Б. А. ИЛЬИШ
Строй
Современного английского
Языка
Учебник по курсу теоретической грамматики для студентов педагогических институтов
На английском языке) ИЗДАНИЕ ВТОРОЕ
ИЗДАТЕЛЬСТВО „ПРОСВЕЩЕНИЕ"
ЛЕНИНГРАДСКОЕ ОТДЕЛЕНИЕ
ЛЕНИНГРАД 1971
Сканирование, распознавание, вычитка:
Аркадий Куракин, г. Николаев, июль 2004 г.
{ark # mksat. net}
Только для использования с некоммерческой целью студентами и преподавателями в учебном процессе.
Орфография из ам. переведена в британскую.
Исправлено ок. 15 опечаток. В частности: prepositoin (178), adressed (183) (2), stylistical (232), conjunctious (267), prepositoinal (283), Dickens’s (302), froom (310) interpretaiton (328), actoin (329), Enlgish (351).
—4 -71
Допущено Министерством просвещения СССР
в качестве учебника для студентов
педагогических институтов по специальности 2103
«Иностранные языки».
PREFACE TO THE SECOND EDITION
In preparing this edition, care has been taken to bring the text of the book up to date and to introduce the reader to some outstanding problems of modern linguistics. One of these concerns the relations between morphology and syntax, on the one hand, and paradigmatic and syntagmatic phenomena, on the other. Recent discussion of this problem has also immediate connection with the treatment of the notion of "sentence". Much attention has accordingly been given to this set of problems in the appropriate places.
Some corrections have also been made in various parts of the book.
Its main purpose remains unchanged. It is meant to encourage the students to think on the essential problems of English language structure and to form their own views of the relevant questions.
B. Ilyish
September 1970
PREFACE TO THE FIRST EDITION
This book is intended as a textbook for the theoretical course on English grammar forming part of the curriculum in our Universities and Teachers' colleges. Its main purpose is to introduce the student to the many linguistic problems connected with grammatical structures and to the modern methods applied in dealing with them. I have endeavoured, as far as was possible, to point out the essence of the problems, and to state the arguments which have been, or may be, put forward in favour of one view or another. This should enable the reader to form a judgement of his own on the question involved and on the respective merits of the various solutions proposed.
It will be found that in many points the views expressed here differ from those laid down in my earlier work on the subject, published in Russian in 1948.1 I have not thought it necessary or expedient to point out in every case the motives which have brought about these changes. The development of linguistics in the last few decades has been so quick and so manifold that a new insight has been gained into practically all the problems dealt with here, and into many others as well, for that matter. This of course was bound to be reflected in the contents of the book and in its very structure.
I have tried to avoid mentioning too many names of scholars or titles of books, preferring to call the reader's attention to the problems themselves. Some hints about authors have of course been given in the footnotes.
1 Б. А. Ильиш, Современный английский язык, изд. 2-е, 1948. 1*
4 Preface
A few words may not be out of place here concerning the kind of work students may bo expected to do in their seminar hours. This may include, besides analysis of modem texts from theoretical points of view treated in the book, reports on the same problems, and discussion of views held by various authors. Some of these problems will probably lend themselves more readily than others to such discussion; among them, the following may be suggested: parts of speech in English; the category of case in nouns and pronouns; the stative; aspect; the perfect and the problem of correlation; voice; prepositions and conjunctions; types of sentences; types of predicate; secondary parts of a sentence; asyndetic composite sentences. Of course much will depend in each case on the teacher's own choice and on the particular interests expressed by the students.
My sincere thanks are due to the chair of English grammar of the Lenin Pedagogical Institute, Moscow, and the chair of English philology of Leningrad University, for the trouble they took in reviewing the MS, and also to Mr William Ryan, postgraduate student of Oriel College, Oxford, who went through the MS and suggested many improvements in the wording of the text.
Ilyish
INTRODUCTION
THE PURPOSE OF THE BOOK
The purpose of this book is to present a systematic study of the grammatical structure of Modern English. It presupposes a sufficient knowledge on the part of the reader of the practical rules pertaining both to the morphology and to the syntax of the language. Thus, we are not going to set out here the ways, for example, of forming the plural of English nouns, or those of forming the past tense of English verbs. It will be our task to give an analysis of English grammatical structure in the light of general principles of linguistics. This is going to involve, in a number of cases, consideration of moot points on which differing views have been expressed by different scholars. In some cases the views of scholars appear to be so far apart as to be hardly reconcilable. It will be our task to consider the main arguments put forward to sustain the various views, to weigh each of them, and to find out the most convincing way of solving the particular problem involved.
What the student is meant to acquire as a result of his studies is an insight into the structure of the language and an ability to form his own ideas on this or that question. This would appear to be a necessary accomplishment for a teacher of English (at whatever sort of school he may be teaching), who is apt to find differing, and occasionally contradictory, treatment of the grammatical phenomena he has to mention in his teaching. Such are, for example, the system of parts of speech, the continuous forms of the verb, the asyndetic composite sentences, etc.
In the course of the history of linguistics many different views of language and languages have been put forward. It is not our task to discuss them here. Suffice it to say that the treatment of a language as a system was characteristic of the grammarians of the 17th century (see, for instance, the French "Grammaire générale de Port-Royal", a grammar published in 1660). Though this was not a linguistic work in any modern sense, it was based on the assumption that the state of a language at a given period was a system and could be treated as such. This view of language structure was then abandoned in favour of a purely historical outlook until the. early years of the 20th century, when the Swiss scholar Ferdinand de Saussure (1857—1913) laid the foundations of a new linguistic theory acknowledging the study of a system of a given language as such. l De Saussure's views were then developed and modified by various schools of modern linguistic thinking. Part, at least, of his views of language were adopted, with certain reservations, by
1 P. de Saussure, Cours de linguistique générale, Geneve, 1922.
6 Introduction
the bulk of Soviet scholars. It is on the basis of this view that a theoretical investigation of the grammatical system of a language at a definite point of its history becomes possible and fruitful.
A peculiarity of the modern trend of linguistics is the desire to arrive at results independent of the view of a particular scholar. There can hardly be any doubt that the ability to arrive at such results would mark a significant advance in linguistics, which has far too long been suffering from conflicts between contradictory views put forward by various authors and disputed by others. As far as can be foreseen at the moment, the area of objective results not to be disputed will gradually increase at the expense of the debated area, which, however, can hardly be expected ever to disappear altogether. In discussing this or that particular problem in this book, we will try to define what can be said to be firmly established and what remains controversial.
A word is necessary here about the limits of grammar as part of a language's structure and the other aspects (or "levels") of language, viz. the phonetic (phonological) and the lexical.
It need hardly be emphasised that a language is a whole consisting of parts closely united. The linguist's task is, accordingly, to point out the demarcation line separating those aspects or levels from one another, on the one hand, and the connections between them, on the other. This is by no means an easy task, as we shall more than once have occasion to observe. Our subject is the grammatical structure of English, and we shall have to delineate the borderlines and connections between grammatical structure, on the one hand, and phonetics (phonology) and the vocabulary, on the other.
LANGUAGE AND SPEECH
The distinction between language and speech, which was first introduced by Ferdinand de Saussure in his book on general linguistics, has since become one of the cornerstones of modern linguistics. Though differences of opinion still persist in the exact delineation of the boundaries between the two spheres, its general idea has been accepted by most scholars.
-Language, then, is the system, phonological, lexical, and grammatical, which lies at the base of all speaking. It is the source which every speaker and writer has to draw upon if he is to be understood by other speakers of the language.
Speech on the other hand, is the manifestation of language, or its use by various speakers and writers of the given language. Thus what we have before us, in oral or in written form, as material for analysis, is always a product of speech, namely something either-pronounced or written by some individual speaker or writer or, occasionally, a group of speakers or writers. There is no other way
Language and Speech 7
for a scholar to get at language than through its manifestations in speech.
As we are here concerned with grammar only, we will not dwell on the problem of a language system in phonology, orthography, and lexicology, but we will concentrate on the system of grammar and of its manifestations in speech, where of course it can never appear isolated from phonology and lexicology.
Thus, in stating that English nouns have a distinction of two numbers, singular and plural, and that there are several ways of expressing the category of plural number in nouns, we are stating facts of language, that is, elements of that system on which a speaker or writer of English has to draw.
Similarly, the statement that in English there are phrases of the pattern "adverb + adjective + noun", is certainly a statement about language, namely, about the syntactical system of English on the phrase level. Thus, in building such concrete phrases as, very fine weather, extremely interesting novel, strikingly inadequate reply, etc., a speaker draws, as it were, on the stock of phrase patterns existing in the language and familiar to its speakers, and he fills the pattern with words, choosing them from the stock of words existing in the language, in accordance with the thought or feeling, etc., that he wants to express. For instance, the concrete phrase, strikingly inadequate reply, is a fact of speech, created by the individual speaker for his own purposes, and founded on a knowledge, (a) of the syntactical pattern in question, and (b) of the words which he arranges according to the pattern.
It may perhaps be said, with some reservations, that the actual sentences pronounced by a speaker, are the result of organising words drawn from the language's word stock, according to a pattern drawn from its grammatical system.
So it appears that the, material which a scholar takes up for investigation is always a fact of speech. Were it not for such facts of speech, whether oral or written, linguistic investigation would not be all possible. It is the scholar's task, then, to analyse the speech facts which are at his disposal, in such a manner as to get through them to the underlying language system, without which they could not have been produced.
NEW METHODS
The last few years have seen a rapid development of various new methods of linguistic investigation, and there is a great variety of views as to their merits.
Briefly, the three main positions in this field may be summarised as follows:
(1) Some scholars think that the new methods now appearing mark the beginnings of linguistics as a science and that everything
8 Introduction
that was done earlier in linguistics belongs to a "prescientific age".
(2) Other scholars are sceptical about the new methods and think that they tend to lead linguistic science away from its proper tasks and to replace it by something incompatible with its essential character.
(3) There is the view that the new methods mark a new period in the development of linguistics, and should be tried out, without implying that everything done in earlier periods should therefore be considered as valueless and "prescientific".
Without going into details about this discussion we will merely. state that the view mentioned last appears to be the most reasonable one and the one likely to prevail in the long run, as has more than once been seen in the history of different branches of learning.
We will therefore keep in our treatment of English grammatical structure many ideas and terms inherited from traditional grammar, such as, for instance, the theory of the parts of speech and parts of the sentence, and at the same time point out what new light is shed on these problems by recently developed methods, and what change the formulation of the very issues should undergo in the light of the new ideas. It will not be too much to say that a considerable number of familiar statements about grammatical facts cannot now be upheld without essential modification, and it would be pointless to ignore this fact. On the other hand, much of what is convincing and useful in the new views has not yet attained a shape which would make it convenient for presentation in a textbook like the present. It will therefore be our task to introduce the reader at least to some of these problems, and to help him prepare for reading the numerous special treatises on these subjects.
What appears to be most essential in the light of new ideas which tend to make linguistics something like an exact science, is a distinction between problems admitting of a definite solution which can be convincingly demonstrated and cannot be denied, and problems admitting of various opinions, rather than of a definite solution. This must not be taken to mean that problems of the second kind should be abandoned: they should be further discussed and their discussion is likely to be fruitful. The point is that an opinion, which can exist side by side with another opinion, should not be presented as a final solution admitting of no alternative. It is especially in the sphere of syntax that problems admitting of various opinions rather than of definite solutions are to be found.
Although in some cases the line between the two sets of problems may be rather hard to draw, the basic difference between them should be always kept in mind. This will help the student to put both the problems themselves and the views of different authors in the proper perspective.
Ore Grammatical Statements 9
In discussing grammatical categories, we shall often have to mention oppositions, that is, pairs of grammatical forms opposed to each other in some way. A simple case in point is the opposition between the singular and the plural number in nouns, with their definite meanings: one as against more than one.
It is often found that of two members of an opposition one has quite a definite meaning, whereas the meaning of the other is less definite, or vague. This is found, for instance, in the opposition between the forms was writing and wrote: the meaning of the form was writing is quite definite, while that of the form wrote is hard even to define. The terms usual for such cases are, "marked" and "unmarked". Thus, the form was writing is the marked, and the form wrote, the unmarked member of the opposition. We shall have more than one occasion to apply these terms.
ON GRAMMATICAL STATEMENTS
As the teaching of a language to foreigners requires the formulation of rules which the learner has to observe if he is to speak and write the language correctly, practical grammars, written both by speakers of the language in question and by foreigners, tend to be excessively strict in laying down what is "inadmissible" in the language. Numerous specimens of exaggerations may be found practically in every grammar book.
Let us consider a few of the most characteristic examples of such exaggerations.
It is frequently laid down as a rule that verbs of perception, such as see, hear, feel (in the meaning 'experience'), also those denoting emotions, such as love, like, hate, etc., cannot be used in any of the continuous forms.
This rule, thus bluntly formulated, is not borne out by actual usage. All of these verbs can, under certain circumstances, be used in the continuous forms though of course they are less commonly used in these forms than, say, verbs of physical action, such as walk, beat, strike, jump, run, etc., or verbs of position in space, such as stand, sit, lie, hang, kneel, etc. To be sure, was seeing is a much rarer form than was running. And yet was seeing is not impossible, nor is was hearing, was liking, etc., and also was being, e. g.. in the sentence He was being polite to you. In a similar way, the verb feel can be used in the continuous form, as for instance in a question addressed to a sick person: Are you feeling better to-day?
Another example of a rule formulated too bluntly is that about the use of tenses in a conditional if -clause. It usually runs something like this: "In a conditional if -clause the present tense is used instead of the future." There are two points to be noted here.
10 Introduction
(1) The expression "instead of the future" has no reasonable sense at all. What is meant here is that if the action mentioned in a conditional if -clause refers to the future the present tense of the verb is used. (2) Besides, the rule, thus formulated, is much too strict, and requires some modification. If it is taken literally at its face value, it should mean that in a clause of this type the groups "shall + infinitive" and "will + infinitive" are completely inadmissible. This, however, is rather far from the truth, at least, in so far as the group "will + infinitive" is concerned. This group may, in fact, be found in conditional if -clauses. The verb will apparently has a certain trace of its lexical meaning preserved, but the group nevertheless is an analytical form of the future tense, as will be seen from the following example: Twenty thousand francs for you, Madame, if you'll stop breathing on my neck and go away. (R. WEST) Thus, an absolute prohibition of the use of "will + infinitive" in conditional if -clauses proves to be a misstatement of the facts of the language. With reference to the group "shall + + infinitive" the statement appears to be more true. Indeed it is hard to find examples of such a use, and the rule may be laid down with a very high degree of probability.
Careful observation of the facts of the language and attention to their possible stylistic colouring (compare also p. 354 ff.) will often help to modify some too strict prohibitions and assertions of impossibility to be found in grammarians' statements.
SOME GENERAL CONSIDERATIONS ON THE STRUCTURE OF ENGLISH
It is a very common statement that Modern English is an analytical language, as distinct from Modern Russian, which is synthetical. Occasionally this statement is slightly modified, to the effect that English is "mainly analytical" and Russian "mainly synthetical". These statements, on the whole, are true, but they remain somewhat vague until we have made clear two important points, viz. (a) what we mean by "analytical language'', and (b) what are the peculiar features distinguishing Modern English from other analytical languages, for instance, Modern French. It would be a gross error to suppose that English and French, being both analytical, are exactly alike in their grammatical structure.
The chief features characterising an analytical language would seem to be these:
(1) Comparatively few grammatical inflections (viz., case inflections in nouns, adjectives, and pronouns, and personal inflections in verbs).
(2) A sparing use of sound alternations to denote grammatical forms.
Order of Discussion 11
(3) 
A wide use of prepositions to denote relations between objects and to connect words in the sentence.
(4) Prominent use of word order to denote grammatical relations: a more or less fixed word order.
Now, features distinguishing the Modern English language from, say, Modern French, are also fairly numerous.
Without going into more minute details, it may be pointed out here that English adjectives are not inflected for either gender or number, whereas French adjectives are, or that English has no future tense formed without auxiliary verbs, whereas French has one, or again, that in English the attributive adjective (with a few exceptions) comes before its noun, whereas in French such an adjective (with a few exceptions, too) comes after it, etc. These examples may be sufficient to show that by calling the English language analytical we do not give an adequate description of its structure. We shall arrive at that adequate description only at the end of the book.
ORDER OF DISCUSSION
The order in which we are going to deal with the problems of English grammatical structure is roughly the following.
First, we shall have to attempt an approximate definition of the boundaries between morphology and syntax in present-day English.
Then, at the start of our morphological investigation, we shall have to establish the morphological resources of the English language, viz. the morphemes and other means of expressing morphological categories.
Our next point will be a general survey of the system of word classes (the so-called "parts of speech"), and a detailed investigation of the structure of each of them in particular. That will be the end of the morphological section.
The syntactical part will consist of two very unequal items: the theory of phrases and the theory of the sentence. These parts are bound to be unequal because the theory of phrases (in its syntactical aspect) seems to be the least developed element of English grammar, whereas the theory of the sentence has a long-drawn-out and fruitful history.
The phrase theory will have to deal with the various types of phrases (noun and verb, verb and adverb, etc.) in their grammatical, as distinct from their lexical, aspect. The theory of the sentence will include a review of the types of simple sentence and parts of the sentence, and of the various types of composite sentences.
At the end we shall try to give a general view of Modern English grammatical structure on the basis of the preceding investigation.
12 Introduction
Wherever it may seem desirable and illuminating, we shall draw parallels between the English language and other languages, notably Russian and German, pointing out both resemblances and discrepancies between them. This ought to help the reader acquire a more profound insight into the peculiarities of the language he is specialising in.
MORPHOLOGY AND SYNTAX
Though the difference and the boundary between morphology and syntax seem obvious enough as a matter of principle, drawing a clear-cut line between them in a given language sometimes proves to be a task of some difficulty. Let us consider a few cases of this kind in Modern English.
The usual definition of morphology, which may be accepted as it stands, is this: Morphology is the part of grammar which treats of the forms of words. 1 As for the usual definition of syntax, it may be said to be this: Syntax is the part of grammar which treats of phrases and sentences. 2
These definitions are based on the assumption that we can clearly distinguish between words and phrases. This, however, is far from being the case. Usually the distinction, indeed, is patent enough. E. g., indestructibility is obviously a word, long as it is, whereas came here, short as it is, is a phrase and thus falls under the heading of syntax. But now what are we to make of has been found? This is evidently a phrase since it consists of three words and thus it would seem to fall under syntax, but it is also a form of the verb find and thus it would seem to fall under morphology.
Of course many more examples might be given of a phrase being at the same time a form of a word. It is obvious that we have here a kind of overlapping of syntax and morphology. It seems most advisable to include all such cases under morphology, considering the syntactical side of the formation to have been put, as it were, at the disposal of morphology.
The problem becomes more complicated still if we take into account such formations as has been often found, where one word (often) comes to stand between two elements of the form of another word (find). Such formations will have to be considered both under morphology and under syntax.
There are also other cases of overlapping which will be pointed out in due course. All this bears witness to the fact that in actual
1 We will not consider here those definitions of morphology which include word-building.
2 Different authors have differing views on the relations between the two parts of syntax.
Morphology and Syntax 13
research work we do not always find hard-and-fast lines separating phenomena from each other, such lines as would make every single phenomenon or group of phenomena easy to classify. More than once we shall have to deal with more involved groupings which must be treated accordingly. For the present the usual preliminary definition of the borderline between morphology and syntax must suffice.
There is also another way of approach to the problem of distinguishing between morphology and syntax.
Let us take as an example the sentence Could you take me in to town? (GALSWORTHY)
The word take which is used in this sentence can be considered from two different viewpoints.
On the one hand, we can consider it in its surroundings in the sentence, namely in its connection with the word you, which denotes the doer of the action, with the word me, which denotes the object of the action, etc. This would be analysing the syntagmatic connections of the word take.
On the other hand, we can consider take as part of a system including also the forms takes, taking, took, taken; we can observe that this system is analogous, both in sound alternation and in meanings, to the system forsake, forsakes, forsaking, forsook, forsaken, and, in a wider perspective, to the system write, writes, writing, wrote, written; sing, sings, singing, sang, sung, etc., and in a wider perspective still, to the system live, lives, living, lived; stop, stops, stopping, stopped, etc. This would be analysing the paradigmatic connections of take, and this gradually opens up a broad view into the morphological system of the language. It should be emphasised that this view is basically different from any view we might obtain by analysing the syntagmatic connections of the form in the sentence. For instance, the connection between took and wrote is entirely unsyntagmatic, as a sequence took wrote is unthinkable.
It may be said that, in a way, morphology is more abstract than syntax, as it does not study connections between words actually used together in sentences, but connections between forms actually found in different sentences and, as it were, extracted from their natural surroundings.
In another way, however, morphology would appear to be less abstract than syntax, as it studies units of a smaller and, we might say, of a more compact kind, whereas syntax deals with larger units, whose types and varieties are hard to number and exhaust.
The peculiar difficulty inherent in the treatment of analytical verb forms mentioned above, such as have done, will go, etc., lies in the fact that they have both a morphological and a syntactical quality. They are morphological facts in so far as they belong to the
14 Introduction
system of the verb in question, as the auxiliary verb adds nothing whatever to the lexical meaning expressed in the infinitive or participle making part of the analytical form. But the same forms are facts of syntax in so far as they consist of two or three or sometimes four elements, and occasionally some other word, which does not in any way make part of the analytical form, may come in between them. It is true that in Modern English possibilities of such insertions are not very great, yet they exist and must be taken into account. We will not go into details here and we will only point out that such words as often, never, such words as perhaps, probably, etc. can and in some cases must come between elements of an analytical verb form: has always come, will probably say, etc. Since it is impossible that a word should be placed within another word, we are bound to admit that the formation has... come is something of a syntactical formation. The inevitable conclusion is, then, that has come and other formations of this kind are simultaneously analytical verb forms and syntactical unities, and this obviously means that morphology and syntax overlap here (see above, p. 13). This is perhaps still more emphasised by the possibility of formations in which the auxiliary verb making part of an analytical verb form is co-ordinated with some other verb (usually a modal verb) which does not in any way make part of an analytical form, e. g. can and will go. This would apparently be impossible if the formation will go had nothing syntactical about it. 1
According to a modern view, the relation between morphology and syntax is not so simple as had been generally assumed. In this view, we ought to distinguish between two angles of research:
(1) The elements dealt with; from this point we divide grammatical investigation into two fields: morphology and syntax.
(2) The way these elements are studied; from this viewpoint we distinguish between paradigmatic and syntagmatic study. Thus we get four divisions:
1 a paradigmatic morphology b syntagmatic morphology
2 a paradigmatic syntax b syntagmatic syntax
According to this view, whenever we talk of parts of speech (substantives, adjectives, etc.), we remain within the sphere of morphology. Thus the statement that an adjective is used to modify a substantive, or that an adverb is used to modify a verb, is a state-
1 The same applies to the Russian language: there, too, a word can come in between the auxiliary and the infinitive, as in the formation буду завтра заниматься, and the auxiliary may be co-ordinated with another verb, as in хочу и буду заниматься.
Grammar and Word-building 15
ment of syntagmatic morphology. Syntax should have nothing to do with parts of speech: it should only operate with parts of sentence (subject, predicate, etc.).
Of these four items, the first and the last require no special explanation. Paradigmatic morphology is what we used to call morphology, and syntagmatic syntax is what we used to call syntax. The two other items, however, do require some special comment. Syntagmatic morphology is the study of phrases: "substantive + substantive", "adjective + substantive", "verb + substantive", "verb + adverb", etc.
Paradigmatic syntax, on the other hand, is a part of grammatical theory which did not appear as such in traditional systems. Paradigmatic syntax has to deal with such phenomena as
My friend has come.
My friend has not come.
Has my friend come?
My friend will come.
My friend will not come.
Will my friend come?
My friends have come.
My friends have not come, etc.
All these are considered as variation of one and the same sentence.
It would seem that the term sentence is here used in a peculiar sense. As units of communication My friend has come and My friend has not come are certainly two different sentences, as the information they convey is different. To avoid this ambiguity of the term sentence, it would be better to invent another term for "paradigmatic sentence". However, inventing a new term which would be generally acceptable is very difficult. In this book we shall use the term sentence in its old communicative sense.
GRAMMAR AND WОRD-BUILDING
The relations between word-building, grammar, and lexicology have not yet been made quite clear. By and large three views have been expressed: (1) word-building is part of lexicology, (2) word-building is partly at least a matter of grammar, (3) word-building is a special sphere intermediate between lexicology and grammar and occasionally encroaching upon either.
According as one or another of these views is endorsed, word-building is either ignored in a book on grammar, as something lying beyond its sphere, or it is treated of in grammar book to some extent, at least.
The difficulty of the question is illustrated by the very fact of such different views being taken by scholars.
16 Introduction
We will not here take up the question in its entirety, as it is obviously a question of general linguistics rather than of English linguistics, and we will merely state some points which we will follow in our treatment of the matter.
A complete enumeration of all suffixes and prefixes existing in a language and used to build words cannot be the task of a grammar. The meaning of such word-building suffixes as, e. g., -ness or -er for nouns, -ful or -less for adjectives, etc., cannot and should not be considered in grammar, any more than grammar can give a complete list of nouns, adjectives, etc. The grammatical aspect of word-building is, that words belonging to a certain part of speech are (or can be) derived by means of certain morphemes, chiefly suffixes (but in a few cases also prefixes), vowel alternation, and so forth.
From this viewpoint it is essential to note that a few word-building morphemes are unambiguous, that is, a word containing them is sure to belong to a certain part of speech, whereas others are ambiguous, that is, the morpheme is not in itself sufficient to make sure that a word belongs to a definite part of speech.
We need not give here any complete list of affixes of either type. A few typical examples will be all that is needed.
Affixes unambiguously showing to what part of speech a word belongs are very few. Among them is the suffix -ity for nouns. In such cases as scarcity, necessity, peculiarity, monstrosity, etc., there is no doubt that the word is a noun. In the sphere of adjectives there is the suffix -less and the suffix -ous. For instance, useless, harmless, fatherless, meaningless can be identified as adjectives by the mere fact of their having this suffix, and so can the words copious, hazardous, luminous, callous, ubiquitous, and so forth.
In the sphere of verbs we may note the suffix -ise (also spelt -ize) as an unambiguous sign of a word being a verb: cf. crystallise, immunise, organise, mobilise, vaporise, and the like.
Most word-forming morphemes are ambiguous, that is, they do not with certainty point to any definite part of speech but leave some choice which has to be decided by other criteria. The wideness of the choice varies with different morphemes. Thus, for instance, the suffix -ful leaves us only one alternative: the word can either be an adjective, which is the more usual case (useful, careful, truthful, masterful, needful, sinful, etc.), or a noun, which is much rarer (handful, spoonful, mouthful, pocketful, roomful, etc.). It will be readily seen that the second type is limited to formations in which the first element denotes some physical object having a certain volume.
In a similar way, the suffix -ment leaves open the choice between noun and verb, of which the first is much more frequent: compare
Grammar and Word-building 17
the nouns instrument, tenement, merriment, government, sentiment, pigment, basement, and the verbs implement, regiment, augment. It will be seen that most of these have homonyms among the nouns. It might perhaps be argued that -ment itself is a noun-forming, not a verb-forming suffix, and that verbs like implement have been formed from the corresponding nouns without any suffix at all. This may be true, but it is irrelevant: the fact remains that in contemporary English we have both nouns and verbs containing the suffix -ment followed by no other word-forming suffix.
Other suffixes may leave us a choice between three or more possibilities, for instance the suffix -ly leaves open the choice between adjectives, adverbs, nouns, modal words, and particles. We shall give a few examples of each category. Adjectives in -ly: orderly, friendly, comely, sickly, masterly; adverbs: kindly, safely, generally, merrily, joyfully; nouns: daily (a newspaper published every day; a woman coming in as daily help), orderly (a soldier assigned to an officer for carrying messages); modal words: possibly, probably, certainly, presumably, admittedly; particles: exclusively, merely, solely. If modal words are not accepted as a separate part of speech, or if words like merely are included among adverbs, the number of possibilities will be reduced by one or two items. But even so we shall have to admit that the suffix -ly is of comparatively little value for determining the part of speech to which a word belongs.
Prefixes are only rarely found to distinguish one part of speech from another.1 Here are some well-known examples: endear v. vs. dear adj., enlarge v. vs. large adj., enmesh v. vs. mesh n., behead v. vs. head n.,2 belittle v. vs. little adj.
A few more examples of this kind may be found, but there is not a single prefix to show definitely to what part of speech a word belongs. For instance, the negative prefix in- may be found in nouns (independence, intransigence), in adjectives (independent, intransigent, inconclusive), in verbs (incapacitate) and in adverbs (independently, inconclusively, inconsistently), so that as an indication of a part of speech it is valueless.
The prefix under- is also to be found in nouns, adjectives, and verbs, for instance, understudy, undersecretary (nouns), underfed,
1 We are not here concerned with the historical origins of this state of things, and therefore we do not dwell on the fact that, for instance, the verb behead comes from Old English beheafdian, which was derived from the Old English noun heafod by means of a suffix as well as of a prefix, nor do we make similar remarks about the verb endear, etc. However such a state of things may have originated, the fact remains that in Modern English the two parts of speech are distinguished by the prefix alone.
2 It might be argued that there is a verb head as well. But the meanings of the two verbs are so very far apart that this argument does not seem convincing.
18 Introduction
underdeveloped, underdone (adjectives), undervalue, underestimate, undermine (verbs).
Other means of word-building are vowel alternations and consonant alternations. However, these are so limited in their application that the presence of this or that vowel or consonant in a word can never be a sure sign of its belonging to a definite part of speech. For example, the alternation [u:] — [i:], in spelling, oo — ee, is found in a few noun-verb groups (doom — deem, food — feed), but it does not follow that the vowel [i:] (spelt ee) is a sign of a verb: there are numerous words belonging to other parts of speech having this vowel in their root: spleen, beech, deed are nouns; keen, green, deep are adjectives, etc. We need not give any more examples. The same is also true of consonant alternations, for example, the alternation [k] — [tS] in such pairs as speak — speech, break — breach, etc. Important as they are from a lexical viewpoint, their grammatical significance is next to nil.
Thus the grammatical aspect of word-building, at least in English, is rather unimportant; the main phenomena of word-building belong to the sphere of lexicology.
In concluding our observations on word-building we may note some so-called nonce-words formed without any suffixes. Thus, in the following example a nonce-word, namely a verb, is formed from an adverb without any suffixes and it is characterised as a verb merely by its surroundings in the sentence. This is a dialogue between a mother and her daughter who was rather late in coming back home from school. "Then where have you been? It's late." "Nowhere." "What?" "Nowhere." "Don't nowhere me. I know how long it takes to walk home from school." (WOODHILL) The third nowhere is shown to be the infinitive of a verb by its position between don't and me. Its meaning is clear from the context. Don't nowhere me obviously means much the same as, Don't say "nowhere" to me, or, Don't try to deceive me by saying "nowhere". In the following example even an unfinished sentence consisting of two words is treated in this way: "Now, Dora — " he began. "Don't you 'Now, Dora' me!" she said in a loud voice, frantically striking the sides of the chair with her hands. "I just can't stand this any longer! I just can't!" (E. CALDWELL) Such formations are not very frequent, and they are conversational rather than literary.
Our study of Modern English morphology will consist of four main items, viz. (1) essentials of morphology: general study of morphemes and types of word-form derivation, (2) the system of parts of speech, (3) study of each separate part of speech, the grammatical categories connected with it, and its syntactical functions.
Part One
Morphology
Chapter I
ESSENTIALS OF MORPHOLOGY
MORPHEMES
The morpheme is one of the central notions of grammatical theory, without which no serious attempt at grammatical study can be made. Definition of a morpheme is not an easy matter, and it has been attempted many times by different scholars. Without going into particulars of the discussions that have taken place, we may briefly define the morphemes as the smallest meaningful units into which a word form may be divided. For instance, if we take the form writers, it can be divided into three morphemes: (1) writ-, expressing the basic lexical meaning of the word, (2) -er-, expressing the idea of agent performing the action indicated by the root of the verb, (3) -s, indicating number, that is, showing that more than one person of the type indicated is meant. Similarly the form advantageously can be divided into three morphemes: advantage + ous + ly, each with a special meaning of its own.
Two additional remarks are necessary here: (1) Two or more morphemes may sound the same but be basically different, that is, they may be homonyms. Thus the -er morpheme indicating the doer of an action as in writer has a homonym — the morpheme -er denoting the comparative degree of adjectives and adverbs, as in longer. Which of the two homonymous morphemes is actually there in a given case can of course only be determined by examining the other morphemes in the word. Thus, the morpheme -er in our first example, writer, cannot possibly be the morpheme of the comparative degree, as the morpheme writ- to which it is joined on is not the stem of an adjective or adverb, and so no comparative degree is to be thought of here.
(2) There may be zero morphemes, that is, the absence of a morpheme may indicate a certain meaning. Thus, if we compare the forms book and books, both derived from the stem book-, we may say that while books is characterised by the -s -morpheme as being a plural form, book is characterised by the zero morpheme as being a singular form.
In modern descriptive linguistics the term "morpheme" has been given a somewhat different meaning.1 Scholars belonging to this trend approach the problem from this angle: If we compare the four sentences: the student comes, the students come; the ox comes, the oxen come, it will be seen that the change of student to students is
1 See, for example, H. A. Gleason Jr., An Introduction to Descriptive Linguistics, 1955, Chapter V.
22 Essentials of Morphology
paralleled by the change of ox to oxen. That is, the meaning and function of the -en in oxen is the same as the meaning and function of the -s in students. On this account the -s and the -en are said to represent the same morpheme: each of them is a morph representing the morpheme, and they are termed allomorphs of the morpheme. Furthermore, as in the word goose the form corresponding to students and oxen is geese, where nothing is added, but the root vowel is changed, the morph representing the morpheme in this case is said to be the very change of [u:] into [i:] (graphically, oo and ее). Thus the morpheme, in this case, has three allomorphs, (1) -s, (2) -en, (3) [u:] -> [i:].
This latter item entails some difficulty, as a morph is shown not necessarily to be a material entity, that is a phoneme, or a combination of phonemes; sometimes it may be a change of one phoneme into another.
Similarly, in the past tense of verbs the morpheme of the past tense has two allomorphs, (1) -ed, (2) the change of vowel, as [ai] —> [ou] (write — wrote), [i] -> [ae] (sing —sang), etc.
We will apply the term "morpheme" only to material units including zero.
In grammar, we are of course concerned with the grammatical, or structural, meaning of morphemes: we do not here study the meanings of root morphemes, which are necessarily lexical, and as to derivation morphemes, i. e. those which serve to build words, we are only interested in them in so far as they are grammatically relevant, and that is the case if they show that the word belongs to a certain part of speech, and if they serve to distinguish one part of speech from another. This grammatical significance of derivation morphemes, if it is there at all, is always combined with their lexical meaning. For instance, if we take this pair of words: write v. and writer n., the derivative morpheme -er has a grammatical significance, as it serves to distinguish a noun from a verb, and it has its lexical meaning, as the lexical meaning of the noun writer is different from that of the verb write.
Inflection morphemes have no lexical meaning or function. There is not the slightest difference in the way of lexical meaning between live and lived, or between house and houses. However, an inflection morpheme can acquire a lexical meaning in some special cases, for instance if the plural form of a noun develops a meaning which the singular form has not; thus, the plural form colours has a meaning, 'flag', which the singular form colour has not. These are cases of lexicalisation (compare below, p. 36).
There is in Modern English a case where a boundary line between inflection and derivation is hard to draw, and a morpheme does duty both ways. This is the morpheme -ing with its function of a suffix deriving verbal nouns and of an inflection serving to
Morphemes 28
form a gerund, which is one of the non-finite verb forms. This appears to be quite a special case in English, and it does not seem to find any parallel in Russian.
Terminology
It should be noted that there is some confusion in the use of the terms "suffix" and "inflection" or "ending".
According to one view, the term "suffix" is taken in a wide sense, and applied to any morpheme coming after the root morpheme, whether it is derivative or inflectional. If this view is endorsed, an inflection is a special kind of suffix, since it falls under the general definition of a suffix just mentioned.
According to another view, the term "suffix" is taken in a narrow sense, and applied to derivational post-root morphemes only. In that case an inflection is not a special kind of suffix but a morpheme of a different kind, having no lexical meaning of any sort.
We will adhere to this latter view, as it seems better to have a clear distinction than to use the term "suffix" in a vague sense.
There is also some slight vacillation in the use of the term "inflection" (or "ending"), and it is connected with the twofold use of the term "suffix" considered above.
According to one view, the term "inflection" (or "ending") is applied to any morpheme serving to derive a form of a word and having no lexical meaning. So the morphemes characterising the infinitive, for instance, in Russian or German, will be termed inflections.
According to another view, the term "inflection" (or "ending") is only applied to morphemes expressing case and number in nominal parts of speech, and to those expressing person and number in verbs. From this viewpoint the morpheme characterising the infinitive in Russian or German or the morpheme characterising the gerund in English would not be an inflection or ending, whereas the morpheme -s in forms like writes, buys, would be one.
We will apply the term "inflection" to every morpheme serving to derive a grammatical form and having no lexical meaning of its own.
Last not least, a question should be mentioned in this connection, which concerns adjectives and adverbs, namely that of degrees of comparison: are formations like longer, longest forms of the adjective or the adverb long, or are they different words? This means: is there a difference in the lexical meaning between long, longer, and longest, or is there not? This question has been treated in various ways, but the view seems to be prevalent that there is no difference in the lexical meaning here. We will take this view, too, and class the morphemes used to derive degrees of comparison among inflections (see below, p. 58).
24 Essentials of Morphology
TYPES OF WORD-FORM DERIVATION
These fall under two main headings, (a) those limited to changes in the body of the word, without having recourse to auxiliary words (synthetic types), (b) those implying the use of auxiliary words (analytical types). Besides, there are a few special cases of different forms of a word being derived from altogether different stems.
Synthetic Types
The number of morphemes used for deriving word-forms in Modern English is very small (much smaller than either in German or in Russian, for instance). They may be enumerated in a very short space.
There is the ending -s (-es), with three variants of pronunciation, used to form the plural of almost all nouns, and the endings -en and -ren, used for the same purpose in one or two words each, viz. oxen, brethren (poet.), children.
There is the ending -'s, with the same three variants of pronunciation as for the plural ending, used to form what is generally termed the genitive case of nouns.1
For adjectives, there are the endings -er and -est for the degrees of comparison.
For verbs, the number of morphemes used to derive their forms is only slightly greater. There is the ending -s (-es) for the third person singular present indicative, with the same three variants of pronunciation noted above for nouns, the ending -d (-ed) for the past tense of certain verbs (with three variants of pronunciation, again), the ending -d (ed) for the second participle of certain verbs, the ending -n (-en) for the second participle of certain other verbs, and the ending -ing for the first participle and also for the gerund.
Thus the total number of morphemes used to derive forms of words is eleven or twelve, which is much less than the number found in languages of a mainly synthetical structure.
It should also be noted that most of these endings are mono-semantic, in the sense that they denote only one grammatical category and not two or three (or more) at a time, as is the case in synthetic languages. For example, the plural -s (or -es) denotes only the category of plural number, and has nothing to do with any other grammatical category, such as case.
It would, however, mean oversimplifying matters if we were to suppose that all English inflectional morphemes are monosemantic. This is certainly not the case with the -s (-es) of the third
1 The problem of the genitive case will be dealt with in Chapter III (see p. 41 ff.).
Types of Word-form Derivation 25
person singular. It expresses at least three grammatical categories: person (third), number (singular), and mood (indicative). In certain verbs it also expresses the category of tense: thus, in the form puts only the -s shows that it is a present-tense form.
Sound Alternations
By sound alternations we mean a way of expressing grammatical categories which consists in changing a sound inside the root. This method appears in Modern English, for example, in nouns, as when the root vowel [ae] of the singular form man is changed into [e] to form the plural men, or similarly the root vowel [au] of mouse is changed into [aı] in mice, and a few more cases of the same kind.
This method is much more extensively used in verbs, such as write — wrote — written, sing — sang — sung, meet — met — met, etc. On the whole, vowel alternation does play some part among the means of expressing grammatical categories, though its part in Modern English has been much reduced as compared to Old English.
Analytical Types
These consist in using a word (devoid of any lexical meaning of its own) to express some grammatical category of another word.
There can be no doubt in Modern English about the analytical character of such formations as, e. g., has invited or is invited, or is inviting, or does not invite. The verbs have, be, and do have no lexical meaning of their own in these cases. The lexical meaning of the formation resides in the participle or infinitive following the verb have, be or do. Some doubt has been expressed about the formations shall invite and will invite. There is a view that shall and will have a lexical meaning.1 We will not go into this question now and we will consider shall and will as verbs serving to form the future tense of other verbs. Thus, have, be, do, shall, and will are what we call auxiliary verbs, and as such they constitute a typical feature of the analytical structure of Modern English.
While the existence of analytical forms of the English verb cannot be disputed, the existence of such forms in adjectives and adverbs is not nowadays universally recognised. The question whether such formations as more vivid, the most vivid, or, again, more vividly and most vividly are or are not analytical forms of degrees of comparison of vivid and vividly, is controversial. We can only say here that if these formations are recognised as analytical forms of degrees of comparison, the words more and most have to be numbered among the analytical means of morphology.
1 See below, p. 87.
26 Essentials of Morphology
Suppletive Format ions
Besides the synthetical and analytical means of building word forms in Modern English, there is yet another way of building them which stands quite apart and is found in a very limited number of cases only. By a suppletive formation we mean building a form of a word from an altogether different stem. Examples in point are, the verb go, with its past tense went; the personal pronoun I, with its objective case form me, the adjective good with its comparative degree form better, and a few more. We consider, for instance, go and went as, in a way, two forms of one word, because in the vast majority of verbs the past tense is derived from the same stem as the present or infinitive, e. g. live — lived, speak — spoke, etc. It is against this background that the units go and went come to be considered as forms of one word, formed from different stems. In the morphological system of Modern English suppletive formations are a very insignificant element, but they concern a few very widely used words among adjectives, pronouns, and verbs.
Such, then, are the means of deriving the forms of words in Modern English. We shall have to ascertain the exact meaning and function of each of them as we proceed on our survey of the parts of speech.
Chapter II
PARTS OF SPEECH (General Survey)
The problem of parts of speech is one that causes great controversies both in general linguistic theory and in the analysis of separate languages. We shall have to examine here briefly a few general questions concerning parts of speech which are of some importance for Modern English.
The term "parts of speech" (as well as the corresponding terms in Russian, German, French, and other languages), though firmly established, is not a very happy one. What is meant by a "part of speech" is a type of word differing from other types in some grammatical point or points. To take the clearest example of all, the verb is a type of word different from all other types in that it alone has the grammatical category of tense. Thus, while it is perfectly reasonable to ask, "What is the past tense of the word live?" (the answer of course is, lived), it would make no sense to ask, "What is the past tense of the word city?" or "What is the past tense of the word big?" Those words just have not got any past tense, or any tense whatever, for that matter: the notion of tense cannot be applied to them. Tense is one of the distinctive features characterising the verb as against every other type of word. However, the question is much less simple with reference to some other types of words, and a general definition of the principles on which the classification of parts of speech is based becomes absolutely necessary.
We cannot here go into the controversy over these principles that has lasted a considerable time now, and we will limit ourselves to stating the principles of our classification and pointing out some difficulties inherent in it.
The principles on which the classification is based are three in number, viz. (1) meaning, (2) form, (
