AN INVESTIGATION OF NON-STANDARD ENGLISH SYNTAX IN 12-YEAR OLD COLOURED CHILDREN

Non-standard English (NSE) syntactic constructions occurring in the natural spdech, behaviour of a group of 20 12-year old Coloured children were identified and analysed. Three syntactic classes were isolated as having contained NSE constructions used significantly by the group: auxiliary/copula verb forms, number-verb agreement and tense. A bi-dialectal sentence repetition task was then adminstered to the same group and to a matched group of White children, in which the two groups were compared on their ability to reproduce sentences containing (1) NSE constructions used significantly by the Coloured group, and (2) the standard English (SE) forms of these constructions. Results indicated that while Coloured Ss were able to reproduce many of the SE constructions, their overall performance on these items was significantly inferior to that of White Ss. Conversely, their repetition of NSE items was significantly superior to that of White Ss. The implications of these findings for the clinical language assessment of Coloured children were discussed.

Recent sociolinguistic research into the language differences of children from economically disadvantaged and culturally-different sub-communities has necessitated a re-evaluation of current approaches to the assessment and diagnosis of language 'disorders in these children.While procedures for the identification and treatment of language deviance and delay are well described in the literature, there is evidence of increasing concern among speech and language professionals that these procedures may be inappropriate, or at least inadequate, where the child concerned is one from a dialectal sub-community whose language manifests certain phonological, lexical or syntactic differences from that of ' the standard language of the larger community 1 ' 14 ' 15 The South African Journal of Communication Disorders, Vol. 28, 1981   Reproduced by Sabinet Gateway under licence granted by the Publisher (dated 2012) Non-Standard Syntax of Coloured Children 69 Two major factors have formed the basis for this concern.1.Firstly, there is a lack of documented information on the language norms prevailing in many linguistic sub-communities.Without this knowledge, it is difficult to differentiate between socially induced language differences and true language pathology.2. Secondly, the commonly-used measures of linguistic behaviour are based on the norms of a "standard" language, and hence run the risk of indicating language deviance or delay in a child who may simply be using rules which conform to his own dialect.Research in the field of sociolinguistics provides some indications as to how these shortcomings might be overcome.In particular, the notions of socio-cultural constraints on language use and "communicative competence" 9 have implications for both the sampling of language data and assessment of the dialectally-different child.
Recent advances in the study of language as a social phenomenon have indicated that an adequate structural description of a speaker's language cannot be provided without consideration of the effects of socio-cultural factors on the speaker's language use. 6' 7 ' 9 ' 10 ' 16 These include such factors as the attitude of speakers toward their language varieties, the roles and status of speakers, the relative formality of the situation or topic of the message, all of which may have a systematic effect on a speaker's linguistic competence in that they will affect his choice of speech style or his use of one particular syntactic form over another.Evidence of systematic variation in language use as a function of contingencies in the linguistic and social environment has led to the suggestion that it is part of a speaker's knowledge or competence of his language that he "knows" implicitly which linguistic rules to apply in which social contexts.This in turn has led to a reformulation of the traditional, Chomskian notion of linguistic competence to a more broadly defined "communicative competence" which includes not only a speaker's knowledge of the rules of grammar, but also the ability or competence for the appropriate uses of language in different contexts. 3• 9 • 10 The integral role of socio-cultural factors in language use afforded by this definition has important implications for the sampling of linguistic data from dialectally-different children.It makes clear that the type of speech behaviour obtained in sampling will vary greatly according to the setting and circumstances in which it is elicited.It means that in order to obtain reliable data concerning the natural speech behaviour of children, the language clinician must know a great deal about the contextual variables which govern speech production and control for them accordingly.The notion of competence for use of language has further implications for the use with dialectally different children of assessment tests based on elicted imitation (EI).More specifically, it has helped to explain the differences in performance found between children from different dialectal communities on these tasks.

Die Suid-Afrikaanse Tydskrif vir Kommunikasieafwy kings, Vol. 28, 1981
Reproduced by Sabinet Gateway under licence granted by the Publisher (dated 2012) Sentence repetition or EI tasks have been widely used by investigators of child language and speech clinicians alike as a means of determining the capacity in young children developing language, for comprehension and expression. 8' 17 • 13 They are based on the premise that a child will correctly reproduce only those sentence structures which are a part of his linguistic competence: 17 Where an item is beyond his level of competence, the child is said to filter the sentence he hears through his own productive system, to reproduce it using the rules he knows. 5ociolinguists who adopted EI tasks as a source of information on the language behaviours of non-standard English (NSE) speakers, found evidence that a similar type of "rule-filtering" occurred with much older (adolescent) speakers of NSE dialects. 10' 13 Labov 10 and Baratz, 2 in separate studies, found that when adolescent Black NSE speakers in the United States were given standard English (SE) sentences to repeat, the majority tended to re-encode the sentences into NSE form in their responses.Thus, for example, given the sentence "I asked Alvin if he knows how to play basketball", the majority of speakers repeated it as "I asks Alvin do he know how to play basketball". 10In terms of Hymes 9 notion of communicative competence, these speakers might be said to have hacf the tacit knowledge for the SE rule, in that the meaning of the sentence was preserved in their responses, yet did not show the ability or competence for use of the SE construction.Furthermore, Baratz's 2 findings indicated that White English speakers were significantly superior to Black speakers in repeating SE sentences, while Black speakers were far superior to Whites in repeating NSE constructions.
It was concluded from these findings that the language assessment of NSE-speaking children should include measures of their knowledge of NSE in addition to their knowledge of SE.However, current language tests based on the EI technique, including the Carrow Elicited Language Inventory 5 and the Northwestern Syntax Screening Test, 11 fail to account for the influence of dialectal differences on performance.They are based solely on SE norms, and hence are liable to score grammatical complexity as being absent when in fact it may be present, but its dialect is not that of the test instrument. 1 It was the contention of the present study that an effective approach to the language assessment of the dialectally-different child requires: 1. Knowledge of the linguistic forms used in the dialect of the child's community, and v •*' 2. The availability of assessment tools which are sensitive to the dialectal variations present in the child's language.The need to meet these requirements was seen to be particularly relevant to the language assessment of children from the Englishspeaking "Coloured" community of Cape Town 18 .That the speech of this group is characterized by certain phonological, lexical and grammatical differences is clearly apparent to the outside observer.Yet the lack of documented information on its linguistic forms 14 make

The South African Journal of Communication Disorders, Vol. 28, 1981
Reproduced by Sabinet Gateway under licence granted by the Publisher (dated 2012) it unclear as to whether there can be said to be a distinct pattern of the language spoken in this community, while at the same time, our measures of language proficiency are based on the norms of a White, middle-class, "standard" English speaking population, and hence of questionable validity to this group.This study attempted a partial fulfillment of these needs, by investigating the presence of NSE syntactical constructions in, the language of Coloured children, and the implications of these for the use with the Coloured child of one type of assessment tool -viz.test based' on elicited imitation.METHOD AIMS 1.The elicitation and analysis of samples of natural speech behaviour from a group of 12 year old English-speaking Coloured children, to identify the significant presence and/or trends of NSE language patterns.2. Administration of a bi-dialectal sentence repetition task to the same group of Coloured children and a matched group of White children, in which the two groups are compared on their ability to reproduce sentences containing: (a) NSE syntactic constructions used significantly by the Coloured group, and (b) the SE forms of these constructions.-SUBJECTSPart I: Ss comprised a group of 20 Coloured school-children, of which 10 were boys and 10 girls.Ages ranged from 12.0 to 13.0 years.All Ss came from English-speaking homes and attended English-medium classes at their schools.All were from low socio-economic status families, as determined by parental occupation.All were judged by their teachers to possess average intelligence and normal speech, hearing and language abilities.Ss were representative of 2 broad areas of the Cape Peninsula.
Part II: A matched group of 20 English-speaking White schoolchildren, judged to be representative of White South African SEspeaking children, was added to the experimental population.These were matched to the first group by applying the same criteria with regard to age, sex, home-language, intelligence and speech, hearing and language abilities.Parental occupations fell within the range of the middle socio-economic group.

Reproduced by Sabinet Gateway under licence granted by the Publisher (dated 2012)
Experimenter (E).Language was elicited in 3 stimulus situations: 1. Unstructured conversation with the E. 2. Responses to a set of questions related to 5 stimulus photographs.To ensure consistency in the method of eliciting language and to ensure a minimum number of utterances from each S, five directions were supplied for each of the 5 enlarged photographs.These were constructed (after Leonard 12 ) so as to elicit present, future, past and future conditional tenses.3. Conversation with the Ε on 4 topics, each generated by a prescribed set of questions.Each session was recorded onto tape by the writer.Language samples from each child were transcribed for the purpose of isolating NSE constructions used by each S.These were then categorized according to grammatical class for examination of possible trends of NSE usage.A manipulation of the Chi-square formula was used to determine which NSE constructions were used significantly by the group as a whole.These comprised the "critical constructions" to be used in Part II of the study.
Part II: For each of the 'critical' NSE constructions, a NSE sentence containing that construction was devised, as well as a corresponding SE sentence containing the standard form of the construction.These sentences, randomized for presentation, formed the basis of the sentence repetition,'task administered approximately 2 weeks later to the same group of '''Coloured" Ss as well as the control group of White Ss.The task was administered to the Coloured group by the same Coloured Ε used in Part I, while the writer served as Ε for the white group.Ss were seen individually by the Es.Responses were recorded in writing immediately, as well as tape recorded, to ensure accuracy.Repetitions were scored in two categories according to whether the critical construction of a sentence was "repeated verbatim" by the S or "altered to opposite dialect".Chi-square analyses were used to determine for each group, which SE and NSE sentences were "repeated verbatim" by a significant number of Ss in the group, and which were "altered to opposite dialect".Also, Chi-square analyses were used to compare the responses of the two groups to SE and NSE sentences.T-tests for independent samples were used to determine whether a significant difference existed between the two groups in the total number of verbatim repetitions of (1) SE and ( 2 0.05), NSE constructions were used by a significant number of Ss.The 7 constructions which fell above this point were: (1) absence of the auxiliary "are".
(3) absence of auxiliary "have".(4) absence of 3rd person singular present tense marker (-s) (5) absence of regular past marker (-ed).( 6) simple present replacing future conditional tense.(7) simple future replacing future conditional tense.These constructions were grouped together with the remaining NSE constructions in context of the syntactic class to which they belonged.A total of 9 syntactic classes were found to contain NSE constructions; for purposes of the present article, however, only those classes containing NSE forms used significantly by the group (the "critical constructions") are presented.
Certain forms of 'BE' and 'HAVE' -"are," "has" and "have"were deleted in present tense constructions (see Table I).A single tendency may underlie all of the items listed in Table I, in that absence of auxiliary/copula forms occurred only for those forms of 'BE' and 'HAVE' which in SE are contractible.The SE contraction rule removes all but the final segments of these auxiliaries in present tense forms.It seems that Coloured speakers may be applying the contraction rule, and then applying a further rule for deletion of the segments remaining after contraction.Furthermore, deletion occurred most frequently in instances where the preceding subject was a pronoun.Taken together, these factors suggest that where contraction is expected in SE, Coloured dialect may delete, and that the probability of deletion is greater when the subject is a pronoun.

Reproduced by Sabinet Gateway under licence granted by the Publisher (dated 2012)
Number-verb agreement NSE forms in this category are listed in Table II.These forms may all be indicative of a single tendency among Ss to eliminate number distinction in verbs and demonstrative pronouns.This tendency appears to be a simplifying process, but may be directly related to the influence of Afrikaans syntax: (1) There are no inflected verb-endings for the 3rd person singular in Afrikaans; hence the same verb form is used for all persons.By eliminating 3rd person singular present tense markers, Coloured English speakers achieve a similar simplicity of form.(2) Ss use of English "is" and "was" regardless of number, reflects a similar pattern to Afrikaans where there is only one present tense form of the verb "to be" ("is") and one past tense form ("was"). ( 3) The use of singular demonstratives with plural nouns may be similarly explained, since there is no plural form for demonstratives in Afrikaans.Reproduced by Sabinet Gateway under licence granted by the Publisher (dated 2012) nearly every S where the full forms of these tenses were used.Furthermore, where reduction to the present tense occurred, it may have been directly related to phonological environment: For example, formation of the regular past in SE involves simply the addition to the verb base of /t/ or Id/, both sounds which tend to be lost in certain kinds of word-final consonant clusters, whether part of a past tense formation or not. 162) The simple future was frequently used in place of the future conditional tense -possibly indicative of a preference to use a more tangible form of the verb, rather than the abstract conditional construction.Numerous further NSE constructions were found in the categories of negation, question forms, adverbials, adjectives, pronouns and prepositions.Since none of these categories contained NSE forms used by a statistically significant number of Ss, they have been omitted from discussion for purposes of the present article.
From the preceding discussion, it seems evident that the NSE forms produced by Coloured Ss cannot be considered without reference to the structural linguistic and socio-cultural factors which may have entered into their realization.It was seen that many of the NSE forms might be considered the result of grammatical "interference" from Afrikaans, to which the Ss under study can be assumed to have had considerable exposure, Afrikaans being the dominant language of the Coloured community.The possible influence of phonological environment on certain of the forms was also referred to.

Die Suid-Afrikaanse Tydskrif vir Kommunikasieafwykings, Vol. 28, 1981
Reproduced by Sabinet Gateway under licence granted by the Publisher (dated 2012) A third consideration was that certain NSE forms might be seen as reductions to a simpler form.Certain of the NSE constructions coincide at least in superficial form to stages in the language development of children in a SE community.This is true of the forms used by Ss comprising morpheme omission -omissions which are also found, interestingly, in many American NSE dialects.Brown 4 , however, has stressed that it should not be concluded from this that these forms constitute "immature" versions of SE.He proposes that many morphemes, by virtue of the fact that they are frequently redundant to the meaning of an utterance, may be especially vulnerable to deletion.He finds no evidence that the 'misconstructions' of certain dialects are simpler than the standard constructions, merely that they are different.Finally, it was evident from the preceding linguistic analysis that the use of NSE forms by Ss was not categorical.Thus, for nearly every NSE construction used, instances could be found in which the SE form of that construction was used.For this reason, the syntactical differences shown by Coloured Ss should not be taken to indicate the consistent presence or absence of certain features, but might rather be viewed as a matter of differences in relative usage.In this regard, it should be noted that certain of the NSE forms outlined above may be detected in the colloquial speech of many White South African English speakers (a case in point being the omission of auxiliary/copula "are").It seems probable that there are social constraints which govern the probability of occurrence of NSE forms in White South African English as opposed to Coloured English.It is likely that the prestige value attached by White speakers to the standard form of the language serves to constrain their use of non-standard variants except in their most casual, unmonitored speech.And equally likely that the more a Coloured speaker strives towards the values of the highest status societal group, the more his language will tend towards the standard variety.These factors would make it doubtful as to whether a standard pattern of non-standard usage could be said to exist in the Coloured community.The variability shown by Ss in this study in their use of NSE forms appears to support this view.
It is suggested, therefore, that rather than forming a separate dialect with distinct NSE forms, so-called Coloured English might be seen as part of a spectrum of South African English, which has at its one extreme White, middle-class SE, and at the other extreme the dialectal variations of the Coloured lower-class.In this view, Coloured English forms part of the same continuum as SE, sharing the vast majority of its rules, yet containing certain extensions and modifications of these rules which may be variably applied depending on factors in the linguistic and social context.,

Part II i
Results of statistical analyses on the sentence repetition task are Results indicated that, in general, when given SE sentences, White Ss repeated them verbatim, and that when given NSE sentences, they re-encoded the sentence in terms of their own SE rules to produce it in SE form.However, given the constructions "auxiliary 'are'" and "copula 'are'", in either SE or NSE, White Ss showed some tendency to apply the NSE rules in their responses.Although this tendency was not statistically significant in either case, it seems that the rules for auxiliary and copula deletion are not necessarily peculiar to the speech of Coloured children, but may be a part of the White child's linguistic behaviour as well.
Would the Ss in each group be consistent in their "translations" from the unfamiliar dialect?Examination of responses by each S in the two groups to SE and NSE sentences indicated that Ss from both groups were consistent in their responses, in that when they translated a given sentence to their own dialect, all applied the same SE or NSE rule in their responses.This result appears to provide evidence for the validity of NSE forms as representing a structured, yet different set of rules.
PROCEDURE Part I: Language samples were obtained from each of the 20 Coloured Ss in recorded sessions, in which Ss were interviewed by a Coloured Die Suid-Afrikaanse Tydskrif vir Kommunikasieafwykings, Vol. 28, 1981 the Chi-square formula to (determine the frequency of occurrence of NSE constructions which would be necessary to yield a significant Chi-square value, it was determined that above 14 (p <The South African Journal of Communication Disorders, Vol.28, 1981

TABLE I :
Non-Standard auxiliary/Copula verb forms in Coloured children

TABLE II :
Non-standard number-verb agreement in Coloured children The South African Journal of Communication Disorders, Vol. 28, 1981

TABLE III :
Non-standard tense usage in Coloured children

The South African Journal of Communication Disorders, Vol. 28, 1981 discussed in
relation to the following questions which this part of the study aimed to explore: Would Coloured Ss reproduce sentences containing SE constructions used irregularly by them?Or would they produce sentences conforming to the NSE syntactical patterns identified in Part I? Results of Chi-square analyses indicated that Coloured Ss showed considerable facility at reproducing verbatim many of the SE constructions.It seems likely that the relative formality of the testing situation in this part of the study, as well as the nature of the task itself, both of which were conducive to a more careful and monitored style of speech, may have served to favour application of the standard rule in this task for some Ss.In spite of these constraints, however, some proportion of Coloured Ss showed a tendency to re-encode each of the SE constructions into NSE form in their responses -indicating that they had decoded the sentence, then re-encoded it in the form they might have used in formulating the sentence themselves.Would the repetition of certain NSE forms be restricted to Coloured Ss, or might White Ss show a similar tendency to produce NSE forms?
If the NSE forms used by Coloured Ss were a matter of random, unordered variation, one would not have expected consistency among the NSE responses.Could one group be said to be significantly superior to the other at repeating (1) SE and (2) NSE sentences?Results of t-tests used to compare the total number of verbatim repetitions by each group to each of the 2 sentence types revealed that Coloured Ss were significantly superior to Whites at reproducing correctly NSE items, and that White Ss were significantly superior to Coloured Ss in reproducing SE sentences.The main trends of the results of Part II may be summarized as follows.While Coloured Ss showed considerable facility at reproducing Die Suid-Afrikaanse Tydskrif vir Kommunikasieafwykings, Vol. 28, 1981Reproduced by Sabinet Gateway under licence granted by the Publisher(dated 2012)