Structural and Transformational Grammars 3 страница

¹ L. Вloomfield. Op. cit., p. 190. See also: О. С. Ахманова и Г. Б. Микаэлян. Современные синтаксические теории. М., 963, pp. 22—23.

a given environment (such as man = good boy) was substitution repeated time and again ¹. Distributional analysis and substitution were not something quite novel in English grammatical theory. Occurrence of an element relative to other elements, now generally referred to as "distribution", has been involved in almost every grammatical statement since Antiquity ². But the difference between the traditional and structural approaches consists in that the former did not rely upon this method as part of an explicitly formulated theory, whereas modern linguistics has given recognition, within the theory of grammar, to the distributional principle, by which traditional grammarians were always guided in practice. The same is true of substitution. This is an entirely-formal method for discourse analysis arranged in the form of the successive procedures.

Starting with the utterances which occur in a single language community at a single time, these procedures determine what may be regarded as identical in various parts of various utterances. And this is supposed to provide a method for identifying all the utterances as relatively few stated arrangements of relatively few stated elements.

Z. S. Harris, E. A. Nida and other American linguists of Bloomfieldian school concentrate their attention on formal operations to discover and describe the features and arrangement of two fundamental linguistic units: the phoneme and the morpheme as the minimal unit of grammatical structure. Like Bloomfield, they attach major importance to spoken language laying emphasis on the fact that writing is a secondary visual representation of speech.

Language came to be viewed not as an aggregate of discrete elements but as an organised totality, a Gestalt which has a pattern of its own and whose components are interdependent and derive their significance from the system as a whole. In F. Saussure's words, language is like a game of chess", you cannot add, remove or displace any element without effecting the entire field of force.

Z. Harris presents methods of research used in descriptive, or, more exactly, structural, linguistics. It is, in fact, a discussion of the operations which the linguist may carry out in the course of his investigations, rather than a theory of the structural analysis which results from these investigations.

P. Roberts and W. N. Francis, following Ch. Fries and H. A. Gleason, are to a large degree concerned with studying patterns of organisation, or structures. They hold the view that linguistics, like physics and chemistry or, say, geology or astronomy, must be preoccupied with structure.

Returning to the traditional names of parts of speech P. Roberts and W. N. Francis establish four major classes of words and several groups of function-words, proceeding from the criteria of distribution

¹ See: Z. S. Harris. Methods in Structural Linguistics. Chicago, 1961, pp. 15—16.

² See: P. Diderichsen. The Importance of Distribution Versus Other Criteria in Linguistic Analysis. Copenhagen, 1966, pp. 270—271; see also: L. L. Iоfik, L. P. С h а k h о у a n. Readings in the Theory of English Grammar. L. 1972, p. 37.

of words, the morphological characteristics of words and their correlation.

The analysis of English structure made by P. Roberts and W. Francis presents a major linguistic interest as a significant contribution to modern linguistic thought.

It is to be noted, however, that some of their statements are devoid of logical consistency.

The classification of words into parts of speech given in these books is open to doubt and questioning because in identifying the linguistic status of words P. Roberts and W. N. Francis, like Ch. Fries, proceed from essentially different criteria: the major classes of words are classified in terms of their formal features and function words — in terms of meaning.

What seems also erroneous and devoid of logical foundations is excluding meaning from this sphere of linguistic analysis.

According to W. N. Francis, there are five devices which English speakers make use of to build words into larger organised combinations or structures. From the listener's point of view, these five are the kinds of signals which reveal the patterns of structural meaning in which words are arranged. As a summary of his assumptions, W. N. Francis lists them describing briefly as follows:

1. Word Order as the linear or time sequence in which words appear in an utterance.

2. Prosody, i. e. the over-all musical pattern of stress, pitch and juncture in which the words of an utterance are spoken.

3. Function words or words largely devoid of lexical meaning which are used to indicate various functional relationships among the lexical words of an utterance.

4. Inflections, i. e. morphemic changes — the addition of suffixes and concomitant morphophonemic adjustments — which adapt words to perform certain structural functions without changing their lexical meaning.

5. Derivational contrast between words which have the same base but differ in the number and nature of their derivational affixes ¹.

The classes of words established by P. Roberts and W. N. Francis do not coincide.

In W. N. Francis' classification there are four parts of speech: Noun, Verb, Adjective and Adverb. Pronouns are treated as two subclasses of nouns, called pronouns and function nouns. The group of pronouns comprises eight words whose importance far outweighs their number. These are: I, we, you, he, she, it, they and who.

The main groups of function-nouns are eight in number (including some stereotyped phrases) plus some unclassified ones (not all the following lists are complete):

a) Noun-determiners: the, a/an, my, your, her, their, our, this/ these, that/those, its, one, two... ninety-nine, many (a), more, several, both, all, some, no, every, (a) few, other.

¹ See: W. N. Francis. The Structure American English. New York, 1958, p. 234.

b) Auxiliaries: can/could, may/might, will/would, shall/should, must, dare, need, do, had better, be, get, have, keep (on), used, be going.

c) Qualifiers: very, quite, rather, pretty, mighty, somewhat, too, a bit, a little, so more, most, less, least, indeed, enough (real, awful, that, some, right, plenty), no, still, much, lots, a (whole) lot, a (good, great) deal, even.

d) Prepositions:

(1) Simple: after, among, around, before, concerning, etc.

(2) Compound: along with, away from, back of, due to, together with, etc.

(3) Phrasal: by means of, in front of, on account of, etc.

(e) Coordinators: and, not, but, nor, rather, than, either... or, etc.

(f) Interrogators:

(1) Simple: when, where, how, why (whence, whither), whenever, etc.

(2) Interrogative pronouns: who, which, what, whoever, whichever, whatever.

(g) lncluders:

(1) Simple: after, although, how, lest, since, etc.

(2) Relative pronouns: who, which, that, when, where, whoever, etc. (h) Sentence-linkers:

(1) Simple: consequently, furthermore, hence, however, moreover, nevertheless, therefore.

(2) Phrasal: at least, in addition, in fact, etc.

There are also function verbs in Francis' classification which stand in place of a full verb-phrase, when the full verb has been expressly stated or strongly implied in the immediate linguistic context or the non-linguistic context.

We cannot fail to see that applying formal structural methods of analysis which seem to be more objective than semantic criteria, grammarians come to somewhat different results.

In terms of N. Chomsky's theory of syntax, sentences have a surface structure and a deep structure, the latter is more complicated, being based on one or more underlying abstract simple structures.

In certain very simple sentences the difference between the surface structure and the deep structure is minimal. Sentences of this kind (simple, active, declarative, indicative) are designated as kernel sentences. They can be adequately described by phrase or constituent structure methods, as consisting of noun and verb phrases (the so-called P-markers, the NP's and VP's). According to syntactic structures, kernel sentences are produced by applying only obligatory transformations to the phrase-structure strings (e. g. the transformation of affix + verb into verb + + affix in the present tense, hit -s, etc.). Non-kernel or derived sentences involve optional transformations in addition, such as active to passive (the boy was hit by the man). But later interpretations of the transformational theory have made less use of this distinction, stressing rather the distinction between the underlying "deep structure" of a sentence and its "surface structure" that it exhibits after the transformations have been applied. Transformational operations consist in rearrangement, addition, deletion and combination of linguistic elements.

Phrase structure rules form a counterpart in the theory of generative grammar to two techniques of linguistic analysis (one old and one rather new).

In the words of E. Bach, the old practice is the schoolroom drill of parsing, that is, of assigning grammatical labels to parts of a sentence. In a schoolroom drill the following analysis might occur:

The man article noun	gave verb	me pronoun	a article	book noun
whole subject		indirect object		direct object
		whole	predicate

The other technique — in reality only a more sophisticated version of parsing — is so-called immediate constituent (IC) analysis. It attempts to break down constructions into subparts that are in some sense grammatically relevant.

The theory of transformational grammar begins by making fundamental distinction between two kinds of sentences: kernel sentences and their transforms. Kernel sentences are the basic elementary sentences of the language from which all else is made. All constructions that are not basic are transforms, i. e. they are derived from the basic ones by certain grammatical rules. Transformations can change and expand the kernel in many ways to form the great variety of sentences possible in a given language.

The system of any language contains a rather small number of basic sentences and other structural elements (such as morphemes and phonemes). All the other linguistic forms, sentences of different structure, are derived (generated) from these basic (kernel) elements by certain regular derivation rules involving different kind of operations. This understanding of the system of any language is, in fact, the main assumption of the transformational grammar.

The two basic problems of the T-grammar are: a) the establishment of the set of kernel or basic structures, and b) the establishment of the set of transformation rules for deriving all the other sentences as their transforms¹.

A transformational rule is a rule which requires or allows us to perform certain changes in the kernel structure: rearrangement of linguistic elements, so-called "permutation", substitution, deletion, the use of function words, etc.

The transformational rules show how to derive something from something else by switching things about, putting things or leaving them out and so on ².

It is to be pointed out that transformational analysis applied in teaching on different instruction levels can hardly be considered as altogether quite novel. Transformational relations involved in tense-formation and passive forms, for instance, were, in fact, always presented as devices of obligatory transformations on the morphological level. The

¹ See: Z. S. Harris. Co-occurrence and Transformation in Linguistic Structure. "Language", v. 33, No. 3, 1957.

² See: P. Roberts. English Syntax. New York, 1964, p. 97. - 3

recognition of brought as the past tense of bring, and similarly be brought as the passive of bring, depends primarily on relating large numbers of sentences and on the analysis of collocations between nouns and verbs in the sentences.

Such are also number and person transformations or, say, different kind of transformations which were applied implicitly in traditional grammar on the syntactic level depending on the purpose of communication: constructing negative transforms, changing an affirmative sentence into a question, transformations which produce exclamatory sentences, etc.

Deficiencies of various kind have been discovered in the first attempts to formulate a theory of transformational generative grammar and in the descriptive analysis of particular languages that motivated these formulations. At the same time, it has become apparent that these formulations can be extended and deepened in certain ways.

N. Chomsky's Aspects of the Theory of Syntax ¹ is a notable attempt to review these developments and to propose a reformulation of the theory of transformational generative grammar that takes them into account. The emphasis in this study is syntax; semantic and phonological aspects of language structure are discussed only insofar as they bear on syntactic theory.

The author reviews the general orientation of all work in generative grammar since the middle fifties. His specific intent is to determine exactly how this work is related — in its divergencies as well as its connections — to earlier developments in linguistics and to see how this work relates to traditional issues in psychology and philosophy.

N. Chomsky implicitly relates his grammar to language teaching and learning by associating his results with traditional grammars. He mentions that these do not give explicit rules for putting words together into sentences, although they give enough rules of word concord, examples and so on, to allow the student to do this intuitively. N. Chomsky gives no rules for putting sentences together to make discourses, but leaves this to the intuitions of the learner. His aim is to put forward the rules to generate all possible sentences of a language in terms of a given set of morphemes. In his words, any language has a finite set of available morphemes, but an infinite set of sentences; this shows definite hypostatisation of the unit "sentence".

Transformational grammar involving a reorientation of linguistic theory has naturally given rise to vigorous controversy in linguistic studies, and much still remains to be done in language learning to evaluate its potentialities adequately. It is to be expected, however, that the theory of T-grammar will continue to develop and contribute to general linguistic study by solving some important previously overlooked issues.

The structural procedures of modern descriptive theory are used by Soviet linguists to identify the nature of some linguistic facts. It must, however, be emphatically stressed that in some questions our standpoint is essentially different. Some American linguists are known to

¹ See: N. Chomsky. Aspects of the Theory of Syntax. Cambridge, Massachusetts, 1965.

advocate rigorous separation of levels and a study of language as an autonomous system. Such abstraction seems altogether erroneous and brings little scientific order to language learning; dogmatic assumptions of this kind are always responsible for the distortion of linguistic facts. This approach seems to have already been abandoned by most structuralists (Z. Harris, N. Chomsky).

What is also open to criticism is setting absolutely apart synchronic and diachronic aspects of linguistic units. In language reality the two aspects are organically related and as such cannot be always absolutely isolated. Regrettable mistakes occur if this is overlooked.

There are a number of European schools of linguistics, and the differences between them are in some instances rather significant. The linguistic theories which they hold have, in fact, been developed in a variety of ways.

With the diversity of view-points within descriptive linguistics, it is not surprising that English descriptive grammar is not as a type uniform. Sometimes grammarians differ in the view of language that underlies them. Some of grammars differ only in terminology, in stylistic conventions of statement, or in other basically inconsequential matters. For the most part there is a variation in many directions, with intergradations in linguistic analysis. But despite a considerable divergency of their aims and linguistic approaches there is a certain continuousness in different English grammars observed in their keeping up the grammatical tradition. The foundations of the English grammatical theory were laid already in the first part of the prescriptive grammar, though its morphological system was based on Latin and syntactic concepts depended largely upon rhetoric and logic.

The prescriptive normative grammar has the longest tradition and is still prevalent in class-room instruction. Its most important contribution to grammatical theory was the syntactic system developed in 19th century.

Though much has been done, the three types of scientific English grammars have not yet succeeded in creating any quite independent and new grammatical systems.

R. W. Zandvoort's Handbook of English Grammar (1957—1965) is a descriptive grammar of contemporary English. It deals with accidence and syntax, leaving aside what belongs rather to idiom and is not amenable to general statement. It likewise eschews historical digressions; synchronic and diachronic grammar are, in the author's opinion, best treated separately. In this, as in other respects, R. Zandvoort confesses himself a pupil of Kruisinga, whose Handbook of Present-day English, despite certain extravagances in its fifth and final edition, he considers to be the most original and stimulating treatment of English syntax.

* * *

A major contribution to the development of modern linguistics has been made by Soviet scholars.

The accomplishments of Soviet linguists in the theory of English structure are presented by the great wealth and variety of individual

studies of numerous problems treated in various monographs, grammar books and work-papers which appeared during this period and have been noted in our bibliography.

Linguistic studies of Modern English structure made by Soviet scholars contain most valuable information about the language as system and have notable merits in the grammatical theory making its study more illuminating and contributing to a scientific understanding of language development. Such are, for instance, the monographs and books edited in this country in 50-60-ies by V. N. Yartseva, A. I. Smirnitsky, O. S. Akhmanova, Y. N. Vorontsova, B. A. Ilyish, N. N. Amosova, I. P. Ivanova, I. V. Arnold and others.

Most perceptive and useful treatments coordinating and deepening the grasp of the language will be found in V. N. Yartseva's monographs and scholarly accounts made at a special academic level, with much new insight on the subject in the light of modern linguistics.

A valuable source of significant information revealing important aspects of language in discussion of syntax and morphology will be found in well known A. I. Smirnitsky's grammar books.

A major stimulus to intensive studies of the theory of English structure in Soviet linguistics was the research of our scholars in recent times. This has brought new accomplishments in modern grammatical theory which are original, significant and practical. Investigations of recent years gain an important insight into the structural methods of linguistic analysis, syntactic description, in particular. Such are the grammar books edited by O. S. Akhmanova, V. N. Yartseva, L. Barkhudarov, L. L. Iofik, Y. O. Zhluktenko, G. G. Pocheptsov and others.

Current work in grammar attempts to provide the insight into semantic aspects of syntax, the processes of sentence formation and their interpretation, the processes that underlie the actual use of language.

Investigations of Soviet scholars throw much additional light on numerous aspects of language encouraging fresh attempts not only in the theory of English structure but also comparative studies of grammar (V. N. Yartseva, Y. O. Zhluktenko).

The structural procedures of modern descriptive theory are widely used by Soviet linguists to identify the nature of some linguistic facts of different levels of the language.

Important observations are presented in A. Korsakov's book where we find the description of the system of the English verb, revealing to the student the way in which the language actually works. The book is not only intended to show the student how the English tenses are actually used. It is also helpful as an introduction of some methods and ways of linguistic analysis.

Various aspects of grammar have been described in a considerable number of dissertations defended in this country on specialised topics, such as semantic aspects of syntax, the grammar of English nominalisations, synonymic correlation of linguistic units, comparative study of languages, etc. to which we turn the attention of the student with suggestions for further reading.

Дата добавления: 2014-12-29