Is 'Bad English' Dying Out? A Diachronic Comparative Study of Attitudes Towards Creole Versus Standard English in Trinidad

This article deals with attitude change towards Creoles in the anglophone Caribbean. For socio-historical reasons, English-lexicon Creoles traditionally form the 'L' variety (= low prestige) in a diglossic relationship with Standard English. The low degree of linguistic distance (Abstand) to their lexifiers has furthermore contributed to the notion of Creoles as "corrupt" versions of the European languages and their consequent low prestige. A comparison of two quantitative empirical studies (Winford 1976 and Mühleisen 1993) on teacher attitudes in Trinidad suggests that these negative evaluations based on the perception of Trinidadian English Creole (TEC) as an inferior form of English are disappearing. The outcome of the questionnaire-based research shows that respondents' linguistic self-confidence plays an important role in their judgements. While a strong domain consciousness exists among the respondents, the acceptance of TEC as a "language in its own right" has increased significantly. The article concludes by considering possible implications of this attitude shift and a call for a discursive model for language attitude research.

1 Introduction

Attitudes towards English-lexicon Creoles in the Caribbean are typically ambivalent as shown by a number of systematic studies (e.g. Haynes 1973, Winford 1976, Rickford 1983, Mühleisen 1993, Beckford-Wassink 1999). That speaker attitudes are also publicly debated may be illustrated with two statements from a Trinidad newspaper:

If one takes the statements above as representing the extreme points of contrasting beliefs and value judgements on Trinidadian Creole one can roughly estimate the wide spectrum of possible views in that society. The fact that these opinions were voiced in a newspaper shows the prominent position language attitudes hold in the public discussion. One may assume, however, that those attitude holders who "go public" advocate particularly strong positions at the respective margins of the attitude continuum. Besides, they may not be representative in that they almost exclusively rely on aesthetic value judgements rather than pointing to the usefulness of Trinidadian English Creole (TEC) in different spheres of life. Thus the functional divide between Standard English and Creole in a typically diglossic situation in a society where Creole represents both the "symbol of powerlessness and degeneracy [...] [and the] symbol of solidarity and truth", (Rickford 1985) is not accounted for in these statements.

The vehemence of emotions and beliefs connected with English-lexicon Creoles in the Caribbean may become clear if one follows St. Clair's postulate that "to understand fully how language attitudes develop, it may be necessary to reach back into the past and investigate the social and political forces operating within the history of a nation" (1982: 164). Contradictory attitudes towards Creole may consequently be explained, at least in part, by the violent colonial past. That the negative evaluation may be subject to change, along with a redistribution of functions in a gradually eroding diglossia, is suggested by a recent article (Shields-Brodber 1997) on the Jamaican language situation. Not surprisingly so: as Edwards (1982: 26) notes, the social context in which speech evaluations (or rather speaker evaluations) occur is not a static entity. As it changes diachronically, one should expect to see changes in evaluation patterns too, and in fact these may be useful indicators of larger adjustments in social perceptions. Language attitude studies have largely ignored the dynamic aspect of their subject matter, possibly also because the question remains how such an attitude change is initiated and how it could be measured and documented.

One of the problems with attitude studies is that the very concept of "attitude" is a relatively vague one which encompasses a whole range of meanings – from openly stated opinions to actual behaviour. One definition is that attitudes are an individual's relatively enduring disposition to respond favourably or unfavourably to an object, person, event or institution etc. There is also some overall agreement that attitudes have an evaluative character. Attitudes therefore contain elements of belief as well as varying degrees of factual knowledge (or what the holder takes to be factual knowledge) which together form the cognitive component of attitudes, next to other elements, the affective (expression of feelings) and the conative (intention to act) component.

Attitudes are therefore difficult to evaluate as one entity, especially since, as Edwards (1982: 20) points out, there often exists an inconsistency between assessed attitudes and actions presumably related to these attitudes. He also stresses the frequent confusion between belief and attitude and concludes quite rightly that "many 'attitude' questionnaires are, in fact, 'belief' questionnaires". However, there is reason to assume that the cognitive component (knowledge and belief) plays a most vital role in attitude formation under certain conditions. Ajzen (1989: 247) states that "attitudes are not merely related to beliefs, they are actually a function of beliefs, i.e. beliefs are assumed to have a causal effect on attitudes" [italics in the original, S.M.].

Applied to language attitudes in the anglo-creolophone context, beliefs about Creole languages, what the attitude holder takes (or took) to be factual knowledge, may have contributed enormously to the traditionally negative attitude formation towards Creole varieties in the Caribbean, seeing them as corruptions of the European languages to which they are lexically related. This is particularly prevalent in those language situations which have traditionally been described as a post-creole continuum where the Creole language is in direct competition with its lexifier (for a more detailed discussion of this, cf. Alleyne 1994: 11–13). The interplay of linguistic autonomy, functional elaboration and language attitudes may thus become evident if we look at the level of stigmatisation of English Creoles which compares unfavourably with Creole languages not clearly associated (and in competition) with one distinct European language such as Papiamentu. As Görlach (1996: 165) notes on a discussion of decision determinants for "language-ness", however, "it will become clear that distance (abstand) and functional range and standardization (ausbau) alone are not sufficient, but that attitude is the decisive factor in making a decision [...]."³ Positive attitudes, in turn, are usually a prerequisite for standardization processes to increase the perception of autonomy (e.g. by choice of orthography) and elaboration (e.g. by use in formal registers).

2 A Comparative Perspective on Language Attitudes in Trinidad

There are no systematic longitudinal studies on language attitudes⁴ in the anglophone Caribbean which could shed light on the dynamic dimension of language attitudes. In view of this research deficit, a comparative approach along diachronic lines was taken up by Mühleisen (1993) as a follow-up study of Winford's (1976) study. In the following, both the methodological design and the findings of the respective studies will be presented.

The follow-up research itself was conducted 22 years or roughly one generation later than the Winford study: the original was conducted in 1970 on a group of trainee-teachers at colleges in the two largest towns in Trinidad, San Fernando and Port of Spain. Using a direct investigation method, a questionnaire, Winford's questions aimed at a speech evaluation of different groups and at the functional distribution of language varieties and their acceptability in different domains. His sample was broken down into the two main ethnic groups in Trinidad (African/Indian) and residency (urban/rural). Winford found that there was a great deal of awarenenss of the varieties used in Trinidad and furthermore that there was substantial agreement on the identification of a "less correct" variety in rural areas and a "more correct" variety used in town areas. Equally notable was the difference between Afro-Trinidadians and Indo-Trinidadians concerning the evaluation of and the confidence in their own speech. Indo-Trinidadians, especially from the rural population, were most insecure with regard to their speech – perhaps not surprisingly so: The Indian population was the last major immigrant group to arrive in Trinidad, and was thus situated at the bottom end of the social structure for a considerable period of time. Moreover, they also maintained their distinct language, Trinidad Bhojpuri for quite some time which only lost its significance by a gradual acculturation to the wider society in this century.

Some of the changed conditions became apparent at the very beginning of the attempt to design a follow-up study and proved to be problematic yet they may be important to note:

• One of the problems was related to categories that might have been very valid in 1970 but appeared to be inapplicable in 1992. For example, a simple urban/rural dichotomy can not represent the residential structure in Trinidad today. To live in a rural place in 1970 Trinidad might have been a different experience in terms of interaction with the wider society than it is today with an ever increasing availablity of means of transportation and telecommunication.
• Winford's respondents were trainee teachers and presumably quite homogeneous in age group (up to about 25 years old) and qualifications. Since the primary concern of the follow-up study was to examine possible changes, age seemed to be a most relevant factor; qualification, sex and their possible role were also considered. Therefore, it was necessary to consider this in the follow-up research design in addition to Winford's extra-linguistic factors (ethnic group and residence group).
• Another problem was related to the acceptance and use of evaluative labels: the pre-pilot study showed very clearly that some of the expressions Winford used in 1970 could not be applied anymore. For example, in the Winford study the respondents were asked to list the things that they or other people considered "bad English" in the speech of Trinidadians. All the pre-pilot respondents in 1992 (who were people of different age groups, occupations and educational backgrounds) refused to acknowledge the label "bad English" and, as a consequence, some of the original questions had to be altered or modified accordingly. The necessity to rename the subject matter of the study may be a case of a "political correctness" term change. Whether this went along with a change of language attitudes was to be substantiated in a more systematic way.

  
  

  
  

  ---

  PhiN 15/2001: 47

  
  

• The acceptability of TEC in certain domains – the media and the classroom – seemed to be particularly interesting considering the developments which have taken place in the last twenty years; therefore some more specific questions were added (such as "acceptability of Trinidadian in Local TV programmes" or "sopa operas") at the expense of general questions like "language use in mass media".

For the follow-up study, a questionnaire was designed with a mixture of open and closed questions. Closed questions had either yes-no choices or a range of choices according to the extent of the respondent's agreement with each item. Space for comments was provided with almost every question. The questionnaire was divided into 4 parts:

• Part A of the questionnaire gathered demographic information on the respondents.
• Part B aimed at an evaluation of Trinidadian Creole in general and and evaluation of the speech of various subgroups (e.g. ethnic group).
• Linguistic self-confidence and the respondents' identification with TEC was examined in part C, as well as the functional distribution of the different codes in interactional domains.
• Finally, an open question gave the respondents the opportunity to state their opinion on the role of TEC today and in the future.

The workability of the questionnaire was then tested in a pilot-study with a random group of school teachers. In the following, the outcome of the study will be presented in a close comparison to Winford's (1976) study. The structure of the questionnaire (part A–D) is reflected in the organisation of chapter 3, starting with general demographic information on the respondents, followed by results of specific evaluative questions (3.1–3.4) to responses to the open question (3.5).

3 Language Attitudes in the Nineties: the Follow-up Study

For the sample of the 1992 research, primary and secondary school teachers were chosen as professional group (i) for reasons of comparability with the Winford study and (ii) because teachers seemed to be a very significant group for research on language attitudes: the fact that school is one institution which is largely responsible for maintaining the organisational structure of society manifested in attitudes and impacts of language attitudes in education makes the role of the teacher in the classroom crucial both for the change or the perpetuation of language attitudes as well as a reflection of societal attitudes.

A sample size of 100 respondents was chosen to ensure an adequate representation of different subgroups. Primary and secondary school teachers were to be represented in equal parts (50: 50). The sample was a random sample from a specific area, but stratified on the basis of the distribution of male and female teachers and their allocation to primary and secondary schools. This choice for a stratified random sample was made in order to minimise the possibility of an under-representation of subgroups.

The male-female ratio of primary school teachers in Trinidad is 29: 71 %, in secondary schools the distribution is 47 % male (M) and 53 % female (F) teachers⁵. Accordingly, the questionnaire was handed out to 15 M and 35 F teachers in primary schools and to 23 M versus 27 F secondary school teachers. The return rate was 90 %. As a consequence, the actual numbers differ slightly from the aspired ones: 15 (M): 34 (F) in primary schools; 17 (M): 24 (F) in secondary schools. The sample was divided into four age groups: up to 25 years (Age 1), 26–35 years (Age 2), 36–45 years (Age 3) and 46–55 years (Age 4). More details are shown in the following figures:

To ensure a fair representation of the different school types and the different denominational affiliations of the schools, six primary schools and two secondary schools of a specific area were involved in the sample:

Primary Schools:

1. Roman Catholic Girls' Primary School
2. Roman Catholic Boys' Primary School
3. Hindu Primary School
4. Anglican Primary School
5. Presbyterian Primary School
6. Government Primary School

Secondary Schools:

1. Junior Secondary School (El Dorado)
2. Grammar School (Hillview Boys' College)

All of these schools are situated in and around Tunapuna, one of the small towns along the principal traffic route in the northern part of Trinidad, the Eastern Main Road. The Eastern Main Road connects Port of Spain in the northwest with Trinidad's third largest town, Arima, in the middle of the northern part. The whole area belongs to the district St. George, the district with the highest population. The small towns along the Eastern Main Road between Port of Spain up to about Tacarigua/Arouca form an almost uninterrupted conurbation with good transportation facilities (public bus, maxi taxis and route taxis which run regularly along the Eastern Main Road) and good shopping facilities. The area would thus fall into the category characterised as "semi-urban". Tunapuna is a relatively old settlement; the railway which was built in the early twentieth century in the vicinity of Tunapuna attracted a large number of immigrant workers from other Caribbean islands. Nowadays, the proximity to the University of the West Indies is a cause for new influx of people, migration and a upward social movement.

Apart from the two main ethnic groups in Trinidad – African and Indian – the categories "mixed" and "European" were represented by relatively small numbers.

It is a commonplace that research done by an outsider of the community will distort the results because of his/her very presence a consequent unnatural behaviour of the target group. In this case, however, it was felt that the fact that the researcher could not be easily classified as a member of one or the other subgroup of the community was rather an advantage than a disadvantage since it may have kept respondents from accomodating to the researcher. The fact that the researcher's mother tongue was not English may have also lowered the pressure for the respondents to prove their ability to speak "correctly". The return rate of 90 % of the questionnaires shows a high rate of participation of the teachers.

3.1 Respondents' Definition of Trinidadian Creole Characteristics

This question aimed to find out how the respondents define Trinidadian Creole, what features they see as "typical" properties of the language and which characteristics they would most readily accept as "English" (which could be an indicator of the respondents' notion of Standard Trinidadian English). The teachers were asked to give one or more examples or statements. Altogether, a number of 293 examples or statements were made. The question was difficult to score, especially because of a difference in "linguistic knowledge" of the teachers. Some simply stated, e.g. "pronunciation", some gave examples like: ""De" for "The" or "Dis ting" for "This thing", others gave a detailed account of Trinidadian speech using elaborated linguistic terminology, like e.g. "no diphthongization, final consonant reduction and palatalization". In some cases, differences between grammatical and phonological features were not recognised, for instance in statements like "Not pronouncing 'ed' of past tense or using present tense for past tense of verbs".

Morpho-syntactic features were the ones cited most frequently, followed by phonological and lexical features. The single most frequently mentioned group of features were phonological characteristics (scores given in brackets):

• consonantal and vocalic markers (number of responses in brackets) (36):
  e.g.: dese, dose, dis, dat instead of these, those, this, that, or gyal for girl
  
• quot;reduction of words" and final consonant cluster reduction (29):
  e.g.: doh for don't, sleepin' for sleeping, was'e for waste
  
• pronunciation/intonation in general (19)

The features that were stated on a lexico-semantic level were grouped together under:

• special lexical items (22), e.g.: to lime or to mamaguy
  
• proverbs (11), e.g.: Crapaud smoke yuh pipe
  
• special expressions (10), e.g.: We (or oui) as tag (I does go there, we)

Morpho-syntactic features were cited as follows:

• omission of copula (26), e.g.: She sleeping
  
• lack of subject-verb agreement (20), e.g.: John have two brothers
  
• lack of inversion (19), e.g.: You('re) coming to the show?
  
• reduplication (15), e.g.: A true true story
  
• pronouns (12), e.g.: All-you
  
• possessive pronouns (12), e.g.: She car
  
• use of 'eh' for negation (11), e.g.: She eh there
  
• quot;grammar" in general (8)
  
• no past tense marker/ 'de' as p.t. marker (8), e.g.: We walk to school yesterday; He de gone
  
• 'does' + verb as pres. tense habitual or cont. (7), e.g.: He does sleep
  
• 'dem' as pluralizer (6), e.g.: de boy and dem / dem girls
  
• "syntax" in general (5)
  
• 'go' as futuremarker (5), e.g.: I go tell you later
  
• double negation (5), e.g.: You don't know nothing
  
• hypercorrection (4), e.g.: I uses to go to church
  
• lack of genitive marker (4), e.g.: My brother book
  
• 'ain't' as negator (2), e.g.: She ain't sure
  
• lack of passive construction (1)
  
• not to for 'do not' (1)

The list of cited items shows that the respondents seem to have a fairly good idea of what the characteristics of Trinidadian Creole are. Some of the features would be understood as "typically Trinidadian" (like, for instance, 'eh' as negator), other features are pan-Caribbean creole features (like, for example, lack of subject-verb agreement). No instances of syntactic characteristics were given where a semantic rather than a formal difference between TEC and English can be noted (e.g. different use of had + V + en in TEC (refers to simple past) and English (where remote past is indicated).

Grammatical features were rated more often "not acceptable" (as English) (65 % of all morpho-syntactic features) than lexical (49 %) or phonological (59 %). This is not surprising considering the fact that some lexical items and phonological properties are even part of a localised acrolect (Trinidadian Standard English). The distribution was as follows:

Some of the respondents noted in this first question, that "Trinidadians from different backgrounds use varying vernacular", thus indicating an awareness that Trinidadian Creole is not one homogeneous lect and that variation in TEC depends on external factors.

Comparison with Winford (1976): In Winford's corresponding question, the respondents were asked to give examples for "bad English" in the speech of Trinidadians. Examples included, for instance, "bad construction sentences" (11), "use of slang" (12), "reduction of words" (9) or "wrong use of tense" (8) or, more specifically, "'th' pronounced as 't" or 'd'" (17). The characteristics singled out as "Trinidadian" are therefore similar to those given in 1992, albeit less detailed. However, the respondents in the Winford study appear to be much more ready to evaluate characteristics of Trinidadian Creole as negative, as "bad English". As explained in the description of the pilot study (ch. 2), this question had to be changed in the 1992 questionnaire. The nature of the question then has a different quality: the mere non-acceptance of a feature as English does not necessarily have to mirror a negative attitude towards Trinidadian Creole, but could also be interpreted as an awareness of a difference in systems. In contrast to 1970 a lot of the 1992 respondents seemed to be well aware that there are two distinct forms rather than one representing a "corruption" of the other. This notion will also be manifested in more detail in the last (open) question of the questionnaire.

3.2 Respondents' Opinion on and Evaluation of Speech Differences of Specific Groups

A subsequent set of questions aimed at the evaluation of Trinidadian speech within the English-speaking Caribbean and, within the Trinidad speech community, at an evaluation of the speech of different subgroups in the community.

1. Questions on Trinidadian Speech and Other West Indian Speech:

Practically all (except a single one) respondents agreed that there are differences in the speech of Trinidadians as opposed to other West Indians. The majority of them specified that pronunciation and intonation was the main distinguishing feature, a smaller number also saw differences in grammar and lexicon. When it came to evaluating the different varieties, more than one third (31) perceived Trinidadian speech to be better than the speech of other West Indian countries, 28 respondents thought it was equal, 21 stated that it depended on the country it was compared with and only one thought it was worse (9 respondents did not reply). Ethnic group membership was the most distinctive determinant in this response: Whereas only 23.3 per cent of people of African descent claimed that Trinidadian speech was better, the majority of the Indian respondents (52.3 %) chose this option. Of those who commented, only few respondents singled out specific countries as positive or negative examples (positive: Barbados (3); negative: Jamaica (3), Guyana (1)), 12 teachers gave no value judgment in their comments, among them some who commented that they thought it was unfair to compare different speech varieties since each variety has its special characteristics. Eight respondents commented that they found the speech of most other countries worse than Trinidadianese and nine replied that the evaluation depended on other factors (e.g. education) than nationality.

Comparison with Winford (1976): The responses in the earlier study were somewhat similar: All of Winford's respondents agreed that there were differences between the speech of Trinidadians and of other West Indians. Also, the features that were singled out as chief distinguishing ones (pronunciation and "accent") correspond with the above stated. However, more respondents in Winford's study thought that the speech of other West Indians was better than Trinidadian speech.

2. Questions on Urban-Rural Differences in Trinidadian Speech:

Ad (i): The awareness of speech differences within the community proved to be similarly notable: There was an overall agreement, that differences between town and country speech existed (90 % of all respondents). Respondents from small towns and semi-urban areas scored even slightly higher (100 % and 94 % respectively) than respondents with rural or urban background (81 % and 80 %). Three rural and two urban respondents replied that there were no speech differences; another three informants could not decide for either option and stated "sometimes" (2) or "not any more".

Comparison with Winford (1976): There was an even greater agreement among Winford's respondents on this question (96 out of 99 respondents).

Ad (ii): Differences between respondents with urban and rural background could also be observed in the evaluation of the speech of people in their own community. People from urban communities were inclined to give more positive replies than people from rural places ("Good" was stated by 75 % of the urban residents; small town: 71 %; semi-urban: 50 %; rural: 31 %). The remaining respondents evaluated the speech in their community as "fair", a single one (from a semi-urban area) stated "bad". Slight differences between respondents of African descent and Indian descent could also be noted: 61 per cent of the former rated the speech in their community as "good", whereas only 51 per cent of the latter applied the same to their neighbourhood (three of the five respondents who specified themselves as "mixed" and the one respondent of European descent ticked "good"). Age group No. 3 (36 to 45 years old) had the highest positive evaluation rate (65 %), age group 1 (up to 25 years) the lowest one (40 %).

Figure 1: Evaluation of Speech Communities by Ethnic Group, Residential Background and Age Group [Percentual Distribution]
* Number of respondents in this category is less than 10 % of the total sample – hence results have to be treated with caution.

Quite a number of respondents commented on this question. 20 teachers suggested that other factors than an urban-rural location were also important determinants for the speech in the community and that a community was usually not a homogeneous construct, again indicating an awareness of the variability of speech.

Comparison with Winford (1976): In the 1970 study, many more respondents rated the speech in their own community as "bad" (35 out of 75), the same number of ratings as "fair to good" (35), another 5 respondents were "non-committal". There was a marked difference between respondents from urban and rural areas but also between respondents of different ethnic groups. Among those who gave negative ratings were: 24 rural versus 11 urban residents and 25 Indians versus 9 Africans. In contrast, a more positive evaluation was made by only 5 rural versus 30 urban residents and 15 Indo- versus 20 Afro-Trinidadians.

Ad (iii): In reply to the question, whether or not the respondents would recognize a speaker from a town area versus a speaker from a country area, the majority of respondents (63 %) recognized speech differences "in most cases", 15 % replied with "yes" and 19 % with "usually not".one respondent answered with "no" and another one replied with "sometimes"(This option was not given in the questionnaire). There were no significant differences in the replies of the different subgroups, except perhaps, that people from semi-urban areas showed more inclination to tick "yes" and respondents with a rural background had more "usually not"-responses. Some respondents commented on what they saw as a factor in reducing speech differences between town and country speech: "In Trinidad now more and more people are becoming exposed to S.E. through TV, etc.,so they speak better." (Rural Indian respondent, age group 3).

Comparison with Winford (1976): Approximately 50 per cent of Winford's respondents stated that people would recognize them as being town-dwellers or country-dwellers. The other half answered in the negative, some were undecided. Respondents from rural areas had a slightly higher negative response percentage than people from urban places. Interesting was the response of one Indian country-dweller who mentioned that he varied his speech in the town, so that people could not tell where he was from. This indicates not only a great awareness of speech differences, but is obviously related to the result of the preceding question where especially Indians attribute little prestige to their communities in rural areas.

Ad (iv) & (v): The next two questions asked respondents whether they thought that town people, or country people, liked the way they spoke. The responses for the speech of urban speakers was largely positive (88 % of all respondents gave positive replies, whereas 11 % had ambigeous opinions and one respondent did not think that town people liked the way they spoke). The highest frequency of a positive evaluation had the subgroups "small town" (92%) and Indian (93%). Markedly less positive was the evaluation of the speech of country residents (67 % of all responses were positive, 26 % ambigous and 12 % negativ) Generally, fewer respondents thought that country-dwellers liked the way they spoke than they thought town-dwellers did. Comments like "Although they may not be ashamed of it, they would prefer if that accent was not there" or "Yes, except in instances where they are confronted by a town speaker" were more frequent for the country speech evaluation than for the town speech evaluation. However, the gap between these two evaluations was far less significant in the rural category than in the urban one:

Age group 3 had the highest positive evaluation rate for both, town speech and country speech, age group 2 had the highest rate of ambigeous replies. There was no significant contrasts between the different qualification groups.

Comparison with Winford (1976): No breakdown was given in the 1970 study on the responses to these questions. There seemed to be a general acknowledgment in the replies "that there are speech differences between town and country, and that generally these are the differences between more, and less, "correct" speech." (p. 56) However, some respondents replied that the acceptance of their speech depended on the variety they chose, and the people they spoke to, and said that they varied their speech accordingly. Speech differences due to factors such as education or social status among speakers from either country or town areas were also recognised.

3. Questions on Ethnic Differences in Trinidadian Speech:

Speech differences between the two major ethnic groups (African and Indian) in Trinidad were included, even though there was no corresponding question in the Winford study. It was considered to be interesting, however, since observation⁶ indicated a great awareness in this respect also.

Ad (i): The replies to the question whether respondents thought there was a difference in the speech of Afro-Trinidadians and Indo-Trinidadians varied considerably: Almost one third (29 %) of the respondents were definitely positive, 41 % stated "in most cases", whereas 26 % stated "usually not" and 6 % answered in the negative. In some of the comments it was mentioned that with an increase of education among the East Indian population and a decline of Trinidad Bhojpuri the differences became less important.

The breakdown into the various subgroup categories showed that

• Afro-Trinidadians had a slightly higher "yes" and "in most cases" rate than Indo-Trinidadians.
• Respondents from small towns and rural areas had the highest negative ("usually not") reply rate.
• Teachers with a Teacher's Certificate had the lowest negative rate, trained graduates the highest one.
• There were no marked contrasts in the replies of different age groups.

Ad (ii): As characteristics of differences between the speech of African and Indian Trinidadians, pitch, "accent" and pronunciation (39 responses) were pointed out most frequently, followed by differences in vocabulary and special expressions (29 examples), seven respondents also mentioned grammatical differences.

Ad (iii): When asked to judge whether the speech of Afro-Trinidadians or Indo-Trinidadians was better or whether the speech was "equally good or bad", most respondents did not attach any positive or negative values to ethnic speech differences (78 % of all respondents). 19 per cent of the teachers said that Afro-Trinidadian speech was better, 3 per cent considered the speech of Indo-Trinidadians to be better. Out of the respondents of African origin, almost one third (31 %) preferred their own way of speaking, none of the Afro-Trinidadians considered Indo-Trinidadian speech as better. Only a few Indians (6 % of the Indian resp.) judged the speech of their own ethnic group better, whereas twice as many (12 %) thought that Afro-Trinidadian speech was better.

The notion that (especially country-) Indians in Trinidad showed more insecurity in their abilities to speak "correctly" and had a less positive attitude towards their own speech than Africans had, is also expressed in the Winford study. One of the causes for this could be the delay in education the Indians have had in Trinidad. That this attitude has changed or is in the process of changing can be shown in the distribution of responses, different age groups gave:

It is obvious in this table that with an increase of age the respondents show more inclination to evaluate the different varieties. The response, that the speech of Africans is better, was also over-represented among urban respondents than among other ones (36 % of the urban respondents gave this reply). No significant differences between respondent with different qualifications could be observed.

4. Questions on Education as a Factor for Speech Differences:

Ad (i): The distribution on the reply of this question was as follows: "yes, definitely": 24 %; – "in most cases": 63 % – "usually not": 10 % – "not at all": 2 % – "sometimes" (not given as an option in the questionnaire): 1 %. Afro-Trinidadians seem to consider this factor more important than Indo-Trinidadians do (38 % of the Africans versus 17 % of the Indians stated "yes, definitely"). In their comments, 20 teachers stressed education as the most important determinant for speech, 5 pointed out that education is one factor among others and 11 respondents mentioned that the speech of the individual varied (e.g. due to who they spoke to, situation and topic), e.g. "It depends on the personality of the person and environmental background, e.g. some highly educated people use the vernacular most times except offically. On the other hand you have the "show offs" who are not very educated and speak the 'Queen's English' most times." (Indo-Trinidadian respondent). There was no corresponding question in the Winford study.

5. Questions on Social Status as a Factor for Speech Differences:

Ad (i): According to the opinion of the teachers who were interviewed, social status appears to be reflected less in the speech of a Trinidadian, than education does. ("Yes, definitely": 17 % – "In most cases": 59 % – "usually not": 18 % – "not at all": 5 % ("sometimes": 1 %)) Whereas there was hardly any difference in the judgement of Indian respondents between education and social status, African respondents saw social status markedly less reflected in the speech of Trinidadians. Commenting on this question, 11 respondents stressed the importance of social status as a significant marker of Trinidadian speech, 10 suggested that other factors were equally important and 8 teachers said that it depended on the situation and the circumstances. Again, no corresponding question in the Winford study.

To sum up the results of this part of the questionnaire (part B) one may conclude that there seems to be great awareness of speech differences of the various subgroups among the respondents. In addition the outcome suggests that respondents are very conscious of the variability of speakers according to various factors including domains and situation. Speech differences between "town speech" and "country speech" were generally recognized – "country speech" was evaluated less positive than "town speech". However, when asked whether the speech of the individual could be recognised in terms of a town/country background, the respondents were less certain of the significance of this factor than they were, for instance, of educational level as a determinant of the speech of a Trinidadian. The perception of speech differences between Afro-Trinidadians and Indo-Trinidadians was the category where the responses showed the greatest diversity. The fact that age played a considerable role in the evaluation of the speech of Indo-Trinidadians versus Afro-Trinidadians reflects both that these differences are becoming less significant and that negative values previously attached to the speech of Indo-Trinidadians are weakening. Educational level was regarded to have a more decisive influence on speech than social status; the variability of speakers according to the circumstances of the speech event was pointed out repeatedly.

3.3 Linguistic Security and Evaluation of Respondents' Own Speech

A further set of questions aimed to find out the respondents' opinions about their own speech, whether they had ever felt any need to change their way of speaking, whether they had been in situations where they had felt that their English was not adequate, and what possible influences they had had for any change or "improvement". It was considered important to find out whether of not the respondents felt linguistically secure and whether this linguistic security had any influence on the code selection they were asked to perform in the subsequent section.

1. Questions on Linguistic Background of Respondents:

The first two question asked for information on the linguistic background of the respondents. It can be safely assumed that some form of Trinidadian Creole is the first language of all respondents, the other languages which were previously spoken in Trinidad being restricted to only very small groups of speakers. However, it was considered whether influences from one or more of these languages (like e.g. Bhojpuri or French Creole) their parents may have spoken and whether or not the respondents themselves continued to speak these languages could have had an effect on their attitudes towards TEC, Standard English and the language(s) their parents speak/spoke. The fact whether or not the respondents continued to speak the language of their parents could again be the result of attitudes towards these languages and towards English as the language of education and social advancement: A respondent of mixed ethnic background (African/Chinese/Spanish/ French/Amerindian) confirmed this notion with her comment "My father was able to speak Spanish (because of his mother) and Patois. My mother could speak Patois. They were very careful not to teach us."

Ad (i): Almost one third of the respondents (31 %) stated that their parents spoke other languages than English or TEC. 18 respondents (= 21 %) mentioned "Hindi" or Bhojpuri, another 6 stated Patois, one Spanish and two Patois and Spanish as languages their parents speak or spoke. Whereas the ethnic distribution held no big surprises (although one person of African descent stated "Hindi"; the other ones with "Hindi" or Bhojpuri background were, of course, of Indian origin, the respondents with Patois- or Spanish-speaking parents were of African or mixed descent), it was somewhat astonishing that the distribution in the different age groups was nearly the same. One would have thought that among older respondents there was a higher rate of people with a different linguistic background than in age groups 1 and 2. This proved to be not the case.

Ad (ii): Out of those respondents who claimed they spoke and understood another language, there were seven persons who claimed to have an active command of Bhojpuri/Hindi, another six stated that they were only able to understand it. For some of those who asserted to be able to speak or understand Patois/French or Spanish, it is difficult to decide whether these languages were related to their linguistic heritage or not. Quite a number of the teachers had an active and/or passive command of Spanish, the language which is taught most often as a second language in Trinidad. The number of those respondents who continued to speak the language of their parents proved to be too small and the significance of their competence in these languages seemed to be too unpredictable, so that the outcome of this part of the questionnaire was not pursued any further.

Comparison with Winford (1976): The number of respondents who had parents with a different linguistic background than English/TEC was considerably higher, especially for Hindi/Bhojpuri. 33 out of 38 country Indians and 20 out of 30 town Indians stated that their parents spoke "Hindi" at home. Eleven of the country respondents had a working command of "Hindi", seven only understood it. Of the town respondents, eight were able to speak "Hindi", three only understood it to some extent.

2. Questions on Respondents' Own Speech Evaluation:

Ad (i): The first question was answered positively by 67 per cent of those who replied (86 replies). 31 per cent stated that Trinidadian Creole was not their most natural speech form, two per cent stated that they were not sure. The breakdown for the different ethnic groups showed that there were strong differences between Indo-Trinidadians and Afro-Trinidadians on this question: Almost half of the African respondents (48 %) answered with "no", while only 21 per cent of the Indians thought that TEC was not their "most natural" language. Slightly more male than female respondents replied affirmatively, the same was true for respondents from rural and small town areas as opposed to urban and semi-urban respondents. Age appeared a most decisive role in determining the answers to this question:

* Number of respondents in this category is less than 10 % of the total sample – hence results have to be taken with caution

Comparison with Winford (1976): In the 1970 study, the overall response to this question seemed quite similar to the 1992 outcome: The majority (51 out of 84) of respondents said that Trinidadianese was their most natural medium of speaking. 20 repondents stated that they preferred to use "correct English", and 4 respondents saw English as their most natural way of speaking. Another nine respondents pointed out that both, Trinidadianese and English, were required at different times. However, quite a lot of Winford's respondents chose not to answer this question. In contrast to the results of the breakdown above, the vast majority of those who commented that they preferred "correct" English, were rural Indians in Winford's study, the group which, in reply to other questions, showed the greatest insecurity concerning their own speech.

Ad (ii): In the evaluation of the respondents' own speech more than three quarters (77 %) of the teachers rated their speech as good, 23 per cent considered it to be "fair". None of the respondents rated their speech as "bad". Slightly more African than Indian respondents had a positive opinion concerning their own speech (African: 80 %; Indian: 75 %; *Mixed: 60 %; *European: 100 %). Female respondents also attached a slightly better value to their own speech (72 % of the male and 79 % of the female respondents rated their speech as good). Age group 3 (36–45 years old) had the highest percentage of positive replies;the qualification breakdown showed that those teachers with a Teacher's Certificate (usually primary school teachers) had the lowest rate of "good"-replies (70 %). Perhaps the most striking contrast was to be found in the urban/rural breakdown:

Figure 2: Own Speech Evaluation by Residential Background.

Comparison with Winford (1976): 48 out of 85 respondents said they had no difficulties in speaking correct English, while 37 (= 43.5 %) stated that they had, either as a general rule, or sometimes. The differences in response between the subgroups here (Indian-African; rural-urban) are quite notable. The majority of those who reported difficulties were Indians from rural areas.

Ad (iii): The total number of respondents who stated that they had tried to change their way of speaking was slightly less than one third (30.7 % or 27 out of 88). Most pronounced was the diversity between the different qualification groups:

Figure 3: Speech Change by Qualification.

Most teachers who fell into the category "other" had more than one qualification, for instance, a teacher's certificate and a university degree in education. Perhaps one could characterise them as highly ambitious and draw conclusions as to the connection between these ambitions and their (attempted) speech change. Except for the youngest age group there was a marked decrease of those who stated that they had tried to change their speech with an increase in age. More male than female respondents and more urban than rural teachers reported attempts to change their way of speaking. There was hardly any difference in the responses of the two big ethnic groups.

Comparison with Winford (1976): The majority of Winford's respondents (47) reported that they had tried to change or improve their speech at one time or another, 32 trainee teachers said they had not. Another nine stated that they thought of their speech as acceptable "Trinidadian", but stated that there were some occasions on which they would attempt to improve it.

---

PhiN 15/2001: 63

Ad (iv): Almost half of the respondents (49 %) claimed that had not had pressure or influence from outside to alter or improve their speech. 14 % responded with "yes" to this question but gave no specification. As specific influences to change or improve their speech, respondents cited education (26 %), parents (8 %), abroad (1 %) and other (1 %). Some of the respondents commented on the sometimes lifelong effect of education as the most decisive influence: "Teachers knowing the problem with speech try [to change or improve one's speech, S.M.]. My parents were teachers. My mother, 81 years old, still corrects any wrong grammar while I'm speaking." (female respondent, age group 4, who rated her speech as "fair").

Comparison with Winford (1976): Almost all of the trainee teachers in this study said they had received strong encouragement from their teachers to "improve" their speech. By far less reported encouragement from their parents. Some said that they themselves tried to "improve" their speech, so there had been no need for their parents or teachers to "correct" their speech.

Ad (v): In response to this question, 72.2 % of the respondents stated that they had never been in a situation where they felt their English to be inadequate, 28.8 % said they had had such an experience. Some of the affirmative respondents specified the situations in their comments. Among the situations which were most stated were "public situations" (16.7 % of the "yes" responses), "abroad" (16.7%) and "academic environment" (11.1 %). Residence groups show the most marked difference in this category, followed by sex, age groups and ethnic groups:

Figure 4: Assessment of Speech Inadequacy (English) in Certain Situations by Sex, Residential Background, Age Group and Ethnic Group

---

PhiN 15/2001: 64

Comparison with Winford (1976): Among the trainee teachers in this study, there was a substantial divergence between the numbers who reported difficulties with "correct" English, and those who said that they had been in an embarrassing situation due to an inability to speak "correctly". While a considerable number of respondents stated the former, only 19 out of those 60 who responded to this question reported experience of the latter kind. Winford explained the reason for this gap by pointing out that the language skill requirements in a formal situation may not be as high as the "ideal standard of correctness" the respondents had in mind when evaluating their own speech.

Ad (vi): The notion that Standard English can also be inappropriate at times and that its use may be interpreted as a "show-off" behaviour in certain situations was already established in the respondents' comments on variability in 3.2. While 44 % of the respondents claimed that they had been in a certain situation where people reacted in a hostile way when they spoke Standard English, 56 % stated that they had never had this experience. Some respondents explained the negative reactions they had experienced, e.g. "As if I was showing off." (Rural Indian respondent, age group 4) or "People often misconceive the use of Standard English as a way of being pretentious or insincere." (Semi-urban Indian respondent, age group 2).

A slightly higher rate of Indian than African respondents reported an experience of this nature. A marked difference could be observed in the responses of male versus female teachers: More than half of the male respondents (60 %), but only 35 % of the female respondents had encountered a situation where their use of Standard English was regarded as inappropriate.

Comparison with Winford (1976): In Winford's study, the respondents were asked whether people ever annoyed them speaking "correctly". Four out of 64 respondents reported being annoyed at the use of "correct English" as a general rule. The remaining 60 respondents were divided between those (30) who said that they were never annoyed by the use of "correct English", some adding that, rather, they admired it, and those (30) respondents who stated that they were annoyed with people whose English was "affected".

As a brief summary of the findings in this section of the questionnaire one may note that most respondents saw Trinidadian Creole as their "most natural way of speaking", increasingly so among the younger respondents. The speech evaluation of the respondents' own speech was overall quite positive and seemed to be less dependent on geographical and ethnic factors than in Winford's study, where mostly rural Indian respondents reported difficulties in speaking quot;correctly". Contentment about and self-confidence in one's own manner of speech can be assumed where respondents rate their own speech as "good" and have never felt embarrased about their speech in formal situations.

---

PhiN 15/2001: 65

One could argue that the strong difference between, for instance, urban and small town respondents are the result of a difference in exposure to formal, "public" situations. This, however, is refuted by the result that small town residents show even more self- confidence that urban dwellers do. Furthermore, the fact that all respondents belong to the same professional group and, despite qualification differences, they all had some kind of formal training (and therefore may have been in similar kinds of "public"situations or "academic environments") reduces the significance of possible exposure in different resident groups. The rating of the different subgroups for question (ii) (how the respondents theoretically evaluated their speech) and (v) (how they experienced their ability to speak "well" in an actual situation) did not always correspond. However, there were considerable overlappings. The subgroups which rated highest in terms of contentment and self-confidence were:

• Trained graduates;
• Small town residents;
• Age group 3 (36–45 years old);
• Females;
• Of African origin.

3.4 Domains of Language Use

Comments to questions of evaluation of different subgroups (3.2) showed a considerable awareness that the language variety one chooses depended much on interactional variables such as addressee, topic, etc. A specific section in the questionnaire consisted of a number of questions where the respondents were asked to make a linguistic choice – between Trinidadian Creole and Standard English – in different domains:

1) Questions on Linguistic Choices on an Interpersonal Level

2) Questions on Linguistic Choices in the Classroom

3) Questions on Linguistic Choices in Public Spheres

Ad 1): The table shows the overall distribution of Creole versus Standard choices on an interpersonal level. Interesting here was that, although this option was not explicitly given, quite a few of the respondents ticked both, the Creole and the Standard option. This could be interpreted as an awareness that there was variation and switching between the codes even within a specific situation.

---

PhiN 15/2001: 66

Table 7: Language Choice on an Interpersonal Level
		Creole	Standard	Both*
a)	With Parents	76.1 %	17.0 %	06.8 %
b)	With Spouse	72.2 %	22.8 %	05.1 %
c)	With Children	28.2 %	61.2 %	10.6 %
d)	With Friends	65.5 %	19.5 %	14.9 %
e)	At Work, With Colleagues	22.7 %	59.1 %	18.2 %
f)	At Work, in Classroom	03.4 %	91.0 %	05.6 %
g)	In Church	09.5 %	79.8 %	10.7 %
h)	When "Liming"7	84.9 %	09.3 %	05.8 %
i)	When Quarrelling	79.5 %	12.0 %	08.4 %
j)	Telling Jokes	92.1 %	05.6 %	02.2 %
k)	New Acquaintance	10.2 %	84.1 %	05.7 %
l)	When Introduced to a Trinidadian Abroad	30.6 %	62.4 %	07.1 %

* Not given in questionnaire.

Perhaps most striking in this table is the rating for the use of TEC "with children": While the majority of the respondents chose TEC as the preferred code in interaction with parents (76.1 %), spouse (72.2 %) and friends (65.5 %), only 28.2 % stated that they would use Trinidadian Creole with children. It seems that parents still feel that Standard English as the language of education and social advancement is of more use to their children than Trinidadian Creole. Somewhat surprising was also the difference in responses to category k) and l): While only 10 per cent of the teachers stated they would use TEC when making a new acquaintance, three times as many said they would prefer TEC to Standard English when meeting a Trinidadian abroad. This result indicates that the respondents ascribe more importance to TEC as a language of identity and solidarity in an unfamiliar environment than in Trinidad (where one's identity as a Trinidadian does not have to be stressed). On the whole, the outcome of Table 12 shows quite clearly, that most of the categories where Creole was rated higher than Standard were "private" or "informal" situations where intimate persons were involved (except for category c!), whereas Standard English was dominant in those situations of more formal character (like work, church, in interaction with strangers). It is also notable that the vast majority of respondents chose Standard over Trinidadian Creole "at work, in classroom", since this result is very inconsistent with the responses to questions on language choice in the classroom.

---

PhiN 15/2001: 67

The correlation with extralinguistic factors showed that there was an overall difference between the two main ethnic groups: Indo-Trinidadians chose Creole over Standard far more frequently than Afro-Trinidadians. The differences were most pronounced in the first three categories, in "family matters". Here, the most striking difference occurred in category c): While only 10 per cent of the African respondents said they would apply TEC with children and 81 per cent stated that they would use Standard ("Both": 9 %), 43 per cent of the Indian teachers interviewed selected Creole in this category while 47 per cent chose Standard ("Both": 10 %). There was also a significant difference between respondents of different qualification. Teachers with a Teacher's Certificate (= mainly primary school teachers) chose Creole less than their colleages of other qualification groups in nearly all categories. Sex and age seem to play a role in determining a domain-consciousness in specific situations: Female respondents selected Standard slightly more often than their male colleages in more formal situations. When examining the breakdown of the different age groups, it has to be noted that in some cases the situation may be essentially different for respondents of different age groups: A 26 year old teacher may have a different relationship to his/her colleague than a 55-year-old and may therefore perceive the situation as less formal than the older colleague (while three quarter of the age group 4 respondents said they would apply Standard English with colleagues at work, less than half of the age group 2 respondents selected this option). With decreasing age, an increasing percentage of respondents stated a preference for TEC also in other more formal domains. This suggests that there may be a change in the rigidity of character of formal and informal domains.

Comparison with Winford (1976): The outcome of corresponding questions in Winford's study was strikingly similar in the overall response. Almost all his respondents stated that they would use "correct English" at work and in church, the vast majority agreed that they would use TEC when "liming", telling a joke or quarreling. A substantial majority also reported using Trinidadian Creole to their parents and to their friends, while a smaller number chose "correct" English. However, unlike in the result of later research, more Indians than Africans reported to use Standard English in "informal" situations.

Ad 2): Only 3.4 % of the respondents stated that they would use Trinidadian Creole in the classroom in the previous part, 5.6 % said they would use both codes. However, when asked in more detail, the teachers seemed to be aware that there is considerable code-switching and/or mixing of the codes at times. The following options were given in the questionnaire and are ranked according to frequency:

---

PhiN 15/2001: 68

Qu: For which purposes would you use Trinidadian Creole in the classroom?

	a)	to break classroom tension	76.7 %
	b)	to facilitate student comprehension	74.4 %
	c)	to increase student interest and response	60.0 %
	d)	to prevent students from getting bored	48.9 %
	e)	for personal conversation with students	43.3 %
	f)	for reprimanding and scolding	21.1 %
	g)	to keep discipline and control	18.9 %
	h)	other purposes	06.7 %

Different preferences could be observed between primary and secondary school teachers and between male and female respondents:

Table 8: Trinidadian Creole in the Classroom
		Primary School	Secondary School	Male	Female
f)	repremand. and scolding	26 %	15 %	35 %	14 %
g)	discipline and control	35 %	11 %	35 %	11 %
e)	personal conversations	36%	53 %	50 %	40 %

A higher percentage of primary school teachers than secondary school teachers said they would use Trinidadian Creole "for reprimanding and scolding" and "to keep discipline and control", consequently in situations, where TEC is used rather agressively. Secondary school teachers had a significantly higher rate of using TEC "for personal conversations" than primary school teachers. Male teachers stated more frequently than female teachers that they would use TEC in most categories, the difference was most significant in the two categories where negative connotations would be attached to the Creole language. However, the overall relatively low rating in these two categories suggests that those teachers who claimed they would apply TEC here would also do so for other purposes (like e.g. "to facilitate student comprehension" or "to break classroom tension" )

Comparison with Winford (1976): In the Winford study there were two corresponding questions. The respondents were asked a) whether they thought teachers should use Trinidadianese to assist in explaining things in class, and b) whether they thought Trinidadianese should be taught instead of English. Of 61 respondents who replied to the first question, 39 were in favour of a use of Trinidadianese in the classroom, the other 22 were opposed to it. 53 respondents replied to the second question. Of these, 43 were against the idea of replacing English with TEC, while only two teachers thought it should be replaced, and the other eight stated that both should be taught.

---

PhiN 15/2001: 69

Ad 3): This question was designed to determine the acceptability rate of Trinidadian Creole in various public sectors, like arts, media, political and academic life. The media sector received special attention in this part of the questionnaire, because the media is one of the domains in Trinidad where external influence, e.g. American tv stations, is most prominent. It was therefore interesting to observe whether there was a difference in the acceptability rate in different genres of print- and audio-visual media. The following table shows the responses of the total sample:

Table 9: Acceptability Rate of TEC Use in Different Media Genres
	Appropriate	Inappropriate	Humerous
Literature	44.2 %	14.0 %	41.9 %
Plays	58.1 %	1.2 %	40.7 %
Music	64.1 %	19.2 %	16.7 %
Radio, music programme	41.9 %	50.0 %	8.1 %
Radio, local issues	46.2 %	50.0 %	3.8 %
Radio, political programme	10.1 %	84.8 %	5.1 %
Radio, sports programme	9.3 %	94.0 %	1.2 %
Newspaper, reports	4.8 %	94.0 %	1.2 %
Newspaper, comments	14.5 %	81.9 %	3.6 %
TV – local issues	31.3 %	63.8 %	5.0 %
TV – documentary programme	7.3 %	92.7 %	0.0 %
TV – political programme	13.8 %	82.5 %	3.8 %
TV – soap operas	51.9 %	21.0 %	27.2 %
Politicians, electoral campaign	46.4 %	38.1 %	15.5 %
PM, inauguration speech	6.3 %	94.9 %	1.3 %
Politicians, budget announcement	3.8 %	94.9 %	1.3 %
University lectures	7.6 %	91.1 %	1.3 %
International conferences	3.7 %	96.3 %	0.0 %
Inter-regional events	6.0 %	94.0 %	0.0 %

---

PhiN 15/2001: 70

The table shows that while there was considerable agreement that the use of Trinidadian Creole in literature, plays and music was not inappropriate (even though a large portion of the respondents seems to see TEC as a "humerous" characteristic in these genres and may therefore not think of it as appropriate in a "serious" context), in other sectors of public life and in different genres in the media Trinidadian Creole was clearly not appropriate, according to the respondents' statements. In a third share of items, the responses were very divergent: respondents disagreed most significantly in their views on the appropriateness of Trinidadian Creole in items like "radio, music programme", "radio, local issues" and "politicians, electoral campaign", less notably in "tv – local issues" and "tv – soap operas". The correlation with extralinguistic factors showed that male and female respondents, and again the two main ethnic groups had the most significant differences in their responses.The responses of female teachers differed much more according to the specific sector in question than the responses of their male colleagues did, i.e. a higher proportion of women then men found Creole appropriate in the art forms and in the "local" media genres, at the same time, a higher percentage of females than males thought Creole to be inappropriate in those sectors where politics were involved or in an "international scenario".

Figure 5: Acceptability Rate in Selected Public Spheres by Male and Female Respondents

Indian respondents had an overall higher "appropriate" rate than Africans. Indo-Trinidadians were more inclined to state that Creole was an appropriate code in the art forms, whereas Afro-Trinidadians tended to see Creole as humerous in this question.

Comparison with Winford (1976): Two of the questions asked in the Winford study were roughly comparable with this section. Respondents were asked 1) what they thought of the use of Trinidadianese in literature and 2) what they thought of the use of Trinidadianese in certain articles in the newspaper and in certain advertisements over the radio. In reply to the first question, 83 out of 108 respondents (= 76.8 %) thought the use of Trindadian Creole was "appropriate", eight (= 7.4 %) thought it was "humerous" and 17 (= 15.7 %) regarded it as "undesirable".

---

PhiN 15/2001: 71

Therefore the percentage of those who stated it was undesirable to use TEC in literature corresponds largely with the 1992 result, however, far less repondents in 1970 than 1992 thought that TEC was a means of humerous effects in these genres. The second question was answered "appropriate" by 89 out of 107 respondents (= 83.2 %). Six trainee teachers (= 5.6 %) thought the use of Trinidadian Creole in the mass media was "humerous", while 12 (= 11.2 %) regarded it as "undesirable". Thus, far more respondents thought the use of TEC was suitable in the media in 1970 than the respondents in 1992 did. However, it must be noted that radio advertisements and newspaper articles (it is not quite clear whether specific newspaper articles were the basis of investigation in the former study) do not cover a large proportion of media genres.

To sum up this section, a functional distribution of the codes in interpersonal formal and informal situations remained similarly strong in the present study as had already been shown in Winford's research in 1970. Extralinguistic factors which seemed to influence the responses in this section most were ethnic membership and qualification: Respondents of East Indian origin and graduates chose Creole over Standard more often than Africans and teachers qualified with a Teacher's Certificate. In contrast to the Winford responses where Indians – who at the same time had a low self-esteem regarding their own speech – often stated they would use SE in informal situations, this particular group had a higher "Creole"-rate in practically all interpersonal situations. Obviously, one's own linguistic security and perhaps greater variability in speech is somehow related to a conscious acceptance of Creole in specific situations. Some of the interpersonal situations seem to have changed or are in the process of changing in terms of the formality which is attributed to them: e.g. in interaction with colleagues and in church, Creole was increasingly chosen by the younger respondents. Whereas Creole was hardly chosen over Standard English as the language to be used in the classroom, the majority of respondents admitted using Creole in the classroom for specific purposes. In public spheres Creole was largely accepted in the art forms and Standard English was the preferred code in most other sectors of public life. In specific genres of the media and politics the preferences were ambiguous. The responses of female teachers differed significantly more than those of male respondents in their distribution of the codes. Women tended to regard Creole as the language of art forms and in local contexts while at the same time they rejected Creole in international contexts and in matters of politics to a higher degree than men did.

3.5 Opinions on the Role of Trinidadian Creole in Trinidad in General

The last question was an open question and concerned respondents' views on the role of Trinidadian Creole in Trinidad today and in the future. The responses to this question ranged from detailed statements to short essays. The following points were most frequently mentioned (number of repondents is given in brackets):

---

PhiN 15/2001: 72

1.	TEC is a means of effective communication and it is the first language of Trinidadians. (32)
2.	TEC is a means of identity in Trinidad. (18)
3.	TEC is expressive in oral communication. (18)
4.	TEC is not appropriate in written communication or as the language of education. (13)
5.	TEC should be encouraged in literature, music, etc. (11)
6.	TEC should be applied/accepted more in general. (11)
7.	TEC should not be applied/accepted more in general. (10)
8.	People have become more variable in their speech and both, Standard English and TEC should be encouraged. (10)
9.	TEC is inappropriate and not useful in international contexts. (9)
10.	There are pros and cons for the use of TEC in formal education. (6)
11.	TEC is appropriate in written communication and formal education. (5)
12.	TEC is appropriate in literature, music, etc. (5)
13.	TEC is a means of national unity. (4)
14.	TEC is not appropriate as a means of oral expression. (4)
15.	There are pros and cons for the use of TEC as a means of effective communication. (3)

Open questions are difficult to evaluate, also since they may provide room for expression especially for those respondents who have particularly strong views on the subject whereas those who are relatively undecided may not feel the need to add anything to what they have already stated in the closed-type responses. However, it can be said that the given responses provide a fairly good idea of what the respondents regarded as being positive or negative in the use of TEC in general. Some examples may illustrate this best:

1.

"Trinidadian Vernacular is very important in every-day use; however, I feel that the people know how much to use it – as a result of the increased number of people who attend Secondary School today. It would continue into the future. However, I do not foresee it becoming the accepted language of the region. This is so because of the difficulty it would present to the outside world. Personally, I feel the Trinidadian Vernacular is much more attractive and picturesque than Standard English." (Resp. 008: M/Age 3/Pr./TC/Semi-urban/Ind.)

---

PhiN 15/2001: 73

2.	"I think the Trinidad Vernacular is important because to many of us it is the first language that we learnt. It is also the language that we can use comfortably with our friends and family. But we have to realise that Trinidad is not the little island it used to be many years ago. [...] If we want to be able to communicate effectively with the world out there, we have to realise that the Vernacular is inappropriate. We need the common base – Standard English – to be internationally accepted wherever we go whether it is to University in Barbados or Cambridge or England. I think it is important for us to recognise our Vernacular as a language but we must be aware of its limitations. As a teacher it is my aim to help students use SE language properly and fluently through constant practice without destroying their spontaneity and creativity." (Resp. 004: F/Age 2/Pr./TC/Semi-urban/Afr.)
3.	"Trinidadian vernacular has its place in the arts. For those who have no other form of language for communication and lack the means to change, then there is little choice but to use such language. Understanding is vital to communication. If we are to keep a proper link with the world then we must be understood by our partners in trade or other interest. In this case Trinidadian Vernacular is useful only in the local setting." (Resp. 082: M/Age 4/Sec./TC/Semi-urban/Afr.)
4.	"Trinidadian Vernacular can enrich speech and comprehension so much, it will continue to play an important part in our development. Historically, it is important, since many words, sayings, grammatical expressions tell us a lot about our past. It can be very "rootsy" and thus have a depth in expression! We need to keep the good, like sayings of old and throw out that which will hinder communication/comprehension with others not indigeneous to Trinidad or this region. Vernacular has its place. Its status must change. The wealth and warmth of our people, our uniqueness must be allowed to be transmitted through our vernacular." (Resp. 075: F/Age 3/Sec./Other/Urban/Afr.)
5.	"Not everyone has a command of Standard English. Hence, it is easier and most comfortable, and effortless to communicate at any level of education (people). Young infants must communicate in this way at first in order for learning to take place: This is how they/we as Trinidadians communicate best, while Standard English is being read, written, and then used. Hence, in later years we can read, write good English and speak vernacular. I see vernacular as a separate language, with its own grammar, diction, sentence structure and words not found in the English dictionary. The influence of many cultures have created a unique vocabulary in native language. This gives Trinidadians a unique personality not easily found elsewhere. Here people of all races may speak in vernacular, hence Trinidadian. The role of vernacular today/future? No need for discussion. It is separate, natural, and used whenever/wherever necessary, and has its good for future continued easy communication of all people of this country. In song, poetry and otherwise the learning of Std. English has become normal and traditional without us realising it. Thus we most times, if not at all, do not write or spell how we speak. It makes good fun reading vernacular when it is written by someone." (Resp.045: M/Age 2/ Sec./ Other/Urban/Mixed)

---

PhiN 15/2001: 74

6.	"It is an integral part of our society and is rich in our linguistic tradition. It has a vibrant role to play and should not be described as 'Bad English' since it is a language in its own rights. I think people have to be better educated about the richness and evolution of Trinidadian Creole and its importance in their everyday lives. The media especially has an important role to play." (Resp. 029: F/Age 1/Pr./TC/Semi-urban/Ind.)
7.	"The Trinidad vernacular identifies the people of Trinidad. It can be used in plays, poetry, literature on the whole. It has its own intrinsic richness and beauty and it is part of Trinidad's cultural heritage. The Trinidad Vernacular is a rich mixture of Indian, African, Arawak, Carib, French, Spanish and English. Such rich language, whether accepted or not by world standards cannot be allowed to be relegated to 'second class'. Our Trinidad writers must continue to promote the vernacular and our children must be encouraged to use it and to be comfortable with it. Of course, they must also use the Standard for formal writing and speaking." (Resp. 044: F/Age 3/Sec./TG/Urban/Euro.)
8.	"Trinidadian Vernacular is a dying form. It will be replaced by Standard English. The Standard in terms of grammar is the form that is mostly used in school; in formal situations it is insisted upon. The Vernacular is not respected as a reflection of refinement in culture. As we move towards refinement in culture, the language will become more refined. Vernacular is a compromise between a lost language and the present reality in which people find themselves. May remain in art forms (e.g. calypso)." (Resp. 086: M/Age 3/Sec./TG/ – /Afr.)

In summary, it was notable here that most respondents see the limitations of Trinidadian Creole in so far as it does not function as a language of international communication while only in example no. 6 TEC is rejected per se, for its "lack of refinement". The possibility to seek job opportunities outside Trinidad (and/or the Caribbean) has played a major role in the Caribbean and may have been of even higher importance in the early nineties with their economic recession and a higher unemployment rate than in the 1970's when the "oil money" brought in a certain international independence. It can be concluded that one's ability to speak or master Standard English may be more important again today than in the early seventies. All in all, many of the comments reveal quite clearly that, while most of the teachers regard Trinidadian Creole as "having its place" in society (and also in the classroom), the fact that TEC is not fully accepted in other public spheres (for instance, in professional life) makes it necessary for the teachers to continue the emphasis on the teaching of Standard English. However, it is also obvious that the increase in variability of speech does not result in the loss of significance of Trinidadian Creole but has rather led to a greater domain consciousness.

---

PhiN 15/2001: 75

4 Conclusion

Like in other social science studies, studies on language attitudes cannot be reproduced under exactly the same conditions. What has, however, become evident is the fact that, as social reality changes, the expressions and terminology for politically and socially sensitive definitions change accordingly. While this may be an interesting development in itself – after all it requires a reflection process on the emotive content of formerly used terms – most of all what needs to be emphasized is that, along with the terminology, the concepts of evaluation seem to be disappearing. This gradual change in language attitudes is not limited to Trinidadian society but may be a general trend in the Anglo-creolophone Caribbean, as Shields-Brodber's (1997: 62–63) recent observations on changing language attitudes in Jamaica would substantiate. She states that despite traditional ideas still prevailing about Jamaican Creole as an inferior relative to English,

there are many others [speakers, S.M.] who affirm Jamaican Creole as the language of national identity, the first language of the majority and like any other language, available for, and easily facilitating the self expression and communicative purposes of its speakers, all of whom may never either achieve mastery of English or suffer any undue consequences as a result (1997: 63),

and she connects this change in evaluation to a gradual change in language use, with JC being increasingly used by speakers of all levels of the socio-economic hierarchy in their public/formal communication, "thereby providing it with newfound legitimacy in that domain" (1997: 63). As JC adopts more and more domains formerly reserved for Standard English, the traditional diglossic relationship is being steadily eroded and transformed. Therefore it is, in fact, SE rather than Jamaican Creole that undergoes a process of marginalisation.

Considering the data above, one can safely assume that in Trinidad, too, the gloomy prediction of the last informant's quote ("Trinidadian Vernacular is a dying form") cannot be confirmed. Nor would the findings here suggest that – according to the subjective perception of the speakers – Standard English will lose its functions at the expense of TEC. On the contrary: while the viewpoint of what is the appropriate code may have shifted in some domains, the awareness of a functional distribution is as great as ever. What appears to be at the verge of disappearance, however, is the negative evaluation of TEC on grounds of false notions of "correctness" and "incorrectness" as well as aesthetic value judgements based on the belief that one variety is a corrupt form of the ("pure") other. So, while both TEC and Standard English are there to stay, the notion of TEC as "Bad English" truly seems a dying form.

---

PhiN 15/2001: 76

The forces involved in such a change in language attitudes are manifold and can be taken to range from socio-economic and political influences to a change in language policy in education. In view of the respondents' statements it seems that the latter has played an important role as regards the cognitive component of attitudes towards TEC. If we take Winford's statement from 1976, that

[...] it seems clear that most respondents were not at all conscious of the fact that the Creole variety of Trinidadian English has its own grammatical system which operates according to different rules from those to Standard English. Informants generally show a great willingness to interpret what are essentially grammatically correct Creole structures as 'corruptions' of 'good English' (1976: 51)

the difference to the the awareness of TEC as a separate linguistic system in the later study is remarkable. Following Ajzen's (1989: 247) hierarchical model of attitudes, one may thus attribute considerable influence of this on the language attitudes as a whole.

It seems that there is much room for further research in order to fully understand the processes of change in language attitudes and language use in the Anglo-creolophone Caribbean which so clearly defy the earliest predictions about a unidirectional continuum. There are, however, limitations in what a quantitative correlational attitude study can achieve. While the comparative approach presented above is one step towards a more dynamic and multidimensional view on language attitudes, the role of Creole as a social and cultural practice in various discursive contexts still offers a vast field of exploration.

References

Ajzen, Icek (1989): "Attitude structure and behaviour", in: Pratkanis, Anthony et al.(eds.), Attitude structure and function. Hillsdale, N.J.: Lawrence Erlbaum, 241–74.

Alleyne, Mervyn (1994): "Problems of standardization of Creole languages", in: Morgan, Marcylina (ed.), Language and the social construction of identity in Creole situations. Los Angeles: Center for Afro-American Studies Publications, 7–18. (= CAAS Special Publication Series, 10)

Beckford-Wassink, Alicia (1999): "Historic low prestige and seeds of change: Attitudes toward Jamaican Creole", in: Language in Society 28: 1, 57–92.

---

PhiN 15/2001: 77

Edwards, John R (1982): "Language attitudes and their implications among English speakers", in: Ryan Bouchard, Ellen & Howard Giles (eds.), Attitudes towards language variation. Social and applied contexts. London: Edward Arnold, 20–33.

Görlach, Manfred (1996): "And is it English?", in: English World Wide 17: 2, 153–174.

Haynes, Lilith M (1973): Language in Barbados and Guyana: attitudes, behaviours and comparisons. PhD Dissertation, Stanford University.

Kloss, Heinz (1967): "'Abstand languages' and 'Ausbau languages'", in: Anthropological Linguistics 9: 29–41.

LePage, Robert B. & Andrée Tabouret-Keller (1985): Acts of identity. Cambridge, etc.: Cambridge University Press.

Mühleisen, Susanne (1993): Attitudes towards Language Varieties in Trinidad. Unpublished MA-Thesis, FU Berlin.

Rickford, John A. (1983): "Standard and non-standard language attitudes in a creole continuum", in: Society for Caribbean linguistics, Occasional Paper No. 16, Trinidad.

Rickford, John A. (1985): "Symbol of powerlessness and degeneracy, or symbol of solidarity and truth?", in: Greenbaum, Sidney (ed.), English language today. Oxford & New York: Pergamon Press, 252–61.

Shields-Brodber, Kathryn (1997): "Requiem for English in an 'English-speaking' community: the case of Jamaica", in: Schneider, Edgar W. (ed.), Englishes around the world, 2: Caribbean, Africa, Asia, Australasia. (Studies in Honour of Manfred Görlach.) Amsterdam: Benjamins, 58–67.

St. Clair, Robert N. (1982): "From social history to language attitudes", in: Ryan Bouchard, Ellen & Howard Giles (eds.), Attitudes towards language variation. Social and applied contexts. London: Edward Arnold, 164–74.

Winford, Donald (1976): "Teacher attitudes toward language variation in a Creole community", in: International Journal of the Sociology of Language 8, 45–75.

---

PhiN 15/2001: 78

Notes

¹ A pelau is a West Indian dish of mixed ingredients.

² 'Dougla' is used as a derogatory term in Trinidad to refer to a person of mixed (African/Indian) heritage.

³ Kloss' (1967) terms Abstand (linguistic distance) and Ausbau (functional elaboration) are increasingly used in this context.

⁴ However, LePage's and Tabouret-Keller's Acts of Identity (1985) offers a rich source of case histories and qualitative research.

⁵ Figures taken from: Annual Statistical Digest 1990. Port of Spain, Trinidad: Ministry of the Economy, Central Statistical Office, 1992.

⁶ It was observed that many Trinidadians claim to recognise the geographical and/or ethnic background of a speaker by their speech, sometimes even the school they went to, when for instance speaking to a stranger on the phone.

⁷ Encompasses a whole range of leisurely activities with friends, e.g. chatting, going to a party, etc.