second language proficiency and acculturation : an ...

SECOND LANGUAGE PROFICIENCY AND ACCULTURATION : AN INVESTIGATION OF THE EFFECTS OF LANGUAGE STATUS AND INDIVIDUAL CHARACTERISTICS' Richard ClCment University of Ottawa

Abstract The present study investigates the relationship between language status and individual differences in attitudes and motivation as they relate to proficiency and acculturation in a second language. All respondents were Francophone university students with either a minority (n = 110) or a majority background (n = 183). They were requested to fill out a questionnaire including measures of ethnolinguistic vitality, attitudes and motivation as well as to participate in an interview aimed at assessing their oral proficiency in English asa second

language. Analyses of the data show that minority group members evidence more self-confidence in their ability to use the second language and greater proficiency in the second language than majority group members. Level of acculturation was a function of proficiency in the second language and an interactive function of language status and frequency of contact. Correlational analyses revealed that proficiency and acculturation were most strongly associated with self-confidence. Attitudes and motivation neither had an important influence on language outcome nor were they influenced by language status. These results are discussed for their theoretical implications and with referelice to second language education programmes.

E a r l y investigations of the determinants of second language acquisition have emphasised t h e role of psychological factors. Language aptitude, for example, has long b e e n recognised (Henmon, 1929) as a determinant of proficiency (for review, see G a r d n e r , 1985). Likewise, early social psychological approaches have emphasised t h e role of individual determinants: attitudes a n d motivation (see Gardner, 1960; G a r d n e r & Lambert, 1972). I n contrast with this emphasis on individual characteristics, much of the conceptualisation proposed as t h e basis for the empirical study a n d application of bilingual education revolves around socio-contextual considerations (cf. Bibeau, 1982; Edwards, 1979, 1981; Hamers & Blanc, 1983; Paulston, 1980) such as the majority o r minority status of the locutors. This last factor appears to be an i m p o r t a n t determinant of whether bilingual education will result in the addition of a new language a n d culture (i.e. additive bilingualism) o r the erosion and eventual loss of t h e first language a n d culture (Le. subtractive bilingualism). T h e present 0261-927X/86/04/027 1-20$02.50/0 JOURNAL OF LANGUAGE A N D SOCIAL PSYCHOLOGY

27 1 Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009

0 1987 R. Clement Vol. 5. No. 4. 1986

272

LANGUAGE AND SOCIAL PSYCHOLOGY

research is therefore designed to investigate the relationship between individual characteristics and socio-contextual factors as they relate to second language proficiency and acculturation. A number of social-psychological models of second language acquisition and proficiency have incorporated socio-contextual dimensions as explanatory constructs (e.g. Ball, Giles & Hewstone, 1984; Gardner, 1985; Giles & Byrne, 1982; Schuman, 1978a, b). ‘The present study was conducted within the scope of a model identified as ‘ClCment’s model’ by Giles & Byrne (1982). ‘Clhent’s model’ follows the empirical and theoretical path traced by Lambert (e.g. 1967,1974) and Gardner (1979; Gardner & Smythe, 1975) and incorporates, as well, elements of the intergroup theory formulated by Giles, Bourhis & Taylor (1977). ClCment’s model (1980, 1984; ClCment & Kruidenier, 1985) proposes that second language proficiency is influenced by the individual’s motivation. Motivation is determined by two processes, respectively influenced and activated by two aspects of the environment: (a) the relative ethnolinguistic vitalities of the first and second language groups, and (b) the frequency of contact with the second language group. T h e first process - the priniary inotivational process - includes two antagonistic aspects: iritegrativeriess which corresponds to the affective orientation towards the second language community and willingness to become similar to its valued members, and fear of assimilation which corresponds to the fear that joining the second language community might result in the loss of the first language and culture. Integrativeness and fear of assimilation are said to be determined by the relative ethnolinguistic vitality of the first and second language group. T h e concept of ethnolinguistic vitality was originally introduced by Giles et al. (1977) as a description of the language status of ethnic groups. It involved three aspects: demographic representation, socio-economic status and institutional support. Accordingly, small language groups (low demographic representation) whose members have low socio-economic standing and whose language and culture are not recognised by existing social and political agencies (low institutional support) would have low ethnolinguistic vitality. The converse would be true of groups with high ethnolinguistic vitality. Giles et al. (1977) also proposed that felt or subjective ethnolinguistic vitality might have more influence on language behaviour than actual vitality evaluated via objective indices. In Clkment’s ‘model, ethnolinguistic vitality is construed as subjective and relative. Individuals are classified according to their perception of the relative vitality of first and second language groups. Thus, their position on a continuum could vary from a perception of the total ascendancy of the firs: language group to 2 perception of the total ascendancy of the second language group. In the case where the first language group has relatively low ethnolinguistic vitality, members of that group should be attracted towards the second language group and demonstrate a high degree of integrativeness relative to members of the majority group. Also, because of their relatively weak position, members of minority groups should be less fearful of assimilation than members of majority groups. For minority group members, assimilation would give access to the goods and services Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009

SECOND LANGUAGE

273

controlled by members of the majority and should, therefore be desirable rather than feared. T h e primary motivational process is the most proximal determinant of motivation in unicultural settings. When contact with the second language group is possible, however, the tendency resulting from the interplay of integrativeness and fear of assimilation infuences the secorzdary rnotivationalprocess which then becomes the most immediate determinant of motivation. The latter process refers to the relationship between the quality and frequency of inter-ethnic contact and the individual’s self-confidence in the capacity to use the second language in an adaptative and efficient manner. Greater quality and frequency of contact are related to greater self-confidence. Within the context of the present study, it would be expected that majority members having less frequent contact with members of the minority group than the converse, would also evidence less self-confidence with their usage of the second language. Since self-confidence is directly related to motivation and proficiency, they would also evidence less motivation to learn the second language and less second language proficiency than minority group members. Finally, Clement’s model proposes that the level of acculturation into the second language culture should be a joint function of contextual characteristics and second language proficiency. Because language is a necessary medium of acculturation, little difference in level of acculturation should be found at lower levels of proficiency between majority and minority groups or high and low contact groups. Given relatively high proficiency in the secorld language, minority group members should, however, evidence more acculturation than majority group members. Also, the magnitude of this latter difference should be directly related to amount of contact between the two groups. T h e adequacy of the causal sequence proposed by ClCment has been generally confirmed by ClCnient & Kruidenier (1985). Contrary to their own predictions, however, they report results whereby both integrativeness and self-confidence evidence a direct relationship to motivation. The explanation they proposed was that the sample studied might not have been in a continuous and close contact with the second language group. As a result, motivation to learn and use the second language would have been influenced by both the primary process (as for individuals living in unicultural environments) and the secondary process (as hypothesised, for individuals living in close contact with the second language group). In the present study involving Canadian Francophones with English as a second language, contact with the second language group is obviously present, if not on a person-to-person basis, at least via the media and cultural artefacts imposed by the North American context. To the extent that majority and minority status are, in part, determined by the relative demographic representation of first and second language groups, being part of the majority group corresponds to the low availability of the second language group. Following Clement & Kruidenier’s (1985) conclusion, it would, therefore, be expected that majority and minority groups would evidence different motivational processes. Specifically, the primary process should be more related to motivation in the majority group than in the minority group. Conversely, the secondary process should evidence a stronger Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009

274


relationship to motivation in the minority group than in the majority group. T h e present research was, therefore, designed to test some implications of Clement’s model for the influence of contextual factors on both the strength of tendencies and the social-psychological processes related to second language proficiency and acculturation. In addition, this study features an attempt at assessing communicative competence in the second language via oral production. Previous studies of second language proficiency have oftcr? used teacher ratings, school marks or standardised test results as measures of proficiency. These measures are designed within the context of a particular curriculum that may promote objectives not directly relevant to communication situations of concern t o social psychologists. In this study, an effort is made to devise a situation fostering the naturalistic communication of the respondent with a second language peer about issues of immediate concern and familiarity to both. Furthermore, the assessment of proficiency was accomplished via a grid aimed at measuring oral production with the emphasis placed on communicative competence (cf. Canale & Swain, 1980; Wesche, 1981) and intelligibility rather than grammatical and syntactic perfection.

Method Subjects Respondents participating in this study were 293 University of Ottawa Francophone students (121 males; 172 females) aged between 18 and25 yearsof age. The University of Ottawa is Canada’s most important bilingual university, in terms of both its students population and the range of its programmes. It is situated in the nation’s capital, Ottawa, a city where Francophones and Anglophones represent, respectively, 19% and 67% of the population. The City of Ottawa is located in Ontario, a province where Anglophones and Francophones represent, respectively, 77% and 6% of the population. T h e principal employer in Ottawa is the Canadian federal government whose departments and agencies are officially bilingual as institutions. Likewise, the University of Ottawa charter stipulates its bilingual nature as well as its particular mission to serve the French fact in Ontario. Students registering at the University of Ottawa expect and must, in addition to course requirements, demonstrate a satisfactory level of bilingualism (French and English) in order to obtain their degree. Participants were selected on the basis of their answer to a short questionnaire concerning the town or village in which they had spent most of their life, their age, mother tongue and language used most often. The respondents whose mother tongue was not French and who reported not usually using French were eliminated from the sample. Information concerning locality was used to compute, for each individual, an index of the relative importance of the second language speaking group in their environment. Specifically, the ratio of the number of Anglophones (i.e. those declaring English as a mother tongue) to the number of Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009

SECOND LANGUAGE

275

Francophones (i.e. mother tongue French) as tabulated by Statistics Canada was computed for all respondents. A ratio smaller than one defined a niajoriry Francophone whereas a ratio larger than one defined a mitzoriry Francophone. In fact, the ratios obtained by majority group members (n = 183; 84 males, 99 females) were all below 1 with a median of 0.09. The ratios obtained for the minority group members (11 = 110; 37 males, 73 females) varied between 1 and 103 with a median of 3.38. Cross-tabulation of this demographic index with the province of origin of the respondents indicated that the majority-minority classification corresponded to a Quebec-Ontario classification in 92% of the cases. That is to say that those individuals who were classified in the majority group were also, for the vast majority, residents of QuCbec, a Canadian province with French as its only official language and the consequent political and institutional support for that language (see Bourhis, 1984). Those individuals who were classified in the minority group were mostly residents of Ontario, a Canadian province with English as its only official language. The majority-minority classification, therefore, corresponded to demographic and institutional support differences.

Materials Participating students answered an attitude and motivation questionnaire and underwent an English oral proficiency interview. The questionnaire included 22 scales which were adapted for this particular population, notably by eliminating references to second language courses as they are irrelevant to the present participants. The 22 scales were reduced to nine composite scores on the basis of previous factor analytic research (cf. Clement, 1978; ClCment & Kruidenier, 1985). Scales with different metrics o r variance were standardised before being added into the same composite index. The variables are listed according to the composite index in which they are included. Estimates of internal consistency reliability (Cronbach alpha), are provided for those scales c’ontaining more than o n e item. Description of most of the variables used can be found in Clement, Smythe & Gardner (1976), Clement & Kruidenier (1985) and Labrie & Clement (1986). Perceived ethnolinguistic vitality was assessed using the Subjective Ethnolinguistic Vitality Questionnaire (SEVQ) developed by Bourhis, Giles & Rosenthal (1981) and modified and translated for the French respondents of this study. Specifically, the labels ‘Greek’ and ‘English’ were replaced by ‘French’ and ‘English’, respectively; subjects were asked to answer the items with reference to the situation which prevails in their province of origin. Following Labrie & ClCment (1986), two composite indices, A and B hereafter were derived from answers to the SEVQ.

A . Perceived Vitality of French 1. Socio-Economic Status (alpha = 0.69) 2. Demographic Representation (alpha = 0.75) 3. Institutional Support (alpha = 0.80) Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009

276

LANGUAGE A N D SOCIAL PSYCHOLOGY

B. Perceived Vitality of English 4. Socio-Economic Status (alpha = 0.74) 5. Demographic Representation (alpha = 0.78) 6. Institutional Support (alpha = 0.86)

C. Integrativen ess 7 . Integrative Orientation (alpha = 0.73) 8. Attitude toward English Canadians (alpha = 0.80)

D. Fear of Assimilation 9. Threat of Assimilation (alpha = 0.78) 10. Fear of Learning English (alpha = 0.75) E. Motivation 11. Attitude Toward Learning English (alpha = 0.85) 12. Motivation to Use English (alpha = 0.86) 13. Desire to Learn English (alpha = 0.68)

F. Qiiality of Contact 14. Quality of Contact (alpha = 0.81) G. Self- Confidelice with Eriglish 15. Self-Confidence with English (alpha = 0.93) 16. Self-Rating: Writing English 17. Self-Rating: Understanding English 18. Self-Rating: Reading English 19. Self-Rating: Speaking English 20. English Use Anxiety (subtracted from other scales; alpha = 0.88)

H . Frequency of Contact 21. Frequency of Contact I . Accrtltiiratiott 22. This 10 item scale was designed specifically for this study. Each item presents the respondent with a communication situation (e.g. reading a newspaper, watching television, talking with a friend) to which helshe must react by choosing one of three answers corresponding to his/her actual language usage: mostly in French, eqitally in Freiwh and Etiglislr and tnostly iti Etiglish. T h e total score varies between 10 and 30, a high score indicating relative assimilation into the English culture, an intermediate score, integration of both languages and cultures and a low score, relative uniculturalism (alpha = 0.82).

J . Oral production T h e respondent's oral proficiency was assessed via an interview led by one of two trained female Anglophone students of the same level and age as the Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009

SECOND LANGUAGE

277

majority of respondents. The interview itself was taped with the permission of the participant and was conducted entirely in English and included three tasks. During the first task, the respondent was requested to communicate the time schedule of the library to a make-believe class. In the second task, the respondent described to the interviewer the situation portrayed in a wordless comic strip containing a sequence of four images. In the third task, the subject was presented with the planned and actual budget of a student. He/she was then asked to comment and offer recommendations while adding, for example, details concerning his/her budget. For each task, the interviewer had to evaluate on five point scales five aspects pertaining to the oral communicative competence of the respondent:

1. Precision. This aspect refers to the extent to which the participant’s mastery of linguistic structures is acceptable to the average Anglophone. 2. Range. This aspect refers to the extent to which the participant mastersa vast enough choice of idioms and structures to produce a varied and interesting discourse. 3 . Size. This aspect refers to the extent to which the subject is able to produce a lengthy and structured discourse as opposed to a series of incoherent sentences. 4. Ittdependertce. This aspect refers to the extent to which the respondent is able to maintain a conversation with the interviewer without constantly needing explanations or clarifications. 5. Pruniirzcintion. This aspect refers to the level of French accentuation of the respondent’s speech. A low evaluation would indicate a heavily accented speech, difficult to understand for Anglophones. For these five aspects, the inter-rater reliability computed on a sub-sample representing a random selection of 30% of the respondents from 0.72 to 0.89 with a median of 0.87. T h e inter-correlations between all of the ranged sub-scales being well above 0.80, it was decided to add the sub-scale scores, thus creating a single Oral Production index.

K . Pruficiertcy Students participating in this study were required to write a general English proficiency test in the context of the bilingualism requirements of the University of Ottawa. The general proficiency test includes a cloze test and tests of reading and listening comprehension. This score corresponds to what appears in the student’s file kept at the University of Ottawa Modern Language Centre. In this case, sub-tests are aggregated without standardisation. Procedure T h e short preliminary questionnaire used in the selection of respondents was filled out during the general proficiency test. On the basis of their answer and written permission to be contacted for this study, respondents were contacted by Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009

278


telephone and requested to participate, for a fee, in the two phases of this study. During the first phase, the respondents answered the attitude and motivation questionnaire in groups of varying size. This phase took approximately 45 minutes. Prior to the session, instructions on the cover page of the questionnaire were read aloud to the students, ensuring them of the confidentiality of their answers and of their right to withdraw from the study should they find any question to be objectionable. A t the end of the session, they were requested to make an appointment with the interviewer for phase two of the study. T h e interview (phase two) took place in a private office, furnished with a desk, two chairs and a tape recorder installed in plain sight. The participant was greeted by the interviewer upon his/her arrival. He/she was invited into the office and subsequently assured that the performance o n the test would not affect their academic standing. The respondent was then asked permission to record the interview and the interviewer proceeded with the description of the first task. The interview was completed within 15 to 20 minutes. At the end of the interview, the participant was given $5 for participating in the study.

Results T h e present study was designed to assess the effects of contextual factors on the social psychological tendencies and processes relating to second language proficiency and acculturation. Two series of analyses were undertaken and are reported separately hereafter. The first series tests the statistical significance of the differences between the means of the majority and minority groups on the variables identified in the Method section. The second series involved correlational analyses and sought to identify differences between the two groups in terms of the pattern of relationship between the dependent variables.

Comparison of means T h e analysis of differences between means involved three distinct statistical models pertaining respectively to: (a) Ethnolinguistic Vitality scales, (b) the Acculturation scale, and (c) the other Attitude, Motivation and Self-confidence scales.

Etliriolirigiristic Vitality scales A two-way analysis of variance using as factors Status (majority vs. minority) and Target Group (Fraucophones vs. Anglophones) as repeated measures was computed on each of the subscales of the SEVQ? Results of the analysis of the Socio-economic Status subscale show that only the interaction between Status and Target Group was statistically significant [F(1,285)= 52.07, p

8 32 -1

(26.16)

(27.93)

26 W

-1

Anglophones Francophones TARGET GROUP

(B) Demographic Representation 44

>-

r2 30 t > u- 32 0

-I

$!

26

W J

1 (29.04)

(21.18)

20

(30.A1

( 24.23 1

Anglophones Francophones TARGET GROUP Downloaded from http://jls.sagepub.com at UNIV OF OTTAWA LIBRARY on July 21, 2009


280

(C)

Institutional Support

2 38 t-

3 LL

0

x

(39.93)

>!=

32 .

MAJORITY m MINORITY

(40.34)

(32.30)

( 33.59)

I

1

Anglophones

Francophones

TARGET GROUP than the Francophone target group ( M =.28.26) by both status groups [F(1,287) = 10.26, p