The relation between relatively specific and absolutely specific units.

The reader must have noticed by now that style is what differentiates a given sublanguage from all other sublanguages. In other words, style is formed by absolutely specific units.

Thus one understands that, for instance, 'scientific prose style' (I. Galperin), 'colloquial style' (I. Arnold), and, let us say, 'tele­graphic style' are particular features: sets of words, word combinations, sentence patterns and text structures to be met with in scientific texts, colloquial texts, and telegrams respectively. Slylistics, therefore, is interested, practically speaking, only in specific units of sublanguages.

But if so, one may ask, what function is performed by relatively specific units? Here we must return to our recent statement: the number of sublanguages is indefinite; they may be as many or, again, as few, as we think advisable.

Nothing, for instance, stands in the way of differentiating only two sublanguages in English. We shall call one of them 'the sublanguage of official intercourse', the other being that of 'informal intercourse'. In this case we obtain two intersecting ellipses as shown below (Fig. 2).

Fig. 2

Each of the two ellipses comprises an infinite number of ellipses of narrower scope representing more particular sublanguages. Thus the sublanguage of formal intercourse will embrace, for example, the sub­language of " straight new" stories, the sublanguage of commercial cor­respondence, the sublanguage of technical instructions, of science, of diplomacy, etc. The sublanguage of unofficial communication will in­clude the colloquial sublanguage, the sublanguage of low colloquial in­tercourse jargons, dialects. Fig. 3 shows this subdivision of the two sublanguages.

Fig. 3

It stands to reason that, firstly, the scope of the non-specific sphere is by no means constant, but changes in accordance with the number of the sublanguages we single out for research. Compare Figs. 1 and 2 to see that in the second instance the overwhelming majority of linguistic units are non-specific. It follows, secondly, that what was relatively specific in Fig. 1 becomes either non-specific or absolutely specific when the number

of sublanguages is reduced. In Fig. 2 there is no space whatever for relatively specific units: they have become absolutely specific.

One example will suffice to understand the principle of relativity in stylistic classifications. The words individual, person, man, chap, guy are more or less synonymous, as all of them imply the meaning 'a human being' (of male sex — except the word person used with reference to either sex). It is obvious then that only the word man is absolutely non-specific. Chap and person, in their turn, are also very widely used English words, but of course the former is a little 'lower', and the latter 'higher', than man. The same may be said about the words guy and individual, respectively. Now, if we intend to establish several sublanguages to sort out every unit where it really belongs, the words person and chap will get somewhere in the intersection of the ellipses, i.e. in the relatively specific spheres of the high-flown and low colloquial sublanguages, whereas the words individual and guy will occupy the extreme, non-intersecting periphery of their respective ellipses.

But the picture changes as soon as we have decided to differentiate only two sublanguages — those of official and unofficial intercourse. Here, only the highest and the lowest words — individual and guy ac­quire (or preserve) the status of absolutely specific units; the rest of them (person, man, chap) become non-specific.

Note. Care should be taken to understand that the schemes illustrate only the theoretical principle, not the actual stylistic structure of language. The figures are merely to help visualize the interrelations (coincidences and individualities) of sublanguages. Language as it is cannot be presumed to consist of regularly identical ellipses. A sublanguage may happen to have certain units in common not with its geometrical neighbour, but with a remote one. Easily imaginable and obviously true to life is a sublanguage of an extremely limited sphere of use. Its schematic representation would include fragments of the neutral, relatively specific, or absolutely specific parts of the circle; its form could be quite different from that of an ellipse; it might be a figure of an irregular shape, occupying some space inside the circle, far from its boundaries or its centre. The reader is requested to picture the sublanguage of, say, a streetcar conductor, who uses only professional formulas and gives some information to occasional passengers concerning the locality.

Sublanguages¹⁷ and styles. Now we can approach the problem of defining the notion of style. It has been several times hinted above that style is what differentiates one text or one homogeneous group of texts from other texts (or other groups). To be more exact, we deal here not with texts, but with sublanguages underlying them. Since we also know what is non-specific (common to all, devoid of characterizing function)

and what is specific (particular, characterizing) we can easily understand the most general linguistic definition of style. It reads: Style is specificity of sublanguage.

This definition appears unexpectedly short, and the reader accustomed to elaborate, heavy-going, all-inclusive definitions could perhaps doubt its validity.

Indeed, habitual discourses on style have much more complicated wording. Suffice it to inspect the classical definition of style, the one belonging to V. V. Vinogradov. This scholar treats style as " socially cog­nized and functionally conditioned internally united totality of the ways of using, selecting and combining the means of lingual intercourse in the sphere of one national language or another, a totality corresponding to other analogous ways of expression that serve different purposes, perform different functions in the social communicative practice of the given nation".¹⁸

This definition is not only overburdened with such self-evident characteristics as " lingual intercourse in the sphere of one national language or another" or " in the social communicative practice of the given nation", but it leaves, besides, a number of problems without an answer.

First of all, is the social cognizance of style really its distinctive feature? In other words, is style only what is acknowledged by society as a special, separate style? If it be so, do linguists' discussions on the number of styles in language influence the final result in any way? Further: what is to be done in view of the fact that a considerable number of native speakers does not " cognize" or at least cannot give a comprehensive definition of this or that style? What percentage of the population should be aware of the existence of a certain style to make it socially acknowledged as a legitimate style? And, finally, was Pushkin's style a socially cognized style before it was researched and masterfully described by V.V. Vinogradov in his book, The Style of Pushkin?

Vinogradov's definition of style, like many other definitions, identi­fies style with the totality of characteristics of the lingual form without differentiating relevant and irrelevant features of that form. Properly speaking, the definition makes no attempt at differentiating what char­acterizes the given type of speech from what is common to several types or even to every type of speech, to every sublanguage.

Following Vinogradov's definition, we should be obliged to describe all the features of texts investigated, irrespective of their importance, which is practically never done in stylistic descriptions. Analysing, for instance, the sentence / ain't never done nothing no stylist would mention the fact that its subject is the pronoun of the first person singular, or

that the direct object is expressed by the word nothing. Although these statements are correct, they do not disclose any particularities of the sentence analysed. Instead, he will certainly single out the specific features of it, namely: the use of the 'ungrammatical' (non-standard) form ain't, which stands here for haven't; and the triple negation (ain't, never, nothing), considered inadmissible by grammarians (most normal negative sentences must contain one negation: / haven't ever done anything). Only the form ain't and the triple negation are specific features putting the sentence to low colloquial, uncultivated speech class.

Stylistic colouring and stylistic neutrality. Since style is the specificity of a sublanguage, it is self-evident that non-specific units of it do not participate in the formation of its style. Units belonging to all the sublanguages are stylistically neutral. Thus we observe an opposition of stylistically coloured specific elements to stylistically neutral non-specific elements.

The essence of stylistic perception consists in mental confrontation of what one hears (or reads) with one's previous linguistic experience: to understand a verbal message means not only to decipher the sense of each linguistic unit (and the sense of their combination), but at the same time to evaluate the units and the total from the viewpoint of their appurtenance to either the neutral (non-specific) sphere of language or to the stylistically coloured sphere. The stylistic colouring, in its turn, is nothing but the knowledge where, in what particular type of communication, the unit in question is current. On hearing, for instance, the above-cited utterance, / ain't never done nothing, we compare it with what we know about standard and non-standard forms of English, and this will permit us to pass judgement on what we have heard or read.

We can further state that stylistic colouring, as well as stylistic neutrality of linguistic units is the result of their distributional capaci­ties. The term distribution, widely used in the second half of the twen­tieth century, implies the possibilities of combining the given unit with its immediate environment. Distribution is the totality of environments of the unit. See, by way of illustration, the forms haven't and ain't. The former has nearly universal distribution: it can be used in all types of oral communication, except the official ones, in which the form have not is preferable. The distribution of the latter (ain't) is confined to subcolloq.uial, uncultivated types of speech; it was shown above in the sentence with several negations.

What the layman's experience shows him is the distributional potential of the unit, in the linguist's conception it is the place occupied by the unit in the system of sublanguages.

To help visualize our discussion of stylistic colouring and stylistic neutrality, we could say that stylistically coloured units (bookish, solemn,

poetic, official or, on the contrary, colloquial, rustic, dialectal, vulgar) have each something like a label on them — some " inscription", a kind of " trade-mark" showing where the unit was manufactured, where it generally belongs, and in what collocations it is proper or improper. Hence, we can say, stylistically coloured units are definitely characterized.

Coming, however, to inspect the units commonly called neutral, what can we state concerning their " trade-marks"? Do they really have none? Do they only denote without connoting? The answer will be in the negative. Like any other unit, a neutral one is bound to possess connotations in its semantic structure. But its connotations (its " labels", or " trade-marks") are manifold, are, in fact, innumerable. No one can tell how many labels showing the sphere of its currency the word water has. We have met this word in thousands of combinations, in various spheres of intercourse. Therefore, its connotations being numerous and varied (sometimes even opposite in the stylistic class they belong to), the general result is their mutual annihilation.¹⁹ The resultant connotation is indefinite, i.e. neutral.

Neutrality and norm. Quite a number of prominent scholars abroad, as well as in this country, along with other definitions of style, come to the conclusion that style may also be defined as deviations from the lingual norm. In their opinion, what is stylistically conspicuous, stylistically relevant, stylistically coloured is a departure from the norm of the given national language. Such or similar statements may be found in the works of Michael Riffaterre, E. Saporta, M. Halliday. Needless to say, they all substitute the word norm for the word neutrality.

But perhaps both norm and neutrality are words of the same meaning? Perhaps the statement, 'what is stylistically coloured is a departure from the norm', makes sense and should be accepted?

To answer this question, we ought to know the exact meaning of the word norm. Obviously the notion of norm implies pre-established and conventionally accepted parameters (i.e. characteristics) of what is evaluated. Let us again have recourse to extralinguistic analogies. What do we call the norm of bread rationing, say, in times of war? It is only the weight of the ration and its quality (established sort of bread), but never the real, concrete portion of bread a soldier has received. What is the qualification norm in sports? Of course, not the action of the sportsman corresponding to the requirements, but their abstract characteristics: the number of kilograms for weight-lifters, the number of centimetres for jumpers, the number of seconds for sprinters.

The same is true with regard to linguistics. The sentence / haven't ever done anything (see above) is not the norm itself, but conforms to the literary norm, being the realization of the latter. And what shall we say about the sentence / ain't never done nothing? To be sure, it deviates from the literary norm (from standard English), but it fully conforms to the

requirements of the uncultivated part of the English-speaking population: they merely have their own conception of norm!

There are as many norms as there are sublanguages. Each sublanguage is subject to its own norm. To reject this statement would mean admitting abnormality of everything that is not neutral. If style were departure from norm, in this case only ABC-books or the texts of the first lessons of English handbooks for foreigners would be considered " normal". Everything else, anything that manifests peculiarities of whatever kind, would have to be condemned as " abnormal". Shakespeare, Dickens, Galsworthy, O. Henry, Dreiser, scientific and technical texts, announcements and advertisements, orations, headlines, telegrams and everyday speech — all this would be for the most part " abnormal" if we were to believe M. Riffaterre and his colleagues.²⁰

This is absurd, of course. One should not confuse what is neutral with what is normal. The characteristic feature of norm in language is its plurality. There never has been one single norm for all.²¹

Borderlines or borderlands of sublanguages. The sublanguages rep­resented above by clear-cut ellipses inscribed in circles show merely the general principle of the relationship between areas occupied by absolutely specific (style-forming), relatively specific, and non-specific (neutral) linguistic units. This has been done to facilitate visualization. The borderlines between sublanguages are only theoretically assumed to be clear-cut. Linguistic reality is much more complicated than its description.

First of all, there exist no objective criteria for classifying units that fit into more than one class. The greatest difficulties arise, of course, in the sphere of phonetics: we can never be sure whether a slight change in the quality of a vowel or a consonant is an individual peculiarity of a native speaker or whether it is as far from normal as to be a foreigner's accent, a mispronunciation of speech sounds. The problem seems easier with morphemes, words, and word combinations, because they are discrete units: no one could mistake one unit for another. And yet this circumstance does not make the problem easier either. Morphemes, words, word combinations, and sentences are indeed individual and discrete, but their stylistic quality is by no means as definite as we have seen above, where examples were specially selected to illustrate stylistic identities, differences, and contrasts. There is certainly no doubt as to stylistic colouring when we oppose the bookish morphemes sub-, super-, ultra- to the neutral ones: un~, re-, less, -ful. The same is valid with regard to units of higher ranks. We remember that the word go is neutral, chap and daddy are colloquial, hereof or whereupon are unmistakably official and so on. It is, further, a well-known fact that the use of the so-called Nominative Absolute (My brother coming home, we sat down to dinner) is confined to written

forms of speech; the construction is also employed in official forms of oral communication, but sounds too pretentious in colloquial speech (just as the Russian причастные and деепричастные обороты). We can be quite sure about some sentence forms:

He will certainly never become a good writer, (neutral) For him to become a novelist of note is sheer impossibility. (bookish)

Good writer, he? Not likely! (colloquial)

But perhaps more often convenient to corroborate the theory accepted ■ here do we observe cases when one is not sure where the unit in question belongs. Are the words swell (in the sense of 'very good'), corking (of the same meaning) colloquial or subcolloquial, i.e. belonging to slang? Are the words and expressions smeller ('nose') or old bean only subcolloquial (low colloquial) or vulgar? The answer certainly depends on the viewpoint of the language user. Similarly, shall we refer words like democracy or constitution to the superneutral (literary, slightly bookish) or to the neutral layer? Of course, they are hardly ever used by little children, but is this fact any criterion at all? Every grown-up person who speaks English knows them.

Our task becomes especially difficult if we take into account the in­evitable divergence of opinion. Practically every language user has his own favourites and pet peeves in the world of words and expressions. Every speaker (and hearer) passes his own j udgement on them. What one considers to be neutral another takes for stylistically coloured — some people place the linguistic units higher or lower than do others.

Therefore we may come to the conclusion that there are no strict borderlines between sublanguages as well as within them (between ab­solutely specific, relatively specific, and neutral spheres). What was pre­sented above as strict borderlines is more likely to be 'borderlands', that is, 'strips of uncertainty', or perhaps we had better name them 'zones of tolerance' (or 'tolerance zones'). The term implies that the units of such a zone are just tolerable in both neighbouring spheres: using one of them, we may hope not to depart from the norm of the sphere in question.

Fig. 4 shows the borderlines between the sublanguages and their spheres as tolerance zones: strips of no definite width.

The reader will have noticed that the borders of the circle itself (i.e. of the national language as a whole) are also indefinite, unclear. Why? Because we never know for sure where the given national language ends and another language begins; it is impossible sometimes to say whether a borrowed word, a borrowed word combination, a construction has already become part and parcel of this language (although it is still felt as foreign), or whether it remains an often used quotation from a foreign language. How about the now world-famous Russian words perestroiha and glasnost? Are they brand-new English words of Russian coinage or merely exotic lexical units denoting phenomena of great social importance for this country? What should we say as to the status of such words in Russian as дисплей, дизайн, брифинг (display, design, briefing)! Have they become Russian words?

Another circumstance which renders the stylistic status of certain units indefinite is the fact that they change their stylistic qualities with the lapse of time. There always are units in actual use which to some of us seem obsolescent or utterly obsolete, yet from the viewpoint of others (mostly of those of the older generation) they appear quite normally bookish or even neutral. On the other hand, the process of coining and borrowing new linguistic units (especially words and expressions) is always going on. A number of innovations acquire the " rights of citizenship" imperceptibly. The majority of newcomers, however, look ostentatious, unwonted or even monstrous, and one can never tell when the moment might come for them to turn into habitual, usual, normal words and expressions.

This much will probably suffice to understand why borders separating sublanguages are not borderlines, but rather borderlands: tolerance

zones.