Monday, December 19, 2005

Bidialectal instructional theory - part 1

PSY-F 8-554

Term Paper
Robert K. Philips
8 June 1979

It has become more acceptable, with the decline of behavioristic paradigms in psychology, to ask questions regarding the nature of knowing (Weimer, 1973). Two conceptual approaches have contributed to the rise in acceptability: information processing, and the notion of the generative grammar in linguistic theory (Chomsky, 1965). In the former, it has been posited that the information processor has an internal representation (Rickards, 1978), and in the latter there is a "deep structure" which represents meaning. In the present problem, it is not intended to challenge these assumptions; the question that is appropriate is "what is learned ?"(Bransford & Franks, 1972).
It has been argued that the choice of methods and modes of instruction are secondary concerns in instructional psychology, and that the determination of "performance criteria" is of more vital importance (Burris, 1976). The present concern is to attempt to provide a more sound theoretical basis for the validation of an existing instructional methodology (Craig, 1966), namely, English language teaching of Creole speakers in the West Indies. Although evaluation has shown that there is a modicum of success in the method, it is contended that it would be worthwhile to characterize the memory structures at various stages in the language learning process.
A linguistic metaphor has been used to describe the state of knowledge of an expert processor of a subject matter area. In this problem, "grammar" is appropriate in its literal sense. Grammars of Creole Languages have been documented (e.g. Bailey, 1966), and have been utilized in the design of instruction. The position taken here is that this is insufficient, that in addition to linguistic analysis, there must be "psychological analysis." It could be argued that there is no need to pursue the construction of a model of knowledge or memory appropriate to a specific language. It is not being contended here that extant models (Collins & Oujilian, Anderson & Bower, 1973, e.g.) are inappropriate, but it must be admitted that their generalizability or universality has not been determined. Of course, it is unrealistic to use the results of a single experiment to construct a model of semantic memory appropriate for a Creole speaker, but such differences as might be determined can be of use in revising existing models.

In the Caribbean, the Creole languages derived in part from English or
French, are spoken by an estimated 70 to 80% of the population (Craig, 1976).
The origins of these languages are debated among linguists but of Jamaican
Creole in particular, Le Page (1968) claims:
Characteristically the Creole vernacular may be described in terms of interference between West African linguistic structures and those of seventeenth century English, with continuing subsequent influence from the English of the day as a model language, and from other languages of the region through contact. p. 434
Clearly, one problem is that of discriminability. It has been noted that as a spoken language [Creole) has its own phonological, grammatical, lexical and semantic structures but these "differ, often quite sharply" from Standard English (Le Page, 1968). The dialect-continuum poses a practical problem for instruction because it is not possible to locate a speaker at a fixed position on the scale. Moreover the speaker has the capacity to vary his position systematically. Craig (1976) notes that this poses problems for those who would opt to instruct in the vernacular. Even if English phonology were applied, the language to which the written word would correspond would be the dialect of a few speakers.
The major problem of discriminability is that "the language of West Indian children possesses a vocabulary which is almost completely English but at the same time possesses a grammar which diverges sharply from English" (Craig,1966).
Critically, the "dichotomy in language ability creates the illusion that the English being taught in the classroom is known already" (Craig, 1966). This illusion occurs as much for the student as the largely untrained teacher (Le Page, 1968). It is evident then, that the problems which arise here are not simply those which occur in foreign language learning. [Engle (1975) notes that bilingual education of minorities is not a simple issue, and that the psycholinguistic bases have not been determined.]

A partial foreign language instructional approach has been developed for the teaching of English. It was considered to be appropriate, in that (1) contrasts of Jamaican Creole (JC) and Standard English (SE) structures to be taught, would be made, (2) the English contrasts would be graded, and (3) methods and materials appropriate for practicing the graded English patterns would be made. However three factors militated against a foreign language approach.
1)There are several types of SE in the society which are acceptable; these are, mainly, Standard Jamaican English (SJE) and Standard British English (SBE).
2) English and Creole have a number of common features.
3) There are intermediate dialectal variants between JC and English.
In the first instance, the teacher has to accept a wide variety of correct forms, whereas in the foreign language approach, there is usually a "single well defined standard." An example of the flexibility required is evident from the following question.
"Why you are making these lines so tall?"
There is no subject-verb inversion, which makes the sentence ungrammatical in SBE, but grammatical in SJE. In addition "tall" is equivalent to "long." It is clear then, that grammatical, lexical and phonological deviations have to be considered acceptable.
For Creole production, Craig (1966) shows how the second and third factors (peculiarities) are demonstrated.
/a 'mi: buk dat’ is equivalent to
/iz 'mi: buk,
and to
/iz mai buk
Further, on the continuum, the standard productions would be
It's my book.
That’s my book.
In addition, there are the "extreme JC" forms:
/ a fi mi: buk dat/
/ dat a fi mi: buk/
Despite the apparent high state of linguistic entropy, some patterns can be identified, and provide a starting point for an instructional procedure. Four classes of English patterns have been isolated.
Class A. Patterns already known. That is, JC speakers know how to use these spontaneously in their own informal speech.
Class B. Patterns used under stress. These may have been learned, without becoming firmly habitual, through traditional school teaching, through short contacts with SE speakers, through intermittent exposure to mass media, etc.
Class C. Patterns known passively. That is, JC speakers would understand these owing to context, if used by other speakers, but JC speakers would not themselves be able to produce them.
Class D. Patterns not yet known.
There is a striking similarity between the idea of "Class B" patterns--those produced under stress, in interaction with authority figures--and the "School register." Houston (1969) notes that "a register is a range of styles of language, which have in common their appropriateness to a given situation or environment."
The School register is careful or corrected speech, arising from the children 15 accurate perception that a different kind of language is expected of them (Houston, 1970, p. 27).
Class C patterns are similar to those which enable non-Standard Black American speakers to "perceive, abstract and reproduce the meaning of many standard forms which they do not produce" (Labov and Cohen, 1973).
The aim of the instructional strategy is to take learners into Class D patterns. There are two main components of the strategy, "free talk" and "controlled talk." Briefly, free talk serves the purpose of enabling the children to express themselves by any means at their disposal, so speech at any position on the continuum is permissible. Where necessary, the teacher will speak in the vernacular. The exercise is not completely unstructured however, as there are specific linguistic structures selected for use in the discussion of topics, which reflect the interests, developmental level and cultural environment of the children (Craig, 1976).
In controlled talk, techniques such as substitution and transformation practice, controlled dialogues and drama are used, but there are "target structures" from the passively known and unknown patterns, which the learners
"are forced to use." Reading and writing are done only for those patterns that have been learned.
Apparently, the major psychological or theoretical input into the instructional method is behavioristic.
The built-in resistance of the second-dialect learner to [concentration on the ordered and sequenced teaching of new language elements) is countered by the carry-over of his free-talk interests into other activities by the constant reinforcement passing from one activity to the other, and by the encouraged possibility of newly learned language gradually infiltrating into free talk, become a part of it and becoming gradually augmented (Craig, 1976, p. 113).
It is clear that the intention is to change the surface patterns of the Jamaican Creole speaker to make them approximate a standard pattern. The major task is to provide contexts for the elicitation of previously imitated patterns or observationally acquired ones. It is not assumed that there is any necessary relationship between the concepts represented by the child and the language he uses. This viewpoint has close parallels in the position taken by Houston, who argues that the Child Black English speaker knows language:
The level which must be remediated or changed is that of systematic performance. It is easily demonstrable that the children's language problems are not knowledge problems but performance problems, and the task of language change should on the whole be not to give the children new language knowledge, aside from lexicon, but to alter their performance, so it becomes acceptable to educated speakers (Houston, 1970, p. 33).
For Child Black English, Houston's stance is probably correct. As she and others have noted, the differences between Black English and Standard American English are more phonological than syntactic. Differences such as the use of "have" and "be," (Labov and Cohen, 1973) are minor compared to the differences between Jamaican Creole and Standard English.