Seven Stories Press

Works of Radical Imagination


With Alan Hodge

Introduction by Patricia T. O'Conner

In late October 1939, Robert Graves wrote to Alan Hodge: “I have begun a new book, about English.” Graves and Hodge had recently completed a social history of the between-wars period called The Long Week-End. Now they embarked on this new project, “a handbook for writers of English Prose,” to be called The Reader Over Your Shoulder.

The world was in total upheaval. Graves had already fled Majorca three years earlier at the start of the Spanish Civil War. As they labored over their new writing project, Graves and Hodge witnessed the fall of France and the evacuation of Allied forces at Dunkirk. In early September 1940 began the bombing of London by the German Luftwaffe, a concentrated effort to destroy the resolve of the English people. Graves’s and Hodge’s idea was simple enough: at a time when their whole world was falling apart, the survival of English prose sentences, of writing that was clear, concise, intelligible, had become paramount if hope were going to survive the onslaught. They came up with forty-one principles for writing, the majority devoted to clarity, the remainder to grace of expression. They studied the prose of a wide range of noted authors and leaders, finding much room for improvement. Quoting grammarian and bestselling author Patricia T. O’Conner from her new introduction, “With a new war to be won, the kingdom couldn’t afford careless, sloppy English. Good communication was critical.”

The book they would write would turn out to be one of the most erudite, and at the same time one of the most spontaneous and inspired, ever to take on the challenge of writing well. O’Conner in her introduction describes The Reader Over Your Shoulder as nothing less than “the best book on writing ever published.” The present edition restores, for the first time in three-quarters of a century, the original, 1943, text, which in subsequent printings and editions had been shortened by over 150 pages, including much of the heart of the book.


Buying options

“The best book on writing ever published.”

“A never-ending pragmatic pleasure.”

The Reader Over Your Shoulder is subtitled A Handbook for Writers of English Prose, but it is also an inspiration for readers . . . . The questions Graves and Hodge ask, the objections they raise to the particular sentences exhibited, are never pedantic; they arise from a decision to take the prose seriously.”

“It is one of the three or four books on usage that deserve a place on the same shelf with Fowler.”

“Robert Graves is a Renaissance figure, among the most generous, self-willed, unseemly and brilliant writers of our century”

blog — September 07

The Peculiar Qualities of English

The e-book of Robert Graves's The Reader Over Shoulder is free on our site through September 11, 5PM EST. Please find below an excerpt from the first chapter.

The most ancient Eu­ro­pean languages—­those that have longest avoided infiltration by other languages—­are the most complicated in their grammar and syntax. The age of a language can be roughly guessed by a count of its declensions, conjugations, moods, tenses, voices, cases, genders and numbers. Latin is clearly less ancient than Greek, since it has no ‘­middle voice’, no ‘dual number’ and no ‘optative mood’—­thus in Latin at least seven words are needed to express the sentence ‘If only you two thieves had drowned yourselves’, but in Greek only four. French is clearly less ancient than Latin, since it has no separate neuter gender and does not decline its nouns; also, its conjugations are far simpler. En­glish is clearly less ancient than French: except for its pronouns, it is ­free of gender differentiation.

Grammatical simplicity is the mark of a vernacular. The word ‘vernacular’, formed from the Roman historian Varro’s phrase vernacula verba, ‘unliterary expressions used by slaves or serfs’, has often since been used loosely to mean ‘the native language of a peasantry’; but few of Varro’s slaves can have been native Italians—­they may have been Greeks, Bithynians, Africans or Germans—­and the language they spoke among themselves was a mixed lingo, sometimes called ‘camp Latin’, which ­later developed in Italy into modern Italian, in Spain into Spanish, in France into French. Properly speaking, then, a vernacular is a lingo, or language of domestic con­ve­nience, compounded of the languages spoken by master and alien slave. It has a less complicated grammar and syntax than the languages from which it springs, but rapidly accumulates words as the slaves become freedmen, and their ­children are born as freemen, and fi­nally their great-­grandchildren, marrying into their masters’ families, are accepted as cultured ­people with full rights as citizens. The historical origins of German, which is not very primitive in structure, are obscure; modern German, however, is not a vernacular in Varro’s sense, but a late artificial compound of several kindred dialects, with a far smaller vocabulary of early borrowings from the Latin and Greek than the other languages of Western Eu­rope.

En­glish is a vernacular of vernaculars. It began in the eleventh ­century as the lingo used between the Norman-­French conquerors and their Anglo-­Saxon serfs, and though it became a literary language in the ­fourteenth ­century has never crystallized in the way that Italian, French and Spanish have done. A proof of this is that no writer of En­glish would be credited with a perfect literary style merely ­because he had exactly modelled himself on some native paragon—­say, Addison in ­England, or Emerson in the United States—as Italians, Spaniards and Frenchmen might be ­after modelling themselves, respectively, on Boccaccio, Cervantes and Bossuet. To write En­glish well, it is generally agreed, is not to imitate, but to evolve a style peculiarly suited to one’s own temperament, environment and purposes. En­glish has never been jealously watched over by a learned Acad­emy, as French has been since the seventeenth ­century; nor protected against innovations ­either by literary professionalism, as with Italian, or, as with Spanish, by the natu­ral decorum of the greater part of ­those who use it. It is, indeed, an im­mense, formless aggregate not merely of foreign assimilations and local dialects but of occupational and ­house­hold dialects and personal eccentricities.

The general Eu­ro­pean view is that En­glish is an illogical, chaotic language, unsuited for clear thinking; and it is easy to understand this view, for no other Eu­ro­pean language admits of such shoddy treatment. Yet, on the other hand, none other admits of such poetic exquisiteness, and often the apparent chaos is only the untidiness of a workshop in which a ­great deal of repair and other work is in pro­gress: the benches are crowded, the corners piled with lumber, but the old workman can lay his hand on what­ever spare parts or accessories he needs or at least on the right tools and materials for improvising them. French is a language of fixed models: it has none of this workshop untidiness and few facilities for improvisation. In French, one chooses the finished phrase nearest to one’s purpose and, if ­there is nothing that can be ‘made to do’, a long time is spent getting the Works—­the Acad­emy—to supply or approve a new model. Each method has its own advantages. The En­glish method tends to ambiguity and obscurity of expression in any but the most careful writing; the French to limitation of thought. The late Sir Henry Head was once preparing an address on neurology for a learned society in Paris. He wrote it in what he hoped was French, but took the precaution of asking a French professor to see that it was correctly phrased. The manuscript was returned marked: ‘pas français’, ‘pas français’, ‘pas français’, with suggested alterations; but almost ­every ‘pas français’ could be matched with a ‘pas vrai’, ­because the amendments in français impaired the force of the argument.

As for the view that En­glish is illogical: it certainly differs greatly in character from French, Italian, Spanish and German, which are claimed to be logical languages. ­These are all able codifications of as much racial experience as can be translated into speech: theoretically, each separate object, pro­cess or quality is given a registered label and ever afterwards recognized by label rather than by individual quality. Logical languages are therefore also rhetorical languages, rhe­toric being the emotionally persuasive use of labels, with ­little concern for the ­things to which they are tied. En­glish has always tended to be a language of ‘conceits’: that is, except for the purely syntactical parts of speech, which are in general colourless, the vocabulary is not fully dissociated from the imagery out of which it has developed—­words are pictures rather than hieroglyphs.

Matthew Arnold, who as a critic did insufficient justice to the peculiar genius of the En­glish language, suggested in his essay on the ‘Influence of Literary Academies’ (1875) that:

‘The power of French Lit­er­a­ture is in its prose-­writers, the power of En­glish Lit­er­a­ture in its poets. While the fame of many poets of France depends on the qualities of intelligence they exhibit, qualities which are the distinctive support of prose, many of the prose-­writers . . . depend wholly for their fame on the imaginative qualities they exhibit, which are the distinctive support of poetry.’

The truth is that the French are not plagued by their meta­phors tending to get out of hand and hamper the argument; whereas En­glish writers of prose or poetry find that, so soon as a gust of natu­ral feeling snatches away the merely verbal disguise in which their phrases are dressed, the pictorial images stand out sharply and ­either enliven and enforce the argument or desert it and go on a digressive ramble. En­glish writers seldom have any feeling for purity of literary form in the Classical sense: it is both their strength and their weakness that imaginative exuberance breaks down literary restraint.

‘Fixed’ En­glish, which may be dated from Dr. Johnson’s Dictionary, completed in 1755, fulfils the need of a safer, less ambitious language arranged on the same system and dedicated to the same uses as French—­a language of agreed preconceptions. ‘Fixed’ En­glish makes pos­si­ble a French-­English, English-­French, or a German-­English, English-­German dictionary. Each foreign label has its En­glish counterpart: ‘Glory’ is matched (not very satisfactorily) with ‘la gloire’ and ‘der Ruhm’; ‘le matelot’ and ‘der Matrose’ with ‘sailor’. ‘Fixed’ En­glish compares well enough with other languages, but is often more mechanically, and therefore more correctly, employed by foreigners than by ­those whose mother-­tongue it is and who are always inclined to slip back into ­free En­glish. ‘Fixed’ En­glish is an easy language to learn, like colloquial Arabic; but of ­free En­glish, as of scholarly Arabic, no wise person ­will ever claim final mastery—­there is no discovered end to ­either language. ‘Fixed’ En­glish is never more than momentarily fixed. The conventional, hotel-­manager En­glish that foreigners learn is always a ­little stilted and a ­little out of date by the time that the book from which they learn it is published; and twenty years ­later the book ­will read very quaintly.

En­glish, ­whether ‘fixed’ or ‘­free’, has certain unusual advantages in structure. In the first place, it is almost uninflected and has no genders. The Romance and Germanic languages, not having had occasion to simplify themselves to the same degree, still retain their genders and inflections. ­There is no logical justification for genders. They are a decorative survival from a primitive time when the supposed sex of all concepts—­trees, diseases, cooking implements—­had to be considered for the sake of religious convention or taboo. Yet even new scientific words have to decide, so soon as coined, on their hy­po­thet­i­cal sex. Writers of the Romance and Germanic languages have an aesthetic objection to a genderless language. But when a language is used for international exchange of ideas the practical disadvantages of gender are generally admitted to outweigh its decorative qualities. Gender is illogical, in being used partly to express ­actual sex, e.g. le garçon, la femme, and partly to dress words up, e.g. la masculinité, le féminisme; le festin, la fête. If one does wish to give sex-characteristics to concrete objects or abstractions (as, for example, masculinity to ‘sword’ and ‘pen’ and femininity to ‘Parliament’), the existing gender is an ­actual hindrance to any such renewal of my­thol­ogy. An inflected tense has a certain beauty from which writers in ­these languages refuse to be parted; but, for merely practical uses, an inflected tense such as je serai, tu seras, il sera, nous serons, vous serez, ils seront seems unnecessary to Britons and Americans, particularly since the French have dropped noun and adjective inflections almost as completely as they have.

The eventual disappearance of Norman-­French from ­England ­after the Conquest was never in doubt once Anglo-­Saxon had been simplified to meet the needs of the French-­speaking invaders. Anglo-­Saxon was deficient in words to fit the new methods of trade and government, and ­these had to be borrowed from French, which had a closer connexion with Rome, the source of all con­temporary civilization. Passing the stage of Broken Saxon, the new vernacular developed an easy grammar and syntax, a modification of Anglo-­Saxon, but with French turns wherever a ­legal or literary subtlety was needed. The vocabulary, though enormously enriched with Norman-­French and Latin words of advanced culture, remained Anglo-­Saxon in foundation: En­glish words of Anglo-­Saxon origin, though not half so numerous as Romance words, are used about five times as often. One feature of the happy-­go-­lucky development of En­glish was that adjectives ­were made to do ser­vice for nouns, nouns for verbs, and so on; ­until by Elizabethan times it could be said that all parts of speech in En­glish ­were interchangeable.

This interchangeability is a ­great help to accurate expression; for example, where an adjective formed in the usual way from a noun has wandered slightly from its original sense. If one wishes to discuss the inflections of a verb and does not wish to write ‘verbal inflections’, ­because ‘verbal’ means ‘of a word’ rather than ‘of a verb’, one is ­free to write ‘verb inflections’, using a noun for an adjective.

Further gains to En­glish in this early period ­were: the wide choice of prefixes and suffixes which the pooling of the wealth of both languages gave, the use of auxiliary words to help out the verb tenses, and the new freedom won by prepositions. ­There is a greater richness of prepositions in En­glish than in any other language of Western Eu­rope: for instance, the French de has to bear the ­whole burden of the En­glish prepositions ‘of’, ‘from’, ‘out’, and à of ‘at’, ‘to’, ‘till’, while German has no separation between ‘of’ and ‘from’; ‘into’ and ‘out of’ are double prepositions with no equivalent in ­either French or German.

The British have long been conscious of the extreme subtlety of their language. James Clarence Mangan, in his humorous essay My Bugle and How I Blow It, published in the early 1830s, wrote about one use of the preposition ‘in’:

‘I am the Man in the Cloak. In other words, I am by no manner of means the Man of the Cloak or the Man ­under the Cloak. The Germans call me “Der Mensch mit dem Mantel”, the Man with the Cloak. This is a deplorable error in the nomenclature of that other­wise intelligent ­people: ­because my cloak is not part and parcel of myself. The cloak is outside and the man is inside, but each is a distinct entity. I admit you may say, “The Man with the Greasy Countenance”, thus also Slawken-­Bergius (vide “Tristram Shandy”) calls his hero “The Stranger with the Nose”, for, however long, the nose was an integral part of the individual. With me the case is a ­horse of a dif­fer­ent colour. I do not put my cloak on and off, I grant, but I can do so when I please: and therefore it is obvious to the meanest capacity that I am the Man in the Cloak and no ­mistake.’

Mangan’s objection to the German idiom could have been strengthened by an opposite objection to the French ‘l’homme à la redingote’, where greater emphasis is laid on the cloak than on the man. Mangan, as he says, is not ‘the Man of the Cloak’; yet, ‘l’homme dans la redingote’ is no more French than ‘Der Mensch in dem Mantel’ is German. And nearly three centuries before Mangan Sir Philip Sidney had written in his Apology for Poetry:

‘En­glish giveth us ­great occasion to the right use of ­matter and manner, being indeed capable of any excellent exercising of it. I know some ­will say that it is a mingled language. And why not so much the better, taking the best of both the other? Another ­will say it wanteth Grammar. Nay, truly it hath that praise, that it wanteth not Grammar, for Grammar it might have, but it needs not; being so easy of itself and so void of ­those cumbersome differences of Cases, Genders, Moods, and Tenses, which I think was a piece of the Tower of Babylon’s curse, that a man should be put to school to learn his ­mother tongue. But for the uttering sweetly and properly the conceits of the mind, which is the end of speech, that it hath equally with any other tongue in the world: and is particularly happy in compositions of two or three words together, near the Greek, far beyond the Latin: which is one of the greatest beauties can be in a language.’

The growth of En­glish as a common language of conqueror and conquered had one ­great disadvantage: the slowness with which it arrived at a common convention for the pronunciation and spelling of words. Neither the French nor the Anglo-­Saxons could form their mouths properly for the management of the ­others’ language, particularly of the vowels; yet the French scribes had to write Saxon words in their ­legal rec­ords, and the Saxon scribes in submitting accounts in writing had to adopt a convention of spelling which their masters would
understand. Each district, too, had its dif­fer­ent dialect. East Midland fi­nally became the dominant one, but words ­were brought into it from other dialects with dif­fer­ent spelling conventions; when at last a general convention was agreed upon, it was (and remains) a tissue of inconsistencies—­the most serious handicap to En­glish as a language for international use.

The termination -­ough, for instance, occurs in words as ­differently sounded as ‘rough’, ‘bough’, ‘cough’, ‘thorough’, ‘though’, ‘through’, and ‘hiccough’: the gh represents what was once an Anglo-­Saxon guttural corresponding with the Greek letter chi. This guttural the Norman-­French could not, or would not, pronounce: ­either they made an f sound of it or they sounded the vowel and left out the consonant altogether. It is probable that this habit became fash­ion­able among ­people of Saxon blood who wished to pass as members of the ruling caste, and except in the North, where it lasted much ­later and still survives in many dialect words, the gh passed out of the spoken language and, in the written, remains merely as an historic relic. To the dif­fer­ent pronunciation of vowels in dif­fer­ent parts of the country, ­these -­ough words are sufficient witnesses; as in the ch in ‘Church’, ‘Christian’, ‘chivalry’ and ‘pibroch’ to the inconsistency of the spelling convention. The trou­ble was that the scribes had only twenty-­six letters (of which x, k, and q ­were redundant) to express forty-­three common sounds. They tried vari­ous methods, such as doubling a consonant to show that the vowel which came before was short, e.g. ‘batter’, and putting a final ‘e’ ­after a consonant to show that the vowel was long, e.g. ‘bone’. None of ­these methods could be used consistently while the pronunciation was still so vari­ous; and the scribes did all that could be expected of them, short of inventing a new alphabet.

Spelling conventions have changed continually ever since and are not yet stabilized. The word ‘mutton’, before its spelling was thus fixed early in the reign of James I, had been spelt: ‘moltoun’, ‘motoun’, ‘motone’, ‘motene’, ‘motonne’, ‘motton’, ‘mouton’, ‘muton’, ‘muttoun’, ‘mutown’ and ‘mutten’. ‘Button’, stabilized at about the same time, had had a still greater range of spelling variations including ‘botheum’ and ‘buttowne’. Yet even ­after this newly borrowed Romance words in ‘-­on’ ­were not similarly En­glished as ‘-­on’: balon, marron, musqueton, salon, ­were spelt ‘balloon’, ‘marroon’, ‘musketoon’, ‘saloon’. This ‘oo’, which also represented the narrow vowel in the word ‘good’, was then confused with the broader ‘oo’ sound as in ‘food’ and ‘moon’. In more recent times the En­glish have ­either spelt and pronounced the newly borrowed words in -­on in the French way, if they are words of limited use, such as soupçon and raison d’être; or, if they are words capable of popu­lar use, they have En­glished them much as they did the words first borrowed. So bâton, as in ‘conductor’s baton’, is made to rhyme with fatten; and on the barrack-­square échelon has come to rhyme with ‘stretch along’. However, on the ­whole the English-­speaking ­peoples have become more willing to pronounce and spell foreign words as they are spelt and pronounced in their countries of origin. Broadcasting assists this tendency: for example, ‘garridge’ for ‘garage’ would have become general in Britain but for the French pronunciation insisted on by the B.B.C. This new re­spect for accuracy of pronunciation has made ­things rather worse than before. For instance, the Irish word ‘lough’, not long borrowed, has added one more pitfall to the pronunciation of words in ‘-­ough’. Conscientious En­glish travellers to Ireland try to manage the sound, but the dictionaries compromise by directing that it ­shall be pronounced ‘lok’—­‘k’ is at any rate a guttural and nearer to the original sound than ‘f’.

Certain advantages have been derived even from ­these confusions. Where ­there has been more than one pronunciation of a word, it has very often been split into two words, each devoted to a dif­fer­ent sense and usually spelt differently. ‘Through’ and ‘thorough’ provide an instance of differentiation both of spelling and pronunciation; ‘flower’ ‘flour’ ‘gest’ and ‘jest’ of differentiation in spelling only. It sometimes happens that the same word is borrowed more than once from the same language, the first borrowing having already changed in sense. Thus the second borrowing becomes a new En­glish word and, the spelling convention having meanwhile also changed, is easily distinguished from the early borrowing. The word ‘saloon’ was the eighteenth-­century En­glish equivalent of the French salon, meaning a reception-­room in a palace or ­great ­house. In early Victorian times it began to be vulgarized, and now has come to mean merely a large room in a commercial establishment. Salon, borrowed again, means the reception-­room of a lady of fashion, where wits and notabilities assem­ble for mutual entertainment. Similarly, ‘antick’, with its more modern forms ‘antic’ and ‘antique’, pres­ent very dif­fer­ent views of ­things antiquated.

In accepting En­glish, one must accept the inconsistent spelling cheerfully, not only for the practical ser­vice it has given in the differentiation of meaning, but on its own account, as one learns to accept and even love the irregularities of a friend’s face. ­There is to the En­glish eye something distasteful in phonetic spelling. Attempts to force it upon the language, though supported by all the logic in the world, are unavailing—­‘because of the ugly look the words have—­too much “k” and “z” and “ay”.’ One would have less objection to phonetic or ‘simplified’ spelling if it could be introduced retrospectively in one’s grand­father’s days; but nobody likes to make such sacrifices for the sake of his grandchildren.

As the regularizing of spelling and pronunciation on a rational basis has never succeeded, so the permanent limiting of par­tic­u­lar words to par­tic­u­lar senses, the fixing of En­glish, has never come to much. ­There have been professorial panjandrums who have undertaken the reformation of the language as their life’s task; ­there have even been short periods, usually ­after a time of national disturbance, when the governing classes have had fits of tidiness and thought about putting the dictionary into better order. But En­glish has always resisted attempts to cramp its growth. Alexander Gil (Headmaster of St. Paul’s School when Milton was a pupil) wrote in 1619 in his Logonomia Anglica to complain of the ‘new mange in speaking and writing’, for which he held Chaucer originally responsible.

‘O harsh lips! I now hear all around me such words as common, vices, envy, malice; even virtue, study, justice, pity, mercy, compassion, profit, commodity, colour, grace, favour, ac­cep­tance. But whither, I pray in all the world, have you banished ­those words which our forefathers used for ­these new-­fangled ones? Are our words to be exiled like our citizens? Is the new barbaric invasion to extirpate the En­glish tongue? O ye En­glishmen, on you, I say, I call in whose veins that blood flows, retain, retain what yet remains of our native speech, and what­ever vestiges of our forefathers are yet to be seen, on ­these plant your footsteps.’

He was not heeded, nor ­were the literary oligarchs of the early and ­middle eigh­teenth ­century who tried to impose ‘decorum’ permanently on En­glish. We find Jonathan Swift vainly protesting in his Bickerstaff Papers (1723) against: the ‘abundance of polysyllables introduced by the late war, speculations, operations, preliminaries, ambassadors, communication, circumvallation, battalions’; and at the same time against such slang as ‘banter, uppish, bamboozle, kidney, mob, sham, bully, shuffling and palming’—­just as in October 1941 a Mr. Faning wrote to the Daily Telegraph:

‘Sir—­I view with concern the increasing introduction of German words into our language—­such words as “blitz”, “panzer”, “luftwaffe” and ­others.

Surely it is sad to reflect on the ease with which Germany can invade our language, if not our shores!

If no check is put upon this practice ­these horrible words ­will become incorporated into our En­glish vocabulary. German is uncouth and hardly fit for a civilized nation.’

Nobody ­will ever succeed in annulling the natu­ral right, which ­every English-­speaking person can claim, of borrowing words from foreign languages, coining words, making new distinctions in sense between alternative forms of the same word, resurrecting ancient words, or using what­ever grammatical oddities he may please—­provided he can satisfy his neighbours, colleagues, or readers, that the coining, the distinction, the borrowing, the resurrection, or the grammatical oddity is necessary. The proof of necessity ­will be that they adopt the form themselves.

The practice of making new words by declaration is of long standing. For example, the word ‘mumpsimus’, meaning ‘an erroneous doctrinal view obstinately adhered to’, was first put into currency by Henry VIII, in a speech from the throne in 1545. He remarked, ‘Some be too stiff in their old mumpsimus, ­others be too busy and curious in their sumpsimus.’ He was referring to the familiar story of a priest who, on being reproved for reading in the Mass ‘quod in ore mumpsimus’ instead of quod in ore sumpsimus’ (‘which we have taken in our mouths’), his Missal being miscopied, replied that he had read it with an m for forty years, ‘and I ­will not change my old mumpsimus for your new sumpsimus.’ The word has held its own since, though the doctrinal sense has lost its importance compared with the scholastic sense: it now means ‘an established manuscript-­reading which, though obviously incorrect, is retained blindly by old-­fashioned scholars’.

In several countries the declaration of an individual is not enough to ­settle a word, even though this is built up regularly from native material or correctly formed from Latin or Greek. The declaration has to be endorsed by the Ministry of Fine Arts or the recommendation of a university committee or learned society. The En­glish word ‘defeatism’ is formed from the French word défaitisme current in 1915, which is not officially French: that is to say, in the early Twenties Marshal Foch, as a member of the Académie Française, vetoed its adoption into the Dictionary, on the ground that it was an un-­French concept and intolerable. It appears apologetically, however, in A. Dauzat’s Dictionnaire Etymologique, 1938, as: ‘1904, appliqué aux Russes; vulgarisé 1915.’

A passage from Lewis Carroll’s Alice Through the Looking-­Glass.

‘Humpty-­Dumpty said: “­There’s glory for you.” “I ­don’t know what you mean by ‘glory,’ ” Alice said. Humpty-­Dumpty smiled contemptuously. “Of course you ­don’t till I tell you. I meant, ‘­There’s a nice knock-­down argument for you!’ ” “But ‘glory’ ­doesn’t mean ‘a nice knock-­down argument’,” Alice objected. “When I use a word,” Humpty-­Dumpty said in rather a scornful tone, “it means just what I choose it to mean, neither more nor less.” “The question is,” said Alice, “­whether you can make words mean so many dif­fer­ent ­things.” “The question is,” said Humpty-Dumpty, “which is to be Master—­that’s all.” Alice was too much puzzled to say anything, so ­after a minute Humpty-­Dumpty began again. “­They’ve a temper, some of them—­particularly verbs, ­they’re the proudest—­adjectives you can do anything with, but not verbs—­however, I can manage the ­whole lot of them! Impenetrability! That’s what I say!”

“Would you tell me, please,” said Alice, “what that means?” “Now you talk like a reasonable child,” said Humpty-­Dumpty, looking very much pleased. “I meant by ‘impenetrability’ that ­we’ve had enough of that subject, and it would be just as well if you’d mention what you mean to do next, as I ­don’t suppose you mean to stop ­here the rest of your life.” “That’s a ­great deal to make one word mean,” Alice said in a thoughtful tone. “When I make a word do a lot of work like that,” said Humpty-­Dumpty, “I always pay it extra.” “Oh!” said Alice. She was too much puzzled to make any other remark. “Ah, you should see ’em come round me of a Saturday night,” Humpty-­Dumpty went on, wagging his head gravely from side to side, “for to get their wages, you know.” ’

Lewis Carroll himself added several words to the language; and not only in the United States, where novelty of expression is widely exploited as a proof of national vigour, but in ­Great Britain too, enterprising ­people are every day ­doing the same. The greater part of their inventions has only a local and temporary currency; but a residue satisfies a national demand and sticks in the general vocabulary. Wars, politics and the popularization of scientific inventions are the chief ­causes of new words gaining wide currency; often ­these are forgotten again within the year. (One cannot yet tell ­whether the word ‘quisling’ ­will outlast the Nazi technique of preparing for the invasion of a country by po­liti­cal infiltration.) En­glish dictionaries are collections of pre­ce­dents, rather than official code-­books of meaning.

In the United States, intelligent writers sometimes use a far less rigidly grammatical prose than British-­born writers would dare. This is from an article (1941) by Otis Ferguson, about ‘The Spirits of Rhythm’, a negro orchestra:

‘[First it was] . . . Douglas Daniels and Leo Watson, who sang, played and messed around with their feet down any cab rank in St. Louis, to coax pennies ringing on to the sidewalk. . . . ​

They started with ukeleles, ­because at the time of “Yes, We Have No Bananas” ­every kid ­going to school had about a dollar’s worth of ukelele ­under his arm, whack it ­every chance he got. They started innocent of ­music as any kind of
exercise and even to-­day, as Doug says, ­there ­isn’t a one of them could read a note as big as that ­table. But they have an active book of arrangements that can keep them ­going all night; and while it’s all in their heads and something they have got up through playing together, you ­couldn’t start out old and get a book like that for less than a thousand dollars.

Their act is what the rec­ord labels call vocal with inst. accomp., the inst. at pres­ent being a guitar, two tipples (a small-­size guitar, using only the first four strings) and a double-­bass. But that gives you no idea. The truth is, with just the four of them playing along they have more balance, depth and lift than any of the bands they are supposed to “relieve” on the stand. . . . ​

Twenty desks of violins in your local grade-­school symphony may make more noise than the four instruments of the Budapest Quartet without getting the power in ­music, and a half-­acre of Goldman Band can take Nola at around 380 to the minute without getting the sense of motion of one good two-­finger piano player ­doing Lazy. . . . ​

They picked up Teddy Bunn in Washington. Teddy Bunn is the pres­ent guitar player. He is a perfectly terrible man, but has the mostest fun, and seems to be resigned to ­music, though ­there was a time when somehow or other he got hold of an electric guitar and he was playing electric all over the place, and ­those ­were not happy days. He is one of the most brilliant men in the business, and keep your pockets sewed up.’

This is good En­glish in its unconventional way. Mr. Ferguson is taking the same liberties with prose as ‘The Spirits’ took with ­music, just to make his write-up fit the subject—­there are negroish conversational inflexions throughout; for example, the abrupt ‘whack it ­every chance he got’, the emphatic ‘­there ­isn’t a one of them’, and the swift change of tone in ‘and keep your pockets sewed up.’ He admires ‘The Spirits’ as showing musical genius and ‘plays along’ with them without finding their racy speech any more comical than their ­music. Some of ­these inflexions may one day become incorporated in the main vocabulary of En­glish.


Robert Graves (1895–1985) was a preeminent English poet, novelist, critic, translator, and scholar of classical mythology. He served in World War I—an experience recounted in his 1929 autobiography, Good-Bye to All That—and later became the first professor of English literature at the University of Cairo. Best remembered today for his acclaimed historical novels about the Roman emperor Claudius, I, Claudius and Claudius the God, his other books include Count Belisarius, Lawrence and the Arabs, They Hanged My Saintly Billy, Homer's Daughter, The Golden Fleece, The Siege and Fall of Troy, Ann at Highwood Hall, The White GoddessThe Hebrew Myths, and Collected Poems. He is also the co-author of The Reader Over Your Shoulder: A Handbook for Writers of English Prose. 

Patricia T. O'Conner, a former staff editor at The New York Times Book Review, is the author of five books on language, most recently Origins of the Specious, written with her husband, Stewart Kellerman. Her first book, Woe Is I, has half a million copies in print and will soon appear in a fourth edition. She and Mr. Kellerman blog about the English language at