I'l. M. 169

2. Since there is no constant connection, no stable interdependence between words, phrases, sentences and the surrounding world, it is only natural that one and the same object may be called different names by different speakers and in different situations. To name an object, we must mentally place it in a suitable class, at the same time recalling the name of this class, the corresponding word or expression, or sentence. But every real object has an infinite number of characteristic features, some of which are objectively important, others are secondary, inconspicuous, unimportant for most people, but very essential for the speaker. Who can prevent one from mentioning a secondary feature if one wants to mention just this feature and hopes to be understood?

Now it must be clear why any object of speech can have innumerable denominations. The feature chosen by the speaker to name depends on his attitude to the object; it also depends on his particular communicative intention. Let us assume that the object of nomination is a certain man. The word man (or its stylistically specific counterparts, such as chap, guy, fellow, person, individual) can very well serve the purpose of identification: no listener will mistake the object of discourse for a house, a table, a car or anything else. But the speaker is free to use any other denominations that suit his purpose better. The same object can figure in speech as young gentleman, our next-door neighbour, my own darling, that greenhorn of yours, bloody blind bat (the last title could have been bestowed on a clumsy pedestrian by a furious driver).

All these denominations of our imaginary young man (as well as an unlimited number of others — my fellow countryman brother. Sergeant, her only son, blockhead, and so on) are not arbitrary, and not devoid of any connection (real or imaginary) with certain features, qualities, actions, etc. of the subject of speech. Words and expressions traditionally used with reference to a certain class of objects can be transferred and applied to a representative of quite a different class — yet this is always done in accordance with certain semantic laws. It is highly improbable that anyone would address the above-mentioned young man using the words traffic, yesterday, window-sill, minute, or snow.

Paradigmatic semasiology and onomasiology establish a classification showing semantic types of transfer of names and logical laws underlying them.

The problems of meaning (semasiology) and nomination (onomasiology) are essentially different from the branches of stylistics described in preceding chapters.

The first distinctive feature. Phonetics, morphology, lexicology, and syntax have clear-cut, formally limited fields of research. They only treat of phonemes (both segmental and suprasegmental), morphemes, words, and syntactic structures respectively. Semasiology, for its part, pays little or no attention to the differentiation of levels: semantically identical (or

similar) phenomena may occur in morphemes, words, phrases, sentences. Only phonemes do not concern semasiology, as they do not have extralingual meanings of their own.

The second distinctive feature. Students of the level-forming disci­plines (phonetics, morphology, lexicology, and syntax) are mostly inter­ested in the stylistic implication of units; they separate stylistically sig­nificant units from neutral ones, and attempt to find out to what sub­language the former belong. The meanings of elements and the sense of utterances are, so to speak, on the periphery of their attention. In opposition to them, onomasiology and semasiology specially deal with 'renamings', 'transfers of names', i.e. with whatever brings about a radi­cal change in the substance of the text.

All kinds of transfer of denominations (from a traditional object to a situational object) bear the name of tropes (from the Greek tropos 'turning')- This is the basic term of paradigmatic onomasiology, which studies only tropes and nothing else.

Every trope, as distinct from a usual, traditional, collectively accepted denomination of the object demonstrates a combination, a coincidence of two semantic planes (actually, of two different meanings) in one unit of form (one word, one phrase, one sentence). A trope, then, is a linguistic unit (word, phrase, sentence, paragraph, text) with two senses, both felt by language users. On hearing the exclamation Oh, you pig! (with reference to a person) the listener is aware of the traditional, original meaning of the word pig (the well-known domestic animal) and its actual reference which imparts an additional sense to the word — 'an untidy, greedy, or rude person'. The word acquires a new meaning, but its original meaning also remains (otherwise it would be senseless to use the word!).

Hence, the psychological essence of a trope is just the prominence given to two units of sense in one unit of form. Only the double meaning creates what is called an image: we observe a trope only if we see both meanings. If, however, the original meaning is obliterated, or at least no longer associated with the secondary one, there is no trope any more, although there may have been one when it was first created. There is no trope in leg of a table, neck of a bottle, foot of a hill, hand of a clock and the like. No one thinks of human legs, necks, feet, or hands when using these expressions. So they are a kind of 'etymological tropes' (metaphors), 'dead' tropes that are dealt with in lexicology, not in stylistics.

The use of tropes is, properly speaking, a false, erroneous qualifica­tion of an object. It is a device inconsistent with the primary reason for being (raison d'etre) of language, whose purpose and motto is 'calling a spade a spade'.

As already mentioned, tropes serve to create images that combine notions and as a result express something different from them both. A

i'

cardinal property of an image is its genetic and ontological independence from lingual expression. An image, as a psychic phenomenon, arises before and outside its verbalization: imaginative perception of analogies, connections, contrasts of reality, overestimation or undervaluation of its properties — all these acts of cognition can take place without language. Music, like the Japanese ikebana (art of flower arrangement) are metaphorical throughout. Metaphors (allegories and symbols) lie at the foundation of painting, sculpture, architecture. The principle of metonymy — a detail in the foreground, a detail instead of the whole — is typical of the cinema. Exaggeration and restraint (hyperbole and meiosis — see below) can be seen in dancing. It is just this indifference of tropes to the means and forms of their expression, to language in general, that necessitates searching for a purely logical classification of types of renaming.

The problem has been discussed for many centuries, the tradition going back to antiquity. Aristotle in his Poetics treats the figurative use of words, not yet differentiating metaphor and metonymy. The literature on tropes is immense, but the majority of scholars were not interested in presenting them as a generalized system. Rare exceptions include A. Bain, an English philologist of the nineteenth century, who gave a very satisfactory explanation of the psychological essence of certain tropes.¹ Most authors, however, either never attempt to solve the problem or propose purely subjective classifications (some of them describe tropes and other stylistic devices in alphabetical order²).

To achieve a more or less adequate treatment of the problem, it must be reduced to the simplest task possible. A very essential difficulty, which may have been taken into account, is the fact that stylistic terminology is the product of many epochs. Some of the terms denote very general, cardinal notions; others name particular phenomena that arise only if the material is viewed from a peculiar angle, sometimes inconsistent with the original aims of stylistic research.

One of these stylistic notions that do not match the system of paradigmatic onomasiology is the epithet. The term is used everywhere and, of course, has a perfect right to exist. But it is not a purely onomasiological term, nor a semasiological one either, since it has syn­tactic limitations: it is known that an epithet is an expressive attribute or adverbial modifier. No subject, object, or predicative can be an epi­thet. The notion, therefore, is a mixture, a hybrid — something partly semantic and partly syntactic. We cannot expect to find a place for the epithet among the tropes, for after all it is not a trope, although it may be metaphorical, metonymic, or ironical. To help the reader understand the situation better, we shall take a case that has nothing to do with linguistics, but presents the same logical problem. Let us assume that we

are to establish a general professional classification of all those working in an industrial enterprise. We might divide them into workers, technicians, engineers, clerks, and managers. This seems logical enough. But what happens with our classification if we enlarge it by only one additional class: tall, fair-haired, unmarried engineers and clerks? Such a group of people could, of course, be singled out (for instance, by the police or by an unmarried woman looking for a husband), but it is clear that such a group cannot be part of the strictly professional classification. The same with the epithet.

Similarly, the general system of tropes (which we are yet to discuss) cannot include specific varieties of tropes: personification, allegory are types of metaphor; synecdoche is the most primitive kind of metonymy; certainly connected with the latter are also symbol and periphrasis; litotes is a syntagmatic way of expressing meiosis. We shall return to these terms, but there is no place for them in the universal, generalizing scheme of renominations (transfer of names).

The multiplicity of concrete acts of renaming can be reduced to a strictly limited number of types.

Whenever we name an object or characterize a situation, we either follow the usual, collectively accepted, rules of naming, or deviate from them. If we are guided by the rules (saying what everyone would say), there is no transfer, there is nothing for stylistics to analyse in our speech act. If, however, we deviate from accepted standards, we can do it either quantitatively (1) or qualitatively (2).

1. What is a quantitative deviation? It is either saying too much,
overestimating the dimensions of the object (the intensity of its
properties), or else it is saying too little, undervaluing the size of the
thing, its importance, and so on.

Picture a situation. In answer to the question Have you got any money on you? the person addressed replies: Yes, I have three dollars. There is nothing in this reply for stylistic analysis (provided the man is telling the truth).

Another situation. The answer to the same question is: Oh, yes, lots! This time the stylist pricks up his ears: the owner of the three dollars is obviously exaggerating his wealth.

One more possibility. The reply is: Yes, just pennies though. Here, the speaker understates what he really has (three dollars, after all!).

The change from what ought to have been said to what really was said is in both cases purely quantitative in nature.

2. By qualitative differences between what is expected to be said and
what is actually said we mean a radical difference between the usual
meaning of a linguistic unit and its actual reference ('occasional
meaning'). The shout Hey you, green coat! You left your handbag does

not address a coat of green colour, but the woman who wears it. The angry remark A perfect ass, really! contains a noun very far in meaning from nouns denoting a human being. The word fine in the bitter statement A fine friend you are! means exactly the opposite.

One more remark. Since all stylistic devices are traditionally called figures of speech (although it is better to use the term only with reference to devices consisting of more than one element, for which see the next part of the book), we shall call tropes figures of replacement. For every trope is really a replacement: the language user discards the usual name of the object and replaces it with another.

Now we can discuss our classification. Figures of replacement (tropes) are first of all divided into two classes: figures of quantity and figures of quality.

The former consist of two opposite varieties: overstatement (hyper­bole), i.e. exaggeration, and understatement (meiosis), i.e. weakening.

The latter (figures of quality) comprise three types of renaming:

a) transfer based upon contiguity, upon a real connection between the
object of nomination and the object whose name is used; the corresponding
term is metonymy;

b) transfer by similarity (likeness, affinity) of the two objects (real
connection non-existent); the term is metaphor;

c) transfer by contrast: the two objects (actions, qualities) are di­
ametrically opposite; the term is irony.

This classification is visualized in the scheme below:

Fig. 7

Figures of quantity demonstrate the most primitive type of renam­ing. Their basis is inexactitude of measurement, disproportion of the object and its verbal evaluation.

Hyperbole. This trope — exaggeration of dimensions or other proper­ties of the object — is an expression of emotional evaluation of reality by

a speaker who is either unrestrained by ethical conventions or knows that exaggeration would be welcome. Quite naturally, the main sphere of use of hyperbole is colloquial speech, in which the form is hardly ever controlled and the emotion expressed directly, without any particular reserve.

Many colloquial hyperboles are stereotyped:

A thousand pardons. I've told you forty times. He was fright­ened to death. I'd give worlds for it. Haven't seen you for ages.

In colloquial speech, expressions of this kind are the natural outcome of uncontrolled emotions or just habit. In any case, the listener is seldom affected by a stale hyperbole: neither the listener, nor sometimes even the speaker notice the exaggerations; no one takes the words at their face value.

But it is the other way round in works of poetry or fiction, where exaggerations serve expressive purposes and achieve their aim: they are noticed and appreciated by the reader, though he also, as in the previous case, does not take them seriously.

An expressive hyperbole, as distinct from trite ones (used in everyday speech), is exaggeration on a big scale. There must be something illogical in it, something unreal, utterly impossible, contrary to common sense, and even stunning by its suddenness. True, the commonly used stock of phrases also comprises hyperboles, very strong because of their absurdity: see, e.g. the phrase in less than no time or the Russian без году неделя? But witty as they are, they are known to everybody, whereas individual creations strike us by virtue of their novelty.

One of the characters of a book by Derbridge " murmured such a dreadful oath that he would not dare to repeat it to himself". In the story, Mrs. McWilliams and the Lightning, Mark Twain thus creates the picture of general merriment:

" One after another those people lay down on the ground to laugh — and two of them died."

Having matter-of-factly mentioned this, the author pretends to be serious, going on: " One of the survivors remarked...".

It is evident that paradoxical, illogical hyperboles are employed for humoristic purposes. Here is an example from another famous American humorist:

" There I took out my pig... and gave him such a kick that he went out the other end of the alley, twenty feet ahead of his squeal." (O. Henry)

It must not be lost sight of that the lines quoted were written half a century before the era of supersonic jets!

Linguistic means of expressing exaggeration are varied. So, for in­stance, certain tautologies (pleonastic, overburdened structures

expressing one idea twice) are examples of hyperbole, as in the following instance:

" One does not know whether to admire them, or whether to say 'Silly fools'." (Christie)

Very often, however, hyperbole is combined with metaphor (see below in this chapter). The metaphor in such cases demonstrates a gigantic disproportion between what is characterized (named) and the characteristics given:

" And talk! She could talk the hind leg off a donkey! " (Peters) The famous American detective-story writer James Hadley Chase is so fond of this device that he uses practically the same hyperbole in his book Hit and Run two times:

" 'You can come down to the station and make a complaint if that's the way you feel about it, ' he said in a voice that could have peeled rust off the keel of a ship."

" One of them said in a voice that could have loosened a rusty nut off the propeller of a liner: 'Hey! You! Where do you think you are going? '"

Another hyperbole from the same book will suffice to show how immense the exaggeration can be:

'" Thank you for your help, sir.' This to Aitken. 'And for yours too.' His small eyes moved to me. Then, in a silence you could lean on, he plodded across the terrace, went down the steps to the waiting police car."

Meiosis, or understatement. This trope is the logical and psychological opposite of hyperbole. It is lessening, weakening, reducing the real characteristics of the object of speech. In other words, it is a device serving to underline the insignificance of what we speak about.

A typical meiosis is, for instance, the current expression It will cost you a pretty penny which in reality implies not a penny, but perhaps many pounds (or dollars), certainly a large sum of money.

Here follow two examples of meiosis taken from The Buyer from Cactus City by O. Henry. A New Yorker, owner of a big firm, talks condecendingly to the buyer mentioned in the title of the story:

" And what did you think of our little town? " asked Zizzbaum, with the fatuous smile of the Manhattanite.

His provincial visitor admits the achievements of New York, but tries hard to remain patriotic:

" You've got good water, but Cactus City is better lit up." His host shoots at him another sarcastic meiosis:

" We've got a few lights on Broadway, don't you think, Mr. Platt? "

One should be careful not to confuse meiosis with certain varieties of hyperbole, which is sometimes done by professional linguists who say that a cat-size pony (= a very small pony) or a drop of water (= not much water) are examples of meiosis.⁴ Their opinion cannot be accepted. It is meiosis only when the speaker understates normal or more than normal (e.g. big) things. That is what we see in a pretty penny, little town (of New York), a few lights on Broadway. When, however, the object spoken about is really small or insignificant, and the expression used to denote it strengthens and emphasizes its smallness and insignificance, we have a hyperbole, not a meiosis: he lives a stone's throw from here (he lives near indeed, but not as near as a stone's throw — cf. the Russian рукой подать); just a moment, please (very soon, but certainly not in a moment); the same in the expression before you could say Jack Robinson. Let us have one more example of hyperbole which could have been mistaken by other authors for meiosis, since the idea discussed is that of small amount:

" She sang listlessly as if she were bored with the whole thing, and the applause she collected could have been packed into a thimble, without overflowing." (Chase)

Viewing the matter psychologically, we can state that hyperbole ap­peals directly to the imagination, being itself a direct expression of emotional extravagance. The essence and the basis of meiosis is some­what more complicated and refined. Meiosis may also be regarded as a kind of strengthening through apparent weakening. The speaker confides in the intelligence of the addressee: the latter is expected to discern the speaker's intentional modesty, the obvious contrast between what he says and what he thinks, what he means to say.

Meiosis has no definite formal expression (on litotes see the next section); various linguistic means serve to express it: I was half afraid you had forgotten me. I kind of liked it. She writes rather too often. I am not quite too late.

A humorous effect is observed when meiotic devices (words and phrases called 'downtoners' — maybe, please, would you mind, etc.) co-occur with rough, offensive words in the same utterance: It isn't any of your business maybe. Would you mind getting the hell out of my way?

It is widely known that understatement (meiosis) is typical of the British manner of speech, in opposition to American English in which

hyperbole seems to prevail. G. McKnight remarks that the word rather is a 'super-superlative' in England (the reader must have noticed that McKnight's own word super-superlative is sheer hyperbole!). Whenever an English gentleman means to say You have amazed me, he merely asks: Really?; his lavishing praise of a thing is: Not so bad? Not at all so bad! Of course, this opinion about Britain is an overstatement: it characterizes the rules of behaviour of the upper classes only: besides, even an aristocrat uses as many overstatements as he likes. Therefore, the following story characterizes the general opinion rather than the real state of things.

An English girl and an American girl climb a steep mountain in the Alps. The English girl says: It's a bit exhausting, isn't it? The American echoes: Why, sure, it's terrific/!!

Litotes. This term denotes a specific form of meiosis, not an inde­pendent trope. Litotes is expressing an idea by means of negating the opposite idea. Thus if we intend to say with his assistance, we turn this into its opposite (by making the construction negative): without his assistance, and then we negate it again, saying: not without his assistance. What is the result? The result is double negation, and from mathematics we know that two minuses make a plus.

But our last statement is faultless only with reference to mathematics. In language, the result is indeed affirmative, but the meaning obtained is weakened: not without his assistance is weaker than with his assistance. That is why litotes produces a meiotic effect.

The negation may be doubled in different ways. The previous example turned the preposition with into its opposite without (which itself expresses a negative idea), and then we added the negative particle not. In the next example two negative affixes (prefixes) make a litotes within one word:

" Jeff is in the line of unillegal graft. He is not to be dreaded by widows and orphans; he is a reducer of surplusage." (O. Henry) The word unillegal aptly characterizes the " professional" activity,

the speciality of the swindler Jefferson Peters.

A variety of litotes employs instead of two negative elements the

negation of the antonym of the idea to be expressed. The effect of

" weakening" is the same. A highly current sample of this kind of litotes

is the word-combination not bad, in which bad (the antonym of. good), being

negated, results in something weaker than just good.

Litotes is very frequent in English — at any rate it seems to be used

more often than in Russian. Examples:

"... she was not unlike Morgiaiia in the 'Forty Thieves'." (Dickens) " And Captain Trevelyan was not overpleased about it." (Christie) " A chiselled, ruddy face completed the not-unhandsome picture."

(Pendelton)

Less obvious examples:

" You wouldn't exactly call Warley heavily industrialized." (Braine)

" His suit... had... that elasticity disciplined only by first-rate tailoring which isn't bought for very much under thirty guineas." (Braine)

In comparing the two opposite quantitative tropes — hyperbole (overstatement) and meiosis (understatement) — it might be mentioned here that they are not mutually exclusive: they often alternate in the same narrative. It looks as if saying too much urged the narrator to compensate the next time, that is to disclose less than expected. Sometimes, the ostentatious use of the two stylistic devices betrays a propensity to mannerism, to impressing the reader.

Illustrative of both habitual use of the tropes discussed and purposely, expressive ornamentality are detective novels by Raymond Chandler.

Thus, one of his female characters is condescendingly characterized in passing as having once been younger. In the same novel, a policeman assures the narrator (a private detective): " He [the policeman's superior] doesn't like you any more than we do."

The understatements quoted seem matter-of-fact, cursory, and almost inconspicuous. So are some of Chandler's overstatements, as in the following scrap of a telephone call from the police station:

" Marlowe? We'd like to see you here, in the office." " Right away? " " Or sooner."

But the writer's manner changes as he turns from the policemen's ready-made jokes to his imaginary narrator's personal experience. Chan­dler's exaggerations become uncommonly bold and striking with their suddenness:

" I was stunned like a dervish, weak like an old motor, defenceless like a beaver's belly, and as sure of success as a ballet dancer with a wooden leg."

The strongest of the four assimilations comes last, making the whole a climax (the meaning of the term is explained in the chapter on syntagmatic semasiology, p. 155).

See also one of Chandler's absurd hyperboles:

" His grey face was so long that he could wind it twice round his neck."

Mentioning in the same book (The Little Sister) a writing desk the size of a tennis court, Chandler goes on to admire the overflow of anything the rich can afford, letting the reader catch a mental glimpse of a multimillionaire's way of life. The narrator's fantasy takes a gigantic leap:

" The serviceman was bent under the load of drinks which he had to carry across the terrace to the swimming pool, the size of Lake Huron, only much less dirty."

At times, Chandler's manner of exaggerating becomes circumlocutory (i.e. periphrastic). One of his characters, consuming what they claim to be a New York steak in a third-rate snack-bar, wonders why New York should be mentioned at all. For, as he says, everybody knows: it is in Detroit, not New York, that tyres are manufactured.