GRAMMAR

Date: 2015-10-07; view: 525.

V. EXTRACTS FOR READING

THE PARTS OF SPEECH

IV. CREATIVE WRITING

1. Write a fairy-tale about parts of speech based on the problems in their system. Begin like this: "Once upon a time…".

2. Write a poem to introduce parts of speech to your school pupils. Make use of the example below.

A 'noun' is the name of anything,

As 'school', 'garden', 'boy' or 'thing'.

Adjectives tell the kind of noun,

As 'great', 'small', 'pretty', 'white' or 'brown'.

Instead of nouns the pronouns stand,

As 'he', 'she', 'us', 'your arms', 'my hand'.

Verbs tell us something being done:

'To read', 'count', 'laugh', 'carry', 'run'.

How things are done the adverbs tell,

As 'slowly', 'quickly', 'ill' or 'well'.

Conjunctions join the words together,

As 'men and women', 'wind and weather'.

The prepositions stand before

A noun, as 'by' or 'through the door'.

The interjection shows surprise,

As 'Oh! How pretty! Ah! How wise!'

Three little words you often see

Are articles 'a', 'an' and 'the'.

The whole are called the 'parts of speech',

Which reading, writing, speaking teach.

1.

F. PALMER

a. PARTS OF SPEECH

The traditional grammars often began with a statement of the 'parts of speech', which today would be called 'word classes'. According to most grammars there are eight parts of speech. They are (with typical examples):

noun (howling, wolf, flock, terror)

pronoun (I, you, he, which)

adjective (this, the, fourth, each, untidy)

verb (see, retire, laugh)

preposition (on, in, to)

conjunction (and, but, because)

adverb (much, deservedly, partly, merely)

interjection (alas)

This kind of classification goes back to the Greek philosophers Plato and Aristotle, though the first really clear statement comes from the most famous of Greek grammarians, Dionysius Thrax, who produced a grammar of Greek in about 100 B.C. Dionysius also recognized eight parts of speech. Six of them were identical with those listed above; the only difference in the more modern list is that it has distinguished nouns and adjectives and added the interjections, omitting the participle and the article as separate parts of speech. It is never very clear in the grammars why these eight parts of speech must be stressed; they do not appear to be an essential prerequisite for the rest of the grammar, but are introduced rather as an academic exercise. Like so much of what is to be found in these grammars they are there because of the classical tradition and because their justification was never seriously challenged. There are, however, quite serious objections to this traditional classification.

First, the definitions are largely notional and often extremely vague. It would in fact often be quite impossible to judge from these definitions whether a particular word was a noun, a verb or an adjective without knowing the answer already! We have discussed Nesfield's definition of a noun <…> and seen how totally circular it was. His definition of the verb is even worse because it is utterly uninformative – 'A verb is a word used for saying something about something else'. Do not most words say something about something else? His definitions of the pronoun and the adjective are a little better: 'A pronoun is a word used instead of a noun', 'An adjective is a word used to qualify a noun.' They are almost definitions in purely grammatical terms, as they should be, but they are still not precise enough. If we consider the pronoun we can see that many kinds of word may be used instead of the noun John:

John came this morning.

A man / Someone / You-know-who / The aforementioned came this morning.

In the definition of the adjective what does 'qualify' mean? Precede? Either precede or follow? Is John's an adjective in John's book? Is there an adjective in the people there? It is interesting to note that the definitions given by Dionysius Thrax almost two thousand years ago were formal, based largely upon morphology (though we may well question whether morphology ought to determine word classes). It was for this reason that he placed noun and adjective in the same class, because in Greek both have case endings.

Secondly, the number of parts of speech in the traditional grammars seems to be quite arbitrary. Why eight? Probably because Dionysius Thrax had eight. The adverb in particular is a most peculiar class. It is quite clearly a 'rag bag' or 'dustbin', the class into which words that do not seem to belong elsewhere are placed. This is easily illustrated by considering very and quickly, both of which are traditionally considered to be adverbs. They have almost nothing in common as shown by the following pairs of possible and impossible sentences (the latter marked with asterisks):

He ran away quickly.

*He ran away very.

He is very good.

*He is quickly good.

He has a very fast car.

*He has a quickly fast car.

We can overcome this problem if we wish by recognizing two different classes, one of 'adverbs' – quickly, beautifully etc. the other of 'intensifiers': very, fairly, quite, etc. But we are then faced with the problem of the status of other rather similar words like usually. Is this an adverb in the same (though restricted sense) in which quickly is? Again let us look at some sentences:

*He ran away usually (but He usually ran away).

He ran away quickly.

He is usually good.

*He is quickly good.

*He has a usually fast car.

*He has a quickly fast car.

It would appear, then, that we need far more classes than eight. The American scholar C. Fries suggested in his book The Structure of English that English had four parts of speech and fifteen groups of 'form words'. The four parts of speech he labelled classes 1, 2, 3 and 4, but they are clearly what would normally be called 'nouns', 'verbs', 'adjectives' and 'adverbs'. Fries was at pains to warn the reader against this identification of his classes with these traditional parts of speech, because he wanted to insist that his theoretical framework was different from that of the traditional grammars. We need not be too concerned about this; there is much to be said for retaining the old terms simply because it is so much easier to remember them, provided we are careful not to forget that we are using them differently. A particularly striking point about Fries's classification is that what traditional grammars call adverbs are treated in no less than five of his fifteen groups as well as providing the total membership of his class 4. The correspondences between the two systems of classification are:

F R I E S N E S F I E L D E X A M P L E

Class 4 Time now

Place here

Number once

Description slowly

Group C Affirming or denying not

Group D Quality, extent or degree very

Group H Place (introductory) there (is / are …)

Group I Interrogative when

Group L Affirming or denying no

Nesfield, moreover, divides his interrogative adverbs into those of time, place, number, description, quality or degree, and cause or reason. This should make us wonder first why the interrogative adverbs are 'form words' (Group I) while the corresponding non-interrogative ones are 'parts of speech' for Fries (compare, for instance, where and here), and secondly why Fries does not subclassify his Class 4. These 'adverbs' do not all have identical functions, apart from the obvious differences in their meaning. There are difficulties with Fries's classification, but at least he showed that the eight traditional parts of speech are not satisfactory.

It is, however, quite impossible to write a grammar of a language without setting up word classes. For not only will the grammar be stated largely in terms of the function of classes of words, but it is essential to indicate in the dictionary precisely how a word functions in the language, and this is done by indicating its class, e. g. that elephant is a noun and depend a verb. We do not need, however, to define each word class before we start. Rather, the definition of the word class is established by what is said about it in the grammar. In other words if you want to know what a noun is you will have to see what the grammar has to say about the functions of the noun (and if you want to know which words can be called nouns you must, of course, look to the dictionary). The functions may well be many and varied. The noun boy, for instance, has a different function in each of the following sentences, and the grammar must specify quite clearly what these functions are:

The boy has come.

I've seen the boy.

I gave it to the boy.

This is the boy's mother.

But by making quite clear that words such as boy have these (and other) functions, the grammar provides a definition of the noun. We do not, therefore, need the notional definitions even if we could use them. Similarly we do not give much weight to morphology. What is important about nouns is not that they have singular and plural forms (boy, boys), but that they have syntactic functions such as those illustrated above <…>.

The traditional classes constitute, like so many traditional grammatical points, a fairly reasonable interpretation of the English language. The two most important classes are noun and verb. Every complete sentence <…> must contain at least one word from each class. Thus a sentence such as Birds fly is the smallest possible. Apart from the -er and -est forms of adjectives, words of these two classes alone have morphological features, the noun with two forms and the verb with up to five (take, takes, took, taken, taking). The most important sub-classification of the noun is into countables and uncountables <…>, though the pronoun which has special features of its own <…> is also, perhaps, to be regarded as a sub-class of the noun, since its functions are largely the same. The most important sub-classification of the verb is into auxiliaries and full verbs <…>.

Another major class is the adjective with two main functions, attributive and predicative, as illustrated by the little boy and The boy is little respectively. But the traditional term 'adjective' includes words that are best regarded as members of a different class. The articles (the, a), possessive pronouns (his, her, my, their, etc), demonstratives (this, that, these, those), plus words such as all, some, neither, which precede the adjectives and most of which are never used predicatively, are treated today as 'determinatives' or 'determiners'.

The adverb we have already discussed. The preposition functions with a noun or rather a noun phrase <…> as in to John, on the table, with the resulting sequence of words functioning like an adverb of time or place. Finally, conjunctions are words that link sentences, though we should, perhaps, distinguish those that coordinate and those that subordinate <…>.

This does not account for all words of English. There are others such as not, who, yes, that have very special functions and belong either to a class of their own or to a class with very few members. It is hardly worth while treating these as word classes. The grammar will specify how they function and the dictionary can do no more than refer to the grammar for this specific item. Some scholars have, in fact, made a distinction between 'full' words and 'empty' words (or between 'lexical' and 'grammatical' words) to distinguish such words as these – and others such as the pronouns and, perhaps, the conjunctions or even the prepositions – from those classes such as noun and verb which contain a vast number of 'lexical' items. But there is no precise line of division between them.

Even on this approach, however, several problems arise. The first results from the fact that some word classes have a variety of functions, but, unfortunately, not all words of the class seem to have all the functions. This problem is well illustrated by the adjective. It will be remembered that we spoke of its attributive and predicative functions as in the little boy and The boy is little. But there are words that may be used predicatively but not attributively. Examples are well, glad and possibly ill as well as many words that begin with a-: afraid, asleep, alike, awake, abroad. We can say:

The boy was well / glad / abroad / afraid, etc.

but not

*the well /glad / abroad / afraid, etc. boy…

The status of ill is not wholly clear. Can we or can we not say the ill boy? Glad, moreover, has an attributive function but only in the 'fixed phrase' which is surely a linguistic 'fossil' – glad tidings. On the other side there are some words which may appear in the attributive position but not in the predicative position. Examples are main, mere, utter:

the main decision but not *The decision was main,

mere ignorance but not *The ignorance was mere,

utter darkness but not *The darkness was utter.

Note also that we can say:

the top shelf but not *The shelf was top,

but both:

the top boy and The boy was top.

There are some words which appear in both positions but do so with quite a different meaning. There is the old joke of the lady who wanted a painting of her late husband and the artist who replied that he did not mind waiting for him. We may also note, for a difference in meaning:

the right girl / The girl was right.

my old friend / My friend was old.

As we have seen, there is a similar problem with a heavy smoker which is not related to The smoker was heavy and similarly with a hard worker, a poor loser, a bad singer, etc. For many of these, in fact, it is quite clear that the related sentence is not of the type The smoker was heavy but He smokes heavily, i.e. that the attributive adjective is not in any way to be associated with a predicative adjective but with an adverb.

We shall have similar difficulties if we treat the so-called ordinal numerals – first, second, last, etc. – as adjectives. For we cannot relate:

The Frenchman was first /second / last with

the first / second / last Frenchman.

But, perhaps, we might not treat these as adjectives at all but as determinatives along with the, this, some, etc.

We might, because of this, be tempted to say that only predicative functions should be used as a true test of the adjective, or at least to regard this as the basic function. In fact, in the early form of transformational grammar <…> it was thought that the good boy should be derived from The boy is good. This should have the advantage that we could deal with the heavy smoker in a different way by deriving it from He smokes heavily, the attributive function thus being derived in different ways in English. But again there are difficulties. We can easily go too far. We might be tempted to say that in that case abroad and asleep were true adjectives, though of a special kind which in attributive position followed their noun instead of preceding it as in people abroad, children asleep. In view of the fact that most adjectives in French follow the noun this would seem quite plausible. It then might follow that here and there are also adjectives in view of:

The men are here / there.

The men here / there.

Equally we shall have to treat in the garden as an adjectival phrase in view of:

The children are in the garden.

the children in the garden.

Clearly we should be doubtful about this statement. In some recent linguistics works words such as abroad are treated as 'adjuncts' rather than as adjectives; semantically most of them are distinguished as place and time markers. But there are some words of this kind that may occupy the position before the noun as well as after it – upstairs, inside, etc.:

the upstairs room / the room upstairs

the inside wall / the wall inside.

Are these adjectives in the first example but adjuncts in the second?

There are other features that may be associated with adjectives and perhaps used as criteria for establishing this class. First, they may be preceded by words like very. This could be said to be quite an important criterion; we can clearly establish that interesting is an adjective while singing is not by using the very test:

The book is interesting / very interesting.

The bird is singing / *very singing.

The interesting / very interesting book.

The singing / *very singing bird.

The occurrence of very in very interesting is important because without it we might be tempted to say that in This book is interesting, is interesting is all part of the verb as in The bird is singing, though we would normally expect a verb such as interesting to have an object as in This book is interesting me. We find that many verbs of a similar type, frighten, shock, please, etc. function in just the same way – This book is frightening / shocking and also very frightening / shocking. We must say, then, that with these the -ing form is a true adjective – formed by derivation. But to return to our previous examples, we find we can say The boy is very glad / well / afraid but not <…> very abroad nor, except perhaps in joking form, <…> very asleep or very awake. This test does not, then, give us any clear answers. It would permit us to treat afraid as an adjective but to exclude abroad, although on the previous predicative / attributive tests these would seem to pass or fail together. It would also exclude upstairs, inside, etc., which have the maximum freedom of occurrence – before the noun, after the noun, and in the predicative position, yet would include top and bottom which are far more restricted:

the top shelf / the very top shelf,

but not

*The shelf was top.

Clearly, then, occurrence with very will not give us a clear decision whether a word is or is not an adjective.

A further possible criterion for the adjective is that it has 'comparative' and 'superlative' forms – nice, nicer, nicest and beautiful, more beautiful, most beautiful. It should be remembered first, however, that some of the words described as adverbs have similar forms fast, faster, fastest, and quickly, more quickly, most quickly. This test will exclude most of the 'doubtful' words – here, there, abroad, asleep, etc., downstairs, inside, etc., top and bottom, as well as main, mere, utter and in the attributive function right, late (but not old – an older friend). And it will permit us to treat heavy and bad as adjectives in heavy smoker and bad singer. But is this perhaps not too restrictive a criterion? We shall be left with only the really 'central' class of adjectives.

The difficulties we have been discussing arise from the fact that not all words that we want to put in a single class seem to have all the functions of that class. There is another problem that arises from what is almost the converse situation, that some words seem to belong to more than one class. Examples of this are steel and cotton in Steel is strong, a steel bridge and Cotton comes from Egypt and a cotton shirt. In the first of each pair the words would seem to be nouns, in the second they seem to be adjectives. Some decades ago there was a famous argument over the question whether these were adjectives or nouns being used as adjectives. One answer that is sometimes given is that we must distinguish between the word class and the function – that we here have the word class noun, and the function 'modifier' which is not to be identified with adjective. All we have to say then is that nouns may be used as modifiers as well as in their more usual function as 'heads' of noun phrases. Steel would be a modifier in steel bridge though a head in fine steel, with fine as a modifier. But this is largely a terminological way out; we could have decided to say that steel is an adjective as well as a noun. There is a further possibility, that it is an adjective being used as a noun. This may seem less plausible for steel but more plausible for poor in The poor are always with us or for blues in the blues. Are not poor and blue adjectives being used as nouns? There is no definite answer to this problem but the most plausible solution is that steel, cotton, etc. are nouns, but 'nouns of material', and that such nouns may be used in an adjectival function to mean 'made of …' and that similarly adjectives like poor, wicked, good, etc. (in fact all adjectives that can be followed by people), may be used with the function of nouns in the poor, etc. The colour adjectives may be used in a similar way with the additional feature that they then have both singular and plural forms, since we have the blue and the blues, but only the poor and not the poors. We have to make a decision, then, to place these words in the class of noun or adjective but to sub-classify to show that they sometimes have functions of the other class. The decision to treat them all as 'basically' nouns or adjectives is one that can certainly be justified in a practical way. It is simpler to list steel, cotton, etc., as a sub-class of nouns with adjective-like function rather than list them twice, first as nouns and then as adjectives.

While we are considering this function of nouns acting as adjectives we ought to consider the status of the first words in bus stop, shoe polish, bread shop, etc. The point here is that in English any noun can be placed before another noun with a kind of prepositional meaning – 'a stop for buses', 'polish for shoes'. We can extend the list of nouns so placed together almost ad infinitum bus stop girl, bus stop girl inquiry, bus stop girl inquiry row, etc., though these are more typical of newspaper headlines than of ordinary English. But here we do not need to talk about adjectival function, because this is a function of all nouns. Complexes of this kind are a feature of grammar. The noun-noun sequence is, perhaps, a particularly English feature, but raises no new problems of description. Note, in particular, the adjectival function of steel in steel bridge and the purely nominal complex of steel company. The stress marks this distinction quite clearly since in the latter the main stress falls on steel (compare also silver in silver box to mean either 'box made of silver' or 'box for silver').

There are other words that we would wish to consider as belonging to several different classes. Love and work in English seem to be both noun and verb, safe and choice to be both adjective and noun. Round seems to be a noun, an adjective, a verb, a preposition and an adverb – a round of toast, a round ball, to round the Horn, round the mulberry bush and make it go round! We could of course, if we wished, say that these belong to the rather special classes of noun-verb, adjective-noun, noun-adjective-verb-preposition-adverb. This would certainly save us listing them more than once in each class. This would certainly save us listing them more than once in each class. This solution was quite seriously suggested by Hockett who recognized seven major classes – N, A, V, NA, NV, AV and NAV – plus an eighth, particles. But in practice this is a pointless difference. We shall never need to list all the members of each class. Rather we shall list (in the dictionary) all the words of the language and nothing at all is gained by stating that work belongs to the class NV rather than that it belongs to the classes N and V.

A final problem involves us in, again, consideration of morphology. It has been assumed throughout this section that word classes are classes of words in the sense in which C A T is a word, not cat and cats. It is for this reason that some scholars (beginning with Dionysius Thrax) have used morphological criteria for setting up word classes. Hockett, for instance, divided the words of Nootka (North American Indian) into inflected and uninflected, a purely morphological division. But it is obvious that the main purpose of setting up word classes is to state how they function in the language, to class the words in terms of their syntactic function, and this we have been trying to do in this section. The status of steel, cotton, blue, poor was very largely discussed in terms of the places in which they occur (though we also noted the morphological point of blue / blues). But a problem arises from the fact that some of the forms of the same word have quite different syntactic functions, functions that relate to different word classes. 'Participles' and 'verbal nouns', for instance, such as singing, playing in English, or amans, amaturus, amandum, amatus, amare in Latin, seem to be forms of verbs, but to have the function of nouns or adjectives. Do we then say they are not inflected forms, but derived forms like singer? Are they nouns or adjectives derived from verbs, since derivation often involves change of word class? The answer must surely be 'no'. Just as we can class steel as a noun and permit it to have adjective-like function, so we must write the grammar in such a way that certain forms of verbs (not all the forms of the whole verb – and this is why we have a different situation from that of steel) may function as nouns or adjectives. There is a rather striking situation in the Cushitic (Ethiopian) language Bilin. The genitive ('possessive') forms of the noun function like adjectives and have all the morphological features of the adjective. For instance, the word for man is gerwa and the word for man's is gerwix^w but gerwix^w is not only the genitive of gerwa but is also an adjective in the nominative case. If we want to say to the man's mother, not only does the word for mother have to be placed in the dative case, but also the word for the man's, since it is an adjective; the form is gerwix^w∂d. In this language forms of nouns not only function like adjectives syntactically, but also regularly have the morphology of the adjective, as well as retaining the morphology of the noun! And the same is true of verb forms which have adjectival function. The Bilin system is very neat and there is no difficulty, provided we do not ask whether its forms are 'really' verbs and nouns or 'really' adjectives. They belong to the class of verb or noun because that is how we must mark them in the dictionary, but our grammar will show that both syntactically and morphologically they have the function of adjectives.