Theoretical and Pedagogical Grammars
In one of his recent articles Chomsky (1981b) talks about two major shifts of focus: the first was a shift of focus from language to grammar which started in 1955 with the publication of "The logical structure of linguistic theory" and spanned over a period of twenty-five years, before it found a theoretical formulation in Chomsky's book 'Rules and Representations'. The second shift of focus was from rule-systems, as explicated in 'Syntactic Structures' (1975) and 'Aspects of a Theory of Syntax' (1965), to system of principles such as the Binding Theory and the Projection Principle ('Lectures on Government and Binding', 1981a).
The first of the two shifts of focus mentioned above, i.e., from language to grammar, considers language as derivative and secondary in status. According to Chomsky the notion of language, as compared to the notion of grammar, is a more abstract one, and is even epiphenomenal in existence. For instance, when we talk about Hindi language, we are led to regional and sociolectal variants, standard and non-standard forms, different styles and registers, etc. Since all these are covered under the rubric of language (Hindi), one has to generalize at different levels, on the basis of these variable data, in order to arrive at the notion of a language.
When talking about Hindi, for instance one can have three levels of generalizations (a) Hindi, which ahs in its ambit the different dialects, e.g., Braj (with literary giants like Surdas and Bihari), Avadhi (with literary geniuses like Jayasi and Tulsi Das), and Khari Boli (with literary artists like Jai Shankar Prasad and Nirala). These dialects collectively constitute the institutional reality of Hindi language; (b) a narrower view of Hindi, which embraces the regionally unmarked koine of India, and functionally operates as a link language, a sort of pan-Indian language; and (c) the view of Hindi as a cultivated form of Khari Boli which functions as the interlanguage for its dialect speakers, i.e., as an associate mother tongue. Even in the narrowest sense, Hindi is not free of its institutionalized language styles such as High Hindi (based on the tatsam lexical resources of Sanskrit language), Cultivated Urdu based on the tatsam Perso-Arabic Lexical resources), and Hindustani which is based on the tadbhav native lexical resources. The situation can be illustrated schematically as follows:
PASTE THE SCANNED IMAGE
In the above schema:
ABX=Hindustani with the base of Hindi (a variety which is acquired and not learnt);
ABC=Hindi with ACX as the cultivated, learned variety;
ABD=Urdu with BDX as the cultivated, learned variety;
and
ABCD=The total range of Hindi, provided one considers Urdu to be linguistically and stylistically a variant of Hindi.In the context of the foregoing, it is interesting to note what Chomsky has to say on the notion of Language:
In fact the notion Language night turn out just to be a useless notion. For example, if we fix a certain level of acceptability, then this internally represented system of grammar generates one set, and we say O.K. that is the language. If we fix the level a bit differently, the same grammar generates a different set, and we can say that this is the language. There is no meaningful answer to the question: Which is the real Language? (Chomsky in Huybregts and Riemsdijk, 1982 : 107).
The second shift of focus, from rule-systems to systems of principles, required the notion of 'Core grammar' and the concept of extensions and modifications of such a grammar. In the initial phase of his transformational generative model Chomsky defined Universal Grammar in terms of systems of rules wherein the system was a reflection of the biological endowments of the human mind. These biological endowments were regarded, then, as a function that maps a body of experience into a particular grammar: the function itself was commonly called 'Universal Grammar'. In the latest phase of the development of Chomsky's theory, Universal Grammar is said to be expressed in part, as a system of principles known as 'Core Grammar' and, in part, as a theory of permissible extensions and modifications of the Core Grammar.
According to Chomsky, the Core Grammar is that component of the human mind-brain which is genetically determined in its initial configuration, and this may be called the initial state of language faculty. It is distinctly marked for two parts of theory - (a) a theory of principles and parameters, and (b) a theory of markedness. The latter makes it possible to incorporate more of the idiosyncratic elements. Although idiosyncratic elements themselves are controlled by innate principles of the grammar, it is through an interaction between the theory of principles and the theory of markedness that the initial state of language faculty is transformed into a more or less steady state leading to the notion of full grammar.
This second shift of focus from rule systems to systems of principles, as a consequence, makes most of the results of mathematical linguistics virtually empty. "The same is true of the mathematical theory of learnability in so far as its results depend on the presumed infinity of the class of grammars" (Chomsky, 1981b).
Whatever be the focus in a universal Grammar, and whatever shifts might have taken place, one thing that has remained constant is the fact that theoretical grammars deal centrally with the structure of language. As a matter of fact, the role of a theoretical linguistic grammar may be viewed as being two-fold: to privide a theory of the structure of language per se and to provide the description of the structure of a language. While the centrality of the question of structure has always been recognised, the related and more crucial question of the exact locus of 'structure' has been answered differently by different linguists. Where does the structure of Language really exist - in the Linguists' mind, in language (corpus) itself or in language faculty, i.e., in the linguistic consciousness of the language-user, or some where else?
There are al least three distinct approaches to Universal Grammar which differ primarily due to the fact that they assign different 'locus' to Language-structure. According to the Firthian view the structure of a language is merely the analyists' construct in order to systematise the data. Firth held that "there are no facts in linguistics until the linguist has made them; they are ultimately, like all scientific facts, the products of imagination and invention". It is for this reason that "Linguistics assumes no categories in rebus, no system inherent in the material awaiting discovery" (Quote from Allen, 1957). The locus of structure, according to this approach, is the linguist, and it yields General Grammar (GG) which is the product of invention by linguists in the form of constructs and categories.
Quite opposite to the above is the view held by Hockett who observes "his (linguist's) purpose in analysing a language is not to create structure, but to determine the structure actually created by the speakers of the language. For the scientist, then, linguistic structure refers to something existing quite independently of the activities of the analyst; a language is what it is, it has the structure it has whether studied and analysed by the linguist or not" (Hockett, 1984). According to this view point the locus of structure is the corpus and, hence, the linguist is concerned with discovering what is already present in the corpus itself. This approach yields Discovery Grammar (DG) which is the result of the discovery of structure via inductive generalizations "which tend to hold true in more than a chance number of comparisons (such as symmetry and sound system) or which state tendencies to approach statistical limits across languages or in one language across time" (Greenberg, Osgood and Jenkins, 1963 : XVIII).
Chomsky has an altogether different notion about the locus of structure. According to him whatever deep-seated structure a language attests is a function of intrinsic intellectual organization of a speaker. It is therefore definable only in the context of universal cognitive traits which produce data and help us to understand the structure and system that lies deep behind the corpus. According to Chomsky (1965) "certain highly abstract structures and highly specific principles of organization are characteristic of all human languages, are intrinsic rather than acquired, play a central role in perception, as well as, in production of sentences, and provide the basis for the creative aspect of language use". Thus, theoretical Universal Grammar, in Chomsky's view, contains the linguistic rules which are structure-dependent, where a structure itself can be construed only in terms of innate property of the human mind. The Chomskian approach yields Innate Grammar (IG) which is the system of principles, conditions and rules that are elements of all human languages as a necessary structure or innate property of the human mind, which is the locus of structure.
Linguistic grammars such as DG and IG make assertions of putative language universals which, by their very nature, are summary statements of those basic characteristics which are shared by all human beings. Both propose two types of universals - one in respect of logical structure (i.e., formal), and the other in respect of content (i.e., substantive), and both of them implicate a theory of language learning. Contrary to this, Firthian GG explicitly rejects the validity of universal grammar and any specific theory of learning. According to Firth "Every analysis of a particular language must of necessity determine the value of ad hoc categories to which traditional names are given" (Firth, 1957 : 21).
Descriptions of any language based on different models not only differ in their emphasis on different aspects of language, but even diverge drastically in form and function or goal and orientation. Secondly, these models with different goals and orientation and with descriptions of a language in different form and presentation, make different claims for Learning Theory. For example, transformational model of a grammar as conceived by Chomsky entails a learning theory based on cognitive psychology. Basic premises of this theory are in direct conflict with those of classical behaviourist approaches - i.e., Paviov's classical conditioning theory, Hulls' habit strength theory, Skinners' instrumental-motivation theory (which have hitherto been the foundation-stone for structuralists). According to Chomsky: 'Learning theory has, for the most part, concentrated on what seems a much more marginal topic, namely, the question of species-indedependent regularities in acquisition of items of a "behavioural repertoire" under experimentally manipulabel conditions' (Chomsky, 1965 : 56-7).
The most diverging aspects of the two Learning Theories (LT) can be summed up as below:
The three types of grammar that we have been under the general category of theoretical grammars which study language for language's sake, deal with the question: "What is language", and are concerned with universals across languages. Despite what has been said earlier about GG, it is as much theory-oriented as IG and DG are.
It is worth mentioning at this point, that somehow all theoretical grammar deal primarily with an idealized conception of language. This led Chomsky to make a distinction between tow aspects of grammar viz., competence and performance. Competence grammar is the system of rules which the native speaker has internalized so that he is able to produce and understand an infinite number of sentences. It is though this competence grammar that he is able to recognize grammatical mistakes and ambiguities. This competence grammar is based on an idealized conception of language which is opposed to the notion of performance, i.e., a specific speech utterance. Chomsky has been insistent in holding that linguistic before him have been mainly concerned with specific utterances of actual speech (performance in a corpus) instead of dealing with competence grammar.
While the distinction between competence and performance has been widely accepted, literature on the subject ahs become increasingly critical of or doubtful about the sharpness of the boundaries between these tow concepts. Speech features such as intonation pose a special problem in as much as it is difficult to decide whether to assign them to competence or to performance. As far as performance is concerned, one has to distinguish between two aspects: (a) those elements which are totally irrelevant to the abstract rule-system, such as hesitation, pauses, etc., arising out of lapses of memory or biological limitations of a speaker, or due to socio-psychological factors involved in speech production - all these elements have to be disregarded in writing the theoretical grammar of a given language;(b) the use of the tacit knowledge of language structure in different socio-cultural contexts. The notion of Communicative Competence as developed by Dell Hymes (1970) focusses on the native speaker's ability to produce and understand sentences which are appropriate to the socio-cultural context in which they occur. According to him the term 'competence' refers to the general capabilities of a person, and is dependent upon both a tacit knowledge, i.e., competence grammar and the ability for its use, i.e., performance grammar.
Hymes' proposed notion of Communicative Competence is, in fact, symptomatic of a third shift of focus, i.e., a shift from grammar to pragmatics. This shift can be seen in the works of Gordon and Lakoff (1971), Searle (1975), Grice (1975), etc. This shift came about as a reaction against the view that accepts form or structure as a primary in itself, and tends to depreciate use and function as far as linguistic theory is concerned. it is to be noted here that recent researches in child language acquisition demonstrate that children are born with a communicative capacity which gets elaboration and differentiation both before and after a child makes use of verbal communication (Bullowa, 1979; Bullowa, Jones and Bever, 1964; Bateson, 1971; etc.). Chomsky talks about innate language capacity and language acquisition device. In hisview the role of experience is only to cause an activation of the innate schematism and its specified manifestations in a particular manner. The new-born was never looked upon as having work by Bullowa and Bateson clearly demonstrates that the child from its very birth behaves as a social communicative being, and that much before 'proto-conversation'.
The above shift of focus implies that in linguistic studies the central issue is function and communication rather than form and structure. We would like to emphasize here that a Communicative Grammar (based on Communicative competence) is not merely an extension of a linguistic grammar that includes in its ambit the additional component of performance grammar; it rather involves a change in theoretical perspective on language and grammar. Since the difference between linguistic and communicative grammar is based on two different kinds of orientation it is desirable that we grasp those salient points which diametrically oppose each other in the two types of grammar. We summarise below these differences in a tabular form:
A theoretical grammar centres round the notion of what Chomsky would call 'Knowledge of Language'. The knowledge of language can be properly characterized by means of a system of rules that assigns structural descriptions to linguistic expressions. Thus, knowledge of language gets tied up with the notions of grammar, knowledge of grammar and the acquisition of grammar. The communicative orientation to the knowledge of language adds one more component to it, viz., how the knowledge of grammar is put to use. However, all these notions are restricted to knowledge of language, and do not have reference to teaching/learning of the ability which creates in a person this knowledge of language. The creation of the ability to encode and decode linguistic messages is the concern of Pedagogical Grammar (PG). It is in this context that Sol Saporta makes a distinction between a scientific grammar (i.e., a theoretical grammar) and a peadagogical grammar. He says, "A scientific grammar enumerates the grammatical sentences of a language and provides each with a structural description and a semantic interpretation. The pedagogical grammar ideally attempts to develop the native speaker's ability to recognize and produce sentences. That is, "a speaker can accept an arbitrarily selected sequence of elements in his vocabulary, determine whether or not it is a sentence in his language and, if so, assign to it its correct structural description and semantic interpretation" (Sol Saporta, 1966).
While it is true that the three types of grammar are related to one another, there is an essential difference in their goals, as well as, in their intrinsic properties. The accountability of a universal grammar is centred round a theory of language, while that of a language-specific descriptive grammar is centered round the corpus. As distinct from these two, a pedagogical grammar has accountability in terms of utility, communicability and teachability/learnability. A universal grammar may be termed theory-oriented (i.e., universals across languages), a language-specific grammar as corpus (text)-oriented (i.e., generalised statements across data), and a pedagogical grammar as learner-oriented (i.e., based on learning strategies and learner behaviour).
Certain facts about pedagogical grammars may be mentioned at this juncture. A pedagogical grammar falls within the domain of applied linguistics. It may be pointed out here that applied linguistics is a field of activity wherein the aims of studying languages go beyond the study of language for its won sake. This implies a clear distinction between theoretical linguistics and applied linguistics - a theoretical linguist is thus a creator of theories about language, while the role of the applied linguist is that of a consumer of theories (Allen, 1973 : ix). However, we should also remember that applied linguists are those who exploit the gains of linguistics knowledge, keeping in view the requirements of those who are in need of such knowledge. Seen in this perspective, applied linguistics is not merely linguistics or the use of linguistics any more than medicine is chemistry or, as Bolinger (1968) has said, "the use of chemistry".
It is in this light that the new approach to second language/Foreign language teaching shifts the focus of attention from the formal grammatical to the functional, communicative characteristics of a language. Language teaching is now being oriented to enable the students to use the language for special purposes. Thus, now instead of talking about the teaching of a language in general terms, one witnesses the teaching of contextually-defined language variants (LSP, i.e., language for Special Purposes). Courses in LSP are devised on the premise that different elements of language structure are sensitive to different kinds of situational features and, thus, language structures which remain the same on the deep level, get realized differently in form and texture on the surface level.
Being applied in nature, the form and functions of pedagogical grammars differ because of certain variables of application. The following three variables are worth mentioning:
1) Teacher/Learner Variable
There orientation of a pedagogical grammar differs depending upon whether it keeps at its centre the teacher or the learner. A teacher-oriented pedagogical grammar takes a detour via contrastive linguistics, while a learner-oriented pedagogical grammar takes a detour via error-linguistics.2) Need/Use Variable
There are different uses of language such as literary scientific, etc. Earlier it was assumed that the differences between them were restricted to lexis or to the frequency of occurrence of certain grammatical forms. However, now it has been established that they represented not different forms of language but different ways of using the same language. Grammatical structures, thus, partake different communicative functions and, in the process, get oriented differently. Based on different use and usages pedagogical grammars also get differently oriented as far as functionality of grammatical structures, usage of lexical items and cohesion in textual composition are concerned.3) Methodics Variable
This is concerned primarily with the types of dialogue that exist between teacher and student, between teaching machine and student or between student and student, and the choice of media such as programmed learning, audio-lingual materials, exercise books, etc. All of these have a bearing upon the selection, gradation and presentation of linguistic items in a pedagogical grammar.These three variables can be integrated into a problem-process perspective, yielding thereby the following schema:
Pedagogical grammars thus, have a direct relationship with theoretical, as well as, language-specific descriptive grammars. However, a pedagogical grammar, in addition, includes many more things. A pedagogical grammar, as stated earlier, is applied in nature, and exists in its own right. We cannot look at this discipline as merely a sub-set of linguistics, since linguistics is concerned primarily with the code of a language, while pedagogical grammar deals with the use of code made maximally effective in language teaching/learning operation and hence, the nature of its grading is not always congruent with the ordering of rules in a theoretical or language-specific grammar.
