Advanced Grammar And Syntax

Advanced Grammar And Syntax

Why an “informational” English grammar?

GRAMMARS AND THEIR CONTEXTS
Each type of English grammar arose in circumstances we may identify as: a) the grammar designed to make it easy for English learners of Latin to master the written forms of what, till recently, was the high status European academic language, or to learn the matching forms of modern European languages described in Latin terms. This is the traditional grammar b) particular selections from traditional grammar. These arose to satisfy (especially spoken) “communicative” needs in circumstances of person-meets-person (actually elite foreign language learner meeting elite native speaker). We may attribute these selections to development of the late 19th Century railway networks c) theoretical grammars deriving from current linguistic theory d) a “bottom-up”, informational grammar to meet a demand for (especially “machine”) translation. (c) and (d) are related. Translation presents repeated problems anticipated in linguists’ relaxation of a formerly rigid idea of “structure”, their recognition of the grammar of the “arguments” of particular words, and their idea that a spoken or written form at one point in a sequence of forms may be represented by a “trace” (not heard or seen) at another. Compare: 1 She took a biscuit from the tin and munched it noisily 2 She took a biscuit from the tin and closed it noisily Whenever a construction of the kind take + noun + from + noun appears, the grammar of a following it reference is informational; in the examples, from the association of 'biscuit' and 'munch', 'tin' and 'close'.
Introduction For speakers of English, the grammar is unnoticed. However, when it has alternative translations according to the grammatical gender of “biscuit” and “tin” (e.g. in French le and la) how would one instruct a computer to make the choice? The obvious way would be to tag biscuit and munch with 'edible', tin and close with 'container'. Such informational tags are essential to the grammar of translation. Machine translation shows up multiple instances of such “bottomup” grammar, in which information, underlying native speaker usage, has to be explicit; so that a computer knows how to translate (Malay) dia as (English) he or she or it; and whether to translate (English) we as (Malay) kita or kami. However, I will show that informational grammar is not just particular, as in the above examples, but general. To begin with, I need to develop the ideas i that the outmoded (a) style grammar is currently applied in both teacher and learner English courses ii that clinging to an out-dated rationale is a reason for our profession not developing an informational style of grammar matching the present day world language and its learner needs. I go on to consider grammatical forms as clues, enabling listeners/ readers: a) to direct attention selectively b) to recognize information continuity/discontinuity. I use few special terms. One is designatum, 'what is designated by a form', 'what a form is a sign for', 'what a form is a clue to'. Informational reference is shown by single inverted commas: ' '. Prague School linguists used the term “Functional Sentence Perspective” to describe the point of view from which particular sentences are organized. I do not see “the sentence” as an informational unit, therefore use “Functional Perspective”, chiefly to maintain the Prague School link; I could have used “Informational Perspective”.
Essays in Informational English Grammar Sinclair (1972) writes, introducing his English Grammar, that “nearly all modern grammars work “downward” from the sentence to the smaller units” and that “distinctions of meaning arise from systemic contrast”. That is to say, the grammar sets off with “English” as a given, selfcontained field for formal study. The Cobuild (1991) Grammar under Sinclair’s editorship does so too. However, (1) English must represent the general functions of all languages (2) communication in English must take place in the way it does in all languages (3) outcomes to the study of English grammar are subject to the limits to study outcomes generally (4) as stated above, study of English grammar is subject to influence from social circumstances, especially from the grammarians’study traditions. It goes without saying that grammarians are not obliged to consider such things; but then, I think “nearly all modern grammars” overlook factors which are central to an English grammar for teachers. And, incidentally, the overlooked factors listed above lead one, eventually, to challenge all three assumptions nearly all modern grammarians are said to accept: “the sentence” as a primary unit, the “working down” procedure, and the centrality of “systemic contrasts”. I will, therefore, outline factors (1) − (3) above, then take up a methodological issue relevant to (4).
GRAMMARS AND THEIR CONTEXTS
Grammars and their contexts The general functions of language Ethologist Lorenz (1974) and geneticist Jones (1991) point out that each of the world’s 5 − 6,000 languages functions to make communication easy, among its language community members; equally, to make communication difficult outside them: all languages have in-built means of keeping strangers (and learners) recognizably “out”. One cannot realistically study features of a language ignoring a primary, biological communication-hindering function. Communication theorists Shannon and Weaver (1949) point out that messages in a natural language are communicated in “noisy” circumstances. Their diagram is: information source transmitter receiver information destination received signal noise signal noise source transmitted signal Figure 1 Elements in Communication Each language must give protection against noise, the protection taking the form of redundancy in all aspects of its spoken and written form sequences. If there were no formal redundancy, all forms would have equal status (like the numbers from 0 to 9). A listener to a message would be like a listener to a string of number forms, having to take them in one by one, unable to predict how the sequence is likely to develop. In fact, being able to predict the development of sequences (knowledge of sequence probabilities) is part of a person’s knowledge of a language’s grammar. It is likely to be unobserved under a “systemic contrast” approach. However, redundant and partly redundant features, being redundant or partly redundant, need not and do not coincide from language to language. Many such features appear to groups of learners to be, as in fact they are, from their nature, non-systemic; they are the means of keeping strangers (and learners) “out”. 4 Essays in Informational English Grammar How communication takes place: “observer” , “encoder” and “decoder” grammars “Observer grammar” Sinclair’s introduction sets his grammar in a descriptive or observer tradition. Observer grammarians spend years working out and describing the systemic contrasts they present; however, the time available to grammarians puts them in a different position, vis-à-vis a language, from its writers and speakers (henceforward encoders) and from its readers and listeners (decoders), needing grammars for use under time pressure. “Encoder grammar” As people under time pressure speak or write sequences of language forms (henceforward texts), they usually feel two obligations towards decoder listeners and readers. A main encoder obligation is to indicate how, in a text, more and less significant information is distributed; the other to indicate (decoder expected) information continuity, or (decoder unexpected) discontinuity. In an informational grammar, these, not the systemic contrasts of an observer grammar, are broad features for study. An “encoder grammar” involves, not in the first place text product, but processes resulting in the text. The processes involve access to forms in the encoder’s memory store, influenced by at least two of three information feedback sources, at the encoding point, at the transmission point and from the decoder(s). information source transmitter receiver information destination 1 2 3 Figure 2 Communication Feedback Circuits “Decoder grammar” A “decoder grammar” requires sensitivity to an encoder’s clues to the distribution of more and less significant information, and to continuity and discontinuity clues. Grammars and their contexts 5 Decoders may happily leave to encoders whatever measure of redundant grammatical features they include in their texts; usually, there is no decoder advantage to perceiving them. “Learner grammar” Learners are decoders, and are usually required to be encoders using language forms they have learned. However, representation of the learning itself requires a reversal of the left-to-right representation of communication. Learning results, when it does, from learners’ exploration of a “field” before them, in the first place to distinguish significant from insignificant features within it. In other words, learners must have the initiative to “scan” the language forms they encounter; and representation has to show a right-to-left process: receiver information destination attention motivation distraction Figure 3 The Learner Viewpoint There is room for a clash of interest when efficiency-seeking learner brains scan material for significant features and their course designer, following observer grammar, presents for learning an uncontrolled assortment of redundant and significant features. Limits to description of structure Structuralist Lévi-Strauss (1962) distinguished “major limb” from “twig” level structure; saying one cannot expect system regularity at the twig level, and must expect irregularities to obscure major limb structure. The theme will be developed; not least as it concerns an aspect of observer grammarians’ methodological weakness with respect to irregularity. A methodological issue From the rarity of occurrence of contrasting features in successive grammatical units, grammarians, wanting examples, are tempted to take them from unconnected, contrasting text snippets (nowadays 6 Essays in Informational English Grammar using a computer), or, as innumerable grammarians do, and as I did, with the “biscuit − tin” example, invent their own. Using snippets and making up their own examples put grammarians in a poor position to be aware of the limits to their “downward” study. Actually, present day English language teachers should have a right not to expect this particular methodological weakness still to be common; Kruisinga (1941) already identified three styles within descriptive grammar: i the “older method of providing sentences made up for the purpose” ii The use of “genuine sentences detached from their context” iii his own use of “a number of passages of English prose representing colloquial as well as literary English”. The relevance of systemic contrasts using examples “made up for the purpose” or taken from non-continuous text is relevance to the grammarians. With “examples made up for the purpose” they do not necessarily take into account how representative or unrepresentative the examples are. With “genuine examples taken out of their context”, they take for granted that systemic contrast existed between forms that came to the mind of the person writing or talking and forms either not coming to that person’s mind; or coming to that person’s mind and being rejected. In either case, the mind of a reader or listener decoder is unlikely to have been concerned with absent elements. Since a sequence “made up for the purpose” does not make a spoken or written text (in the way a sequence does when a speaker or writer has non-grammarian purposes), information about the distribution of forms in texts is not represented in the “systemic contrast” with which grammarians work. This, for a teacher’s grammar, indeed, for a grammar, is a drawback. Methodological weaknesses in “systemic contrast” grammar 1 Some traditional “systemic contrasts” are “systemic” because grammarians classify separately the language forms which might be expected to but do not enter into the contrast. For instance, they state a meaning contrast between Grammars and their contexts 7 “continuous” and “simple” verb forms, then separately list “verbs of perception” (etc.) not showing the contrast. — Ignoring the number of text occurrences of the “verbs of perception”. 2 Often, a contrast is made to appear systemic by the grammarian disregarding large-scale meaning and function overlap. Even at word level, this and that are more often interchangeable than not; in narrative, 'next' may be expressed by now as well as by then (or next). Most vocabulary and grammar contrasts refer to scale ends, and do not mention large “in-between” areas, in which speakers and writers choose forms without awareness of contrast. 3 It is rare for a single form to have a single function, or for a function to have expression through a single form. Either way, systemic contrasts become parts of complexes: played When I was young, I would play chess every day used to play 4 Where there is a “systemic contrast”, the resulting “smaller units” are often, incorrectly, assumed to contribute equally to communication, and be equally learnable. The “working down” from the sentence tradition de Saussure (1909) said language shows two kinds of relations, shown in a school Substitution Table: 1 2 3 She walked room The girls rushed into the kitchen Miss Brown sidled auditorium Table 1 “Horizontal” and “Vertical” Relations de Saussure identified a) the horizontal relations, those between successive selected forms from Columns 1, 2 and 3, making a type of clause: and b) the vertical relations, making the forms in each Column (1, 2, and 3) a type of forms, i.e. one of the “smaller units” within the clause type. 8 Essays in Informational English Grammar The “vertical style” grammar then acquires sub-units. Here are two clauses: 1 2 3 i She likes Miss Brown ii Miss Brown adored *she Table 2 Establishing Subclasses Unlike Miss Brown and the girls, she of Table 1 won’t fit in Column 3 of Table 2. She must have subclassification. Similarly, subclassification is needed from the presence of the in Column 3 of Table 1 and its absence before Column 1 items; subclassification of the ed and s endings in Column 2 of Table 2 when the grammar writer contrasts them with the stem forms (like adore). For millions of learners, the subclassification is redundant. It is therefore unfortunate that, traditionally, the subclasses are given names, and invite “downward working” grammarians to state, for each named subclass, what its function is, contrasting one with the other: “noun” versus “pronoun”; “subject” form versus “object” form, “proper noun” versus “class noun”; “past” tense versus “present”; “third person singular present” versus “other person present”. Little of the subclassification need have a significant place in even native speaker decoder grammar. And learners as encoders show limited tolerance for it; indeed, the nature of learning (exploration, to distinguish significant from insignificant features in a field) virtually requires learners to disregard subclassification which does not match that of their mother tongue. A systemic contrast account, with associated contrast of function, and associated class and subclass names, misrepresents the degrees of informational (in)significance of the forms and their rôle in communication. I now need to state the local reasons for traditional observer grammarians giving particular attention to such “vertical” rather than to “horizontal” relations.