, of course, freer play in the language of the uneducated and of children than in the speech of those who are more restrained by conventional "rules." Thus, the man in the street is likely to substitute drink, drank, drank for drink, drank, drunk, and swim, swam, swam for swim, swam, swum; similarly, young children instinctively prefer runned to ran and comed to came.

The strong verbs in Old English fall into six classes, according to their ablaut series; in addition, there is a seventh group made up of reduplicating verbs, which may or may not have gradation. All of these are represented in Modern English. Principal parts of a representative example of each of the six ablaut series, in Old and Modern English, follow. It will be noted

142 The Development of Modern English

that the four principal parts of Old English have been

reduced to three in Modern English, since the distinction

between the second and third forms (the preterits, sing-

ular and plural) has long since been lost; further, as

we have seen, the three variants often tend in Modern

English to be simplified to two, and the principal parts

therefore, to be no more complicated than those of weak

verbs.

Old English Modern English

1. drīfan, drāf, drifon, drifen drive, drove, driven

2. cēosan, cēas, curon,44 coren44 choose, chose, chosen

3. singan, sang, sungon, sungen sing, sang, sung

4. beran, bær, bæron, boren bear, bore, borne45

5. etan, æt, æton, eten eat, ate, eaten

6. scacan, scōc, scācon, scacen shake, shook, shaken

Reduplication of the initial consonant to form preterit

tenses occurred in many Teutonic verbs, but the only

trace of this in Old English is the resulting contraction

of certain verbal forms. The vowel sequence of a fair

number of Old and Modern English verbs is neverthe-

less the result of their former initial inflection.

Typical representatives of the reduplicating (or

seventh) group are the following:

Old English Modern English

bēatan, bēot, bēoton, bēaten beat, beat, beaten

blōwan, blēow, blēowon, blōwen blow, blew, blown

44 In these two forms, r has taken the place of s by a substitution called

rhotacism. To state the matter more fully, the Indo-European s became

voiced [z] in these forms because the preceding vowel was not accented

(Verner's Law): later, as in Old English, this voiced s passed to r. The

Modern English forms of course show the analogical replacement of s.

Compare the past participles of two other verbs of this group, lēosan

(lose) and frēosan (freeze): lorn and frore survive as archaic forms in

Modern English, but for general use they are replaced by lost and frozen.

Was and were (<O. E. wæs and wēron) show a similar alternation, pre-

served to this day.

45 Modern English born is a variant form, differentiated in meaning.

English Inflections

143

Old English Modern English

feallan, fēoll, fēollon, feallen fall, fell, fallen

grōwan, grēow, grēowon, grōwen grow, grew, grown

Weak verbs in Old English have the preterit in -de or -te, and the past participle in -d or -t; in Modern English the same verbs end, in both preterit and past participle, in -d, -ed, or -t. This is the surest criterion of the distinction between strong and weak. Other indications differentiating them may be confusing: internal inflection, for example, may occur in the weak verbs as well as in the strong, though in the former it is due to mutation or umlaut rather than gradation or ablaut; then, too, the formerly distinctive suffix of the strong past participle, the -en ending, has been lost in many verbs. Weak verbs are divided, in Old English, into three groups, of which only two are represented by many verbs, the third having been reduced to just four verbs.

The following examples may be given:

Old English Modern English

Class 1. bringan, brōhte, gebrōht bring, brought

hīeran, hīerde, gehīered hear, heard

settan, sette, gesett set, set46

tǣcan, tǣhte, getǣht teach, taught

Class 2. clǣnsian, clǣnsode, geclǣnsod cleanse, cleansed

lufian, lufode, gelufod love, loved

seglan, seglode, geseglod sail, sailed

Class 3.47 habban, hæfde, gehæfd have, had

libban, lifde, gelīd live, lived

secgan, sæde, gesæd say, said

46 The so-called invariable verbs of Modern English represent the ultimate simplification of principal parts. As we have seen, four forms must be given for the Old English strong verbs and three for the weak; in Modern English, the four strong forms are reduced to three and sometimes to two, while the three weak forms are reduced to two and sometimes to one. Other invariable verbs are burst, cast, cut, hit, hurt, let, put, shed, shut.

47 Only one other verb of this class survived in Old English: hycgan (think).

144 The Development of Modern English

The foregoing classification does not include certain verbs which, though numerically few, are in use extremely important. Two irregular groups must therefore also be mentioned: the mi-verbs and the preterit-presents. The former are the sole survivors of another type of conjugation than that which gives the strong and weak verbs of Old English. Indo-European verbs used either -ō or -mi as the ending of the first person present indicative singular. Almost all English verbs are of the former type, but to the mi-class belong a few of our most important irregular verbs: be, do, go, and will. These are verbs that go back eventually to the classification of such Greek verbs as εἰμί, τίθημι (in Latin sum, λúω, φέρω. Only one form of Modern English, am, retains a trace of the -mi ending, but in Old English dōm (I do) is also found—in the Mercian and Northumbrian dialects. These four so-called anomalous verbs —be, do, go, and will—are for the most part irregular only in the present tense, but in be and go it should be noted that more than one root is used in the conjugation. Is (or are) and was are not from the same source as be, while go formerly used as its preterit a separate form, ēode (later yede), and now uses still another form, went, originally a preterit of wend.

The other group of irregular verbs comprises the preterit-presents, so called because the present tense of each of them is a former preterit that has superseded the old present and is now a present in use.48 The old

48 Just as Latin novi and Greek οἶδα have become present in meaning (= I know); Old English wāt is exactly parallel.

English Inflections

145

strong preterit has then usually been succeeded by a new, weak form. The preterit-presents that are retained in Modern English are ought, can, dare, shall, may, must. The first and last have no separate forms for the preterit; the others have could, dared (formerly durst, but now usually regularized), should, and might. Ought and must are peculiar in that, in their histories, preterit forms have twice become present in use: the original strong preterits, āh and mōt, became present in meaning and were succeeded by weak preterits, āhte and mōste (Mn. E. ought and must); these new preterits in turn became present in use as well as past—as is true also of durst. There is consequently ambiguity in the modern meaning of ought, must,49 and durst. Durst for past use has been largely replaced by the new form dared; but such a substitution is scarcely possible for ought50 or must, since these verbs have already exhausted the two possible types of preterit formation, the strong and the weak. All these preterit-present verbs are conjugated in the present indicative without -s in the third person singular; in other words, like forms of the preterit tense rather than of the present. Dare, however, is a partial exception, since dares (by analogy of regular verbs) is now more usual than dare.

The simplification of inflectional endings may now be summarized. The present and preterit endings of the typical Old English verb were as shown at top of p. 146.

49 “He must go,” for example, formerly meant “He had to go,” but now ordinarily means “He has to go.”

50 There is the additional reason, in this instance, that the regular verb owe, owed, owed comes from the same source as ought and has been differentiated from it in meaning.

146 The Development of Modern English

INDICATIVE

PRESENT PRETERIT51

Sing. Plur. Sing. Plur.

1. —e —að —ede —edon

2. —est “ —edest “

3. —eð “ —ede “

SUBJUNCTIVE

1, 2, 3. —e —en —ede —eden

In addition, there were also these forms: two imperatives, the singular without ending for strong verbs and in —e for weak verbs, and the plural in —að; the infinitive in —an and the gerund in —anne; the present participle in —ende and the past in —ed or —od (for strong verbs, —en). The past participle was prefixed by ge-, normally, in the weak verbs, and also often in the strong verbs.

In Middle English, these were the typical endings:

INDICATIVE

PRESENT PRETERIT52

Sing. Plur. Sing. Plur.

1. —e —e(n) —ede —ede(n)

2. —est “ —edest “

3. —eth “ —ede “

SUBJUNCTIVE

1, 2, 3. —e —e(n) —ede —ede(n)

51 The forms given are for the weak verb. The strong preterit was more complicated, since two different stems are involved: one for the first and third singular, and the other for the remaining four forms of the indicative; for example, ic cēas, pū cure, hē cēas, wē (gē, hīe) curon. In the subjunctive, only the second stem is present: cure for the singular, curen for the plural.

52 Again the forms of the strong verb were somewhat different from the forms given, which are those of the weak. The preterit singulars were 1 and 3: —, 2: —e; the plurals, 1, 2, and 3: —e(n). Usually, the two preterit stems have been leveled to one—more commonly, the one originally singular. Occasionally, however, even as late as Chaucer, both are preserved, so that we have such a conjugation as this: singular, 1 and 3, bigan, 2, bigonne; plural, bigonne(n). In the subjunctive, as in Old English, the singular is usually (but not always) made by attaching —e to the preterit plural stem, and the plural by attaching —e(n).

English Inflections

147

The imperatives and infinitives are as in Old English, except that a regularly weakens to e; the plural imperative, accordingly, ends in -eth, and the infinitive in -e(n). The gerund is usually the same form as the infinitive, but sometimes a distinctive form like to seyne or to done occurs in Chaucer. -Ing(e) replaces -ende as the ending of the present participle, possibly through the influence of verbal nouns in -ing(e) [from O. E. -ung(e)].53 The usual endings of the past participle are, for weak verbs, -(e)d or t, and for strong verbs, -en (which is, however, sometimes syncopated to -n and sometimes altered to -e). The old prefix to the past participle, ge-, has been weakened to y-, and appears with many verbs, both strong and weak. Incidentally, the fact that y- in Middle English is not invariably the sign of the past participle (the O. E. infinitive forms gepeon (prosper) and gesēon (see), for example, occur in Chaucer as y-thee and y-see) may have tempted later writers of archaic tendencies to employ it in other verbal forms quite without any historical basis for doing so; this seems the explanation of such an error as Milton's "star y-pointing pyramid."54

In further explanation of the transformation of the inflectional endings of Old English into those of Middle English, one or two additional remarks may be made. The -(e)s ending (in place of -eth) for third person singular present indicative appears in Chaucer but rarely, and sometimes distinctly as a dialectal form; its regular use in the later language is explained as a taking over of a Northern ending, rather than as a development of the Midland -eth. Likewise, the Middle English -e(n) (in

53 Emerson, p. 376.

54 "On Shakespeare," 1. 4.

148 The Development of Modern English

the plural of the present indicative) can scarcely be explained as an outgrowth of Old English -að; rather it would seem that when a distinctive form was needed for this purpose (because -að would normally evolve into -eð, the same ending as that of the third person singular), the subjunctives in -e(n) and possibly the strong preterits of the indicative in -e(n) (<O. E. -on) furnished the model.

The normal Modern English endings may be listed for purposes of comparison:

Indicative

PRESENT Sing. Plur. 1, 2, 3. — — — 1, 2, 3. — — —

PRETERIT55 Sing. 1. —ed " 2. " " 3. " " Plur. —ed " " "

Subjunctive

PRESENT 1, 2, 3. — 1, 2, 3. —

PRETERIT 1, 2, 3. —ed 1, 2, 3. —ed

The imperatives, singular and plural, and the infinitive are without ending. The present participle ends in -ing, and the past in -d, -ed, -t, -n, or -en. Archaic or poetic language occasionally substitutes the ending -est (always preceded by thou) in the second person singular present and preterit indicative, and the ending -eth in the third person singular present indicative.

If, then, we summarize the leveling of verbal endings in Modern English, we observe that only one personal ending, the -s of the present indicative third person, has survived. To offset this impression of simplicity, however

55 Once more, the strong verbs are distinct from the weak in the preterit. Now, however, only a single preterit stem is used, which is made distinct from the present by internal vowel change and which takes no ending.

ever, it is to be pointed out that Modern English has

built up, on foundations laid in Old and Middle English,

a most elaborate, though scarcely complicated, system

of compound forms. The additions include a future

tense by means of auxiliaries; a complete passive voice;

what amounts to a new mood, the potential subjunctive

(with can, must, may, and so forth), as well as a recon-

struction of the subjunctive-optative; and a differentia-

tion into three distinct types of statement—simple or

indefinite, progressive or definite, and emphatic. To

illustrate the last point, it may be remarked that Modern

English may vary such phrases as “I give,” “I walk,”

“I gave,” and “I walked” in either of two ways: “I am

giving,” “I am walking,” “I was giving,” and “I was

walking”; or “I do give,” “I do walk,” “I did give,” and

“I did walk.” Thus, while Modern English on the one

hand has reduced the number of different forms to but

six for strong verbs and five for weak (as compared with

a maximum of twenty-five in Gothic and fourteen in

Old English), it has also greatly increased the number and

variety of verbal phrases. The history of the English

verb in general is therefore a particularly striking instance

of the progress from synthesis toward analysis, of the

substitution of invariable and separable form-words for

longer one-word inflections.

Now that the review of the parts of speech in which

inflection occurs has been concluded, it is appropriate

to examine in greater detail a question already suggested,

and one which, as has just been remarked, the develop-

56 A typical presentation of modern verbal conjugation, that in House

and Harman’s Descriptive English Grammar (New York, Prentice-Hall,

1931), takes ten pages in its enumeration of the forms now in use in a

single verb

150 The Development of Modern English

ment of the verb particularly places before us. Is the evolution of a type of inflection that is chiefly analytical rather than synthetic, in the species of word-formation that it implies, to be thought of as beneficial or detrimental? Parenthetically it may be remarked that unless one believes that everything happens for the best in this best of all possible worlds, it is not necessary to believe that every change which on the whole is beneficial must be felt to be unqualifiedly so. Nor, on the other hand, is it necessary to believe that loss at one point implies loss at every point. Changes in language, like changes in other human institutions, are much more likely to be compounded of both profit and loss. Our question, then, is not necessarily whether inflectional simplification is all profit or all loss, but rather, granted that one is likely to entail some measure of the other, whether the advantages outweigh, or do not outweigh, the disadvantages. It is clear, of course, that if we have faith in the effectiveness of Modern English, we must believe that the advantages are, on the whole, more important than the disadvantages. Even though our conclusion is thus anticipated, it is still of value to pursue the inquiry to the extent of ascertaining, as definitely as possible, just what has been gained and what has been lost as a result of the long-continued process of inflectional leveling. In doing this, we shall not confine ourselves to the differences between older and more recent aspects of the English language, but shall at least glance at the larger question of the advantages and disadvantages of a language that is largely analytic as compared to one that is largely synthetic.

If, then, we look first at the elements of loss that are incidental to inflectional simplification, the first general-

ization that suggests itself is this: The sound changes

that are brought about in the evolution of Modern

English inflection have meant a sacrifice of musical

quality. There is no figure of speech, for example, in

the familiar statement that Chaucer's language is more

musical than ours. It is a matter susceptible of definite

proof. Such a line as "And smale foules maken melodie"

is superior in musical content to the Modern English

line that represents the translation of the words into their

present forms and sounds-"And small fowls make

melody"—for the very specific reason that in the

modernization, four vowels have been lost,57 while all

the consonants but one are retained. And vowels are

essentially musical notes, while consonants are brief

noises. To decrease the percentage of vocalic sounds

and increase that of consonantal sounds is therefore to

make the language more harsh. In the line just con-

sidered, one may observe how the omission of vowels

takes away a certain tripping melody and substitutes

uglier consonantal combinations. The point may be

summarized in the assertion that inflectional leveling

frequently means the loss of vowels; this is especially

striking in the comparison of Middle and Modern English,

since an Old English inflectional ending consisting of

vowel plus consonant is often represented in the Middle

period by the vowel alone.

A more fundamental disadvantage, however, since it

affects the utility rather than the beauty of language, is

that inflectional leveling has brought about a greater

57 To be strictly accurate, it should be added that one of the original

vowels is replaced with a diphthong—but this scarcely affects the present

point. Further, the loss of the single consonant is not altogether typical,

for make is as frequent as maken in Middle English.

152 The Development of Modern English

danger of ambiguity. When words are stripped of their inflectional endings, all sorts of duplications in form are created. The mere appearance or the mere sound of the word sometimes gives no clue as to whether it is, for example, an infinitive or an indicative form of the verb, or whether it is a verb in the present or the preterit tense, or even whether the form is a verb or a noun. Perhaps the most extreme examples are the invariable verbs of Modern English. Such forms as cast, cut, and let are thoroughly ambiguous: they may represent absolutely any use of the verb except the third person singular present indicative and the present participle58; and they may be either verbs or nouns. There is, of course, something to be said for the simplicity of expression that this kind of duplication of forms implies. At present, however, we are concerned with the fact that in reading Modern English one sometimes has occasion to wonder whether they cut, for example, means "they are cutting" or "they were cutting," or whether (to take another type of duplication) we read implies the sounds [red] or [ri:d].

An important result of the increasing duplication of forms in the development of English has been the greater and greater necessity of observing a uniform and stereotyped order of words in the sentence. Modern English is more definitely restricted than Old English, and much more than such a language as Latin or even Modern

58 The ambiguity of let, incidentally, goes further than this, since two separate Old English verbs have grown together into the present form. Curiously, they had almost opposite meanings—“to allow” and “to hinder.” Both senses are present in early Modern English; the Biblical “let or hindrance” and Shakespeare’s “Thy kinsmen are no let to me” (Romeo and Juliet, II, ii, 69 [q1.]) , and “I’ll make a ghost of him that lets me” (Hamlet, I, iv, 85) represent the less familiar meaning.

English Inflections

153

German, to the one order: subject, predicate, object. The reason, of course, is that paucity of inflectional and unambiguous forms makes it necessary to adhere strictly to a prescribed sequence of our largely invariable words.

It may be replied that there is no loss here; that if, to avoid ambiguity, we are compelled to use just one word-order, so much the better, in the interests of simplicity.

But are there no useful purposes served by varying the order?

It would seem that highly inflected languages have a decided advantage in this respect, in that graceful variations in expression and subtly differentiated shades of emphasis are possible in a measure which is denied to the slightly inflected language.

To justify this assertion definite comparisons are necessary.

If we wish to state in English that Romulus founded Rome, there is only one straightforward way of doing it: “Romulus founded Rome.”

To reverse subject and object—“Rome founded Romulus”—is to violate idiom, if not truth.

The Latin order that most nearly corresponds to the English subject, predicate, object is subject, object, predicate: Romulus Romam condidit.

Here, however, the order is not necessitated by the ambiguous appearance of the subject and of the object, for Romulus is clearly marked as a nominative and Romam just as clearly as an accusative.

The Latin order, in fact, is subject to variation to express minor shades of emphasis.

Thus, Romam Romulus condidit is not so much “Romulus founded Rome” as “Rome was founded by Romulus,” but it has the advantage of avoiding the awkward passive

59 The illustration is suggested by Classen, Outlines of the History of the English Language, pp. 15 and 16.

154 The Development of Modern English

construction; and while Condidit Romulus Romam can be turned into English as "The founding of Rome was done by Romulus," the eight English words are at best a clumsy paraphrase of the three Latin ones.

A further illustration of a kind of rhetorical effectiveness that is impossible to the language of few inflections is such a Latin sentence as this: Astra regunt homines, sed astra regit Deus. The neat parallelism of the Latin is not to be taken over in English, for "The stars rule men, but God rules the stars" varies the position of "the stars," while "The stars rule men, but the stars are ruled by God" introduces the passive voice in the second clause and thus sacrifices a certain unity of expression.

It is evident, then, that there is a real disadvantage to English in being restricted, through inflectional simplicity, to a regular word-order. That writers of English have found this restriction irksome is attested by the occasional attempts, particularly in poetry (where the exigencies of rhyme and metre may be more responsible than a desire for grace or emphasis), to get away from it. A stock illustration is this line in Gray's Elegy:60

And all the air a solemn stillness holds,

where it is impossible to tell which of the two nouns that precede the predicate is subject and which is object. Inversions of all sorts, in prose as well as in poetry, are likely to run the risk of similar ambiguity; at best, they occasion difficulty or hesitation for the reader.

The necessity for a set word-order in Modern English extends to the position of modifying words in the sentence. Modifying words must not only be placed in close proximity to the words modified; almost always

60 Quoted, for example, in Jespersen, Language, p. 345.

they directly precede such words. The split infinitive,

as we shall see,61 exhibits this principle in operation and

taking precedence over another convention. The usual

position of the qualifying adjective is another illustration.

Such a line of Chaucer's as this:

Thebes with his olde walles wyde62

typifies the older, freer order—the noun posed between

the two modifying adjectives, not preceded by both.

Here the vestige of adjective declension, the plural final

-e's, suffices to attach them visibly to the plural noun.

It is the loss of these endings that eventually makes it

necessary to alter the position of the adjectives, so that

their function is made evident by position alone, not at

all by form. In Modern English there is, of course, no

other order than “its old, broad walls.” Occasional

exceptions like Kipling's title Captains Courageous and

such advertising phrases as “suburb superb” and “house

beautiful” call attention to themselves by their very

oddity.

So much for the debit side of the ledger. What has

been pointed out is that inflectional simplification

involves above everything else the danger of ambiguity

and necessitates à careful observance of strict rules as to

the order of words in the sentence, in order to lessen that

danger. Before leaving this point, however, it may be

added that even highly inflected languages are not

entirely free from the possibility of ambiguity. It is

clear that no language can escape this danger so long as

61 P. 504.

62 “Knight's Tale,” l. 1880. Cf. also a favorite Chaucerian phrase:

“my blisful lady free” (i.e., “my blessed, generous lady”).

156 The Development of Modern English

there is any duplication of form. The Latin phrase Malo malo malo malo, translated by the jingle

I'd rather be

In an apple tree

Than a bad man

In adversitee,

may serve to illustrate, rather grotesquely, the possibilities of duplication of forms in a highly inflected language. The reader may be interested, too, in working out the various possible translations of a sentence suggested by Jespersen63: Horatius et Virgilius poetae Varii amici erant. The amazing variety of possible meanings here is of course dependent on two facts: -ae and -i endings; and the Latin word-order is not at all a fixed one. We are not to suppose, then, that a high state of inflection is a guarantee against ambiguity any more than we are obliged to feel that the resulting economy in numbers of words is an unqualified blessing.

Turning to the positive merits of the language of few inflections, it would appear that, from the present point of view, the most conspicuous advantage that a language like English enjoys over one like Latin is the comparatively small place given to the expression of grammatical or syntactical ideas. If simplicity can sometimes be connected with ambiguity, it can also be connected with the suppression of the irrelevant. When we say "the inflections disappeared," we express the idea of plurality once, and once only, in the -s ending of the noun; but a more highly inflected language would express it twice more-in the form of the article and in

63 Language, p. 343.

English Inflections

157

the ending of the verb. The needless repetitions, going

far beyond the demands of clarity, such as are common in

a synthetic language, are largely eliminated. Bradley64

cites, as a “somewhat extreme instance” of the obtrusion

of grammar on the attention, the Latin phrase duōrum

bonōrum virōrum. Here the attention is certainly unduly

distracted from the really important parts of the phrase,

the du-, bon-, and vir-, to the three sonorous (and

accented) -ōrum's, which after all mean nothing more

than “masculine, genitive, plural.” Bradley felicitously

labels the relative absence of this sort of thing in English

its “noiseless grammatical machinery.”

The Latin phrase just quoted suggests still other

aspects of the contrast between the synthetic and the

analytic types of expression. If, beneath

duōrum bonōrum virōrum,

we set

of two good men,

we observe that the greater diffusion of English, usually

assumed to exist, is present, if merely numbers of words

are counted; there are, that is to say, four words of

English as against three of Latin. Actually, however,

more space is required for the Latin phrase, and more

syllables (in the proportion of nine to four) are present

in it. Jespersen65 ingeniously advances the formula

axyz + bxyz + cxyz (in which a, b, and c stand for root

ideas, and x, y, and z for inflectional ideas) to suggest

the synthetic, and (a + b + c)xyz to suggest the ana-

lytic, method of expression. The Latin phrase mentioned

above is a concrete illustration of the first formula, while

64 Making of English, p. 77.

65 Language, p. 351.

158 The Development of Modern English

the English "of two good men" is true to the analytic ideal in doing away with the repetition of the common factors. Thus, case is expressed but once, through the preposition of; number is expressed but once, through the plural form of men; and gender is not expressed at all—and there is nothing to regret about its absence.

The elimination of all inflectional distinctions in the adjective, the definite article, and the possessive pronoun is perhaps one of the most obvious and least debatable advantages of inflectional simplification. Surely it is a clear gain for English to be able to say "the wife and children" rather than la femme et les enfants, or "my father and mother" rather than mein Vater und meine Mutter. The gain as compared with German is greater than as compared with French, for German has, in addition to the distinction for number that it shares with French, a more elaborate set of distinctions for case. Thus mit oder ohne cannot be used with the same freedom as with or without or sans ou avec, for the reason that mit is followed by the dative case and ohne by the accusative. English is, however, superior to both French and German in the greater ease with which the parts of speech just mentioned can be used—for still another reason: the leveling of inflectional distinction has been accompanied by the substitution of natural for grammatical gender. "For this irrational distinction," as Logan Pearsall Smith puts it,66 " . . . is yet found, as a survival of barbarism and a useless burden to the memory, in all the other well-known languages of Europe."

Among the advantages accruing to the language through inflectional leveling it seems justifiable, on the

66 The English Language, p. 14.

whole, to list the shortening of words that has largely

been brought about in this way. To strip the roots of

words free from syllables that express frequently irrel-

evant grammatical relations must be reckoned a benefit.

And if the effect of this, especially on our native vocabu-

lary, is to produce many monosyllables, the process still

need not be regretted. After all, the Anglo-Saxon

monosyllable is quite likely, in practice, to be ranged

side by side with the borrowed polysyllable, and the con-

trast is often admirable and effective. Then, too, every

student of English literature must be aware of the

splendid use that has been made of monosyllables

alone. The matter will be referred to again, so that

this point need not be labored here; but perhaps a single

example may be mentioned. The marvelously poignant

last line of Milton’s sonnet “To his Deceased Wife”:

I waked, she fled, and day brought back my night

surely owes part of its effectiveness to the slow movement

that the succession of weighty monosyllables entails.

To sum up, then: it has been argued that the simplifica-

tion of English inflections, proceeding of course hand in

hand with phonetic obscurings and with the discarding of

grammatical gender, has meant the triumph of analysis—

a triumph that differentiates English in degree if not in

kind from other European tongues. The elimination of

the superfluous and the development of the most econom-

ical and direct means of expression are victories of no

mean consequence. One last question still suggests

itself about the process as a whole: Has inflectional

simplification run its course in English, or are we to

regard the language of the present and the future as being

160 The Development of Modern English

susceptible of it in the same measure as the language of

the past?

The latter part of our question evidently answers

itself. There can scarcely be the same measure of

inflectional leveling in the future that there has been in

the past for the obvious reason that there is far less room

for it in Modern English than there was in Old English.

Theoretically, of course, there is still opportunity for it in

many places. We might, for example, make all plural

nouns end in -s, and all preterits of verbs end in -ed:

we might say mans instead of men, and ringed instead of

rang. Here, however, it is evident that the powerful

counter-influence to the trend toward regularity and uni-

formity is habit or custom, particularly well-established

and unassailable in forms that are in constant use.

Revolutionary changes of the sort suggested are not in

the least likely in any time that we can at present foresee.

What can be expected in the way of inflectional simplifi-

cation is less spectacular: the reducing of three principal

parts for verbs like swim and drink, for example, to two

by the leveling of the distinctive form of the past partici-

ple to agree with that of the simple preterit. Colloquial

language, as we have seen, has already made these

changes. Colloquial speech tends also to eliminate the

distinction between the two auxiliaries of the future

tense, and to destroy a special accusative form for who.

Perhaps, too, the tendency, sometimes evident even in

educated speech, to level the distinctive third person

singular of doesn't he to don't he, on the analogy of the

other forms of the conjugation, may be thought of as an

additional illustration. On the whole, however, we are

probably justified in thinking of inflectional leveling as a

historical phenomenon that has well-nigh reached the

limit of its operation. It may, in the future, wear away this or that inflectional excrescence that still protrudes,

but it is bitterly opposed by the teaching of the schools, and this, with the smaller field that is still open to its

influence, will restrict its results to exceedingly limited dimensions. Theoretical considerations of economy, it

must be added, have small weight against educational conservatism and long-established custom. It can be

argued that the -s ending of the third person singular present tense of verbs in general and the anomalous

forms of the verb be are impediments to the free and facile expression of thought.67 Likewise, as has more

than once been pointed out, it would be logical to eliminate these and those as special plural forms (on the analogy

of the or of my, your, or his). Nevertheless, there is not the slightest likelihood of these changes taking place.

On the whole, even from the angle of theory, the best and most desirable simplifications have already been

made. Inflectional leveling will be less successful in the future precisely because it has been so thoroughly suc-

cessful in the past.

References for Further Reading

The histories of the English language referred to on page 37 (Chapter II) or page 64 (Chapter III), and, in addition:

Aiken, Janet R., English, Present and Past, New York (Ronald), 1930.

Bloomfield, Leonard, Language, New York (Holt), 1933.

Curme, G. O., Syntax, New York (Heath), 1931.

Jespersen, Otto, A Modern English Grammar, Pts. II-IV, Heidelberg (Winter), 1922, 1927, 1931.

_______________, Essentials of English Grammar, New York (Holt), 1933.

_______________, Language, New York (Holt), 1924.

67 Cf. Jespersen, Language, pp. 335 and 336.

CHAPTER VI

English Sounds and Their History

FOR even the most cursory survey of the sounds of our speech and their use in the language of the past and the present, recourse to a phonetic alphabet is virtually inevitable. The reason, of course, is that while all existing alphabets are inadequate to the task of indicating, with complete accuracy, the actual sounds of speech, and while the spelling of all languages is to some degree unphonetic, these shortcomings are, in Modern English, peculiarly acute. It is to be observed, however, that the difference is rather one of degree than of kind.

Perhaps we are accustomed to think of Italian as standing at the opposite end of the scale from English, among modern European languages, in that its spelling is as reliable a guide to its pronunciation as that of English is unreliable. But if a phonetic alphabet implies a single and invariable meaning for every symbol, one has only to observe the inconsistent use of c and g in such proper names as Galli-Curci and Giovanni Boccaccio in order to realize that Italian too falls short of the phonetic ideal.

It is equally clear, nevertheless, that if the spelling of Italian is not altogether satisfactory, from the present point of view, that of Modern English may fairly be described as totally, and indeed almost ludicrously, unsatisfactory. To use a phonetic alphabet for the discussion, on the printed page, of English sounds, past and present, is therefore imperative.

162

English Sounds and Their History 163

A phonetic alphabet may be invented for more than one purpose. It may be intended as a complete substitute for our ordinary alphabet; the eventual goal of the systematic movement for the reformation of English spelling, promoted by the Simplified Spelling Board, is the replacing of our present alphabet with a phonetic one.

More frequently, however, the use of a phonetic alphabet has the less ambitious purpose of serving, purely as a temporary expedient, to present the sounds of speech in a clear and unambiguous fashion. Such is the use that will be made of a phonetic alphabet in the following pages.

For our purpose, too, it is unnecessary to aim at the highest degree of precision in differentiating among sounds. What phoneticians describe as a "broad" rather than a "narrow" system of notation will therefore serve.

The basis of such a system, almost necessarily, is the alphabet of the International Phonetic Association, with whatever minor modifications seem desirable for the needs of an elementary discussion of phonetics, which will emphasize, of course, Modern English and, more particularly, present American pronunciation.

Ideally, one might begin with an entire reconstruction of the present alphabet. The Roman letters that, in the sixth century, superseded the runic alphabet used by the Germanic invaders are not particularly successful in expressing the relation between one consonantal sound and another.

It is true that m and n do imply, by their appropriate similarity in form, a certain real parallelism in the sounds which these symbols indicate. These two, however, are the exception. As we shall see in a moment, t-d, k-g, f-v, and so forth, stand for paired sounds that ought, logically, to be represented by pairs of similar but differentiated symbols.

The symbols of Pitman

164 The Development of Modern English

shorthand ( ʃ , — , ʌ ) are, from this point of view, very much better. The conventional symbols, however, are so thoroughly established in general European use that the International Phonetic Alphabet could scarcely discard such of them as have a relatively standard meaning. For vowels too, most European languages (not, however, Modern English) agree fairly well on certain basic sounds that can be represented by generally understood symbols. The so-called Continental values are therefore the starting point for the indication of the vowels in our phonetic alphabet. To these letters must be added, for the vowels, additional symbols to stand for some at least of the many vocalic sounds that the five basic symbols do not unequivocally represent; and, for the consonants, additional symbols to stand for elementary sounds—like those represented by sh (in she) and ng (in sing)—that nevertheless have no place in the usual alphabet.

A possible alternative to such a scheme as that of the International Phonetic Alphabet is to follow the system used in many of our best dictionaries, and to indicate pronunciation so far as possible without respelling—by employing the diacritical markings, for example, of the Websterian system of notation. Phoneticians, however, are almost unanimous in decrying the use of these markings as unscientific and as thoroughly misleading in that they continually suggest a relation that does not exist between present-day spelling and pronunciation. Their familiarity to many Americans, which has come about through their use in dictionaries, can only be deplored. Whatever small trouble the memorizing of the phonetic alphabet entails, the reader may be assured that the time spent in mastering it will be well spent, for

English Sounds and Their History

165

it will give him the key to reading most of the recent works on phonetics that are best worth reading. How-

ever, before proceeding to the description and classifica-

tion of English sounds, which will involve the use of phonetic symbols, it will be convenient to glance very

briefly at the physiology of sound production.

It is conceivable that speech sound may be produced by the inhalation as well as the exhalation of the breath; certain African tongues, for example, are said to employ

different varieties of inhalation, and even in English, its use in certain exclamations is not unfamiliar. This possi-

bility may, however, be virtually excluded if our discussion is to be limited to the normal sounds of English speech.

We may assume, then, that the sounds of speech are produced by breath which is emitted from the lungs and

which passes through the larynx and finally emerges from the oral or the nasal cavity. If the breath is

entirely unmodified in its passage, the result is an inaudible, or almost inaudible, sound. This is the sound of

ordinary breathing; when the breath is emitted with more force than usual, it becomes the sound [h].1 To produce

any other speech sound, however, there must be modification of the air which is expelled from the lungs. This

modification may take place in either or both of two places: the larynx and the resonance chamber—by

which latter term is meant the general cavity of the mouth and of the nose.

Before continuing with the idea of modification, however, it will be useful to go back a step. The breath,

which may be modified in the larynx or in the resonance

1 Brackets are used here and elsewhere to indicate that the letters included within them are to be understood as phonetic symbols.

166 The Development of Modern English

chamber, is furnished by the lungs. This then is the

first function of the lungs: to provide the raw material

of speech sound. Their second function is to furnish the

breath in varying degrees of force, to impart to speech

the quality which we know as stress. This, as we have

seen, is of primary and peculiar importance in the

Teutonic languages, since the Teutonic accent is entirely

a matter of stress, and since the English habit of over-

stressing certain syllables and understressing others has

largely contributed to the inflectional leveling of Modern

English. The whole system of Teutonic versification,

too, is built on the distinction between the stressed and

the unstressed syllable.

From the lungs, the air passes through the bronchial

tubes and the trachea (windpipe), to the larynx or voice-box.

The breath then emerges from the glottis, or rift between the

membranes of the larynx. These membranes are called

the vocal cords. If the glottal rift is wide, the breath

passes through without causing the vocal cords to vibrate,

and the resulting sound is merely unmodified breathing.2

If, however, the rift is made smaller by the tautness of

the vocal cords, the cords vibrate and the quality

known as voice is imparted to speech. This happens in

the production of all vowels, and in the production of the

voiced, as opposed to the breath (or voiceless), con-

sonants. The action of the vocal cords likewise imparts

another general quality (besides voice) to the breath

used in language. This is pitch, the distinction between

a high and a low sound (as these terms are used in music).

The rate of vibration of the vocal cords controls the

2 Or [h], as noted above.

English Sounds and Their History 167

pitch, rapid vibration making for high sounds and slow vibration for low sounds. The reader is probably familiar with the way in which a violin string may be made to give forth a higher and higher tone as its length is diminished by the fingers of the player, and as the rate of vibration is thereby increased. However, the parallel with the vocal cords is misleading if it suggests a similar picture of strings freely vibrating in space: the vocal cords are of course membranes attached to the sides of the larynx. It should be added too that while distinctions of pitch are of considerable importance in many languages—the original Indo-European accent is thought to have been a matter of pitch rather than stress, and certain languages (like Chinese) utilize pitch3 as a means of distinguishing the meaning of one word from another—it plays in English, and in the Teutonic languages generally, a much less significant part than stress.

So far, we have been concerned with general qualities—stress, voice, and pitch—that are imparted to speech sound by the action of the lungs or of the larynx. The resonance chamber, on the other hand, is the place where the individual differences that separate one consonant from another and one vowel from another are made. The breath passes from the larynx into the pharynx, and is then expelled from the mouth or, in the case of the nasals [m], [n], and [ŋ], from the nose. The principal modifiers of the oral cavity and those that give the final differentiations to speech sound are these: the soft palate (or velum), which can be raised to shut off

3 In the Cantonese dialect, eight degrees of pitch are said to be distinguished.

168 The Development of Modern English

the nasal cavity; the hard palate, which is the roof of the mouth from about the middle to the gums; the gums (or alveoli); the teeth, upper and lower; the tongue; and the lips. The parts that these organs play in the formation of speech sound will be indicated in the description and classification of the vowels and consonants.

To divide the sounds of speech into vowels and consonants is the usual custom, but the basis of this distinction is by no means as obvious as might, at first sight, appear. Though it is evident that all vowels are voiced, it clearly will not do to describe vowels as the voiced sounds and consonants as the breath sounds, for such consonants as [b] and [v] are also voiced. On the other hand, it just as clearly will not serve to describe vowels as the prolonged sounds and consonants as the momentary sounds, for such consonants as [f] and [s] are capable of indefinite prolongation. The most useful approach toward separating the vowels from the consonants, on the whole, would seem to be to conceive of all speech sound as due to some degree and kind of obstruction which is offered to the breath. If the obstruction is entirely in the larynx, and the breath passes through the resonance chamber without audible friction, the resulting sound is a vowel. If, on the other hand, there is obstruction in the resonance chamber which causes complete or partial stoppage of the breath, and hence audible friction, the resulting sound is a consonant. Vowels are then, on the whole, open or unobstructed sounds, and consonants, on the whole, narrowed or obstructed sounds. Yet [h], the one sound entirely unobstructed (even in the larynx) and therefore completely open, is nevertheless listed among the consonants. For a demonstration of the shadowy line that

English Sounds and Their History

169

separates the vowel from the consonant, one may also consider how the sound indicated in spelling by the

symbol r may have so little friction that the consonant imperceptibly shades into the vowel.

The differentiation just suggested has, however, a sufficient degree of reality—to say nothing of the force

of common habits of thought—to commend its use here to the extent of classifying consonants and vowels sepa-

rately rather than as a single body of sounds. Con-

sonants are the more readily classified, because audible friction implies a fairly definite place and kind of obstruc-

tion that can, usually, be adequately identified. The three questions that anything like a complete description

of a consonant makes necessary are these: (1) Do the vocal cords vibrate, or do they not? (2) Is the breath

completely or only partially stopped? (3) Where and how is the final modification provided in the resonance

chamber?

The first two questions involve no special difficulty or need for elaboration. With a very little practice, it is

easy to discern whether a given consonant is voiced or not. The readiest test,4 perhaps, is to frame one's

vocal organs to pronounce a given consonant; do not make the sound audible, but observe whether there is

or is not tension in the larynx. Thus when one is about to, but does not, pronounce [b], he should be able to

feel strain in two places—the lips and the throat. The corresponding breath consonant [p] provides the same

feeling in the lips, but not in the throat. A similar contrast is produced by [d] and [t], [g] and [k], [v] and

4 Another test that is frequently recommended is to pronounce breathed and voiced sounds while one's ears are stopped; a buzzing sound should be audible in the latter case, but not in the former.

[f], [z] and [s], and so forth. This is the distinction between voice and breath. Likewise, the difference between the stop and the continuant is readily apparent. The breath is temporarily but completely checked in pronouncing such consonants as [b], [p], [d], and [k]; after its stoppage, it is allowed to burst forth in a brief explosion. These are the stops. The continuants,5 on the other hand—[s], [v], [ʒ], and so forth—are made by narrowing the aperture through which the breath emerges, without closing it entirely, so that the breath comes forth as a continuous rather than a momentary sound.

To answer the third question—that of the how and the where—implies a complete classification of consonants. Before giving this, it will be well to distinguish the two sets of terms that must be used in this twofold description: those that name the organs active in the articulation, and those that state the manner in which the articulation occurs. A simple system of terminology, for the first point, includes (with certain subdivisions) such terms as labials, dentals, palatals, and velars; for the second point, it includes—besides stops and continuants—such terms as nasals, laterals, and fricatives. To go into further detail, the labials or lip-sounds include both bilabials, or sounds in which both lips are active, and the labiodentals, or sounds formed between the

5 The term continuant is avoided by some writers because it seems to imply that only these sounds can be varied in length, whereas actually "stops" can be held as long as "continuants," and in this sense a "stop" may be a "continuant." To avoid whatever contradiction exists here, fricative more satisfactorily designates most of the sounds that are often called continuants; but fricatives do not include laterals or nasals. What is to be emphasized, at any rate, is that the distinction between stop and continuant is less absolute than the terms imply.

English Sounds and Their History

171

lower lip and the upper teeth; the dentals are more accurately subdivided into the less common (in English) pre-dentals, sounds in the production of which the tip or blade of the tongue6 touches the upper teeth, and the alveolars (or post-dentals), sounds articulated against the gums or teeth-ridge by the blade or tip of the tongue; the palatals are the sounds made by the front of the tongue against the hard palate; and the velars are those made by the back of the tongue against the soft palate. As to the other set of terms, those descriptive of the manner rather than the place of articulation, the nasals or nose-sounds are those in the production of which the soft palate is lowered so that the breath passes out through the nose only, the oral cavity being completely closed at some point; the laterals or side-sounds are those in which an obstacle in the middle of the mouth makes it necessary for the breath to emerge from the two sides rather than the center; the fricatives are those in which the narrowing of the passage above the tongue makes the breath escape with a whistling, puffing, or hissing sound.

From the points of view that have been indicated, then, the principal consonants of English may be arranged as in the following diagram.7 Where two consonants are enclosed in parentheses, the first is the breath sound and the second the voiced sound; the others, those given separately, are all voiced.

6 For phonetic purposes, the upper surface of the tongue is conceived as having three principal parts: the part opposite the gums when the tongue is in the position of rest is called the blade, and this includes the extremity, or tip; the part opposite the hard palate is called the front, and the part opposite the soft palate is called the back.

7 This classification follows chiefly, with minor modifications and simplifications, those given by Professor Daniel Jones in The Pronunciation of English, p. 9, and An Outline of English Phonetics, p. 15.

172 The Development of Modern English

Labial Dental Palatal Velar

Bi-labial Labio-dental Pre-dental Alve-olar

Stops (p, b) (t, d) (k, g)

Con-tinu-ants Nasal m n ŋ

Lateral l

Fricative (f, v) (θ, ð) (s, z) (ʃ, ʒ) r8 j

The sounds least satisfactory to classify according to this scheme, and hence those omitted for the present, are the thoroughly ambiguous [h] and the semivowels [w] and [ʌ]. It might be added that [j] and [r], in their usual American treatment, are perhaps better described as semivowel glides (tongue-glides, however, rather than lip-glides like [w] and [ʌ]), and therefore are with some misgiving classified on our chart as fricative consonants. By a glide is meant a consonant in the articulation of which the vocal organs are continually changing; it is neither a pure stop nor a pure continuant, and it bears indeed almost as much affinity to a vowel as to a consonant. Further attention will be given to the sounds just mentioned after the other consonants have been described in some detail.

Of the phonetic symbols that have been used in the foregoing catalog, [p], [b], [t], [d], [k], [g], [m], [n], [l],

8 This symbol stands, not for the rolled or trilled r, as it does in the International Phonetic Alphabet, but for the American r described on p. 178.

English Sounds and Their History

173

[f], [v], [s], [z], [w], and [r] are familiar from their use in

the ordinary spelling of English. [j], however, has a

special use—that commonly represented in the usual

spelling by y rather than by j (which is phonetically a

double consonant, [dʒ]). The only new symbols are

[ʌ], [ŋ], [θ], [ð], [ʃ], and [ʒ]. They all stand for familiar

sounds in English pronunciation, but the first five are

represented in the conventional spelling by two symbols

each, while the last lacks a corresponding letter or letters

in the alphabet. For phonetic purposes, therefore, new

symbols have had to be created. [ʌ] corresponds to the

spelling wh (in Old English, hw), and [ŋ] to ng (where the

two letters represent a single sound, as in sing). [θ]

and [ð] are, respectively, the voiceless and voiced values

of the digraph th—the first as in thigh, the second as

in thy. [ʃ] is conventionally represented by sh, as in

shall, wash, and so forth; its voiced equivalent is [ʒ],

the sound occurring in such words as pleasure and azure.

The consonants will now be reviewed in somewhat

greater detail and grouped, in part at least, according

to another terminology. We may begin with the series

of stops. [p] is the voiceless lip-stop and [b] the corre-

sponding voiced sound. [t] is the voiceless alveolar

point-stop (that is, the stop formed by the point or tip

of the tongue touching the upper gums); [d] is the corre-

sponding voiced sound. In French and German, [t]

and [d] are dentals in the stricter sense, for in their

formation the point of the tongue touches the teeth

rather than the gums. The remaining stops are [k]

and [g], respectively the voiceless and the voiced back

sounds—that is to say, sounds articulated by the back

of the tongue against the soft palate. This is the mean-

ing of “velar,” and accounts for the position of [k] and

[g] in our chart. A more complete classification, however, would have to discriminate between the initial consonants of coop and keep, and similarly between those of goose and geese. Coop and goose include the sounds we have described (back soft-palatal stops), but the corresponding sounds in keep and geese are not velar but hard-palatal—in other words, the stops articulated by the front of the tongue against the hard palate. The explanation, of course, is that the precise value of [k] or [g] is determined by its vocalic surroundings: the sound shifts from back to front with the accompanying vowel. In our alphabet, therefore, [k] and [g] are not completely unambiguous symbols.

The nasals will be referred to next. All are voiced and continuant. [m] is a lip, [n] a point, and [ŋ]—variable like [g]—a front or a back, sound. In each case, to put it differently, there is a closure of the oral cavity: by the lips, by the point of the tongue against the gums, by the front of the tongue against the hard palate, or by the back of the tongue against the soft palate.9 While the oral cavity is thus completely stopped, the breath is allowed to escape through the nose. It will be noticed also that the conformation of the oral cavity for [m] is like that for [b], and that [n] and [d] and [ŋ] and [g] likewise are parallel. This explains why it is that when one has a cold, the nasals tend to be replaced by the corresponding (voiced) stops; one says [spriŋ] instead of [spriŋ], and (bai do:z] instead of [mai no:z], for the reason that the stoppage of the nasal passages transforms what should be nasals into oral and stopped sounds. As has been said, the symbol [ŋ] is necessarily used in order

9 For the last differentiation, compare the [ŋ] in sing with that in song.

English Sounds and Their History

175

to represent a single sound by a single symbol. The conventional spelling is unsatisfactory, too, in that the ng may sometimes stand for [ng] as well as [ŋ]; the former is represented in the pronunciation of finger, and the latter in that of singer.

It will be noticed also that "dropping the g"11 is far from being the explanation of what happens when "going hunting" is transformed into "goin' huntin'"; this is of course not the dropping of a consonant, but the substitution of one nasal for another ([n] for [ŋ]).

The one lateral consonant is [l]. It may be described as a voiced alveolar point continuant. Normally, in its production, the tongue-tip is placed against the roots of the upper teeth and the blade against the gums just above them, with a channel left on each side of the tongue for the passage of the breath. With some speakers, however, the breath emerges from one side only. [l] is a consonant that sometimes approximates a vowel, for its formation varies with the vocalic setting. Thus, while the point continues to touch the gums, the back of the tongue may assume various vowel positions. This is particularly noticeable when [l] is final, as in such a word as people. The pronunciation, sometimes called "dark l," and represented by the spelling peepul, illustrates the u-type of [l]—the sound, that is to say, that approximates the vowel [ʌ]. The vocalic character of [l] is also exemplified in the fact that it may carry a syllable: phoneticians recognize a syllabic [l] in battle, little, and

10 In German, however, Finger and Singer are alike in this respect.

11 In Galsworthy’s Maid in Waiting, this tendency in Aunt Em’s speech is explained through her having been at a school "where an undropped g was worse than a dropped h"; it is neatly put but phonetically inadmissible.

176 The Development of Modern English

apple which is parallel to syllabic [n] and [r]12 in such

words as heaven and dinner. What makes this possible

is of course the fact that [l] has both comparatively high

sonority and relatively little friction.

The fricatives form the most numerous group of

consonants. In the pronunciation of [f] and [v], the

lower lip and upper teeth are brought together, and the

breath emerges either through the openings between the

teeth or, if this is impossible, through a narrow opening

between the edges of the teeth and the lip. [f] may

therefore be called the voiceless lip-teeth continuant

and [v] the corresponding voiced sound. In [θ] and [ð],

the point of contact is between the upper teeth and either

the tip or the blade of the tongue. The air stream passes

either through a narrow opening between the tongue and

teeth or through the interstices of the teeth. [θ], then,

may be described as the voiceless blade-teeth (or point-

interdental) continuant; [ð] is the voiced equivalent.

It will be observed that the formation of [f] and [θ] and of

[v] and [ð] is much alike: in both pairs of sounds, the

column of air emerges either from between the upper

teeth or through a narrow passage between the teeth-

edges and either the lower lip or the tongue blade. This

accounts for the frequent substitution by children of [f]

for [θ] and [v] for [ð], in words like think and with.

The gums as well as the teeth are used in the formation

of the sibilant13 fricatives [s], [z], [ʃ], and [ʒ]. There are

more ways than one to form each of these sounds, but

12 E. Kruisinga (Handbook of Present-Day English, Part I, p. 23)

distinguishes a group of sounds that form “an intermediate class between

consonants and vowels.” These sounds—[m, n, ŋ, l, r]—he calls by the

suggestive but impossibly ugly term “vowellikes.”

13 Or “hissing.”

English Sounds and Their History 177

the attempt will be made to describe the typical formation

of each. For [s], the tongue-tip approaches but does not

touch the upper teeth, while the edges of the blade are in

light contact with the gums; the hissing sound is produced

by the air being expelled in a narrow stream above the

tongue and between the edges of the lower and the upper

teeth. [s] may be called the voiceless blade-gum con-

tinuant, and [z] the voiced equivalent. The other

sibilants, [ʃ] and [ʒ], may be labelled similarly, the

former of course being the voiceless and the latter the

voiced sound. In the formation of [ʃ] and [ʒ], however,

the tongue is lower and farther back than for [s] and [z],

and its point is more blunted. The channel over the

blade for the emission of breath is therefore broader and

deeper, and the breath emerges with a “hushing” rather

than a “hissing” sound. The two sounds [ʃ] and [ʒ]

are represented in conventional spelling in a variety of

ways, the one appearing, for example, in shoe, sugar,

chandelier, mission, motion, and anxious, and the other

in delusion, azure, leisure, and massage.

No other consonant presents so much difficulty and

so much variety as that represented in the conventional

spelling by the symbol r. This symbol stands, in differ-

ent words or in the pronunciation of different localities,

for sounds that vary from one made in the back of the

mouth by the tapping of the uvula (the “Northumbrian

burr”) to one that has so little friction that it has no

consonantal quality at all. It frequently stands, indeed,

for a sound that is altogether lost in Modern English.

Hence the statement, always puzzling at first to Ameri-

can readers, but appearing in all English books on pho-

netics, that “in Standard English r is not sounded when

followed by another consonant, or when final (unless

178 The Development of Modern English

followed by a word beginning with a vowel).” By “Standard English,” English phoneticians mean, of

course, that of Southern Englishmen, or, as Professor Jones14 more explicitly puts it, “that of Southern Eng-

lishmen who have been educated at the great public boarding schools.” Incidentally, an American may

perhaps be permitted to express wonder that the pronunciation of so small a proportion of the English-

speaking world can still be “standard.”

The symbol r, then, stands for a great number of

sounds, variously designated as “burred,” “rolled,” “trilled,” “semi-trilled,” “vocalic,” “glide,” and “lost.”

For the present purpose, since we are concerned primarily with the pronunciation of present-day American English,

it will be necessary to distinguish only between the two varieties of r that are common in general American

practice. The more energetic and distinct sound, that designated here as [r], may be called the point-gum

voiced fricative. In its production, the tongue-point lightly touches the upper gums; friction is created when

the breath forces itself through this point of contact. This sound is present in the usual American pronuncia-

tion of r both initially and when preceded by a consonant—as in such words as ring and draw—and also when

occurring between vowels, as in American. The other sound, [.I], is a fainter echo of [r] and tends to replace

it in the speech of many Americans. It may fairly be doubted whether [.I] is accurately to be designated a

consonant at all. In its production, the tongue point merely rises toward the gums, and both the point of

contact and the friction incidental to [r] are absent. It

14 Outline of English Phonetics, p. vi.

English Sounds and Their History

179

seems, however, proper to recognize this weak r as present

in the general American pronunciation of such words as

bar, board, and never ([baːɹ], [bɔːɹd], [nevəɹ]), and to

distinguish between a weakening of the r and its complete

disappearance; in the practice of many Southerners and

New Englanders, for example, the last three words are

made [baː], [bɔːd], and [nɛvə].

The trilled (rolled) or semi-trilled r is so infrequent in

American speech as to make it unnecessary to use a

special symbol for it. Englishmen, however, ordinarily

use the semi-trilled r (accomplished by a single tap of

the tongue against the hard palate just above the gums)

when r is intervocalic; thus, the word American as

pronounced by an Englishman is different from the same

word as spoken by an American in the single trill given

to the r as well as in the sharper enunciation of the e.

On the other hand, Englishmen commonly omit the

final r (as in car, star, and far) which is made [ɝ] by Amer-

icans. A prolonged trill or roll of the r is characteristic

of Scotch pronunciation, but the prolonged r heard in

the speech of Americans of the Middle West and the

North Central states is not usually a trilled but a hard-

palatal or back r,15 made by curving back the point of the

tongue toward the middle of the hard palate. One last

variety of the American r may be noted: that associated

with the speech of New York City and its imme-

diate vicinity. This, called by Kipling the “unre-

producible slid r,”16 and especially noticeable in such

words as earl, girl, first, and hurt, is usually indicated

(though not quite accurately) by dialect writers through

15 Cf. Krapp, Pronunciation of Standard English in America, New York

(Oxford), 1919, pp. 23 and 24.

16 “My Sunday at Home.”

180 The Development of Modern English

a substitution of the diphthong oi, phonetically [ɔɪ], for

r (and the vowel preceding).

A final point about the consonant r is perhaps worth

adding—its affinity for l and its tendency to interchange

with it. The symbols l and r do not perhaps, at first

sight, seem to stand for sounds that should be likely to

be confused, but we have seen that [r] and [l] are pro-

duced in very similar ways. It is not to be wondered at,

then, that those to whom one or both sounds are strange

should have difficulty with them. The tendency of the

Chinaman in speaking English to transform all r's into

l's—Fliday for Friday, 'Melican for American—is well

known; to a Japanese ear, too, it is evident that l and

r are identical.17 Helen Keller, learning to pronounce

not, as is generally done, by imitating the spoken sounds,

but rather by learning how the vocal organs frame the

sounds, found that to discriminate between l and r

was the final difficulty to be overcome.18 The phe-

nomenon that has been called the "shift of liquids"19—

the tendency of r and l to interchange—is a helpful

principle in etymological study, for it at once relates

words that otherwise seem quite distinct from one

another. A striking instance is the group of words for

"star" in various European languages. Star is at once

recognized as a cognate of Latin stella if one has this

principle in mind—and even more surely if one recalls

that the Middle English form is sterre (O. E. steorra).

Then there is the Greek ἀστήρ (familiar in English in

17 Louise Taylor, "Across the Pacific—Cerebrally," Atlantic Monthly,

Vol. 149, No. 3 (March, 1932), p. 349.

18 Story of My Life, New York (Doubleday), 1903.

19 Liquid is a term sometimes applied to l and r alone, and sometimes

to other vowel-like consonants.

English Sounds and Their History

181

the name of the flower aster and in such compounds as disaster and astronomy); the Latin form, too, is familiar

in English words like stellar and constellation; the French étoile is from an Old French esteile, still used in English

as the proper name Estelle side by side with the Latin Stella; the German form retains the r: Stern. Other

instances are the Modern French word for “nightingale,” rossignol, as compared with the Old French lossignol;

and the English word colonel, still spelled with the l though pronounced with the r. Finally, variant forms

of proper names, such as Mary and Molly, Katharine and Kathleen, may be mentioned.

Like [r] (in some of its pronunciations), [j] is a tongue-glide which approximates a vocalic sound. It is, there-

fore, frequently called a semivowel. In its production, the breath passes through a narrow aperture between the

front of the tongue and the hard palate. Since the vocal cords vibrate, the sound may be called a voiced front

fricative. [j] is most clearly a fricative consonant when it is followed by the high front vowel [i:], as in the word

ye. Often it seems to be merely a manner of beginning a vowel20 and to take its color entirely from the particular

vowel which it precedes. It is evident, for example, that before a back vowel, as in yawl [jɔ:l], the [j] is not in the

same position as in ye [ji:]: it is not only further back, but it is also definitely less a fricative consonant, for it has the

open, unobstructed character of the vowel.

There remain (besides the ambiguous [h]) the sounds [w] and [ʌ], often also called, like [j], semivowels. The

former, particularly, is obviously very like the vowel [u:], since in its formation, the back of the tongue is

20 Kenyon, American Pronunciation, Ann Arbor (Wahr), 1924, p. 78.

182 The Development of Modern English

raised toward the soft palate and the vocal cords vibrate. [w] is frequently described as the voiced lip-rounded back fricative, but it is doubtful whether it should be described as a continuant consonant in view of the fact that to prolong it is impossible. What is prolonged in the sounds associated with the spelling w is the following vowel (since w, like y, occurs only before vowels). It is, perhaps, more accurate to think of it as a lip-glide than as a fricative. [w] is the sound indicated not only by the spelling w but also by wh, so far as the usual pronunciation of Southern England is concerned. Jespersen says, indeed, that "a great many 'good speakers' always pronounce [w] and look upon [hw] as harsh or dialectal."21

It is clear, however, that there is a counter-tendency, in England as well as in America, to cultivate the pronunciation [M] in most words spelled with wh. The present writer's observation does not, however, coincide with that recorded in most American works on phonetics—to the effect that [M] is the usual American pronunciation in most words spelled with wh. The schools, indeed, make a gallant effort (just how laudable must be a matter of opinion) to introduce [M], but their efforts seem on the whole crowned with no more success, in natural American pronunciation, than somewhat similar efforts in favor of [ju] as opposed to [u:].22

21 A. Modern English Grammar, Part I, 13-51. Wyld remarks, Short History of English, New York (Dutton), 1927, p. 33, that the other sound, [M], has been "introduced comparatively recently—within the last thirty and forty years—apparently through Scotch and Irish influence, backed up by the spelling. Many excellent speakers of Standard English never use the sound at all." Jones observes (Outline of English Phonetics, p. 4) that it is more commonly heard, in the educated speech of Southern England, from ladies than from men; and the editors of the Concise Oxford Dictionary believe that, in the same region, it is chiefly confined to "purists in pronunciation."

22 Cf. p. 236.

English Sounds and Their History

183

The symbol [M], used here in deference to very general practice,23 would seem to imply that the sound it designates is closely allied to [w]—different from it, in fact, only in being voiceless rather than voiced. This is, indeed, the usual description of it: the voiceless lip-rounded back fricative. It may, however, be questioned whether the symbol [hw] (used, for example, by Jesperson and by Kenyon) is not a more accurate designation of the sound under consideration. Professor Kenyon’s description of the sound24 as beginning “with a quick increase in the force of the breath identical with that found in the sound of [h]” seems to the present writer entirely just. If so, it is accurately characterized by the double symbol [hw] as being essentially [w] with an initial increase in stress. Conceivably, however—especially in unusually careful or emphatic enunciation—the sound may be the voiceless fricative which [M] is intended to designate. Its use in England, as has been said, is often deprecated as artificial and pedantic,25 and it is also perhaps less commonly to be heard in unforced American speech than is usually admitted.

As to [h], the last of the consonantal symbols to be discussed, Professor Krapp’s observation26 that “there are as many varieties of [h] as there are varieties of vowels before which it may stand” may be quoted. Evidently [h] is always voiceless rather than voiced; however, considering its distinctly “open” character, its usual designation as “fricative” is somewhat unsatisfactory.

23 In the works, for example, of Jones, Ripman, Grant, Krapp, and Mrs. Aiken.

24 American Pronunciation, p. 76.

25 Ripman observes, “It is taught by professors of elocution, and is therefore commonly heard at recitals and also at amateur theatricals.” The Sounds of Spoken English, New York (Dutton), 1914, p. 39.

26 Pronunciation of Standard English in America, p. 15.

184 The Development of Modern English

What friction there is varies from that made by the narrowing of the glottis by partly closing the vocal cords, to that made by the sudden pressure of the front of the tongue against the hard palate. The former sound is that which occurs in the ejaculatory Ha!; the only difference between the sound of this exclamation and that of Ah! is that in the one the glottis at first is half-open and later closed, while in the other (the vowel [a:]) the glottis is kept closed. In the glottal [h], the tongue is in the normal or neutral position, but an entirely different sound (though here designated by the same symbol) may be made by the contact of the tongue with the hard palate. This is the front fricative [h] that occurs in he [hi:]. Still more strongly fricative is the initial sound in such words as hue, huge, human; here it approximates, indeed, the ch of German ich. Much more commonly, however, [h] is merely unmodified breath, appropriately designated "rough breathing" by the Greeks, and perceptible only—especially medially, in such a word as behind—by the increased stress with which the syllable following it is pronounced. [h] may usually, therefore, be thought of less as a distinct sound than as a manner of beginning another sound.

Vowels lack the definite points of contact or friction that characterize consonants and are therefore much more difficult to classify and enumerate. To the question of how many vowels exist in the pronunciation of Modern English, the answer must vary according to the type of pronunciation described and, what is even more important, according to the fineness of the distinctions made. Thus, Sweet27 has evolved no less than seventy-

27 Sweet, Henry, Primer of Phonetics (3rd ed.), and Sounds of English, London (Oxford).

English Sounds and Their History 185

two distinct phonetic symbols to stand for a corresponding number of vocalic sounds. It is clear that a much

less elaborate system of notation will have to suffice for our present purpose, and that only vowels that are

readily perceptible as distinct sounds, and sounds current in contemporary American speech, can possibly be given

a place. The readiest starting point for our discussion of vowels will be the five symbols used in the conventional

alphabet: a, e, i, o, and u. The meaning of these letters as phonetic symbols is, however, not at all accurately

suggested by their use in Modern English spelling. The values-the so-called 'Continental' ones-are much

nearer those associated with these letters as used in Old English, or in German, Italian, or Latin. In other

words, [ɑː] is the sound of the first vowel in father;

[eː] is the sound of the vowel in ate; [iː] that in seen;

[oː] that in note; and [uː] that in boot. To these, the symbols of two other fundamental sounds for English

pronunciation should be added at once: [æ] for the vowel in hat, and [ɔː] for the vowel in awe. These seven sounds

may now be associated with their positions and their formation in the mouth.

The four types of classification most readily applied to vowel sounds are high-low, front-back, tense-slack,

rounded-unrounded. These terms are in large measure self-explanatory, but a word may be added as to the

application of the several distinctions. Raising and lowering the lower jaw accounts for differences in height;

thus, [iː] and [uː] are pronounced with the jaws nearly closed, while [eː] and [oː] require a wider opening, and

[æ] and [ɔː] still a wider one. The first pair may therefore be called the high sounds, the second the mid-

sounds, and the third the low sounds. The same pairs

186 The Development of Modern English

of vowels may be used to illustrate the front-back distinction: when [i:] and [u:] are pronounced in succession, it will be observed that the tongue is extended forward for the former, and retracted, with its tip sunk and its back raised, for the latter; similar differences hold between [e:] and [ɔ:] and between [æ] and [ɔ:].

The tongue positions thus indicate that [i:], [e:], and [æ] are front vowels, and [u:]; [ɔ:], and [ɔ:] back vowels; the point is that in the first group the tongue-front is the highest part, and in the second the tongue-back.

There are also central, or mixed, vowels, made with an intermediate part of the tongue in the highest position.

To these we shall return in a moment. Of the seven sounds enumerated in the preceding paragraph, the one most doubtfully classified is [ɑ:]; it has been disputed whether this vowel, in English pronunciation, is properly classified as mid-back or low-back, or even whether it is a back vowel at all.28

In addition to high-low and front-back, two other types of classification have been mentioned and will be referred to briefly. The rounded-unrounded distinction gives no special difficulty.

The most unmistakably rounded vowels are, in order, [u:], [ʊ] (to be defined presently), and [ɔ:]—all back and either high or middle vowels. [ɔ:] is either rounded or unrounded in American speech, while the other sounds, both those already referred to and those to be referred to, including the front, the mixed, and some of the back sounds, are

28 Cf. Noël-Armfield, G., General Phonetics (4th ed., 1931), p. 31. On the other hand, Mrs. Aiken classifies it as not only back, but low, as compared to low-mid [ɔ] and [ɔ:] (English Present and Past, p. 146).

To the present writer, the sound appears to be lower, but also slightly more of a front sound, than [ɔ:]. This position agrees with that assigned to [ɑ:] by Kenyon (p. 85) and by Jones (Outline of English Phonetics; p. 21).

English Sounds and Their History 187

unrounded. The final differentiation, that between

the tense and slack (or lax) sounds, depends on the degree

of muscular tension of the tongue. In this book, the

colon is used to designate vowels that are tense, which

are usually also the "long" vowels, as that rather

unsatisfactory term is generally understood. It will

be convenient to define most of the symbols that have

not yet been explained, in enumerating the various

"tense-slack" pairs of vowels. Thus, the tense [i:]

has its slack, and slightly lower, counterpart in [I],

as in hit; and a similar difference in tensity and position

holds between the tense [e:] and the slack [ɛ], as in set.

So too in the back of the mouth, the [u:] of soon is

both more tense and higher than the [ʊ] of good. There

is no necessity of using the two symbols [o:] and [o],

since in American speech, at any rate, the vowel seems

the same in quality and equally tense, whether in an

accented syllable or not. [ɔ:] and [ɔ], however, are

perceptibly different, the former being tense and nor-

mally occurring only in stressed syllables, while the latter

is the slack sound, regularly shorter than [ɔ:], which

occurs in such words as autumnal and auspicious.

Another tense vowel that must be mentioned is the long

[ɛ:], which occurs only before r, in such words as there,

fare, and so forth; it is to be distinguished from [e:] not

only historically but also physiologically, since it is a

slightly lower front sound. There are also these slack

vowels: [ʌ], as in hut; [a], the intermediate or compromise

sound of the vowels in such words as calf, dance, and

path as heard in an American pronunciation that avoids

both [ɑ:] and [æ]29; and [ə], the obscure or colorless

29 For a discussion of this sound, which probably occurs much less

frequently than is generally supposed, see p. 234.

188 The Development of Modern English

sound that is very generally current in the modern

pronunciation of unstressed syllables (as in the first

of about and the second of China). Finally, one other

tense vowel should be added: the [ʌ:] of “r-less” speech,30

in such words as bird and hurt ([bʌ:d] and [hʌ:t]).

We may now place the vowel sounds we have enumer-

ated in a chart indicating their positions relative to each

other and to the part of the oral cavity in which they are

made. Tense vowels (which are usually also “long”

vowels) are designated, as has been said, by the colon.

It must be understood, however, that there is no una-

nimity of opinion among phonologists as to several of

the positions indicated here.

So far we have assumed that vowels are simple and

single sounds. In practice, however, and more fre-

quently in the pronunciation of the South of England

(as opposed both to Northern English and to American

pronunciation), this is by no means the invariable rule.

It is perhaps not too much to say, indeed, that in the

pronunciation of Modern English in general, all the

30 This is almost identical with the sound often represented by [ɜ],

a symbol which is therefore omitted here in the interest of simplicity.

31 See the previous paragraph.

English Sounds and Their History 189

"long" vowels tend to become diphthongs, completely or in part. The distinction is this: a pure vowel is made with one breath impulse and with the resonance chamber in one position; a diphthong is made with one breath impulse but with an alteration of the resonance chamber in the process. There are a number of partial diphthongs, then, especially in Southern English pronunciation, which consist of a pure vowel sound followed by a much weaker vowel glide. These occur in such words as day, home, hear, and four ([deɪ, hoʊm, hiə, fɔə]). A curious illustration of this tendency in Modern English pronunciation is the advice sometimes given to German students of English to pronounce such a word as home as if it had two distinct syllables [ho: + ʊm], in order to counteract the Germans' tendency to use the vowel sound natural to them, the pure [o:] only.

The full diphthongs of Modern English, then—those in which the two vocalic elements are unmistakable—are these four:32 [aɪ], [aʊ], [ju], and [ɔɪ]. Partial diphthongs ([oʊ] and [eɪ]) are also sometimes heard in the United States, and regularly in Southern England, in such words as go, snow, day, and play. It remains for us to define the full diphthongs somewhat more completely. [aɪ]33 is the familiar sound, commonly but quite erroneously termed "long i," which occurs

32 The colon is omitted, as it is also in the diphthongal sounds just mentioned, for the sake of simplifying the symbols; but a more accurate notation would include it. In three of the four "full" diphthongs to which reference is made (and also in [eɪ] and [oʊ]), the sign of tenseness would come after the first vowel, since all these are "falling" diphthongs; that is, diphthongs in which the beginning rather than the end is stressed. The exception is [ju], more accurately [ju:]—a "rising" diphthong.

33 Krapp (Pronunciation of Standard English in America and The English Language in America) prefers the symbol [aɪ].

190 The Development of Modern English

in such words as ice, high, buy, and right. The fact that

it is a diphthong rather than a single vowel sound is

demonstrated by an attempt to prolong it; if this is

done, only the second element, the [ɪ] (which in this

case becomes [i:]) is likely to be prolonged. [aʊ] is

anothersound of wide occurrence, appearing in such words

as house, bough, and now. [ju] is perhaps not strictly

to be described as a diphthong—though it tends to

alternate, in American speech, with the true (and falling)

diphthong [ɪu]—but practical convenience, as well as

the fact that the [j] is only slightly consonantal, argues

for its inclusion here. It is the sound that occurs in

such words as pew, music, view, and beauty. [ɔɪ] is

the falling diphthong that occurs in words spelled with

oi or oy, such as boy, boil, and join.

When we turn from a description and classification of

the sounds of contemporary English to a review of the

history of English sounds, two questions may very

well be asked: How do we know that the sounds have

changed? and How can we be confident of what the

changes are? Since the invention of the phonograph

and the “talkie,” we have, it is clear, means of perma-

nently recording the pronunciation of a particular time

and place (or individual); but it is equally evident that

before the late nineteenth century, any such infallible

and mechanical methods were entirely lacking. There

are, however, certain other ways of reproducing the

pronunciation of an earlier time with some degree of

accuracy. One way is through a study of the rhymes

of the poets, and other uses of like sounds, such as puns

and word plays. Another is through investigating old

records that preserve the spelling of the illiterate. Still

another is through comparing English sounds with those

English Sounds and Their History 191

of related languages, in which pronunciation has been comparatively stable. Illustration of the use of these methods will be given, but it may be observed parenthetically that it is too much to assume that the pronunciation of earlier periods can be reproduced with absolute fidelity. A great deal of work, for example, has been done on the pronunciation of Chaucer, and much has been learned about it; nevertheless, the result of it all is justly summarized by Professor Manly in these terms:34 “If Chaucer could hear a good student read his poetry, the pronunciation would probably seem like that of a foreigner, but he would, we hope, be able to understand what he heard.” We are not, then, to be too sanguine about the infallibility, for every detail of pronunciation, of the methods that have been employed.

The study of the rhymes of poetry clearly has wide applicability. If a sound that occurs at the end of a line can be identified with certainty, it follows, granting that the poet is a conscientious craftsman, that the sound of the final syllable of the rhyming line must be parallel. Changes in pronunciation since the eighteenth century, for example, are to be observed in a study of the rhymes of such poets as Pope and Cowper. The neat anticlimax of this couplet

Here thou, great Anna! whom three realms obey,

Dost sometimes counsel take—and sometimes tea35

falls deplorably flat in present-day pronunciation, but the lines must once have rhymed perfectly. If, as can be demonstrated, the word obey has remained virtually unchanged in its sounds, tea has altered greatly. The

34 Introduction to J. M. Manly’s edition of the Canterbury Tales, New York (Holt), 1928, p. 89.

35 Rape of the Lock, Canto III, ll. 7 and 8.

192 The Development of Modern English

old pronunciation, evidently, was not [ti:] but [te:]; the comparative method substantiates the conclusion, [te:] being also the pronunciation of the word in such languages as French and German and in certain dialectal forms of English—Irish, for example. A similar change is shown by the lines of Cowper's well-known hymn,

God moves in a mysterious way, . . . , He plants his footsteps in the sea . . . ,

for the vowel of sea has gone from [æ] through [e:] to [i:], and in the eighteenth century it was still in the middle stage. In Pope's famous attack on Addison, these lines occur as a rhyming couplet:36

Dreading e'en fools; by flatterers besieged, And so obliging that he ne'er obliged.

Once more, a surviving dialectal pronunciation helps to make it clear that the sounds of obliged were once parallel with those of besieged. The reverse of this situation is often given by the conventional rhymes called "eye rhymes"—the employment, that is to say, of words spelled alike but sounded differently, where the poet's only justification is the fact that at some former time the words really did rhyme. It would seem, however, a rather illegitimate extension of poetic privilege to rhyme such words as star and war for no other reason than that such a seventeenth-century poet as Henry Vaughan did so. Lines like

My soul, there is a country Far beyond the stars, Where stands a wingèd sentry All skilful in the wars

indicate a parallelism in sounds that no longer exists.

36 Epistle to Dr. Arbuthnot, ll. 207 and 208.

English Sounds and Their History

193

The second method of inferring older pronunciation, an investigation of records that preserve the spelling of the illiterate, or, more accurately, the semi-literate, perhaps needs no special demonstration. It is obvious that one whose spelling is untrammelled by education will write English words differently, and more phonetically, than one handicapped—from this point of view—by notions of orthodoxy. Professor Krapp has thrown a great deal of light upon seventeenth- and eighteenth-century American pronunciation by studying old town records. Such spellings as arst for asked and clack for clerk are better clues to pronunciations in actual use than are statements of grammarians and lexicographers, who are likely to be concerned with how words should be pronounced rather than how they are pronounced.

The use of a third method, a comparison of English sounds with those of related languages, has already been suggested. It is this that is back of the reconstruction of the earliest sounds of English, and that leads to the assumption that, since the Teutonic languages generally have the values designated as the "Continental" ones, English once possessed these sounds too. Our spelling, from most points of view so bad, is sometimes of real service here in indicating earlier (and common Teutonic) pronunciations. Thus oo is now usually interpreted as [u:], [ʊ], or [ʌ], and ee as [i:] or [ɪ]; but the modern spelling clearly indicates that the earlier sounds were [o:] and [e:]. If, then, as has frequently been said, we spell words not as they are now pronounced but as they were formerly pronounced, the practice has at any rate the incidental value of throwing occasional light on older habits of pronunciation. At the least, it makes clear certain general directions of phonetic change.

194 The Development of Modern English

In sketching the development of English sounds, the consonants, which have changed on the whole relatively little, may be disposed of first. Old English spelling is sufficiently phonetic to justify taking its symbols as our starting point. The following letters, then, have normally been used, from Old English times on, to designate consonantal sounds that are quite unambiguous and stable: b, p, d, t, l, m, n, w, and x.37 Two other letters of the Old English alphabet—f and s—were used to designate both voiced and voiceless continuants; the Old English alphabet did not include v or z, except occasionally in late loan-words or foreign proper names, but these symbols were eventually borrowed to designate the voiced equivalents of voiceless f and s, which then in turn tended to be used (not quite consistently) to designate the voiceless sounds only. In addition to v and z, mention should be made of q and j as later additions to the original alphabet. The former, q, occurring in Modern English always in the combination qu, is, of course, due to French influence on the spelling38; no phonetic changes are involved in the rewriting of Old English cwic (living), for example, as quick. The symbol k, too, is rare in Old English, its sound being usually designated by c. Modern English j represents an evolution that will be referred to later, when Old English c and g are discussed.

Four other Old English symbols may be grouped together: h, r, ð, and p. The first two, h and r, stand for sounds that have tended to be weakened or to disappear. Old English h was always sounded, and when

37 In Old English, as in Modern, x = ks.

38 The combination qu was borrowed at a time when its pronunciation in French was [kw], not, as it became later, [k].

English Sounds and Their History 195

closing a syllable, it was a palatal or a velar continuant,

like German ch in ich or in Nacht. Later French borrow-

ings had a weak h which has remained silent in a few

words like heir, honest, and hour, and sometimes (espe-

cially in the pronunciation of England) in such words as

humble, humor, hotel, and human. The r of Old English, a

trilled or perhaps sometimes a “curved” or “cerebral”

sound, has, as we have already seen, often been made

much less energetic, and frequently disappears alto-

gether, in modern pronunciation. As to the other two

symbols, ð and þ, it is quite clear that Old English

had both the voiceless and the voiced th that Modern

English retains, though ð and þ were apparently used

indifferently for both sounds. When intervocalic, ð (þ)

would inevitably become voiced, like f and s (cf.: oven

[<O. E. ofen], risen, and other).

The two remaining symbols of the Old English alpha-

bet, c and g, were both used quite variously. The one,

c, while never having the [s] sound it occasionally has

in Modern English, varied, as it does in Modern English,

from a front to a back stop according to its vocalic

environment, and was even, in certain positions, a palatal

continuant. This is the origin of Modern English ch

or tch [tʃ] of chew (<O. E. ceowan), drench (<O. E.

drencean), witch (<O. E. wicce), and pitch (<O. E.

pic). In the last three words, the fully dentalized

pronunciation, [tʃ], was probably used even in early

Old English. Palatalization also took place in the Old

English combination sc, this becoming, before the end

of the Old English period, the simple continuant [ʃ].

Thus, Old English sceall, scur, wascan, and fisc became

respectively shall, shower, wash, and fish. A few words

may perhaps be added on somewhat similar changes in

196 The Development of Modern English

later stages of the language. In early Modern English,

the [S] sound developed also in words that had formerly

been pronounced as [s] or [t + j] such as mansion,

ocean, patience, and nation. Later still, the correspond-

ing voiced continuant [3] replaced what had formerly

been the double sound [z + j]. There is no satisfactory

symbol for this consonant in modern spelling, nothing

corresponding to the sh for [S], but the sound itself is

familiar enough, being found in such words as measure,

usury, azure, and osier.

The letter g of the Old English alphabet was, like c,

used with a variety of meanings. It was likewise (as

is still the so-called "hard g" of Modern English) both

a front and a back stop, and it was also, before front

vowels and in certain other positions, a continuant [j].

The last pronunciation explains such Modern English

words as year, young, and yet from the Old English

gēar, geong, and gīet: before a front vowel, the consonant

became a palatal continuant. In the later development

of the prefix ge, the consonantal y gave way to the vocalic;

Old English gecleopod, for example, became yclept,

and a parallel change took place in Modern English

enough (<O. E. genōh; cf. Ger. genug), Modern English

handiwork (<O. E. handgeweorc), and so forth. The

combination cg (or double g) probably had in Old

English still another sound—[d3]; Old English ecg and

its Modern English descendant, edge, are therefore

phonetically identical. The similar spelling of such other

words as bridge, cudgel, hedge, and ridge indicates a parallel

history, but it should be noted too that the [d3] sound

of Modern English is quite as generally represented by

the symbol j, a borrowed letter which became familiar

in Middle English in Romance loan-words such as joy

and join, and by the “soft” g. The sound [dʒ], however

represented in spelling, is thus very common in Modern

English in both native and borrowed words. Although

in Modern French g before front vowels is pronounced

[ʒ], rather than [dʒ] as in Old French, even recent Eng-

lish borrowings from French tend to alter the [ʒ] to

[dʒ]; observe the popular tendency to make the final

consonants of massage and garage like those of message

and courage.

To summarize the foregoing account, it may be pointed

out that to the consonantal sounds of Old English, the

later development of the language has added only two

single consonants, sh [ʃ] and zh [ʒ], and one double

consonant, ch [tʃ]. The new symbols added—v, z, q,

and j—involved no additional sounds, though the last

(phonetically [dʒ]) has become, as we have seen, a far

commoner sound in Modern English than it was in Old

English, when it was represented by the combination

cg (or gg). Moreover, the loss of ð and þ from the later

alphabet did not mean a disappearance of sounds from

the language, since both voiced and voiceless th are

familiar in Modern English. It is nevertheless a fact

that our consonantal sounds have not been quite so

stable as this summary might indicate, for Old English

sounds have sometimes been dropped altogether, and

certain alterations have taken place on a large scale.

The details of some of these changes will now be referred

to.

The most conspicuous losses are palatal and velar

consonantal sounds, frequent in Old English but rare in

Modern English. The virtual disappearance of these

values for the symbols g and h in Modern English helps

to make present pronunciation strikingly different from

198 The Development of Modern English

that of Old English (and German). The consonant r,

too, as we have seen, has become in modern English and

American pronunciation something quite different from

what it was. Initial k (in Old English, usually c) when

preceding n has likewise changed its character: the spell-

ing of words like knight, knave, knit, knee, and knead

reminds us that here is a frequent loss. Initial w in

the combination wr in modern spelling (in words like

wrath, wretch, wrist, write, and wrong) carries a similar

suggestion. Parallel and, to modern habits of pronun-

ciation, likewise difficult combinations are the hl and the

hr in such Old English words as hlāf, hlūd, and hlystan,

and hring, hrōf, and hrōst. In these cases, however,

Modern English has dropped the initial consonant in

spelling as well as in pronunciation: the modern forms are

loaf, loud, and listen, and ring, roof, and roost.

The change of g and h, referred to above, is a type of

the phenomenon known as vocalization, the change of a

consonant into a vowel. Loss of palatal or velar con-

tinuants in such words as Old English heah, ðurh, niht,

and bōg gives Modern English high, through, night, and

bough, in which the spelling still testifies to the former

consonantal quality of the final sounds. This vocali-

zation was not complete until the early Modern English

period. In some words, moreover, it did not take place

at all, though even here the palatal-velar continuant

was lost; in laugh and tough, for example, it was trans-

formed into the labiodental continuant. Next to g

and h, the consonants most likely to be vocalized, or

completely assimilated39 to the vowels, were naturally

39 There is no hard and fast line of demarcation between “vocaliza-

tion” and “assimilation,” any more than there is between the vowel and

the consonant. In ordinary use, however, “assimilation” emphasizes

English Sounds and Their History

199

those most like vowels; the semivowels w and y and the

"vowellikes" m, n, l, and r have therefore been particu-

larly subject to this process. We have already referred

to the vocalization of the Old English prefix ge [>ye >y

(i̯)]; as for w, to compare the spelling and pronunciation

of Modern English two, who (<O. E. hwā), sword, and

answer is to suggest a like development. The disappear-

ance of the [l] sound in such words as walk, talk, folk,

half, calf, and calm is a kindred process, as is the wide-

spread loss of r in present-day pronunciation. Not only

l and r but also m and n are, very often, sufficiently

vocalized in present pronunciation to enable them to

carry a syllable; they are endowed, in other words, with

some of the quality of a vowel. In such words as heaven,

able, bottom, and maker, the syllabic or vocalized pronun-

ciation of the final consonant has probably been in vogue

for the whole of the Modern English period.

A number of the other changes to which English

consonants have been subject must be referred to briefly.

Assimilation in the strict sense—that is, absorption of

one consonant in another of like quality—is illustrated

in the loss of n in words like damn, hymn, autumn, and

column, and in others like mill (<O. E. myln) and kiln

(so spelled, but usually pronounced [kil]). Here the

neighboring m or l has absorbed the n. A phenomenon

that has operated much more widely is that known as

"voicing." In late Middle or early Modern English

many formerly voiceless consonants were transformed

into the corresponding voiced sounds. This develop-

ment affected consonants (not initial) with voiced sur-

roundings and following an unstressed or weakly stressed

the disappearance, and "vocalization" the replacing, of a sound; and of

course only "assimilation" can be applied to vowels.

200 The Development of Modern English

vowel. Through it, these changes took place: [f] > [v],

[θ] > [ð], [s] > [z], [tʃ] > [dʒ], and, later than the

others, [ks] > [gz]. A few illustrations follow:40 of

(now with [v], but in Middle English, with [f]), the

stressed and voiceless form being retained in adverbial

off; the, they,41 and so forth (in Middle English, with

[θ], but now with [ð]); the s ending of plural and posses-

sive singular nouns, and of third person singulars of

verbs now frequently pronounced [z]; Modern English

knowledge, cabbage, and partridge, as compared with

Middle English knowleche, cabach, and partriche. The

reverse process, "unvoicing," is also to be found. The

dental preterit ed has been regularly unvoiced whenever

the loss of its vowel has brought the final d into juxta-

position with a voiceless consonant; thus the final

[d] of walked and voiced itself has been unvoiced to

[t], while that of wagged and raised has remained [d].

Perhaps the next most frequent type of unvoicing is

the change of final [ð] to [θ]. This also is frequently

due to the loss of a vowel; in Modern English youth,

earth, and both, the final [θ] replaces an older [ð], a sound

inevitable in Middle English youthe, earthe, and bothe.

Other instances of consonantal change might be added.

In various ways, consonants have been added, dropped,

and interchanged. The vagaries of r, some of which

have already been mentioned, will serve to illustrate sev-

eral of these possibilities. We have already referred to

the frequent interchange of l and r—"shift of liquids"42—

40 For a full account, see Jespersen, Modern English Grammar, Part I,

6.511–6.90.

41 The voicing of these initial consonants seems due to their frequent

position between vowels in the sentence.

42 See p. 180.

English Sounds and Their History

201

when one of these consonants occurs but once in a word;

there is also another type of interchange, due appar-

ently to "dissimilation," when two r's occur in a word.

Modern English pilgrim, laurel, purple, and marble,

for example, correspond to Latin peregrinum, laurarium,

purpur, and marmor, and the substitution of l for one of

the r's has taken place either in Old French or in Middle

English. "Metathesis," or change of position within

the word, not infrequently affects r—the neighboring

vowel has changed positions with r in these words,

sometimes very early in their history: bird, fresh, grass,

third, thirty, through, wright, and wrought. The addition

of an unetymological r is seen in bridegroom (cf. Ger.

Bräutigam and O. E. guma, "man"), cartridge (cf.

Fr. cartouche), and hoarse (<O. E. hās); a noteworthy

case of its loss, which here occurred before the end of the

Old English period, is Modern English speak and speech

(cf. early O. E. spreċan, and Ger. sprechen). Other

consonants than r may, of course, be subject to similar

wanderings: an additional or "excrescent" n, for example,

is to be observed in nightingale (cf. Ger. Nachtigal),

passenger (cf. Fr. passager), messenger (still messager

in Caxton); moreover, as we have already seen,43 the

dropping of an n, especially when final, is very common

indeed. Similarly, d has been frequently added—sound

(cf. Chaucer's soun), thunder (cf. Ger. Donner, Fr.

tonnerre), hoarhound (cf. M. E. horehoune), and the

vulgar drownd for drown; it has likewise occasionally

been dropped: lawn (<M. E. laund) and scan

(<scand).

43 P. 199.

202 The Development of Modern English

The foregoing account of English consonants should not be interpreted as indicating that the consonants have greatly altered in the development of the language from Old to Modern English. On the contrary, though it has been possible to point out many alterations in individual words and to see certain larger shiftings that apply to groups of words, it may still be affirmed that the consonantal system of Modern English is not widely different from that of Old English. The contrast with the vowels is most striking: the symbols for the consonants still mean, for the most part, exactly what they meant in Old English, but the symbols for the vowels have been given an utterly different meaning. Moreover, almost all the changes that have taken place in the use of the consonants have been readily accompanied by new spellings of the words in which the changes occurred; but the alterations in the vowels of English words have been so great that spelling has lagged far behind pronunciation and is now hopelessly unphonetic. All that can be done here in sketching the history of vowel-sounds is to indicate certain general trends.

There is, fortunately, one very clear trend that has had most far-reaching consequences upon the development of our vowels. The “great vowel-shift,” as Jespersen has termed it,44 consists of the surprisingly uniform raising of all Old English long, or tense, vowels; the necessary and quite regular exception to the upward direction of the shift is that the two vowels that were already high—[i:] and [u:]—split into the diphthongs [aɪ] and [aʊ]. The general course of the shift may be represented as shown at the top of page 203.

44 Modern English Grammar, Part I, 8.11.

English Sounds and Their History

203

Front

i:

e:

ɛ:

æ

Back

u:

o:

ɔ:

a:

If a cause may be assigned for this phenomenon, it would seem to be the very human tendency of laziness in operation–pronouncing the vowel with as little effort as possible, and hence with a smaller and smaller opening of the lips. To give typical illustrations of the results, by pairing Old English words and their Modern English descendants,

[æ] > [i:] sǣ,45 sea; clǣne, clean; hǣlan, heal

[e:] > [i:] grēne, green; tēð, teeth; sēcan, seek

[i:] > [aɪ] īs, ice; wīf, wife; hȳdan, hide

[a:] > [o:] bān, bone; bāt, boat; wrāt, wrote

[o:] > [u:] tōð, tooth; sōna, soon; dōm, doom

[u:] > [aʊ] mūð, mouth; hūs, house; ūt, out

These shifts will be detailed somewhat more particularly.

The change from [æ] or [ɛ:], through [e:], to [i:] affected not only Old English words spelled with ǣ (in Mercian, generally ē), but those with ēa as well; for usually this diphthong simplified to the single vowel [ɛ:]. This accounts for the present pronunciation of words like beacon, east, and stream, in which the spelling

45 The Old English vowel æ was apparently not unlike the vowel of Modern English hat or man lengthened; however, it would perhaps be more nearly accurate to designate it [ɛ:], and to give the shift referred to here as from [e:] to [i:].

still preserves the Old English diphthong ēa. The

[ε:], or "open" e sound, as it is often called, was kept

in both classes of words until Elizabethan times. In the

seventeenth century, probably, the [ε:] was replaced

by [e:], and vowels from Old English æ were no longer

distinguished from those from Old English e. Words with

Old English ē had, in the meantime, been preserved with

the original [e:] unchanged. That the distinction

between [ε:] and [e:], or "open" and "close" e, was a real

one in late Middle English is illustrated by the rhymes

of Chaucer, in which the two types of e are serupulously

distinguished. The Old English diphthong ēo also

developed, for the most part, like close e; for it ordinarily

lost its second element—hence, Modern English deep,

thief, and be (<O. E. dēop, þēof, and bēon). Throughout

the Middle English period, both "open" e (<O. E. æ or

ēa) and "close" e (<O. E. ē and ēo) were designated

indifferently by e or ee; but, as remarked above, the two

sounds were distinct until well within the modern period.

The subsequent development, beginning, it is supposed,

in the eighteenth century, was a raising to [i:] of all the

varieties of e that have been mentioned. In late Modern

English, especially in the pronunciation of England, this

[i:] tends to be diphthongized; the sound may be indi-

cated by the combination [ɪj]. It is also rather fre-

quently shortened to [ɪ]: been, for example, is in American

pronunciation [bɪn] more frequently than [bi:n], and

league and eagle tend to be [lɪg] and [ɪgl].

In the back of the mouth, the shift [ɑ:] to [ɔ:] to [o:]

to [u:] parallels what has just been noted for the front

vowels. There is the difference, however, that the usual

raising of the lowest vowel [ɑ:] has been to [ɔ:] only.

The highest vowel, [u:], has been developed ordinarily from

English Sounds and Their History

205

an Old English [o:], not from [a:] or [ɔ:]—whereas the highest front vowel, [i:], has been developed both from

Old English [ε:] (or [æ]), and from Old English [e:]. [a:] must regularly have gone through the middle stage

[ɔ:] before [o:] was reached; [ɔ:] was, accordingly, the usual Middle English pronunciation of the vowel in

words that had been spelled in ā in Old English and were now o. This was the so-called “open” o, which

existed side by side with the “close” o with no difference in the Middle English spelling, just as the “open”

and “close” e were indistinguishable. The “close” o was, of course, the vowel of many Old English words

spelled with ō, which persisted unchanged in Middle English. The Modern English history of the o’s, how-

ever, is different from that of the e’s in that the Old English [ɔ:] (M. E. [ɔ:]) changed to [o:] and remained

that in Modern English; whereas the Old English [o:] (M. E. [o:]) was raised to [u:] in early Modern

English. The spelling has retained the o or, more commonly, the double o, of Middle English, in spite of

the fact that that vowel is now of utterly different quality. Thus, we have the absurdity of spellings like

food and goose, in which [u:] is designated by double o. This [u:] is frequently “shortened” (more accurately,

made lax and slightly lowered) in the standard pronunciation of such words as good and hook, and in a widely

used pronunciation of room, roof, root, and the like. Occasionally, it is lowered and “fronted” to the “mid-

mixed” position of [ʌ]; such a word as blood has evidently gone through all these stages: [blo:d], [blu:d], [blud],

[blʌd].

Vowels that were already in Old English in the highest position—the front [i:] and the back [u:]—split into

206 The Development of Modern English

diphthongs that have become the Modern English [aɪ] and [aʊ]. Though the point is disputed, Jespersen⁴⁶ argues that it was at this upper end that the great vowel shift began; in other words, that it was the diphthongization of [i:] and [u:] that allowed [e:] and [o:] to move upwards and occupy the places formerly held by [i:] and [u:]. Words with the [i:] sound included words spelled in Old English in y as well as in i, for the originally distinct vowel designated by y—rounded like the German ü—had conformed, within the Old English period, to that designated by i. The pure [i:] probably became diphthongal before the end of the Middle English period and assumed the present pronunciation perhaps in the seventeenth century. [u:] seems to have developed in a parallel way. It is interesting to observe that in this case, diphthongization has not taken place in the Scotch and Northern English dialects: house is still [hu:s] in Scotch pronunciation, exactly as it was in Old English.

The lax, or “short,” vowels did not as a rule take part in the upward shift. Their development has less regularity and simplicity about it for the additional reason that lax vowels have been more affected by their consonantal environment. Only the most conspicuous and uniform developments can, therefore, be summarized here.

Old English a (as distinguished from ā) had, apparently, the sound of the a in artistic, as distinguished from the a of art. This sound remained throughout the Middle English period, but was finally fronted, probably in the seventeenth century, so that we have such Modern English words as sand, thank, have, and saddle, from

⁴⁵ Modern English Grammar, Part I, 8.12-8.22.

English Sounds and Their History 207

Old English sand, panc, habban, and sadol. Some of these originally short a's followed another development: they first lengthened in Middle English to [ɑː], after Old English ā had become [ɔː]; and much later, in Modern English times, they went up and forward to [eː].47 This is the history of an exceedingly large number of Modern English monosyllables, in which the spelling still bears witness to the original a of the stem; ape (<O. E. apa), shake (<O. E. sceacian), knave (<O. E. cnapa), name (<O. E. nama), and bathe (<O. E. baðian) are a few examples. The present pronunciation of this sound tends to introduce a glide, or [ɪ] sound, after the [eː]. Another low vowel of Old English, the æ of such words as sæt, glæd, æx, and bæt, has undergone a still more curious history. It was evidently a front vowel in Old English, became a back vowel (short [ɒ]) in Middle English, and went forward again to [æ] in Modern English.48 The Modern English spelling is the same as the Middle English, but the pronunciation (except for certain Northern dialects, where [a] is still to be heard) has gone back to the [æ] of Old English.

Two other short vowels of Old English—the front vowels [ɛ] and [ɪ]—have survived with little or no change in Middle English and Modern English whenever they remain short and occur in stressed syllables. The spelling is usually unchanged: ebb (<O. E. ebba), help (<O. E. helpan), length (<O. E. lengð), step (<O. E. steppan), west (<O. E. west), and so forth. Before certain nasal combinations, however, [ɛ] has been replaced with [ɪ];

47 Probably first to [æː], then to [ɛː], and at last to [eː] or [eɪ].

48 It is so curious a history that its truth has been doubted. For a defense of the theory, see Jespersen, op. cit., Part I, 8.63; for a recent expression of doubt, see Mrs. Aiken, Why English Sounds Change, p. 93.

thus, Old English sengean has become Modern English singe, Old English hlence Modern English link, and Old English streng Modern English string.

In one or two instances, the modern spelling retains the e in spite of the fact that the sound has long since become [i]; this is true of England (<O. E. Englalond).

Modern [l], however, usually represents the very regular survival of Old English short i; a few illustrations are drink (O. E. drincan), fish (O. E. fisc), if (O. E. gif), in (O. E. in), rib (O. E. rib), and this (O. E. ðis).

In Old English, the symbol y, originally representing a sound distinct from that represented by i, finally came to stand for the same sound.

As a result, short y has fallen in with short i, just as long ȳ fell in with long ī; and as ȳ took on, with ī, the sound [ai] in Modern English, so y, with i, remained unchanged, except for the spelling.

Modern English fill (O. E. fyllan), pit (O. E. pytt), and sin (O. E. synn) represent this survival.

There remain the short back vowels o and u. Short o of Old English was phonetically [ɔ], not a shortened value of [o:].

This sound has persisted with little or no change to the present time, so far as the “standard” pronunciation of England is concerned.

Words in point are ox, fox, pot, got, and so forth.

Except for New England, however, the usual American pronunciation is not [ɔ]. but a shortened form of [ɑ:]—that is to say, the first vowel of artistic.

Curiously enough, it is also an American tendency to lengthen the Old and Modern English [ɔ] to [ɔ:] (instead of introducing the [ɑ] sound) when the [ɔ] is followed by [ŋ] and sometimes when it is followed by [d], [g], or [k].

Americans pronounce long with a vowel that is perceptibly longer than that used by Englishmen, and sometimes god and hog in a way that dialect writers render

gawd and hawg. Old English and Middle English short

u ([ʊ]) has usually developed into Modern English [ʌ], a

"mid-mixed" sound that is without the rounding char-

acteristic of both [uː] and [ʊ]. This is illustrated in

love (<O. E. lufian), sun (<O. E. sunne), wonder

(<O. E. wurdor), and so on. The [ʊ], however, is

frequently preserved in the neighborhood of labial

consonants; Modern English full and pull illustrate this

possibility.

The original diphthongs of Old English have been

lost, as such—usually through the emergence of the

predominating element as a single vowel; we have seen,

for example, how ēa and ēo conformed with ǣ ([ɛː]) and

ē ([eː]), respectively, and developed in the same way.

It might be added that the corresponding short ea and eo

simplified in a parallel way. Moreover, of the several

new diphthongs that were developed in Middle English,

all but two have been made monophthongs in Modern

English. Thus, [aʊ] in such a Middle English word as

faught, and the closely akin [ɔʊ] of thought, have altered

to [ɔː]; while in other cases a diphthong made up of

[ɔː] + [ʊ]—as in soule (soul)—has simplified to [oː].

Likewise, Modern English [eː] often represents a sound

that was a diphthong in Middle English; in fact, two

separate diphthongs are represented here, for the spell-

ings ai and ei stand for sounds that are rhymed in Chaucer

but that in earlier Middle English were distinct. Day,

may, sail, play, and way are of this group. The excep-

tions to the process of monophthongization are the

borrowed oi ([ɔɪ])—in French loan-words like join, joy,

and toil; and [ju], which has a double source—Middle

English eu (or ew) and iu. The eu was originally [ɛʊ],

found in words like few and new; and the iu was originally

210 The Development of Modern English

[Iu], found in words like humor and rude. Even here, monophthongization has taken place to some extent in Modern English; for while few and humor are invariably pronounced with [ju], most Americans substitute the simple vowel [u:] in new. It may be in order here to emphasize once more the idea that what we commonly call the simple or pure vowels of Modern English are really not monophthongs in practice. Modern English lacks, in a strict sense, the pure vowels of Old English or of Modern German; Modern English [e:], [i:], [o:], and [u:] all tend, in degrees varying in different parts of the English-speaking world, to be diphthongal.

Reference has been made, so far, to vowels that occur in stressed syllables and in native words. It has seemed unnecessary to complicate this survey by treating of borrowed words, for the reason that, in the main, the vowels of such words follow the development of native words of like sound. But it should not be imagined that anything like the complete history of even the native vowels has been given. For one thing, the separation into three categories-long (or tense) pure vowels, short (or lax) pure vowels, and diphthongs-is somewhat unreal. Something has been said of the interchange of single vowels and diphthongs; and much more might be said of the lengthening of short vowels and the shortening of long ones. More striking than such changes, however, and so far-reaching in its consequence that the entire character of English pronunciation has, through it, been altered, is the general obscuring that has occurred in the vowels of unaccented syllables. The matter has already been touched on49 as one of the fundamental

49 See p. 29.

English Sounds and Their History

211

causes of inflectional leveling. Here it may be briefly recapitulated from the angle of the phonetic changes involved.

Before the end of the Old English period, every unaccented a, e, o, and u tended to become a vowel that was spelled e and was presumably pronounced [ə]. These are typical illustrations: O. E. oxa, M. E. oxe; O. E. foda, M. E. fode; O. E. heorte, M. E. herte;50 O. E. swēte, M. E. swete; O. E. nacod, M. E. naked; O. E. wundor, M. E. wonder; O. E. lufu, M. E. love; and O. E. sunu, M. E. sune. Unaccented i ([I]), however, tended to remain unchanged through Middle and in Modern English; note the second syllable of English, shilling, and evil (though here the l has become syllabic, and the [I] consequently lost, in colloquial pronunciation). The [I] has also usually been preserved in Middle English borrowings from French: peril, punish, service. When e ([ə]) was final in Middle English—as it is in most of the above illustrations—it was eventually lost; hence the emergence of the modern monosyllables ox, food, heart, sweet, love,51 and son. Certain endings in which [ə] was followed by a consonant—especially the possessive and plural -es and the preterit -ed—have regularly syncopated, so that here too [ə] is lost. The exceptions perhaps should be noted: a noun ending in a sibilant, such as [s], [z], [ʃ], or [tʃ], does not syncopate its ending (busses, vases, rushes, ditches); nor does a verb ending in the dental [t] or [d] (wetted, wedded). Equally as important as the frequent loss of [ə] that has just been noticed is the tendency, beginning in the sixteenth century and

50 In this pair and the next, the final e is in O. E. [ɛ], and in M. E. [ə].

51 The fact that the spelling of this word (and many others) retained final -e did not, of course, affect its pronunciation.

212 The Development of Modern English

universal in the pronunciation of present-day English,

to level all vowels unprotected by accent to either [ə] or [ɪ]. The general rule may be stated thus: Every

Modern English vowel occurring in a syllable that receives neither primary nor secondary stress approximates one or the other of the relaxed vowels [ə] and [ɪ].

This trend is combated by the conservatism of dictionaries, which has done something to perpetuate

spelling-pronunciations; but in natural, colloquial speech, it has full sway. It will be our task in the following

chapter to survey the extent of its victory.

References for Further Reading

Aiken, Janet R., English, Present and Past, New York (Ronald), 1930.

, Why English Sounds Change, New York (Ronald), 1929.

Emerson, O. F., History of the English Language, New York (Macmillan), 1894.

Flom, G. T., Introductory Old English Grammar and Reader, New York (Heath), 1930.

Jespersen, Otto, A Modern English Grammar, Part I, “Sounds and Spellings,” Heidelberg (Winter), 3rd ed., 1922.

, Essentials of English Grammar, New York (Holt), 1933.

, and Pedersen, H., Phonetic Transcription and Notation, London (Oxford), 1926.

Jones, Daniel, Outline of English Phonetics, New York (Stechert), 3rd ed., 1932.

, The Pronunciation of English, Cambridge (Eng.), 2nd ed., 1914.

Kruisinga, E., A Handbook of Present Day English, Part I, “Sounds,” Utrecht (Kemink en Zoon), 4th ed., 1925.

Moore, Samuel, Historical Outlines of English Phonology and Morphology, Ann Arbor (Wahr), 1925.

Noël-Armfield, G., General Phonetics, Cambridge (Eng.), 3rd ed., 1924.

English Sounds and Their History

213

Soames, Laura, Introduction to English, French, and German Phonetics, London (Macmillan), 3rd ed., 1913.

Wright, Joseph, An Old English Grammar, Oxford (Milford), 3rd ed., 1925.

Wyld, H. C., A History of Modern Colloquial English, New York (Dutton), 1920.

———, Historical Study of the Mother Tongue, London (Murray), 1906.

CHAPTER VII

Contemporary Pronunciation

WHEN we turn from a survey of English sounds as they have existed in the past to an attempt to state the pronunciation of the present, we are at once confronted by several difficult questions. Is present usage, all over the English-speaking world, sufficiently uniform to admit of satisfactory generalization? If there are important deviations, can any one brand of pronunciation claim a special place as "correct" or "standard" speech? If it cannot, is there any such thing as a single standard of pronunciation? Is it necessary to recognize several standards, and more than one standard within the United States? How valid, after all, is the assumption of standard pronunciation, in any sense of the term? These—and others like them—are questions that are inevitably and quite legitimately asked, and they are questions that the present chapter must cope with. It will be more convenient, however, to postpone any definite formulation of their possible answers until the end of the chapter, and in the meantime to set forth some of the facts on which these answers should be based. We shall be concerned, then, first to examine the status of contemporary pronunciation, and later to inquire what theory of pronunciation or attitude toward pronunciation is required by the facts as they exist. It must be recognized, however, that statement of facts and theoretical generalization from the existing facts

214

Contemporary Pronunciation

215

cannot be completely separated; nor does it seem desirable that they should be.

The logical way to begin a discussion of the contemporary pronunciation of English would seem to be to define what is often conceived of as its central focal point, the "Received Standard" of England. This conception of a peculiarly favored brand of pronunciation, to be heard in its purity from the lips of educated residents of Southern England, is one that is, quite naturally, set forth with special enthusiasm by scholars who are native to, or who have acquired, this type of speech; it must be admitted, moreover, that it is a conception that has a certain degree of reality. Professor Daniel Jones's definition of this brand of pronunciation has already been quoted.1 It may be useful to repeat it here, in another and a fuller form:2 "that most usually heard in everyday speech in the families of Southern English persons whose men-folk have been educated at the great public boarding schools." Professor Jones, that is to say, conceives of "Received Standard" as a geographical as well as a class dialect. Another eminent English phonetician, Professor H. C. Wyld—to whom, indeed, we owe the term "Received Standard"—formulates the idea somewhat differently: " . . . it is not confined to any locality, nor associated in any one's mind with any special geographical area . . . Received Standard is spoken, within certain social boundaries, with an extraordinary degree of uniformity, all over the country."3 To Professor Wyld, then, the essence of

1 P. 178.

2 Quoted from the Introduction (p. vii) to Jones's English Pronouncing Dictionary.

3 History of Modern Colloquial English, p. 2. In the latest (1932) edition of his Outline of English Phonetics, Professor Jones has sub-

216 The Development of Modern English

"Received Standard" is that it is exclusively a class dialect—not, as those who follow Professor Jones believe, a class dialect spoken within a geographically limited territory. The two definitions cannot of course be completely reconciled, though it might be added that Wyld too believes that "perhaps the main factor in this singular degree of uniformity is the custom of sending youths from certain social strata to the great public schools. If we were to say that Received English at the present day is Public School English, we should not be far wrong."4 Both definitions have been quoted in order to indicate two somewhat different (and almost equally influential) English conceptions of what is meant by "Standard English." To the present writer, it appears that, although the term is Wyld's, the definition of "Received Standard" as formulated by Professor Jones is the more useful and the more accurate—in other words, that the term serves best to describe a form of speech that is limited geographically as well as socially, and limited geographically almost entirely to London and what are known as the "Home Counties."

It is clear, at any rate, that American pronunciation has far more in common with the speech of Northern England than with that of Southern England. It has been more than once observed that it is Northern rather than Southern British pronunciation that corresponds with the General American brand,5 and that it is

stituted the phrase "Received Pronunciation" for what was formerly "Standard English Pronunciation."

4 Op. cit., p. 3.

5 For an illustration of this, see the three versions of the fable "Wind and Sun" as given in the pamphlet "The Principles of the International Phonetic Association," pp. 20–22 (a supplement to Maître Phonétique, Sept.–Oct., 1912). These are phonetic transcriptions of Southern

Contemporary Pronunciation

217

with Eastern (or New England) speech that the “Received Standard” has its affiliations. The former correspondence, it may be said, applies generally to the speech of

Lowland Scotland as well as to that of the north of England. American radio audiences, for example, have observed that Ramsay Macdonald sounds very much

more like an American than like a “typical,” or Southern, Englishman. The most obvious connecting link between Northern British and General American pronunciation is

undoubtedly their common preservation of both final r and r before a consonant; alike in “Received Standard” and New England speech, these r's are vocalized or lost.

The speech of the Northern Englishman further resembles that of the average American in its use of [æ] or [a] as the vowel of such words as ask, laugh, and class; the

Londoner and the New Englander alike substitute [ɔ:].6 In other points of difference7 between London and Northern English—for example, the greater tendency

in Southern English to make long vowels such as [e:] and [o:] into diphthongs—it is again very clear that the characteristic American pronunciation is considerably

nearer to that of the north of England than to that of the metropolis.

Before proceeding to generalize about American pronunciation or to indicate its three main varieties, as

British English, Northern British English, and American English pronunciation. The transcriptions are also reproduced in Krapp, Pronunciation of Standard English in America, pp. 208-211, and in Ripman, Sounds of Spoken English, pp. 135-138.

6 On the other hand, Southern and American English sometimes agree on [æ] where Northern British uses [a] or [ɑ:]; this is true, for example, in the transcriptions cited above, for the words traveler and wrapt.

7 Many of these are pointed to by R. E. Zachrisson in his paper “Northern English or London English as the Standard Pronunciation,” Anglia (1914), Vol. XXXVIII, pp. 405-432.

218 The Development of Modern English

they are now almost invariably recognized, it seems advisable to formulate a certain attitude with reference to its very existence. How does it come about that cultivated American pronunciation is at all different from cultivated British pronunciation? That it is different—for it seems impossible to avoid dogmatism here—must be uncompromisingly maintained. The Wyld theory of a “Received Standard” for all of England is carried to what seems to the present writer its logical and at the same time its fatal conclusion in Palmer, Martin, and Blandford’s Dictionary of English Pronunciation with American Variants. These writers paraphrase Wyld’s dicta thus, their additions being the parenthetical matter.8

It is not any more the English of London, as it is sometimes mistakenly maintained, than it is of York (or we might add: New York), or Exeter, or Cirencester, or Oxford, or Chester, or Leicester (we might add: Harvard, or Chicago, or New Orleans, or San Francisco). In each and all of these places, and in many others throughout the length and breadth of England (we might replace “England” by “the English-speaking world”), Received Standard is spoken among the same kind of people, and it is spoken everywhere, allowing for individual idiosyncrasies, to all intents and purposes, in precisely the same way.

Now this, to put it bluntly, is simply nonsense. What we are at present concerned with, however, is the attitude that seems to be implied, by such a view, toward “American variants.” These are evidently to be deprecated as unfortunate strayings, wilful or unwitting, from a theoretical standard for the English-speaking world: the

8 “Introductory,” p. xvii. Reprinted by permission of the D. Appleton-Century Co.

Contemporary Pronunciation

219

"Received Standard" or "public school" English.

Historically they are nothing of the sort. For one thing,

at the time of the American Revolution, "Received

Standard" was certainly both less influential and less

uniform than it is now. Discarding, then, the idea that

cultured American pronunciation is either a form of, or a

departure from, the "Received Standard," our original

question remains: What is it, and how does it come about

that it differs from that of England? More specifically,

since to most minds there are obvious and clear-cut

differences—Are its characteristic marks of native

development or are they survivals from older speech

habits in the English language?

This query may be best answered in the words of the

scholar most qualified to speak on such a point, Professor

Krapp: "If one did not fear to affirm a universal positive,

one might say that in every case the distinctive features

of American pronunciation have been but survivals from

older usages which were, and in some instances still

are, to be found in some dialect or other of the speech of

England."9 These survivals of older usage are undoubt-

edly better preserved in the South than elsewhere, but

there is every reason to suppose, as indeed Professor

Krapp also indicates,10 that the speech of the seventeenth-

century Colonists of New England was much more like

that of their Virginian contemporaries than the speech of

their descendants today is like the present Southern

speech. The Colonists of both New England and Vir-

ginia were chiefly from London and the adjoining

Midland and Southern counties, and both groups

9 The English Language in America, Vol. II, p. 28.

10 P. 34.

unquestionably spoke some more or less dialectal variant of Southern English. New England, more subject to influences from both England and other parts of America, has retained fewer seventeenth- or eighteenth-century peculiarities than has Virginia; but in a number of ways-most conspicuously in its treatment of r—Southern American speech is still associated with that of New England in its resemblance to the speech of Southern England. Later colonizers of other parts of America, coming from other parts of the British Isles than London and the adjacent territory, have presumably been influential in transplanting a speech more like that of the North than that of the South of England. It remains, however, something of a paradox than the earlier colonization of America was accomplished by the speakers of Southern English, however dialectal, and that nevertheless the General American pronunciation is demonstrably closer to the Northern British standard.

It is necessary that such a term as “General American” be specifically defined. As has already been suggested, the convention of dividing American pronunciation into three great zones or areas is now well established. This does not, of course, imply that within any one zone there is absolute uniformity; in one sense, it is arguable that there are as many brands of pronunciation as there are individual speakers—a proposition advanced in Shaw’s Pygmalion. Evidently, within the limits of a single state—Virginia is a notable example—there may be striking variations. At the same time, the concept of a threefold division among the speakers of American English is as familiar to the man in the street as to the phonetician; and as it is commonly accepted by both, it

Contemporary Pronunciation 221

would be superfluous to argue for it. It means in general that American speakers use either Southern English, the English of New England, or finally that which for lack of a better name has been called "Western" or "General." The more precise geographical limits of the three areas may conveniently be indicated, as a preliminary to inquiring what these three types of speech have in common and also wherein their tendencies chiefly differ.

The Southern type of American English is marked off with fair precision, though its limits are not precisely coterminous with those of the (politically) "Solid South." States on the borderlines do not always fall satisfactorily into one group or the other; but Maryland, Arkansas, southern Missouri, and eastern Texas would seem to belong linguistically to the South, while Delaware, West Virginia, and the remaining sections of Missouri and Texas fall within the General (or Western) zone. The Southern influence has also been observed to spread somewhat north of the Ohio River and to affect parts of Ohio, Indiana, and Illinois. New England, though markedly divided within itself, falls as a whole in a separate category, sometimes called the "Eastern"; and this category is often held to include a definite section (social rather than geographical) of New York City. The rest of the United States, comprising the Middle Atlantic states and the Middle and the Far West generally, uses the General or Western type of pronunciation. "Linguistically speaking," as Professor Kurath observes,11 "the ‘West’ begins on the Hudson

11 "American Pronunciation," Society for Pure English, Tract No. XXX, p. 286.

222 The Development of Modern English

River, and in some respects as far east as the Con-

necticut River." To avoid the absurdity of calling the

speech of New York and New Jersey "Western,"

however, it would seem both more convenient and more

accurate to designate this great area, containing roughly

two-thirds of the population of the Union, as that in

which the "General" manner of speech is to be found.

There is every reason to regard this pronunciation as the

typical or characteristic American speech.

We have, then, to investigate the distinguishing features

of American pronunciation, and especially that of the

"General" type, as opposed to British pronunciation.

It will be understood, of course, that the intention is to

compare pronunciations on parallel cultural levels; that

is, not to oppose the speech of the educated in one

country to that of the uneducated in the other, but rather

to contrast the pronunciation of British and American

speakers of the educated class. First of all is the elusive

matter of intonation. The general rule is unquestion-

ably that American speech shows far less variation in

musical pitch or timbre than British. Between the

speech of the rural Middle-Westerner, however, and the

speech of the British speaker of "Received Standard,"

this difference is very much more considerable than that

between the speech of the urban resident of the Middle

Atlantic states and the speech of the inhabitant of

Northern England. Granting that in its cadences, as in

other matters, General American speech is more similar

to Northern than to Southern English, the average

differences are still likely to be perceptible. Taken at

the two extremes, even within the limits of the speech of

the educated, they are very striking indeed. The

English voice is more rapid, more vivacious, more excited

Contemporary Pronunciation 223

—it may seem to an American listener—over nothing in particular. A casual remark about the weather12 is delivered with more apparent gusto than an American puts into the most heartfelt utterance. The words may be the same, but the tune is entirely different. On the other hand, the American voice is slow, less varied both in pitch and in rhythm, and consequently much more monotonously droning.

From the eighteenth century on, American speech (especially that of rural New England) has been labelled by both British and native observers as characterized by a “drawl” and a “twang.” The “drawl,” differing in degree, to some extent, in time and place, is evidently the extreme of what has just been described as a characteristic feature of General American utterance—its level, flat, and deliberate character. What is meant by “twang,” in all probability, is the quality of nasality that often associates itself, again more frequently in the rural districts (both of New England and the Middle West) than in the cities, with the “drawl.” “The word ‘twang,’” Professor Krapp points out, “often means nothing more than merely ‘flavor.’”13 It is in fact, we may add, confused with “tang.” Nevertheless, the term “nasal twang” undoubtedly has a real meaning, and describes an ingredient in American speech that once was more conspicuous and more general than it is now. Of course, the habit of nasalizing vowels is no more of indigenous growth in America than are other distinguishing features of American speech. It once flourished

12 Professor F. N. Scott’s little paper “The Father Tongue,” The Standard of American Speech, New York (Allyn), 1926, pp. 16–18, illustrates this difference rather neatly.

13 The English Language in America, Vol. II, p. 14.

224 The Development of Modern English

in England (and to some extent still does), as Krapp indicates,14 and was satirized as a Puritan failing in Hudibras and elsewhere.15

Connected with the “drawl” rather than the “twang” is a further conspicuous difference between American speech and British. It seems to the present writer to be going too far to say of American speech, as has recently been said,16 “The stresses are levelled out and all syllables have almost the same value . . . .” But if this is exaggeration, it is the exaggeration of a real truth, and an important one for us if we are to distinguish between American and British habits of speech. The point is that British English has more and more tended to emphasize the heavy stress accent characteristic of the Teutonic languages in general; it has obscured, and often telescoped, unaccented vowels. American English has, in many classes of words, done this to a less degree, and has indeed preserved distinctive vowel qualities through a habit of secondary accent that has been discarded in British English. This is particularly noticeable in polysyllables ending in -ary, -ery, and -ory. These are Middle English borrowings, for the most part, of Old French words that are ultimately Latin. The main accent in Old French was ordinarily on what is now the penultimate syllable, but in Middle English it shifted to the fourth rather than the second syllable from the

14 Ibid., p. 23; and compare his reference to Ripman, Sounds of Spoken English, p. 14.

15 For references to dramatic satire directed against the Puritans’ habit of speaking “i’ th’ nose,” see Myers, A. M., Representation and Misrepresentation of the Puritan in Elizabethan Drama (Philadelphia, University of Pennsylvania, 1931), pp. 20, 36, and 85.

16 Larsen and Walker, Pronunciation: A Practical Guide to American Standards, New York (Oxford), 1930, p. 13.

end, as native words were commonly accented on the first syllable if this syllable was a root. However, a secondary accent remained on the penultimate syllable, and this secondary stress is preserved in General American can practice though it has been lost, probably within the last century and a half, in British speech.17 Our ultra-refined radio announcers who stress only the first syllable of words like dictionary, ordinary, secretary, laboratory, and cemetery, and telescope the rest are presumably either ignorant of this law or convinced of the plausibility of a “Received Standard” for the whole of the English-speaking world. It is to be regretted that the British habit is making headway in our schools as well.

If the “drawl” is to be accounted a vice of American pronunciation, surely the greater intelligibility that results from the secondary stress may fairly be reckoned a virtue that does something to compensate. The frequent testimony of foreigners, even those who have learned their English in England, that they can understand an American’s English more easily than an Englishman’s seems most plausibly explained through the slower tempo of American speech, and particularly, its habit of secondary accent. This habit, it may be added, affects certain other words besides those in -ary, -ery, and -ory already referred to. A footnote in Nicklin’s Standard English Pronunciation18 in regard to circumstances is somewhat amusing to an American: “Some speakers seem to have let themselves be influenced by Gilbert’s rhyming

17 Kenyon, American Pronunciation, pp. 156–164, has a fuller account of the matter, with lists of words in which American and British practice vary.

18 P. 72. Reprinted by permission of the Oxford University Press.

226 The Development of Modern English

of this word with chances and dances, without observing that such rhymes no more pretend to guide us to the Standard pronunciation than the rhyme with folly suggests that we should remove the emphasis from the first syllable of melancholy." As to the latter point, it is interesting that the rhyme folly and melancholy that seemed good to Milton (Il Penseroso, ll. 61-62) is still equally good in American speech, though the Englishman prefers ['melənkəli]. But to suggest that any British deviation from the practice of accenting circumstances on the first syllable only is due to a lyric in the Mikado is surely to exaggerate the influence of the rhymes of comic opera librettos. If Gilbert has contaminated British taste on this point, is he also responsible for gal (in place of girl), which, in an equally familiar lyric in Iolanthe, he rhymes with liberal?

Here, however, a word of caution may be in order. The statement that, in certain groups of words, American pronunciation retains a secondary stress that British pronunciation discards should not be interpreted as implying that American pronunciation fails to share the general trend of Modern English speech toward obscuring unaccented vowels. Indeed, with but minor qualifications, one who generalizes about American speech is quite justified in accepting for American pronunciation also the rule that Mr. Nicklin formulates on this point: "It is hardly too much to say that in nearly every word of more than one syllable all the vowels except those on which the principal stress accent falls are now sounded either as [ə] or as i in fit, and this however the word is spelt."19 Failure to observe this principle, as we shall

19 P. 71.

Contemporary Pronunciation

227

have occasion to note again, leads to many pedantic

"spelling-pronunciations," not a few of them, alas,

dignified by the authority of our dictionaries.

To return to the differentia of General American

speech. Apart from the distinctive trends that have

already been noticed—the less varied pitch, the slower

tempo, and the greater frequency of secondary stress—

there are of course several particularly conspicuous

differences in the treatment of individual vowels and

consonants. Two of these have already been mentioned,

in speaking of points of contact between General Ameri-

can and Northern British,20 and conversely, points of

difference between General American and “Received

Standard”: the sounding of the r finally and before

consonants; and the low front [æ] (or [a]) in such words

as ask, bath, dance, half, pass, and so forth. Others that

may be added are the use of the simple vowel [u:] instead

of the diphthong [Iu] or [ju] in such words as duke, duty,

new, student, tube, and so on; and the preference for the

short unrounded [ɑ] rather than the slightly rounded

short [ɔ] in words like hot, not, rock, and possible. Still

another point that is usually mentioned in such a cate-

gory as this is the following one: “In white, wheat, whale,

etc., the w is always preceded by a voiceless fricative.”21

The writer can merely record that this use of [ʍ] rather

than [w] (where the spelling is wh) frequently does not

take place, so far as his observation of General American

goes. He has asked many persons whether they pro-

nounced the wh as [ʍ] in such words as those referred to.

The answer has usually been, “I try to,” or “I’ve been

20 P. 217.

21 Kurath, op. cit., p. 284.

228 The Development of Modern English

taught to"—pretty clear evidence that [ʌ] is not in natural, unforced use.

Before discussing further the general tendencies of contemporary American pronunciation, it may be worth while to point out the several leading differences that separate Southern and New England speech from General.

Southern and New England pronunciation are in the main alike, and different from General, in their treatment of r. Both sections share the Southern British habit of obscuring or dropping the r unless it is protected by a following vowel.

It is probably more common in New England than in the South, however, to find even educated speakers interpolating a final r before a word beginning with an initial vowel, as in "idear of" or "I sawr (h)im."

The reason seems to be that the habit of using the final r when the next word begins with a vowel—“far off” and “far away” as spelled, but “fa(r) distant”—is sometimes carried over even to words that do not contain the final r, when these words are followed by an initial vowel.

In both New England and the South, also, the [ɝ] of General American in such words as bird, hurt, shirt, and worth, is replaced, as in Southern British, by [ɜː]; the tongue-point, that is to say, no longer rises in the direction of the gums, and r has lost even a semblance of consonantal quality.

As to the special marks of New England speech, most conspicuous, probably, is the natural use of the so-called broad a ([ɑː]) in such words as answer, dance, fast, half, and so forth.

Elsewhere in the United States, this [ɑː] is likely to be due to affectation or Anglomania.

The substitution of [ʊ] for [uː] in words like proof, soon, and spoon is likewise more marked in New England than in American pronunciation generally.

A third distinc-

tive mark of New England speech—especially that of

eastern New England—is the tendency to round partially

the vowel of hot, not, rock, spot, and the like, in a way

reminiscent of Southern British, rather than to use the

perceptibly different unrounded [ɑ], as is the usual Ameri-

can practice. New Englanders, too, are likely to use

this same short and slightly rounded [ɔ] sound rather than

the prolonged and fully rounded [ɔ:], common to most

American speech, in such words as coffee, cost, long, lost,

and song.

Southern American speech is markedly uniform in its

unaffected use of the diphthong [ɪu] or [ju] in duke, new,

tune, and so on, rather than the [u:] characteristic of

American speech generally. Words like door and more

are also given sounds that are quite distinctive. The

final r of course is lost, as it is in New England also,

but the vowel sound is not [ɔ:]; rather, it varies from the

more cultured combination of [o:] (somewhat lowered)

plus an obscure following vowel—the sounds indicated

by dialect writers as doah and moah—to the Negro and

lower-class white [ɔ:]—the sound indicated by do’ and

mo’. The latter pronunciation makes door and dough,

and more and mow, indistinguishable. A third pecul-

iarity, not universal in the South but noticeable in

common and sometimes in cultured speech, is the employ-

ment of the broad [ɑ:] in words in which r is preceded by

either of the diphthongs [ɑɪ] or [ɑʊ]; thus, fire and hired

are given as [fɑ:, hɑ:d] (as well as [faɪə, haɪəd]) and

our and flowers become [ɑ:, flɑ:z]. The same vowel,

somewhat shortened, frequently takes the place of

[ɑɪ] in other positions also, notably in [ɑ] and [mɑ] for

I and my, when these pronouns are unemphatic. A

prevalent misconception of Southern speech is that which

230 The Development of Modern English

attributes its more extreme divergences from General American to the influence of the Negro. A little reflection serves to indicate that the chief influence must have been in the other direction. There are after all far fewer distinctively Negro traits than Northerners usually assume: probably the Negro's difficulty with the th sounds (leading to [d] for the voiced and [f] for the voiceless sound) is the only important one not paralleled in the speech of the "poor white." Professor Kurath's words sum up the situation: "The negro has his own cadences and his unmistakable voice, but in matters of pronunciation he has simply learned and preserved dialectal peculiarities of his former masters that have been levelled out among the educated Southerners."22

From this brief account of American speech and its three chief varieties, we may turn to a more specific discussion of the general trends it exhibits today, and to an inquiry into what one's attitude toward them may reasonably be. It will not always be necessary to separate American speech, or that most typical phase of it which has been designated General American, from the broader field of contemporary English pronunciation as a whole. Only a minor qualification, for example, is desirable when what suggests itself as the very first characteristic of the present-day pronunciation of English is mentioned. This is a tendency toward relaxed pronunciation, a tendency that has its roots in the heavy Teutonic stress accent and that has resulted in over-emphasis on stressed, and underemphasis on unstressed, syllables. Obscuring and loss of formerly distinctive

22 P. 296. Several of my illustrations of Southern speech have been taken from this summary of it.

Contemporary Pronunciation

231

vowel sounds, and dropping of consonants in phonetically difficult combinations (the second often a consequence of the first), are both to be found in Old English; but they have become more marked characteristics of Middle and Modern English. The degree to which these processes have occurred is often concealed by our archaic spelling.

What is perhaps not so often perceived is that the wide extent of both is often concealed by our dictionaries as well. There will be occasion to return to this topic a little later. Here it seems advisable to indicate that pronunciations that depart from the spelling in dropping or obscuring consonantal and vocalic sounds are not necessarily to be apologized for. They simply follow a well-established trend in our language that is but slightly more marked in present-day British practice than in American. If it were not so, we might still be saying hlāfweard (loaf-ward) for lord, and daegesēage (day’s-eye) for daisy.

To some, however, it seems easy to perceive the fact of relaxed or obscured pronunciation (particularly that recorded in the spelling) for the past, but difficult to accept the continuing consequences of the same trend upon the language of the present. Such persons are behind even our dictionaries—in pronouncing a p in raspberry, for example, or a t in Christmas. Actually, these relaxed pronunciations are very common indeed, and in perfectly reputable practice: towards as a mono-syllable, forehead as [fɔrɪd], and waistcoat as [weskit] are a few random illustrations. The [ɪ] of the unaccented syllables of the two words quoted last is a sound of wide occurrence in contemporary pronunciation; it occurs in the best speech, for example, in mountain, senate, and subject, however the spelling varies and no matter whether

dictionaries may rule otherwise. Still more prevalent, of course, is the neutral [ə]; as has already been suggested,

it is almost safe to assume that every unstressed vowel approximates [ə] or [ɪ]. To put it differently, it is almost

a necessity to have either secondary accent or distributed stress, if other vowel sounds than [ə] or [ɪ] are to be

preserved. Thus, in the more rapid British pronunciation of registrar and bursar, the [ə] sound is lost in the

final syllables, which thus have the vowel of sir; but the usual American practice is to preserve the [ə] through

secondary accent in the one case, and through distributed or hovering stress in the other. British pronunciation,

as we have suggested, is in general somewhat more subject to the type of change under discussion than American.

Thus, medicine and venison have lost their middle vowels in British practice but not in American; business, it may be added,

seems to be going the same way, in both countries.

One further point about relaxed or obscured pronunciation may be made: It tends to occur even more thoroughly

in phrases than in single words. This is strikingly true of the conventional formulas of speech that are used again

and again. The principle evidently is that the more familiar the phrase, the less distinctly its separate sounds

need to be pronounced. Hence, phrases become single words or single sounds. Goodbye, from God be with you, is

a classical illustration. The evolution, in English, of the Latin mea domina is likewise noteworthy: in Middle

English, it was madame (three syllables, the middle accented); but in Modern English, it has become successively madam, ma'am,

and even 'm (as in Yes'm, and the Southern American No'm). In the last, and surely the ultimate, simplification, the five vowels and four

Contemporary Pronunciation

233

consonants of the original phrase have been reduced to a

single consonant. The process of course is not peculiar

to English; to compare the pronunciation of the French

monsieur23 or mademoiselle with the spelling is to suggest

the parallel way in which similar formulas have been

simplified in another language. Again, however, the

greater extent of relaxed pronunciation in English is

fairly evident; and it is quite explicable in view of the

more thorough changes in English pronunciation in

general.

The distinctive peculiarities of General American in

individual vowels or consonants may now be enlarged

upon. Most characteristic of all, probably, is the all but

universal use of the low flat [æ]—or sometimes the com-

promise [a]—in words which in the usual British pronun-

ciation have [ɑ] or [ɑː]. In many of these words, the

sound precedes the combination of a nasal and a con-

tinuant;24 half, laugh, grass, bath, path, dance, and answer

are a few examples. It may be stated unequivocally

that, except for eastern New England and tidewater

Virginia, the normal American pronunciation makes the

vowel of such words as these identical with that of hat and

man. In spite of this, the notion still lingers in some

American minds that the “flat” sound should be aban-

doned for the supposedly more elegant “Italian” a.

The history of the sounds in question should do some-

thing to dispel this idea. The [ɑː], in the first place, is

not of very great antiquity in these words; Old English

long a [ɑː] of course did not remain unchanged, but

developed into Modern English [ɔː]. Further, the usual

23 Sometimes spelled, indeed, m'sieur or m'sieu'.

24 For a full enumeration, see Kenyon, American Pronunciation,

pp. 100–102.

234 The Development of Modern English

evolution of Old English short a [ɑ] was, in open syllables, into Modern English [e:], and in closed syllables, into Modern English [æ]. The latter development has occurred in such words as cat, man, and so forth. In the early Modern English period, then, [ɑ] or [ɑ:] was not used in either English or American pronunciation, except dialectally. The first dictionary recognition of these sounds (in such combinations as those mentioned above) comes almost at the end of the eighteenth century; until about that time, [ɑ] was usually regarded as the vulgar and [æ] the polite sound.25

The story of [ɑ] and [æ] is further complicated by the attempt of early nineteenth-century grammarians and lexicographers, in both England and America, to introduce a compromise vowel [a]. It is curious to remember that this was at first advocated as an alternative to a more vulgar [ɑ], if one was not willing to go the whole way and say [æ]; in the later nineteenth century, [a] has been advocated in just the other way—as an alternative to a more vulgar [æ], if one was not willing to go the whole way and say [ɑ]. Opinions differ very widely as to the extent to which [a] has been adopted in American speech; and of course there is considerable difference in the interpretation of the symbol itself. If we take it to mean a sound that is very much nearer [ɑ] than [æ], there is every reason to feel that its use in General American is as artificial as that of [ɑ] itself. On the other hand, if we take it to mean a sound that is much nearer to [æ] than to [ɑ], there is much less reason to be certain of its arti-

25 Professor H. M. Ayres points out that this distinction still persists in the English of Bermuda: [æ] is the polite, and [ɑ] the low-class pronunciation. See his article “Bermudian English,” American Speech, Vol. VIII, No. 1 (February, 1933), pp. 3-10.

Contemporary Pronunciation

235

ficiality. It sometimes seems to be taken indeed to be what [æ] properly should stand for—the low front monophthong; and by “flat” a, a sound perhaps justly reprehended, is meant a diphthongized (and sometimes nasalized) combination that is not far away from [e: + ə]. The tendency seems to be growing to regard [a], in the sense of a strictly half-way compromise between [ɑ] and [æ], as thoroughly pedantic and unnatural.26

Discarding, then, the possibility of [a] (in the sense just defined), and confining the choice, in such words as bath, calf, dance, past, and the like, to [ɑ] and [æ], need there be any uncertainty as to present practice? That the question should be raised at all testifies to the strength of the New England tradition in our dictionaries and in our national life. For it is obviously the deference to the standard of New England (complicated, possibly, by a touch of Anglomania) that caused [ɑ] to be cultivated elsewhere in the United States, especially in the last decades of the nineteenth century. In New England itself, the rightness of [ɑ] for general American purposes is occasionally maintained even more recently. The following pronouncement is all too typical of this blatant sectionalism: “The broad a seems likely to come in. Some New Englanders, as we have seen, overdo it; but unquestionably, aunt, laugh, last, past, and fast, as pronounced in New England, have the weight of author-

26 See Kenyon, American Pronunciation, p. 102 (and his references to the attitudes of Jespersen, Sweet, and Jones); also, a more recent article by Kenyon, “Flat A and Broad A,” American Speech, Vol. V, No. 4 (April, 1930), pp. 323–326. Krapp observes of this sound (Comprehensive Guide to Good English, pp. 2 and 3): “There is no word in the language in which it is unequivocally illustrated, and there is no region where it is naturally spoken.”

236 The Development of Modern English

ity so distinctly on their side that laff and ant are fast becoming provincial."27

Though the preference for [a] lingers on in the very restricted territory already indicated—and occasionally in young ladies' seminaries and in schools of elocution outside—there is no longer any doubt that [æ] is the normal American sound. Even within New England, and in England itself, [a] is losing ground. A very practical reason for not making the attempt to cultivate it may be given to any not to the manner born who seek to acquire it: It will surely be used in the wrong places. Even if they avoid the most grotesque error—that of dropping [æ] altogether, and making hat and cat, for example, almost if not quite identical with hot and cot—how can they possibly remember that class, pass, and grass have the [a], but not mass and gas; that dance is different from askance and entrance, example and sample from lamp and amble, and plastic from plaster? They may recall that vast and mass have different vowels, but what will keep them from the all too revealing [væst] [mas], instead of the “correct” [vast] [mæs]? To put it on the lowest ground, that of expediency, it is an attempt that had better not be made.

When we turn from [æ] in certain groups of words, as a mark of American speech, to [u:] (rather than [ju])28

27 Greenough and Hersey, English Composition, New York (Macmillan), 1917, pp. 354 and 355. But, alas for human hopes! In 1932, it evidently seems to the same writers less certain that the broad a will "come in"; for, in their revision of this book, now called Writing Well, they have reduced this statement to the innocuous (if meaningless) advice: "Care should be taken in pronouncing have, aunt, laugh last past, and fast." (P. 451.)

28 To avoid complicating the point, in what is after all a highly controversial matter, I do not attempt to distinguish between the rising diphthong [ru] and the falling diphthong [ju]. For a careful treatment of this distinction in American speech, see Kenyon, pp. 125-129.

Contemporary Pronunciation

237

in another group, we are confronted by a somewhat similar state of affairs. Here again most Americans normally use the one sound, but they do so in spite of lexicographic prohibition and often (much more than in the other matter) with an uneasy feeling that they are falling short of the best standards. Words about which this question especially arises are those in which the vowel (or diphthong) is preceded by a pre-dental or alveolar consonant (except l and r). Words in which the preceding consonant is a labial (or [k] or [h]) give no difficulty: pew, beauty, music, few, view, cute, and huge are pronounced only with the [ju]. Normally l and r are followed by the simple vowel, in England as well as in America; lucid, lure, grew, true, and rule are typical instances. This is probably more generally true, however, in respect to British pronunciation, for r than for l. But Nicklin29 observes that “ . . . in lute, the more careful pronunciation is the rarer, and the majority of speakers confuse the word with loot.” To an American, the “more careful” actor’s pronunciation of this line in the opening soliloquy of Richard III: “Capering to the lascivious lute” sounds decidedly stagey. It may be added that there is no doubt about the [u:] when l is preceded by another consonant: blue, clue, flute, and glue are examples. Words of the group in which British and General American usage are sharply divided are those in which the stressed vowel is preceded by s, z, sh, t, d, th, or n. Even here, however, a further discrimination may be made. After the first three consonants just mentioned, it would appear that [u:] is gaining ground, in England as well as in America. To quote Nicklin

29 Op. cit., p. 29.

238 The Development of Modern English

again (page 29): “Thus, while the stage until quite lately persisted in speaking of Syew' zan (Susan), modern English has abandoned the painful attempt and says quite frankly Soo'-zan.” For suit, [sjut] was John Barry-more's pronunciation of the word in:

Nor customary suits of solemn black,

but it was, to an American ear, decidedly artificial. For these words as a group, American usage almost always favors [u:], and British usage may have begun to follow suit.

The real cleavage is in those words in which the stressed vowel follows t, d, th, or n. Typical words are stew, tune, tube, due, duke, duty, enthusiasm, thews, new, nude, and nuisance. British usage recognizes only [ju] for all these and similar words. This is also the usual prescription of dictionaries and schools in America. There can be no doubt, however, that [u:] is in far more general use in actual practice, and that it is gaining ground. This is so even in New England, where once [ju] was more prevalent-so very prevalent, indeed, that [ju] sometimes lingers on in rural New England even where [u:] is the only sound in cultivated speech; to and two, for example, are sometimes made [tju], and do, due, and dew given alike the pronunciation [dju]. Only in the South does [ju] remain in general natural use, alike in rude and in polished speech.

The fact that [u:] is the usual American pronunciation, in the words just indicated, and the additional circumstance that it is undoubtedly winning wider and wider acceptance in cultivated speech, would seem to be sufficient proof that it cannot be rejected as characteristic only of vulgar speakers. That a different practice

Contemporary Pronunciation

239

obtains in England should surely not be regarded as making it necessary for Americans to abandon [u:] and studiously cultivate [ju]. Some Americans do, of course —and frequently with as ludicrous results as when the same speakers substitute [a] for their native [æ]. If the appeal to individual authority carries more weight with such people than does the force of general (including good) usage, let them consider that W. D. Whitney, the greatest student of language that America has yet produced, went on record with the observation that in his pronunciation the vowel of tube, new, and so forth, was "frankly and unmistakably" identical with that of food.30

General American English is characterized further by the use it makes of [ɔ] and [ɔ:]. The tense [ɔ:] of awe and law is heard in many words in which British pronunciation favors a slack [ɔ] that approaches [ɑ]. Three interesting groups of such words are these: (1) words in which the vowel is followed by the voiceless continuants [f], [s], and [θ]; (2) words in which it is followed by [n] or [ŋ]; and (3) words in which it is followed by the stops [k] and [g], and a few words in which it is followed by [d]. Examples of the words in question are: (1) coffee, cough, office, soft, Boston, cost, lost, sausage, broth, cloth, and moth; (2) gone, long, song, strong, throng, and wrong; and (3) lock, mock, rock, stalk, bog, dog, fog, frog, hog, log, God, rod, and sod. In the first two groups, [ɔ:] (sometimes relaxed to approximate [ɔ]) is the typical American pronunciation; the use of [ɑ], to many Americans, savors of affectation. For the third group, there is

30 Elements of English Pronunciation, New York (Scribner's), 1874, p. 220.

240 The Development of Modern English

considerable variation in General American pronunciation. Only the words stalk, dog, and (perhaps) God

commonly have [ɔ:] rather than [ɑ] in the pronunciation of the Middle Atlantic states; while farther west, [lɔ:g],

[hɔ:g], and [mɔ:k] are familiar.

On the other hand, [ɔ] does not appear in General American, though it does in the speech of the South and

of eastern New England (as in British pronunciation) in words in which the vowel appears before other voiceless

stops than [k]. Such words are fop, got, hot, not, spot, top, and watch. For these, the General American value

of the vowel is [ɑ]. This sound sometimes appears likewise, in General American use, in such miscellaneous

words as forest, on, continent, follow, proposition, pomp, orange, doll, and horrid. Of these, it may be said that

the [ɒ] changes to [ɑ] as one goes west from the Middle Atlantic seaboard, the alteration beginning in western

Pennsylvania.

Another case of divided usage is the alternation of [u:] and [ʊ] in words spelled with the double o. The

general tendency here is very clear: Old English ō [o:] quite uniformly became [u:], though the spelling absurdly

retained the oo symbol, which thus came to take on another meaning. However, it is evident that the tense [u:] tends to shorten to the lax [ʊ], and sometimes to

lower and go forward to [ʌ]. The four possibilities are illustrated in these words: (1) brooch (the exceptional

survival of the original Old English value of the vowel, paralleled also in a few proper names like Moore and

Roosevelt, in the latter of which the Dutch sound is the same as the Old English one), (2) food, (3) good, and (4)

blood. The chief variation in present-day usage is between [u:] and [ʊ] in such words as broom, Cooper, hoof,

Contemporary Pronunciation

241

hoop, proof, roof, room, soon, soot, and spoon. Satis-

factory generalization seems impossible here; one can

see that the trend is from [u:] to [ʊ], but it is constantly

held in check by the frequent presumption that [u:]

is the “correct” sound. In British usage, for example,

only [u:] seems to be sanctioned in hoof, hoop, roof, and

perhaps root and soot; but in the United States, there is

considerable sectional diversity in all these words.31

Professor Krapp's opinion that for broom, room, and

spoon, the [ʊ] sounds are “rare and the influence of

spelling now seems to be decisive against them”32 is

both dogmatic and wide of the mark; for why should the

“influence of the spelling” be any more decisive for room

than for book or good? One may compare with this

Nicklin's remark33 that:

. . . room . . . occurs so frequently in compounds such as

bath-room, bed-room, and dining-room, where the stress accent

on the first syllable tends to shorten the vowel in the second

part of the compound, that by a reaction, apparently, the

uncompounded word has come to have the short ʊ sound.

Two other words are in danger, for whatever cause, of getting

their vowel sound similarly shortened—viz., soon and broom.

To the present writer, it seems that [ʊ] in both room and

broom is well established in General American, and the

same sound in soon in New England.

To the foregoing list of individual sounds in which

some question of present-day practice arises, two others

31 See Grandgent's article, Modern Language Notes, Vol. VI, pp. 458 ff.

(1891). His results are summarized in Kenyon, pp. 123–124.

32 English Language in America, Vol. II, p. 149.

33 Pp. 29–30. Reprinted by permission of the Oxford University Press.

And see the Appendix to Nicklin's book for an interesting list of words

arrayed in three columns according to whether the pronunciation in

the “Standard dialect” is [Λ], [ʊ], or [u:] (pp. 100 and 101).

242 The Development of Modern English

—this time consonants rather than vowels—might be added. These are the sounds represented, in the symbols of conventional spelling, by wh and r. The questions that both suggest have, however, already been discussed. In general, it may be repeated that the “Received Standard” deprecates both [ʌ] and [ɹ]; General American, on the other hand, at least theoretically goes in for [ʌ] and (with more misgiving) [ɹ]. Pride of course can be taken in either [ʌ] or [w]. Thus an adherent of Southern British finds with satisfaction that his speech retains the simplification of hw to w that was made in the South but not in the North, while the usual American tendency, as has been already suggested34 is to cultivate [ʌ] where it is often not the sound natural to the speaker. So too with r. Professor Wyld, for example, looks with distinct approval upon the “r-less” nature of contemporary Southern British, while a Northerner (like many an American) finds this very characteristic a sign of foppishness, if not downright effeminacy. “Devoid of the rasping sound of inverted r before consonants” is one of the points in Professor Wyld’s catalog of the desirable qualities of his own brand of pronunciation;35 but a Scotchman looks at this tendency with something like horror: “Abhor the practice of men who will not bring out the letter r.”36 Not a few Americans, on the other hand, are prone (another testimony to the strength of the New England tradition) to feel that to suppress the r is to speak in a more distinguished manner.

34 P. 227.

35 Historical Study of the Mother Tongue, New York (Dutton), 1906, p. 23.

36 Mair, William, Speaking, Edinburgh (Blackwood), 1908, p. 47. The writer attributes this advice to Spurgeon—“An Englishman [speaking] to Englishmen.”

Contemporary Pronunciation

243

It may occur to some readers that all this discussion of variation in contemporary pronunciation has been conducted with very little reference indeed to an authority usually evoked when a dispute upon pronunciation arises —the dictionary. Why is it necessary to go elaborately into varying habits of speech in different places, when we have a court of highest appeal that will at once settle all questions of correct usage? In a word, to put the question in the form it usually takes in practice, why not “look it up in the dictionary”? To answer this question as specifically as possible, to point out why “the dictionary” is not completely satisfactory as the arbiter of pronunciation, must now be our task. If anything can be done to shake the blind faith with which many a consulter of “the dictionary” approaches that supposedly sacred book, something useful will, in the opinion of the writer, have been accomplished.

To “look it up in the dictionary,” in the first place, implies either that there is only one dictionary or that the verdict of all dictionaries is one and the same. Actually, there are many dictionaries, of varying degrees of reliability and usefulness, and there is, further, a very considerable diversity in the matter we are just now concerned with—the pronunciations recorded. We may very easily, that is to say, “look it up” in the wrong dictionary. It comes as a shock to an American using the Oxford, for example, to find that the preferred pronunciation of ate is [ɛt]. The best of all dictionaries for such questions as etymology may thus be a most inadequate index to accepted American pronunciation.37 The

37 Of H. C. Wyld’s new and excellent Universal Dictionary (New York, Dutton, 1932), Professor Kemp Malone remarks, Modern Language Notes, Vol. XLVIII, No. 6 (June, 1933), p. 379: “. . . the American

244 The Development of Modern English

Shorter Oxford (1933), indeed, specifically disclaims making allowance "as a rule, for dialectal, colonial, or American varieties [of pronunciation]."38 Nor is this difficulty completely removed if we grant merely that differences in national standards do exist, and therefore confine our choice, for questions of pronunciation, to American dictionaries only. It is still a far from remote possibility that the pronunciation one finds will be a sectional rather than a truly national one.

The sectional bias of the leading American dictionaries, particularly Webster's, takes the form of frequently preferring the New England, rather than the General American, variant, in words that have no one nationally accepted pronunciation. The realistic method of dealing with such words would certainly seem to be to indicate three different pronunciations (General American, Southern, and New England) for those words in which differences do in fact exist in the cultivated speech of these three sections. Only in this way could the dictionary approach a real record of divided usage as it actually exists. Moreover, only in this way could the dictionary keep abreast of the present state of knowledge as to American pronunciation. If it be objected that this solution is a counsel of perfection, that to attempt it would be an impracticable extension of the present scope of our dictionaries, it is still possible to reply that the solution adopted—not to be sure in theory, but frequently in practice—is almost the worst conceivable

reader must be on his guard, for Mr. Wyld makes a curious distinction between sounds and meanings: in dealing with the former he ignores American usage, whereas in dealing with the latter he is careful to record everything American that comes his way."

38 Introduction, p. ix.

245

Contemporary Pronunciation

one. For if one of the three sections has a better right

than another to be considered, in its speech, as approxi-

mating national status, it is obviously that which speaks

General American. It is little short of grotesque to

find in the latest edition of Webster's Collegiate the

assertion that "the generally accepted sound" of the

vowel in words like calf, half, and salve is that of mar.

This dictum is, to be sure, qualified by the admission that

the compromise [a] or "even ă [as in hat] are also common

in such words." But the real status of the sounds

in question, as we have seen, is that [a] is artificial or

pedantic only, that [a] is virtually limited to Eastern

New England and Eastern Virginia,39 and that [æ]

is in every other respect the one sound in national use.

To represent [a] as "the generally accepted sound"

is a violent distortion of the facts in the case.

This preference for New England usage, perhaps less

marked in our other leading dictionaries than in Webster's

but still all too generally familiar, can be explained

only as the survival of a former deference, outmoded now

in most other matters, to New England as the center of

our national culture, and to Boston as the "hub of the

universe." It is coupled perhaps with another and a

more ancient survival: the Colonial attitude of sub-

servience to English tradition and authority. Since, in

matters of pronunciation, the usage of Eastern New

England and that of Southern England are frequently at

one, both influences affect our dictionaries in the same

way, and both operate to disqualify them as reliable

39 Even here, it would seem that usage, among both cultured and

uncultured speakers, is sharply divided between [æ] and [a]. See Krapp,

English Language in America, Vol. I, p. 39, and his reference to Primer,

Pronunciation of Fredericksburg, Va.

246 The Development of Modern English

guides to American pronunciation. Professor Kurath

has stated the result forcibly, but not extravagantly:40

Why should the dictionaries deliberately ignore this diver-

gence [that of the West, the East, and the South]? Why

should they take it upon themselves to force the standard of

one area on the entire country? How fruitless their efforts

have been! If it be the aim of scientific investigation to find

out the facts and to make them known, should we not expect

our better dictionaries to record faithfully how the edu-

cated classes actually speak, instead of indulging in their

make-believe?

The first cause of dissatisfaction with the result of

"looking it up in the dictionary" is thus the very real

possibility of discovering either an un-American or a

sectional pronunciation. It might have been added too

that the mere factor of time militates against the accuracy

of dictionaries. Not only will an early edition of a

standard dictionary occasionally, and quite inevitably,

record a pronunciation that has been superseded in actual

usage, but even the latest edition must necessarily be,

to some slight degree, behind actual speech. Even were

dictionaries less conservative than in truth they are,

one might still argue that, in the very nature of the task,

every dictionary is out of date as soon as it is issued.

Living speech is an ever-flowing stream that is continually

passing by the stationary and unchanging record of its

course that the dictionaries preserve for us. The

record is of great value, to be sure, but it must not be

mistaken for the thing which it records.

A second general reason for feeling that dictionaries

give but inadequate guidance on pronunciation is the

fact, honorable as it is to dictionary-makers. that there is

40 Op. cit., p. 281. Reprinted by permission.

Contemporary Pronunciation

247

no consensus of dictionary opinion as to what is the best pronunciation of many words. There is of course no one dictionary that can lay unchallenged claim to the place of "Supreme Authority"—though one, in its advertising, does usurp that proud title—and there is, in point of fact, very considerable disagreement among the best of them as to the sounds of many words. Both Webster's New International and Funk and Wagnalls's Standard include, in their unabridged editions, long lists of words for which the leading dictionaries give differing pronunciations. Nor should it be supposed that differences of this sort occur only in unessential details or in uncommon words. One who has thought of the advice on pronunciation given by "the dictionary" as being available, unchanged, in any dictionary, will be astonished by the variety of the preferences exhibited in several different dictionaries. To test the matter (and a), let the reader investigate, in four or five of the best dictionaries, the pronunciation of adult, advertisement, aerial, again, alkali, aristocrat, associate, and asthma.

But this is by no means the whole of the case against our leading dictionaries. Not only do they disagree among themselves, but as a group they fail, in greater or less degree, to give the full record of the obscuring of unaccented vowels and the blurring of consonantal combinations that are phonetically difficult—both of which, as we have seen, take place on a large scale in actual speech, cultured as well as uncultured. The dictionaries are all too ready, in a word, to recommend spelling-pronunciations, with the result that what the dictionaries present is not current usage, but rather the sounds of a former day that are embalmed in an archaic

248 The Development of Modern English

spelling. Richard Grant White observed that "it is in

the delicate but firm utterance of the unaccented vowels

with correct sound that the cultured person is most

surely distinguished from the uncultured," and something

very like this is the standard recommended today, at

least by implication, by Dr. F. H. Vizetelly, editor-in-

chief of the Standard, when he writes: "Unfortunately,

we have with us a large class of persons who speak

without thinking how our words are spelled, and who,

therefore, squeeze all the juice out of our speech by

refusing to enunciate carefully all the niceties of sound

that the words contain."41 As the necessary antidote to

this, let the reader note the words of Professor Wyld

condemning such an attitude " . . . a rigid appeal to the

spelling—the very worst and most unreliable court for the

purpose."42

The full consequences of connecting spelling and pro-

nunciation, as most of our leading dictionaries do, have

not yet been suggested. Almost inevitably, inaccurate

or merely theoretical43 pronunciations are constantly

41 Preface to A Desk-Book of 25,000 Words Frequently Mispronounced,

New York (Funk and Wagnalls), 4th ed., 1929, p. viii.

42 A History of Modern Colloquial English, p. 18.

43 Even the Oxford is not beyond reproach in this matter. In its

latest form, the Shorter Oxford (1933), there is an introductory explana-

tion (p. ix) to the effect that twelve vowel sounds, which are given with a

separate symbol for each, occur in certain words only as "primary or

ideal" values, "as in rhetorical utterance, in singing, and in cases of

deliberate or affected precision"; otherwise, and "in normal speech,"

the value of each one is [ə]. Yet these twelve variants, which are further

described as occurring only in "the historical and ideal pronunciations"

of the words in question, are the only ones that are given when these

words are pronounced in their alphabetical order in the dictionary.

Surely this is misleading: the average user of the dictionary, who in at

least nine cases out of ten has not read the introduction, naturally

assumes that he is being given pronunciations that are not "historical"

or "ideal" but current and realistic.

Contemporary Pronunciation

249

being presented, especially when the effort is made to indicate pronunciation without respelling. The Websterian diacritical markings, which are, unfortunately, the one code at all generally familiar in our schools, continually suggest quite untrue relationships between the written aspect and the sound of the word. Their use can only be deplored. Would anyone suspect, for example, that the second vowels of novĕl and finăl (so marked) are in present practice actually one and the same—an [e] so thoroughly obscured that it tends to be lost in [l]? The other vowel that prevails in unaccented syllables is of course [ɪ]; but in the Websterian code it appears as ĕ in ĕvent and as ă in senăte. The case of the last is particularly interesting, for it is one of the keywords appearing on every other page of Webster's Collegiate. Does not its use suggest that the pronunciation of senate is to be distinguished in some subtle fashion from that of sennit or of Sennett? In actual use, all three are identical.

One of the unfortunate results of the dictionaries' weakness for spelling-pronunciations is that teachers are likely to accept them and to insist that their pupils use them. Professor Kenyon44 tells of a teacher who carefully drilled her pupils to pronounce the noun subject with the 'full sound of e as in let,' and then, passing to another topic, observed that she would change the ['sʌbdʒɪkt]. This is, one may be sure, an all too typical consequence of depending on the spelling. Not, of course, that dictionaries are always to blame. Theoretically, they recognize that spelling follows pronunciation, not pronunciation spelling. They do not advocate, for

44 Op. cit., p. 4.

250 The Development of Modern English

example, that such words as busy, colonel, debt, does,

English, knife, said, two, and women be pronounced as the

spelling might suggest. That other teacher, referred to

by Professor Fries,45 who instructed her pupils to

pronounce laughter like slaughter and daughter (i.e.,

[ˈlɔtər]), because it was similarly spelled, could not blame

the dictionary for her mistake. In a less extreme way,

however, pronunciations suggested by the spelling rather

than by usage are not uncommon in dictionaries. A few

random illustrations are the unstressed syllables of these

words: carpet, engage, fidget, goodness, preface, mistress,

necklace, and useless. The usual dictionary tendency is

to represent the unstressed vowel by a symbol—ĕ or ă,

for example—adapted from the conventional spelling

of the word, rather than to indicate the [ɪ] that is actually

heard in speech.

It will be recognized that at least part of the remedy

for this failure of the dictionary is plain. If the system

of indicating pronunciation without respelling leads to

inaccuracies, the obvious improvement is to employ a

phonetic alphabet and respell all words phonetically.

That this is what ought to be done is scarcely to be

doubted, but there are more serious practical obstacles

than perhaps appear at first sight. For Webster's

New International in its next edition to change its entire

system of indicating pronunciation would be, for one

thing, an enormously expensive undertaking. Even

apart from considerations of this sort, there is grave

doubt as to whether the great mass of the users of the

dictionary would want their familiar, though unscien-

tific, symbols done away with. Our leading American

45 Op. cit., p. 59.

dictionaries are, after all, primarily commercial enterprises that cater to a public whose habits and prejudices must be considered. Webster's chief American rival, the Standard, while using a second and a phonetic alphabet—not, unfortunately, a very close approximation to the International Phonetic Alphabet—still retains the Websterian symbols, doubtless as a concession to the deeply rooted habits of its patrons.46

To return to our catalog of the shortcomings of "the dictionary." More can be said than that individual dictionaries fail as a guide to present-day, national pronunciation; that dictionaries as a group vary among themselves; and that they record inadequately the modern tendency to obscure unstressed vowels. Besides all this, the user of the dictionary should be aware that, in the very nature of the task, the record of the dictionary is that of individual, isolated words—not of words in their ordinary groupings. The pronunciation of the word does, in practice, vary with its setting in the phrase or the sentence. Spoken language is made up rather of phrase-groups than of separate words. It therefore follows that a single word may very well have two (or more) pronunciations, which vary with the amount of emphasis given to the word in different contexts. Dictionaries show differences of this kind only very rarely; to distinguish two the's ([ðiː] and [ðə]), as most dictionaries

46 Wyld's recent Universal Dictionary has adopted a similar policy. I cannot forbear quoting Professor Malone's comment on this (op. cit., p. 379): “. . . a phonetic respelling for the literate, preceded by a makeshift Websterian respelling for the illiterate. . . . although we may be sceptical of the value, commercial or otherwise, of the makeshift respellings, we must be grateful for the presence, alongside them, of phonetic spellings which the hapless editor can look at without shame, and the student can use with safety.”

252 The Development of Modern English

aries do, is to apply this principle; but such an instance is very exceptional. Yet a parallel variation in the

pronunciation of other words is very common indeed in natural, colloquial speech. Professor Kenyon lists47 a

group of very familiar words that have, in good pronunciation, two or more variants. These variations are

sometimes the result of the stressed or unstressed character of the word in the phrase, but sometimes also the

result of its immediate phonetic environment. Thus, that will be either [ðət] or [ðæt] according to whether it is

the unstressed conjunction or the stressed demonstrative. Observe how the pronunciation varies in the phrase,

"He said that that was correct." A contemporary novelist,48 by the way, attempts to make the spelling

record these pronunciations, by employing that for the conjunction and thatt for the demonstrative. The other

type of variation, that caused by the immediate phonetic environment, is illustrated in Kenyon's citation of six

possible pronunciations for and: the stressed forms [ænd] and [æn] in, respectively, "and, indeed, I should"

and "both Jane and James are here"; and the unstressed forms [ənd], [ən], [nd], and [n] in "snow and ice," "cup

and saucer," "head and arm," and "rod and gun." When variation exists to this extent in good usage, it is

not difficult to understand that the comparatively rigid record of the dictionaries must necessarily fail to echo

the real accents of speech.

The point that the pronunciation given in the dictionaries is sometimes theoretical rather than realistic might

be developed further. Dictionaries, in spite of their

47 Op. cit., pp. 151-157.

48 V. Sackville-West, in Family History (New York, Doubleday, Doran, 1932).

protestations to the contrary, not infrequently yield to

the temptation to tell what the pronunciation of a word

should be, rather than what it is. The spelling of the

word, or its etymology, or an analogy which is suggested

with other words-all of these are possible factors in

the dictionaries' occasional recommendation of a pro-

nunciation that is simply at variance not only with wide

usage but even with good usage. There undoubtedly

is such a thing, in other words, as a "dictionary pro-

nunciation." In such terms one may perhaps describe

the dictionaries' preference for the accent on the first

syllable of arbutus, frontier, and peremptory, for the

accent on the second of abdomen, for the sounding of the

first c [k] in Arctic, for [e:] rather than [æ:] as the accented

vowel in armada, for the [g] rather than the [dʒ] in

oleomargarine, and for oboe as [ˈoʊboʊ, ˈoʊboʊɪ, or ˈoʊbɔɪ].

It is only fair to add, however, that the more popular

pronunciation is sometimes grudgingly admitted-occasion-

ally prefaced by "often mispronounced as." As a

final illustration of this attitude of the lexicographer,

with its implication that his dictum takes precedence

over usage, let the reader ponder these words on inter-

esting: "The word is never correctly pronounced as if

consisting of three syllables. No dictionary or work on

pronunciation indicates any such pronunciation as

in'-tris-ting."49

A word of caution may fittingly be attached to the

foregoing enumeration of reasons for feeling that "the

dictionary" is an inadequate guide to good pronuncia-

tion. These pages will have been completely misinter-

49 The italics are mine. The quotation is from F. H. Vizetelly, How

to Use English (New York, Funk and Wagnalls, 1932), p. 378.

254 The Development of Modern English

preted if they have been thought to advise against using

dictionaries. It is not a less extensive, but a more intelli-

gent use of dictionaries that is being advocated—and a

recognition of what one can expect and what one should

not expect to find in their pages. Dictionaries, it cannot

too often be repeated, are not the lawmakers; they are

merely the law-recorders. For the reasons that have

been stated, their record may easily be fallible. To

put it differently, they do not so much create as indicate

the standard of pronunciation; and again, their inter-

pretation of the standard may be in error.

Where, then, is the standard of pronunciation to be

found? How may it be stated? It is easy to say that

the standard is set by good usage; but what constitutes

good usage? Who are the speakers whose habits of

pronunciation may be felt to indicate what good pro-

nunciation is? Are they sufficiently well recognized

and is their speech sufficiently uniform so that one may

say that the pronunciation used by this or that segment

of the population of the English-speaking world is good

pronunciation, and all else is bad?

It is evident that, among the possible answers to

questions like the last, there are two that are opposite

and mutually contradictory. One is the conception of a

very definite standard, that known as Received Standard

or Public School English. Definitions of this we have

already quoted.50 Those who hold this belief conceive

of the standard of pronunciation as being set by a highly

limited group of speakers (limited socially or limited both

socially and geographically). It perhaps should be

emphasized that faith in this particular brand of pro-

50 Cf. p. 215.

Contemporary Pronunciation 255

nunciation as constituting a single standard, and a standard already adhered to in practice, is not confined to Southern Englishmen. A fervent—though, to the present writer, wholly mistaken—plea for it is made in Miss M. E. DeWitt's book significantly entitled Euphon English and World-Standard English in America.51

The other extreme view is that there is no such thing as a standard of pronunciation—that throughout the English-speaking world, speech is too varied to permit of any standard at all. This theory has, perhaps, been set forth most persuasively by Professor Lounsbury.52 His formulation of it deserves to be quoted at some length:

Where exists that perfect standard which all orthoëpists assert or imply that they have furnished, but in the representation of which in numerous particulars no two of them concur?

From what quarter are we to look for the coming of this infallible guide for whose arrival we are all longing? It seems never to have occurred to any of the compilers of dictionaries, and to but few of those who consult them, that the simple solution of the whole difficulty is that in the matter of pronunciation there is no standard of authority at all. Nor, as things are, can there be. Pronunciation must and will vary widely among persons of equal intelligence and cultivation. A dictionary which sets out to establish on a solid base an authoritative standard is bound to take into account the practice of the whole body of educated men the world over who are entitled to consideration. How is that to be ascertained? The mere statement of the fact shows its physical impossibility. It is a task beyond the power of any one person or any number of persons to accomplish.

Even this is not the worst. If everybody worth consulting could be consulted, we should still be left in precisely the same state of uncertainty as before. . . .

51 New York (Dutton), 1924.

52 T. R. Lounsbury, The Standard of Pronunciation in English, pp. 212–216. Reprinted by permission of Harper and Brothers.

256 The Development of Modern English

Still this belief in the existence of a standard authority is one that will die hard even with the educated class. With the semi-educated class it will never die at all. . .

If one had to choose between the two opposite conceptions—a standard of pronunciation that is extremely limited, and no standard at all—the second would probably have more to commend it. Yet many feel that there is such a thing as a standard of pronunciation based upon good usage, though it is at once to be admitted that the standard must be elastic enough to allow for sectional and even individual variations, and that the standard is not one and the same for all occasions and purposes. As to the last point, it seems feasible to hold that good usage varies with the more or less formal character of the occasion. A greater number of “relaxed” pronunciations and of elisions and syncopations will naturally be admitted into ordinary conversation than would be appropriate for oratorical declamation.

Just as there are certain differences to be observed between the spoken and the written language in general, just so both spoken and written language have separate planes and degrees, each with its own tone and atmosphere, that mark the gradual transition from the least to the most formal occasion for speaking or for writing. There is such a thing as good usage, however, for the less formal as well as for the more formal occasions. Much misunderstanding is possible here. For one thing, “colloquial” should not be taken as a term of contempt. The “colloquial” word or the “colloquial” pronunciation of the word is as much in place in familiar conversation as the “formal” word or the “formal” pronunciation would be out of place.

Contemporary Pronunciation

257

It remains for us to discuss one or two somewhat different attempts to find a brand of pronunciation that may be considered as embodying good usage. There is, for example, the practice of our best actors and actresses. May not stage pronunciation be accepted as a practical model, as in Germany the Bühnenaussprache is felt to be? There are several reasons for deciding that this is at best a most dubious guide. For one thing, in the very nature of the case, stage pronunciation must be audible and intelligible at far greater distances than conversational English need be; it tends therefore to be too declamatory and oratorical to serve as an appropriate guide for ordinary purposes. Then too, our American actors are often decidedly under British influence in speech. There is a tradition here that is quite at variance with the trend of General American. A good part of our stage pronunciation is international in character, and far more British than American. Again, some of our leading actors are notoriously bad speakers, and some are given to types of pronunciations that are obviously ill-suited to imitation. One of our most distinguished actresses, who must be nameless here, has been cruelly but not altogether unjustly chided by critics for her “waterfront voice.” And one of the most promising young actors, famous for his portrayal of wistful adolescents, is almost as well known for his flat and lifeless enunciation. All in all, there seems small reason to seek the embodiment of good usage on the American stage.

What of the radio? Here, very evidently, is a most potent instrumentality for putting a standard of pronunciation before millions of listeners. It is clear too that radio pronunciation—the speech, that is to say, of the professional broadcaster—does tend more and more to be

258 The Development of Modern English

of uniform character. One can scarcely doubt, moreover, that it is widely influential. Indeed, it seems quite probable that to the broadcaster's following of British standards we owe the recent popularity of the single accent in secretary, laboratory, and so forth. But it is patently artificial, as was illustrated by the announcer who conscientiously strove to refer to the Kentucky Derby as the ['dɑrbi:], but occasionally relapsed to ['dɝbi:]. Something like that is just the criticism. The broadcaster's pronunciation, far too often, is self-conscious, affected, and stilted53 to such a degree as to disqualify it from affording a standard of good usage. It is not the speech of a man who naturally speaks well as much as that of one who is making every effort to enunciate elegantly.

For all that, the radio and its almost equally influential relative, the talking picture, are certainly worthy of attention if one is to speculate about the future of English pronunciation. That the talking picture has seriously alarmed English people by exposing them to the horrors of the "American accent" is too well known to need comment. Out of their indignation and out of the American producers' willingness to conciliate their prejudices, it seems altogether likely that a compromise between American and British habits of speech, for "talkies" intended for international circulation, will increasingly be effected. Perhaps it may justifiably be feared that in this compromise, partly by reason of the British tradition in the pronunciation current on our stage, the prestige of Southern English speech will have

53 For an excellent treatment of this point, see the editorial "Broadcasting English," Saturday Review of Literature, Vol. VII, No. 28 (Jan. 31, 1931).

Contemporary Pronunciation

259

an undue share. Certainly it would seem, as has just

been indicated, that a similar fear for the pronunciation of

American broadcasters is being more and more justified.

The standardization of radio pronunciation, and the

standardization, in part at least, after Southern English

models, is a most interesting phenomenon. In addition

to the usual factors that give this brand of speech a

special prestige even in America, there is to be noted

something else: the more careful and more intelligent

supervision that has been given to the development of

radio programs in England, as compared to their hap-

hazard growth in America. The English evidently take

their radio more seriously. This is as true of the matter

of the broadcasters’ pronunciation as of other things.

American as well as English readers will find of great

interest the British Broadcasting Corporation’s “Recom-

mendation for Pronouncing Doubtful Words.”54 For our

present purpose, it is pertinent to remark that this

pamphlet specifically observes,55 “Unfortunately speech

is not capable of rigid measurement, and there is no

standard of pronunciation.” This assertion, however, is

followed by the enunciation of other principles that

qualify it in a vital way:

Out of the broad band that comprises all district and class

variants, there is emerging a considerably narrower band of

variants that have a very great measure of similarity. This

narrow band of types has more features in common with

Southern English than with Northern English. Those who

speak any one variety of the narrow band are recognized as

educated speakers throughout the country. They may

broadcast without fear of adverse intelligent criticism.

54 Reissued with criticisms, edited by Robert Bridges, as Tract No.

XXXII, Society for Pure English (1931).

55 P. 9. Reprinted by permission of the Oxford University Press.

260 The Development of Modern English

A fairly close approach, after all, to the doctrine of the

"Received Standard" of Southern English! It is only

fair to add, however, that the introduction and the

proposals themselves are conceived in a spirit that is

free from any suspicion of the dogmatic. Americans will

be interested to see that Sir Robert Bridges, the chairman

of the committee56 that made the recommendations,

dissents rather frequently both from his committee and

from the other critics whose opinions he has invited, and

occasionally favors a pronunciation that is more com-

monly American than English.

Probably one is justified in seeing in the radio a force

that must eventually make for a nearer approach to

uniform pronunciation over the whole of the English-

speaking world. We are living, however, too early in its

development to be at all certain of just how complete

this standardization will ever be. All that can be said at

present is that so far we must look elsewhere for a

formulation of the standard of pronunciation. And

here we return to our question: Where is the standard of

pronunciation at present to be found? It may be more a

matter of instinctive faith than something that is sus-

ceptible of demonstration, but the answer that seems

most satisfactory is the one that finds standard pronunci-

ation to be determined by the practice of good speakers

all over the English-speaking world. "Standard" in

this sense is evidently not equivalent to "uniform"; for

it is clear that there are not only divergences as between

British and American usage, but also sectional diver-

gences within the United States. The point is that both

56 Consisting of Forbes-Robertson, Daniel Jones, A. Lloyd James,

Bernard Shaw, and Logan Pearsall Smith.

types of divergence, and others like them for other parts

of the English-speaking world, may exist within the

boundaries of good pronunciation.

How then is the individual to determine what is

prescribed as the pronunciation for individual words and

phrases by this somewhat elastic international standard?

In the last analysis, the method must be the individual's

own observation of what constitutes good practice in his

section of the English-speaking world. Clearly it will

not do to restrict the application of "section" very

narrowly: sectional must not be identified with local.

The danger in that direction is that the sectional become

the parochial or the provincial. Yet even this danger is

to be preferred to that threatened by the opposite course,

that of accepting a rigid and uniform and (from the

General American point of view) alien standard. Here

lies the way to artificiality, to affectation, and to pedantry.

One word more, to qualify and also to safeguard the

position just set forth. Because the observation of the

individual is necessarily circumscribed, he should avail

himself of whatever help is afforded by reputable dic-

tionaries and handbooks of usage. Only, as we have

insisted perhaps to weariness, he should do this always

with a full realization that the usefulness and the relia-

bility of such guidance is limited. Direct observation

of the spoken tongue as employed in good practice must

ever be paramount. To supplement this observation

is the proper function of the guides we have mentioned.

If this version of good pronunciation and the method of

arriving at it has any validity, it follows that the grossest

possible error is to reverse the positions of the primary

sanction and its supplement. A pronunciation is not

"correct" because it may be found in a dictionary;

262 The Development of Modern English

rather, it may (or should) be found in a dictionary because good usage has already determined that it is "correct."

References for Further Reading

The books by Jones, Wyld, Kruisinga, and Jespersen listed on pages 212 and 213, and, in addition:

Fries, C. C., The Teaching of the English Language, New York (Nelson), 1927, Chap. III.

Grandgent, C. H., Old and New, Cambridge, Mass. (Harvard), 1920, including "Fashion and the Broad A" (pp. 25-30), "The Dog's Letter" (pp. 31-56), and "New England Pronunciation" (pp. 121-149).

Grant, William, The Pronunciation of English in Scotland, Cambridge (Eng.), 1913.

Jones, Daniel, An English Pronouncing Dictionary, New York (Dutton), rev. ed., 1924.

Kenyon, J. S., American Pronunciation, Ann Arbor (Wahr), 1924.

Krapp, G. P., Modern English, Its Growth and Present Use, New York (Scribner's), 1909, Chap. V.

, Pronunciation of Standard English in America, New York (Oxford), 1919.

, The English Language in America, New York (Century), 1925, Vol. II.

Kurath, H., "American Pronunciation," Society for Pure English, Tract No. XXX, London (Oxford), 1928.

Lloyd, R. J., Northern English, Leipsig (Teubner), 2nd ed., 1908.

Lounsbury, T. R., The Standard of Pronunciation in English, New York (Harper's), 1904.

Nicklin, T., Standard English Pronunciation, London (Oxford), 1925.

Palmer, H. E., Martin, J. Vietor, and Blandford, F. G., A Dictionary of English Pronunciation with American Variants, New York (Appleton), 1926.

Ripman, Walter, The Sounds of Spoken English, New Version, New York (Dutton), 1914.

CHAPTER VIII

Spelling and Spelling Reform

THERE is something inevitable in the association of "spelling" and "spelling reform," a suggestion that the visible representation of speech is never quite adequate or quite satisfactory—that its more rigid substance is continually in fresh need of adjustment to the changing contours of the spoken tongue. Does this way of putting it, however, express the true and the sole relationship between pronunciation and spelling? Is writing, in other words, to be conceived as existing for no other purpose than to represent the sounds of language? This is frequently accepted as an axiom by the spelling reformer; but that it is in need of serious and even vital qualification can scarcely admit of doubt. One way of approaching afresh the problem of how spelling and pronunciation are related is to go back as far as possible into the early history of language, and summarize the evolution of the alphabet. To do this would seem justified not only by the light that the story throws upon the interrelations of the three primary aspects of language—the visible symbol, the sound, and the meaning—but also by the intrinsic interest of the story itself. Against the background of such an account, which must, for the present purpose, emphasize the development of the written symbols of our own language, we shall later be in a better position to discuss the sometimes puzzling problems that result from the very unsatisfactory

263

264 The Development of Modern English

relationship at present existing between English pronunciation and English spelling.

Systems of writing, viewed in their chronological development, are thus summarized by Pedersen:1

Pre-alphabetical systems.

(1) Word- and syllable-script.

(2) Syllable-script.

Alphabetical systems.

(1) The Semitic alphabet (a syllable-script, which to us may seem a consonant-script).

(2) The Greek alphabet (vowel- and consonant-script).

Of the earliest type of pre-alphabetical system it may be questioned whether in strictness it deserves the name "writing" at all. The ideographic stage is rather drawing than writing, inasmuch as the interpretation of narrative drawings can be put into a variety of words.

Picture-writing, which was apparently developed independently by different peoples in widely separated parts of the world, came eventually into connection with the spoken tongue, though theoretically it might have grown up apart from it—as the sign language of the American Indians had an existence separate from the spoken dialects. Since it must be assumed, however, that spoken language preceded written language, it seems well-nigh inevitable that sooner or later the two would come into contact with each other. For it would be unthinkable that peoples possessing a somewhat fully developed set of auditory symbols would continue indefinitely to work out a new set of visual symbols in complete independence of the other. Drawing became writing when the picture ceased to stand for an idea and

1 Linguistic Science in the Nineteenth Century, p. 142.

came to stand for a combination of sounds. The tran-

sition was undoubtedly gradual and came about, in all

probability, "by means of the rebus method . . . by

the same sort of process which we apply when we

allow a picture of the sun to stand for the first syllable in

sundry, or let the picture of an eye stand for the pronoun

I."2 But of course even after the written symbol came

to represent the sounds of some one spoken word by

which an idea could be expressed, it also directly signified

the idea itself. It was not until the written sign stood

only for a sequence of sounds (irrespective of its meaning)

that the transition from word-script to syllable-script,

or from ideogram to phonogram, was complete. This

was the beginning of the true alphabet.

The full possibilities of the alphabet, however, were not

disclosed until the clumsy device of using a symbol for a

syllable (consisting perhaps of several consonants and a

vowel) gave way to the use of a separate symbol for each

individual sound. The Semitic alphabet was a syllable-

script which represented the consonants definitely but

left the vowels vague, so that it appears to consist of

consonant-characters only. Characters were no longer

so numerous as in the older scripts—the Egyptians

bequeathed to a Western Semitic people a means of

reducing their number—but the alphabet, as we under-

stand the term, was not yet quite evolved. The final

contribution was that of the Greeks: the resolution of

syllables into their separate vowels and consonants, and

the use of individual characters for the sounds thus

analyzed. All known alphabets are descended from the

Semitic, but only the Greeks revolutionized it in the way

just indicated, by discarding its syllabic nature. Of the

2 Pedersen, op. cit., p. 143.

266 The Development of Modern English

Mediterranean peoples who learned the Greek alphabet,

one race alone—the Romans—continued its use from

pre-Christian times through the entire Christian era.

Through its inherited use in the Romance languages and

through its borrowing by other Indo-European groups

such as the Teutonic, the Latin alphabet has long since

dominated the modern world.

Of the four stages in the evolution of writing just

outlined—the pre-alphabetic word-script and syllable-

script, and the Semitic and the Greek alphabets—it has

been observed by Henry Bradley that “the penultimate

stage . . . has a peculiar interest.”3 Bradley’s acute

analysis of the significance of the third stage, for the light

it throws upon the general relations of spoken and written

language, has very great interest. Arabic, he points out,

is to this day usually written as a consonant-script,

though vowel marks may be added. Yet it is evident

that the consonantal outlines can be apprehended, by

one familiar with the language, as readily as if the full

phonetic indication of the word were given—perhaps

more readily, because the eye takes in the simpler symbol

(if sufficiently familiar) more quickly and just as com-

pletely. Just so, the modern stenographer can read the

consonantal outline of familiar words, which have omitted

the vowel-markings that the Pitman system provides,

just as swiftly as the fuller ones. Medieval Latin manu-

scripts, as Bradley further suggests, were frequently

written with nearly every word abbreviated, and yet

were undoubtedly read, by one who had mastered the

art, with greater rapidity than they would have been

if the words had been written in full.

3 On the Relations between Spoken and Written Language, p. 6.

Spelling and Spelling Reform

267

Bradley's point is, of course, that it matters not a jot to the accomplished reader whether his native language is phonetically spelled or not; "what is important is that the group of letters before him shall be that which habit has led him to associate with a certain word." It might even be added that it is not necessary, always, to qualify this assertion by limiting it to one's native language. Certain Latin abbreviations that are commonly used in English have just as immediate meaning to the English reader as native English ones: e. g. does not require to be translated into "exempli gratia" and then "for example," or i. e. into "id est" and then "that is." In both instances, the eye conveys the meaning to the mind without even the suggestion of a phonetic middle stage. To put it more fully, there are two approaches from visual symbol to meaning: the one by way of the phonetic sign, the other direct. Graphically stated, the longer way is this (A standing for visual symbol, B for sound, and C for meaning):

The shorter route is, of course:

Just now often the second supersedes the first, it is probably impossible to say. Certainly in reading poetry and in reading such prose as that of Milton or Browne, DeQuincey or Coleridge, it may be assumed that the

268 The Development of Modern English

middle stage exists; it would assuredly be deplorable if it did not. Probably even in the rapid reading of the less poetic types of prose there are some words that call up sound-pictures and are audible to the mental ear. But it seems equally probable that for many words there is no such association; many written or printed words, that is to say, would seem to symbolize their ideas directly.

This digression has been intended to attack the proposition, often regarded as axiomatic, that the sole function of writing is to represent the sounds of language. One might argue further, as Bradley does,4 that not only do systems of writing that were originally phonetic tend to become ideographic with regard to their use or function, but in addition that English, together with other cultivated written European languages, has become partly ideographic with regard to structure: for example, in such matters as the capitalization of proper names, the use of quotation marks and of the apostrophe to denote possession, and (most strikingly) the ideographic employment of different spellings to discriminate between words once phonetically distinct but now homophones—rain, rein, and reign; or, oar, and ore; so, sow, and sew, and so forth. Some of these ideas we shall have occasion to refer to again when discussing the history of spelling reform. It is necessary now, however, to return to our summary of the evolution of the alphabet, which we left in order to comment on the permanent significance of the stage just preceding the creation of the Greek alphabet. We shall be concerned more particularly with the development of the symbols used in our language, and with the general trend of English spelling.

4 P. 10.

Spelling and Spelling Reform

269

The Latin adaptation was, as we have seen, the most important offshoot of the Greek alphabet. From the Latin alphabet descended the runic letters common to all the Teutonic peoples.5 The Old Norse word rún meant “secret” as well as “letter”—a suggestion of the esoteric character of early writing and its exclusive possession by a priestly cult, as well as its supposedly magical properties. The West Germanic peoples are thought to have learned the art of writing from the Gauls and to have received from them the Latin letters, which, however, they greatly transformed. It is probable that the theory of a common indebtedness to the Gauls, rather than any direct association, explains whatever likeness exists between the Teutonic runic alphabet and the Ogham alphabet of the Irish, a much more complete transformation of the Latin letters.6 Among the Teutonic peoples, the runes quickly gave way to the standard Latin alphabet, except in Scandinavia and Britain. In Britain—unlike in Scandinavia, where the number of runic symbols was decreased—the original runes were added to. But after the Christianization of Britain—that is, beginning about 600—the runic characters were for the most part very rapidly superseded by the Latin alphabet. The Anglo-Saxons nevertheless retained, for some time, a few of the Teutonic characters, especially the symbol for the sound of w and one of the symbols for the sound of th. It is the usual custom of modern editors of Old English texts to retain only the latter,7 substituting

5 The alphabet used in the oldest Teutonic literary memorial is, however, an exception. For his Biblical translation into Gothic, Bishop Ulfilas created an alphabet compounded of both runic and Greek letters.

6 Cf. Pedersen, pp. 229–239.

7 Flom, in Old English Grammar and Reader, employs another letter strange to the modern alphabet. This is a character like 3, used for the

270 The Development of Modern English

the modern form of w for the runic "wen." To the "thorn" letter, p, used for both the voiceless and the voiced values of th, the crossed d, ð, was later added to signify the voiced th only. This distinction, however, was made but very rarely in the manuscripts; almost always, as in modern printings of Old English texts, p and ð were used interchangeably. Two digraphs that have later been dropped were used for vowel sounds by the Anglo-Saxons: æ, for a sound familiar through the whole history of English and represented in modern phonetic alphabets by a similar symbol, though eventually discarded in conventional spelling in favor of simple a; and œ, for a sound like German ö, and soon discarded for e. The vowel y and the diphthongs ea and eo also corresponded to symbols of the runic alphabet.

The earliest spelling of English that has been preserved is thus seen to have been not very greatly different from that of Modern English in the symbols that it employed. Apart from the few preservations of runic characters, the principal differences are the absence of j and q, and the rare (usually very late) use of k, v, and z.8 As has already been indicated, the spelling of Old English was far from being perfectly phonetic, though it represented a much nearer approach to correspondence between symbol and sound than does that of Modern English. The first cause of the lessening correspondence between spelling and pronunciation that has marked the general development of English is, of course, the radical change in

palatal, as opposed to the velar, g. The Insular script, of the native documents in Old English, employed this symbol for all varieties of g, and made use as well of characters for r and s that differ from those used in Latin manuscripts.

8 See also page 194.

Spelling and Spelling Reform

271

pronunciation and the inability of spelling to keep up

with it. English pronunciation has altered in such a

revolutionary way that it has always been difficult for

spelling to keep abreast. Sporadic attempts have been

made in age after age to bring spelling abreast of pronun-

ciation; but their always partial success has lagged

far behind the necessities of the case. In considering the

general trends of both conscious and unsystematic change

in spelling, it will be necessary to specify some of the

causes-in addition to the primary one just mentioned-

that have brought about the present very unsatisfactory

relationship between spelling and pronunciation.

The familiar assumption that Old English was entirely

phonetic in its spelling is not quite in accordance with

the facts. The Latin alphabet, for one thing, was a

borrowing that did not entirely fit the need-for that

matter, the Latin alphabet itself was not, from the angle

of theory, ideal. Such modifications as were necessary in

adapting the Latin letters to Old English words are

perhaps to be regarded as the first in the long series of

adjustments and compromises one encounters in the

history of English spelling. Another cause of difficulty

was the fact that important dialectal differences existed

within the language and tended to prevent spelling from

changing as pronunciation changed. Still, it is probably

sufficiently near the truth to think of Old English as being

as nearly phonetic as Latin was or as Italian or Spanish is.

The innovation of v and z, appearing in late borrowed

words, was a minor step in the direction of the phonetic

goal, though failure to discriminate between p and ð as

symbols of the voiceless and voiced sounds must be

reckoned a counterbalancing evil.

First in time among the larger factors that have wrought havoc in English spelling was the influence of Norman French. New letters (j and q) were introduced, or at any rate first given wide use; new phonetic values were attached to the old symbols, or the old symbols were used in new combinations; e.g., c was written for s, o for u, and ou for ū. French influence did not cease with the centuries following the Conquest. To it, to give only a few illustrations, we owe such spellings as tongue for what would more sensibly be tung (like lung; compare the German cognate, Zunge); guess for what Chaucer spelled gesse; and such other anomalies as programme, catalogue, and quartette.

But spelling was still, until late Middle English times, quite largely a field in which the individual might display whatever eccentricity he pleased. Orm, whose Ormulum appeared soon after 1200, is one of the first of spelling reformers; his contribution was to double the consonant after short vowels and thus, after a fashion, indicate pronunciation somewhat more accurately. The invention of the printing press, however, marks the second great event in the history of English spelling, and the remote beginning of modern standardization. It was a kind of standardization, curiously, that made whatever confusion existed before appear unimportant. In the bitter words of Professor Lounsbury, "Upon the introduction of printing, indeed, English orthography entered into that realm of Chaos and old Night in which it has ever since been floundering; it then began to put on the shape it at present bears, 'if shape it may be called that shape has none.'"9

9 English Spelling and Spelling Reform, p. 272.

Spelling and Spelling Reform

273

Several reasons may be advanced to account for the unfortunate effect that the early printing press had upon spelling. One was the circumstance that most of the earliest printers came from Holland. These men followed certain Dutch analogies, and introduced absurdities like ghost, gherkin; ghospel, ghossip; ghess, ghest. The illustrations have been given in three groups, to suggest the three different lines of development that these gh words later took: the first pair retained the unnecessary h; the second pair dropped it and restored the Old English g; the third pair was later "Gallicized" by the substitution of gu for gh. Probably of more importance than such arbitrary respellings of individual words was the setting up, by different printing houses, of separate and inconsistent rules for spelling in general—a practice inherited in some measure by their modern successors. Moreover, the first printers were not only aliens, unfamiliar with English pronunciation; quite certainly, they were in general men of less education than their predecessors, the copyists of medieval manuscripts. Medieval authors sometimes complained of the incompetence and carelessness of their scribes—Chaucer's humorous exasperation with "Adam, his owne scriveyn" is a well-known instance—but undoubtedly these scribes were as a class far superior in knowledge and culture to the early printers. Spelling fell increasingly into the hands of semi-literate men; authors were frequently, as they have been ever since, misrepresented by the typesetter.

Through the sixteenth and seventeenth centuries, uniformity in spelling—however irrational as regards its relation to etymology or to pronunciation—was increasingly approached by the efforts of the printing

274 The Development of Modern English

houses. At the same time, since dictionaries generally regarded as authoritative did not yet exist, there was no court of higher appeal to adjust the varying rules of separate printing houses. Diversity necessarily continued—not, however, the kind of diversity in spelling that represents variation in pronunciation so much as a diversity occasioned by the individual whims or theories of different "stylebooks." Spelling nevertheless tended to become more nearly uniform, as one printing establishment influenced another; but its uniformity consisted largely in agreement upon spellings that had long since ceased to represent pronunciation. Renaissance spellings, standing often for a still older pronunciation, were more often than not permanently fastened upon the language.

The important consequence of the situation just described is that, from early Modern English times, the gulf between spelling and pronunciation has rapidly widened. Pronunciation changed freely—it had often changed, in fact, before the more or less permanent spelling was established—while spelling remained, for the most part, static. Here are the seeds for the creation of two languages—spoken and written. Whatever may be thought of the result, it can scarcely be denied that something like this is the present state of affairs: We possess a language that appears, on the printed page, in a form so different from the way in which it is spoken that the only adequate way for the dictionary to indicate the pronunciation of Modern English words is to respell them in a phonetic alphabet. One should not let familiarity with this situation conceal its oddity.

One most important step in the evolution of our spelling has not yet been fully indicated. The spelling

Spelling and Spelling Reform

275

itself, as we have seen, is much more the creation of

printers than of authors. But the stamp of approval

set upon it by the dictionaries of the middle and late

eighteenth century was very far-reaching indeed. It

was this that made the approximate uniformity that

had been achieved by printers up to the eighteenth

century very nearly absolute. The individual who

figures most largely in this consummation is, naturally,

the first really authoritative lexicographer—Samuel

Johnson.

It was no part of Johnson’s purpose to establish new

and radical principles in English spelling. The almost

contemporary dictionary of the French Academy (1762)

reformed the spellings of something like five thousand

words—more than a quarter of the whole number it

included; but Johnson’s dictionary of 1755 sets out, in

more conservative fashion, to bring order out of the chaos

of English spelling by ironing out existing inconsistencies

rather than by innovations. It is significant of his

attitude that he specifically asserted that he preferred, in

the few changes he had made, to go back to old models:

I have attempted few alterations, and among these few,

perhaps the greater part is from the modern to the ancient

practice; and I hope I may be allowed to recommend to those

whose thoughts have been, perhaps, employed too anxiously

in verbal singularities, not to disturb, upon narrow views, or

for minute propriety, the orthography of their fathers.

Johnson’s aim, then, as he set out to “ascertain”

English spelling, was primarily to make it consistent

and uniform, and preferably by conformity with long-

established custom. As he sought to control irregularity

by applying the principle of analogy, he was often forced

to compromise with the latter purpose. That he was

276 The Development of Modern English

very frequently unsuccessful even in the limited endeavor to be consistent is well known. It may be of interest to rehearse a few of the amusing inconsistencies he fell into, some of which unfortunately are fastened all but irrevocably upon the language. These may be shown by indicating pairs of words in which inconsistent spellings have been used: moveable, immovable; downhil, uphill; distil, instill; sliness, slyly; deceit, receipt; deign, disdain; install, reinstal; anteriour, posterior; interiour, exterior.

From the time of Johnson on, the dictionary has been a conservative and a standardizing agency for the spelling of the language as well as for its other aspects. Some of Johnson's mistakes and inconsistencies were corrected by later lexicographers, but many, such as deign (by the side of disdain) and receipt (by the side of deceit), were not. Though no single lexicographer of later times has enjoyed the prestige of Johnson in his era, there can be no doubt that dictionaries as a whole have, since his day, occupied a position of authority that gives unquestioned weight to their decisions on spelling. The belief that, with but the rarest exceptions, there can be but a single correct spelling for one word, and the premium placed on "correct" spelling as one of the readiest indications of the individual's education and culture, were both firmly established by the end of the eighteenth century. Spelling is no longer commonly regarded as a proper field for individuality or experimentation. This is not to imply that later lexicographers have been as ill-equipped for the task of regulating spelling as was Samuel Johnson, or that their attitudes necessarily resemble his in conservatism. On the contrary, dictionaries have often given aid and comfort to spelling reform; the early work

Spelling and Spelling Reform

277

of Noah Webster and the whole policy of the Standard are striking instances. It remains true, nevertheless, that the public's (particularly the American public's) ready acceptance of the authority of dictionaries, in spelling as in other matters, is a force that on the whole works for both conservatism and standardization.

It is time to resurvey our subject, the history of English spelling, from the point of view of the efforts, individual or otherwise, that have been put forth at various times to reform it. "Reform" in this sense need not be taken as synonymous with "improvement"; it stands, rather, for deliberate attempts, whether wisely or foolishly conceived, to change the spelling of words or groups of words that have seemed to the "reformer" in need of renovation. The two motives that actuate such efforts are likely to be interest in pronunciation and etymology. As a result, spellings are changed either to bring them into closer union with the sounds of the words in question, or to disclose more definitely their derivation (real or supposed). The duplicative Orm already referred to was an early advocate of phonetic reform. On the other hand, there arose in Renaissance times a movement to change the spellings of many English words for a very different reason: to make them conform to what was assumed to be their etymology.

It was one of the less fortunate by-products of the humanistic movement that scholars and writers should have deemed it good to respell English words in order to make them conform to Latin or Greek analogies. Often this was done with words that, though originally Latin, had been borrowed not from Latin but from French, with the changes in spelling and in pronunciation that had taken place in Vulgar Latin and Old French.

Thus debt was respelled to indicate its remote ancestor,

the Latin debitum, and doubt was similarly treated to show

a connection with Latin dubitum; the Middle and early

Modern English spellings dette, det, doute, and dout, better

both phonetically and etymologically, were discarded. In

these words, the b has never been pronounced. The p in

receipt has a similar history: Elizabethan scholars changed

the earlier receit or receyt, as they did also conceit and

deceit (though here with less permanent results). Sometimes,

the changed spelling has in turn affected the pronunciation,

as when perfect (for Chaucer's parfit) was made to look like

its Latin grandfather rather than its French father. This

zeal for etymology becomes amusing as well as deplorable

when the etymology is utterly mistaken. A poor defense can

be made, on the ground of remote relationship, for the b of

debt, the c of perfect, and the p of receipt; but even this

limited excuse does not serve to justify the s of island (not

even remotely from Latin insula or French isle, but from an

Old English igland, which became iland in Middle English and

early Modern English), or the g in foreign or sovereign. The

last two words are particularly instructive. Sovereign comes

from Old French sourain, in turn derived from the Latin

superanus [< super (above)]. It was erroneously associated

with reign [< Lat. regnare (rule)], and the mistaken analogy

not only affected sovran (so spelled by Milton), but accounts

for the spelling of foreign as well. Foreign is really from

Old French forein [from Vulgar Lat. foraneus, eventually from

Lat. foras (out of doors)].

It would be misleading to give the impression that all efforts

to renovate English spelling in the sixteenth and seventeenth

centuries were in the interests of classical

etymology. On the contrary, there were many noteworthy attempts to simplify spelling rather than complicate it, and to simplify it in the direction of a closer bond between spelling and pronunciation. One of the earliest to state the principle of purism in vocabulary was Sir John Cheke, the first Regius Professor of Greek at Cambridge. Consistently enough, he advocated reform in spelling as well as in the choice of words,10 giving preference, in both departments, to the Anglo-Saxon as opposed to the classical. Cheke's friend and Cambridge associate (like him, too, a secretary of state), Sir Thomas Smith, worked out a more systematic scheme of spelling reform, in the shape of a Latin treatise (1568) advocating the adoption of an alphabet with additional letters as well as the use of diacritical marks. John Hart (1569) and William Bullokar (1580) also published treatises on orthography—the latter a scheme, like Smith's, of a new alphabet, consisting of thirty-seven letters. In the seventeenth century, Doctor Gill (1619) and Bishop Wilkins (1633) put forth similar proposals for a reconstructed alphabet. None of these more ambitious proposals bore much practical fruit11; but simplifications, often widely adopted, in the spelling of individual words were sometimes made by writers and scholars. Such were many of the recommendations of James Howell's Grammar (1662), including the replacing

10 His best-known statement with regard to the English language (1557) is quoted in McKnight, Modern English in the Making, pp. 118 and 119.

11 A nineteenth-century analog to such schemes is that of Isaac Pitman, the creator of phonographic shorthand. Pitman's proposal, worked out in its final form in 1843, was an alphabet of forty letters, including sixteen new symbols.

280 The Development of Modern English

of logique with logic, warre with war, sinne with sin, witt with wit, and so on.

We have already glanced at Samuel Johnson’s zeal for etymology. In his case, as so often in our own day, as well as in the eighteenth century, this meant more particularly the desire to make English spelling conform to Latin and Greek analogies. “Etymology” is almost equivalent to “classical etymology,” for the scientific study of the Teutonic languages had yet to be born. When Johnson does recommend a “Saxon” spelling it is more than likely to be a mistaken one, historically. Thus, he insists on the final k in such words as critick, historick, musick, and prosaick on the ground that the truly English spelling should “always have the Saxon k added to the c.” As Lounsbury remarks,12 “The ‘Saxon k’ was the lexicographer’s personal contribution to the original English alphabet.” If such an elementary fact in the history of English orthography could be unfamiliar to him, it is no wonder that Johnson’s enthusiasm for etymology should have had curious results upon the spellings he advocated. As has already been pointed out, he preferred to recommend spellings that had already been in vogue; in practice, this meant all too frequently throwing the weight of his authority upon the side of artificial respellings, based upon real or imaginary classical analogies, that had been perpetrated by Renaissance scholars.

The rigidity of English spelling from the eighteenth century on, and its general failure to record the many changes in pronunciation that have since occurred are thus in large measure to be laid at the door of Samuel

12 Op. cit., p. 292.

Spelling and Spelling Reform

281

Johnson. Not that there has been an utter lack of

individuals and groups of people who have felt that it

is more important to have spelling indicate the sounds

of words than their "etymology." An eighteenth-

century protestant against the status quo in spelling was

Benjamin Franklin, who found time among his multi-

farious activities to compile, in 1768, a "Scheme for a

New Alphabet and Reformed Mode of Spelling."

Franklin's purpose differed from that of his contempo-

raries who were publishing spelling and pronouncing

dictionaries in that he proposed, not to make pronun-

ciation follow spelling, but rather, by means of a phonetic

alphabet, to alter the spelling to suit the prevailing

pronunciation. This radical reform, however, proved

abortive, for Franklin felt "too old to pursue the plan,"

and his proposals were never published. It was left

for another American to make the first effective move

against the system largely inherited from Johnson.

This was Noah Webster, the most important lexicogra-

pher since Johnson. His advocacy of simplified spelling

has been chiefly responsible for the accepted differences

between British and American spelling. Some of his

activities may repay more detailed attention.

In early life Webster was strongly influenced by

Benjamin Franklin, though not to the extent of adopting

the phonetic system which Franklin attempted to have

him carry through. The extreme nationalism prevalent

in many American circles in the last decades of the

eighteenth century is voiced in Webster's American

Spelling Book, 1783: "For America in her infancy to

adopt the present maxims of the old world, would be to

stamp the wrinkle of decrepit age upon the bloom of

youth, and to plant the seed of decay in a vigorous

constitution." With increasing age, however, and—perhaps it may be added—with the increasing prospect

of appealing to a British as well as an American public, Webster's patriotism became less flamboyant. Never-

theless, though in his great dictionary, the first edition of which came out in 1828, he recognized the fact that

"the body of the language is the same as in England," he recommended many spellings that deviated from

British practice. Later editions of the dictionary and later editions of the spelling book (the latter also an

enormously influential work, for more than fifty million copies were sold) are alike much more conservative than

the early ones. From a reformer of orthography Webster had become content to be a reformer in orthography.13

Still, the milder type of reformation was sufficient to cause violent protests, and Webster (and his later editors)

often receded before the storm. Some of the simplifications that were retained and that are now no longer

questioned in American usage are the -or rather than the -our ending in such words as honor, labor, and the like;

the -er rather than the -re in center, meter, and so forth; and the single l (or other consonant) in such words as

traveled, traveling, and so on.

Webster's simplifications were not necessarily phonetic simplifications. Two other motives took precedence,

indeed, over any attempt at phonetic shortening: the desire for etymological correctness, and the desire for

uniformity. Distrusting the etymological lore of his predecessors as worthless, he evolved a new theory

of the relationship of languages for himself and quite largely, in Sir James Murray's words, "out of his inner

13 The terms are Lounsbury's (p. 240).

Spelling and Spelling Reform

283

consciousness." The results, of course discarded in later editions of the dictionary that bears his name, were quite naturally devoid of value, and where, as was frequently the case, they affected the spellings recommended, they could only be deplored. Orthographic changes that he introduced with the idea of consistency and uniformity in mind (as opposed to etymological theory) must be somewhat differently regarded. Changes of this kind frequently removed such absurdities as those in which Doctor Johnson's dictionary abounded, and they have not seldom been approved by later usage.

On the whole, however, it must be said that the spelling reform achieved by Noah Webster was not on a great scale. Webster lacked, after all, the firmer basis for both phonetic and etymological simplification that the work of later generations of linguistic scholars was to supply. It was not until the last decades of the nineteenth century that the science of linguistics achieved anything like its present status. From scholars trained in this newer school, the recent and contemporary movements aiming at spelling reform have come. The activities of Isaac Pitman and A. J. Ellis just before the middle of the century14 were a forerunner and a symptom of the temper of the scholarly world on spelling;15 but the first concerted step was the action of the American Philological Association in 1875 in appointing a committee to consider the whole matter of spelling reform. From this, and similar action on the part of the Philological Society (of England), sprang the Spelling Reform Association and the British Spelling

14 See Handbook of Simplified Spelling, Part I, pp. 9-12.

15 W. D. Whitney published in 1867 several articles advocating a thorough reform in spelling.

284 The Development of Modern English

Reform Association. In 1906, the Simplified Spelling Board was organized to conduct the more active propaganda financed by Andrew Carnegie, and in 1908 the Simplified Spelling Society was formed in England. Since all three of these societies—Spelling Reform Association, Simplified Spelling Board, and Simplified Spelling Society—are still (1934) in existence, it may be of interest to review briefly their aims and achievements.

The somewhat different tactics that underlie the proposals of the three organizations are succinctly stated in the periodical Spelling,16 which is issued by them jointly. They may be still more briefly summarized here in the sentences that follow. The Simplified Spelling Board believes that progress may best be achieved by seeking, as the first step, the widest possible adoption of a very limited number of new spellings, on the theory that a campaign of gradual education is necessary before new spellings so numerous as to transform the appearance of the printed page can conceivably be accepted by the general public. The Simplified Spelling Society would go further: while keeping the present alphabet of twenty-six letters, it recommends for a teaching device in elementary schools an entirely new and consistent scheme of using the existing symbols, with no additional letters or diacritics. (It apparently recognizes a distinction between a revised spelling as a method in general education and a revised spelling for general use; for the latter, so far as the present is concerned, such comparatively small changes as the Simplified Spelling Board proposes are all that can hopefully be attempted.) The Spelling Reform Asso-

16 Vol. I, No. 1 (March, 1925), and Vol. 2, No. 5 (March, 1931).

ciation accepts no compromise with the ideal goal:

it stands for thoroughgoing renovation, as opposed to

merely partial simplification; accordingly it advocates

a complete phonetic alphabet, with such new letters

as a minimum-precision, one-sign-one-sound alphabet

may require.17 It is of course quite possible for one

individual to sympathize with all three of these purposes

—to accept the first as a practical initial stage, the

second as an educational measure, and the third as the

ultimate objective. In practice, it seems necessary to

add, the three schools of thought on spelling reform are

not so sharply divided into water-tight compartments

as this account of the strategy of each might seem to

imply.

Not only are the three principal organizations for

spelling reform in England and America in large measure

united for a common purpose; more than this, they have

joined in welcoming the most recent recruit to the cause

of English spelling reform. This is the Swedish project

known as “Anglic,” primarily a scheme to make English

more adaptable for world-wide use by renovating its

spelling. Anglic spelling is not phonetic, nor does it

add new letters to the alphabet. It aims to bring order

out of the general chaos of conventional spelling by

“generalizing the most common or serviceable of the

existing spelling variants, introducing at the same time

a few new digraphs (uu, dh, and zh), which have to do

17 In 1928, the Spelling Reform Association proposed a phonetic alpha-

bet “for immediate use in the teaching of English, particularly in the

elementary grades, and as a practical working basis for eventual phonetic

spelling reform of English for general use.” It was adopted also by the

Simplified Spelling Board “for immediate use as a phonetic key alphabet

for English wherever respelling to indicate pronunciation is required.”

(Pamphlet of May, 1928.)

286 The Development of Modern English

the duty of new letters."18 Though Anglic obviously runs counter in important particulars to the purposes of

the older organizations, it was nevertheless endorsed by representatives of all three of them (and of the

International Phonetic Association as well) at the World English conference held in London in June, 1930. The

conferees agreed "to give the Anglic movement their full support and coöperation . . . both as an inter-

national auxiliary language and as a basis for reform of English spelling for the English-speaking world."19

It should be added, however, that the Simplified Spelling Society, whose program is more directly in rivalry with

that of Anglic, is perhaps less enthusiastic about Anglic than are the other two organizations. Sir G. B. Hunter,

chairman of its committee, states that the committee "have considered, but do not advocate, Professor

Zachrisson's Anglic as an improvement of our spelling; though some of the most eminent members of the com-

mittee approve and are willing to adopt Anglic."20

In summarizing the recent history of spelling reform, considerations of brevity make it necessary to limit our

attention chiefly to the activities of one organization, the American Simplified Spelling Board. It has already

been indicated that the parent society, the Spelling Reform Association, and the sister society in England,

the Simplified Spelling Society, have not been quiescent.

18 Pamphlet published by the Anglic Fund, Uppsala, Sweden, 1931. For a fuller statement, see the larger booklet, Anglic, a New Agreed Simplified English Spelling, also published by the fund (1930). Discussions of Anglic are contained in the various numbers of Spelling for 1931; in American Speech, Vol. VI, No. 5 (June, 1931), pp. 378-380; and in the Bookman, February, 1931 ("Or Shall we go Anglic?" by Mrs. Aiken).

19 Spelling, Vol. II, No. 5 (March, 1931), p. 7.

20 Spelling, Vol. II, No. 6 (June, 1931), p. 19.

Spelling and Spelling Reform

287

The recent memorial of the last named (June, 1932) to the British Board of Education is an evidence of its continued activity. This was a petition, bearing the signatures of some seven hundred members of the faculties of British universities, for an inquiry into the simplification of English spelling. The plea was based on the conviction that the increasing world-wide use of English is chiefly hindered by "its inconsistent and difficult spelling." But the most conspicuous activity in spelling reform during the present century undoubtedly has been that which followed the formation of the Simplified Spelling Board in 1906. To that activity we now return.

The very title of the new organization implied that the strategy of the campaign against the conventional spelling had been modified. It was no longer to be "reform," but "simplification"—obviously a concession to the strength of popular prejudice and conservatism. But even the policy of "cutting off the dog's tail an inch at a time instead of all at once"21 did not disarm hostile criticism on the part of public and press. When President Theodore Roosevelt, with characteristic impetuosity, officially announced his conversion to the cause by ordering the Public Printer to use in Government documents the new spelling of certain words, a most amazing storm broke. Congress threatened to withhold the appropriation for the printing of executive department publications if the order was not countermanded.

21 Handbook of Simplified Spelling, Part II, p. 39. In this policy, all adherents of spelling reform, quite naturally, did not, and do not, concur. Sir G. B. Hunter has recently used this very phrase in a derogatory sense, Spelling, Vol. II, No. 4 (December, 1931), p. 59; he concludes, "Save us from trusting to gradual improvement."

288 The Development of Modern English

The President was forced to limit his reform to White House correspondence. On the whole, the controversy that ensued probably did the cause of simplified spelling more harm than good. The movement had more publicity than it has had before or since, in the memory of those now alive, but it was publicity of an unfortunate kind.22 Simplified spelling, if not exactly considered a mere foible of Roosevelt's, was all too generally regarded as the creation of a few misguided cranks. Newspapers ordinarily dealt with it only to misrepresent and ridicule it. An unusually witty reference to it was the famous single-word editorial of the New York Sun on the occasion of President Taft's inauguration in March, 1909; glancing not at the new but at the old president, the editor was able to cram an astonishing amount of malicious satisfaction into the four letters: "Thru."

In the years that have followed the first campaigns of the Simplified Spelling Board, progress has been, it must be confessed, disappointingly slow. Dictionaries, especially the Standard, have given aid; but the public acceptance of dictionary permission has been somewhat surprisingly hesitant. Newspapers, though some have lent partial sanction to the proposal of the Board, have on the whole maintained their original attitude of contemptuous, and frequently thoroughly ignorant, criticism. Only a few weekly or monthly periodicals—notably the Literary Digest—have adopted a real measure of simplification in their own columns. Many educators have given complete or partial assent to the program outlined by the Board, but many more are ultraconservative, to the extent of refusing to accept

22 A lively account of the controversy will be found in Mark Sullivan's Our Times, Vol. III.

from their pupils even the milder deviations from spelling orthodoxy. The National Education Association in 1907 and again in 1916 adopted certain of the Board's recommendations, but in 1921 withdrew a large measure of this sanction. Last, but perhaps most important, the man on the street has in large measure inherited an attitude of complacent and ill-informed ridicule of the whole project, an attitude which the newspapers of the last quarter-century have done a great deal to foster. A dispassionate observer must be struck, in retrospect, with the spectacle of small results achieved by great and valiant efforts.

Turning from the history of spelling reform to the arguments for and against it, one's first impulse is to assert that almost everything can be said for it, and exceedingly little against it. Yet it is clear that the somewhat less tangible and certainly less logical objections that have commonly been urged against it are nevertheless very powerful. Only thus can its small measure of success be accounted for. To simplify our summary of the case for spelling reform and the case against it, it will be necessary to limit our attention chiefly to the program that has been outlined by the Simplified Spelling Board. As a preliminary step, the recommendations that the Board advocates for immediate adoption must now be summarized. They have been drawn up as the following four principles:23

23 Handbook of Simplified Spelling, Part I, pp. 18 and 19, with the omission of some of the examples. Here, as elsewhere in the chapter (when matter in reformed spelling has been quoted), I have altered the spelling to conform to that used in this book generally. It is, however, an indication of the small extent of the changes recommended for immediate adoption that, of seventy-eight words quoted, I have changed only three (fonetic, thru, and ar).

290 The Development of Modern English

1. When current usage offers a choice of spellings, to adopt the shortest and simplest. Examples: catalog, not catalogue; center, not centre; gage, not gauge; honor, not honour; maneuver, not manoeuvre; mold, not mould; quartet, not quartette; rime, not rhyme; tho, not though; traveler, not traveller.

2. Whenever practicable, to omit silent letters. Examples: activ, not active; anser, not answer; definit, not definite; det, not debt; frend, not friend; hart, not heart; scool, not school; shal, not shall; thru, not through; yu, not you.

3. To follow the simpler rather than the more complex of existing analogies. Examples: aker, not acre; buro, not bureau; deciet, not deceit; enuf, not enough; spritely, not sprightly; telefone, not telephone; wize, not wise.

4. Keeping in view that the logical goal of the movement is the eventual restoration of English spelling to the phonetic basis from which in the course of centuries and through various causes it has widely departed, to propose no changes that are inconsistent with that ideal. [No examples of the last principle are given. It may, however, be illustrated through the fact that elsewhere in the Handbook24 the Board deprecates such spellings as fite, lite, and nite (for fight, light, and night) on the ground that, though shortenings, they are unphonetic shortenings and would perpetuate a wrong principle of notation: the use of final silent e to indicate the quality of the preceding vowel. On the other hand, to restore delite and spritely is a somewhat different matter; these spellings are recommended not as innovations but as older and better spellings than the present ones.]

It is necessary to add that the arguments advanced by the Board refer not only to the minor changes recommended for immediate adoption, but also to the completely phonetic spellings that are sought as the eventual goal. Some of the chief claims made for simplified spelling, then, are as follows: it will save much time and expense in elementary education, and in writing,

24 Part II, pp. 5 and 6.

Spelling and Spelling Reform

291

typewriting, and printing; it will improve and tend to standardize pronunciation; it will remove the greatest barrier to the Americanization of our foreign population and to the use of English as an international language.25 These three propositions will be considered separately and in the light of other arguments growing out of them.

The practical gain in time and money would doubtless be enormous. Dr. Godfrey Dewey, secretary of the Board, has computed that the present conventional spelling (as compared, in this instance, with a single one-sound-one-symbol phonetic alphabet) wastes at least a billion dollars a year,26 and consumes a full year of the school life of every English-speaking child. As to the latter point, it is arguable that even worse than the loss in time is the needless distraction and "atrophy of logical faculties" experienced by the child. Many another father has echoed these sentiments of the great sociologist, W. G. Sumner:

I have two boys who are learning to spell. They often try to spell by analogy, thus using their brains and learning to think. Then I have to arrest them, turning them back from a rational procedure, and impose tradition and authority. They ask me, "Why?" I answer, "Because your father and others who have lived before you have never had the courage and energy to correct a ridiculous old abuse, and you are now inheriting it with all the intellectual injury, loss of

25 Most of the eight "reasons for simplified spelling" as advanced in the Board's pamphlet "Rules and Reasons" (1919) and elaborated in the Handbook, Part II, are here summarized.

26 The unsympathetic may feel that the reductio ad absurdum of this type of argument is the statement of Professor Zachrisson that to English spelling are due "increased unemployment and the going down of trade not only in England and the United States but in the whole world" (Anglic, p. 18).

292 The Development of Modern English

time, and wasted labor which it occasions. I am ashamed that it should be so."27

The new spelling, which would certainly be more easily learned than the old, would be an infinitely better index to pronunciation. If it be objected that spelling that followed pronunciation would vary as pronunciation varies, the retort is of course obvious: that the ideal of uniform "correct" spelling is an utterly false one. On the other hand, it is plausibly maintained that variation in pronunciation would be lessened as the corresponding variation in spelling made it more conspicuous. The effect of phonetic spelling would quite conceivably be to center attention on pronunciation in a new way, to promote discussion of the variants in current use, and to work toward a greater standardization after the pattern of the most approved models. Complete standardization of pronunciation would never take place, nor does it seem desirable that it should.

Finally, there seems no good reason to doubt that the spelling of English is the one great barrier in the way of its acceptance as the international auxiliary language, and that the same cause most effectively hinders the operation of the melting pot in the United States. On the subject of English as a world speech, it is perhaps worth recalling that long ago the great German linguist, Jacob Grimm, congratulated other Europeans that the English had not yet discovered that only one thing prevented the universality of English from being completely apparent: its "whimsical, antiquated orthography." Today, with the English language perceptibly nearer the indicated goal, the great majority of the students of language would certainly agree that it

27 Spelling, Vol. I, No. 1 (May, 1887), p. 16.

Spelling and Spelling Reform

293

is still the same handicap that prevents English from attaining it completely. It is of course with this in mind that the campaign for Anglic has been instituted.

If these arguments have any great measure of truth, what can possibly be said on the other side? What is most commonly urged is undoubtedly what the Board, perhaps justifiably, refers to as the "etymological bugaboo." This is the notion prevalent among amateur defenders of the status quo in spelling that to change it would be to "destroy etymology"; what they really mean is that it would obscure the derivation of words. No scholar of repute any longer holds this point of view.

It is necessary, indeed, to go back to Archbishop Trench's English Past and Present (1855) to find a presentation of it by any competent writer; and Sir James Murray expressed, in 1880, the opinion that Doctor Trench himself "if leisure had been given him to keep pace with the progress of science, . . . would now have been second to no one as a spelling reformer." The belief voiced by Trench (and to quote it is to voice the argument in its most nearly respectable form) is that to retain the etymological spellings of words is to help a large class of readers—but neither the grossly ignorant nor the accomplished scholars—because to keep these spellings is to enable such readers "to recognize the words which they are using, whence they came, to what words in other languages they are nearly related, and what is their properest and strictest meaning."

We cannot take the space that would be required to examine all the fallacies stated or implied in such a dictum, but, since the belief is still widely held, it may be worth while to indicate some of them.

294 The Development of Modern English

The existing spelling, in the first place, is a most unreliable index to the derivation of the word. This has already been demonstrated, and it has likewise been suggested that to simplify the present spelling would very frequently restore an earlier and a more nearly etymological, as well as a more nearly phonetic, spelling. Further, the help afforded by the present spelling (even if etymologically defensible) is useful only to those acquainted with the languages from which the words are derived. Archbishop Trench was thinking of those readers-proportionately a much smaller group now-whose knowledge of Latin and Greek would be useful for their understanding of English. On this basis, spelling is to be kept safe for only a very limited fraction of our democracy. But if as further prerequisites for the full appreciation of our present spelling we add acquaintance with such other important studies as Old English and Old French, it is evident that the minority has dwindled very much more. Yet, how can anyone defend a spelling on the ground of its being etymologically correct and etymologically revealing unless he knows the etymology involved? The advertising slogan "Patronize your naborhood store," may be objected to on other grounds, but does nabor really mean less than neighbor unless one knows that neighbor literally is "nigh-dwell-er"? The chances are that a negligible fraction of objectors to this spelling have this information. The "etymological bugaboo," on the whole, has very little indeed to commend it. Professor Skeat, the editor of the best of etymological dictionaries, has appropriately disposed of it in the observation: "In the interest of etymology we ought to spell as we

Spelling and Spelling Reform

295

pronounce. To spell words as they used to be pronounce. To spell words as they used to be pro-

nounced is not etymological, but antiquarian.'

More weight may justly be given to a second objection

to spelling reform: the fact that many homophones

would no longer be differentiated on the printed page.

Doctor Bradley, by no means an opponent of spelling

reform, has pointed out how very commonly "difference

of spelling, once phonetically significant, has become a

mere ideographic device."28 The useful distinction

(to the eye) between such pairs of words as night, knight;

son, sun; and bare, bear would be lost. It is true that

there would be a certain compensation for this in the

visual differentiation of words now spelled alike but

pronounced differently, such as the present and past

tenses of read; bow meaning "knot" and "fore part of

boat," tear meaning "to rend" and "water from the

eye," and so on. At the same time, it is evident that

this second list is far more difficult to prolong than the

other. The proponents of spelling reform are perhaps

in general too prone to neglect the gulf, awkward as

it is from more than one point of view, that simply must

be recognized as existing between spoken and written

English. As Bradley further points out, a great many

words one encounters on the printed page are but rarely

pronounced; the first, and almost the sole, function of

their spellings is to convey meaning to the eye and mind

of the reader. This they do without calling up the

question of pronunciation at all. There is some reason,

therefore, to feel that frequently more would be lost

than gained if words now spelled differently, though

pronounced alike, were made identical in appearance.

28 Op. cit., p. 11.

296 The Development of Modern English

A kindred objection is that the connection between words related in meaning would be obscured by a phonetic spelling. Vowel sounds are almost invariably altered when the accent shifts from one syllable to another: observe what happens to all three of the vowels of photograph when the word is made into photographer.

Yet it is a practical convenience that the two words should look alike and be recorded together in dictionaries, as should also history and historical, injury and injurious, and so on. Sometimes the pronunciation is altered even by an ending that does not shift the primary stress of the word; conspire and conspiracy, nature and natural, and zeal and zealous afford instances.

To use different symbols for the vowels in cases like these would confuse children and foreigners learning the language. There is a good deal to be said (apart from strictly "etymological" considerations) for a spelling that makes words that are alike in meaning look alike even though they have developed widely different sounds.

So far as that goes, there is something even in the "etymological" argument of the anti-reformers—especially if it is limited to words that appear only on the printed page and that carry their meaning to the eye of the educated reader, who alone is concerned with such words.

R. W. Chapman has well pointed out how much would be sacrificed by a phonetic spelling of such words as adolescent, chryselephantine, hypotenuse, and ichthyophagous.29

Still, there is danger that the plea for the retention of familiar spellings on the ground that they are familiar (and even, in one way or another, logical) may blend

29 In his essay, "Thoughts on Spelling Reform," in The Portrait of a Scholar, pp. 89–108. He reinforces many of Bradley's points, and is, like him, not unsympathetic with a certain measure of reform.

Spelling and Spelling Reform

297

into the less defensible position that they should be retained because they are beautiful. One aspect of our prejudice in favor of things familiar is our conviction that there is a superior beauty and appropriateness in such labels as are already established. Professor George Saintsbury’s well-known outburst of exasperation30 takes dangerous ground when the writer remarks, “If the pestilent folk who call themselves Spelling Reformers or Simplifiers ever thought of the beauty of words—which, I believe, they honestly profess either to know nothing about or to disregard entirely—it would be worth while offering them a crux or two in the department of proper names.” Yet, even in the department of proper names—which actually, be it said in passing, the spelling reformers leave severely alone—it is clear that many of Mr. Saintsbury’s preferences are purely subjective. As he is quite aware, his dislike of “Ann” and admiration for “Anne” are not shared by everyone; and, as he himself admits, Thackeray liked “Anny,” which Mr. Saintsbury thinks hideous (though he thinks “Annie” quite pretty). This finding of beauty and ugliness, particularly the former, in the mere outward forms (or even the sounds) of words is often a most curious delusion. As an antidote to Mr. Saintsbury, one should read Max Beerbohm’s delightful essay on “The Naming of Streets.”31 No demonstration of the position that “what you take to be beauty or ugliness of sound is indeed nothing but beauty or ugliness of meaning” could possibly be more convincing than his calling attention to the “beauty” of ermine, manor house.

30 In his Last Scrap Book, New York (Macmillan), 1924.

31 Yet Again, London (Heinemann), 1930, pp. 193–207.

298 The Development of Modern English

gondola—and then exhibiting the “ugliness” of vermin, warehouse, scrofula! To be sure, in saying this Mr. Beerbohm is not primarily concerned with spelling, but many are undoubtedly convinced that the appearance of the first group is as beautiful as the sound is.

In considering the “esthetic” argument against spelling reform we have turned from the logical to the illogical form of objection. At the same time, this delusion that the familiar is appropriate and beautiful and the unfamiliar ugly (or at least comical) is clearly very deeply rooted and extremely influential. Many of those who dislike simplified spelling do so for no better reason than that it “looks queer” and is therefore distasteful to them. To the objection, “I don’t like it,” the Spelling Board admits inability to reply.32 Frequently too, this type of opposition is reinforced by mere inertia. Some who might overcome their distaste for the queer appearance of reformed spelling are kept by both conventional and lazy habits of mind from doing so. After all, they reason, the benefits of the reform would be for the succeeding generations, not for ours. Why should we make matters smoother for our children than we have found them—particularly as it would entail our living in an era of readjustment and distracting change? To this unheroic but exceedingly human mental attitude would seem to be due a large part of the responsibility for the slow progress of spelling reform—precisely the same factor that prevents speedy progress in other fields. Perhaps it is not too pessimistic to say that this one cause chiefly accounts for what may be called the glacier-like speed of the movement, though

32 Handbook, Part II, p. 38.

Spelling and Spelling Reform 299

after all there may be some doubt as to whether that term is fully appropriate: a glacier is slow but eminently sure, and one may reasonably be uncertain as to whether, even eventually, spelling reform will attain a real measure of success.

To sum up. It has been contended that the logic of the case is chiefly (but not quite exclusively) on the side of the advocates of systematic spelling reform. Not, however, that they are altogether the children of light, and their opponents the children of darkness. There is a good deal of Samuel Johnson's characteristic common sense in his observation that, "Some have endeavored to accommodate orthography better to pronunciation, without considering that this is to measure it by a shadow, to take that for a model which is changing while they apply it." Amplified, this is a just and permanent argument against a thoroughly phonetic spelling. We have questioned, particularly, the reformers' assumption that spelling exists solely to indicate pronunciation; and we have found support for our contention in the temperate yet incisive comments of Henry Bradley and R. W. Chapman. Further, it seems fair to say that there has sometimes been an excess of zeal in the terms used by the reformers to describe the absurdities and anomalies of conventional spelling.

Sir William Craigie points out in the preface to his English Spelling, Its Rules and Reason that the impression that "English spelling is a hopeless chaos" is one that can and should be corrected; yet the enthusiastic reformer would find that description of it unsatisfactory only because he thought it too mild. The late Professor Joseph Wright, successor to Max Müller in the chair of comparative philology at Oxford, has likewise observed that "our

300 The Development of Modern English

English orthography . . . far from being devoid of law and order . . . is considerably more systematic than would appear at first sight," and that the "ordinary general reader" is mistaken in thinking of it as "a thing born of ignorance, grown up haphazard, and existing by pure convention without rhyme or reason for its being, or method in its madness."33 These quotations have been included in order to indicate that at least some highly qualified students of the language have felt that the depreciation of our spelling familiar to the ordinary user of it (and still more to most scholars) is very frequently exaggerated. The final evidence that the orthodox spelling is not quite hopeless would seem to be the fact that Anglic, described as achieving "most of the practical advantages of complete and exact spelling reform," can still leave unchanged "sixty to seventy-five per cent of the words on an average printed page."34

The foregoing pages have been quite misinterpreted, however, if they have been understood to assert that in general the case for spelling reform is weaker than the case against it. On the contrary, what has been said to disqualify or modify the position of the spelling reformers and what has been said to reinforce some of the objections of their opponents has been said simply because the latter position so evidently is on the weaker side. Spelling reform—more specifically the program of the Simplified Spelling Board—the writer believes to be a cause amply fortified by reason; the fact that it has had only slight success he likewise believes to be due much more to ignorant prejudice and sheer inertia

33 Preface to Wright, Joseph and E. M., An Elementary Historical New English Grammar, London (Oxford), 1924.

34 Pamphlet published by the Anglic Fund, op. cit.

Spelling and Spelling Reform

301

than to any serious weakness of its own. When every

just deduction for overstatement on the part of the

reformers has been made, the controversy still appears

very largely that of reason against folly, and the account

of it must be considered a melancholy narrative that

reflects exceedingly little credit upon human intelligence.

One more question remains: What can the individual

writer who believes in the advantages promised by

spelling reform do to promote its cause?

It is a curious commentary on the whole matter that

some of the most zealous of spelling reformers have

confined themselves to preaching, and have practiced

what they preached not at all. Says Doctor Bradley,35

" . . . even Professor Skeat, one of the most impas-

sioned advocates of reformed spelling, seems never, in

his published writings, to have ventured to spell any

single word otherwise than in the conventional fashion,

nor, if I may judge from his many letters to me, did he

do so in his correspondence." Similarly, one of the most

militant of our American reformers, Professor Lounsbury,

accuses himself in these terms: "Were I not myself

inconsistent and lazy and several other disreputable

adjectives, I should write fonetic instead of phonetic.

This I cheerfully admit."36 If "impassioned advocates"

of the cause thus hesitate to add practice to precept, even

in the smallest and least questionable changes, need

they be surprised that there is but small response to their

exhortations? For it is clear that the man in the street

has even more cause than the scholar to hesitate: from

him "reformed" spellings would be interpreted (as they

35 Op. cit., p. 33.

36 Op. cit., p. 287.

302 The Development of Modern English

would not from Lounsbury or from Skeat) not as advanced, but merely as illiterate.

Granting that reform must come gradually, as the Simplified Spelling Board believes, how many and what sort of innovations can the individual use without subjecting himself to violent criticism? It is to be admitted at once that the principle of changing first what most needs changing will scarcely apply. Proper names are, from the point of view of the reformer, among the least satisfactory, orthographically, of all classes of words. Yet even the extreme proposals of the Simplified Spelling Society stop short of meddling with their spellings. Surnames like Jonson, Johnson, More, Moore, Spencer, and Spenser are (in this instance, in the history of English literature) identifying labels in their present spellings; to change these spellings would be to confuse the identities they represent. Geographical terms likewise have a traditional background that it seems, even to the reformer, inexpedient to disturb.37

No, it is not the principle of first attacking what are phonetically the worst offenders that can be adopted; rather, it is the policy of making the initial changes those that are least conspicuous and hence least disturbing.

In practice it is admittedly not easy to follow this policy. Let us see how it applies to the four general proposals of the Board that have already been quoted. In doing this, we shall be outlining an attitude toward reformed spelling that compromises with the compromises of the Board. The first principle, to use the short-

37 For a solution of the varied and complicated problems involved in the spelling of foreign geographical terms in an English context, see the excellent "First Report on Foreign Geographic Names," published by the United States Geographic Board (Washington, 1932) and available, for ten cents, from the Government Printing Office.

Spelling and Spelling Reform

303

est and simplest of any variants sanctioned by current usage, is one that, in theory at least, can be questioned only by the confessedly reactionary. Even here, however, there is occasionally some doubt (which dictionaries, with varying policies toward simplified spelling, do not completely resolve) as to just what choice current usage does in fact permit. Still, it is clear that no one need hesitate to employ center, check, gram, honor, license, maneuver, plow, quartet, or traveler. To use any of these spellings is merely to adopt what has long since become, in American practice at least, the better orthography.

One may sympathize, indeed, with the feeling that there is something suspiciously like Anglomania in an American's use of centre, cheque (or checque), gramme, honour,38 quartette, and so forth. But it unfortunately does not follow that spellings that would seem to be completely analogous have actually that status: though the writer prefers the newer forms, he must admit that theater has not quite the standing of center, nor cigarett that of quartet. There is a widespread objection, too (which again the present writer does not share), to catalog rather than catalogue, and rime rather than rhyme. On the whole, however, the illustrations the Board provides for its first principle give few qualms even to the conservative-minded.

It is no very bitter dose, after all, to swallow even tho, the least generally used, perhaps, of the spellings included in this group, but still one that is unquestionably sanctioned by the test of present usage.

When we turn to the second and third principles,39 however, we realize that we are on different ground. If

38 For a most elaborate discussion of the often misstated issues involved in words of this type, see Lounsbury's chapter "The Question of Honor" (op. cit., pp. 194–237).

39 See p. 290.

304 The Development of Modern English

we use the spellings recommended under these heads,

we are not merely progressive; we need to be pioneers,

active crusaders. Perhaps only thru, of all the examples

given in the two lists, will pass muster—in the sense of

being sufficiently inconspicuous, by reason of a fairly

widespread use, and therefore not unduly disturbing.

To many, thru (like tho) is something of a shibboleth.

It is theoretically better than through, and sufficiently

sanctioned in practice; yet somehow one shies away from

it. As to the other words, human inconsistency and the

common lack of even a small measure of courageous

enterprise are all too clearly exemplified in the general

rejection of spellings that are both simpler and historic-

ally preferable like det and tung and those that involve

such an apparently innocuous simplification as the drop-

ping of the silent, and completely misleading, e of suffixes

like ite and ive (e.g., definit, favorit, activ, positiv).

Of the third principle, to follow the simpler rather

than the more complex of existing analogies, perhaps a

specific example should be commented on in some detail.

Let us consider the recommendation to substitute f

for ph. The words affected are all of ultimately Greek

derivation and the “etymological” spelling would hence

be ph. But this ph has been superseded by f in a few

English spellings that are unquestionably standard,

such as fancy and fantasy (from the same Greek source,

which may be transliterated as phantasia). There is

therefore good precedent for making the corresponding

simplification in many other words of Graeco-Latin

descent, as is done indeed in a more phonetic language

like Italian, in which filosofia corresponds to English

philosophy. Nevertheless the “simpler analogy” has

been invoked in vain, save in rare instances. Fantom

Spelling and Spelling Reform

305

has a certain vogue, and so, in scientific work especially, has sulfur, but that is about all. The same writer who uses sulfur would hesitate at fosforous, where the f appears twice and, through its use as initial symbol especially, is more conspicuous. Again it would appear that the conspicuous changes are the difficult ones.

Lounsbury, as we have seen, was apologetic about his failure to write fonetic for phonetic. Small wonder, then, that a spelling that changes both initial and final symbols, like fonograf for phonograph or fotgraf for photograph, appears to the ordinary writer quite unthinkable. One’s comment on both the second and third proposals must take the form of expressing regret that so little has been accomplished, or seems likely to be accomplished, in either of these directions.

The fourth and last general recommendation of the Board is a negative one: in brief, the caution that new spellings should not be mere simplifications, but simplifications that are consistent with the eventual purpose of restoring spelling to a phonetic basis. It is a caution that is necessary and useful. Amateur enthusiasts for spelling reform frequently seem to prefer the unauthorized to the authorized innovations. They indulge, for example, in such undesirable novelties as lite for light, nite for night, or (still worse) thot for thought. The common-sense warning is very evident: If you wish to be progressive in spelling, be sure that the new forms you use have a really authoritative basis. Look them up, for example, in the Standard dictionary, which includes the recommendations of the Simplified Spelling Board (sometimes as alternatives, equally recommended, to the conventional spellings), or study them in the Handbook of the Board. Even so, you will not be free from

criticism, nor will your motives, in all probability, be

clear to every reader.

The last consideration brings us once more to one of the

central difficulties, so far as the individual is concerned,

in the whole problem of spelling reform. He may do

successful battle with the inconsistency, conservatism,

and laziness of his own nature, and conscientiously and

consistently follow a definite program of reform, such

as that recommended by the Board. This is much to

grant; but even if he is successful so far, it is quite certain

that he will not be applauded. His motives indeed will

probably be labeled a desire to be conspicuous or, at

least, "different." If he is writing for publication, he

will be thwarted by the printer. In the end, it is quite

likely that he will give up all but the most minor items

in his program.

It is a melancholy note on which to end this account

of the struggles for a more rational spelling. Greater

optimism, however, is scarcely possible for even a sym-

pathetic observer who sees, rather, in the more systematic

attempts at reform during the last quarter-century,

only a slight acceleration of the speed with which the

written symbol, hopelessly outdistanced, has always

pursued the spoken sound. The inequality of the race

is due both to the unfair handicaps imposed upon the

pursuer (so far as English is concerned) in its early stages,

and to the greater suppleness of the quarry. Perhaps

the last point further implies that the small success

of systematic spelling reform in English is an illustration

of the extreme difficulty, verging on impossibility, of

bringing about changes in language from above and from

without, rather than from below and from within.

Better spelling, so far as the present outlook goes, seems

Spelling and Spelling Reform

307

likely to come about almost as slowly and painfully and

in almost the same small measure as it has in the past.

It is to be hoped, however, that time will prove this view

too dark, and that the portents are really more propitious

than they appear to be. In the meantime, even the

minor changes referred to above, which are, taken all

together, not so inconsiderable after all—those that the

individual can make in his own writing without exposing

himself to very severe criticism—will be in some degree

helpful. At the very least, the student of the language

can—perhaps one may say, should—refrain from the

criticism that is largely born of ignorance, and maintain

an attitude of tolerance, if not of sympathy, as he regards

the activities of more enterprising souls than himself.

References for Further Reading

Bradley, Henry, On the Relations between Spoken and Written

Language, with Special Reference to English, London

(Oxford), 1919.

Chapman, R. W., “Thoughts on Spelling Reform,” The

Portrait of a Scholar, London (Oxford), 1922.

Craigie, W. A., English Spelling, Its Rules and Reasons, New

York (Crofts), 1927.

Handbook of Simplified Spelling, New York (the Simplified

Spelling Board), 1920.

Lounsbury, T. R., English Spelling and Spelling Reform, New

York (Harper’s), 1909.

Pedersen, H. (trans. by J. W. Spargo), Linguistic Science in

the Nineteenth Century, Cambridge, Mass. (Harvard),

Spelling (1887-1894), official organ of the Spelling Reform

Association.

Spelling (1925 and 1931), published jointly by the Simplified

Spelling Society, the Simplified Spelling Board, and the

Spelling Reform Association.

CHAPTER IX

Sources of the Vocabulary

IN the chapters just preceding, we have been concerned with words chiefly on the side of their sounds and their forms. The study of the vocabulary implies that primary attention will be given to a more vital aspect—their meanings. It may well be argued that to this aspect every other should properly be subsidiary. There can be little question, at any rate, that this is the most human and the most popular approach to the study of language. However, it is to be granted at once that the meaning of the word should not be isolated from its history. For anything like an adequate introduction to the study of vocabulary, we must consider also the origins of our word-stock and the various processes of word-formation. These two subjects would seem, indeed, to demand attention even before we approach the study of meaning (and change of meaning). With some overlappings, all the words in our dictionaries may be said to fall into one of two classes: they are either created or borrowed. The formation of the English vocabulary may thus be surveyed from two points of view. Here it will be convenient, though in some ways it is less logical, to discuss first the various sources of word-borrowing, together with the relation of the native and borrowed elements in the vocabulary, and to reserve for later treatment the subject of word-formation. Both

308

Sources of the Vocabulary

309

topics will pave the way for subsequent more detailed attention to the meaning, and especially the changing meaning, of words.

The core of the English vocabulary is of course Anglo-Saxon and Teutonic. Those words of Modern English that have prototypes in Middle English and Old English, and that trace their ultimate descent from the Indo-European parent tongue, are the very heart of the word-stock. Examples of such words are the lower numerals—one, two, and three; the most intimate of the family relationships—father, mother, and brother; and the common phenomena of nature—night, dew, and star. The relatively small place filled by such words in Modern English dictionaries is by no means an accurate index of their true importance—a point to which we must later return. Here it will be useful to emphasize the idea that resemblances between these English words and words in other languages that are like them in form, sound, and meaning are by no means to be explained as due to borrowing in either direction. The English (or the foreign) word is not a derivative of the other. Their relationship is rather that of cognates. When, however, one considers the Modern English vocabulary as a whole, it is quite evident that there are far more examples of another kind of relationship between English and foreign words. Borrowed words, to put it differently, are far more numerous in our modern dictionaries than native words. The policy of our language toward word-borrowing has on the whole been that of the open door. English words have their sources in foreign tongues that are, or have been, spoken in all corners of the world. It must now be our task to classify the major and some of the minor sources of borrowing.

310 The Development of Modern English

Three languages have contributed such extensive shares to the English word-stock as to deserve more detailed attention than the others. These are Greek, Latin, and French. Together they account for so overwhelming a proportion of the whole borrowed element in our vocabulary that the rest of it seems insignificant. Practically, too, it is often difficult or impossible to determine the immediate source of the borrowings, for Greek words are likely to be Latinized in form before they are made English, and Latin words are likely to be Gallicized. Thus texture, known to be a borrowing direct from Latin, looks as though it came from the French texture rather than the Latin textura; figure, the immediate source of which is not clear, may just as well have come from the Latin figura as from the French figure which it more nearly resembles. So too, telegraph, actually a recent coinage in English direct from the Greek words τῆλε (far) and γράφω (write), would seem rather to have come from the French télégraphe, in turn apparently from a Latin *telegraphus.1 Theater is an example of a word that is ultimately Greek, but one that was borrowed from the Greeks by the Romans, and in turn taken from Latin into French and from French into English. The spelling theatre more definitely suggests the French word as the immediate source of our English one. In the following account of the Greek, Latin, and French elements in English, we shall confine ourselves as much as possible to direct borrowings.

1 Greenough and Kittredge, pp. 95 and 96, and 49 and 50. Professor Bloomfield (Language, p. 494) neatly summarizes the factors that account for the English form of classical borrowings (even synthetic creations like telegraph): “Since the Romans borrowed words from Greek, we can do the same, altering the Greek word in accordance with the Roman’s habit of Latinization, plus the Frenchman’s habit of Gallicizing Latin book-words, plus the English habit of Anglicizing French learned words.”

Sources of the Vocabulary

311

When this restriction is applied to the Greek element in Old English and in Middle English, there is almost nothing left. Three of the few very old loan-words from Greek are church, devil, and angel. It is almost certain, however, that each was borrowed through the medium of Latin—even church, a pre-Christian borrowing—

ultimately a Greek word meaning “Lord's (house)”—and one that the English retained in preference to any derivative of the Latin ecclesia. Devil is likewise a Continental borrowing, far older than the migration of the Teutonic peoples to Britain, but presumably a borrowing directly from the Latin diabolus rather than from the Greek διάβολος (slanderer). Angel likewise comes from Latin angelus rather than Greek ἄγγελος (messenger); the Old English form, like the Modern German, was engel, the modern spelling and pronunciation being due to the later influence (amounting virtually to a re-borrowing) of the Old French angele. Among the Latin words introduced by the Roman (or by the Irish) missionaries into Old English are the following ecclesiastical terms that had been originally borrowed from Greek: abbot, alms, clerk, monk, pope, priest, and synod. Later medieval borrowings, also of course through the medium of Latin and usually French as well, are such words as diet, geography, logic, physic, rhetoric, surgeon (for “chirurgien”), and theology. Not until after the humanistic movement revived the study of Greek in western Europe did Greek words begin to enter the English vocabulary in great quantities, and even then they were likely to be Latinized in spelling and in their terminations—a convention that has usually been followed since. Thus, a Greek k is replaced with c, u with y, and the Greek suffix -os with -ous.

312 The Development of Modern English

Even in the last three centuries, the great flood of Greek words that have been incorporated within the English vocabulary has not always implied direct borrowing from Greek. Much of the modern scientific vocabulary is international in character, and often the word has been created, from Greek words or roots, in French and then adopted in English; such words as barometer and thermometer are examples. Thus, we frequently have a curious kind of indirect indebtedness to Greek as well as an odd combination of word-borrowing and word-creation. The Greek language has been used for the creation of the new terms that the expansion of the various sciences has made necessary, primarily perhaps because it is particularly well suited for the making of compounds. Latin and the Romance languages have this quality in much less degree; and English, so far as the native vocabulary goes, has largely lost the disposition for it that it once possessed. It is of advantage, too, even for a language that, like German, indulges freely in compound-making, to use for scientific purposes the largely international vocabulary that has been formed from Greek words and roots.

How pervasive is the Greek element, direct or indirect, in the Modern English vocabulary may best be illustrated by noting how, in a single department of art or science, words of this type tend to accumulate. Not only do modern drama and dramatic criticism stem directly from the practice of the Greeks and the Poetics of Aristotle; their very terminology reminds one of this descent at every turn. Thus drama itself is Greek; so also is its place of performance, the theater; so likewise are its species, like comedy, tragedy, and melodrama, and its parts or elements, like catastrophe, climax, dialog,

episode, epilog, peripety, prolog, and scene. In the various

natural sciences, the preponderance of Greek words is

just as striking; it is perhaps sufficient to mention merely

the names of such fields as bacteriology, botany, histology,

physiology, physics, and zoölogy (all Greek labels) in order

to suggest how the Greek language has permeated their

various specialized vocabularies. This effect is quite as

striking, perhaps even more so, in fields of science that

have been recently developed. Terms used in modern

medicine or allied subjects will serve to illustrate:

adenoids, osteopathy, pediatrics, psychiatry, and psycho-

analysis are a few of the hundreds of such words. The

highly specialized terminology of recent research in

electricity-what has been called "Schenectady Greek"

-is a jargon that is often completely mysterious to the

non-Greek layman: a group of typical examples, made up

of words utilizing the same termination, includes

dynatron, kenotron, phanotron, pliotron, magnetron, and

thyratron.2

While borrowings, and particularly direct borrowings,

from the Greek have been largely confined to Modern

English, our language has been exposed to Latin influence

throughout its whole history. We have already glanced

at the Latin influence on the Old English vocabulary,

an influence which began before the migrations of the

Teutonic tribe to Britain and received its chief accelera-

tion through the later Christianization of the island.3

The actual number of Latin borrowings is far smaller

than might have been expected, for the general policy of

2 These are suggested by an article "Boom Times in the Laboratory,"

New Republic, Vol. LXVII, No. 868 (July 22, 1931), pp. 253-256.

3 P. 44.

314 The Development of Modern English

Old English was clearly antagonistic to external influence. After the Conquest this policy toward Latin was gradually altered, largely through the established practice of borrowing from French; and when in the thirteenth century English reasserted itself as a literary language, Latin began to be freely levied upon to amplify and enrich its word-stock. Many of the pre-Renaissance borrowings from Latin are doubtless due to English translations of the Vulgate Bible, especially the fourteenth-century versions of Wyclif and Purvey. While Old English translators of the Bible had usually turned Latin words into native ones that rendered them almost literally—mildheortnisse (mildheartedness) for Latin misericordia (mercy), and ðrynnes (three-ness) for trinitas (trinity)—Wyclif introduced such Latinisms as generation, persecution, and transmigration.4

The revival of classical learning in the sixteenth century first swelled the numbers of English words borrowed from Latin to overwhelming proportions. Whereas the number of Middle English loan-words from Latin is difficult to estimate, because of the habit of treating even a word direct from Latin as if it came through the medium of French, there can be no doubt that the great mass of borrowings in early Modern English came directly from Latin, even though the convention of Gallicizing the form of the word persisted. Romanic borrowings in Middle English are on the whole much more likely to be from French than from Latin; in Modern English, they are much more likely to be from Latin than from French. Recent borrowings from

4 Cf. McKnight, G. H., English Words and Their Background, pp. 111-113.

Sources of the Vocabulary

315

French are of course frequent enough, and usually the word carries an unequivocally French appearance; but their number is far less than the number of borrowings direct from Latin. The habit of treating direct Latin borrowings as if they were French words has continued, but there is nevertheless a fair number of Latin loan-words in Modern English that have escaped this tendency and appear in English unchanged.5 A few specimen words, illustrating also the great diversity in the parts of speech and the inflectional forms borrowed, are these: affadavit, agendum, alibi, animal, bonus, deficit, exit, extra, fiat, item, maximum, memento, memorandum, omnibus, posse, propaganda, quorum, sponsor, terminus, verbatim, veto, and via. Though it is perhaps impossible to get anything like an accurate estimate of the whole number of Latin words borrowed in English, an estimate of another kind that has been widely accepted is that of Professors Greenough and Kittredge,6 who, through computing the proportion of words in a in Harper's Latin Dictionary that have been taken over into English, have arrived at the conclusion that English “has appropriated a full quarter of the Latin vocabulary, besides what it has gained by transferring Latin meanings to native words.”

The enormous proportion of Latin loan-words in Modern English dictionaries of course betokens a reversal of the originally conservative attitude of the language toward word-borrowings. We have seen in general how the more liberal policy was developed in Middle English

5 E. L. Johnson, in Latin Words of Common English (pp. 63-87), gives many such examples.

6 Words and Their Ways, p. 106.

316 The Development of Modern English

times and became the more or less fixed practice of Modern English. A more detailed treatment of the history of the language from this point of view, however, would have to indicate many currents and eddies. There have always been two opinions as to the propriety of adding to the resources of the vocabulary by levying freely upon other languages. In the Old English period there must have been some inclination to borrow from without; otherwise the Old English vocabulary would have remained exclusively Teutonic. On the other hand, when the Renaissance habit of adorning the English page, or even English conversation, by a very liberal infusion of Latinisms began to assert itself, there were many voices raised in protest. Shakespeare's satire, in Love's Labor's Lost, of the "inkhorn" terms used by the Latinists is generally familiar. It is perhaps more surprising to find a writer so deeply indebted to Latin as Sir Thomas Browne making the same kind of protest when he remarks, "If elegancy still proceedeth, . . . we shall, within a few years, be fain to learn Latin to understand English."

In the era of the Renaissance, as has already been suggested,7 the victory lay with the Latinist; at the same time such protests as those just mentioned were useful in tempering the extravagances of the policy of free borrowing. There were extravagances of the opposite sort too. Sir John Cheke's purely Anglo-Saxon version of a part of the New Testament affords a curious illustration of the diametrically opposed theory. In general, it is clear that the later compromise effected between the extremes of both positions has been one in which the

7 P. 66.

Latinist has had the better of it. But the Saxonist has not been completely downed. He sometimes succeeds in getting an English word established, or reëstablished, in the face of a competing Latinism. Professor Jespersen has traced the curious history of handbook.8 the Old English handbōc disappeared in the Middle English period, when its place was taken by the Latin-French manual and even (in the sixteenth century) by the Greek enchiridion. When handbook was re-introduced, in the nineteenth century, it was treated as an objectionable neologism, "so accustomed had the nation grown to preferring strange and exotic words." In spite of its superior intelligibility—manual after all merely implies the idea of "book"—handbook has only very recently won wide acceptance. A somewhat similar story is that of foreword. The scorn of the Latinist is voiced in Hilaire Belloc's dictum:9 "what Anglo-Saxons call a foreword, but gentlemen a preface." There will be occasion later to return to the debatable ideal of "purity" in vocabulary.

To complete an enumeration of the major sources of word-borrowing, the French element must be added to the Greek and the Latin. The influence of Norman French began to operate even before the Conquest; but the great influx of French borrowings took place considerably later, after Parisian or Central French had become a more important source than the Norman dialect. There are certain doublets in English that testify to borrowing the same French word twice, once in its Norman (or Picard) form and once in its Central

8 Growth and Structure, pp. 49 and 50.

9 The Path to Rome, New York (Putnam), 1915, p. ix.

318 The Development of Modern English

form. Thus the pairs cattle and chattel and catch and

chase illustrate the different treatment in the two French

dialects of what was once the same initial consonant.

Likewise, originally Teutonic words with initial w that

had been borrowed in French passed into English in two

forms: the Norman dialect retained the w, but in Central

French this was transformed into gu (or g). English has

sometimes borrowed both variants, so that we have such

doublets as warrant and guarantee, warden and guardian,

and reward and regard.

Variant English forms of the same borrowed root are

not of course always to be explained in the way just

mentioned. To digress a moment, it may be of interest

to see some of the other possibilities, especially those

connected with the Greek, the Latin, and the French

sources. Double borrowings of words originally Greek

are these pairs, in which the first member is the earlier

borrowing and the second the later: balm, balsam; blame,

blaspheme; diamond, adamant; fancy, phantasy; priest,

presbyter; and palsy, paralysis. Some of these had

variant forms in French—for example, blâmer and blas-

phémer—and English has borrowed both; in other cases—

for example, adamant, phantasy, and paralysis—the

second borrowing is a learned one, direct from the Greek.

Again, corpse is a borrowing from Old French, while

corps is a re-borrowing of the same word with the s silent,

as the Modern French pronunciation has it. But the

doublets captive and caitiff almost reverse this situation:

for caitiff is a borrowing from Old French, while captive,

much nearer the form and sound of the Latin original,

is the second borrowing. On a larger scale, the possibili-

ties of multiple borrowing are strikingly exemplified

in the history of the Latin root that appears in regem

(king). The adjective form, regalis, gives English regal, a direct borrowing. But in the usual development of Old French from Latin, intervocalic consonants are lost; hence real is a Middle English borrowing from Old French. (Another real in English represents Latin rem (thing); the real here referred to, however, occurs as the name of the coin real, and in real-m.) Modern French roi has evolved, with the usual phonetic changes, from Latin regem, and again the corresponding adjective has been borrowed in English royal. Hence regal, real, and royal are triplets, as are also legal, leal (preserved in the Scotch song, "Land o' the Leal"), and loyal. The use to which Latin discus (itself a borrowing from Greek) has been put in English is even more extensive: the Old English direct borrowing is preserved in dish, a Middle English variant gives us desk, and a borrowing of the Old French form gives dais; seventeenth-century direct borrowings of the Latin word are disk (disc) and discus, the last popularized anew when the Greek sport of throwing the discus was revived. Duplication, together with the variation in sound and form that results from borrowing at different times and through different channels, may be further illustrated in capital, chief, and chef; candelabra, chandler, and chandelier; gentile, gentle, genteel, and jaunty.

A few examples of the kinds of words borrowed by Middle English from French were given in an earlier chapter. It would, however, be erroneous to assume that loan-words from this source are to be found only in particular categories like these. Actually the English vocabulary is permeated by French everywhere. Many—indeed, most—of the earlier borrowings are not felt to be alien in any sense. The common impression that

320 The Development of Modern English

monosyllables in English are of native origin only is refuted by the long list that could easily be compiled of

French loan-words of but one syllable. A very few typical illustrations, limited to words beginning with c,

are the following ones, which probably the great majority of English speakers would consider to be as thoroughly

"English" as any words that could be mentioned: cage, calm, car, catch, cause, cell, chain, chair, chance, change,

charge, chase, cheer, choice, claim, clear, close, coast, cost, course, crime, cry.

At the same time, it is obvious that borrowings from French do tend particularly to accumulate in certain

fields. The first large group of borrowings was made up of ideas associated with religion and the church. Impor-

tations of the eleventh century are these: cell, chaplain, charity, evangelist, grace, mercy, miracle, nativity, paradise,

passion, sacrament, and saint. These words, in general standing for the more formal and outward aspect of

Catholicism, have been suggestively contrasted10 with borrowings of the thirteenth century that testify to a

more inward and personal aspect of religious faith: anguish, comfort, conscience, devotion, patience, pity,

purity, and salvation are witnesses to the religious revival on the Continent and the advent of the preaching

friars in England. Other fields than religion in which borrowing from French was early active are of course

law, government, military affairs, and the general usages and conventions of polite society. Legal terms

of early date in addition to those already mentioned are assets, bail, bailiff, embezzle, lease, and perjury. Typical

loan-words for ideas relating to national government are

10 By L. P. Smith in The English Language, pp. 165 and 167.

chancellor, country, exchecquer, govern, minister, power, and

reign. French borrowings that stand for ideas associated

with war and things military are particularly interesting.

Army had to compete with three Old English words:

here (the harrying body) (the term applied by the Anglo-

Saxons to the Danish force), fird (the traveling body)

[<faran (go)], and the neutral werod [(body) of men]; it

was not until the fifteenth century that the older terms

were completely displaced. But a great deal of the

French military vocabulary was introduced very early.

To illustrate the scope and variety of words of this class, a

few specimens may suffice: assault, company, enemy,

hauberk, lance, lieutenant, mail, navy, sergeant, soldier, and

troops. It may perhaps be significant of the fons et

origo of militarism that the practice of borrowing its

terms from French has been by no means confined to

Middle English. Captain and colonel, for example, are of

considerably later date than lieutenant and sergeant.

Other sixteenth- or seventeenth-century borrowings are

attack, barricade, campaign, cannonade, commandant,

corps, cuirassier, dragoon, march, and massacre. Some of

these represent the renewed general activity in borrowing

from the French which came about through the special

French influence in the reigns of Charles I and Charles II.

This influence upon the English language and English

literature continued, of course, into the eighteenth

century. In 1710, we find Jonathan Swift complaining

that "the war has introduced abundance of polysyllables

which will never be able to live many more campaigns."11

11 The Tatler, No. 230 (Sept. 28, 1710). Swift, however, is wrong

about the date of some of these words: "the war" he refers to must be

the War of the Spanish Succession (1701-1713), but the Oxford English

Dictionary cites earlier uses of all but the first three of the list.

322 The Development of Modern English

The terms he cites are speculations, operations, preliminaries, ambassadors, palisadoes, communication, circumvallation, and battalions. A curious example of the folly of such confident prophesying about language! None of the words he gives has perished, and only one has been altered, very slightly, in form. Borrowing from the French military vocabulary has continued, of course, up to the present; accessions due to the World War are represented by words like barrage and camouflage.

One other general group of French borrowings has been mentioned, that composed of words standing for the usages and the conventions of polite society. Too much can easily be made of a contrast between “homely” Anglo-Saxon and “polished” French ideas in Middle (and Modern) English; nevertheless, to some extent this contrast does hold good. It has often been pointed out that vigor and sincerity rather than grace and finish are characteristic of the “Anglo-Saxon tradition”12 in literature. In language, the words standing for courtly and graceful ideas are similarly more likely to be of French than of native origin. Courtliness and grace themselves belong here, as do such kindred terms as chivalry and honor. The French introduced more polished table manners, and words pertaining to eating, like dinner, supper (though the homelier breakfast remained), table, fork, plate, and napkin; also, various ways of preparing food for the table, like boil, broil, and roast, as well as the names of different kinds of food.13 French, too, are the general names for clothes—apparel, costume, dress, and garment—as well as most of the more specific

12 Cf. the eloquent chapter, of that title, in J. L. Lowes’s Convention and Revolt in Poetry, Boston (Houghton Mifflin), 1919.

13 Cf. p. 56.

ones dealing with the kinds of dress, their styles and

materials: brassière, décolleté, voile, chemise, and lingerie

are a few recent examples of such words. The earliest

words relating to art are French—art itself, and beauty,

color, design, ornament, and paint—and so are the more

specialized terms of such a field as architecture (for

example, aisle, arch, chancel, column, pillar, porch,

reredos, and transept). Some of the humbler occupations,

like those of the baker, fisherman, miller, shepherd, and

shoemaker have kept their native names; but many

others, especially the more skilled trades and those

that brought their practitioners into contact with the

upper classes, have French designations: the words

barber, butcher, carpenter, grocer, mason, painter, and

tailor are typical.

A final general aspect of the French borrowings in

English remains to be discussed. In what way and to

what extent have these words been naturalized? The

general principle is obvious: the older the borrowing,

the more thoroughly it tends to follow normal English

habits of accentuation and pronunciation. Mere age,

however, is not the sole factor; the extent to which the

word is used in English is also of primary importance.

Thus, the recent importation camouflage was very soon,

in popular usage, strongly accented on the first syllable,

and its final consonant changed from [ʒ] to [dʒ]. Curi-

ously enough, however, garage, a word in still more

general use, is not usually given the completely Anglicized

pronunciation [ˈɡærɪdʒ]. This is what might have been

expected, on the analogy of dozens of words like message,

courage, cabbage, and so forth. Mr. H. W. Fowler14

14 Dictionary of Modern English Usage, London (Oxford), 1927. p. 211.

324 The Development of Modern English

gives this, apparently, as both the common and the desirable treatment of the word; but in American use, at any rate, it is far less frequent than [gə'raʃ] and [gə'radʒ], the latter compromise being the really popular version. The usual statement of American phoneticians that the ordinary pronunciation is ['gæridʒ] would seem to be based on the expectation that the word would follow the normal path, rather than on actual observation of the somewhat exceptional development that it has really undergone.

As has been indicated, the usual expectation is that both borrowings long in use and borrowings of recent date that are widely and popularly employed will be completely made over to conform to English patterns. Many Middle English importations, as we have already seen, are to all intents and purposes apprehended as native English words, so completely has the language assimilated them. This is truest, of course, of thoroughly popular, monosyllabic words, themselves borrowed from everyday, short French ones. But even borrowings that have taken over longer and somewhat less homely words of Old French, though they may sometimes be felt to be aliens, are completely naturalized so far as pronunciation goes. Thorough Anglicization is also the invariable rule for Modern English borrowings from Modern French until about the era of the Restoration. From this time on, however, a deliberate attempt was often made to preserve the original pronunciation; and although this has rarely met with complete success, it is a tendency that has kept the older trend toward thorough naturalization in check (except in the case of the most popular words, which would usually insist on becoming English). Quite recently, one may observe a counter-tendency in

Sources of the Vocabulary

325

cultivated as well as in popular speech: wider recognition of the desirability of making French loan-words that are to be in living use in English speech conform to English analogies. Thus valet is now usually as completely Anglicized as varlet; and trait, long a shibboleth of the difference between British and American pronunciation, tends to be pronounced with sounded final t in British pronunciation too.15

What seems the desirable procedure with the pronunciation of French borrowings in Modern English has already been suggested. The impossibility of attempting to keep them French has been set forth wittily and persuasively by Mr. Fowler:16 “To say a French word in the middle of an English sentence exactly as it would be said by a Frenchman in a French sentence is a feat demanding an acrobatic mouth; . . . it is a feat that should not be attempted.” But if they are not to be kept French, what is to be done with them? Completely Anglicized renderings of many comparatively recent borrowings have, of course, no standing at all. The un-English accentuation of such words as ballet, cadet, caprice, grimace, prestige, and so forth, must apparently be retained. Comfort may be given, however, to any tendency toward Anglicization that makes its way upward from popular to cultivated usage. The accent of even prestige may be changed, perhaps, and the final consonant made [dʒ] instead of [ʒ]. A similar

15 Fowler remarks (p. 660), “The final t is sounded in America but still usually silent in England.” The committee advising the British Broadcasting Committee on pronunciation go further than this in recommending the “American” pronunciation for British use (S. P. E. Tract No. XXXII. Robert Bridges’s comments are on pp. 380 and 381).

16 P. 194.

326 The Development of Modern English

change seems to be taking place in sabotage, and a some-

what different one when technique is repelled technic.

Foyer is given by Fowler as [ˈfwajeː], but [ˈfoijeː] is

unpleasantly common. Surely [ˈfoijər] is preferable to

the hybrid effect of the latter. In a word, a pedantry

all too familiar in both the use and the pronunciation of

French borrowings is one that should be combated.

Aside from Greek, Latin, and French, only Danish

has made a really substantial contribution to the English

vocabulary. So substantial is it, indeed, that could we

be more certain of its precise extent, we might more

justly reckon the Danish (or, to be more accurate, the

Scandinavian) element the fourth of the major sources

rather than the chief of the minor ones. The reader may

be referred to Jespersen’s chapter “The Scandinavians”17

for an account, by one peculiarly qualified to deal with

the subject, of this difficult matter. “Scandinavian”

rather than “Danish” is to be preferred, for it has been

shown that the so-called Danish invaders of England

were often Norwegians. The resemblances between

Old English and Old Norse (or Old Icelandic) were of

course much closer than those that now exist between

English and Danish. It is consequently often difficult

to say whether a given word of Modern English has

developed from Old English or from a Scandinavian

cognate. Sometimes it happens that the same root

appears twice in Modern English, once from Old English

and once from Danish. Such doublets as no and nay,

rear and raise, from and fro, and shirt and skirt illustrate

this situation, the first of each pair being the native

word and the second the Scandinavian. But in whole

17 Chapter IV (pp. 59-86) of Growth and Structure.

(O. E. hāl) and hale, church and kirk, and chest and kist,18 it is equally likely that the second word of each pair represents a native Northern dialect rather than a Scandinavian variant—an additional complication in the difficult task of sorting out the Scandinavian contribution to Modern English.

Occasionally, while it is clear that the native form has survived, it is evident that the meaning is the Scandinavian rather than the native one: thus, dream comes in form directly from its Old English ancestor, but this meant “joy,” whereas an Old Norse cognate had the meaning “vision”; and ploh meant “measure of land” in Old English, while the Old Norse plógr meant “plough.”

The best indication of the peculiarly intimate relation between Old English and Danish is the fact that in all probability the pronominal forms with initial th—they, their, and them—are due to Scandinavian influence. This is the one case in which English has borrowed pronouns from another language.

Indeed, “borrowing” seems scarcely the term for the process of gradual assimilation that we must imagine to have taken place. Two conditions evidently made it possible: the chance of ambiguity that would have resulted from the normal development of Old English hīe, hiera, and him (their likeness to he, her, and him); and the presence in the language already of other pronominal forms in th—this, that, and (later) the.

But borrowing of pronouns even in these circumstances witnesses most vividly to the close interrelation of the Anglo-Saxon and Danish elements in early English.

Many Danish borrowings have, of course, been lost. A certain nautical vocabulary, particularly terms asso-

18 Jespersen, pp. 66 and 67.

ciated with naval warfare, that the Anglo-Saxons, quite understandably, took over from their Viking invaders, has not descended from Old English to Modern English. To some extent this corresponds to a much later group of borrowings from Dutch, and for a similar reason: the English of the eighth and ninth centuries learned ship-

building from the Danes, as their seventeenth-century descendants learned seamanship from the Dutch. Somewhat similarly, early borrowings of Scandinavian law terms have been chiefly lost, and their places taken by the much more numerous law terms from Norman French. Law itself, however, is one of the survivors, as are the compounds outlaw and by-law,19 and a few terms like hustings, riding [originally thriding (third part)], and wapentake (the last two, names for the divisions of English counties).

A few miscellaneous illustrations of the Scandinavian element in English may be added. If we can be reasonably sure that a great many of our most familiar words, like man, father, house, life, summer, come, bring, see, hear, think, sit, stand, wise, well, better, best, and so forth, are strictly English, it is only, as Jespersen points out,20 because the survival of a considerable body of early English literature has made it possible to label them as Old English. Many words were doubtless identical in Old English and Old Norse, and many more only slightly differentiated in sound or form. In addition to the numerous words that may be Scandinavian rather than

19 This meant "town-law," the first element being that which appears also in place names like Grimsby, Haddeby, and Rugby; but popular etymology has distorted its meaning, through confusion with the more familiar sense of by, into "additional or supplementary regulation." Hence, probably, such new creations as by-play and by-product.

20 Pp. 64 and 65.

Sources of the Vocabulary

329

Anglo-Saxon, there are many that are indubitably Scandinavian. These words stand, for the most part, for the everyday, commonplace things of life: there is no suggestion of a differentiation in culture or interests between Dane and Saxon, as to some extent there is between Saxon and Norman. Typical nouns are fellow, gate, haven, husband, knife, root, skin, and window; verbs include bask, call, cast, die, drown, get, hit, scrape, scream, scowl, skulk, take, and want; adjectives are such as happy, ill, loose, low, meek, odd, rotten, seemly, ugly, and wrong.

Like the Celts, the Danes have left a conspicuous record of their occupation of parts of Britain in the shape of place names. According to Isaac Taylor,21 more than six hundred towns in the east of England have names ending in -by, from the Danish word for “town.” Place names in -beck, -dale, -thorp, -thwaite, and -toft are likewise of Scandinavian origin. Besides these marks of the original contact between Anglo-Saxon and Dane, there are rather numerous Scandinavian loan-words of later date: batten, billow, blight, clumsy, and doze—all probably of Scandinavian origin—made their way from the northern and eastern dialects into literary English in the sixteenth and seventeenth centuries; a few words—equip, flounder (the fish), and possibly the verb sound—are thought to have entered English through the circuitous route of Scandinavian elements retained in the speech of the Normans from their original northern tongue and later brought by them to England as part of their French dialect; several comparatively recent borrowings—nag, scamp, fiord, floe, geyser, saga, and ski—represent either later accretions from English dialects or fresh contacts with Scandinavian languages.

21 Words and Places, New York (Dutton), 1909, p. 158.

330 The Development of Modern English

Turning to the more specifically minor borrowed elements in the English vocabulary, it is natural to begin with what is chronologically the oldest: Celtic. The curiously small extent of the Celtic influence (apart from place names) upon the Old English word-stock has already been commented on.22 Besides the words previously mentioned, only a very few others—ass, bin, coomb, dim, and mattock—may be assigned, with any show of probability, to this small handful. Of other words (and also of this group) often listed as Celtic, it is quite likely that some come, not from the Celts of Britain, but from the Irish missionaries to Northumbria, or even that they are Teutonic or Romanic words borrowed in Celtic and later absorbed into English. Later stages of the English language have, however, levied upon the more modern forms of the aboriginal tongues. But the loan-words from Welsh, which we might have expected to be rather numerous, are relatively few: perhaps only flannel and flummery are really common words; others, like coracle, cromlech, and eisteddfod, are distinctly more exotic. Both Scotch-Gaelic and Irish contribute a far greater number of words, which are usually characteristic: clan, claymore, pibroch, slogan, and whisky are typical of the one group, and bog, colleen, pillion, shamrock, and spalpeen of the other.

Aside from English, there are two important modern languages of the West Teutonic group: Dutch and German. Borrowings from both are on the whole so scanty that one need have little hesitation in terming each of these sources a minor one. It should be remembered, however, that the extremely close relationship between Dutch (and Flemish) and English makes it quite con-

22 P. 43.

ceivable that some words that we label Anglo-Saxon

may just as plausibly be explained as of Dutch or Flemish

origin. As we have seen,23 numerous English words are

still nearly or quite identical with their cognates in the

Low German languages; the right of such words to be

considered “pure” Anglo-Saxon is, of course, highly

dubious. One group that may safely be assigned to

Dutch or Flemish influence on Middle English is made

up of such trade terms as pack, from the wool trade;

hops, scum, and tub, probably from the brewers of the

Low Countries; and the following terms of the cloth-

making industry (believed to date from the importation,

in Chaucer’s day, of Flemish weavers into England) : curl,

scour, spool, stripe, and tuck. Far the most important

and homogeneous group of Low German—chiefly Dutch

—loan-words are those that pertain to ships and the

sea: medieval importations are bowsprit, buoy, freight,

keel, leak, lighter, marline, and pump; more recent ones

are such words as avast, belay, boom, bow, commodore,

cruise, dock, keelhaul, yacht, and yawl.24 Besides these

nautical borrowings, there is a fair number of miscella-

neous words that Modern English has taken from the

Low German tongues; a few specimens are foist, free-

booter, groat, heyday, hoist, jeer, plunder, snap, switch,

toy, and wagon. Among the many words that may

possibly be of Low German origin25 are some very familiar

ones, such as boy, girl, bounce, luck, mud, and scoff.

23 P. 36.

24 Most of these illustrations are from Smith, L. P., The English Lan-

guage, pp. 191-193.

25 They are so assigned by Skeat in the appendix to his Etymological

Dictionary. Here, as elsewhere in this chapter when they are utilized,

Skeat's etymologies have been verified, wherever possible, by consulting

the Oxford English Dictionary.

332 The Development of Modern English

High German borrowings are considerably rarer than Low German ones. Bradley,26 however, calls attention to one group of words in which the influence of Modern German is conspicuous: the science of mineralogy has retained in its English vocabulary, with slight modifications, a number of the terms employed in the land of its origin—bismuth, cobalt, gneiss, quartz, shale, and zinc are typical. Another group that is likewise noteworthy is that made up of terms used in eating and drinking; for example, sauerkraut, stein, and lager beer. Such words are, of course, far less numerous than French loan-words in the same field, and they are less completely assimilated in English. The same observations may be made of the occasional importations from German in other departments: such various borrowings as carouse, hinterland, kindergarten, leitmotif, meerschaum, poodle, umlaut, waltz, and wanderlust, though most of them are in common enough use in English, still bear, in some measure, the marks of their alien origin.

Even apart from French, the modern Romance languages, especially Italian and Spanish, have contributed far more to the English vocabulary than have the Teutonic languages that are the closer relatives in our family-tree. Borrowing from Italian naturally began to be active when the culture of the Italian Renaissance impressed itself upon sixteenth-century England. When the traveled Elizabethan Englishman, as Sir Thomas Overbury put it, spoke “his own language with shame and lisping,” it was chiefly with Italian that he filled out the supposed deficiencies or improved upon the crudities of his native tongue. Most of these Elizabethan affecta-

26 Making of English, p. 103.

Sources of the Vocabulary 333

tions, however, have not stuck, any more than have

the Spanish importations that are based upon literary

admiration of Góngora and Guevara. Of those that did

establish themselves, attitude, fiasco, isolate, motto, stanza,

and umbrella are typical Renaissance borrowings from

Italian, and desperado, grandee, negro, punctilio, and

renegade from Spanish.27 As to Italian borrowings in

general, it is well known that the chief indebtedness of

English to this source is for its musical vocabulary.

This extensive stock of words comprises both those in

everyday use, such as piano, alto, soprano, opera, and

tempo, and those that are more definitely specialized or

technical, such as legato, diminuendo, rallentando, pizzi-

cato, allegretto, scherzo, and andante. Words like these

are chiefly of the eighteenth century, the period when

Italian music was introduced into England. A some-

what similar, though smaller, group of Italian borrowings

comprises terms used in the fine arts, like chiaroscuro,

dado, fresco, portico, replica, studio, and torso. Perhaps

half of all our Italian loan-words come to us through

the medium of French; it is this indirect relationship

that obtains, for example, in virtually all the oldest, or

Middle English, borrowings: alarm, brigand, ducat,

florin, orange, and rebec28 (the last two ultimately of

Eastern origin) are typical. But words that are shown

by their very aspect to be direct from Italy have fre-

quently been borrowed throughout the Modern English

period; a few random illustrations are cameo, campanile,

dilettante, extravaganza, lava, macaroni, rifacimento, spa-

ghetti, and virtuoso. Spanish borrowings are less com-

27 Bradley, op. cit., p. 103.

28 Emerson, p. 168.

334 The Development of Modern English

mon; but such loan-words as armada, castanet, cigar, guerilla, hidalgo, matador, mosquito, quadroon, and vanilla testify that this source has by no means been a negligible one.

English-speaking peoples have had commercial (as well as occasional warlike) contacts not only with the Spanish but with the Portuguese, as is evidenced by such importations from the latter as banana, cobra, cocoa, molasses, and pimento.

It is of course evident that a good deal of history, both political and cultural, is implicit in the various groups of loan-words.

This is just as true for the minor sources as for the major ones. The indebtedness of Western European civilization to the speakers of Arabic for the early development of chemistry, mathematics, and medicine is manifest in such borrowings from this source as the following: alcohol, alchemy, alembic, algebra, alkali, attar, cipher, elixir, naphtha, sugar, syrup, and zero.

That the Arabic-speaking peoples were often the intermediaries through whom Greek culture reached Western Europe at a time when Greek itself was all but unknown is suggested by the history of several of these words—but all came through the medium of Arabic.

The devious routes by which Arabic words have become English are illustrated in the history of the doublets cipher and zero: both come from the Arabic sifr (cipher), but the one via Spanish and French, and the other through Vulgar Latin and Italian.

The exotic East, always fascinating to the Western traveler, is conjured up in such other characteristic borrowings from Arabic as emir, bedouin, fakir, gazelle, giraffe, harem, houri, hookah, lute, minaret, mosque, myrrh, sheik, simoom, sirocco, sultan, and vizier.

Sources of the Vocabulary

335

Hebrew, contributes similar pictures of the Orient in loan-words like camel, cassia, cinnamon, elephant, hyssop, and sapphire. A more important indebtedness to Hebrew, however, is to be observed in religious terminology. That many of these words come into English indirectly rather than directly is, of course, natural enough in view of the part that Greek and Latin have played in the tradition of Biblical translation; here too is the obvious explanation of why, in spite of the great influence of the Bible upon the English language, the total number of Hebrew borrowings is comparatively small. Typical words of this sort are amen, cherub, Gehenna, hallelujah, manna, pharisee, rabbi, sabbath, and seraph.

Another Oriental source is Persian, a nearer relative than Arabic or Hebrew, since Persian is not Semitic, but Indo-European. On the characteristic flavor of this group of loan-words, the following rhapsody of Logan Pearsall Smith may be quoted: "To me they glitter like jewels in our northern speech. Magic and Paradise, for instance; and the names of flowers and gems and rich fruits and tissues–Tulip and Lilac and Jasmin and Peach and Lapis Lazuli . . . and Orange and Azure and Scarlet."29

Among other Asiatic countries, only India has contributed a considerable number of words. From its many languages, often very circuitously, we have taken over such terms as bandanna, calico, bungalow, indigo,

29 More Trivia, New York (Harcourt), 1921, p. 85. (The etymologies are, apparently, in one sense sound enough, except of course that lapis is Latin–though lazuli is thought to go back to the same Persian origin as azure. Most of the words, however, are far from being direct borrowings from Persian; magic and paradise, for example are Middle English < Old French < Latin < Greek < Old Persian.)

336 The Development of Modern English

jungle, loot, rajah, and sandal. Next most numerous

are Malay words like amuck, gingham, gong, gutta-percha,

and lory.30 Chinese and Japanese have been levied upon

infrequently: mandarin, pongee, serge, and tea represent

the one source, and hara-kiri, jiu-jitsu, kimono, and

samurai the other. From minor Asiatic or Polynesian

languages have come such words as atoll, bamboo, cheroot,

junk, sago, taboo, tattoo, and teak.

It is clear, on the whole, that from languages that have

contributed but a few words to English we have taken

only words for indigenous things. This is true, for

example, of the borrowings even from a European lan-

guage (and one of the Indo-European system) like

Russian: ikon, knout, pogrom, steppe, tsar, and vodka can

be said to be English only in a limited sense, for they

would scarcely be used except with direct reference to

things Russian. Perhaps, however, the future of the

Russian experiment may determine whether other

Russian loan-words in English will be more commonly

used; certainly, more recent borrowings like bolshevik,

soviet, and intelligentsia (the last of course not ulti-

mately Russian) seem destined to a wider usefulness.

The limited way in which English assimilates most of the

minor sources upon which it levies is perhaps less doubt-

fully to be illustrated through its loans from Turkish:

bey, caftan, effendi, janissary, horde, kiosk, odalisque, and

turquoise remain, for the most part, exotic. Likewise

limited in character are the loan-words from various

African tongues: chimpanzee, gnu, gorilla, kraal, oasis, and

zebra are representative. Even the languages of the

American Indians, with which American English was so

30 Bradley, p. 104.

long in direct contact, have contributed but a few words

in everyday use, such as potato and tobacco—both altered

from Spanish versions of the original native names. A

very much longer list—of which squaw, tepee, tomahawk,

and wigwam would be typical—could be compiled of

borrowings confined to ideas or things peculiar to the

Indians.

Enough has perhaps been said of the major and minor

sources of borrowing that have been utilized by the

English language in general. Clearly the general rule is

that almost every vernacular, anywhere on the globe,

with which the speakers of English have had any

important contact at all, has been levied upon for at least

a few words. Since American English has had its own

contacts with other tongues, it is natural that it should

have absorbed certain foreign words that it does not

share with British English. The most distinctive of

these contacts is of course that referred to above, its

relation to the dialects of the Indian. Canoe, chipmunk,

hominy, mackinaw, moccasin, moose, opossum, paw-paw,

persimmon, sagamore, and succotash are typical loan-

words from this source that have scarcely become part of

the general English vocabulary. The two European

languages that have contributed most to the distinctively

American vocabulary are undoubtedly Spanish and

French. Spanish borrowings are of course to be found

most abundantly in the sectional speech of the South-

west;31 some that are in comparatively general use

in the United States are adobe, broncho, cafeteria, cinch,

corral, mustang, ranch, rodeo, and vamoose. Loans

from French are due chiefly to contacts with that lan-

31 Cf. Mary Austin, "Geographical Terms from the Spanish," American

Speech, Vol. VIII, No. 3 (October, 1933), pp. 7-10.

338 The Development of Modern English

guage in two widely separated regions: Louisiana and the Canadian border. They include such words as bayou, cache, chute, crevasse, lagniappe,32 levee, portage, and prairie.

On the whole, however, it is safe to say that peculiarly American borrowings from foreign languages are far fewer than might be expected. In general, American English has been surprisingly little affected by the speech of immigrants. A sense of social superiority seems to have kept the speakers of American English from borrowing, except very occasionally, from the many languages introduced into the United States in this way. Though the position of natives and newcomers is just reversed, it is evidently the same reason that kept the Anglo-Saxons from levying upon the language of the Celts. The children of immigrants too are likely to scorn as inferior the speech of their parents, and to achieve precisely the same brand of English as that of their contemporaries of old American stock. Jespersen33 demonstrates how this has been the rule with the Scandinavian languages brought into the United States by immigrants. It may be pointed out, of course, that European languages have occasionally continued to exist for generations, always in a form made corrupt by association with English, in sectionally limited parts of America; “Pennsylvania Dutch” and Canadian French are the most conspicuous examples. Such cases are, however, different in that in these dialects the base is the foreign language and the admixture is American English.

32 For an appreciation of the usefulness of this word, see Mark Twain, Life on the Mississippi, Chap. XLIV.

33 Growth and Structure, pp. 77 ff.

Sources of the Vocabulary

339

Where English has been attempted by foreigners, the original speech of the immigrant is all but completely discarded in remarkably short order.

Not only has the speech of comparatively recent immigrants left almost no impression on the general body of American English; the marks of the languages spoken by non-English settlers of Colonial America have also been almost completely eradicated, except as they have been incorporated in the English language in general. Thus a few loan-words from the Dutch of New Netherlands—for example, boss, snoop, and spook—have won currency not only in American but in British English. But a far greater number of borrowings from this source have either disappeared altogether or, in a few instances, remained in an extremely restricted sectional use; in either case, they are not part of the general American vocabulary. It has been the author's observation, for example, that a few such words, still current in New York City, the Hudson Valley, or northern New Jersey, are quite unfamiliar in Philadelphia; in this category are dominie (“minister,” dying out with an older generation), kill (the topographical term), paas (“Easter,” perhaps only in the phrase Paas dyes or Paas eggs), scup (child's swing), and stoop (“porch,” known but not used elsewhere).

On the other hand, as has already been suggested, it is not difficult to cite a number of words of German origin that are part of the general American vocabulary—words, that is to say, that are neither international nor sectional. German borrowings, especially terms relating to food, have been mentioned as far more common in the United States than in England. To those already given, others like frankfurter, hamburger, pretzel, wiener-

340 The Development of Modern English

wurst, and zwiebach34 may be added. German ancestry

is also attributed to such American slang or colloquialisms

as bower (<Ger. Bauer, for the “jack” in cards, for which

knave is preferred in England), dumb (stupid), bum,

loafer, and shyster.35 It is probable, likewise, that a case

can be made for the position that the Irish element in

the American population accounts for certain turns of

expression,36 as well as habits of pronunciation,37 that

American English does not share with the English of

Southern England.38 Here, however, there is a great

deal less certainty than in the matter of German influence.

It might be expected, a priori, that American English

would be powerfully affected by the speech of races far

less akin than the Irish or the Germans to the Anglo-

Saxon stock but also represented in great numbers in the

American population. Has the speech of the Negroes,

the Chinese, or the Jews contributed much to the Ameri-

can vocabulary? The answer clearly is in the negative.

Krapp points out39 that only five or six words of non-

English origin (especially, banjo, hoodoo, pickaninny,

voodoo, and—very dubiously—jazz) can be attributed to

Negro influence, and that the list of English words that

are especially associated with Negro life in America

(like cake-walk, coon, darkey, and mammy) is not very

much longer. Chinese immigration has made almost a

34 Of these, only pretzel (labeled “U. S.”) appears in the Shorter Oxford (1933).

35 Mencken, The American Language (3rd ed.), pp. 106–108, cites these

and others.

36 Cf. Mencken, pp. 108–111.

37 On this side of the matter, cf. Krapp, English Language in America,

Vol. II, pp. 96–97 and passim.

38 Cf. also McKnight, English Words and their Background, pp. 26 and

27, and his reference to Joyce, P. W., English as We Speak It in Ireland.

39 Op. cit., Vol. I, pp. 161–163.

negligible impression on American speech; of words from

this channel, perhaps only chop-suey (which is question-

able),40 fan-tan, joss, kowtow, tong, and the slang yen

(strong desire) are at all generally familiar. Clearly, the

social barrier has operated most strongly in both instances

to keep linguistic influence at the minimum. One can-

not put the case of the third group, the Jews, on quite

the same level; but it is still true that the influence of

Hebrew or of Yiddish upon American English seems

unexpectedly small. Yiddish of course is simply a

dialect of German, and it is quite possible that some

Americanisms attributed to the direct influence of

German should really be ascribed to that of Yiddish;

poker (the game), for example, has been explained in both

these ways, as well as in several others.41 Among the

unquestioned loan-words from the speech of Jews in the

United States are kibitzer, kosher, and mazuma. It is

altogether likely, however, that Jewish influence on

American speech is considerably greater than the number

of loans definitely traced to it would indicate; it would be

strange indeed if the Jewish population of New York City

had made no considerable impression upon the English

spoken there. Like the Irish immigrants of an earlier

epoch, the German, Polish, and Russian Jews who have

come over more recently have probably influenced

American idiom and word-formation to a greater extent

than is usually realized. Mr. Mencken, at any rate, has

no doubt of it; he closes his chapter "Tendencies in

40 One or two of the others are likewise dubious: kowtow is not peculiar

to American English, while joss is explained by the Oxford English

Dictionary as probably from Portuguese deos (god), and as Pidgin English

rather than Chinese.

41 Mencken, pp. 112-113, details six theories of its etymology.

342 The Development of Modern English

American" with the assertion42: "Many characteristic Americanisms of the sort to stagger lexicographers-for example, near-silk-have come from the Jews, whose progress in business is a good deal faster than their progress in English."

To return to the subject of word-borrowing in the English language as a whole. One rather special aspect of the matter has as yet been merely glanced at-the treatment of foreign proper names. Places in which English-speaking settlers have replaced earlier ones that spoke another language have frequently been allowed to retain the earlier names (or some development of them). Thus, as we have seen, there are very many more Celtic place names in Great Britain and Indian place names in the United States than there are common nouns of Celtic or Indian ancestry in the language concerned as a whole. Another angle of the question is the treatment of place names for parts of the world that are not English-speaking. The earlier practice was either to translate them completely, as was possible in such names as Black Forest for Schwarzwald,43 Netherlands for Nederlanden, or to make them as English as possible by recasting their pronunciation and spelling along more familiar lines. Thus, in calling the Italian cities Firenze, Venezia, and Napoli by the names Florence, Venice, and Naples (all suggested by the general European use of French revisions of the native names), the English proceeded in a way not altogether unlike that in which the Germans proceeded when they called the same cities Florenz,

42 P. 212.

43 As might be expected, German has indulged in this practice much more extensively than English ever has; e.g., the German turning of Salt Lake City into Salzseestadt, or Tierra del Fuego into Feuerland.

Sources of the Vocabulary 343

Venedig, and Neapel. But although other languages are likely to persist in the practice of naturalizing foreign place names, the present trend of English is to abandon it. The treatment of French geographical terms will serve to illustrate. Only the most frequently used of such words have well-established Anglicizations; the usual pronunciation of Paris, in an English setting, is the most familiar instance. For most French place names no such equivalents exist. The English or American speaker really does not borrow them at all; rather he makes a valiant, though usually inadequate, attempt to keep the word French. Foreign place names, that is to say, are usually no longer conceived to be susceptible of borrowing at all.

It may be useful to sum up certain impressions of the borrowed element in our vocabulary, especially in contrast to the native element. Such an enumeration of the foreign sources of the English word-stock as the present chapter contains may very easily be misinterpreted. So much has been borrowed that perhaps the impression may be created that almost nothing is left. If it is true that the foreign words in our dictionaries very greatly outnumber the native ones, is it still reasonable to hold that the small Anglo-Saxon core is of considerable importance? To be in a position to deal with this question, a preliminary inquiry into the actual proportions of foreign and native elements must be made.

It should be pointed out, in the first place, that to base a count upon words as they appear in the pages of the dictionary is not the only (and not the most useful) method of computing these proportions. Such a calculation would doubtless reduce the native element to almost negligible dimensions; but it is equally probable that the result would have very little significance, for many of the

344 The Development of Modern English

words in the dictionary are rarely used. A more realistic test is to limit the count to a part of the dictionary that is

in active use. A count of this kind, based on 20,000 words in common, present use,44 makes the Anglo-Saxon

element less than one-fifth, and the Graeco-Latin (including the French) element more than three-fifths.

The larger the number of words selected, the greater would be the preponderance of borrowed words, espe-

cially those from Latin and Greek. Individual vocabularies obviously vary, in a somewhat similar

fashion, with the interests and the education of the person in question.

Another method of calculating the proportion of native and borrowed elements has often been referred

to.45 This is the plan of counting, not different words only (in the dictionary, as a whole or in part, or in a

literary work), but rather counting every word every time it is used (in a piece of writing or in a bit of conversation).

If this is done, the proportions are almost reversed. The reason is plain: almost all prepositions, conjunctions,

pronouns, and auxiliary verbs—exactly the words most often repeated—are of native origin, as are also the

greater number of familiar nouns, adjectives, and verbs. In general, one-fourth of the task of expression in English

is accomplished, so it is said,46 by just nine words—and, the, of, to, a, in, that, it, and is.

44 Reported by R. G. Kent in Language and Philology, pp. 5 and 6.

45 Cf. Krapp, Modern English, pp. 267 ff., and Emerson, History of the English Language, pp. 125 ff.

46 Cf. McKnight, English Words and their Background, p. 158. A study by Godfrey Dewey, “Relative Frequency of English Speech Sounds,”

Harvard Studies in Education, Vol. IV (1923), gives the first nine words used in written matter as the, of, and, to, a, in, that, it, and is. Another

calculation, based on spoken English—French, Carter, and Koenig’s Words and Sounds of Telephone Conversations (Bell Telephone System

Technical Publications, June, 1930)—gives them as I, you, the, a, on, to, that, it, and is. In all three lists, all of the words are of native origin.

Sources of the Vocabulary

345

be, have, it, of, the, to, will, and you—and one-half of the

task by these and thirty-four others, likewise exclusively

of native origin. No wonder, then, if counting every

word every time it occurs in a passage of literature

reveals that the writer of the most highly Latinized style

still employs many more native than borrowed words.

The exact proportion will of course vary both with the

individual writer and with the subject on which he writes.

Professor Kent47 shows that while poets of the last

hundred years have employed only from 10 to 20 per cent

of classical words, the percentage in recent technical

writing sometimes runs as high as 40. Shakespeare48 is

given credit for using 90 per cent of native words and

only 10 per cent of foreign ones, while in the King James

Bible (on the basis of three gospels) the proportion is

94 and 6. In contrast with these proportions are those

for writers of a notoriously Latinized vocabulary:

Samuel Johnson’s figures are 72 and 28, and Gibbon’s 70

and 30. Midway between the extremes stand such

writers as Milton with 81 and 19, Pope with 80 and 20,

Addison with 82 and 18, and Tennyson (who prided

himself on his “Saxon” English) with 88 and 12.

The English vocabulary, then, as it has been used by

different writers and for different purposes, exhibits

considerable diversity in the elements of which it is

composed. The difference between the six per cent

of borrowed words in the King James Bible and the thirty

per cent in Gibbon is really very great—enough, cer-

tainly, to give the two styles quite divergent colorings

and tendencies. This is the more evident when one

47 Pp. 55–57.

48 Cf. Emerson, op. cit., and Krapp.

346 The Development of Modern English

considers that the native words include almost all the necessary machinery of the sentence, the words that are vital to grammatical structure but relatively colorless in meaning. The borrowed words, on the other hand, are almost always essential to meaning. It can easily be demonstrated49 that while to remove the borrowed words from a random passage of English prose or verse may leave the sentence structure intact, it will almost inevitably rob the passage of many of the words that carry the essential meaning. By the exercise of a little ingenuity, it would doubtless be possible to construct sentences of Anglo-Saxon words only (as it would not be possible to construct them of borrowed words); but surely one could not go far in the realm of ideas with such sentences.

This consideration brings us to the disputed question of profit and loss in word-borrowing, and to that of “pure” English as an ideal to be striven for. To the present writer it seems clear that, as in the parallel case of inflectional leveling, it is a matter of balancing greater gains against smaller losses; that if we believe that the language, through its policy of the open door, has on the whole won more than it has lost, we may also believe that this advantage has been gained at a certain cost. A word, then, as to what has been lost.

Perhaps the most obvious disadvantage of the twofold word-stock is that a great many pairs of words that are alike in meaning are totally unlike in appearance and sound. This is particularly the case with nouns and their corresponding (in meaning) adjectives. Eye—ocular; house—domestic; land—agrarian; moon—lunar;

49 Cf. Professor Kent's specimens of “Latinless English,” pp. 156-158.

mouth—oral; son—filial; sun—solar; and sea—marine are

a few of the many pairs of this kind that might be men-

tioned. Virtually the same state of affairs exists with

other pairs in which the noun is a familiar, homely word,

though not a native one, and the adjective is decidedly

more learned: city—urban; country—rural; letter—epis-

tolary; and sound—phonetic are typical. There is some-

thing undeniably awkward in this situation.

It can scarcely be doubted, moreover, that the practice

of borrowing has weakened the resources of the native

vocabulary in certain definite ways. As will be pointed

out again, the earlier capacity of the language for forming

compounds has been seriously curtailed. Not infre-

quently, too, a borrowed compound has driven out a

native one, as when the French borrowing despair

replaced Old English wanhope, with the result that hope

was left without a formal equivalent for the contrasting

idea. Even here, however, it may be argued that there is

still the possibility of effective literary contrast in words

that are opposite in meaning and unlike, rather than

similar, in form; Lamb’s phrase “in hope sometimes,

sometimes in despair, yet persisting ever . . . ” (in

“Dream Children”) illustrates this.

What has just been said is perhaps indicative of how

even what seems disadvantageous about the composite

vocabulary of English may still be turned to good effect.

On the whole, there seems to be no good reason to combat

the common belief that English has profited, far more

than it has suffered, from the practice of extensive word-

borrowing. Its great wealth of synonyms is usually, and

surely with justice, reckoned as one of the noteworthy

assets of Modern English. Finely discriminated shades

of meaning are made possible by the existence of hundreds

348 The Development of Modern English

of pairs of words that are almost, but not quite, exact synonyms. One need consider only paired adjectives like the following (of which, in each pair, the first is native and the second borrowed) to be impressed by the point: hearty, cordial; deadly, mortal; bloody, sanguinary; motherly, maternal; lively, vivacious; watery, aqueous.

Often, there are more than two words among which to choose the one that is exactly right in connotation as well as in denotation. Groups of words like manly, male, masculine, virile; womanly, feminine, female; kingly, royal, regal; and earthly, earthy, earthen, terrestrial will illustrate this situation. It is not contended that the necessity of a choice like this makes English an easy language to handle well; what is contended is that the possibilities of precise and complete expression in English are very great indeed.

A point closely allied to one made above may be developed further. One implication of pairs of adjectives like deadly, mortal; bloody, sanguinary; and fatherly, paternal is that there exist in the English vocabulary two planes-one made up of everyday, familiar words, and the other of rarer, more learned ones. This is not, of course, the same distinction as that between native and borrowed, though it tends to run parallel with that distinction; by and large, that is to say, the native words are the familiar words, and the borrowed words the learned.

On the whole, too, it is clear that the native words are the shorter and the borrowed words the longer. As we have already remarked, the native vocabulary is largely monosyllabic, and the borrowed vocabulary more commonly (though not inevitably) polysyllabic. Since the several divergencies that have been enumerated tend largely to overlap, the general situation may be summed

up as follows: There are two general types or styles of

English expression, one colored by short, familiar,

native words and the other by longer, learned, borrowed

ones. The first style is characterized by directness,

brevity, and plainness, the second by eloquence, grandeur,

and sonority. There is every reason to be grateful that

the English language is not an instrument of one string;

that, on the contrary, two styles exist—for somewhat

different purposes and certainly achieving very different

effects—and together offer infinite possibilities of effec-

tive contrast. Perhaps it will not take us too far afield to

illustrate the more extreme use of each of these styles

under skilful hands.

Ornate, Latinized, polysyllabic diction is represented

at its best in the prose of Sir Thomas Browne. Here is a

typical paragraph from the famous concluding chapter of

Urn-burial, which Saintsbury has called “the longest

piece, perhaps, of absolutely sublime rhetoric to be

found in the prose literature of the world”50:

Pyramids, arches, obelisks, were but the irregularities of

vainglory, and wild enormities of ancient magnanimity.

But the most magnanimous resolution rests in the Christian

religion, which trampleth upon pride, and sits on the neck of

ambition, humbly pursuing that infallible perpetuity, unto

which all others must diminish their diameters, and be poorly

seen in angles of contingency.

Let us set alongside this another passage of seventeenth-

century prose, one distinctly in the other tradition of

style. John Bunyan had, of course, no theory of prose

style; his diction is monosyllabic and Anglo-Saxon

because he knew the King James Bible so fully and

50 Cambridge History of English Language, Vol. VII, p. 275.

350 The Development of Modern English

because he strove to present his ideas most simply and directly. Here are the opening sentences of Pilgrim’s Progress:

As I walked through the wilderness of this world, I lighted on a certain place where was a den, and I laid me down in that place to sleep; and as I slept, I dreamed a dream. I dreamed, and behold I saw a man clothed with rags, standing in a certain place, with his face from his own house, a book in his hand, and a great burden upon his back.

Analysis of only a single aspect of the two styles represented here—the length of the words employed—reveals one explanation of the totally different impressions they create. The fifty-seven words of the passage from Browne include twenty-six polysyllables, divided into ten words of two syllables, seven of three, six of four, two of five, and one word of six. On the other hand, the seventy-two words of the passage from Bunyan include only eight that are not monosyllables; and of these, seven are words of two syllables and only one word contains as many as three syllables.

There is no need whatever to assert that the one type of diction is superior to the other. What is evident, rather, is that the two styles are equally superb in their utterly divergent ways. As Lytton Strachey observes of a somewhat similar comparison he offers51 of the Latinized diction of Sir Thomas Browne as opposed to “Saxon” English: “It would be foolish to claim a superiority for either one of the two styles; it would be still more foolish to suppose that the effects of one might be produced by means of the other.”

51 In his essay, “Sir Thomas Browne,” Books and Characters, New York (Harcourt), 1922, p. 41.

Sources of the Vocabulary

351

Both the ornate and the simple styles have the defects of their qualities. The one may tend toward obscurity or an inflated emptiness of content; the other may be plain to the extent of a poverty-stricken and childlike bareness. What perhaps is most in need of being pointed out is that the more monosyllabic type of English diction has its own beauty and eloquence, and may be and has been used for the finest effects of poetry as well as prose.52 Laments such as those of Jonathan Swift on the barren poverty and flatness that have come to the English language as a consequence of the shortening of words are easily refuted by an appeal to the great poets. To confine ourselves to a single illustration, let us observe the almost completely monosyllabic diction of a famous scene in Shakespeare, called by the finest of modern Shakespearean critics “perhaps the most tear-compelling passage in literature”53—the scene from King Lear in which Lear wakes from sleep to find Cordelia bending over him (Act IV, scene vii). Here are Lear’s first words and Cordelia’s replies:

Lear. You do me wrong to take me out o’ the grave.

Thou art a soul in bliss; but I am bound

Upon a wheel of fire, that mine own tears

Do scald like molten lead.

Cordelia. Sir, do you know me?

Lear. You are a spirit, I know; when did you die?

Cordelia. Still, still, far wide! (ll. 45–50)

Three dissyllables in fifty-four words!—and much the same proportion holds in this later excerpt from the same dialog:

52 E.g., the line from Milton quoted on p. 159.

53 Bradley, A. C., Shakespearean Tragedy, New York (Macmillan), 1930, p. 61.

352 The Development of Modern English

Lear.

. . . nor I know not

Where I did lodge last night. Do not laugh at me;

For as I am a man, I think this lady

To be my child Cordelia.

Cor.

And so I am, I am.

Lear. Be your tears wet? Yes, faith. I pray, weep not.

If you have poison for me, I will drink it.

I know you do not love me; for your sisters

Have, as I do remember, done me wrong:

You have some cause, they have not.

Cor.

No cause, no cause.

(ll. 68-75)

Much the same observation might be made of two great

speeches in the last scene: Lear's words (Act V, scene iii,

ll. 8-19) beginning

No, no, no, no! Come let's away to prison;

We two alone will sing like birds i' the cage . . .

and his dying utterance (ll. 306-312):

And my poor fool is hang'd! No, no, no life!

Why should a dog, a horse, a rat have life,

And thou no breath at all? . . .

Do you see this? Look on her, look, her lips,

Look there, look there

This last speech—according to Bradley, "the most

pathetic speech ever written"54—is utterly different, in

the monosyllabic simplicity of its diction, from the last

words of Hamlet or of Othello (as the same critic points

out); but it is surely at least equally as moving.

The attempt has been made to demonstrate that both

the longer, borrowed words and the shorter, native ones

have their own kinds of usefulness and effectiveness.

In a similar way one might show further that the lan-

54 Op. cit., p. 292.

guage is the gainer through the possession of both

instruments, largely by reason of the possibilities of

contrasting one with the other. The classical illustra-

tion of this contrast is the passage in Macbeth (Act II,

scene 2, ll. 60-63) :

Will all great Neptune's ocean wash this blood

Clear from my hand? No; this my hand will rather

The multitudinous seas incarnadine

Making the green one red.

in the last two lines of which the Latin and the Anglo-

Saxon diction seem deliberately set off in opposition to

each other. It is of course the same contrast, in less

striking fashion, that is always to be found in the English

language in actual practice.

Need any preference be given to the native word over

the borrowed or the borrowed word over the native?

As we have seen, the history of the English vocabulary

discloses the fact that different eras, as well as different

individuals, have answered this question variously.

Roughly speaking, Old English clung to the native word,

Middle English often preferred the borrowed one, while

in the Renaissance the question was hotly debated, and

"Latinist" and "Saxonist" ranged themselves into

opposing camps. But on the whole, the more liberal

attitude toward word-borrowing prevailed in the Renais-

sance and, after the conservative interlude of the

eighteenth century, has prevailed in general through

the Modern English period. Not that it is quite satis-

factory to generalize to this extent. If the eighteenth

century on the whole leans to a simpler and more native

diction, it is to be remembered that its prose literature

contains, in addition to the more Anglo-Saxon style of a

354 The Development of Modern English

Swift, an Addison, or a Goldsmith, the more elaborate manner of a Johnson, a Gibbon, and a Burke.

The Renaissance debate between Latinist and Saxonist has its occasional echoes in our own day. The advice is frequently given to the student of English composition that "other things being equal, the native word is the better word." This is probably sound enough, if one could be sure when other things are equal. Better advice, quite certainly, is that of William Hazlitt, who maintained55 that in the choice of words, the principle should be "the best word in common use." Words should be chosen, that is to say, not on the score of etymology but on that of familiarity and the accuracy with which they convey meaning. It will often happen that the accurate and the familiar word is also the native word. So much the better; but the basis of the choice is still not the etymology. Begin is a better word than commence not because it is the native word and the other the borrowed one, but because, in addition to having precisely the same meaning, it is the more homely and the more familiar. So much faith one may reasonably put in the theory of preferring native to borrowed words. Fortunately it is true, as we have indicated, that the native word, the shorter word, and the familiar word are often one and the same. But this is obviously a very different matter from the policy of seeking out the "Saxon" word as such and substituting it for the borrowed word that is now in more common use. That way lie folly and a pedantry quite as bad as that of the writer of the most inflated academic style. The language is and has long been so thoroughly committed to a com-

55 In "On Familiar Style."

posite word-stock that it can never be restored to the

standard of "purity." To substitute folkwain for

omnibus or steadholder for lieutenant is eccentricity that

borders on lunacy.56

One word more as to the implications of the composite

vocabulary to which Modern English is permanently

committed. While it is true that the native and the

borrowed word-stocks have to some extent their separate

excellences, and even more certain that their combined

resources may be skilfully exploited, it must be admitted

also that there is something unattractive in the picture

of a vocabulary often sharply divided into two parts.

The extent of the gulf may easily be exaggerated. For

one thing, not all borrowed words are on the same basis.

Indeed, in speaking of "native words" it is often cus-

tomary, and certainly justifiable enough, to include with

them Teutonic words of long standing even though not

Anglo-Saxon—especially Danish and Dutch words.

Furthermore, many French borrowings like aunt, beef,

and car are quite obviously not to be compared with

Greek or Latin loan-words like euhemerism,ratiocination,

or hyperesthesia. Nevertheless, it still remains that

there are in general two groups of English words that

have little in common, and that they are made up of the

native words on the one hand and the borrowed ones on

the other. As a result, there is something "undem-

ocratic," as Jespersen has well maintained,57 in a large

portion of the vocabulary, especially the literary vocab-

ulary, of English. Both Jespersen and Bradley have

56 For a statement of the issues that are involved in "Saxonism," see

the article of that title (pp. 514 and 515) in Fowler's Modern English

Usage.

57 Growth and Structure, p. 146. And cf. Bradley's Spoken and Written

English, p. 26.

356 The Development of Modern English

pointed out that some of these are merely eye-words, existing solely on the printed page and without fixed pronunciation. There are many that have a well-established pronunciation but that after all are intelligible only to those with some knowledge of Greek and Latin. Such words can never compete on an equal basis with words that are firmly rooted in an English soil and that strike an immediate response in the hearts as well as the minds of English-speaking peoples. Probably in no other language is it so necessary as it is in English to take definite precautions against using "big" words—words that require some sort of translation in order to be completely understood. It can scarcely be thought of as a source of strength that the danger of "overwriting" and pretentiousness is an ever-present one to him who deals in English words. This is not to deny the variety and the richness that the enormous composite vocabulary of English presumes; it is merely to affirm that even in words, there is such a thing as the embarrassment of riches. There are superfluities and ugly excrescences in our dictionaries that could well be pruned away. Be it repeated, however, that the disadvantages of the situation are of less moment than the advantages. If in English the danger of failing to strike the right note is particularly great, it may still be argued that to the real master of the instrument, the more complicated mechanism offers an opportunity for both rich and clear expression that is quite unequalled elsewhere.

References for Further Reading

The books listed on pages 420 and 421, and also: Johnson, E. L., Latin Words of Common English, New York (Heath), 1931.

Sources of the Vocabulary

357

Kent, R. G., Language and Philology, Boston (Marshall Jones), 1923.

Krapp, G. P., The English Language in America, New York (Century), 1925.

Mencken, H. L., The American Language, New York (Knopf), 3rd ed., 1923.

Smock, J. C., The Greek Element in English Words, New York (Macmillan), 1931.

CHAPTER X

The Making of Words

BORROWING from without is, as has been said, one of the two general methods of building the vocabulary. The other method is creation of new words by utilizing material that already exists in the language, whether its ultimate source is native or foreign. In the present chapter, we shall be concerned with the various types of word-formation or word-creation. Creation in the strictest sense of the term, be it said at once, is exceedingly rare; it is difficult indeed to find examples of words that are called into being from the void. This is not, of course, equivalent to asserting that there are only a few words of unknown etymology. Actually there are many; but for the great majority of such words, what is meant by this description is merely that the clue to their history has been lost. In all probability these words "of unknown etymology" have either been borrowed from foreign sources, or developed, by one of the recognized processes of word-formation, from native or even from foreign material. Pure root-creation is, for the most part, to be associated with the primitive stages of language, of which no record has been preserved. The stock example of a word "created" out of whole cloth in comparatively modern times is gas, a word devised by the Dutch chemist Van Helmont in the seventeenth century and adopted for international use. But it is known that the word was suggested to its creator by the Greek χάος

The Making of Words

359

(chaos), so that it can scarcely be said to be absolutely without etymology.

According to Mark Twain, Eve gave the name do-do to the bird in question for the excellent reason that, to her, it "looked like a do-do." Better authenticated examples of such virtuosity in root-creation are unfortu-

nately all but non-existent. When the need for a new word arises, it almost never occurs to the person who requires a label for a new thing or a new idea that sounds already existing in the language might be arbitrarily arranged into a new pattern. Yet it is evident that there are hundreds and indeed thousands (even if we limit our speculation to monosyllables made up of no more than three component sounds) of words that might be created,1 in the true sense of the term, to fill new needs.

Except for the somewhat dubious case of gas, the only English word commonly cited as a pure root-creation is kodak, a trade name that has won complete dictionary recognition.

There are, of course, many other trade names, with or without dictionary recognition, that have established themselves in wide present-day use. Usually, however, the method of coinage is not the arbitrary selection of sounds, but the combination of the syllables or initial letters of easily identifiable phrases. Such are Uneeda ("You need a"), Odorono, Asco ("American Stores Co."), Nabisco ("National Biscuit Co."), Socony ("Standard Oil

1 The number of possible monosyllables in English was actually calculated by Herbert Spencer; his figure was 108,264. Jespersen, who refers to this, Monosyllabism in English, New York (Oxford), 1929, pp. 8 and 9, finds the total too small, and estimates, as his own calculation, "rather more than 158,000 monosyllables"; " . . . only a small part . . . occur as actual words in the language" (p. 13).

360 The Development of Modern English

Co. of New York''), and so forth. Sometimes, the method is simply to reverse the letters of a word related in meaning, as has been done in Klim (from milk), the name for a brand of lactic powder; in Peredixo, the name of a facial cream, the first syllable of peroxide is kept and the letters of the others reversed. Again, a less radical change in an existing word or phrase may be brought about by respelling it, either more simply (if not more phonetically) or more fantastically. What difficulties in spelling may ensue for a younger generation constantly confronted with advertisements bearing these distortions it is not pleasant to contemplate! Examples of this kind of novelty are Hom-aid (with a pun to complicate matters further) for bread; Shurtred and Phiteesi for shoes, and the like. Ingenuity of this kind has "made its masterpiece" in the punning labels of Ford cars—Tudor (two door) and Fordor (four door).2

So far we have been dealing chiefly with recent, and usually conscious and sophisticated, types of word-creation. There is another general species of root-creation that extends so far back into the earlier stages of language that it was formerly taken3 to be the only kind of primitive speech and hence to explain the very origin of language. This group of words is made up of the "onomatopoetic" terms—words, that is to say, that "make their own names." If it were literally true that certain natural, human, or animal noises make the names by which they are known in language, one would be justified in calling this type of formation, in a very strict

2 Louise Pound, in "Word Coinage and Modern Trade Names," Dialect Notes, January, 1914, gives many additional illustrations.

3 Cf. p. 8.

The Making of Words

361

sense, pure creation. In point of fact, however, no word

of any language ever exactly reproduced the sound which

it described. Hence, it seems more accurate, as well as

less cumbersome, to label this method of word-making

"echoism" (Sir James Murray's term) rather than

"onomatopoeia," and to admit that there is an element

of conventional symbolism in the echoic words as a group.

Yet there is evidently a very real degree of root-creation

in these words. Certainly, for whiz, hiss, fizz, sizzle,

twitter, and titter, we need look up no "etymology" in

the usual sense of the term; the combination of a sibilant

or an explosive consonant with the short front [I] is

admirably adapted to convey the noise the word

embodies, and the word has obviously come into being in

order to echo an already existing sound. Not that all

echoic words are so nearly accurate. Heavier noises

seem less adequately conveyed by such terms as bang,

crash, and whack; but the difference is one of degree

rather than kind, [æ] having evidently been taken to

stand for a somewhat more solid and resounding noise

than [I]. In a similar way, cuckoo may pass for a satis-

factory rendering of the cry made by the bird, and hence

for the name of the bird itself; but bob-white employs a

more far-fetched echoism. Likewise, miaow serves more

felicitously for the cat than bow-wow for the dog.

A great deal more might be said on the use of imitation

-echoic or symbolic-in the creation of words. If it has

not been proved, and it seems altogether unlikely to be

proved, that imitation made the first beginnings of

human speech, it is evident that it still runs through

all human language. Words like mama, papa, and baby

are found in almost all tongues, and the usual explana-

tions of a cognate relationship or of borrowing will

362 The Development of Modern English

scarcely apply.⁴ What is almost certain is that the

labials m (the sucking sound), p, and b are almost the

first sounds any child acquires, and that their repetition,

with the vowel [ɑː] interspersed, gives the words in

question. Imitation, then, on the part of the parents

perpetuates their use and conventionalizes their mean-

ings—meanings that are thereupon gradually acquired

by the child, also through imitation. There is an obvious

use of imitation in such echoic words as have been cited

in the preceding paragraph, whether the particular

interpretation of the sound that forms the word is

peculiar to English (as in fizz, hiss, and titter), or whether

it is a borrowing from another language.⁵ In point of

fact, the imitations that different languages provide for

the noises made by animals do vary widely; compare

English tweet-tweet for the bird's chirp, with Russian

tsif-tsif, or, for the rooster's crowing, English cocka-

doodle-doo, German kikeriki, French coquerico.

When we proceed from echoic imitation to more remote

symbolism we are perhaps on more dubious ground.

Yet, it seems quite certain that sound symbolism is a

factor very much more frequently present in the creation

of words than is commonly appreciated.⁵ It is not

of course confined to our own language, though it is

convenient to take our illustrations chiefly from Middle

and Modern English, which have been particularly rich

⁴ Jespersen, for example, denies that papa and mama in English,

German, Danish, Italian, and so on are borrowed from the same words in

French, though it is conceivable that a French fashion may have dictated

their continued use beyond the age of the nursery (Language, p. 159).

⁵ One of the few adequate treatments of it is Chapter XX of Jespersen's

Language (pp. 396–411); for briefer accounts, see Bradley, Making of

English, pp. 156–159, and L. P. Smith, English Language, pp. 101–104.

I have taken some of my illustrations from all three sources.

The Making of Words

363

in it. A few of the clearer cases of a generally accepted symbolism in sounds are appended. The thin vowel [I] is used again and again to stand for small size or slightness: little, kid, slim, chit, imp, slip, pigmy. Sometimes, it is raised and prolonged to [i:]—a further degree of smallness. Thus, teeny is a child's word for tiny (teeny-weeny means something still more minute), and leetle is, in provincial speech, a more intense kind of little; child (O. E. cild, with the [i:] vowel), wee, and the very common diminutive endings spelled -y and -ie and pronounced [i] are further illustrations. Consonants have their symbolic associations as well as vowels. The use of the breath with some force is suggested by bl6—blow, blast, blub, blab, bubble, bluster; an awkward kind of movement by fl—fling, flounce, flounder, flop, flump, flurry; and the abrupt cessation of sound or movement by the final voiceless stops p, t, and k—pop, clip, snip, rap, pat, crack. Smith7 points out the neat differentiation between the type of noise or action just mentioned and that symbolized by a final sh, in which the sound or movement “does not end abruptly, but is broken down into a mingled mass of smashing or rustling sounds, as in dash, splash, smash, etc.”

Before leaving the subject, it may be well to voice the warning that symbolism should not be looked for everywhere. To do that would be to repeat the erroneous conception of the Greeks as to the etymology or the “true meaning” of the word, the doctrine that there is a

6Perhaps the violent expulsion of breath stands symbolically for the idea of a disgusted rejection in another group of bl words: blasted, blamed, bloody, blessed (ironically used), blowed, blighted, blooming, blistering, blithering, and so forth.

7Op. cit., p. 103.

364 The Development of Modern English

natural correspondence between sound and sense. It is

easy to be deceived here. The words that we are familiar

with are apt to strike us as having an inherent appro-

priateness, in their very sounds, to the ideas for which

they stand. So too the words that evidently are in

some degree echoic or symbolic are likely to seem the

inevitable and perfect combination of sounds for the

designated purpose. What seems to the native speaker

an obvious bit of appropriate symbolism may, however,

leave the foreigner cold; he is likely to prefer the render-

ing of his own language as more accurately descriptive-

as, in fact, in every way a "better" word. The German

will prefer flüstern to the English whisper, while to the

Frenchman both will seem inferior to chuchoter. Such

considerations should lead one to go cautiously both in

finding symbolism and in deeming even unquestioned

examples of it to be the perfect fitting of sound to sense.

In the illustrations that have been cited, it will be

observed how echoism, in the stricter sense, shades into

a more conventional symbolism. Echoism and sym-

bolism (together with the rare case of arbitrary re-com-

bination of sounds) between them evidently account for

a surprisingly large number of words. Words, however,

are still more commonly added to the vocabulary not by

any sort of new creation but by a new use that is made of

words or word-elements already existing in the language.

The general line of demarcation, that is to say, for the

two great types of word-making is that between root-

creation on the one hand and what has been called

"adaptation," or differentiation, on the other.

A type of adaptation no longer active in English, but

in former times responsible for the similarity in form and

sound of a number of groups of words related in meaning

The Making of Words

365

is that known as gradation. In Old English, the so-called ablaut series of the strong verbs are instances of this kind of variation in form with corresponding variation in meaning. Verbs of the modern language like ring, rang, rung and drive, drove, driven perpetuate these gradations, though the number of variants in each series has been reduced from four to three through the loss of one of the two distinctive preterit forms, and often to two, through the falling together of preterit and past participle. But gradation is not confined to the variations of verbs. The principle is rather this: A consonantal framework, like r-d, s-ng, or b-r, stood for a certain generalized root-idea, and was given specific application by the variation in form of the medial vowel. Thus, one gradation-group included the noun song, as well as the verbal forms sing, sang, sung; another included not only ride, rode, ridden, but also road and raid (for the road was the place of riding, and the raid an all too frequent objective—in Old English, “to ride a raid” was an idiom parallel to “to sing a song”). The third, represented by the framework b-r, is one of the oldest Indo-European roots; in English, it appears not only in bear, bore, born and borne, but also in the nouns bairn, birth, bier, barrow, and burden. However, fruitful as this means of building new words was in older stages of the language, it is evident that gradation has long since ceased to be an active principle in word-formation.

Much more common in word-making than gradation, and abundantly active in every stage of the language, is the phenomenon known as “composition.” A compound of course is formed by the act of joining two or more words into a single one, with the result of creating

366 The Development of Modern English

a new unit of expression. The simplest type is that known as the "full-word compound," in which separate and independent elements are still distinguishable. Almost any conceivable combination of the parts of speech may thus be used to form a new word.8 To give a few examples, we have the joining of noun and noun, as in housetop, weekend, and railroad; noun and adjective, as in land-poor, coal-black, and airtight; adjective and noun, as in blackberry, hothouse, and sweetmeat; adverb and noun, as in upshot, overhead, and downfall; adverb and verb, as in overthrow, upturn, and inset; and so forth. The part of speech formed may be different from that of either of the component elements: the adjective shipshape is made up of noun and noun (or verb), the noun upset of adverb and verb, the noun lean-to of verb and adverb. Finally, it is not at all uncommon to find three words forming a compound, as in nevertheless, notwithstanding, hand-to-mouth, and brother-in-law.

One may observe, incidentally, in the compounds cited in the foregoing paragraph, how exceedingly inconsistent is the use that modern practice makes of the hyphen, which occasionally, but by no means always, proclaims the word as a compound. Indeed, there is frequently no real difference between the status of two words printed separately and that of two words printed as one; as Krapp points out,9 the "out of" and "into" of the phrase "out of the frying-pan into the fire" might more logically be treated in the same way. Similarly,

8 For fuller illustration, see Bradley, pp. 113-115; or Krapp, Modern English, pp. 187-189. Bradley's entire treatment of the theory of composition, pp. 111-128, is particularly valuable.

9 P. 189.

The Making of Words

367

different usage with regard to "none the less" and "nevertheless" is merely capricious. The hyphen may be retained in lean-to for the good reason that to do so avoids a difficulty in interpretation; but who can justify the practice of keeping the hyphen (as most dictionaries do) where the to comes first, as in the immediately intelligible and everyday compounds today and tomorrow? The only worse thing is to discriminate between "to-day" and "tomorrow."

Frequent as compounds are in Modern English, it is nevertheless the fact that the language has lost much of the ability it once possessed of forming compounds with absolute freedom. It is evident that in this respect as in so many others, German is nearer than Modern English to the state of affairs in Old English. Our native compounds have often been replaced with Latin or French borrowings that are not compounds, as when treasure replaced gold-hoard and when medicine superseded leechcraft. The very fact that Latin and French lack the compound-making ability of Greek and German helps to account for the decline of the compound in English, for the first two languages have been levied upon much more extensively for the English vocabulary than have the latter two. Something may be said of the types of compounds, from the point of view of the parts of speech making them up, that are obsolete or obsolescent in Modern English. We no longer make free use of the combination of adverb and verb (or, sometimes, a participle or gerund)—the type involved in output, income, uprising, and downtrodden. New formations on the model of down-sitting and forthcoming would probably strike us as excessively Teutonic, though their equivalents would have been quite acceptable in

368 The Development of Modern English

Old English. Bradley notes10 that the Scotch dialect has used this type of formation more actively than standard English, and recent examples of it have come in through this source; such are uptake, upkeep (more favored in American than than British use), and outcome (introduced into literary use by Carlyle and long objected to). The reverse of this formation—that is, the combining of verb and adverb to make a noun—is likewise somewhat rare in recent English; it is usually found rather in colloquial than in formal speech, as in lean-to, dug-out, walk-over, wash-out. The parallel case of noun of agency plus adverb is similarly seldom found, save in less literary language (for example, listener-in, hanger-on, and runner-up); Shakespeare’s “snapper-up of unconsidered trifles” has probably suffered a loss in dignity. This mode of formation, however, unlike the preceding one, has enjoyed what vogue it has had, chiefly within the modern period.

There is no doubt that a compound obviously made up of two distinct elements is often felt to be clumsy and unwieldy. Not infrequently, a compound is taken apart and only one of its elements used, though the sense of the fuller word is retained; thus, “to get a person on the ‘phone” is certainly the more common way of putting the idea, and to “go by rail” more frequent though less elegant than “to go by railroad.” The general fondness of English for clipped economy of expression operates here, as it does when taxi is substituted for taxicab (cab itself is a shortening of cabriolet, and taxi of taximeter) and bus for omnibus. English, incidentally, has no monopoly on this type of shortening, for bus is

10 P. 124.