LANGUAGE IN INDIA

TOWARDS AN UNDERSTANDING OF SENTENCE IN SCIENCE
IN AN INDIAN LANGUAGE CONTEXT

M. S. Thirumalai, Ph.D.

1. Linguistics

Traditionally the core of the discipline of linguistics includes the general theories about language, the techniques and the concepts related to a description of the structures of a language, the principles and processes involved in the changes that take place from period to period in a language, the principles and the concepts involved in comparing languages and the adaptation of all the above for some practical purpose. Thus, we have general linguistics, descriptive linguistics, historical linguistics, comparative linguistics, and applied linguistics respectively. Modern linguistics is characterized by formalism, structuralism and functionalism: Linguistics is structural because, in addition to the above, emphasis is laid upon the contrasts among units and parts of units. In the characterization of linguistic units, first of all, the composition of the unit is taken into account. Secondly, the distributional characteristics of the unit are presented. Every unit of a language occurs only in a restricted number of combinations. And these places of occurrence show the distributional characteristics of the unit. Thirdly, a unit is definable in terms of the use to which the unit is put. This is called the function of the unit.

Within linguistics, there are different schools of thought which differ from each other in many respects. One of the aspects is with regard to how language is acquired by human beings: is social behaviour. Secondly, these scholars differ among themselves on the number and kinds of levels which should be posited in a description of human language. Some scholars posit five levels -phonetic, phonemic, grammatical, morpho-phonemic and semantic, but some others subsume all the above only under three components -phonological, syntactic and semantic. Thirdly, the scholars differ among themselves as to whether the levels posited thus, should be considered independent of one another or they should be considered dependent -in other words, as to whether the description of one level is dependent upon the description of another level. There are many more points on which differences of opinion and assumptions exist among different linguistic approaches to the study of language. These are matters which do not directly concern us here, even though adopting one school in favour of another will definitely have a direct bearing on how we would go about describing the linguistic data.

Linguistics is concerned mainly with an exposition of the structures language employs. The linguistic description concerns itself with the linguistic structures. Assumptions are made about the psychological and neuro-physiological mechanisms of language and yet the main concern continues to be an exposition of linguistic structures.

In their preoccupation with the exposition of linguistic structures, most linguists consider sentence as the unit of analysis. The elements which constitute a sentence and the order in which the elements occur are dealt with. Formal features of the elements are sought. Linguists even consider how one sentence is related to the other in terms of patterns and embedding processes. However, most of their research is sentence-based. The inter-sentential relations that exist between utterances or sentences in a text are also dealt with, but such a treatment does not form the core of linguistics.

In this chapter we first deal with how the unit sentence is employed in the language of science; then, we discuss the inter-sentential relations and how a discourse is formed with special reference to the scientific texts. Later, we discuss also the stylistic variations that we make in the language of science and discuss briefly the characteristics of style sheets employed by research journals. We link these with rhetoric - a science of persuasion.

2. What is Sentence?

Everyone knows what a sentence is. But when it comes to defining it, we all run into difficulty. More than two hundred different definitions are reported to be available. Some of these are as follows: A sentence is a group of words expressing a complete thought. A sentence may be identified on the basis of punctuation - between two periods we have a sentence. Thus, a sentence is defined as a unit followed and proceeded by indefinite pause or silence, with phonetic features to indicate the prepausal positions. A sentence is marked in writing with an initial capital letter in languages in whose scripts a distinction between capital and small letters is maintained. A sentence is marked in writing also by a full stop, a question mark, a mark for exclamation, or by a semicolon.

In speech, a sentence is marked by a characteristic intonation. A potential pause is easily identified between sentences. Hence, a sentence is defined phonologically as a stretch of speech which may be uttered with a complete intonation preceded and followed by silence.

A sentence is defined sometimes on the basis of "sentence feeling." A sentence is defined as having a minimum number of prescribed elements/categories such as predicate, subject, direct and indirect objects, and so on. A sentence is also defined as a portion of speech that puts forward to the listener a state of things (a thing meant) as having validity. A sentence is defined a grammatically constructed smallest unit of speech which expresses its content with respect to this content's relation to reality. A sentence is defined as a relatively complete and independent human utterance -the completeness and humanness being shown by its standing alone or its capability for standing alone, that is, capable of being uttered by itself. A sentence is defined as a group of words joined together by grammatical agreements (relating devices) and which, not grammatically dependent upon any other group, are complete in themselves. A sentence is also defined as a construction of form which, in the given utterance, is not part of any larger construction. On the basis of the dependence or independence of though, a simple sentence is defined as a group of words which expresses a single independent thought. A compound sentence is defined as a group of words which expresses two or more unified thoughts, out of which one is the main or principal thought and others subordinate thoughts.

Sentences are classified into many types using various kinds of variables. On the basis of thought, as mentioned above, sentences are classified into simple, compound and complex types. Sentences are divided into simple, compound and complex on the basis of their structural composition also. A simple sentence has one independent proposition or one independent clause. (A clause is defined as a syntactic construction containing a subject and predicate and forming part of a sentence or constituting a whole simple sentence.) A compound sentence contains two or more propositions or tow or more independent clauses; and a complex sentence consists of one independent propositions or an independent clause and one or more dependent propositions or subordinate clauses. The compound and complex sentences may be combined in the same sentences. Sentences are also classified on the basis of the types of content embodied by the sentences: declarative (statement), interrogative (question), imperative (request or command), exclamatory (emotions), and so on. On the basis of the grammatical patterns also, sentences may be classified into various types such as equational (copula) types, verbal types, dative subject sentence type, passive sentence type and so on. There are certain sentences, called elliptical- elliptical in form or sense, or both. These also form a type of their own.

Thought, meaning, punctuation, structural composition, number or words, kinds of categories, content, and independence of occurrence are, thus, some of the major variables used to define or characterize what a sentence is. Using one to the exclusion of others results always in a definition which does not satisfy. Linguistically oriented definitions insist upon certain structural composition and independence of occurrence, in spite of the "independent" status of elliptical types which go against the requirement of structural composition of "full" sentences. In any case, one intuitively knows what a sentence is, in a language known to him. One is able to recognize a sentence, perhaps with some initial difficulty, even in a language not known to him, because of certain underlying-universal- characteristics of sentences across languages. These universal characteristics are intuitively felt, but hard to characterize formally. However, note that positing such universals in no way fails to recognize or minimizes the importance of the difference in structural patterns of sentences within and across languages.

3. Analysis of Sentence

Grammatical analysis of sentences presupposes a knowledge of the meaning they convey. Parsing still continues to be a major step in the analysis of sentences - the total meaning of a sentence is so compartmentalized that each compartmentalized meaning part has an appropriate formal part. Generally speaking, sentence is considered or identified as consisting of several formal parts and this internal structure is identified, described and explained in the grammatical analysis. There are two meanings, namely, the lexical meaning and the structural meaning. For comprehending the meaning of a sentence one should have a knowledge of both lexical and structural meanings.

The grammar of a language consists of formal devices -forms and their arrangements, meanings and distribution. Both forms and their arrangements have their significance as parts of a system. Sentence in a language is not merely a group of words but a structured string of words. The formal meanings of the arrangements into which words fall in a sentence can be generally understood without an understanding of the lexical meanings. Sentences of a language are innumerable but the underlying patterns of all the possible sentences in a language are finite in number. Through the use of finite patterns, we produce an infinite number of sentences in human languages. The native users of a language have generally no difficulty in identifying what a sentence is and what a sentence is not in that language. They know intuitively well to identify and use the admissible combinations of clusters of words in sentences of that language.

Linguistic analysis generally has sentence as its major, if not the only, focus. A complete sentence or potentially complete utterance is the focus of grammatical analysis. Though certain links may be established between sentences of a conversation or a paragraph in writing (the links based on grammatical devices, in addition to the links based on the continuity of content such as cross reference and anaphora, (See section 7) such links are not the pegs on which grammatical theories are hung. Linguists find that the interrelations of elements within a sentence form a much more organized coherent whole and, so, they consider sentence as the largest unit of grammatical analysis.

In the grammatical analyses of sentences we seek at least three types of information. In the first type, we divide the sentence into various components or categories which in their turn have their owns substrings; in the second type of information, we seek to find out the functions of the components identified in the first processes; and, in the third type of information the features of each individual item will be sought. Thus, as Chomsky (1965) reports, the first one looks at the components of the sentence, the second at how these parts are held to each other and the third at how each form is linked with the structure.

Grammatical categories such as noun, verb, adjective and adverbs can be further subcategorised. This subcategorization is done on the basis of their behaviour in a sentence, in terms of elements with which the lexical items of a category coheres with the other lexical items in the sentence. That is, we identify the features associated with the syntactic context in which an item occurs. A group of items of a category that share the common contexts will be put under that category as a subcategory. For example, the grammatical category verb can be subcategorised into transitive, intransitive, causative, reflexive, etc., on the basis of the contextual features demanded for the used of the above. A transitive verb requires that a direct object occur after it. Transitive verbs in English may be further classified into those that do and those that do not normally permit object deletion. Under the former category come the verbs such as read and eat and under the latter category frighten and keep. In the English language, the verbs are subcategorized with respect to the occurrence and types of verbal complements: subcategorization of a grammatical category is done on the basis of the syntactic behaviour of an item (in terms of its occurrence with other items).

Consider the following sentences. The asterisked sentences are not acceptable for syntactic and/or semantic reasons. The verbs of these sentences do not cohere with the words following, in semantically and/or syntactically acceptable sentences. (We exclude from our purview the metaphorical, mythological and other such uses).

1. I killed a snake.
2. I killed a table.*
3. He walked.
4. He walked a table.*
5. I read a novel.
6. I read a man.*

Hornby, et al., (1963) identify twenty-five verbal patterns in English. Also they identify passive types of constructions for eleven of these twenty-five constructions. All these categories of verbs and syntactic patterns are used in the expression of sciences. That is, almost all the structures of the ordinary language are also used in the language of science. Exceptions are few and may be related directly to the formal nature of the scientific activity and its specialization. To this extent, the ordinary language and the language of science share a large number of features. But the differences between the two lie in the preference exhibited by the language of science.

We will find out these differences very easily if we analyse the language of any given book of science. We would discover that all the patterns given in the above-cited reference are used in that book. However, these patterns would not be found in their "pure" form, as the scientific texts abound in complex sentences. In a textbook, the author's identity or presence is felt thorough the use of we. This sometimes refers only to the author or authors of the book and at times refers to the author/authors as well as the reader. Such identification inclusion is restricted generally to indicate the assumptions of the author/authors and the conclusions they have drawn. It is used also to persuade the reader to go along with the author/authors. We would also notice that the length of the sentences used in a scientific book is greater than the one we find in "non-conceptual" types of publications.

While preference for certain sentence types and the length of the sentence characterize sentence construction and use in a scientific text, the use (in English) of the It construction and passive sentence type contributes to an aura of objectivity. The length of a sentence is generally indirect proportion to the number and complexity of concepts packed into a sentence. A child resorts to an acquisition and mastery of longer sentences along with its mastery of and urge to express complex concepts. Brevity without loss of essential information is a desideratum in the language of science.

It is true, however, that not all the patterns given in the above-cited reference would be found used in a single journal article. Journal articles are of a limited nature in terms of their scope, content and linguistic structures. Even in the case of syntactic structures used in a book of science, there seems to be a preference shown for some structures over others. This preference can be easily demonstrated through a frequency count of the occurrence of different structures used in any book of science. This preference form some structures over others is reflected in journal articles also.

Another characteristic also must be mentioned here. We have referred to a fact that a constancy of meaning for each word used in the language of science is generally demanded. Because of this demand for constancy, alternative available to a user of language of science are quite limited. Generally speaking, only one word out of the likely many in the ordinary language is admissible and is selected. Although such strict adherence and restrictions are not insisted upon in the selection and use of syntactic patterns, in practice such an adherence is quite common in journals articles. It is common in elementary books also. A consequence of this restriction is the conscious avoidance, on the part of the scientists, of figurative speech, satire, metonymy, alliteration, hyperbole, etc., except in model building where such devices are found productive and helpful.

The different kinds of verb patterns which we have listed above can be grouped broadly under five forms of the predicate in English, following Onions (1932). Before we present Onions' classification, some clarification of the notions of subject and predicate is required. A subject denotes the person or thing about which something is said by means of the predicate. A predicate is what is said about the person or thing denoted. This characterization of subject and predicate is intuitively understood and serves practical ends. We find that (most of) the predicate phrases contain a verb and through this verb, predication is conveyed.

In Onion's classification, the first form of the predicate contains only a verb as in sentences, Day dawns and He studied. In the second form of the predicate, the predicate consists of a verb and a predicative adjective, a predicative noun or a predicative pronoun. This form of the predicate includes participles. There is a unity between the verb and adjective, etc. Examples are They were ruined, He was coming, and The warrior lay dead. The third form of the predicate has a verb and an object, which denotes the person or thing to which the action of the verb passes over: I eat apples. The fourth form of the predicate consists of a verb and two objects: We ask you this, and I asked him a question. In the fifth form of the predicate, a verb, an object and a predicative adjective or predicate noun are found: He though himself a man. These five forms may be considered as the principal or fundamental ways of expression in English, from which other manners of expression are derived. Again, all the five forms could be easily recognized in a scientific text. One may also suggest a hierarchy on the basis of intuitively felt frequency of occurrence in the decreasing order: Predicative form 3, Predicative form 2, Predicative form 4, l Predicative form 1 and Predicative form 5.

To continue our discussion of the types of information conveyed by components of a sentence, lexical items can be looked at also from the point of view of what linguists call selectional restrictions. Although two items can cohere in a sentence on the basis of the appropriateness of grammatical categories as well as the appropriateness of the subcategorization of these categories, still such coherences need not make a sentence a correct or a good one.

As for example, the verb in the sentence "I ate a cake" is a transitive verb which can take a direct object. The sentence "I drank a cake" also has a transitive verb followed by a direct object but will not be acceptable, as the coherence of the lexical item drank with the lexical item cake breaks the selectional relations. Consider another example: "My husband is pregnant". This also is not acceptable because, as one might easily find out, the lexical features of the words husband and pregnant are in conflict, although they fulfil the requirements of grammatical categories in the cited sentences.

One can easily cite many other examples. The verbs admire and see require animate subjects; frighten and strike may have, both animate and inanimate subjects; read and know require human subjects. And like these, many restrictions are identified for the use of various categories of words. Such restrictions are not heeded in the newspaper language, in news items as well as advertisements. This led J. R. Firth (1964:118), a well known British linguist, to comment as follows in 1930's: 'If Charles Lamb could be brought back to explain the examples given,' (from newspapers under various headings), 'he would soon be very perplexed and would not get the right kind of certificate. What Dr. Johnson would say could be imagined and might make amusing if only we could find a 1937 Boswell to do it.' Such a breaking of restrictions in the uses of words is found also in the language of science, but the difference lies in the users of the latter making conscious efforts to avoid feelings, social attitudes, prejudices, fears, wishes, fantasies, and ambitions, in the words of Professor Firth, 'to be free as possible from sentiment, prejudice, wishful thinking, from the shackles of the mythical, magical, and fantastic.' Such a breaking of selectional restrictions characterizes the language use in the expression of sciences in English and several other languages.

Languages do differ among themselves as to the plan of selectoinal restrictions imposed on the coherence of different grammatical categories. Accordingly, the extent of ignoring selectional restrictions for special purposes in one language may be different form that of another language. In a language like Tamil where the main verb in a sentence must indicate, among others, the number (singular or plural), the gender (masculine, feminine, common, or neuter) and the person (first, second or third person), scientific concepts and scientific expressions automatically adopt the neuter gender and the third person. Once this choice is made, the choice with regard to the verbal form (the strong or the weak, the transitive or the intransitive) is to be made. More often than not, the weak/intransitive form of the verb is used; if there is to be emphasis and if the action is caused by an agent, then the strong form/transitive form of the verb is used, generally speaking. Thus, the form a verbal root takes also has great relevance in this regard.

4. Ambiguity and Language of Science

Native users of a language have several abilities in terms of the use of their language. These abilities have to be acquired also by a second language learner for a competent use of the particular language they are learning. These abilities include the following:

1. The native users of a language have the ability to distinguish between the grammatical and ungrammatical strings of a potentially infinite number of utterances in the language. This contributes to or is a consequence of their ability to recognize the order in which words occur in the sentences.
2. The native users of a language can often understand the full meaning of a sentence even when there is no explicit occurrence of all the words in a sentence. This contributes to our ability to reconstruct the full meaning of a sentence, even when not all the words are present in a given sentence This contributes also to our ability to anticipate words and complete a half-uttered sentence.
3. We also have an ability to interpret ungrammatical strings even when the elements of interpretation may not be physically present in the string.
4. The native speakers are able to interpret sentences which look exactly alike: I had three books stolen; Flying planes can be dangerous, and so one.
5. Sentences which seem to have similar structural shapes are interpreted correctly:
   The cow was found by the farmer.
   The cow was found by the stream.
6. The above two sentences look as if they have similar structures, but different interpretations are to be given for the phrases by the farmer and by the stream. The first one has an agentive function and the second one a locative function.
7. Though two sentences may seems to have different formal shapes/structures, the native speakers have an ability to interpret them as related:
   The farmer found the cow.
   The cow was found by the farmer.

Utilizing the above abilities native users of a language are able to disambiguate the utterances they hear and/or utter. Science expects from its practitioners an unambiguous language without embellishments. But note that ambiguity in language use is quite a natural process. Ambiguity may be on word level or on sentence level. Some examples are: The police station was right by the bank; The lamb is too hot to eat; and Visiting relatives can be a nuisance. In the disambiguation process, the native speakers are helped by all the abilities listed above, but most important is the knowledge of the subject. Although these abilities are very important in the expression and clarification of scientific ideas, one must bear in mind that whether it is a literary piece or a scientific piece, clarity is very important. That is why the classical traditions I may languages of the orient as well as of the West have identified several techniques to retrieve the information contained in poetical works as well as scientific texts. The classical traditions have identified the errors that a writer/poet should avoid and the "beauties" or stylistic devices that one may adopt to enhance the clarity value of what one writes or composes. Some of these are language-specific in the sense that such errors and beauties are dependent upon the linguistic structures employed. Others are language-independent and applicable to discourses in any language.

The ancient Tamil grammarians, over a thousand years ago, suggested the following:

1. The ten errors: Understatement (use of less number of words than necessary): exaggeration (use of more words than necessary); use of erroneous expressions; ambiguous statement; vacuous statement; digression; progressive loss of compactness in diction and content; and, use of words sans significance.
2. The ten "beauties": Brevity; a brief but lucid statement that is easy to comprehend by all; pleasant to read; use of words that sound well; use of fine and culturally appropriate words; a statement portraying deep thought/susceptible for deep interpretation; an appropriate sequencing; making statements in consonance with a highly valued tradition/form; making statements which express highly desired good things; and, making statements with easily understood examples.
3. The thirty two techniques: Start with an iteration of what is to follow; appropriate organization; state generically; specify the components of what is stated generically; to conclude the way the men of tradition concluded; indicate the (literary) usage/occurrence/ illustration for the (grammatical) rule stated; occasional use of the ancient aphorisms/formulae/ definitions; cite the arguments of others in one's own book; elucidate the details to make explicit the meanings of the words; use of words that are related in form, use or meaning; to make statements which carry more than one meaning; conclude stating reasons for something which has been stated earlier without giving reasons; make concluding statements about something/somebody even when it is not the appropriate place -do this, if the definition offered for something else is appropriate for the present matter also; extend the definition of something somebody to all those to which the definition is appropriate on the at occasion; non-use of obsolete items; acceptance and use of innovations; use again, wherever necessary, the first mentioned information; present later an item that should occur in the beginning; bring out the difference/contradictory or distinguishing facets; to proceed form the contradictory/distinguishing facets to generalized or generic statements; to mention in brief stating that the same will be death with in detail later on; refer back to what is said earlier in order not to make a detailed exposition, at present juncture, of what has already been discussed; accept one form along the several arguments; illustrate one's own principles/definitions with own examples; explain how examples cited fit into the definition/principle offered; resolve ambiguity and choose the appropriate interpretation; make statements in such a way that these imply validity even for those things left unstated; accept/recognize/approve the arguments given in others' books; speak of one's own innovations/contributions repeatedly; conclude in such a way that the content is totally accounted for/exhausted/exploited, when the expression comes to a close; present all the related items in the same place; and, present definitions or principles in such a ay that the meaning of which would be understood through a deep study of these and other definitions/principles.

The above listed errors, beauties, and techniques were suggested for the appropriate use of language, particularly for the expression of grammatical concepts. In Tami and in several other languages, the oldest "scientific" pursuit and expression were made in grammatical and rhetoric studies. The errors and appropriate expressions were a subject matter of intensive study in ancient Greece as part of expositional logic. Equal importance to innovations, discoveries, inventions and the expression of all these in words has been given throughout the ages.

Ambiguity when used purposely as a literary device adds to the quality of a literary piece. The same ambiguity when used in a scientific text, however, hampers comprehension, and is taken generally as a sign of lack of proper grounding in the subject matter, or lack of comprehension of the subject matter by the writer. Resolution of ambiguity and comprehension of what is being offered require some competence -adequacy- in both the writer and the reader.

The language of science is often compared to an equation - a valid equation. The sentence should balance with what is being communicated -intended to be communicated. If it balances, there is no ambiguity. A sentence is to be direct, lean without unnecessary embellishments, with the right type of verb and with smooth transitions form sentence-to-sentence and paragraph-to-paragraph. But, to achieve this, the writer-scientist should have a knowledge of what his words do and do not, can do and cannot do.

Grammar and punctuation, vocabulary, (used of clichés, big words, jargon, high sounding words, etc.) and failure on the part of the author to put himself into the readers' shoes are all factors that contribute to ambiguity, in addition to writer's general inadequacy in handling the syntactic nuances of language, variable styles and the subject on hand. A better writer is aware of the possibility of ambiguity and strives his best to avoid it by baring in mind the audience for whom he is writing and the purpose for the fulfilling of which he is writing. He writes and rewrites striking out unnecessary words. The act of striking out reveals that the thought can still be expressed without the words struck out. Progress in science depends a lot on the clarification of ideas already available and on the clear elucidation of newer ones. Accordingly, one should be able to recognize ambiguity whenever it occurs and must have a clear idea as to how ambiguity can be avoided using proper linguistic structures.

5. Deep Structures, Transformations and Surface Structures

For resolving ambiguity and to arrive at the correct and intended meaning, we must have an idea of the deep structure of the sentence in question. Sentences in any language have both a deep structure and a surface structure. Deep structure gives the meaning of the sentence uttered. Deep structure underlies the surface structure; surface structure is what we actually utter in the form of sentences in our speech and writing, when we communicate with one another. In many- cases the deep and the surface structures of a sentence may be identical, as in I go home. In several cases the surface structure of the sentences may be same or even identical but the underlying deep structures may be different: I had three books stolen; Flying planes can be dangerous.

The deep structure is considered abstract. With the application of rules, called transformational rules, on deep structures, we derive surface structures. That is, a deep structure becomes a surface structure through an application of transformational rules. Note that we do not always speak or write only in simple sentences-simple in the sense that the sentence cannot further be split into smaller sentences. These simple sentences need not be short. The requirement is that there be only one finite verb. (The characteristic formal features of finite verb may differ from language to language-in general, it carries the burden of main action of the sentence.) Length of a sentence can be increased by phrasal (non-sentential type) additions to it. Furthermore employing only short sentences at a stretch would make the text unnatural and less effective means of communication. In fact, more often than not, the sentences we utter in speech or in writing are not simple in the above sense. These sentences are derived through an application of transformational rules on the underlying deep structures of sentences.

Although all languages contain transformational rules which transform deep structures into surface structures, the number and kinds of specific transformations, their scope of applications and the order in which they occur in one language may be different from those of another language. That is, it amounts to saying that the number and kinds of sentence types in a language may be different from those of -another language, because these types, in a majority of the cases, are derived through an application of transformational rules. The identification, description, and explanation of the transformational processes are an activity that continues along with the refinement or innovation of grammatical models of description.

The variety of transformations that a language uses (here, English) may be appreciated from a list of transformations Jacobs and Rosenbaum (1968) discuss in their book: adjective, adverb preposing, affix, agreement, article, auxiliary agreement, auxiliary incorporation, case, cleft sentence, complementizer deletion, conjunction reduction, conjunction shift, contraction, copula, extraposition, identical conjunct reduction, identical noun phrase deletion, identical verb phrase deletion, imperative, indirect object inversion, interrogative, "it" deletion, "it" replacement, manner adverbial, negative adjunction, negative placement, negation, nominalization, non-restrictive clause, non-restrictive relative clause, noun segment deletion, noun suffix, particle segment, particle movement, passive perfect segment, preposition, preposition deletion, progressive segment, pronoun, pronoun deletion, question deletion, reflexive relative clause reduction, relative pronoun deletion, time-place deletion, verb agreement, verbal agreement, verb suffix, and WH-question.

In spite of the variety of transformations one may notice in an individual language, the transformations used in all the languages may be grouped under several formal classes. Although the form, the content and the order of these formal classes are still a matter of dispute among linguists, one is in agreement with Bach (1974: 79) when he says that 'the basic purpose of including transformational rules in a grammar is to make it possible to describe one type of structure directly and to describe other structures as modifications of structures of the first type.' For instance, a recursive rule is applied repeatedly on a given sentence to produce longer and longer sentences. The embedding rule, which also has recursive property, is used to combine two or more sentences and this combination results in a complex sentenced sentence with main and subordinate clauses. The use of joining rules results in a compound sentence.

6. Sentence Types and Preferences

The discussion of transformations leads us to a discussion of the types of sentences. A closer look into the structures of sentences in any text would indicate that the longer sentences of a language are, indeed, structured in the image of a small number of short sentences. The short sentences are expanded in several ways through an addition of phrases/sentences in appropriate places in the given short sentences. This facility enables us to understand sentences that we have not heard or read before. We interpret a new sentence as a sentence based on structures already known to us. The ability to interpret and comprehend the sentences not heard or read before reveals our mastery of basic patterns of sentence structures as well as the means by which these patterns are expanded to derive new sentences.

As we have pointed out, sentences used already or which may be coined and used in a language can be reduced to a small number of patterns-infinite number of sentences through a finite number of patterns. Some linguists consider the patterns common to a large number of sentences as favourite types in that language. The favourite type, however, differs from content to content, context to context and pursuit to pursuit. This variation in preferences can-be identified even within the language of science on the basis of the above-mentioned points. The variation is found also in relation to the manner in which the purpose is achieved, as well as in relation to the subject matter itself. But all these variations accept a central binding force, namely, the impersonal style. What is of utmost importance is the feature of fitness between the linguistic expressions and the requirements of the content. To the extent this fitness is evident and maintained, a linguistic expression will be complete, adequate or inadequate. With the fitness comes accuracy, economy, and compactness, and also the impersonal tone. (But while acquiring this feature of fitness, there are bound to be several stages.' We shall discuss this in another paper.)

Language use in science through English (and in many other languages) seems to prefer the passive voice correspondence of the active voice sentence. (Such a preference is now on the decline). In English, the past perfect tense is used for a description of past work. The past tense is used to describe a procedure and the present tense is used to describe the main contents. The future tense is very rarely used; it is used to indicate predictions of what would occur if a set of objects, processes or events has been followed. It is also used to indicate projections for the future. The imperative is used for analytical procedures, for recipes and for conducting experiments. For example, the expression was heated tells us that the action occurred in the past, that it was performed on one substance and not on several, and that some subject had suffered the action and did not perform it. If the phrase is was boiled, the additional information that the subject which suffered the action was a liquid would also be available. In any case, note that the information as to who did the act was not made explicit. As regards the filling in of the subject slot position, there are three alternatives available while making a matter of fact statements. These are: (a) I filtered the mixture, (b) We filtered the mixture, and, (c) The mixture was filtered. Of the three, (c) is preferred first, (b) second and (a) last. Moreover, the entire experiment, like the general practice in other narratives, would be generally reported in past tense.

Such preferences exhibited by one language need not be found in another language. This has been reported by us earlier. For example, in the place of a transitive verb sentence which has a corresponding passive-like construction, the Tamil language prefers a corresponding intransitive verb sentence for the expression of sciences. Instructions for the conduct of experiments take the form of imperatives in English with the imperative pronoun deleted. In Tamil, imperatives may be honorific. (The addressee may be shown respect as a superior or treated as an equal/inferior through the use of appropriate suffixes to the verbal root.) The instructions for the conduct of experiments, etc., take the form of an imperative sentence type in which the pronouns are deleted. But this imperative construction is different from the regular type which is used for forthright or face-to-face instructions. In this construction which is different from the regular imperative construction, the verbal concord suffixes indicating gender, number and respect (honorific) are not used. But a neutral verbal concord suffix is used, which literally gives the meaning "is desired".

In English, the imperative construction used for face-to-face communication is not different from the one used to give instructions for conducting experiments, etc. In Tamil, however, a distinction, as mentioned above, is maintained. Scientists using German seem to be very much conscious of the sentence length and, as a result, one notices a tendency to keep the sentence as short as possible.

One tries to avoid expansions wherever possible, although materials continue to be presented within parentheses. The extent of presenting certain information within parentheses, however, does not reach the level where the matters presented within series of parentheses may occupy more space in a page than the matters without, as one finds in some English authors like John Lyons, a well-known British linguistic theoretician. In German, the adjuncts, that is, phrasal attributes, to a simple sentence can be many. Of these attributes, adverbial phrases seem to be more. These attributes can make even a simple sentence longer. The number and kinds of verbs used in scientific German are few, as in other scientific languages. Special meanings are attached to verbs, making such usages technical terms. Complex sentences, as in English, are used for definition purposes. However, in German, complex sentences are used mainly for this purpose. Infinitival and participial phrases are frequently used to condense information. Some of these features are shared by all languages and some are specific to German. Thus, each language has its own structural patterns, and favourite types of sentences for various purposes.

The above fact is largely ignored in the sense that even in languages where there is no proper passive voice construction, attempts at the expression of sciences, imitate the practice of preferring passive constructions in English. If clarity, objectivity and an impersonal tone are important characteristics that any scientific expression must have, the same can be achieved without oneself hinging exclusively upon a single sentence pattern. Such a situation was prevalent in the language used for commercial transactions in the past between commercial concerns and customers. A certain amount of formality was felt to be necessary not only to maintain the dignity of the individuals and the concerns but also to underline the seriousness of the transactions and so on. This formality clung to set phrases, set expressions and set formats.

Although one would not say that the considerations mentioned above do not weigh in the minds of the business-letter writers of present day English, one does not fail, however, to notice a conscious effort to underplay the formality tone. Set phrases characterize officialese, even though attempts are made once in a while to make it more informal. But these attempts at reform or refinement do not meet with any success, whether the medium is the mother tongue or is acquired as a second language. In the case of English in India, the vast majority of people who have some exposure to the medium acquire and stabilize their competence through using that language for official or formal purposes. And hence, for them, the standard is only the officialese form of the English language. To quote Professor Firth (1964) once again: 'Most Indian English is badly over-drawn. But it is kept going by the Government, and though it has therefore a certain local currency, it has no gold backing. English literature up to and including Addison is not a suitable security on which to issue current tokens of speech in the twentieth century. Babuism is not by any means confined to India. It is the common danger lurking in all purely literary education, and especially perilous if the languages are alien to the social life of the learners'. The legacy of the past lingers on!

An understanding of the syntactic usage of a language leads us to the selection and use of that sentence pattern which is more natural (in some difficult to define sense), straightforward, and frequent, and which -has sufficient elasticity, productivity and neutrality. This problem of choosing and, if necessary, coining sentence patterns for the expression of sciences is almost solved in developed languages in the sense that these languages have got some established conventions in this regard. In languages which came to be used only recently for the expression of sciences, we notice a tendency to choose and at times even coin a pattern similar in syntactic form and function to the one found in the developed language which acts as the model. Instead of looking for similarity in form and function of syntactic structures between languages, one should look for suitability of a syntactic pattern for the avowed object of communication-communicative intent.

In languages, noun-verb sentence pattern is found. Many other sentence patterns are derived from this form of sentence and its expansions are quite common in English, German and many other languages. The frequency of this pattern of sentence is found to be greater than the frequency of other sentence patterns. In all such patterns, verb has a very important role, in many cases carrying more information than the noun. Because of the pivotal role played by a verb and because of the frequent occurrence of noun and verb sentence patterns, the form of logical propositions also has assumed the same pattern. In fact if there is an -ellipsis-such as the deletion of verb in an expression (see below for further elucidation)-the expression is in some manner considered defective, and a process to recover the- verbal form is always attempted in reconstructing the logical proposition. In languages like Latin, Tamil, Hindi, Italian, Spanish, etc., the verb (verbal construction) alone can constitute a sentence, as its use will include the indication for person, number and gender of the subject noun. It will also indicate the tense of the action. Indication of the subject through the use of a full-fledged subject noun or phrase is, thus, superfluous, although the subject noun/phrase is not always deleted and is at times obligatorily retained to maintain proper references, when ambiguity may be caused by context.

In the language of science, however, only very rarely a sentence is written with only the verb phrase in these languages. In fact the language of science avoids deletions of the sort which may be acceptable in a literary genre. Technical statement is always a completed statement, and not an individual word. Hence deletions are not common. And yet both in the ordinary language and in the language of science some sentences do lack some part or parts, when compared to their full form. These are called elliptical sentences. Ellipsis is very often resorted to in ordinary spoken language, and in poetical and rhetorical language to create some effect on the audience. In colloquial speech ellipsis is resorted to not only as a measure of economy but also to create an impression of precision and brevity. It is found commonly in answers where the complete form of the answer would be a mere repetition of the major part of the question asked. Thus, ellipsis is found commonly in answers to questions, in raising questions and in exclamations.

As far as the language of science is concerned, such elliptical sentences are generally avoided. The language of science, however, makes a frequent use of an ellipsis of a different kind. Ellipsis is easily identified (and is considered a virtue) in the deletion or non-mention of the agent-in the passive construction of science language in English.

Since deletion of the agent is considered contributing to the objective and impersonal style of language of science, many other languages have also followed the practice of deleting the agent. In some languages the sentence patterns that delete the agent may be frequently used even in ordinary language; or they may be assigned some special functions such as topical focus and used in ordinary language. Hence their use in the language of science does not sound artificial or contrived. In some other languages such patterns that delete the agent may have only rare occurrence, and, following the model provided by English, scientists using such languages identify a pattern similar to that of English and make use of the same for the expression of sciences in their languages, although such a pattern may have only a rare occurrence in the ordinary language (and, thus, may sound contrived when used for a longer stretch). The peculiar quality of language of science in such cases must be ascribed to this state of condition also. Moreover, although brevity is very much sought after in the expression of sciences, the urge for precision exercises a control over the urge for brevity in the expression of sciences. This results generally in the occurrence of a specified type or a very few specified types of sentences and this fact also contributes to the contrived nature of syntax. As precision requires specificity and as specificity requires provision for all the factors and processes, sentence in the language of science gets an involved syntactic form.

7. Types of Sentence Expansions

We have referred to expansions of underlying forms of sentences. There are five different ways in which the sentences can be expanded. These are nested, self-embedding, multiple branching, left branching and right branching. A construction is considered nested if in that construction the phrase (A) falls totally within the phrase (B). Thus, the phrase The man who wrote the book that you told me about is nested in the phrase called the man who wrote the book that you told me about. A construction will be considered self-embedded if the construction (A) is of the same type as (B) and is nested in (B). An example of the self-embedding would be who the students recognized in the construction who the boy who the students recognized point out, since both are of the same type-both are relative clauses. Note that there is a significant difference between nested and self-embedded Constructions. In the former, the nested constructions need have no similarity with the one in which it is embedded, whereas in the latter the embedding is done only when the nesting and self-embedding constructions are grammatically similar. In a multiple branching construction we have no inherent structure. All the items stand in their own and are connected to a commonhead. An example would be John, Bill, Tom and several of their friends. In a left-branching structure, the construction is expanded leftward, as in the illustration John's brother's father's uncle. In the right branching construction, the expansion is done rightwards, as in the example This is the cat that caught the rat that stole the cheese.

Languages differ from one another in having and utilizing different types of expansions. As our examples given above indicate, the English language has all the five types. Tamil does not have nested, self-embedding and right-branching constructions at all. Left branching is the most frequent and the most preferred form in languages like Tamil and Hindi. In English as well as in other languages, the preference for and the frequency of these constructions vary from content to content, context to context and pursuit to pursuit.

Expansions are controlled by another factor as well-the length of the entire sentence in general and of the resultant phrase as a whole after the processes of expansion, is also a very important factor in comprehension. Greater the length the lesser is the comprehension and the retention of information. Hence, theoretically, the number of phrases, whether they be of similar or dissimilar structure, that can be packed into a sentence is infinite; but in practice, they are to be kept to the level where the speed and quality of comprehension will not be adversely affected.

Insertions are generally indicated through some formal device in all the languages. We said earlier that a complex sentence has one main clause and more dependent clauses subordinate clauses. Subordinate clauses tell us something about the nature of some other part of the sentence. This is so because the ways in which they are put in a sentence cause them to serve the information given by some other part of the sentence in a subordinate capacity. The main or major versus subordinate status of information supplied is indicated through formal means. In many languages subordinate clauses are marked through some formal device or the other. In English, these devices include the use of words such as when, since, because, if, unless, although, where, after, before, as, until, that, which, who, whom and whose. Moreover, the form a verb assumes when occurring in a subordinate clause may be different from the form it assumes while occurring in a main clause. The choice of a special word order also may be utilized by languages to signify the subordinate clause. In any case, as we have already pointed out, expansion processes may differ from language to language, and within a language also it may be conditioned by content, context, pursuit, etc. Subordination as in important process, since we do not speak or write only in simple sentences. Subordination is important also in expressions like the ones we have in sciences where comparisons are made, conditions are identified and the relations that exist between objects, events and processes are sought after.

8. Discourse

We have, on several occasions in the above pages, indicated that language can be studied in several ways-as a sociological, psychological, biological or a linguistic phenomenon. Within linguistics, we may look at language from the point of view of the eight pedagogical components, namely, phonetic, phonemic, morphological, morphophonemic, syntactic, semantic, lexical and graphemic ones. We may also study language in relation to the content, context, and pursuit of its users. A study of language, in fact, is not generally based on any one of the factors to the exclusion of others-the study generally takes into account a few of the factors together and a comprehensive study aims at including all the above in some proportionate manner dictated by the goals and necessities of the investigator.

Scientific language, unlike the ordinary language, is hardly ever emotive; it is generally informative and symbolic. But its construction and composition, as we have already pointed out, follow the processes of ordinary language. Like the sentences of ordinary language, sentences of the language of science also may have qualities that may not be wholly ascribed to the words that the sentences have. Composition of sentences in all languages involves selection of- words and selection of arrangements into which these words are to be put. One cannot say for certain whether the arrangement or the choice of words is decided upon first. In any case, there are several alternatives of arrangements and of words available to the users of a language. This availability of alternatives forces the users to make a decision, and the capacity for making appropriate decisions varies from one user to another. It also varies on the basis of the content, context and pursuit. Note, however, that alternatives as regards the items of language of science are of a limited number and nature.

We do not speak or write sentences in isolation. The sentences we utter or write are generally a part of a larger whole called discourses cluster of sentences with a focus on a single topic, generally speaking, which is intuitively felt as forming a unit by itself. Thus, a sentence has not only its own internal network of relations formed by the words, phrases, and even embedded sentences that go into its making but also it has its own external relations when it occurs in a particular discourse. These external relations between sentences in a text or discourse are signalled both by linguistic and nonlinguistic means. Several linguistic devices are employed by us to refer back to what has been already suggested, to suggest that what is being presented in a particular sentence is a logical follow-up of what has been presented in preceding sentences, to point out what is being presented in a particular sentence is simply going to be contradicted later, and so on.

Coherence and cogency, readability and to some extent comprehensibility may be dependent both on the linguistic and nonlinguistic characteristics of the discourse. Thus, the external relations of the above sort that exist between sentences in a scientific text should also be considered as part of the characteristics of the language of sciences, in addition to a study of the relations between syntax and lexis (lexical items and their use). A study of scientific language will be enriched not only -by a comparison of one scientific language with another, one functional language with another, but also by a comparison of the same with the spoken/colloquial language. That is, the language of science should be compared within a language across levels of communication and levels of education. Such comparisons will lead to insights into the functions attached to the varieties of a language and this insight should be of great use for the cultivation and enrichment of one's own mother tongue. Be as it may, we shall now present some of the characteristics of scientific discourse.

In our description of sentential structure, we used a term, namely, subordinate clause. The subordinate clause is an internal relation and, thus, contains information which is subsumed by the information carried by the main clause of the whole sentence. Generally speaking, it is the information that is carried by the main clause which is used to establish the communication bridge with other sentences in a discourse. However, while the main clause of a sentence carries the information and acts as the bridge in most of the cases, the subordinate clause is not precluded from having such relations in the discourse. In all the cases, however, a complete understanding is achieved in a discourse on the basis of the understanding of other sentences of the discourse. This is much more so in a scientific text because sentence in a scientific text is there in the text because of the information it carries and not for any ornamental purpose. The sentence has its own information which, however, forms part of the whole in some specified and clear manner.

Discourse analysis is based on the assumption that we can identify and isolate from the sentences that comprise a text or speech the linguistic features which give a coherence to the entire piece. In fact stronger claims are made in some quarters that linguistics must give up its sentence - based grammar approach and switch over to a study of the discourse and an enunciation of discourse grammar. While this proposition continues to be a bone of contention among schools of linguistics, it is true that taking a discourse view of the linguistic aspects of any communication event will bring out characteristics that may not be accessible if we restrict our observation to individual sentences only.

A discourse has something coherent about it. Some of the features that contribute to the cohesion of a text or speech are juxtaposition of sentences or parts of sentences; the presuppositions shared by a writer and a reader, a speaker and a listener; repetition of items; the use of synonymous items; anaphoric references (explained below), and so on. Juxtaposition of sentences by itself does not contribute to cohesion of a text. One can produce a large number of sentences and place them one after another in a text without bringing in, in any significant manner, cohesion to the text. The sentences may mean one thing when viewed as separate and disconnected sentences. But as a text they may not convey anything in an orderly fashion.

We juxtapose sentences in a discourse on the basis of connection between propositions. Juxtaposition may be viewed from two angles, namely, juxtaposition of parts of a sentence and juxtaposition of sentences themselves. The grammatical concepts of co-ordination and subordination may be considered as part of juxtaposition of parts of sentences (internal relations as explained above). In co-ordination we juxtapose two or more identical grammatical structures. Conjunction plays a crucial role here which connects parts of the sentence.

We can also have apposition as in the following examples: Queen Mary, Professor Higgins, and Nehru, the author of The Discovery of India, in which both the components may be considered attributes of ' each other. Subordination helps us to build up the qualifications of the head in addition to fulfilling several other functions, some of which we have referred to earlier.

Conjunction between sentences is an important step in bringing out the cohesion between sentences in a text. This connection is found to be of several types in a scientific text in English. The most frequent ones are termed as the logical sequence, (e.g., thus, therefore), the contrast versus non-contrast, (e.g., however, in fact), the doubt versus certainty, (e.g., probably versus certainly), the non -contrast versus contrast, (e.g., moreover, similarly, also), and the expansion of detail (e.g., for example, in particular). For using the logical sequence-connectives, sentences should be 'in a particular time sequence or deduction sequence for them to be used in making their relation explicit'.

The study by Winter and others (quoted in Winter, 1971) has indicated that within this sequence of sentences instrument relation (thus) and the deduction or reason relation (therefore) have a greater frequency in scientific texts. The contrast sentence connectives, are used to make explicit a contrast or an incompatibility between their clauses. These include the items, however, in fact, yet and nevertheless in English. Out of these items, the frequency of however was found way ahead of the rest.

The sentence connectives of doubt and certainty include the following: The connectives expressing doubt are probably, apparently, perhaps and presumably. Connectives expressing certainty are of course, clearly and actually. These sentence connectives are non-anaphoric (not having back-reference, generally speaking) and 'they do not make explicit a particular clause relation for their sentence, but can strongly predict a contrast relation for it. What they do make explicit is a subjective attitude of doubt or certainty and it is this subjectivity itself which can predict a contrast relation as an intellectual balancing of statements where tentative is balanced by certainty, or certainty by tentative', (Winter, 1971).

The sentence connectives of non-contrast are the semantic converse of the contrast items in the category of the contrast sentence connectives. These 'make explicit a relation of non-contrast or compatibility.' These include items such as moreover, similarity, also, indeed and in addition. If, however, the connectives establish the links between sentences that follow one another and also the links between parts of a sentence, the theme of a sentence that follows another should not be viewed merely as an addition or an expansion even of the theme of a preceding sentence. It is true that a succeeding sentence will add to and expand the information/theme content of the preceding sentence/sentences. And yet, the significance of a succeeding sentence lies mainly in its function/role of 'moving forward the communicated content'.

It is where the language use in science aims at excellence and steals a march over the ordinary language use. This moving forward the communicative content is an absolute requirement in language of science, whereas in other uses of language such a primary importance to this aspect need not be given. In this process of moving forward the communicated content, the first position in a sentence is given to some element which is about to be emphasized. This is a general practice in language use in the expression of sciences in known languages. And this general practice or rather the compulsion of the profession at times leads to a choice of a structure which may be quite natural and appropriate in one language and contrived and artificial in another.

In several places in this chapter we used terms such as anaphora, anaphoric references and back references. Anaphora is an important aspect of discourse. It may be defined as a type of dependency in a sentence when one of its elements refers back to an earlier sentence (or information contained therein) in some manner. Even within the same sentence one can have anaphoric reference. Consider the following example:

1. The boy hurt the boy.
2. The boy hurt himself.

When the phrase the boy given in the sentence in the predicate phrase position is meant to refer to the boy given as the subject phrase, it is automatically transformed into a reflexive pronoun and, thus, we get the second sentence. Reflexives are generally anaphoric within a sentence. One can refer back to an earlier sentence through the use of pronouns, articles, demonstratives, adverbials of place, and through a process of ellipsis. John in an earlier sentence can be referred to as He in the subsequent sentences. Consider the following:

Question: Who is coming?

Reply: John!

Here a process of ellipsis takes place and a mere utterance of the name with an emphatic-assertive intonation indicates that the action referred to in the question is being performed by the person named in the reply. Furthermore, one easily notices in a discourse certain markers in the form of words, intonations, interjections, affixes and so on, which indicate not only the continuity of the episode under narration, but also refer back to events mentioned in earlier sentences. It is noticed that certain forms and kinds of discourses may exhibit some preference for certain kinds of anaphoric devices over others.

The maintenance of internal cohesion of any discourse is necessitated first of all by the need to be understood. The dependencies must be viewed from this angle. If one resorts to an excessive use of anaphoric devices, it is likely to affect comprehension. As communication is both inter-personal and intra-personal, we must use the indicators in appropriate ways if we want to be understood. In some languages like Tamil, Latin, Spanish, Hindi and so on, the main verb of the predicate phrase may be considered as having anaphoric reference to the subject of the sentence, in an extreme sense. The main verb in Tamil is inflected for gender, number and person, and as a consequence, can occur without the subject being mentioned at all in a sentence in the course of a text. But this has its own limitations. The identity of the dramatis personae is not complete only with the identification of their gender, person and number. Their exact identity as persons in a context of situation requires much more than these and hence a discourse soon reveals the identity of the subject in as many respects as possible. Otherwise the communication bridge is bound to be broken.

There are a few other features such as presupposition, paraphrase and lexical organization of sentences which also contribute to the internal cohesion in a discourse. Presuppositions are carried by certain statements, question, etc. "Has John stopped beating his wife?" presupposes that John has one and has beaten her. Presupposition need not be considered only from the angle of shared information between the communicators but also from the angle it gets reflected in linguistic structures. The chosen syntactic patterns and lexical items combine to express the presuppositions of the writer and the speaker. The linguistic philosophers, devote considerable attention to the problem of presupposition and expend their energies to separate the content-presupposition from the linguistic presupposition. Linguists also expend a considerable amount of their energies to debate the extent to which the implications of the notion of presupposition should form part of the linguistic description of a language. Be that as it may, in some specified sense, the scientists are required not to "presuppose" and to make "presupposed statements". They are required to make what is being presupposed as explicit as possible! Hence the language of the scientist is expected to be explicit and straightforward. Paraphrase enables us to use, in the place of using pronouns for subsequent references, noun phrases, which can hold paraphrase relations with the subject of the discourse. This technique of cohesion or, rather, a stylistic device, is generally avoided in the language of science, unless the writer prefers to highlight certain particular aspects of the subject referred to. The connections between lexical items in successive sentences also bring in cohesion in a discourse:

Waikiki. The sand was wet. The crabs were digging. The shells were all around. At a distance was seen a ship.

Note that in the above passage there is no explicit syntactic device to indicate the cohesion between sentences. But certain words give a sense of continuation and the flow from one sentence to another is marked by the choice of lexical items from more or less the same set. This example shows that anaphoric references need not be identified only at the syntactic level through an identification of linking morphemes, phrases, subordinate clauses, etc. There is yet another anaphora, semantic anaphora, which is exemplified in the use of items from the same set. It emphasizes the need to look at the back reference phenomenon of a discourse from the semantic point of view. In fact, there are suggestions that we must posit a semantic deep structure and a surface structure for the discourse. The deep structure will be the chain of events of a narrative (in our case the conception of theory, assumptions, original motivating force and even the design of experiment and anticipated solution) and the surface structure will be the manifestation of events in linguistic structures of a language (in our case, the reporting). The events and the surface structure need not have the same sequence and be in one-to-one correspondence. Languages may differ from one another in the way the underlying semantic deep structure are organized in the surface structure of the discourse.

An important segment of discourse is paragraph. Some consider it only as a form of punctuation, having only the function of giving relief to the tiring eye. But a paragraph has more to it than meets the eye. An ideal paragraph has several functions to perform which include, among other things, its usefulness to indicate the various stages in an exposition, as a prop to the reader to comprehend and retain information and better organize what is being comprehended and retained for easy retrieval, and its usefulness in moving forward the communicative content. If the thought is coherent, paragraph will have greater coherence. But even when the thought is coherent, if it is not matched by appropriate linguistic structures, the paragraph may not reveal any coherence at all. Word links alone cannot give a sense of coherence to the content of a paragraph-especially so in a scientific text. There is a general theme which holds together in some perceptible manner all the sentences that constitute a paragraph. Placement of the theme or focus is dependent upon the goal of the concerned paragraph. If elucidation or enunciation is the goal, the theme sentence may be placed in the beginning of a paragraph. If conclusion or deduction is the goal, the theme sentence may occur at the end of a paragraph. Paragraph exercises demand from the students giving titles for each paragraph of a text presented to them because of the assumption that a paragraph is a coherent whole in itself focussing on a particular point.

The characteristic features listed above are those of an ideal paragraph and such ideal paragraphs need not be written by all or by an author on all occasions. Spotted brilliance is rather a general rule. And yet it is absolutely necessary that there be congruity between the thought expressed and the medium used to express that thought. It is also necessary that there be coherence within the thought expressed and within the linguistic constituents used to express the thought.

The inter-relationship between thought and language is an absorbing subject of investigation in its own right with linguists like Bloomfield taking the position that thought is nothing but the subvocal behaviour which has the progression from oral speech to whisper to underground speech-thought, with linguists like Sapir emphasizing the separate identity of the two while highlighting the fusion taking place in higher order processes of reasoning which places language on a paramount, with epistemologists like Piaget likening the relationship between language and thought to the banks of a river and the river itself while emphasizing the separate identity, separate origins, different functions and developmental processes of language and thought, and with psychologists like Vygotsky emphasizing the social factors and necessities in the emergence of thought and language with a progression from socialized speech to egocentric speech to inner speech. With such differing views and others, the interrelationship receives a good deal of attention of all. Like any phenomenon we meet with in this universe, thought in general, and a thought in particular, must have its own structure.

The view an author, a scientist or a writer takes about the structure of a particular thought should be considered as revealed in his paragraph organization. Thus, while the importance of thinking processes (in conjunction with or without language) for the elucidation, creation and interpretation of sciences cannot be exaggerated, the importance of language, through which, and through which only, scientific ideas are expressed and understood, cannot be minimized.

Rhetoric is the discipline which studies the interplay of thought and language in actual performance. The general characteristics/features of discourse in a scientific text clearly fall within the scope of rhetoric. Hence it will be useful to see the methods of rhetoric.

Rhetoric aims at a theory of how and' why one communicates and at a theory of processes involved in mastering communication. All the living organisms have an urge or an impulse inherent in them to communicate. Their survival within their own species and in the universe depends upon the use of their steering mechanism-communicative tool. How this communicative tool is exploited to achieve the ends an organism sets before himself may be considered falling within the scope of rhetoric. Thus, throughout ages, rhetoric has been conceived as an art of effective expression, as an art of persuasion and as a discipline concerned with the problems of communicating truth and value. This is, indeed, an ancient discipline whose scope included what one may call ancient rhetoric principles, and whose present scope includes modern scientific studies of language and communication.

It is clear to us that in a specific event of communication things do not go always as explicit as they should have, or things do not make the effects for which they were intended. Certain strategies adopted for communication purposes may prove to be more effective than others, on the basis of the goals of communication one sets before himself. Rhetoric aims at studying these processes.

One begins the study of rhetoric with an organism's impulse to communicate with others and proceed from there to the characteristics of audience and strategies to present the ideas. If the written medium is adopted, writing or rewriting processes must be identified. However, the ultimate sources/origins are generally difficult to identify. Along with this uncertainty or impossibility of identifying the ultimate mental sources as to how an idea originated, one is also cautioned not to treat the processes we identify and the stages we posit for the progression of the communicating event as linear in any strict sense.

Before something is said or written in a situation, very often one weighs in his mind the various alternatives available. Though investigation or rather intelligent guessing of what goes on in an organism's mind and what processes an organism employs consciously is made, the major focus of rhetoric studies is on how this intimately personal experience is conveyed to the audience and in the written or spoken word. Speaking and writing are productive processes. In speaking there is a real and immediate presence of an audience; in writing as well as in speaking through mass media like Radio and TV, the audience is assumed. Rhetoric enables us to recognize, identify, describe, explain and characteristics of this assumed audience. To a scientist, rhetoric makes explicit the processes that a writer may or can profitably utilize in the exposition of ideas. The principles and methods of rhetoric (of various schools) closely resemble the principles and methods of scientific inquiry which we elucidated elsewhere.

The characteristics identified in any "ideal" or "good" work, written or spoken, through rhetoric analysis will be the same in both scientific and other materials. Science itself is an art of persuasion-persuading others to accept what is being presented through certain well-laid procedures. Because of this reason, language of science can be also studied from the point of view of rhetoric analysis. Differences in age, sex, social role and several other factors have a bearing upon the interpretation we make of the same event. Rhetoric aims at identifying how one can overcome such hampering factors and arrive at an appropriate interpretation of the event. To that extent, aims of rhetoric closely resemble those of sciences, and rhetoric shares with sciences a long list of problems.

Usefulness of rhetoric analysis for an understanding of the language of science will become more apparent if we consider the point that although expression of sciences is grounded in ordinary language, and that although development of sciences depends upon how the developments have been worded, the concepts themselves may have an existence .of their own. As Einar Haugen, a great American linguist points out, 'the development of mathematics may be seen as an attempt to overcome the weaknesses of natural languages for the purpose of exact and elegant statement. The terminology of science has been successful to the extent that it has been able to surmount the limitations imposed by natural language and produce an inter- and super- linguistic language.... The great discoveries and inventions of Western science have not been expressed in the Western languages at all, but in the special language of mathematics. There is nothing in English as such that-enables us to talk about relativity or atomic energy or the double helix. The popular accounts that most of us read are only approximations to the underlying theories, and the English in which we road them is a new terminology which can be translated precisely into any language regardless of its grammar, syntax or phonology but which in any language, including English, requires some distortion of the ideas involved because these are not precisely expressible in any natural language,' (Haugen, 1977). One may not fully agree with what Haugen says, but it is precisely to hold to the minimum the inevitable distortion element and to get beyond this distortion in retrieval processes, the principles and techniques of rhetoric will be greatly useful. Einstein suggested that science searches for relations which are thought to exist independently of the searching individual. Rhetoric aims precisely at the same outcome.

Some classify rhetoric into macro-rhetoric and micro-rhetoric. The macro-rhetoric covers inter-relationship between writer or speaker, purpose, intent, text, audience and the immediate and historical contexts, and so on. The micro-rhetoric covers formal texture of a work including structures of words, sentences, paragraphs and other devices, which one may call loosely as linguistic devices. Such a division is rather arbitrary, for one needs both macro-and micro-rhetoric to understand the wholesome process of communication. One may also venture to say that while macro-rhetoric covers the phenomena in a general manner, micro-rhetoric covers the communication processes within a specified linguistic group. Even then it is needless to say that the particular' falls into a general pattern and that the general pattern has its existence because of the particular, although it exercises some influence over the particular.

In ancient times the rhetoric was divided into five units and discussed. These included invention, arrangement, style, memory and delivery. Invention is related to 'logic/dialectic which order the process of finding what might be said about a given subject to a given audience'. Arrangement referred to 'the formula for organization of the whole composition based on the divisions of the classical oration'. Style consisted of 'the resources of language-sound and rhythm-and of classified figures of speech and thought'. Memory and delivery drew attention to voice and gesture and provided a 'mnemotechnical system for placing and holding received knowledge', (Piché, 1970).

Modern times saw a revolt against the study of rhetoric in the school system, just as there was a revolt against the study of grammar for grammar's sake in the schools. Emphasis was shifted in early modern period to a study of rules for topic sentences, achieving unity within a discourse, coherence and on mastering several forms of discourse such as description, narration, exposition and argument. That is, formal aspects of a text and processes to achieve these formal aspects came to be emphasized rather than the philosophy of rhetoric as a science of communication. However, this phase is slowly giving way to a better appreciation of the usefulness of techniques of rhetoric to reveal the intimate relationship between language use and thought, to create and maintain enlightened communicative competence and to retrieve the intended information from a mass of misunderstanding.

I.A. Richards, who leads one of the modern schools of rhetoric, advances the Mosaic Usage Doctrine. According to this doctrine any verbal act may be considered a mosaic with pieces of fixed shape and colour-here, linguistic items. Each piece has its placement and use in the mosaic. From this fixed mosaic one can view the shifts caused by various factors of language and experience. To quote Richards, 'understanding them (the meanings) is seeing how the varied possible meanings hang together, which of them depend upon what else, how and why the meanings which matter most to us form a part of our world-seeing thereby most clearly what our world is and what we are who are building it to live in' (Richards, 1942). Richards posits four aspects in the understanding phenomenon-sense (what the speaker actually says), feeling (his attitude toward what he is talking), tone (his attitude toward his hearers and/or readers) and intention (his conscious or unconscious aim). The speculative equipments for comprehending, according to Richards, include indicating, realizing, valuing, influencing, controlling and purposing. All these are involved in any one event of comprehending.

Another modern model of rhetoric can be identified in the works of an American author, namely, Burke (Burke, 1945 and 1955). According to him, rhetoric is rooted in the use of language as a symbolic means of inducing co-operation in beings that by nature respond to symbols. Identification is the key term in his concept. With identification, an organism shifts the focus of the study from persuasion to co-operation based on symbolic action. Burke's model is a dramatistic model. It emphasizes that we need the word that names the scene (the background of the act, situation in which it occurred), the word to indicate what person or kind of person (agent) performed the act, what means or instruments he used (agency), and the purpose. The interaction between all these forms part of the rhetoric investigation.

There are also several other scholars or schools of rhetoric. Some discuss issues from a micro-rhetoric view, adding up to a macro-rhetoric in stages. Richard Weaver (1953) echoes Plato when he looks at rhetoric as the intellectual love of the good and something which shows a better vision. Christensen (1967) suggests four major principles or characteristics for modern English prose. He considers 'composition essentially a process of addition. This addition shifts our attention to the role of modifier. But then it is not the modifiers that carry forward the communication. Both addition and the principle of direction of modification are identified in cumulative sentences. The principle of generality implies that downshifting or back tracting done by the modifiers is in the direction of lower levels of abstraction. The fourth principle, texture, contrasts bare unmodified predications with those of sentences whose modifications or texture is both tense and varied' (from Piché, 1970).

A team of scholars including Becker, Young and Christensen, following the tagmemic school of linguistics propounded by Kenneth L. Pike, offers promising trends in modern rhetoric. They present certain maxims on the basis of which rhetoric analysis of text is proposed to be undertaken. The first maxim is that people conceive of the world in terms of repeatable units. It is true that there may not be exact repetition of environments; inputs and personae may be different. And yet we all act and accept a degree of perceptual constancy. This perceptual constancy is embodied in language through names for objects, processes and events. The second maxim states that units of experience are hierarchically structured systems. We all segment experience into discrete and repeatable units. In segmenting and ordering we may differ from one another, although communication strives towards the retention of the perceptual constancy. We may raise several questions as to whether we are focussing on different parts of the same whole, do the different parts have the same nucleus, what is the level of magnification, what kinds of details are relevant to understanding an experience, what kinds of details will be interesting to the one with whom we are communicating and so on. However, it is easier said than done-what is relevant and what is not relevant is not easy to decide and define. To come to grips with this, one should know the contrastive features (why it is an X and not something else), range of variations and its distribution in larger contexts. Thus, the third maxim states that in order to understand a unit, we should know its contrastive features, its range of variation and its distribution in larger contexts.

The rhetoric analysis influenced by Pike's thinking suggests four stages. These are similar to the ones that a scientist undergoes in his scientific work. The four stages are preparation, incubation, illumination and verification. The preparation stage of the process of inquiry includes initial awareness of a difficulty, formulation of the difficulty as a problem and exploration of the problem. These are all put into a form of languages progression from feeling to conscious analysis is accompanied by putting the idea in a language form. The preparation stage is followed by a period of subconscious activity-incubation stage. In the illumination stage, hypotheses are formed. But note that the relevance, appropriateness and significance of the hypotheses framed are generally in direct proportion to the knowledge of the subject possessed by the investigator. The last stage aims at finding out on the basis of the results as to whether the hypothesis in fact provides an adequate solution to the initial problem. Once inadequacy is revealed, the process of inquiry begins again.

How does one become an effective inquirer? 'To become an effective inquirer it is essential that you develop sensitivity and receptivity to problematic situations' (Young, el al., 1970). When we make a statement two basic components, namely, a problematic situation and an unknown must be made explicit. 'A statement of the unknown is actually a partial description of the solution'. It has two functions: it acts as a guide to inquiry since it describes what we are looking for, and, further, it enables us to know when we have found our solution, since the solution will match the description. The model of the scholars influenced by Pike's thinking advocates that 'a single problem is in fact a cluster of interdependent subordinate problems each of which must be solved before a solution to the larger one can be found'. Discovering the unknown is done through 'successively classifying and reclassifying it as a question of fact, of process, and of relationships.

1. "A question of fact requires that the answer isolate and identify something (who? what? where?)
2. "A question of process requires that answer describe an activity or prescribe a set of operations (How did it happen? How can it be done?).
3. "A question of relationships always involved systems or units in systems. Questions of relationships include:
4. 1. "value questions which ask about the relative positions of units in a value system (what is better?)
   2. "questions of causation and probability (what caused it ? what is its probable effect? which is more likely ?),
   3. "questions of consistency, such as questions of logic and classification (Does the conclusion follow from the evidence ? Do the units have enough in common to warrant being grouped in a single category), and,
   4. "questions of policy (what should be done?)," (Young, et al., 1970).

The fourth maxim suggests that a unit of experience be viewed as a particles static view, or as a wave a dynamic view, or as a field-a network view, or as all three. The fifth maxim characterizes the rationale for rhetoric. It suggests that change between units occur only over a bridge of shared features. Shared features are insisted upon as prerequisites they fall under various categories such as reader's knowledge, reader's values and reader's social relationships. Constraints are imposed upon the writer, particularly on the writers of materials connected with intellectual pursuits as opposed to literary pursuits. (This is not to say that literary pursuit is not of an intellectual type what we emphasize here is the discipline-based knowledge such as those expressed through sciences). Compare the facility in literature, where novelty in form and expression is highly valued, with scientific writing, where conservation, adherence to a single standardized form and expression, is highly valued. What will happen if the freedom of creativity as viewed in literature is exercised by scientists while writing their scientific reports? Can creation of new forms (genres) happen in science writing? Can we express scientific concepts in different forms of literature? What is the extent of flexibility and rigidity in this regard in scientific writing?) Bridges of communication which we seek in scientific reporting have their own features. And in the choice of appropriate features lies the success of the scientist-writer. But note that choice does not imply any sacrifice of the crucial points to be conveyed. A change in the reader is sought to be achieved with the help of, and not ignoring or sacrificing, points at issue.

We give below a programmatic outline suggested by the above scholars as a prop for better communication:

Developing the Message (pp. 237-238)

1. Introduction
2. 1. Direct the reader's attention to the subject or problem.
   2. Explain your experience with it.
3. Background
4. 1. Explain the nature of the problem-its history and causes.
   2. Explain why the problem is important to the reader.
5. Thesis and Explanation
6. 1. State the thesis.
   2. Develop it.
7. Conclusion
8. 1. Explain the implications (social, philosophical, psychological, and so on) of the information.
   2. Summarize your discussion: the problem (2a), your thesis (3a), and your explanation (3b).

Variables: The nature of the message, its length and complexity, the extent of the reader's relevant knowledge, his knowledge of you, and his attitude to you.

For First Draft (pp. 234-235)

1. Introduction
2. 1. Direct the reader's attention to the subject or problem.
   2. Explain your experience with the subject, the reasons why you can write with authority.
   3. Establish bridges with the reader by pointing out shared beliefs, attitudes, and experiences.
3. Background
4. 1. Explain -the nature of the problem-its history and causes.
   2. Explain its relevance to the reader's problems, desires, and interests-the reasons why the problem is important to him.
5. Argument
6. 1. State the major premise. Include any information that is necessary for making it clear and acceptable.
   2. State the minor premise. Include any information that is necessary for making it clear and acceptable. (It is usually the minor premise that needs the most substantial support. Cite authoritative statements, facts, statistics, personal experiences and experiences of others and so on).
   3. State your conclusion.
   4. Demonstrate the superiority of your position by pointing out defects in the inferences or premises of alternative positions. Explain why the alternatives cannot solve the problem; or if they can, why your solution solves it better.
7. Conclusion
8. 1. Explain the implications of the argument, such as the benefits to the reader of accepting it and the undesirable consequences of rejecting it.
   2. Summarizing your argument: the problem (2a), your conclusion (3c), and the reasons for accepting it (3a and 3b).

The sixth maxim suggests that linguistic choices are made in relation to a universe of discourse -from a set of possibilities which in their turn are governed by inter-relationships between sets of possibilities for units. If choice is possible, naturally correction of a choice is also possible. There is a progressive perfection or editing of the choices to arrive at the most appropriate and effective. There is bound to be a norm once the concept of choice is accepted. The deviation from the norm would be of a wide range, the restriction being the retention of contrastive features of an absolute nature.

What is emphasized in a rhetoric analysis of the type we have discussed above is the structural network of whatever is written or spoken, whatever is read or listened to. The structural network begins with the exploitation of shared features, establishing bridges of communication through discovery of these features for effecting a change in the reader or the listener. The processes employed closely resemble those of the techniques of scientific inquiry. It is emphasized that every piece of writing is composed of a hierarchy of patterns. 'Related sentences form paragraphs, related paragraphs form units of a still higher level-sections of essays, complete essays, chapters of books'. And finally the overall pattern is the mental image of the whole structure of an author's, or a scientist's work -'an image that may be clear in his mind when he begins to write or that may evolve gradually through repeated trials made during the act o writing'. The sections on interpretation of data, the deductions from observed phenomena, the' suggested implications for a variety of data and the proposals for future research we find in scientific articles and books all come under this category.

10. Style Sheets

To recapitulate, style is a manner as opposed to matter. Some do not recognize or find any use for the concept of style. In one approach to the study of style, the relationship between the writer and the text forms the basis. This amounts to the characterization of style through a study of the writer's personality and the circumstances in which the text was written. In, another approach, one focuses on the relationship between the reader and the book in which case the responses of the reader become the basis of stylistic analysis. In the third approach we may not worry about the writer, but concentrate only on the work. In yet another approach, style is considered as the result of a choice one makes from among many possibilities. Even within linguistic approaches, we have three types, namely, the norm and the deviation, the addition, and the connotation. Addition of stylistic features is assumed to be performed on neutral expression. The environments in which the linguistic features acquire stylistic values form the subject matter of the connotation approach. Stylistic analysis has been closely associated with literary studies. These studies aim at establishing links between aesthetic responses in the reader and the features of the work. Thus, there are many ways pursued in stylistic investigations. But then what is style? Stylistic variants are generally held to be equivalent ways of saying the same thing. And yet each form of expression has its own style, as already indicated. Science has its own style and within this scientific style, we have, for each discipline of the scientific pursuit, a substyle. The outlines-the formal aspects of these substyles -are generally reflected in the style sheets of journals. These style sheets are, indeed, a combination of the substyle as well as the individual variations journals impose to highlight, from their points of view, the important aspects of reporting.

As stated above, style may be considered a manner as opposed to matter--the ways in which a writer or speaker presents what he has to say to his readers or listeners. Is this variation necessary? To achieve what ends and to create what effects this facility is put into use? Making stylistic variations is a characteristic that we notice in all of us. But is this to be resorted to, if our aim is simply to convey information-pure, simple and bare information? Such a question has been often raised and in the history of expression of sciences, one always notices a tendency not to be "unduly" stylistic in expressing one's thoughts. Certain norms have been laid to bring forth some semblance of order and homogeneity of expressions assumed to be in tune with the assumptions and practices of scientific inquiries, to maintain the tenor of scientific discussions and to enable the readers to correctly grasp the meaning of what is being presented. In spite of such strong tendencies towards order and homogeneity, stylistic variations are easily noticeable between the writings of various scientists and also between the writings of the same scientist depending upon various factors. However, such variations are found more in books than in articles by scientists.

The style sheets of the journals make order and homogeneity their main goal-the nature of order and homogeneity of expressions and format are decided upon by the constraints imposed by matter and by the practice, with reason or no reason, followed by the journal. The restrictions imposed range from the length of the article to the presentation of tables, citation and presentation of references, spellings of words, use of certain phrases, the order of progression from one segment or section to another, rules for quoting, etc., to even (very rarely) the sentence patterns to be used or avoided and content included. Within such restrictions also, an ingeneous writer is given wide scope to make innovations and even embellishments in his manner of expression. It should also be noted again that a style sheet lists sentence patterns that should be or should not be used. Some style sheets ask the writers to avoid certain phrases and expressions but they do not list the sentence patterns that should be avoided. It is the unwritten convention that makes a writer to prefer or not to prefer specified sentence patterns.

*** *** ***

REFERENCES

Bach, E. 1974. Syntactic Theory. Rhinehart and Winston, Inc., New York.

Burke, K. 1945. A Grammar of Motives. Prentice-Hall, New York.

Burke, K. A Rhetoric of Motives. George Braziller, Inc. New York.

Chomsky, N. 1965. Aspects of the Theory of Syntax. M. I. T. Press, Cambridge, Mass.

Christenson, F. 1967. Notes Toward a New Rhetoric. Harper and Row, New York.

Firth, J. R. 1964. The Tongues of Men and Speech. Oxford University Press, London.

Haugen, E. 1977. Linguistic Relativity: Myths and Methods, in McCormack, W. C. and Wurm, S. A. (eds.), Language and Thought: Anthropological Issues. Mouton, The Hague.

Hornby, A. S., Gatenby, E. V., and Wakefield, H. 1963.The Advanced Learner's Dictionary of Current English. Second Edition. Oxford University Press, London.

Jacobs, R. A., and Rosenbaum P. S. 1968. English Transformational Grammar. Blaisdell, Waltham, Mass.

Onions, C. T. 1932. An Advanced English Syntax. Sixth Edition. Routledge and Kegan Paul, London.

Piché, G. L. Language and Composition: Some "New" Rhetorical Perspectives. In Marckwardt, A. H. (ed.), Linguistics in National Programs. The National Society for the Study of Education, Chicago, IL.

Richards, I. A. 1942. How to Read a Page. W. W. Norton and Co., New York.

Weaver, R. 1953. The Ethics of Rhetoric. Henry Regnery Co., Chicago, IL.

Young, R. E., Becker, A. L., and Pike, K. L. 1970. Rhetoric: Discovery and Change. Harcourt, Brace and World, Inc., New York.

Winter, E. O. 1971. Connection in Science Material: A Proposition about the Semantics of Clause Relations, in Science and Technology in a Second Language. C. I. L. T. (Reports and Papers 7), London.

HOME PAGE | A Contrastive Analysis of Hindi and Malayalam | The Language Education Policy of Gujarat Government | Mean Length of Utterance and Syntactic Complexity in the Speech of Mentally Retarded | Towards an Understanding of Sentence in Science in an Indian Language Context | CONTACT EDITOR