The Status and Role of Regional Languages in Higher Education in Pakistan

The Status and Role of Regional Languages in Higher Education in Pakistan Sabiha Mansoor Available online: 29 Mar 2010

To cite this article: Sabiha Mansoor (2004): The Status and Role of Regional Languages in Higher Education in Pakistan, Journal of Multilingual and Multicultural Development, 25:4, 333-353 To link to this article: http://dx.doi.org/10.1080/01434630408666536

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-andconditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

The Status and Role of Regional Languages in Higher Education in Pakistan Sabiha Mansoor Centre of English Language, Aga Khan University, Karachi, Pakistan Pakistan, as a multilingual country, faces numerous problems in language planning in higher education. There are concerns about the limited role of regional languages, lack of required materials in Urdu, and student difficulties in English. The research reported here is a nationwide survey of 2136 students, 121 Subject and English teachers of public and private sector colleges and universities from all capital cities of Pakistan, as well as 63 parents who responded to the questionnaire

. The survey examines the students’ background, their competency and use of mother tongue/regional languages, attitudes to languages, the availability and quality of materials, the role of regional languages in education, as well as language and sociocultural outcomes. Results reveal a language shift in the regional speakers who display low competency and use of their mother tongue/regional languages in formal and informal domains. They also display negative attitudes to their own languages as seen in their preference to study in English and Urdu medium at all levels of schooling. The study recommends a language policy in education that promotes cultural pluralism and also provides state support to the minority languages in Pakistan. Keywords: minority languages, language shift, assimilation policy, case study, ethnolinguistic vitality, subtractive bilingualism

Introduction
A brief analysis of the present language situation and a historical perspective indicates that regional languages have not been provided state support in education at all levels, particularly in higher education, in Pakistan. The issue regarding the status and role of regional languages has not been adequately addressed by the various education commissions set up by different governments to look into the problems being faced by students and teachers in higher education. The Report of the Education Sector Reforms (2001) and the Task Force on Higher Education (2002) set up by General Musharraf have also not addressed the issue of language policy in higher education.

The official policy with regards to language has been to maintain English as the medium of instruction in Higher Education after the country’s independence in 1947, as seen in all educational policies and reports of education commissions and committees set up in this regard (1957Á/1998). This policy is seen as an interim arrangement. The long-term language policy has been throughout to introduce Urdu as the official medium of instruction in higher education once teaching materials have been developed in the national language. Although Urdu was declared the official medium of instruction for schooling (classes 1Á/12) in the public sector soon after the country’s independence, the period assigned to the transfer from English medium to Urdu medium in higher education has varied in various reports, that is 15 years in the 1950s (Sharif Commission, 1959) and five to seven years in the 1970s (University Grants Commission, 1982).

Higher Education in Pakistan
The literacy rate in Pakistan was estimated at 47.1% with 59% males and 35.4% females (Economic Survey: 1999Á/2000). In higher education (which includes undergraduate and postgraduate studies), the participation rate is a mere 3% as compared to the 50% participation rate in the advanced countries (PST, 1999). The present government of General Musharraf is also attaching a great deal of importance to higher education. Higher education has been allocated 24% of all funds dedicated to the education sector in the 10-year period (Planning Commission Ten Year Perspective Development Plan, 2001Á/2011).

Language Situation
Pakistan fits the classic concept of a culturally plural society on many grounds: ethnicity, language and culture.1 Punjabi- and Seraiki-speaking communities make up 54.68% of the population (Punjabi 44.15%, Seraiki 10.53%); the Sindhi-speaking community (15.42%), Balochi-speaking community (3.57%) and Pushto-speaking community (15.42%) correspond to the four provinces of the country Á/ Punjab, Sindh, Balochistan and North West Frontier Province (1998 Census Report of Pakistan). Each group has its own language and cultural differences, often with marked subdivisions within each group (Hayes, 1987). No less than 24 languages and a number of dialects are spoken in Pakistan. Punjab has Punjabi and Saraiki.2 Sindh has Sindhi in rural Sindh, Urdu in urban Sindh and Gujrati among influential minorities.

In the North West Frontier Province (NWFP), Pushto is the language of the majority of the population, though one district, Hazara, uses Hindko. Despite having the smallest population, Balochistan has multiple languages; it has Balochi, Pushto, Brohi and a sprinkling of Seraiki and Punjabi (Haque, 1983: 13 Á/ 18). The national language Urdu is the mother tongue of only 7.57% of Pakistanis. The 1998 Census Report states the population of Pakistan as 145.5 million. Thus, in actual terms, even relatively small linguistic groups are fairly large as compared to the linguistic groups in less populated countries. Numerical strength does not determine majority or dominant languages in Pakistan. The regional languages, despite a large number of speakers, are minority languages. ‘Urdu’ and ‘English’ enjoy ‘high status’ and are reserved for public and official use and also dominate the regional languages Á/ politically, economically, and culturally. Press and media also mainly use the dominant languages Urdu and English. They are also used as alternate medium of instruction in education, and are compulsory subjects in the school and college curriculum.

On the other hand, the regional languages are accorded a low status and are limited to community and home. They play little or no role in the official life of the provinces and their educational role is limited to the primary or secondary level in most provinces. The state’s declared policy has been to promote Urdu as a national language. It is a symbol of national identity and integration to help avoid regional autonomy and separation. This has been the official policy since Pakistan came into existence in 1947. Although the official stance has not been pro-English, English still continues to flourish as the second official language of Pakistan. This fact prompts political analysts to infer that the support to Urdu is only at a rhetorical level and has led critics to state that there appears to be an inherent contradiction between Pakistan’s stated official policy of promoting Urdu as the national official language of Pakistan, and the real policy where the status quo is maintained and English remains for all purposes the language of power (Rahman, 1996, 2002).

The ruling elites, that is, all those with power and influence in Pakistan, like the bureaucracy and military, have command over English through their English medium schooling and training. English is not only the language of the upper classes in Pakistan but fluency in English also facilitates access to the best jobs in the governmental, nongovernmental and international bureaucracy. With regards to the status and role of regional languages, the provisions of 1956, 1962, 1973 Constitutions state (see Khan, 1986): ‘Without prejudice to the status of the National language, a Provincial Assembly may by law prescribe measures for the teaching, promotion and use of a provincial language in addition to the national language.’ The position of the regional languages as seen in article 251 (clause 3), is conditional to the status of the national language. There is a constitutional provision to promote the regional language even in teaching by the Provincial Assembly. However, it has been generally observed that only lip service has been made in this regard. Indeed, it is implied that no measure should be adopted to promote the regional language at the cost of the status of the national language Urdu. In all provinces except Sindh, hardly any legislation has been made to promote the regional languages in the official spheres including education (Abbass, 1998; Rahman, 1999).

Language Spread versus Language Shift
According to Paulston (1980), it is unusual to maintain group bilingualism. The norm for groups, in prolonged contact within one nation, is for the minority group to shift to the language of the dominant group either over several hundred years, or in the span of three generations. Fishman (1972) defines language shift as a change in the language usage patterns of a population that employs more than one language variety. While language shift is always the outcome of language spread, the latter can also occur without language shift i.e. language spread can take place without threatening or replacing the other. According to Lewis (1981), language shift does not mean that knowledge of the forsaken language has disappeared but that the relative competence in the two languages has switched. Cooper (1983) suggests that language spread takes place as lingua franca or as language for wider communication. On the whole such spread is neutral in attitudes when they do so as an additional language but as a new mother tongue the language spread becomes a case of language shift.

Maintenance of the Mother Tongue
Hudson (1980) suggests that assimilationist policies create ‘non-stable’ or ‘subtractive’ bilingual communities. International concern about local language education is evident in the UNESCO report (1953: 6) and the work of Skutnabb-Kangas (2000). Provision for mother-tongue teaching and the use of the mother tongue as a medium in the mainstream curriculum is justified in the UNESCO report (1953). There are also indicators that the maintenance of mother tongue and culture may facilitate the learning of a second language (Skutnabb-Kangas & Toukoumaa, 1976).

The planning process is costly and time consuming but it can be accomplished on the commitment made to the effort in a span of a few years (Fishman, 1972). The vernacular language should be a basis of minimal literacy (Ferguson, 1978) and not used transitionally as a bridge to the national language (as in the case of the USA and the Soviet Union, as per Lewis, 1981). Vernacular education is most appropriate for countries like Pakistan where the average number of years of schooling completed is low and where social and economic pressures ‘mark’ the language (Tucker, 1977: 40). As discussed earlier, the middle- and upperclasses make more use of Urdu and English in Pakistan as compared to the rural classes that mainly use the regional languages.

Mother-tongue Education
According to Pattanayak in Spolsky (1986), the use of colonial languages in Third World education as a substitute for many mother tongues, has created many problems. These include a chasm between the elite and the masses, stunted cognitive growth, and lack of creativity and innovativeness in children as well as a weakening of indigenous cultures. The colonial languages have also led to the slow development of science and technology to meet local needs. In addition they have created an educated elite that is westernised. Tollefson (1991) examines the controversy over using languages other than the dominant language for school education. He argues that in almost all countries, schools adopt one or more languages of instruction.

Students who do not speak the language of instruction are at a disadvantage along with those who speak nonstandard varieties of the dominant language or different languages. Hence the issue of whether or not education plays an important role in employment and in gaining access to political power, mother tongue education plays a significant role in language policy and language education. Skutnabb-Kangas (2000) emphasises that since further development of a mother tongue, especially in more formal domains, takes place within a school system, it becomes important to see what kind of educational language related human rights are provided in various bilingual education systems. Hornberger (1996) suggests the enrichment model that encompasses all bilingual education programmes. It encourages the development of minority languages on the individual and collective levels, cultural pluralism at school and in the community, and an integrated national society based on the autonomy of cultural groups (Hornberger, 1991: 222; in Freeman, 1998).

The Study
The study was designed to focus on some of the key areas in language planning and language education. Both approaches, namely quantitative and qualitative, were considered to examine the suitability for this research study. It was observed that the combined approach was more helpful as suggested by Robson (1993). The main purpose of the study was not only to find out how many number and percentages of respondents held the same or different attitudes, values, opinions and beliefs, but also the reasons for this. In this study Language Education in Pakistan was the CASE and the methods of data collection were: (1) questionnaire survey, (2) interviews, and (3) documentary analysis. The present research was designed to be a ‘real world enquiry’, mainly concerned with contributing to both language policy and practice in higher education in Pakistan. Our concern in this article is with the first source.

The questionnaire The ‘sociolinguistic survey’, suggested by Kaplan and Baldauf (1997: 110) was seen as a good model to adopt in this regard. Four types of data-collecting instruments were developed for conducting the survey. These included: (a) three-structured survey questionnaires divided on the basis of the following populations: students, parents, and teachers; and (b) one English language test to gauge the English language proficiency of students and teachers. The research questions in our study revolved around collecting information on (1) the background of learners (demographic, educational, and language), (2) attitudes to languages in education, (3) availability and quality of materials in different mediums, and (4) learners’ language and sociocultural outcomes.

The sample size of the testing of hypothesis for the prevalence of students regarding their opinion about the role of English in higher education in our population was calculated by taking a proportion of 50% with level of significance (8/) of 5%, a power of 80%, the bound on error (the absolute difference between actual and hypothetical prevalence) of 4% and design effect of 2 (because of two-stage cluster sampling). The minimum sample size of 2450 students was required for this study. The sample of the colleges and universities from Private and Public (Government) as well as General and Professional was taken from the list of degree colleges and universities as listed by the Handbook: Colleges of Pakistan (University Grants Commission, 1999). A decision was taken by the researcher to select only the colleges and universities from capital cities of Pakistan and the federal capital Islamabad.

The staff As the data was to be collected on a large scale (nationwide study), a team comprising two faculty members who assisted with the data collection, a statistical advisor of the university, as well as an editor and two data entry operators was put together. As all the team members were already trained and had been involved in previous research projects, no formal training was required. The research project had the support of the Aga Khan University Seed Money Grant Award, and hence all expenses related to the study were borne by the Award.

Results
The questionnaire responses were coded and entered into the Statistical Package for the Social Sciences (SPSS, Version 6.1.3) and analysed by the Statistical Advisor of the Aga Khan University and the author of this article. The analysis of all quantitative data obtained from the questionnaires was done at a descriptive and inferential level (see Kinnear & Gray, 1999). This article only deals with the results related to the regional languages in education, as revealed by the questionnaire. Background characteristics The sample comprising 2136 students was predominantly from Punjab (Lahore and Taxila) and Sindh (Karachi and Hyderabad), Balochistan (Quetta), followed by the Federal Capital (Islamabad), and the NWFP (Peshawar). The study was able to collect data from 121 subjects and English teachers from all provinces except Balochistan. Only 63 parents’ questionnaires were received from Punjab and Sindh. The profile of the students by different characteristics of institution shows that the majority of them were from colleges (84%) and a smaller number from universities (16%). Also, around two thirds (66%) of all students in the study were from the public sector (colleges and universities).

The majority of the students reported Urdu as their mother tongue (42%), followed by Punjabi (30%), Pushto (14%), Balochi (5%), Sindhi (4%) and others (5%). The results of the study show that the students involved in higher education in Pakistan belong to various socioeconomic backgrounds (income groups, divided into quintiles). Students studying in the private sector have significantly higher monthly household income as compared to students studying in public sector (see Appendix 1). Proficiency and usage in regional languages/mother tongue Many students, especially from Punjab and Sindh, report in the study that they are not fully competent in their mother tongue, both in speaking and writing as seen in Appendix 2. Only 45% Punjabi male and 16% female students report full spoken competency in their mother tongue Punjabi. Again, only 28% Sindhi male and 8% female students report full spoken competency in their mother tongue Sindhi. Very few male and female students (4% Á/20%) from all regions report full written competency in their regional languages Punjabi, Sindhi, Balochi and Pushto.

On the other hand, a large majority of Urdu-speaking students (57% Á/64%), both male and female report full spoken and written competency in their mother tongue Urdu (see Appendix 2). In terms of competency and use of mother tongue, Urdu and English, male students claimed more competencies in mother tongue as compared to female students (see Appendix 2). A significant difference in males and females is found in the competency of speaking regional languages. The result shows that males are more competent than females in speaking regional languages, while no significant difference in males and females is seen in the competency of writing regional languages. It is also seen that regional speakers do not make full use of their mother tongue or regional languages. Less than one fifth (20%) report using their regional language with their families and friends.

Very few students report using their regional language in the formal domain especially with teachers (3%) and fellow students (11%), as well as with office/ bank employees (5%), and while visiting another province (8%) as seen in Appendix 3. Male students also reported a higher use of the mother tongue with family, friends, teachers and office and bank employees than females who reported more use of Urdu in these domains (see Appendix 3). The higher competency and use of mother tongue by male students as compared to female students support the view that females prefer the use of ‘high status’ languages (Urdu) to their ‘mother tongue’ or regional language, as they are more conscious of social status (see Trudgill, 1983). The role of regional languages in education The study revealed that the majority of students from the private sector have studied in English medium of instruction, whereas the majority of students from the public sector have studied in Urdu medium of instruction at different levels of schooling.

As seen in Appendix 4, it is also seen that very few students from both public and private sector have studied in the regional language at various levels of schooling, including primary (8%), middle (5%), secondary (3%), intermediate (1%) and graduate (1%) levels. As seen in Appendix 4, the distribution of students having regional language as a medium of instruction is very low as compared to Urdu and English in both public and private sectors. A decreasing trend of regional and Urdu as mediums of instruction is seen with the increase in the stage of schooling, while there is a reverse trend in case of English in both public and private sectors. A very high significant difference was found in Urdu and English as medium of instruction at different stages of schooling between public and private sectors. This trend of difference also shows that the majority of the Urdu medium students belong to the public sector educational institutions, whereas the majority of the English medium students belong to the private sector educational institutions. At graduate level, the result shows a significant difference between public and private institutions regarding their medium of instruction as reported by the students.

The majority of private institution students reported English medium while the majority of public institution students reported Urdu as their medium of instruction (see Appendix 4). About one fifth of private as well as public institutions students reported both (English and Urdu combined) as their medium of instruction. There were hardly any students who claimed that they were studying in their regional languages including Sindhi. Availability and quality of materials The regional languages are also seen as deficient for educational purposes because of lack of materials, particularly in science and technology. As seen in Appendix 5, both students and teachers report that very little materials are available to them in regional languages for their courses (4% Á/5%) in higher education; they rated these materials as low in quality (5% Á/ 11%). The results reveal that there are very few textbooks or reference books in regional languages at higher levels for Arts or Sciences. Attitudes to English, Urdu and regional languages/mother tongue As seen in Appendix 6, conflicting attitudes ranging from ambivalent attitudes to negative attitudes are displayed by students to their regional language and mother tongue.

The students, parents and teachers generally do not approve of their languages being used as medium of instruction (11% Á/20%) and there is no commitment to practice. It is also considered economically unimportant. Almost no student recommended the regional languages as a compulsory subject (0%), and very few as a medium of instruction (11%), and that was mainly for primary schooling. Again, in terms of attitudes to languages, the results in our study show that all students (male and female) prefer English medium education to Urdu medium at all levels and display a low preference for regional languages as a medium of instruction. However, male students and teachers show a higher preference than females for regional languages as a medium of instruction at primary levels. The results of the study confirm an earlier finding on the low ethnolinguistic vitality of Punjabi students, especially girls, in which Punjabi speakers hold their own language Punjabi, in low esteem (Mansoor, 1993).

As seen in the results of the study, the main reasons for preference of English as medium of education at various levels by students, teachers and parents are mainly instrumental. Studying in English is seen as useful for students in studying abroad (78%) and getting good jobs (63%). The reasons given by all subjects for studying in Urdu are mainly integrative. Students see Urdu as the main link language (67%) and also for its role in promoting national language (67%). The main reasons for studying in regional language are related to learning and it is felt by the students that study in the regional language is helpful in the first few years of schooling (60%), and also because students can learn better in their mother tongue (58%).

As seen in the results, the majority of all students, teachers and parents (59% Á/97%) show a preference to study English as a compulsory subject at all levels followed by Urdu (8%Á/50%). The preference for Urdu as a compulsory subject gets lower in higher levels. Very few subjects show a preference for study of regional language as a compulsory subject even at primary levels (0%Á/22%). The study also revealed that both males (50% Á/96%) and females (72% Á/ 100%) show a strong preference to study English as a compulsory subject at all levels. This is followed by a preference to study Urdu as a compulsory subject but this preference keeps getting lower as the levels get higher. However, more male students and teachers show a preference for learning a regional language at primary (17% Á/28%) and middle levels (8% Á/13%) than females.

Language outcomes The proficiency in regional languages was also found to be significantly different between students of private and public institutions for higher studies and for future employment. Lower proficiency of private sector students in regional language for the mentioned areas (see Appendix 7) was reported. The self-reports of students, teachers and parents also show that the students’ present language proficiency is inadequate to fully meet their English language needs for higher studies (39%), future employment (47%) and social interaction (30%). Again, the students’ language proficiency in Urdu is seen to be inadequate to fully meet their needs for higher studies (47%) and employment (48%).

Proficiency in Urdu is, however, considerably higher to meet their informal needs such as social interaction (57%). In regional languages the situation is more serious with only 18% students reporting a high proficiency to meet their language needs in higher education, 18% for future employment and 23% for social interaction. Socio-cultural outcomes As seen in the results reported in Appendix 8, the urban Punjabi-speaking students display a higher competency in Urdu (75%), as compared to their mother tongue Punjabi (47%). Also, they report higher use of Urdu (66%) than Punjabi (44%), thus displaying a language shift from their mother tongue Punjabi to Urdu, a high status language. The results also show that the urban Sindhi-speaking students have almost similar competency in mother tongue (51%) as in Urdu (49%), but with far more Sindhi students using Urdu with their families (77%) than their mother tongue Sindhi (25%), thus displaying a trend towards a language shift to Urdu.

Consequences of Language Attitudes
The main trends that are seen as a result of the research and its relationship to bilingualism and society can be studied in their wider perspective, along with their possible consequences on language maintenance and language shift. The findings also make evident the outcomes of the prevailing attitudes to languages in education in terms of ‘additive’ and ‘subtractive’ bilinguals. The results of the present study seem to indicate a language spread in the case of Urdu and English. A large number of students are proficient in Urdu and use it dominantly in all spheres, informal and formal. The result of the study makes evident a ‘language shift’ in many urban educated regional speakers, especially the Punjabi and Sindhi students where Urdu has replaced their mother tongue in all spheres of use, both formal and informal. The Punjabi students display a leaning towards subtractive bilingualism on the additiveÁ/subtractive continuum. In learning a highly valorised second language (Urdu), the mother tongue or first language (Punjabi), a devalorised language, is weakened. In many cases Urdu has replaced Punjabi as the first language and even where Punjabi is the first language, dominant use is made of Urdu in not only formal but in informal spheres as well.

In the case of Punjabi girls, this phenomenon is more accentuated. The Punjabi students display negative attitudes to their mother tongue (Punjabi) and a low image of self and language community. They also show little or no preference to study in their mother tongue, Punjabi, as a subject or a medium of instruction. Also, as Punjabi is not a medium of instruction or a compulsory subject in schools, except in some cases at the primary level, cognitive skills are not highly developed in the first language leading to subtractive bilingualism. However, due to soft boundaries between Punjabi and Urdu in which there are varying degrees of mutual intelligibility, devalorisation of the Punjabi language and negative attitudes would seem to be more crucial in resulting subtractive bilingualism. The study also shows a language spread in English due to the highly positive attitudes to English as an international language and Urdu as the national language, displayed by students and teachers, and other factors such as mass media, information technology, commerce etc. An earlier study on attitudes and motivation of students and teachers had indicated the same trend (Mansoor, 1993). The students show a strong desire to study English as a medium of instruction and as a compulsory subject mainly for instrumental reasons.

The students also make use of English in both informal and formal domains despite their limited proficiency in the language. The study reinforces the views of Lambert (1980) on types of attitudes that exist within and between language groups where one language is dominant politically, economically and culturally, and the other is without power and prestige. The negative attitudes to the devalorised language are so amplified by the majority group that members of the minority group downgrade themselves as well. In a country like Pakistan where sentiments of nationalism are very strong, bilinguals often reflect negative attitudes towards the minority language group.

This subtractive form of bilingualism results because the dominant language group (Urdu) is putting pressure on, for example the minority language (Punjabi), to assimilate as quickly as possible. Highly positive attitudes to English have led to its spread despite effort by the government in language and educational policies, especially during Zia’s rule in 1978, to oust English and replace it with Urdu. As seen from the results of the study, it appears that Punjabi students are experiencing negative ethnic identity. They display negative attitudes to their own language (Punjabi) which affords them only unfavourable intergroup comparisons with other language communities Á/ Urdu-speaking and English-speaking communities that enjoy a high status. According to Stubbs (1985), ‘In the long run negative attitudes can lead to language shift’. An important factor in the rapid assimilation of Punjabi by the dominant language (Urdu) is the low prestige and status of the language. Since Punjabi enjoys a low status, its own language community displays negative attitude to the language.

An important factor in language spread or language shift is the relative economic, political and linguistic prestige of the language. The role of language planning is crucial in determining the status and prestige of a language. The choice of Urdu as the national language, where Urdu is seen as a symbol of national identity and national integration, has given a great boost to Urdu. In the case of the spread of English, it is due to the high status it enjoys. It has been the second official language, and is used along with Urdu for all official purposes. English is also a medium of education in the country and a compulsory subject. A major incentive to learn a language is the income. In Brudner’s terms (1972), jobs select language-learning strategies that are to say wherever there are jobs available, people will learn the languages required to access them. In Pakistan, the most lucrative jobs require proficiency in English.

English is also seen as very useful for higher education as all materials are in English. On the other hand, the low ethnolinguistic vitality in the case of Punjabi speakers could be explained by the inferior and relative state of deprivation of the Punjabi language. Punjabi as a regional language has been completely ignored by official policy makers and has been assigned no role in education. No encouragement is given to Punjabi in language or educational planning. Hence, Punjabi, having no official status or role in the province, remains neglected. Apart from Punjabi poetry and literature, little development has taken place in educational materials in Punjabi. Therefore Punjabi is normally regarded as the language associated with illiterate and the poor especially the rural workers. It has been argued that literacy is often an important factor in language maintenance and transmission from one generation to another. The role of written variety and standardisation of language has been emphasised by Kloss (1989).

As Punjabi is not used for purposes of literacy, its status is greatly weakened. This may be strengthened by the covariation principle (Kelley, 1969). Punjabi children from lower status backgrounds perceive those who use Punjabi as occupying a relatively low position and those using Urdu or English as occupying a higher one. Hence, language is seen as the cause of differential status. This may be strengthened by in-group members who are now occupying positions of higher status. It therefore appears that the upper classes and elites are moving away from Punjabi. Urdu and English are seen as languages of a future urbanised society. English is particularly seen as modern and progressive, and attracts the Punjabi speakers, including rural Punjabis, who have moved into urban areas thus displaying a language shift to dominant languages such as Urdu. Results of this study also make explicit the attitudes of parents and teachers. Both parents and teachers display highly positive attitudes towards English.

Working class parents see school as an avenue for the upward mobility of the children. The choice of schooling, like preference of parents to send their children to English medium or Urdu medium schools encourages children to speak the respective language. The phenomenon of parents helping their children to learn the ‘correct’ language (in this study, Urdu and English and not Punjabi) so as not to be stigmatised later in life and to advance socially and materially is widespread. This is the main reason for Punjabi parents not transmitting their language to the next generation. It is seen that the current language(s) for teacher Á/student talk in colleges is English and Urdu. It seems that Punjabi is rarely spoken between college teachers and students although it is their mother tongue.

The results also show a higher use of Punjabi by parents, many of whom have not transmitted their mother tongue to their children. The results also reveal a trend in Sindhi-speaking students to move away from their mother tongue Sindhi. This is evident in them making more use of Urdu than Sindhi not only in the formal domain but also in the informal domain with family and friends. These results are indicative of the lack of attention being paid to Sindhi in terms of improving its status in the official sphere as well as in developing materials for educational purposes. This could gradually lead to a language shift as in the case of Punjabi students, if steps are not taken by the state to address this issue.

State Policies: Assimilation versus Integration
The language shift in many urban educated regional speakers may be seen as a result of the assimilation policy being followed by the state, where Urdu as the dominant language is replacing other minority languages, rather than an integration policy which aims at cultural pluralism, where all languages coexist side by side, without feeling endangered of becoming extinct (see Skutnabb-Kangas, 2000). Pakistan, since its independence in 1947, has been following a policy of giving a tremendous boost to Urdu as the national language. Urdu enjoys a high status and is used along with English for all official purposes.

Urdu followed by English has maximum participation in social institutions: schools and mass media. This helps in the spread of the language. English also enjoys a great deal of institutional support in Pakistan including access to trade and commerce. One of the important factors responsible for rapid assimilation of Punjabi in urban areas is lack of institutional support. The results also indicate that no serious attention or funds are being allocated to develop materials in the local languages except for Sindhi. The results of the study reveal that availability of materials in Sindhi is also very low, keeping in view that Sindh is the only province where the regional language is being used for higher levels of schooling.