The first attempts at producing a grammar of English were made when there were less than ten million speakers of English in the world, almost all of them living within 100 miles or so of London. Grammars of English have gone on being written during the intervening 400 years reflecting a variety (and growing complexity) of needs, while speakers of English have multiplied several hundredfold and dispersed themselves so that the language has achieved a uniquely wide spread throughout the world and, with that, a unique importance.
We make no apology for adding one more to the succession of English grammars. In the first place, though fairly brief synopses are common enough, there have been very few attempts at so comprehensive a coverage as is offered in the present work. Fewer still in terms of synchronic description. And none at all so comprehensive or in such depth has been produced within an English-speaking country. Moreover, our Grammar aims at this comprehensiveness and depth in treating English irrespective of frontiers: our field is no less than the grammar of educated English current in the second half of the twentieth century in the world's major English-speaking communities. Only where a feature belongs specifically to British usage or American usage, to informal conversation or to the dignity of formal writing, are 'labels' introduced in the description to show that we are no longer discussing the 'common core' of educated English.
For this common core, as well as for the special varieties surrounding it, we have augmented our own experience as speakers and teachers of the language with research on corpora of contemporary English and on data from elicitation tests, in both cases making appropriate use of facilities available in our generation for bringing spoken English fully within the grammarian's scope. For reasons of simplicity and economic presentation, however, illustrative examples from our basic material are seldom given without being adapted and edited; and while informal and familiar styles of speech and writing receive due consideration in our treatment, we put the main emphasis on describing the English of serious exposition.
When work on this Grammar began, the four collaborators were all on the staff of the English Department, University College London, and jointly involved in the Survey of English Usage. This association has happily survived a dispersal which has put considerable distances between us (at the extremes, the 5000 miles between Wisconsin and Europe). Common research goals would thus have kept us in close touch even without a rather large unified undertaking to complete. And though physical separation has made collaboration more arduous and time-consuming, it has also - we console ourselves in retrospect - conferred positive benefits. For example, we have been able to extend our linguistic horizons by contact with linguists bred in several different traditions; and our ideas have been revised and improved by exposure to far more richly varied groups of students than would have been possible in any one centre.
It will be obvious that our grammatical framework has drawn heavily both on the long-established tradition and on the insights of several contemporary schools of linguistics. But while we have taken account of modern linguistic theory to the extent that we think justifiable in a grammar of this kind, we have not felt that this was the occasion for detailed discussion of theoretical issues. Nor do we see need to justify the fact that we subscribe to no specific one of the current or recently formulated linguistic theories. Each of those propounded from the time of de Saussure and Jespersen onwards has its undoubted merits, and several (notably the transformational-generative approaches) have contributed very great stimulus to us as to other grammarians. None, however, seems yet adequate to account for all linguistic phenomena, and recent trends suggest that our own compromise position is a fair reflection of the way in which the major theories are responding to influence from others.
As well as such general debt to our students, our contemporaries, our teachers and our teachers' teachers, there are specific debts to numerous colleagues and friends which we are happy to acknowledge even if we cannot hope to repay. Five linguists generously undertook the heavy burden of reading and criticizing a preliminary draft of the entire book:
Dwight L. Bolinger, Bengt Jacobsson, Ruth M. Kempson, Edward Hirschland and Paul Portland. His many friends who have been fortunate enough to receive comments on even a short research paper will have some idea of how much we have profited from Professor Bolinger's deep learning, keen intellect, incredible facility for producing the devastating counter-example, and - by no means least - readiness to give selflessly of his time. The other four critics had qualities of this same kind and (for example) many of our most telling illustrations come from the invaluable files assembled by Dr Jacobsson over many years of meticulous scholarship.
Colleagues working on the Survey of English Usage have of course been repeatedly involved in giving advice and criticism; we are glad to take this opportunity of expressing our thanks to Valeric Adams and Derek Davy, Judith Ferryman, Florent Aarts and Michael Black, as also to Cindy Kapsos and Pamela Miller. For help with specific parts, we are grateful to Ross Almqvist and Ulla Thagg (Chapters 3,4 and 12), Jacquelyn Biel (especially Chapters 5 and 8), Peter Fries (Chapter 9), A. C. Gimson (Appendix II) and Michael Riddle (Appendix III). The research and writing have been supported in part by grants from HM Department of Education and Science, the Leverhulme Trust, the Calouste Gulbenkian Foundation, the Longman Group, the Graduate School Research Committee of the University of Wisconsin-Milwaukee, the University of Göteborg, the University of Lund and University College London.
For what Fredson Bowers has called ‘authorial fair copy expressing final intention’, the publisher received from us something more resembling the manuscript of Killigrew's Conspiracy in 1638: a ‘Foul Draught’ full of ‘Corrections, Expungings, and Additions’. We owe it largely to Peggy Drinkwater's unswerving concentration that this has been transformed into orderly print.
RQ SO GL JS
- Preface v
- Symbols and technical conventions xi
- The English language 1
- The sentence: a preliminary view 33
- The verb phrase 61
- Nouns, pronouns, and the basic noun phrase 123
- Adjectives and adverbs 229
- Prepositions and prepositional phrases 297
- The simple sentence 339
- Adjuncts, disjuncts, conjuncts 417
- Coordination and apposition 533
- Sentence connection 649
- The complex sentence 717
- The verb and its complementation 799
- The complex noun phrase 855
- Focus, theme, and emphasis 935
- Word-formation 973
- Stress, rhythm, and intonation 1033
- Index 1093
THE ENGLISH LANGUAGE
1.1-7 The importance of English
- .1-2 Criteria of 'importance'
- .3-4 Native, second, and foreign language
- .5-7 The demand for English
- .5 The teaching of English
- .6 A lingua franca in science and scholarship
- .7 International character of English
1.8-14 Grammar and the study of language .9-9 Types of linguistic organization
- .8 Sounds and spellings
- .9 Lexicology, semantics, grammar .10-14 The meanings of 'grammar'
- .10 Syntax and inflections
- .11 Rules and the native speaker
- .12 The codification of rules
- .13 Grammar and other types of organization
- .14 Grammar and generalization
1.15-37 Varieties of English and classes of varieties .
- 16-17 Regional variation .
- 18 Education and social standing .
- 19 Standard English .
- 20-22 National standards of English
- .20 British and American English
- .21 Scotland, Ireland, Canada
- .22 South Africa, Australia, New Zealand
- 23 Pronunciation and Standard English
- 24 Varieties according to subject matter
- 25-26 Varieties according to medium
- 27-29 Varieties according to attitude
- 30-32 Varieties according to interference
- .32 Creole and Pidgin
- .33-35 Relationship between variety classes
- 36-37 Varieties within a variety
The importance of English
Criteria of ‘importance’
English is the world's most important language. Even at a time when such a statement is taken as a long-standing truism, it is perhaps worthwhile to glance briefly at the basis on which it is made. There are, after all, thousands of different languages in the world, and it is in the nature of language that each one seems uniquely important to those who speak it as their native language - that is, their first (normally sole) tongue: the language they acquired at their mother's knee. But there are more objective standards of relative importance.
One criterion is the number of native speakers that a language happens to have. A second is the extent to which a language is geographically dispersed: in how many continents and countries is it used or is a knowledge of it necessary? A third is its ‘vehicular load’: to what extent is it a medium for a science or literature or other highly regarded cultural manifestation - including ‘way of life’? A fourth is the economic and political influence of those who speak it as ‘heir own’ language.
None of these is trivial but not all would unambiguously identify English. Indeed the first would make English a very poor second to Chinese (which has double the number of speakers) and would put English not appreciably in front of Hindi-Urdu. The second clearly makes English a front runner but also invites consideration of Hebrew, Latin and Arabic, for example, as languages used in major world religions, though only the last mentioned would be thought of in connection with the first criterion. By the third criterion, the great literatures of the Orient spring to mind, not to mention the languages of Tolstoy, Goethe, Cervantes and Racine. But in addition to being the language of the analogous Shakespeare, English scores as being the primary medium for twentieth-century science and technology. The fourth criterion invokes Japanese, Russian and German, for example, as languages of powerful, productive and influential communities. But English is the language of the United States which - to take one crude but objective measure - has a larger 'Gross National Product' (both in total and in relation to the population) than any other country in the world. Indeed the combined GNP of the USA, Canada and Britain is 50 per cent higher than that of the remaining OECD countries (broadly speaking, continental Europe plus Japan) put together: cf Organization for Economic Co-operation and Development, Main Economic Indicators, June 1971.
What emerges strikingly about English is that by any of the criteria it is prominent, by some it is pre-eminent, and by a combination of the four it is superlatively outstanding. Notice that no claim has been made for the importance of English on the grounds of its ‘quality’ as a language (the size of its vocabulary, the alleged flexibility of its syntax). It has been rightly said that the choice of an international language, or lingua franca, is never based on linguistic or aesthetic criteria but always on political, economic and demographic ones.
Native, second, and foreign language
English is the world's most widely used language. It is useful to distinguish three primary categories of use: as a native language, as a second language, and as a foreign language. English is spoken as a native language by nearly three hundred million people: in the United States, Britain, Ireland, Australia, New Zealand, Canada, the Caribbean and South Africa, without mentioning smaller countries or smaller pockets of native English speakers (for example in Rhodesia and Kenya). In several of these countries, English is not the sole language: the Quebec province of Canada is French-speaking, much of South Africa is Afrikaans-speaking, and for many Irish and Welsh people, English is not the native language. But for these Welsh, Irish, Quebecois and Afrikaners, English will even so be a second language: that is, a language necessary for certain official, social, commercial or educational activities within their own country. This second-language function is more noteworthy, however, in a long list of countries where only a small proportion of the people have English as their native language: India, Pakistan, Nigeria, Kenya and many other Commonwealth countries and former British territories. Thus, a quarter of a century after independence, India maintains English as the medium of instruction for approximately half of its total higher education. English is the second language in countries of such divergent backgrounds as the Philippines and Ethiopia, while in numerous other countries (Burma, Thailand, South Korea and some Middle Eastern countries, for example) it has a second language status in respect of higher education. It is one of the two ‘working’ languages of the United Nations and of the two it is by far the more frequently used both in debate and in general conduct of UN business.
By foreign language we mean a language as used by someone for communication across frontiers or with people who are not his countrymen: listening to broadcasts, reading books or newspapers, commerce or travel, for example. No language is more widely studied or used as a foreign language than English. The desire to learn it is immense and apparently insatiable. American organizations such as the United States Information Agency and the Voice of America have played a notable role in recent years, in close and amicable liaison with the British Council which provides support for English teaching both in the Commonwealth and in foreign countries throughout the world. The BBC, like the USIS, has notable radio and television facilities devoted to this purpose. Other English-speaking countries such as Australia also assume heavy responsibilities for teaching English as a foreign language. Taking the education systems of the world as a whole, one may say confidently (if perhaps ruefully) that more timetable hours are devoted to English than any other subject.
We shall look more closely in the next section at the kind and degree of demand, but meantime the reasons for the demand have surely become clear. To put it bluntly, English is a top requirement of those seeking good jobs - and is often the language in which much of the business of ‘good jobs’ is conducted. One needs it for access to at least one half of the world’s scientific literature. It is thus intimately associated with technological and economic development and it is the principal language of international aid. Not only is it the universal language of international aviation, shipping and sport: it is to a considerable degree the universal language of literacy and public communication. Siegfried Muller (former Director of the Languages-of-the-World Archives in the US Department of Education) has estimated that about 60 per cent of the world’s radio broadcasts and 70 per cent of the world’s mail are in English. The great manufacturing countries Germany and Japan use English as their principal advertising and sales medium; it is the language of automation and computer technology.
The demand for English
The teaching of English
The role of chief foreign language that French occupied for two centuries from about 1700, therefore, has been undoubtedly assumed by English - except of course in the English-speaking countries themselves, where French is challenged only by Spanish as the foreign language most widely studied. Although patriotism obliges international organizations to devote far more resources to translation and interpreter services than reason would dictate, no senior post would be offered to a candidate deficient in English. The equivalent of the nineteenth-century European ‘finishing school’ in French now provides a liberal education in English, whether located in Sussex or in Switzerland. But a more general equivalent is perhaps the English-medium school organized through the state education system, and such institutions seem to be even more numerous in the Soviet Union and other east European countries than in countries to the west. More general still, of course, is the language work in the ordinary schools, and in this connection the introduction at the primary (pre-lycée, pre-Gymnasium) level of foreign language teaching has meant a sharp but almost accidental increase in English teaching and in the demand for English teachers. That is, if a foreign language is to be taught at the primary level, what other language should the French or German schools teach but English? And if children already have some English before entering secondary education, what more obvious than to continue with this particular foreign language, making any other language at secondary level a lower priority option, learned to a less adequate degree?
To take France as an example, in the academic year 1968-69, English was being learned as first foreign language by 80 per cent of secondary school pupils, the nearest rival being German with 16 per cent. When we include those who study it as their second foreign language, we have a total of over two million teenagers studying English in France, a country with a tradition for teaching several other European languages - Spanish in the south-west, Italian in the south-east and German in the northeast.
A lingua franca in science and scholarship
We might refer also to an inquiry recently made into the use of foreign languages by the learned community in French-speaking territories. It transpired that 90 per cent found it necessary to use books in English - and this percentage included scholars whose research lay in the field of French literature. Perhaps even more significant: about 25 per cent preferred to publish their scholarly and scientific papers in English. The latter point is strikingly paralleled in Italy and Germany. About 1950, the Italian physics journal Nuovo Cimento decided to admit papers in languages other than Italian: in less than 20 years the proportion of papers published in Italian fell from 100 per cent to zero and the proportion of papers published in English rose from zero to 100 per cent. A German example: between 1962 and 1968 alone the proportion of articles published in English in Physikalische Zeitschrift rose from 2 per cent to 50 per cent. In both these cases, the change may in part be due to the editors' acceptance of papers by American, British and other English-speaking physicists, but for the most part one would surely be right in thinking that it reflects the European scientists' desire to share their research most efficiently with their colleagues all over the world by means of the twentieth-century lingua franca. Telling evidence of this is provided by the European journal Astronomy and Astrophysics in which two-thirds of the contributions by French scientists are in English, and by the official publication of the Agence Internationale de l’Énergie Atomique, Nuclear Fusion, where all articles are in English, despite the fact that the Agency is subsidized by the French Government.
International character of English
For the foregoing observations, we have deliberately drawn heavily on the work of an outstandingly qualified Frenchman, Denis Girard, Inspecteur Regional de l'Academie de Paris, in order to insure ourselves against the danger of overstating the importance of English, and to assure ourselves of seeing English measured in terms of international values. Not that one is tempted to do otherwise. English, which we have referred to as a lingua franca, is pre-eminently the most international of languages. Though the mention of the language may at once remind us of England, on the one hand, or cause association with the might of the United States on the other, it carries less implication of political or cultural specificity than any other living tongue (with French and Spanish also notable in this respect). At one and the same time, it serves the daily purposes of republics such as the United States and South Africa, sharply different in size, population, climate, economy and national philosophy; and it serves an ancient kingdom such as Britain, as well as her widely scattered Commonwealth partners, themselves as different from each other as they are from Britain herself.
But the cultural neutrality of English must not be pressed too far. The literal or metaphorical use of such expressions as case law throughout the English-speaking world reflects a common heritage in our legal system; and allusions to or quotations from Shakespeare, the Authorized Version, Gray's Elegy, Mark Twain, a sea shanty, a Negro spiritual or a Beatles song - wittingly or not - testify similarly to a shared culture. The Continent means ‘continental Europe’ as readily in America and even Australia and New Zealand as it does in Britain. At other times, English equally reflects the independent and distinct culture of one or other of the English-speaking communities. When an Australian speaks of fossicking something out (searching for something), the metaphor looks back to the desperate activity of reworking the diggings of someone else in the hope of finding gold that had been overlooked. When an American speaks of not getting to first base (not achieving even initial success), the metaphor concerns an equally culture-specific activity-the game of baseball. And when an Englishman says that something is not cricket (unfair), the allusion is also to a game that is by no means universal in the English-speaking countries,
Grammar and the study of language
Types of linguistic organization
Sounds and spellings
The claim is, therefore, that on the one hand there is a single ‘English language’ (the grammar of which is the concern of this book), but that on the other there are recognizable varieties. Since these varieties can have reflexes in any of the types of organization that the linguist distinguishes, this is the point at which we should outline these types of organization (or ‘levels’ as they are sometimes called), one of which is ‘grammar’. When someone communicates with us by means of language, he normally does so by causing us to hear a stream of sounds. We hear the sounds not as indefinitely variable in acoustic quality (however much they may be so in actual physical fact). Rather, we hear them as each corresponding to one of a very small set (in English, /p/, /l/, /n/, /i/, /ð/, /s/...) which can combine in certain ways and not others. For example, in English we have spin but not *psin, our use of the asterisk here and elsewhere in this book denoting non-occurring or unacceptable forms. We similarly observe patterns of stress and pitch. The sounds made in a particular language and the rules for their organization are studied in the branch of linguistics known as phonology, while their physical properties and their manner of articulation are studied in phonetics.
Another major method of linguistic communication is by visual signs, that is, writing; and for English as for many other languages there has been developed an alphabetic writing system with symbols basically related to the individual sounds used in the language. Here again there is a closely structured organization which regards certain differences in shape as irrelevant and others (for example capitals versus lower case, ascenders to the left or right of a circle - b versus d) as significant. The study of graphology or orthography thus parallels the study of phonology in several obvious ways. Despite the notorious oddities of English spelling, there are important general principles: eg combinations of letters that English permits (tch, qu, ss, oo) and others that are disallowed (*pfx, *qi, *yy) or have only restricted distribution (filial v or j occurs only exceptionally as in Raj, spiv).
Lexicology, semantics, grammar
Just as the small set of arabic numerals can be combined to express in writing any natural numbers we like, however vast, so the small set of sounds and letters can be combined to express in speech or writing respectively an indefinitely large number of words. These linguistic units enable people to refer to every object, action and quality that members of a society wish to distinguish: in English, door, soap, indignation, find, stupefy, good, uncontrollable, and so on to a total in the region of at least half a million. These units of language have a meaning and a structure (sometimes an obviously composite structure as in cases like uncontrollable) which relate them not only to the world outside language but to other words within the language (good, bad, kind, etc). The study of words is the business of lexicology but the regularities in their formation are similar in kind to the regularities of grammar and are closely connected to them (cf App I.I ff). Meaning relations as a whole are the business of semantics, the study of meaning, and this therefore has relevance equally within lexicology and within grammar.
There is one further type of organization. The words that have been identified by sound or spelling must be combined into larger units and it is the complex set of rules specifying such combination that we refer to as grammar. This word has various common meanings in English (as in other languages: cf: grammaire, Grammatik) and since it is the subject matter of this book some of its chief meanings should be explored.
The meanings of ‘grammar’
Syntax and inflections
We shall be using ‘grammar’ to include both syntax and the inflections (or accidence) of morphology. The fact that the past tense of buy is bought (inflection) and the fact that the interrogative form of He bought it is Did he buy it? (syntax) are therefore both equally the province of grammar. There is nothing esoteric or technical about our usage in this respect: it corresponds to one of the common lay uses of the word in the English-speaking world. A teacher may comment
John uses good grammar but his spelling is awful
showing that spelling is excluded from grammar; and if John wrote interloper where the context demanded interpreter, the teacher would say that he had used the wrong word, not that he had made a mistake in grammar. So far so good. But in the education systems of the English-speaking countries, it is possible also to use the term ‘grammar’ loosely so as to include both spelling and lexicology, and we need to be on our guard so that we recognize when the word is used in so sharply different a way. A ‘grammar lesson’ for children may in fact be concerned with any aspect of the use, history, spelling or even pronunciation of words.
When grammar is prefixed to school (as it is in several English-speaking countries, though not always with reference to the same type of school), the term reflects the historical fact that certain schools concentrated at one time upon the teaching of Latin and Greek. This is the ‘grammar’ in their name. No serious ambiguity arises from this, though one sometimes comes upon the lay supposition that such schools do or should make a special effort to teach English grammar. But there is a further use of ‘grammar’ which springs indirectly from this educational tradition. It makes sense for the lay native speaker to say
Latin has a good deal of grammar, but English has hardly any
since the aspect of Latin grammar on which we have traditionally concentrated is the paradigms (model sets) of inflections. This in effect meant that grammar became identified with inflections or accidence, so that we can still speak of ‘grammar and syntax’ in this connection, tacitly excluding the latter from the former. And since all of the uses of ‘grammar’ so far illustrated might appear in the speech or writing of the same person, the possibilities of misunderstanding are very real.
Rules and the native speaker
Nor have we completed the inventory of meanings. The same native speaker, turning his attention from Latin, may comment:
French has a well-defined grammar, but in English we're free to speak as we like
Several points need to be made here. To begin with, it is clear that the speaker cannot now be intending to restrict ‘grammar’ to inflections: rather the converse; it would seem to be used as a virtual synonym of ‘syntax’.
Secondly, the native speaker's comment probably owes a good deal to the fact that he does not feel the rules of his own language - rules that he has acquired unconsciously - to be at all constraining; and if ever he happens to be called on to explain one such rule to a foreigner he has very great difficulty. By contrast, the grammatical rules he learns for a foreign language seem much more rigid and they also seem clearer because they have been actually spelled out to him in the learning process.
But another important point is revealed in this sentence. The distinction refers to grammar not as the observed patterns in the use of French but to a codification of rules compiled by the French to show the French themselves how their language should be used. This is not grammar ‘immanent’ in a language (as our previous uses were, however much they differed in the types of pattern they referred to), but grammar as codified by grammarians: the Academy Grammar. There is no such Academy for the English language and so (our naive native speaker imagines) the English speaker has more ‘freedom’ in his usage.
The codification of rules
The ‘codification’ sense of grammar is readily identified with the specific codification by a specific grammarian:
Jespersen wrote a good grammar, and so did Kruisinga
and this sense naturally leads to the concrete use as in
Did you bring your grammars?
and naturally, too, the codification may refer to grammar in any of the senses already mentioned. A French grammar will be devoted very largely to syntax, while accidents of intellectual history in the nineteenth century lead one to treat without surprise the fact that an Old High German grammar (or an Old English grammar) may well contain only
inflections together with a detailed explanation of how the phonological system emerged.
The codification will also vary, however, according to the linguistic theory embraced by the author, his idea of the nature of grammar per se rather than his statement of the grammar of a particular language:
Shaumjan has devised a grammar interestingly different from Chomsky's
It is important to realize that, in the usage of many leading linguists, this last sense of grammar has returned to the catholicity that it had in the Greek tradition more than two thousand years ago, covering the whole field of language structure. Thus, in the framework of formal linguistics, contemporary generative grammarians will speak of ‘the grammar’ as embracing rules not only for syntax but for phonological, lexical and semantic specification as well.
Grammar and other types of organization
Progress towards a more explicit type of grammatical description is inevitably slow and the whole field of grammar is likely to remain an area of interesting controversy. While theoretical problems are not the concern of this book, our treatment cannot be neutral on the issues that enliven current discussion. For example, we would not wish to assert the total independence of grammar from phonology on the one hand and lexico-semantics on the other as was implied in the deliberate oversimplification of 1.8f. Phonology is seen to have a bearing on grammar even in small points such as the association of initial /ð/ with demon-strativeness and conjunction (this, then, though, etc: 2.13). It is seen to bear on lexicology, for example, in the fact that numerous nouns and verbs differ only in the position of a stress (App 1.43, App 11.5):
That is an 'insult
They may in'sult me
But most obviously the interdependence of phonology and grammar is shown in focus processes (cf the connection between intonation and linear presentation: 14.2-7), and in the fact that by merely altering the phonology one can distinguish sets of sentences like those quoted in App 11.20.
The interrelations of grammar, lexicology and semantics are still more pervasive. To take an obvious example, the set of sentences
John hated the shed
John painted the shed
Fear replaced indecision
have a great deal in common that must be described in terms of grammar. They have the same tense (past), the same structure (subject plus verb plus object), will permit the same syntactic operations as in
The shed was painted by John
Did John paint the shed?
It was John that painted the shed
Up to a point they will also permit the permutation of their parts so that the abstraction ‘subject - verb – object’ appears to be an adequate analysis:
John replaced the shed
John hated indecision
But by no means all permutations are possible:
*Fear painted the shed
*Fear hated indecision
*John replaced indecision
To what extent should the constraints disallowing such sentences be accounted for in the grammatical description? Questions of this kind will remain intensely controversial for a long time, and little guidance on the problems involved can be given in this book (cf however 7.37-38).
Grammar and generalization
Our general principle will be to regard grammar as accounting for constructions where greatest generalization is possible, assigning to lexicology constructions on which least generalization can be formulated (which approach, that is, the idiosyncratic and idiomatic). The gradient of ‘greatest’ to ‘least’ in the previous sentence admits at once the unfortunate necessity for arbitrary decision. Confronted with the correspondences:
He spoke these words ↔ The speaker of these words
He wrote these words ↔ The writer of these words
we will wish to describe within grammar the way in which items in the first column can be transformed into the shape given them in the second. But this will leave us with second column items such as
Ø ↔ The author of these words
for which there is no first-column ‘source’. This particular example, we may agree, raises no semantic problem: there is merely a lexicological gap in the language - no verb *auth. But we have also first-column items for which there is no second-column transform:
He watched the play ↔ Ø
Here we cannot account for the constraint in terms of a lexical gap, but we may be very uncertain as to whether it is a problem for lexicology or grammar (cf App 1.24). One further example:
He spattered the wall with oil
He smeared the wall with oil
He rubbed the wall with oil
He dirtied the wall with oil
*He poured the wall with oil
It is not easy to decide whether we should try to account within grammar for the imbalance in relating items from such a set to alternative predication forms (12.62f):
He spattered oil on the wall
He smeared oil on the wall
He rubbed oil on the wall
*He dirtied oil on the wall
He poured oil on the wall
The question is not merely how minimally general must a rule be before it ceases to be worth presenting within grammar but one of much deeper theoretical concern: what, if anything, ultimately distinguishes a rule of grammar from a rule of semantics? Provided that we can remember at all times that such questions remain matters for debate, no harm is done by offering - as we do in this book - some provisional answers.
Varieties of English and classes of varieties
Having established, subject to these important qualifications, the extent to which we may speak of different types of linguistic organization such as phonology, lexicology and grammar, we may now return to the point we had reached at the beginning of 1.8. What are the varieties of English whose differing properties are realized through the several types of linguistic organization?
A great deal has been written in recent years attempting to provide a theoretical basis on which the varieties of any language can be described, interrelated and studied: it is one of the prime concerns of the relatively new branch of language study called sociolinguistics. The problem is formidable, we are far from having complete answers, and all attempts are in some degree an oversimplification. It may help now to consider one such oversimplification for the purposes of this book. First, an analogy. The properties of dog-ness can be seen in both terrier and alsatian (and, we must presume, equally), yet no single variety of dog embodies all the features present in all varieties of dog. In a somewhat similar way, we need to see a common core or nucleus that we call ‘English’ being realized only in the different actual varieties of the language that we hear or read. Let us imagine six kinds of varieties ranged as below and interrelated in ways we shall attempt to explain.
The fact that in this figure the ‘common core' dominates all the varieties means that, however esoteric or remote a variety may be, it has running through it a set of grammatical and other characteristics that are present in all others. It is presumably this fact that justifies the application of the name ‘English’ to all the varieties. From this initial point onwards, it will be noted that nothing resembling a noded tree structure is suggested: instead, it is claimed by the sets of braces that each variety class is related equally and at all points to each of the other variety classes. We shall however return and make qualifications to this claim. The classes themselves are arranged in a meaningful order and the justification will become clear in what follows.
Varieties according to region have a well-established label both in popular and technical use: ‘dialects’. Geographical dispersion is in fact the classic basis for linguistic variation and in the course of time, with poor communications and relative remoteness, such dispersion results in dialects becoming so distinct that we regard them as different languages. This latter stage was long ago reached with the Germanic dialects that are now Dutch, English, German, Swedish, etc, but it has not been reached (and may not necessarily ever be reached, given the modem ease of communication) with the dialects of English that have resulted from the regional separation of communities within the British Isles and (since the voyages of exploration and settlement in Shakespeare's time) elsewhere in the world.
Regional variation seems to be realized predominantly in phonology. That is, we generally recognize a different dialect from a speaker's pronunciation or accent before we notice that his vocabulary (or lexicon) is also distinctive. Grammatical variation tends to be less extensive and certainly less obtrusive. But all types of linguistic organization can readily enough be involved. A Lancashire man may be recognized by a Yorkshire man because he pronounces an /r/ after vowels as in stir or hurt. A middy is an Australian measure for beer - but it refers to a considerably bigger measure in Sydney than it does in Perth. Instead of I saw it, a New Englander might say I see it, a Pennsylvanian I seen it and a Virginian either I seen it or I seed it, if they were speaking the natural dialect of their locality, and the same forms distinguish certain dialects within Britain too. 
It is pointless to ask how many dialects of English there are: there are indefinitely many, depending solely on how detailed we wish to be in our observations. But they are of course more obviously numerous in the long-settled Britain than in the more recently settled North America or in the still more recently settled Australia and New Zealand. The degree of generality in our observation depends crucially upon our standpoint as well as upon our experience. An Englishman will hear an American Southerner primarily as an American and only as a Southerner in addition if further subclassification is called for and if his experience of American English dialects enables him to make it. To an American the same speaker will be heard first as a Southerner and then (subject to similar conditions) as, say, a Virginian, and then perhaps as a Piedmont Virginian. One might suggest some broad dialectal divisions which are rather generally recognized. Within North America, most people would be able to distinguish Canadian, New England, Midland, and Southern varieties of English. Within the British Isles, Irish, Scots, Northern, Midland, Welsh, South-western, and London varieties would be recognized with similar generality. Some of these - Irish and Scots for example - would be recognized as such by many Americans and Australians too, while in Britain many people could make subdivisions: Ulster and Southern might be distinguished within Irish, for example, and Yorkshire picked out as an important subdivision of northern speech. British people can also, of course, distinguish North Americans from all others (though not usually Canadians from Americans), South Africans from Australians and New Zealanders (though mistakes are frequent), but not usually Australians from New Zealanders.
Education and social standing
Within each of the dialect areas, there is considerable variation in speech according to education and social standing. There is an important polarity of uneducated and educated speech in which the former can be identified with the regional dialect most completely and the latter moves away from dialectal usage to a form of English that cuts across dialectal boundaries. To revert to an example given in a previous section, one would have to look rather hard (or be a skilled dialectologist) to find, as an outsider, a New Englander who said see for saw, a Pennsylvanian who said seen, and a Virginian who said seed. These are forms that tend to be replaced by saw with schooling, and in speaking to a stranger a dialect speaker would tend to use ‘school’ forms. On the other hand, there is no simple equation of dialectal and uneducated English. Just as educated English (I saw) cuts across dialectal boundaries, so do many features of uneducated use: a prominent example is the double negative as in I don't want no cake which has been outlawed from all educated English by the prescriptive grammar tradition for hundreds of years but which continues to thrive in uneducated speech wherever English is spoken.
Educated speech - by definition the language of education - naturally tends to be given the additional prestige of government agencies, the learned professions, the political parties, the press, the law court and the pulpit - any institution which must attempt to address itself to a public beyond the smallest dialectal community. The general acceptance of ‘BBC English’ for this purpose over almost half a century is paralleled by a similar designation for general educated idiom in the United States, ‘network English’. By reason of the fact that educated English is thus accorded implicit social and political sanction, it comes to be referred to as Standard English, and provided we remember that this does not mean an English that has been formally standardized by official action, as weights and measures are standardized, the term is useful and appropriate. In contrast with Standard English, forms that are especially associated with uneducated (rather than dialectal) use are often called ‘substandard’.
The degree of acceptance of a single standard of English throughout the world, across a multiplicity of political and social systems, is a truly remarkable phenomenon: the more so since the extent of the uniformity involved has, if anything, increased in the present century. Uniformity is greatest in what is from most viewpoints the least important type of linguistic organization - the purely secondary one of orthography. Although printing houses in all English-speaking countries retain a tiny element of individual decision (realize, -ise; judg(e)ment; etc), there is basically a single, graphological spelling and punctuation system throughout: with two minor subsystems. The one is the subsystem with British orientation (used in all English-speaking countries except the United States) with distinctive forms in only a small class of words, colour, centre, levelled, etc. The other is the American subsystem: color, center, leveled, etc. In Canada, the British subsystem is used for the most part, but some publishers (especially of popular material) follow the American subsystem and some a mixture (color but centre). In the American Mid-West, some newspaper publishers (but not book publishers) use a few additional separate spellings such as thru for through. One minor orthographic point is oddly capable of Anglo-American misunderstanding: the numerical form of dates. In British (and European) practice ‘7/11/72’ would mean ‘7 November 1972’, but in American practice it would mean ‘July 11 1972’.
In grammar and vocabulary, Standard English presents somewhat less of a monolithic character, but even so the world-wide agreement is extraordinary and - as has been suggested earlier - seems actually to be increasing under the impact of closer world communication and the spread of identical material and non-material culture. The uniformity is especially close in neutral or formal styles (1.27) of written English (1.25) on subject matter (1.24) not of obviously localized interest: in such circumstances one can frequently go on for page after page without encountering a feature which would identify the English as belonging to one of the national standards.
National standards of English
British and American English
What we are calling national standards should be seen as distinct from the Standard English which we have been discussing and which we should think of as being ‘supra-national’, embracing what is common to all. Again, as with orthography, there are two national standards that are overwhelmingly predominant both in the number of distinctive usages and in the degree to which these distinctions are ‘institutionalized’: American English and British English. Grammatical differences are few and the most conspicuous are widely known to speakers of both national standards; the fact that AmE has two past participles for get and BrE only one (3.68), for example, and that in BrE the indefinite pronoun one is repeated in co-reference where AmE uses he (4.126) as in
One cannot succeed at this unless tries hard
Lexical examples are far more numerous, but many of these are also familiar to users of both standards: for example, railway (BrE), railroad (AmE); tap (BrE), faucet (AmE); autumn (BrE), fall (AmE). More recent lexical innovations in either area tend to spread rapidly to the other. Thus while radio sets have had values in BrE but tubes in AmE, television sets have tubes in both, and transistors are likewise used in both standards.
The United States and Britain have been separate political entities for two centuries; for generations, thousands of books have been appearing annually; there is a long tradition of publishing descriptions of both AmE and BrE. These are important factors in establishing and institutionalizing the two national standards, and in the relative absence of such conditions other national standards are both less distinct (being more open to the influence of either AmE or BrE) and less institutionalized.
Scotland, Ireland, Canada
Scots, with ancient national and educational institutions, is perhaps nearest to the self-confident independence of BrE and AmE, though the differences in grammar and vocabulary are rather few. There is the preposition outwith ‘except’ and some other grammatical features, and such lexical items as advocate in the sense ‘practising lawyer’ or bailie ‘municipal magistrate’ and several others which, like this, refer to Scottish affairs. Orthography is identical with BrE though burgh corresponds closely to ‘borough’ in meaning and might almost be regarded as a spelling variant. But this refers only to official Scots usage. In the ‘Lallans’ Scots, which has some currency for literary purposes, we have a highly independent set of lexical, grammatical, phonological and orthographical conventions, all of which make it seem more like a separate language than a regional dialect.
Irish (or Hiberno-) English should also be regarded as a national standard for though we lack descriptions of this long-standing variety of English it is consciously and explicitly regarded as independent of BrE by educational and broadcasting services. The proximity of Britain, the easy movement of population, and like factors mean however that there is little room for the assertion and development of separate grammar and vocabulary. In fact it is probable that the influence of BrE (and even AmE) is so great on both Scots and Irish English that independent features will diminish rather than increase with time.
Canadian English is in a similar position in relation to AmE. Close economic, social and intellectual links along a 4000-mile frontier have naturally caused the larger community to have an enormous influence on the smaller, not least in language. Though in many respects (zed instead of zee, for example, as the name of the letter ‘z’), Canadian English follows British rather than United States practice, and has a modest area of independent lexical use (pogey ‘welfare payment’, riding ‘parliamentary constituency’, muskeg ‘kind of bog’), in many other respects it has approximated to AmE, and in the absence of strong institutionalizing forces it seems likely to continue in this direction.
South Africa, Australia, New Zealand
South Africa, Australia and New Zealand are in a very different position, remote from the direct day-to-day impact of either BrE or AmE. While in orthography and grammar the South African English in educated use is virtually identical with BrE, rather considerable differences in vocabulary have developed, largely under the influence of the other official language of the country, Afrikaans. For example, ‘veld’ open country', koppie ‘hillock’, dorp ‘village’, konfyt ‘candied peel’. Because of the remoteness from Britain or America, few of these words have spread: an exception is trek ‘journey’.
New Zealand English is more like BrE than any other non-European variety, though it has adopted quite a number of words from the indigenous Maoris (for example, whare ‘hut’ and of course kiwi and other names for fauna and flora) and over the past half century has come under the powerful influence of Australia and to a considerable extent of the United States.
Australian English is undoubtedly the dominant form of English in the Antipodes and by reason of Australia’s increased wealth, population and influence in world affairs, this national standard (though still by no means fully institutionalized) is exerting an influence in the northern hemisphere, particularly in Britain. Much of what is distinctive in Australian English is confined to familiar use. This is especially so of grammatical features like adverbial but or the use of the feminine pronoun both anaphorically for an inanimate noun (job... her) and also impersonally and non-referentially for ‘things in general’:
The job's still not done; I'll finish her this arvo, but.
(... it this afternoon, however.) '
Are you feeling better?' 'Too right, mate; she'll bcjake.'
('... Absolutely, old man; everything will be fine.')
But there are many lexical items that are to be regarded as fully standard : not merely the special fauna and flora (kangaroo, gumtree, wattle, etc) but special Australian uses of familiar words (paddock as a general word for ‘field’, crook ‘ill’, etc), and special Australian words (bowyang ‘a trouser strap’, waddy ‘a bludgeon’, etc).
Pronunciation and standard English
This list does not exhaust the regional or national variants that approximate to the status of a standard (the Caribbean might be mentioned, for example), but the important point to stress is that all of them are remarkable primarily in the tiny extent to which even the most firmly established, BrE and AmE, differ from each other in vocabulary, grammar and orthography. We have been careful, however, not to mention pronunciation in this connection. Pronunciation is a special case for several reasons. In the first place, it is the type of linguistic organization (1.8) which distinguishes one national standard from another most immediately and completely and which links in a most obvious way the national standards to the regional varieties. Secondly (with an important exception to be noted), it is the least institutionalized aspect of Standard English, in the sense that, provided our grammar and lexical items conform to the appropriate national standard, it matters less that our pronunciation follows closely our individual regional pattern. This is doubtless because pronunciation is essentially gradient, a matter of ‘more or less’ rather than the discrete ‘this or that’' features of grammar and lexicon. Thirdly, norms of pronunciation are subject less to educational and national constraints than to social ones: this means, in effect, that some regional accents are less acceptable for ‘network use’ than others; cf 1.16 Note.
Connected with this is the exception referred to above. In BrE, one type of pronunciation comes close to enjoying the status of ‘standard’: it is the accent associated with the English public schools, ‘Received Pronunciation’ or ‘RP’. Because this has traditionally been transmitted through a private education system based upon boarding schools insulated from the locality in which they happen to be situated, it is importantly non-regional, and this - together with the obvious prestige that the social importance of its speakers has conferred on it - has been one of its strengths as a lingua franca. But RP no longer has the unique authority it had in the first half of the twentieth century. It is now only one of the accents commonly used on the BBC and takes its place along with others which carry the unmistakable mark of regional origin - not least, an Australian or North American or Caribbean origin. Thus the rule that a specific type of pronunciation is relatively unimportant seems to be in the process of losing the notable exception that RP has constituted. 
Varieties according to subject matter
Varieties according to the subject matter involved in a discourse have attracted linguists’ attention a good deal in recent years. They are sometimes referred to as ‘registers’, though this term is applied to different types of linguistic variety by different linguists. The theoretical bases for considering subject-matter varieties are highly debatable, but certain broad truths are clear enough. While one does not exclude the possibility that a given speaker may choose to speak in a national standard at one moment and in a regional dialect the next - and possibly even switch from one national standard to another - the presumption has been that an individual adopts one of the varieties so far discussed as his permanent form of English. With varieties according to subject matter, on the other hand, the presumption is rather that the same speaker has a repertoire of varieties and habitually switches to the appropriate one as occasion arises. Naturally, however, no speaker has a very large repertoire, and the number of varieties he commands depends crucially upon his specific profession, training, range of hobbies, etc.
Most typically, perhaps, the switch involves nothing more than turning to the particular set of lexical items habitually used for handling the topic in question. Thus, in connection with repairing a machine: nut, bolt, wrench, thread, lever, finger-tight, balance, adjust, bearing, axle, pinion, split-pin, and the like. ‘I am of course using thread in the engineering sense, not as it is used in needlework’, one says. But there are grammatical correlates to subject-matter variety as well. To take a simple example, the imperatives in cooking recipes: ‘Pour the yolks into a bowl’, not ‘You should’ or ‘You must’ or ‘You might care to’, still less ‘The cook should ...’ More complex grammatical correlates are to be found in the language of technical and scientific description: the passive is common and clauses are often nominalized (13.34 f); thus not usually
You can rectify this fault if you insert a wedge...
Rectification of this fault is achieved by insertion of a wedge...
More radical grammatical changes are made in the language of legal documents:
Provided that such payment as aforesaid shall be a condition precedent to the exercise of the option herein specified...
and the language of prayer:
Eternal God, Who dost call all men into unity with Thy Son ...
It need hardly be emphasized that the type of language required by choice of subject matter would be roughly constant against the variables (dialect, national standard) already discussed. Some obvious contingent constraints are however emerging: the use of a specific variety of one class frequently presupposes the use of a specific variety of another. The use of a well-formed legal sentence, for example, presupposes an educated
variety of English.  -
Varieties according to medium
The only varieties according to medium that we need to consider are those conditioned by speaking and writing respectively. Since speech is the primary or natural medium for linguistic communication, it is reasonable to see the present issue as a statement of the differences imposed on language when it has to be couched in a graphic (and normally visual) medium instead. Most of these differences arise from two sources. One is situational: the use of a written medium normally presumes the absence of the person(s) to whom the piece of language is addressed. This imposes the necessity of a far greater explicitness: the careful and precise completion of a sentence, rather than the odd word, supported by gesture, and terminating when the speaker is assured by word or look that his hearer has understood. As a corollary, since the written sentence can be read and re-read, slowly and critically (whereas the spoken sentence is mercifully evanescent), the writer tends to anticipate criticism by writing more concisely as well as more carefully and elegantly than he may choose to speak.
The second source of difference is that many of the devices we use to transmit language by speech (stress, rhythm, intonation, tempo, for example) are impossible to represent with the crudely simple repertoire of conventional orthography. They are difficult enough to represent even with a special prosodic notation: cf App 11.21. This means that the writer has often to reformulate his sentences if he is to convey fully and successfully what he wants to express within the orthographic system. Thus instead of the spoken sentence with a particular intonation nucleus on John (AppII.14)
jŏhn didn’t do it
one might have to write
It was not in fact John that did it. 
As with varieties according to subject matter, we are here dealing with two varieties that are in principle at the disposal of any user of English as occasion may demand, irrespective of the variety of English he uses as a result of region and education. But again there are contingent constraints: we do not expect less educated speakers to perform in written English with the facility that educated speakers acquire. This indeed is what a great deal of education is about.
There are contingent constraints of another kind. Some subject-matter varieties of English (legal statutes especially) are difficult to compose except in writing and difficult to understand except by reading. Other varieties are comparably restricted to speech: the transcript of a (radio) commentary on a football match might have passages like this:
Gerson to Pelé; a brilliant pass, that. And the score still: Brazil 4, Italy 1. The ball in-field to - oh, but beautifully cut off, and...
On the other hand, a newspaper report of the same game would be phrased very differently.
Varieties according to attitude
Varieties according to attitude constitute, like subject-matter and medium varieties, a range of English any section of which is in principle available at will to any individual speaker of English, irrespective of the regional variant or national standard he may habitually use. This present class of varieties is often called ‘stylistic’, but ‘style’ like ‘register’ is a term which is used with several different meanings. We are here concerned with the choice of linguistic form that proceeds from our attitude to the hearer (or reader), to the subject matter, or to the purpose of our communication. And we postulate that the essential aspect of the non-linguistic component (that is, the attitude) is the gradient between stiff, formal, cold, impersonal on the one hand and relaxed, informal, warm, friendly on the other. The corresponding linguistic contrasts involve both grammar and vocabulary. For example:
Overtime emoluments are not available for employees who are non-resident...
Staff members who don’t live in can’t get paid overtime...
While many sentences like the foregoing can be rated ‘more formal’ or ‘more informal’ (‘colloquial’) in relation to each other, it is useful to pursue the notion of the ‘common core’ (1.15) here, so that we can acknowledge a median or unmarked variety of English (see 1.35 Note), bearing no obvious colouring that has been induced by attitude. As in
This student’s work is now much better and seems likely to go on improving
and thousands of sentences like it. On each side of this normal and neutral English, we may usefully distinguish sentences containing features that are markedly formal or informal. In the present work, we shall for the most part confine ourselves to this three-term distinction, leaving the middle one unlabelled and specifying only usages that are relatively formal or informal. 
Mastery of such a range of attitudinal varieties seems a normal achievement for educated adults, but it is an acquisition that is not inevitable or even easy for either the native or the foreign learner of a language. It appears to require maturity, tact, sensitivity and adaptability - personality features which enable the individual to observe and imitate what others do, and to search the language’s resources to find expression to suit his attitude. The young native speaker at the age of five or six has broadly speaking one form of English that is made to serve all purposes, whether he is talking to his mother, his pets, his friends or the aged president of his father’s firm. And although even this can cause parents twinges of embarrassment, it is understood that the invariant language is a limitation that the child will grow out of.
The foreign learner is in a somewhat similar position. Until his skill in the language is really very advanced, it is attitudinally invariant, though the particular variety on which he is ‘fixed’ is much less predictable than that of the native child. If much of his practice in English has been obtained through textbooks specializing in commercial training, his habitual variety will be very different from that of the learner who has done vacation work helping on a farm. These are extreme examples, but it is a commonplace to notice an invariant literary, archaic flavour in the speech of foreign students, and even a Biblical strain in the students from some parts of the world. Better this no doubt than an excessively informal usage, but in any case just as the native child’s youth protects him from criticism so does the overseas student’s accent inform his listeners that there are respectable reasons for any inappropriateness in the language variety he uses.
The three-way contrast is not of course adequate to describe the full range of linguistic varieties that are evoked by differences of attitude. Martin Joos considers that we should at least add one category at each end of the scale to account for the extremely distant, rigid (he calls it ‘frozen’) variety of English sometimes found in written instructions, eg
Distinguished patrons are requested to ascend to the second floor
and to account also for the intimate, casual or hearty - often slangy -language used between very close friends (especially of similar age) or members of a family, or used when a speaker feels for any other reason that he does not need to bother what the listener (or reader) thinks of his choice of language. We might thus match the foregoing example with
Up you get, you fellows!
We are thus now in possession of a potential five-term distinction:
(rigid) - formal - normal - informal - (familiar)
One final point on attitude varieties. As with the English dictated by subject matter and medium, there are contingency constraints in the normal selection of attitudinal variety. Just as statute drafting (subject matter) normally presupposes writing (medium), so also it presupposes a particular attitude variety: in this case ‘rigid’. Similarly it would be hard to imagine an appropriate football commentary on the radio being other than informal, or a radio commentary on the funeral of a head of state being other than formal, though both are in the same medium (speech).
Varieties according to interference
Varieties according to interference should be seen as being on a very different basis from the other types of variety discussed. It is true that, theoretically, they need not be so sharply distinguished as this implies. We might think of the ‘common core’ (1.15) in native speakers being ‘distorted’ in one direction where a person is born in Ohio and in another direction if he is born in Yorkshire. The differences in their English might then be ascribed to the interference of Ohio speech and Yorkshire speech respectively on this common core.
But in more practical terms we apply ‘interference’ to the trace left by someone’s native language upon the foreign language he has acquired. Indeed, to be still more severely practical, we apply it only to those traces of the first language that it is pedagogically desirable to identify and eradicate. Otherwise, we should be applying an identical classification to linguistic situations that are sharply different: on the one hand, the recognizable features of Indian English or West African English (undoubtedly inherited from one generation to another) which teachers may be trying to eradicate and replace with speech habits more resembling BrE or AmE; and on the other hand, the recognizable features of Irish English (many of which are the reflexes of Irish Celtic), which are also passed on from one generation to another but which are approved by teachers as fully acceptable in educated Irish use.
The important point to stress is English acquired by speakers of other languages, whether as a foreign or as a second language (1.3-4), varies not merely with the degree of proficiency attained (elementary, intermediate, advanced, let us say) but with the specific native language background. The Frenchman who says, ‘I am here since Thursday’ is imposing a French grammatical usage on English; the Russian who says ‘There are four assistants in our chair of mathematics’ is imposing a Russian lexico-semantic usage on the English word ‘chair’. Most obviously, we always tend to impose our native phonological pattern on any foreign language we learn. The practised linguist is able to detect the language background of his English pupil and this has obvious implications for language teaching in devising drills that will be directed to helping students with the problems that give them the greatest difficulty.
At the opposite extreme are interference varieties that are so widespread in a community and of such long standing that they may be thought stable and adequate enough to be institutionalized and regarded as varieties of English in their own right rather than stages on the way to a more native-like English. There is active debate on these issues in India, Pakistan and several African countries, where efficient and fairly stable varieties of English are prominent in educated use at the highest political and professional level.
Creole and Pidgin
At an extreme of a different kind, there are interference varieties which have traditionally been used chiefly by the less prosperous and privileged sections of a community but which have also been stable over several generations. Political, educational and sociolinguistic thought vacillates as to whether such creolized forms of English (as in Sierra Leone or the Caribbean) should be institutionalized or not. Would Creole speakers benefit from the self-assurance this might give, or (since the elite in their society would still learn a more international English in addition) would the danger be that this would tend to perpetuate their underprivileged status? Here is a sample of Jamaican Creole in an orthography that already suggests partial institutionalization:
Hin sed den, ‘Ma, a we in lib?’ Hie sie, ‘Mi no nuo, mi pikini, bot duon luk fi bin niem hahd, ohr eni wie in a di wohid an yu kal di niem, bin hie unu.’ Hin sed, ‘Wel Ma, mi want im hie mi a nuo mi.’ ‘Lahd nuo, masa! Duo no kal di niem, hin wi kom kil yu.’ Hin sie,’ Wel Ma, hin wi haf fi kil mi.’ [See Note a] 
Creole is normally the principal or sole language of its speakers, being transmitted from parent to child like any other native language. Moreover, for all its evidence of interference from other languages, it is usually more like ordinary English than Pidgin is and gives less impression of being merely a drastic reduction of ordinary English.
Pidgin is technically distinguished from Creole by being essentially a ‘second’ language (1.3), used rather to replace a native language for restricted public (especially commercial) purposes than to conduct family affairs and talk to one’s children. In New Guinea an attempt has been made to raise the status of Pidgin (and its speakers) by institutionalization as ‘Neomelanesian’; a public press, local administration and some education both secular and religious are conducted in it. Here is a sample from the Neomelanesian version of St Mark’s Gospel (‘ Gud Nius Mark i Raitim’), Chapter 13, verse 13:
Na olman bai i bel nogud long yufela bilong nem bilong mi. Tasol man i stap strong oltaim i go i kamap long finis bilong em, disfela i ken stap gud oltaim. [See Note b] 
In this case (as distinct from the Creole example) it would be very dun-cult to spell the passage in conventional orthography, and this is an interesting indication that we are here beyond the limits where it is reasonable to speak of a variety of English.
Relationship between variety classes
In presenting the table of varieties in a schematic relationship in 1.15, reference was made to each stratum of varieties being equally related to all others. In principle, this is so. A man may retain recognizable features of any regional English in habitually using a national standard; in his national standard, he will be able to discourse in English appropriate to his profession, his hobbies, a sport; he could handle these topics in English appropriate either to speech or writing; in either medium, he could adjust his discourse on any of these subjects according to the respect, friendliness or intimacy he felt for hearer or reader. And all of this would be true if he was proficient in English as a foreign or second language and his usage bore the marks of his native tongue. Clearly, as we review this example, we must see that the independence of the varieties is not solely a matter of principle but also, to a large extent, a matter of actual practice.
But to an at least equally large extent the independence does not hold in practice. We have drawn attention to contingent constraints at several points (for example, in 1.29). Let us attempt to see the types of interdependence as they affect the varieties system as a whole. To begin with, the regional varieties have been explicitly connected with the educational and standard varieties. Thus although there is ‘independence’ to the extent that a speaker of any regional variety may be placed anywhere on the scale of least to most educated, there is interdependence to the extent that the regional variety will determine (and hence it dominates in the table, 1.15) the educational variety: a person educated in Ohio will adopt educated AmE not BrE. There is an analogous connection between the interference variety and the regional and educational variety:
someone learning English in Europe or India is likely to approach a standard with BrE orientation; if in Mexico or the Philippines, an AmE orientation.
Next, the subject-matter varieties. Certain fields of activity (farming and ship-building, for example) are associated with specific regions; clearly, it will be in the (especially uneducated) dialect of these regions and no others that the language of daily discourse on such activities will be thoroughly developed. In other fields (medicine, nuclear physics, philosophy) we will expect to find little use of uneducated English or the English of a particular region. In discussions of baseball, AmE will predominate but we will not expect to find the vocabulary or grammar specific to AmE in reports of cricket matches.
Since writing is an educated art, we shall not expect to find other than educated English of one or other national standard in this medium. Indeed, when we try on occasion to represent regional or uneducated English in writing, we realize acutely how narrowly geared to Standard English are our graphic conventions. For the same reason there are subjects that can scarcely be handled in writing and others (we have mentioned legal statutes) that can scarcely be handled in speech.
Attitudinal varieties have a great deal of independence in relation to other varieties: it is possible to be formal or informal on biochemistry or politics in AmE or BrE, for example. But informal or casual language across an ‘authority gap’ or ‘seniority gap’ (a student talking to an archbishop) presents difficulties, and on certain topics (funerals) it would be unthinkably distasteful. An attempt at formal or rigid language when the subject is courtship or football would seem comic at best.
Finally, the interference varieties. At the extremes of Creole and Pidgin there is especial interdependence between the form of language and the occasion and purposes of use: indeed the very name Pidgin (from ‘business’) should remind us that it is of its nature inclined to be restricted to a few practical subjects. Creole is usually more varied but again it tends to be used of limited subject matter (local, practical and family affairs). As to English taught at an advanced intellectual level as a second or foreign language, our constant concern must be that enough proficiency will be achieved to allow the user the flexibility he needs in handling (let us say) public administration, a learned discipline such as medicine with its supporting scientific literature, and informal social intercourse. The drawback with much traditional English teaching was that it left the foreign learner more able to discourse on Shakespeare than on machinery - and chiefly in writing at that. A swing towards a more ‘modern’ approach is hardly welcome if it concentrates on colloquial chit-chat, idioms and last year’s slang. Attempts to teach a ‘restricted’ language (‘English for engineers’) too often ignore the danger in so doing of trying to climb a ladder which is sinking in mud: it is no use trying to approach a point on the upper rungs if there is no foundation.
Our approach in this book is to keep our sights firmly fixed on the common core which constitutes the major part of any variety of English, however specialized, and without which fluency in any variety at a higher than parrot level is impossible What was said in 1.27 about an marked variety in respect of attitude applies also to the varieties conditioned by the other factors such as medium, subject matter and interference. Only at points where a grammatical form is being discussed which is associated with a specific variety will mention be made of the fact that the form is no longer of the common core. The varieties chiefly involved on such occasions will be AmE and BrE; speech and writing; formal and informal. 
Varieties within a variety
Two final points need to be made. First, the various conditioning factors
(region, medium, attitude, for example) have no absolute effect: one should not expect a consistent all-or-nothing response to the demands of informality or whatever the factor may be. The conditioning is real but relative and variable. Secondly, when we have done all we can to account for the choice of one rather than another linguistic form, we are still left with a margin of variation that cannot with certainty be explained in terms of the parameters set forth in 1.15 and discussed in subsequent paragraphs.
For example, we can say (or write)
He stayed a week
He stayed for a week
Had I known
If I had known
without either member of such pairs being necessarily linked to any of the varieties that we have specified. We may sometimes have a clear impression that one member seems rarer than another, or relatively old-fashioned, but although a rare or archaic form is likelier in relatively formal rather than in relatively informal English, we cannot always make such an identification. It might be true for the plural cacti as opposed to cactuses, but it would hardly be true for beer enough as opposed to enough beer, where the former is rarer but probably more used in informal (or dialectal) speech.
It may help to see variation in terms of the relationships depicted opposite, where both the verticals represent a ‘more-or-less’ opposition. The upper pole of the first vertical corresponds to the features of greatest uniformity, such as the invariable past tense of bring in the educated variety of English, or the many features characterizing the main stable common core of the language, such as the position of the article in a noun phrase. The lower pole of the first vertical corresponds to the area of fluctuation illustrated in 1.36. The second vertical represents the situation in which, on the other hand, an individual may indulge in such a fluctuation (I wonder whether one moment and I wonder if a little later), and on the other hand, there may be fluctuation within the community as a whole (one member appearing to have a preference for He didn’t dare ask and another a preference for He didn’t dare to ask: cf 3.2l). This appears to be a natural state of affairs in language. All societies are constantly changing their languages with the result that there arc always coexistent forms, the one relatively new, the other relatively old; and some members of a society will be temperamentally disposed to use the new (perhaps by their youth) while others are comparably inclined to the old (perhaps by their age). But many of us will not be consistent either in our choice or in our temperamental disposition. Perhaps English may give rise to such fluctuation more than some other languages because of its patently mixed nature: a basic Germanic wordstock, stress pattern,4 word-formation, inflection and syntax overlaid with a classical and Romance wordstock, stress pattern (App 11.4), word-formation (App 1.3) -and even inflection and syntax. The extent to which even highly educated people will treat the Latin and Greek plurals in data and criteria as singulars or will use different to and averse to rather than different from and averse from - and face objections from other native speakers of English - testifies to the variable acknowledgement that classical patterns of inflection and syntax (‘differre ab’, ‘aversus ab’) apply within English grammar. It is another sense in which English is to be regarded as ‘the most international of languages’ (1.7) and certainly adds noticeably to the variation in English usage with which a grammar must come to terms.
On English in relation to other languages, see British Council (1969), pp 7-22; Girard (1970); Halls (1969); Muller (1964).
On linguistics and the teaching of English, see Lyons (1968); Halliday, McIntosh and Strevens (1964); Nickel (1971).
On varieties of English, see Avis (1967); Branford (1970); Crystal and Davy (1969);
Hall (1966); Joos (1967); McDavid-Mencken(1963); Quirk (1972); Spencer (1971);
THE SENTENCE: A PRELIMINARY VIEW
2.1-11 Parts of the sentence
.1 Subject and predicate
.2 Operator, auxiliary, and predication
.3-8 Verb, complement, object, adverbial
.4 Complements and objects
.5-6 Categories of verb
.7-8 Categories of adverbial
.9 Some types of subject
.10 Types of sentence structure
.11 Element realization types
2.12-16 Parts of speech
.14 Closed-system items
.15 Open-class items
.16 Stative and dynamic
2.18-26 Sentence processes
.19 Yes-no questions
.20 The focus of a question .
.21-22 Assertion and non-assertion
.23 Negation and question
.24 Other processes
2.27 Relation to later chapters
The attitude of native speakers to other people's dialect varies greatly, but. in general,
dialects of rural and agricultural communities are regarded as more pleasant than dialects of large urban communities such as New York or Birmingham. This is connected, of course, with social attitudes and the association of city dialects with variation according to education and social standing (1.18) rather than region.
The extreme variation that is tolerated in the pronunciation of English in various countries puts a great responsibility upon the largely uniform orthography (1.19) in preserving the intercomprehensibility of English throughout the world. A ‘phonetic’ spelling would probably allow existing differences to become greater whereas -through ‘spelling pronunciation’ with increased literacy - our conventional orthography not merely checks the divisiveness of pronunciation change but actually reduces it.
Some subject matter (non-technical essays on humanistic topics, for example) invites
linguistic usages that we shall refer to as literary; others (law, religion) involve usages that are otherwise archaic, though there is a strong trend away from such archaism in these fields. Poetry also frequently uses archaic features of English, while ‘literary’ English must sometimes be described as poetic if it shows features that are rare in prose. By contrast, technical or learned writing, in showing a close relation to a particular subject matter (psychology, electronics, or linguistics, for example), is often pejoratively referred to as jargon, especially when technical language is used too obtrusively or to all appearances unnecessarily.
The advantages are not all on one side, however; the written medium has the valuable distinctions of paragraph, italics, quotation marks, etc, which have no clear analogue in speech (App ffl.1 ff)-
A further term, slang, is necessary to denote the frequently vivid or playful lexical usage that often occurs in casual discourse, usually indicating membership of a particular social group.
[a] He said then, ‘Ma, and where does he live?’ She says, ‘I don’t know, my child,
but don’t look hard for his name, or anywhere in all the world that you call the name, he will hear you.’ He said,’ Well, Ma, I want him to hear me and know me.’ ‘Lord, no, master! Do not call the name: he will come and kill you.’ He says,
‘Well, Ma, he will have to kill me.’
 [b] And everyone will feel badly towards you on account of my name. But anyone who stays strong right till the end, this person will remain in well-being for ever.
Relationship between variety classes
The distinction between ‘marked’ and ‘unmarked’ relates to the differing degrees of
inclusiveness, specificity and neutrality that two related linguistic forms may have. For example, while he and she are opposed as masculine and feminine respectively, the former can be regarded as unmarked in comparison with the latter since he can include ‘feminine’ more readily than she can include ‘masculine’ (as in ‘Ask anyone
and he will tell you’).