Syntactical relations between the components of a phrase

These fall under two main heads: (1) agreement or concord, (2) government.

Syntactical Relations between the Components of a Phrase 175

Agreement

By agreement we mean a method of expressing a syntactical relationship, which consists in making the subordinate word take a form similar to that of the word to which it is subordinate. In Modern English this can refer only to the category of number: a subordinate word agrees in number with its head word if it has different, number forms at all.¹ This is practically found in two words only, the pronouns this and that, which agree in number with their head word. Since no other word, to whatever part of speech it may belong, agrees in number with its head word, these two pronouns stand quite apart in the Modern English syntactical system.

As to the problem of agreement of the verb with the noun or pronoun denoting the subject of the action (a child plays, children play), this is a controversial problem. Usually it is treated as agreement of the predicate with the subject, that is, as a phenomenon of sentence structure. However, if we assume (as we have done) that agreement and government belong to the phrase level, rather than to the sentence level, and that phrases of the pattern "noun + + verb" do exist, we have to treat this problem in this chapter devoted to phrases.

The controversy is this. Does the verb stand, say, in the plural number because the noun denoting the subject of the action is plural, so that the verb is in the full sense of the word subordinate to the noun? Or does the verb, in its own right, express by its category of number the singularity or plurality of the doer (or doers)?²

There are some phenomena in Modern English which would seem to show that the verb does not always follow the noun in the category of number. Such examples as, My family are early risers, on the one hand, and The United Nations is an international organisation, on the other, prove that the verb can be independent of the noun in this respect: though the noun is in the singular, the verb may be in the plural, if the doer is understood to be plural; though the noun is plural, the verb may be singular if the doer is understood to be singular. Examples of such usage are arguments in favour of the view that there is no agreement in number of the verb with the noun expressing the doer of the action.

The fact that sentences like My family is small, and My family are early risers exist side by side proves that there is no agreement

¹ In some other languages, such as Russian, there is also agreement in case and gender.

² This question was raised with reference to Indo-European languages in general by A. Meillet in his book Introduction a l'étude comparative des langnes indoenropeennes, 6^eme ed., 1924, p. 323, and with reference to the Russian language by A. Peshkovsky (see A. M. Пешковский, Русский синтаксис в научном освещении, изд. 7-е, 1956, стр. 183 сл.).

176 Phrases

of the verb with the noun in either case: the verb shows whether the subject of the action is to be thought of as singular or plural, no matter what the category of number in the noun may be.

Thus, the sphere of agreement in Modern English is extremely small: it is restricted to two pronouns — this and that, which agree with their head word in number when they are used in front of it as the first components of a phrase of which the noun is the centre.

Government

By government we understand the use of a certain form of the subordinate word required by its head word, but not coinciding with the form of the head word itself — that is the difference between agreement and government.

The role of government in Modern English is almost as insignificant as that of agreement. We do not find in English any verbs, or nouns, or adjectives, requiring the subordinate noun to be in one case rather than in another. Nor do we find prepositions requiring anything of the kind.

The only thing that may be termed government in Modern English is the use of the objective case of personal pronouns and of the pronoun who when they are subordinate to a verb or follow a preposition. Thus, for instance, the forms me, him, her, us, them, are required if the pronoun follows a verb (e. g. find or invite) or any preposition whatever. Even this type of government is, however, made somewhat doubtful by the rising tendency, mentioned above (p. 66 ff.), to use the forms me, him, etc., outside their original sphere as forms of the objective case. The notion of government has also become doubtful as applied to the form whom, which is rather often superseded by the form who in such sentences as, Who(m) did yon see? (compare p. 69).

As to nouns, the notion of government may be said to have become quite uncertain in present-day English. Even if we stick to the view that father and father's are forms of the common and the genitive case, respectively, we could not assert that a preposition always requires the form of the common case. For instance, the preposition at can be combined with both case forms: compare I looked at my father and I spent the summer at my father's, or, with the preposition to: I wrote to the chemist, and I went to the chemist's, etc. It seems to follow that the notion of government does not apply to forms of nouns.

Other Ways

In Russian linguistic theory, there is a third way of expressing syntactical relations between components of a phrase, which is termed примыкание. No exact definition of this notion is given:

Syntactical Relations between the Components of a Phrase 177

its characteristic feature is usually described in a negative way, as absence both of agreement and of government. The most usual example of this type of connection is the relation between an adverb and its bead word, whether this is an adjective or a verb (or another adverb, for that matter). An adverb is subordinate to its head word, without either agreeing with or being governed by it. This negative characteristic cannot, however, be said to be sufficient as a definition of a concrete syntactical means of expression. It is evident that the subject requires some more exact investigation. For instance, if we take such a simple case as the sentence,... lashes of rain striped the great windows almost horizontally (R. WEST) and inquire what it is that shows the adverb horizontally to be subordinate to the verb striped, we shall have to conclude that this is achieved by a certain combination of factors, some of which are grammatical, while others are not. The grammatical factor is the fact that an adverb can be subordinate to a verb. That, however, is not sufficient in a number of cases. There may be several verbs in the sentence, and the question has to be answered, how does the reader (or hearer) know to which of them the adverb is actually subordinated. Here a lexicological factor intervenes: the adverb must be semantically compatible with its head word. Examples may be found where the connection between an adverb and its head word is preserved even at a considerable distance, owing to the grammatical and semantic compatibility of the adverb. Compare, for instance, the following sentences: Nobly, nobly Cape Saint Vincent to the North-West died away. (BROWNING) Swiftly he thought of the different things she had told him. (DREISER)

An adverb can only be connected with its head word in this manner, since it has no grammatical categories which would allow it to agree with another word or to be governed by it. With other parts of speech things stand differently in different languages. In inflected languages an adjective will agree with its head word, and even in French and Italian, though they are analytical languages, adjectives agree with their head words both in number and gender. In Modern English no agreement is possible. The same can be said about many other types of phrases.

However, there is another means of expressing syntactical connection which plays a significant part in Modern English. It may be called "enclosure" (Russian замыкание) and its essence is this. Some element of a phrase is, as it were, enclosed between two parts of another element. The most widely known case of "enclosure" is the putting of a word between an article and the noun to which the article belongs. Any word or phrase thus enclosed is shown to be an attribute to the noun. As is well known, many other words than adjectives and nouns can be found in that position, and many phrases, too. It seems unnecessary to give examples of adjectives

178 Phrases

and nouns in that position, as they are familiar to everybody. However, examples of other parts of speech, and also of phrases enclosed will not be out of place here. The then government — here the adverb then, being enclosed between the article and the noun it belongs to, is in this way shown to be an attribute to the noun. ¹ In the phrase an on-the-spot investigation the phrase on-the-spot is enclosed between the article and the noun to which the article belongs, and this characterises the syntactic connections of the phrase.

The unity of a phrase is quite clear if the phrase as a whole is modified by an adverb. It is a rather common phenomenon for an adverb to modify a phrase, usually one consisting of a preposition and a noun (with possible words serving as attributes to the noun). Here, first, is an example where the phrase so modified is a phraseological unit:... that little thimbleful of brandy... went sorely against the grain with her. (TROLLOPE) The adverb sorely cannot possibly be said to modify the preposition against alone. So it is bound to belong to the phrase against the grain as a whole.

An adverb modifying a prepositional phrase is also found in the following example: The funeral was well under way. (HUXLEY) The adverb well can only modify the phrase under way, as a phrase well under is unthinkable. This is possible because the phrase under way, which is a phraseological unit, has much the same meaning as going on, developing, etc.

A phrase may also be modified by a pronoun (it should be noted, though, that in our example the whole phrase, including the pronoun, is a phraseological unit): Every now and again she would slop and move her mouth as though to speak, but nothing was said. (A. WILSON) It is clear that a phrase every now would not be possible. A similar case is the following: Every three or four months Mr Bodiharn preached a sermon on the subject. (HUXLEY) It is quite evident that the whole phrase three or four months is here modified by the pronoun every. This may be to some extent connected with the tendency to take phrases consisting of a numeral and a noun in the plural indicating some measure of time or space as denoting a higher unit (compare p. 38).

The phrase "noun + after + the same noun" may be a syntactic unit introduced as a whole by a preposition, thus: She spent the Christmas holidays with her parents in the northern part of the State, where her father owned a drug-store, even though in letter after letter Eve Grayson had urged and begged her to come to New Orleans for the holidays, promising that she would meet many interesting men while she was there. (E. CALDWELL) That the preposition in introduces the whole phrase letter after letter is evident

¹ Another view is that then is an adjective here.

Equivalent to Prepositions and Conjunctions 179

from the fact that it would not be possible to use the noun letter (alone) after the preposition without either an article or some other determinative, such as, for example, her.

In the following example the preposition with introduces, not a noun, but a phrase consisting of a noun, a preposition (upon) and the same noun repeated. Brown varnished bookshelves lined the walls, filled with row upon row of those thick, heavy theological works which the second-hand booksellers generally sell by weight. (HUXLEY) That the preposition with introduces the phrase row upon row rather than the noun row alone, is evident from the fact that it would not be possible to say... filled with row of those... works... The noun row could not be used without the article, to say nothing of the fact that one row of books was not enough to fill the walls of a room.

Sometimes a phrase of the pattern "adverb + preposition + + noun" may be introduced by another preposition. Compare this sentence from Prof. D. Jones's Preface to his "English Pronouncing Dictionary": For help in the preparation of this new edition I am particularly indebted to Mr P. A. D. MacCarthy, who supplied me with upwards of 500 notes and suggestions. The phrase upwards of 500 notes and suggestions means the same as more than 500 notes and suggestions, and this may explain its use after the preposition with. But the fact remains that a preposition (with) is immediately followed by a prepositional phrase (upwards of).

PHRASES EQUIVALENT TO PREPOSITIONS AND CONJUNCTIONS

Under this heading wo will treat such formations as apart from, with reference to, as soon as, so long as, etc., which quite obviously are phrases rather than words, and which quite definitely perform the same function in a sentence as prepositions and conjunctions respectively.

The treatment of these units in grammatical theory has been vague and often contradictory. Most usually they are treated as prepositions or conjunctions of a special type, variously described as compound, analytical, etc. This view ignores the basic difference between a word and a phrase and is therefore unacceptable. We will stick to the principle that a phrase (as different from a word) cannot be a part of speech and that phrases should be studied in Syntax.

An obstacle to this treatment was the view that a phrase must include at least two notional words (see above, p. 170). As we have rejected this limitation, we can include under phrases any groups, whether consisting of a form word and a notional word, or of two form words, etc.

180 Phrases

Among phrases equivalent to prepositions we note the pattern "adverb + preposition", represented, for instance, by out of, apart from, down to, as in the sentences, "I love you so," she answered, "but apart from that, you were right." (R. WEST) As the cool of the evening now came on, Lester proposed to Aram to enjoy it without, previous to returning to the parlour. (LYTTON) All within was the same, down to the sea-weed in the blue mug in my bedroom. (DICKENS) The phrases equivalent to prepositions (we may accept the term "prepositional phrases") perform the very functions that are typical of prepositions, and some of them have synonyms among prepositions. Thus, the phrase apart from is a synonym of the preposition besides, the phrase previous to a synonym of the preposition before, etc.

Another pattern of prepositional phrases is "preposition + + noun + preposition", e. g. in front of, on behalf of, with reference to, in accordance with, as in the sentences, His friend was seated in front of the fire. (BLACK) Caesar crossed in spite of this. (JEROME K. JEROME) It must be admitted that there may be doubts whether a group of this type has or has not become a prepositional phrase. Special methods can then be used to find this out. For instance, it may prove important whether the noun within such a phrase can or cannot be modified by an adjective, whether it can or cannot be changed into the plural, and so forth. Opinions may differ on whether a given phrase should or should not be included in this group. On the whole, however, the existence of such prepositional phrases is beyond doubt.

Other types of phrases ought to be carefully studied in a similar way, for example the phrase of course, which is the equivalent of a modal word, etc.

The number of phrases equivalent to conjunctions is rather considerable. Some of the more specialised time relations are expressed by phrases, e. g. as soon as, as long as. Phrases with other meanings also belong here, e. g. in order that, notwithstanding that. These phrases may be conveniently termed "conjunctional phrases", though this term is not so usual as the term "prepositional phrases".

There are several patterns of conjunctional phrases. One of them is "adverb + adverb + conjunction" (as soon as, as long as, so long as). The first component of the two former phrases is probably an adverb, though it might also be argued that it is a conjunction. We may say that the distinction between the two is here neutralised.

There is also the pattern "preposition + noun + conjunction", as in the phrase in order that, which is used to introduce adverbial clauses of purpose, or in the phrase for fear that, which tends to become a kind of conjunctional phrase introducing a special kind of clause of cause: For fear that his voice might betray more of his feel-

Phrases Equivalent to Preposition and Conjunction 181

ings, which would embarrass the old lady so involved still with her voyage and getting away to where it would be quiet again, so without such sudden, sick floods of sentiment herself, he simply repeated again how good, good it was to see her... (BUECHNER)¹

It would appear that the treatment of such phrases attempted here does better justice both to their structure and function than a treatment which includes them under prepositions and conjunctions proper and thus obliterates the essential difference between words (parts of speech) and phrases (groups of words).

In passing now from a study of phrases to that of the sentence we are, it should be remembered, proceeding to a different level of language structure. Notions referring to the phrase level should be carefully kept apart from those referring to the sentence and its members. An indiscriminate use of terms belonging to the two levels (as, for instance, in the familiar expression "subject, verb and object") leads to a hopeless muddle and makes all serious syntactic investigation impossible. It must, however, be pointed out that in some cases distinction between the two levels proves to be a very difficult task indeed. ² We will try in such cases to point out whatever can be urged in favour of each of the diverging views and to suggest a solution of the problem.

¹ From the lexicological viewpoint some of these phrases functioning as equivalents of prepositions and conjunctions must certainly be described as phraseological units. This, however, is irrelevant for their grammatical characteristic.

² We shall see this when we come to the problem of the attribute, (p. 222 ff).

Chapter XXIV

THE SENTENCE

The notion of sentence has not so far received a satisfactory definition, which would enable us by applying it in every particular case to find out whether a certain linguistic unit was a sentence or not.

Thus, for example, the question remains undecided whether such shop notices as Book Shop and such book titles as English are sentences or not. In favour of the view that they are sentences the following consideration can be brought forward. The notice Book Shop and the title English Grammar mean 'This is a book shop', 'This is an English Grammar'; the phrase is interpreted as the predicative of a sentence whose subject and link verb have been omitted, that is, it is apprehended as a unit of communication. According to the other possible view, such notices as Book Shop and such titles as English Grammar are not units of communication at all, but units of nomination, merely appended to the object they denote. Since there is as yet no definition of a sentence which would enable us to decide this question, it depends on everyone's subjective view which alternative he prefers. We will prefer the view that such notices and book titles are not sentences but rather nomination units.

We also mention here a special case. Some novels have titles formulated as sentences, e. g. The Stars Look Down, by A. Cronin, or They Came to a City, by J. B. Priestley. These are certainly sentences, but they are used as nomination units, for instance, Have you read The Stars Look Down?, Do you like They Came to a City? ¹

With the rise of modern ideas of paradigmatic syntax yet another problem concerning definition of sentence has to be considered.

In paradigmatic syntax, such units as He has arrived, He has not arrived, Has he arrived, He will arrive, He will not arrive, Will he arrive, etc., are treated as different forms of the same sentence, just as arrives, has arrived, will arrive etc., are different forms of the same verb. We may call this view of the sentence the paradigmatic view.

Now from the point of view of communication, He has arrived and He has not arrived are different sentences since they convey different information (indeed, the meaning of the one flatly contradicts that of the other).

¹ The same may be found in Russian, for instance in some titles of plays by Alexander Ostrovski: Бедность не порок, Свои люди — сочтемся, Не в свои сани не садись.

Classification of Sentences 183

CLASSIFICATION OF SENTENCES

The problem of classification of sentences is a highly complicated one, and we will first consider the question of the principles of classification, and of the notions on which it can be based.

Let us begin by comparing a few sentences differing from each other in some respect. Take, for example, the following two sentences: (1) But why did you leave England? (GALSWORTHY) and (2) There are to-day more people writing extremely well, in all departments of life, than ever before; what we have to do is to sharpen our judgement and pick these out from the still larger number who write extremely badly. (CRUMP)

Everyone will see that the two sentences are basically different. This is true, but very general and not grammatically exact. In order to arrive at a strictly grammatical statement of the difference (or differences) between them we must apply more exact methods of observation and analysis.

Let us, then, proceed to a careful observation of the features which constitute the difference between the two sentences.

1. The first sentence expresses a question, that is the speaker expects an answer which will supply the information he wants. The second sentence expresses a statement, that is, the author (or.speaker) states his opinion on a certain subject. He does not ask about anything, or expect anybody to supply him any information. This difference is expressed in writing by the first sentence having a question mark at the end, while the second sentence has a full stop.

2. The first sentence is addressed to a certain hearer (or a few hearers present), and is meant to provoke the hearer's reaction (answer). The second sentence is not addressed to any particular person or persons and the author does not know how anybody will react to it.

3. The two sentences differ greatly in length: the first consists of only 6 words, while the second has 39.

4. The first sentence has no punctuation marks within it, while the second has two commas and a semicolon.

5. The first sentence has only one finite verb (did... leave), while the second has three (are, have, write).

These would seem to be some essential points of difference. We have not yet found out which of them are really relevant from a grammatical viewpoint. We have not included in the above list those which are quite obviously irrelevant from that viewpoint; for example, the first sentence contains a proper name (England), while the second does not contain any, or, the second sentence contains a possessive pronoun (our) while the first does not, etc.

Let us now consider each of the five points of difference and see which of them are relevant from a purely grammatical point of view, for a classification of sentences.

184 The Sentence

Point 1 states a difference in the types of thought expressed in the two sentences. Without going into details of logical analysis, we can merely say that a question (as in the first sentence), and a proposition (as in the second) ¹ are different types of thought, in the logical acceptation of that term. The problem now is, whether this difference is or is not of any importance from the grammatical viewpoint. In Modern English sentences expressing questions (we will call them, as is usually done, interrogative sentences) have some characteristic grammatical features. These features are, in the first place, a specific word order in most cases (predicate — subject), as against the order subject — predicate ² in sentences expressing propositions (declarative sentences). Thus word order may, with some reservations, be considered as a feature distinguishing this particular type of sentence from others. Another grammatical feature characterising interrogative sentences (again, with some reservations) is the structure of the predicate verb, namely its analytical form "do + infinitive" (in our first sentence, did.., leave..., not left), where in a declarative sentence there would be the simple form (without do). However, this feature is not restricted to interrogative sentences: as is well known, it also characterises negative sentences. Anyhow, we can (always with some reservations) assume that word order and the form "do + infinitive" are grammatical features characterising interrogative sentences, and in so far the first item of our list appears to be grammatically relevant. We will, accordingly, accept the types "interrogative sentence" and "declarative sentence" as grammatical types of sentences.

Point 2, treating of a difference between a sentence addressed to a definite hearer (or reader) and a sentence free from such limitation, appears not to be grammatical, important as it may be from other points of view. Accordingly, we will not include this distinction among grammatical features of sentences.

Point 3, showing a difference in the length of the sentences, namely in the number of words making up each of them, does not in itself constitute a grammatical feature, though it may be more remotely connected with grammatical distinctions.

Point 4 bears a close relation to grammatical peculiarities; more «specially, a semicolon would be hardly possible in certain types of sentences (so-called simple sentences). But punctuation marks within a sentence are not in themselves grammatical features: they

¹ As a matter of fact, our second sentence contains more than one proposition; but this does not affect the basic difference between the two types of sentences.

² We will here provisionally accept the terms "subject" and "predicate" without definition. For a full discussion of these terms see p. 198 ff.

Classification of Sentences 185

are rather a consequence of grammatical features whose essence is to be looked for elsewhere.

Point 5, on the contrary, is very important from a grammatical viewpoint. Indeed the number of finite verbs in a sentence is one of its main grammatical features. In this particular instance it should be noted that each of the three finite verbs has its own noun or pronoun belonging to it and expressing the doer of the action denoted by the verb: are has the noun people, have the pronoun we, and write the pronoun who. These are sure signs of the sentence being composite, not simple.¹ Thus we will adopt the distinction between simple and composite sentences as a distinction between two grammatical types.

The items we have established as a result of comparing the two sentences given on page 183 certainly do not exhaust all the possible grammatical features a sentence can be shown to possess. They were only meant to illustrate the method to be applied if a reasonable grammatical classification of sentences is to be achieved. If we were to take another pair or other pairs of sentences and proceed to compare them in a similar way we should arrive at some more grammatical distinctions which have to be taken into account in making up a classification. We will not give any more examples but we will take up the grammatical classification of sentences in a systematic way.

It is evident that there are two principles of classification. Applying one of them, we obtain a classification into declarative, interrogative, and imperative sentences. We can call this principle that of "types of communication".

The other classification is according to structure. Here we state two main types: simple sentences and composite sentences. We will not now go into the question of a further subdivision of composite sentences, or into the question of possible intermediate types between simple and composite ones. These questions will be treated later on (see pages 200 and 254 respectively). Meanwhile, then, we get the following results:

Types of Sentences According to Types of Communication

(1) Declarative

(2) Interrogative

(3) Imperative

Sentences belonging to the several types differ from each other in some grammatical points, too. Thus, interrogative sentences are

¹ We use the term "composite sentence" in the same meaning as that attributed to it in H. Poutsma's Grammar of Late Modern English, namely as opposite to the term "simple sentence".

1S6 The Sentence

characterised by a special word order (see Chapter XXX). In interrogative sentences very few modal words are used, as the meanings of some modal words are incompatible with the meaning of an interrogative sentence. It is clear that modal words expressing full certainty, such as certainly, surely, naturally, etc., cannot appear in a sentence expressing a question. On the other hand, the modal word indeed, with its peculiar shades of meaning, is quite possible in interrogative sentences, for instance, Isn't so indeed? (SHAKESPEARE)

There are also sentences which might be termed semi-interrogative. The third sentence in the following passage belongs to this type:

"Well, I daresay that's more revealing about poor George than you. At any rate, he seems to have survived it." "Oh, you've seen him?" She did not particularly mark her question for an answer, but it was, after all, the pivot-point, and Bone found himself replying — that indeed he had. (BUECHNER) The sentence Oh, you've seen him? is half-way between the affirmative declarative sentence, You have seen him, and the interrogative sentence, Have you seen him? Let us proceed to find out the precise characteristics of the sentence in the text as against the two sentences just given for the sake of comparison. From the syntactical viewpoint, the sentence is declarative, as the mutual position of subject and predicate is, you have seen, not have you seen, which would be the interrogative order. In what way or ways does it, then, differ from a Usual declarative sentence? That is where the question of the intonation comes in. Whether the question mark at the end of the sentence does or does not mean that the intonation is not that typical of a declarative sentence, is hard to tell, though it would rather seem that it does. To be certain about this a phonetic experiment should be undertaken, but in this particular case the author gives a context which itself goes some way toward settling the question. The author's words, She did not particularly mark her question for an answer, seem to refer to the intonation with which it was pronounced: the intonation must not have been clearly interrogative, that is not clearly rising, though it must have differed from the regular falling intonation to some extent: if it had not been at all different, the sentence could not have been termed a "question", and the author does call it a question. Reacting to this semi-interrogative intonation, Bone (the man to whom the question was addressed) answered in the affirmative. It seems the best way, on the whole, to term such sentences semi-interrogative. Their purpose of course is to utter a somewhat hesitating statement and to expect the other person to confirm it.

Imperative sentences also show marked peculiarities in the use of modal words. It is quite evident, for example, that modal words expressing possibility, such as perhaps, maybe, possibly, are incom-

Classification of Sentences 187

patible with the notion of order or request. Indeed, modal words are hardly used at all in imperative sentences.

The notion of exclamatory sentences and their relation to the three established types of declarative, interrogative, and imperative sentences presents some difficulty. It would seem that the best way to deal with it is this. On the one hand, every sentence, whether narrative, interrogative, or imperative, may be exclamatory at the same time, that is, it may convey the speaker's feelings and be characterised by emphatic intonation and by an exclamation mark in writing. This may be seen in the following examples: Bat he can't do anything to you! (R. WEST) What can he possibly do to you! (Idem) Scarlett, spare me! (M. MITCHELL)

On the other hand, a sentence may be purely exclamatory, that is. it may not belong to any of the three types classed above. This would be the case in the following examples: "Well, fiddle-dee-dee!" said Scarlett. (M. MITCHELL) Oh, for God's sake, Henry! (Idem)

However, it would perhaps be better to use different terms for sentences which are purely exclamatory, and thus constitute a special type, and those which add an emotional element to their basic quality, which is either declarative, or interrogative, or imperative. If this view is endorsed, we should have our classification of sentences according to type of communication (see p. 185) thus modified:

(1) Declarative (including emotional ones)

(2) Interrogative (including emotional ones)

(3) Imperative (including emotional ones)

(4) Exclamatory

This view would avoid the awkward contradiction of exclamatory sentences constituting a special type and belonging to the first three types at the same time.

Types of Sentences According to Structure

(1) Simple

(2) Composite

The relations between the two classifications should now be considered.

It is plain that a simple sentence can be either declarative, or interrogative, or imperative. But things are somewhat more complicated with reference to composite sentences. If both (or all) clauses making up a composite sentence are declarative, the composite sentence as a whole is of course declarative too. And so it is bound to be in every case when both (or all) clauses making a composite sentence belong to the same type of communication (that is the case in an overwhelming majority of examples). Sometimes, however, composite sentences are found which consist of clauses

188 The Sentence

belonging to different types of communication. Here it will sometimes he impossible to say to what type of communication the composite sentence as a whole belongs. We will take up this question when we come to the composite sentence.

Some other questions connected with the mutual relation of the two classifications will be considered as we proceed.

THE SIMPLE SENTENCE

We will now study the structure of the simple sentence and the types of simple sentences.

First of all we shall have to deal with the problem of negative sentences. The problem, briefly stated, is this: do negative sentences constitute a special grammatical type, and if so, what are its grammatical features? In other words, if we say, "This is a negative sentence," do we thereby give it a grammatical description?

The difficulty of the problem lies in the peculiarity of negative expressions in Modern English. Let us take two sentences, both negative in meaning: (1) She did not know when she would be seeing any of them again. (R. MACAULAY) (2) Helen's tremendous spell — perhaps no one ever quite escaped from it. (Idem) They are obviously different in their ways of expressing negation. In (1) we see a special form of the predicate verb (did... know, not knew) which is due to the negative character of the sentence and is in so far a grammatical sign of its being negative. In (2), on the other hand, there is no grammatical feature to show that the sentence is negative. Indeed, there is no grammatical difference whatever between the sentences Nobody saw him and Everybody saw him. The difference lies entirely in the meaning of the pronouns functioning as subject, that is to say, it is lexical, not grammatical. The same is of course true of such sentences as I found nobody and I found everybody. On the other hand, in the sentence I did not find anybody there is again a grammatical feature, viz. the form of the predicate verb (did... find, not found).

The conclusion to be drawn from these observations is obviously this. Since in a number of cases negative sentences are not characterised as such by any grammatical peculiarities, they are not a grammatical type. They are a logical type, which may or may not be reflected in grammatical structure. Accordingly, the division of sentences into affirmative and negative ought not to be included into their grammatical classification. ¹

¹ If we were to accept affirmative and negative sentences as grammatical types, we should find it very awkward to deal with sentences like Nobody saw him or I found nothing: we should have to class them as affirmative. The category of negation does of course exist in the morphological system of the English verb (see above, p. 123 ff.).

The Simple Sentence 189

Before we proceed with our study of sentence structure it will be well to consider the relation between the two notions of sentence and clause. Among different types of sentences treated in a syntactic investigation it is naturally the simple sentence that comes first. It is with specimens of simple sentences that we study such categories as parts of the sentence, main and secondary; homogeneous members, word order, etc. It is also with specimens of simple sentences that we illustrate such notions as declarative, interrogative, imperative, and exclamatory sentences, as two-member and one-member sentences, and so forth. As long as we limit ourselves to the study of simple sentences, the notion of "clause" need not occur at all.

When, however, we come to composite sentences (that is, sentences consisting of two or more clauses), we have to deal with the notions of main clause, head clause, and subordinate clause. Everything we said about the simple sentence will also hold good for clauses: a clause also has its parts (main and secondary), it can also be a two-member or a one-member clause; a main clause at least must also be either declarative, interrogative, imperative, or exclamatory, etc. We will consider these questions in due course.

So then we will take it for granted that whatever is said about a simple sentence will also apply to an independent clause within a composite sentence. For instance, whatever we say about word order in a simple sentence will also apply to word order in an independent clause within a composite sentence, etc.

TYPES OF SIMPLE SENTENCES. MAIN PARTS OF A SENTENCE

It has been usual for some time now to classify sentences into two-member and one-member sentences. ¹

This distinction is based on a difference in the so-called main parts of a sentence. We shall therefore have to consider the two problems, that of two-member and one-member sentences and that of main parts of the sentence, simultaneously.

In a sentence like Helen sighed (B. MACAULAY) there obviously are two main parts: Helen, which denotes the doer of the action and is called (grammatical) subject, and sighed, which denotes the action performed by the subject and is called (grammatical) predicate. Sentences having this basic structure, viz. я word (or phrase) to denote the doer of the action and another word (or phrase) to denote the action, are termed two-member sentences. However, there are sentences which do not contain two such separate parts; in these sentences there is only one main part: the other main part

¹ The Russian terms are, двусоставные and односоставные предложения.

190 The Sentence

is not there and it could not even be supplied, at least not without a violent change in the structure of the sentence. Examples of such sentences, which are accordingly termed one-member sentences, are the following: Fire! Come on! or the opening sentence of "An American Tragedy": Dusk — of a summer night. (DREISER)

There is no separate main part of the sentence, the grammatical subject, and no other separate main part, the grammatical predicate. Instead there is only one main part (fire, come on, and dusk, respectively). These, then, are one-member sentences.

It is a disputed point whether the main part of such a sentence should, or should not, be termed subject in some cases, and predicate, in others. This question has been raised with reference to the Russian language. Academician A. Shakhmatov held that the chief part of a one-member sentence was either the subject, or the predicate, as the case might be (for example, if that part was a finite verb, he termed it predicate). ¹ Academician V. Vinogradov, on the other hand, started on the assumption that grammatical subject and grammatical predicate were correlative notions and that the terms were meaningless outside their relation to each other.² Accordingly, he suggested that for one-member sentences, the terra "main part" should be used, without giving it any more specific name. Maybe this is rather a point of terminology than of actual grammatical theory. We will not investigate it any further, but content ourselves with naming the part in question the main part of one-member sentence, as proposed by V. Vinogradov.

One-member sentences should be kept apart from two-member sentences with either the subject or the predicate omitted, i. e. from elliptical sentences, which we will discuss in a following chapter.³ There are many difficulties in this field. As we have done more than once, we will carefully distinguish what has been proved and what remains a matter of opinion, depending to a great extent on the subjective views or inclinations of one scholar or another. Matters belonging to this latter category are numerous enough in the sphere of sentence study.

¹ А. А. Шахматов, Синтаксис русского языка, стр. 49—50. ²В. В. Виноградов, „Синтаксис русского языка" акад. А, А. Шахматова. Вопросы синтаксиса современного русского языка, 1950, стр. 108 сл. ³ See p. 252.

Chapter XXV

FUNCTIONAL SENTENCE PERSPECTIVE

In studying the structure of a sentence, we are faced with a problem which has been receiving ever greater attention in linguistic investigations of recent years. This is the problem of dividing a sen-tenсe into two sections, one of them containing that which is the starting point of the statement, and the other the new information for whose sake the sentence has been uttered or written. This has been termed "functional perspective". We will illustrate it by a pimple example. Let us take this sentence from a contemporary novel: I made the trip out here for curiosity, just to see where you were intending to go. (M. MITCHELL) Here the words I made the trip out here are the starting point, and the rest of the sentence (for... go) contains the new information. It cannot be said that every sentence must necessarily consist of two such sections. Some sentences (especially one-member sentences) cannot be divided up in this way, and doubts are also possible about some other types. However, most sentences do consist of these two sections and the relation between the syntactic structure of the sentence and its division into those two sections presents a linguistic problem deserving our attention.

Before we go on to study the problem it will be well to establish the terms which we will use to denote the sections of a sentence from this viewpoint.

There have been several pairs of terms proposed for this purpose, such as "psychological subject" and "psychological predicate", "lexical subject" and "lexical predicate", "semantic subject" and "semantic predicate", and others. Some of these are distinctly unacceptable, as they either suggest a wrong view of the phenomena in question, or are incompatible with our general principles for analysing language phenomena.

Thus, the terms "psychological subject" and "psychological predicate", proposed by the German scholar H. Paul, ¹ obviously will not do, as they introduce a notion of individual psychology, which lies beyond the sphere of linguistic investigation: the question we are discussing is not, what individual interpretation an individual reader or hearer may give to a sentence but what is objectively expressed in it, independently of a hearer's personal views or tastes.

The terms "lexical subject" and "lexical predicate", proposed by Prof. A. Smirnitsky,² will not do either, because they appear to take the whole problem out of the sphere of syntactic study and to include it into that of lexicology, which, however, has nothing to

¹ See H. Paul, Prinzipien der Sprachgeschichte, 5. Aufl., 1937, S. 124.

² See А. И. Смирницкий, Синтаксис английского языка, стр. 110.

192 Functional Sentence Perspective

do with it. We are not going to analyse the lexical meanings of individual words, which are treated in lexicology, but the function of a word or word group within a sentence expressing a certain thought; their function, that is, in expressing either what is already assumed or what is new in the sentence uttered.

We would rather avoid all terms built on the principle of combining the already existing terms "subject" and "predicate" with some limiting epithets, and use a pair of terms which have not yet been used to express any other kind of notion.

The pair of terms best suited for this purpose would seem to be "theme" and "rheme", which came into use lately, particularly in the works of several Czech linguists, who have specially studied the problem, notably with reference to the English language, both from the modern and from the historic viewpoint. Among the Czech scholars who have widely used these terms we should first of all mention Jan Firbas, who has developed a theory of his own on the historical development of the English language in this sphere.¹

The terms "theme" and "rheme" are both derived from Greek, and are parallel to each other. The term "theme" comes from the Greek root the- 'to set', or 'establish', and means 'that which is set or established'. The term "rheme" is derived from the root rhe -'to say', or 'tell', and means 'that which is said or told' (about that which was set or established beforehand). These terms are also convenient because adjectives are easily derived from them: "thematic" and "rhematic", respectively.

What, then, are the grammatical means in Modern English which can be used to characterise a word or word group as thematic, or as rhematic? We should note in passing, however, that it will hardly be possible to completely isolate the grammatical from the lexical means, and we shall have to discuss some phenomena which belong to lexicology rather than grammar, pointing out in each case that we are doing so.

The means of expressing a thematic or a rhematic quality of a word or phrase in a sentence to a great extent depend on the grammatical structure of the given language and must differ considerably, according to that structure.

Thus, in a language with a widely developed morphological system and free word order, word Order can be extensively used to show the difference between theme and rheme. For instance, word order plays an important part from this viewpoint in Russian. Without going into particulars, we may merely point out the difference between two such sentences as Старик вошел and Вошел

¹ See J. Firbas, Some Thoughts on the Function of Word-Order in Old English and Modern English. Sbornik prací filosofické fakulty brnenské university, 1959,

The Rheme 193

старик. In each case the word (or the part of the sentence) which comes last corresponds to the rheme, and the rest of the sentence to the theme. It is quite clear that no such variation would be possible in a corresponding English sentence. For instance we could not, in the sentence The old man came in, change the order of words so as to make the words the old man (the subject of the sentence) correspond to the rheme instead of to the theme. Such a word order would be impossible and we cannot make the words old man express the rheme without introducing further changes into the structure of the sentence.

In Modern English there are several ways of showing that a word or phrase corresponds either to the rheme or to the theme. We will consider the rheme first.

A method characteristically analytical and finding its parallel in French is the construction it is... that (also it is... who and it is... which) with the word or phrase representing the rheme enclosed between the words it is and the word that (who, which). Here are some examples of the construction: For it is the emotion that matters. (HUXLEY) Emotion is in this way shown to represent the rheme of the sentence. But it was sister Janet's house that he considered his home. (LINKLATER) Sister Janet's house represents the rheme.

In the following sentence the adverbial modifier of place, here, is thus made the rheme, and the sentence is further complicated by the addition of a concessive though- clause. It was here, though the place was shadeless and one breathed hot, dry perfume instead of air — it was here that Mr Scogan elected to sit. (HUXLEY) Without this special method of pointing out the rheme, it would be hardly possible to show that the emphasis should lie on the word here. In the variant Mr Scogan liked to sit here, though the place was shadeless and one breathed hot, dry perfume instead of air the emphasis would rather lie on the word liked: he liked it, though it was shadeless, etc.

Could it be, he mused, that the reliable witness he had prayed for when kneeling before the crippled saint, the mirror able to retain what it reflected like the one with the dark, gilded eagle spread above it before him now, were at fault in so far as they recorded all the facts when it was, after all, possibly something at another level that more crucially mattered?___(BUECHNER) The phrase emphasised by means of the it is... that construction is, of course, something at another level. The peculiarity of this example is that two parentheses, after all and possibly, come in within the frame of it is... that.

In the following example a phrase consisting of no less than eleven words is made into the rheme by means of the it is... that construction. It was his use of the highly colloquial or simply the

7 Б. A. Ильиш

194 Functional Sentence Perspective

ungrammatical expression that fascinated her in particular, for in neither case, clearly, did he speak in such a manner out of ignorance of the more elegant expression but, rather, by some design. (BUECHNER)As the that is far away from the is, it seems essential that nothing should intervene between them to confuse the construction, and, more especially, no other that should appear there.

The question of the grammatical characteristic of such sentences will be dealt with in Chapter XXXV (p. 276) and Chapter XXXVII (p. 302).

Another means of pointing out the rheme in a sentence is a particle (only, even, etc.) accompanying the word or phrase in question. Indeed a particle of this kind seems an almost infallible sign of the word or phrase being representative of the rheme, as in the sentence: Only the children, of whom there were not many, appeared aware and truly to belong to their surroundings, for the over-excited games they played, dashing in and out among the legs of their elders, trying to run up the escalator that moved only down, and the like, were after all special games that could be played nowhere but in the station by people who remembered that it was in the station they were. (BUECHNER) The particle only, belonging as it does to the subject of the sentence, the children, singles it out and shows it to represent the rheme of the sentence.

It goes without saying that every particle has its own lexical meaning, and, besides pointing out the rheme, also expresses a particular shade of meaning in the sentence. Thus, the sentences Only he came and Even he came are certainly not synonymous, though in both cases the subject he is shown to represent the rheme by a particle referring to it.

'Another means of indicating the rheme of a sentence may sometimes be the indefinite article. Whether this is a grammatical or lexical means is open to discussion. The answer will depend on the general view we take of the articles, a problem we have been considering in Chapter IV. Treating the article here in connection with functional sentence perspective is justified, as it does play a certain part in establishing the relations between the grammatical structure of a sentence and its functional perspective.

Owing to its basic meaning of "indefiniteness" the indefinite article will of course tend to signalise the new element in the sentence, that which represents the rheme. By opposition, the definite article will, in general, tend to point out that which is already known, that is, the theme. We will make our point clear by taking an example with the indefinite article, and putting the definite article in its place to see what consequence that change will produce in the functional sentence perspective.

Let us take this sentence: Suddenly the door opened and a little birdlike elderly woman in a neat grey skirt and coat seemed almost

The Indefinite Article and the Rheme 195

to hop into the room. (A. WILSON) The indefinite article before little birdlike elderly woman shows that this phrase is the centre of the sentence: we are told that when the door opened the person who appeared was a little birdlike elderly woman. This meaning is further strengthened by the second indefinite article, the one before neat grey skirt and coat. Since the woman herself is represented as a new element in the situation, obviously the same must be true of her clothes.

Now let us replace the first indefinite article by the definite. The text then will be Suddenly the door opened and the little bird-like elderly woman in a neat grey skirt and coat seemed almost to hop into the room. This would mean that the woman had been familiar in advance, and the news communicated in the sentence would be, that she almost hopped into the room. The indefinite article before neat grey skirt and coat would show that the information about her clothes is new, i. e. that she had not always been wearing that particular skirt and coat. This would still be a new bit of information but it would not be the centre of the sentence, because the predicate group seemed almost to hop into the room would still be more prominent than the group in a neat grey skirt and coat. Finally, if we replace the second indefinite article by the definite, too, we get the text Suddenly the door opened and the little birdlike elderly woman in the neat grey skirt and coat seemed almost to hop into the room. This would imply that both the elderly little woman with her birdlike look and her grey skirt and coat had been familiar before: she must have been wearing that skirt and coat always, or at least often enough for the people in the story and the reader to remember it. In this way the whole group the little birdlike elderly woman in the neat grey skirt and coat would be completely separated from the rheme-part of the sentence.

This experiment, which might of course be repeated with a number of other sentences, should be sufficient to show the relation between the indefinite article and the rheme, that is, functional sentence perspective.

There are also some means of showing that a word or phrase represents the theme in a sentence. Sometimes, as we have just seen, this may be achieved by using the definite article. Indeed the contrast between the two articles can be used for that purpose.

But there are other means of pointing out the theme as well. One of them, which includes both grammatical and lexical elements, is a loose parenthesis introduced by the prepositional phrase as for (or as to), while in the main body of the sentence there is bound to be a personal pronoun representing the noun which is the centre of the parenthetical as-for -phrase. This personal pronoun may perform different syntactical functions in the sentence but more often than not it will be the subject. A typical example of this sort of

196 Functional Sentence Perspective

construction is the following sentence: As for the others, great numbers of them moved past slowly or rapidly, singly or in groups, carrying bags and parcels, asking for directions, perusing timetables, searching for something familiar like the face of a friend or the name of a particular town cranked up in red and gold... (BUECHNER) After the theme of the sentence has been stated in the prepositional phrase as for the others, the subject of the sentence, great numbers of them, specifies the theme (pointing out the quantitative aspect of the others) and the rest of the sentence, long as it is, represents the rheme, telling, in some detail, whatever the others were busy doing at the time.

Sometimes a word or phrase may be placed in the same position, without as for: The manuscript so wonderfully found, so wonderfully accomplishing the morning's prediction, how was it to be accounted for? (J. AUSTEN) Here the first half of the sentence, from the beginning and up to the word prediction, represents the theme of the sentence, while the rest of it represents its rheme. The pronoun it of course replaces the long phrase representing the theme.

Here are a few more examples of the word or phrase representing the theme placed at the beginning of the sentence as a loose part of it, no matter what their syntactical function would have been if they had stood at their proper place within the sentence. That laughter — how well he knew il! (HUXLEY) There are two possible ways of interpreting the grammatical structure of this sentence. First let us take it as a simple sentence, which seems on the whole preferable. Then the phrase that laughter must be said to represent the theme of the sentence: it announces what the sentence is going to be about. In the body of the sentence itself it is replaced by the pronoun it, which of course is the object. Another possible view is that the sentence is an asyndetic composite one. In that case the phrase that laughter is a one-member exclamatory clause, and the rest of the sentence is another clause.

A somewhat similar case is the following, from the same author: His weaknesses, his absurdities — no one knew them better than he did. Just as in the preceding example, it seems preferable to view the sentence as a simple one, with the words his weaknesses, his absurdities representing the theme.

There are two more points to make concerning functional sentence perspective:

(1) The theme need not necessarily be something known in advance. In many sentences it is, in fact, something already familiar, as in some of our examples, especially with the definite article. However, that need not always be the case. There are sentences in which the theme, too, is something mentioned for the first time and yet it is not the centre of the predication. It is something about which a statement is to be made. The theme is here the starting

The Theme 197

point of the sentence, not its conclusion. This will be found to be the case, for example, in the following sentence: Jennie leaned forward and touched him on the knee (A. WILSON) which is the opening sentence of a short story. Nothing in this sentence can be already familiar, as nothing has preceded and the reader does not know either who Jennie is or who "he" is. What are we, then, to say about the theme and the rheme in this sentence? Apparently, there are two ways of dealing with this question. Either we will say that Jennie represents the theme and the rest of the sentence, leaned forward and touched him on the knee its rheme. Or else we will say that there is no theme at all here, that the whole of the sentence represents the rheme, or perhaps that the whole division into theme and rheme cannot be applied here. Though both views are plausible the first seems preferable. We will prefer to say that Jennie represents the theme, and emphasise that the theme in this case is not something already familiar but the starting point of the sentence.

The same may be said of most sentences opening a text. Let us for instance consider the opening sentence of E.