Internal migration and changes in language repertoire among Sindhi youth

Full Text

1. Introduction

This study focuses on everyday discourse of young and older Sindhi speakers in the internal diaspora and reasons for possible language shift. David (2000, 2001a & 2001b) has reported how Sindhis in the external diaspora have shifted to dominant languages. However, as reported in Abbasi et al. (2020a) diverse language patterns are used by younger Sindhi generation in comparison to older Sindhi generation. Therefore, focussing on both younger and older Sindhis in the city of Karachi, this study seeks to analyse language patterns of both age groups and explore reasons for these language patterns in in the internal diaspora.

Sindh is the historical home of the Sindhis. The Sindhis are a socio-ethnic group originating from Sindh which today is a province in Pakistan. In the wake of the partition of the Indian sub-continent in 1947 and even prior to the partition, Sindhi-speaking Hindus had made their home in India and in numerous other countries. Studies conducted on Sindhi Hindus in Malaysia, Singapore, London, (David 2000, 2001a & 2001b) and Hong Kong (Detramani & Lock 2003) identified a shift away from the habitual use of the heritage language in the Sindhi diaspora.

In contrast, a study on the Sindhi Muslim community in the homeland of Sindh, Pakistan reveals high ethnolinguistic vitality (David et al. 2017). Today many young members of the Sindhi community are moving to cities in Pakistan where Urdu and English are quite popular among the masses, as Urdu which is the national language is used as a Lingua Franca. Rahman (2000: 41–42) states that “Urdu is a symbol of national identity while English is the source of employment”. In this study, we examine the language choices of Sindhi youth who have moved away from their hometown and are working and studying in the city of Karachi.

As these Sindhi speakers are exposed to three languages, Sindhi (in the home domain); Urdu (in social and educational domains) and English (in the educational domain); they (Sindhis) code-switch from one language to another and often  code-mix words from Urdu and English in Sindhi (Abbasi & David 2020). Over the  years, analysis of the discourse on language patterns has been based on code-switching (CS) and code-mixing of two languages (Mugheri et al. 2021). However, the younger Sindhi generation, due to the exposure to three languages, is  moving among three languages. Such an intermixture can be termed translanguaging (Canagarajah 2011, Garcia & Wei 2014, Slembrouch 2018, Simpson 2017, Wei 2018).

Codeswitching occurs in bilingual or multilingual settings. In urban areas, speakers often acquire languages to meet different community needs. Therefore, such speakers can use multiple languages to convey and share their thoughts and emotions with each other. There are several terminologies used to describe language change patterns of bilingual/multilingual speakers in a speech community. According to Myers-Scotton (2006), code switching is a language pattern that involves the use of two language varieties in the same conversation. Muysken (2000) used the term code-mixing to refer to intrasentential CS. In code-mixing, a sentence structure consists of lexical insertion, code alternation and congruent lexicalization. ‘Insertion’ indicates single lexical items being inserted into the frame provided by the matrix language (ML). Alternation implies change of both the ML and lexicon within continuous utterances by the same speaker; while congruent lexicalization means that the “the grammatical structure is shared by languages A and B, and words from both languages are inserted more or less randomly” (Muysken 2000: 8).

The Matrix Language-Frame Model (MLF) of Myers-Scotton (1993) emphasized the notions of matrix and embedded language. The ‘matrix language’ is the dominant one in supplying the morphosyntatic frame of bilingual utterances, while the embedded language has an auxiliary function, in supplying selected content morphemes. According to Myers-Scotton (2002), there are two types of code-switching i.e. classical and composite switching. Classical code-switching involves “morphosyntactic frame from one language: namely Matrix language” (Kheir 2019: 480), while in composite code switching, speakers provide the  “morphosyntactic frame from more than one of the participating languages, resulting in a composite Matrix Language Frame.” (Kheir 2019: 480).

Myers-Scotton’s proposed Matrix Language turnover hypothesis (1998) is of relevance to this study, as it links language shift to an intermediate phase of a changeover between matrix and embedded language. During this changeover phase, both languages play the role of matrix languages but gradually the former embedded language increasingly plays the role of a matrix language (Myers-Scotton 1998).

Similarly, the term translanguaging stresses the complexity and creativity of code-switching and mixing (Canagarajah 2011, Garcia & Wei 2014, Slembrouck 2018, Wei 2018). Wei (2018) and Canagarajah (2011) emphasize that translanguaging often seems to be a major form of linguistic behaviour among multilinguals, rather than a temporary space for the meeting of two or more separate languages. Furthermore, such behaviour does not imply that all the languages in the mix are fully known to the speaker. The lexis and occasional phrase from a language being acquired may nevertheless prove useful or attractive to the user. For instance, Jørgensen’s (2008) study showed how polylingual norm in some Turkish families maximized linguistic heterogeneity through the use of the languages available at the speakers’ disposal.

 Canagarajah (2011: 401) emphasises in his work how multilingual speakers treat the diverse languages that treat the diverse languages that form their repertoire as an integrated system. However, this takes us back to the point that Mesthrie (2001) ¹raised about the degrees and directionality of “unmixing” in different domains and registers or styles. Code-switching as a term has two meanings: (a) as a general cover term for the phenomena under study and (b) as a specific term distinguishing it from code-mixing (Stell 2005). It is traditionally accepted (Mesthrie 2001) that code-switching in the specific sense is clear and explainable in terms of external factors (e.g., a topic, the repertoire of an addressee, speaker hierarchies), whereas mixing involves informal interaction between interlocutors and is linguistically ragged as it involves swaying back and forth. However, the insights of Poplack (1980) Myers-Scotton (1992 &1993), Muysken (2000), Auer (2019) unpack the complexities of code-mixing and code-switching. In this study, the term translanguaging encompasses the mixing and merging of lexical items and syntactical structures from three languages: Sindhi, Urdu, and English.

 Studies conducted on language patterns of bi/multilingual Sindhi speakers have so far focused on language shift and maintenance and attitudes towards language shift and maintenance in the external diaspora (David 2000, 2001a & 2001b). None of the previous studies to the best of our knowledge has focused on the dominant language patterns in young Sindhis in comparison to older Sindhis in Karachi in Pakistan’s Sindh province. Therefore, the present study focuses on the internal diaspora of Sindhis and explores the dominant language patterns used by young Sindhis as compared to older Sindhis. This study contributes to the existing literature by investigating language patterns, daily discourse of two intergenerational Sindhi families and reasons for language patterns and mixed discourse in the urbanized city of Karachi.

 Objectives of the study are:

• To investigate the dominant language and discourse among young Sindhis in Karachi.
• To investigate the dominant language and discourse among older Sindhis in Karachi.
• To know reasons for different language practices and mixed discourse among Sindhis in Karachi.

Therefore, the study aims to answer the following questions.

• What is the dominant language and discourse among young Sindhis in Karachi?
• What is the dominant language and discourse among older Sindhis in Karachi?
• What are reasons for different language and discursive practices among Sindhis in Karachi?

2. Literature review

 Language shift in multilingual societies and the globalized world is a common phenomenon (Eslami et al. 2023, Zabrodskaja 2021). Nesteruk (2010) explored the language behaviour of young immigrant European children in USA by interviewing their parents through semi-structured interviews. The study explored parental attitude, transmission strategies, and barriers to transmission and language maintenance and adjustments to language shift and language loss. Findings reported that parents had positive attitude towards language and cultural preservation as they utilized children’s books, media (cartoons & videos), and folktales in their local language to promote and preserve their heritage language. Similarly, ethnic day-care centres, schools, and community centres play an important role in promoting the language. As far as the language transmission is concerned, the presence of ethnic friends and grandparents provides an opportunity to the speakers of the heritage language to practice their language. The heritage speaking house-working mother also maximises the chances for young children to learn the heritage language; although most of the parents reported that they use the native language, yet some parents emphasized using and learning the English language. Regarding language adjustments by parents, the study provides the case of Russian parents who rely on individuals, family, and community resources to promote the heritage language and are fluent bilinguals. Similar findings were provided by Zabrodskaja (2021) who reported that there were multiple factors responsible for language maintenance of Russian as a heritage language in different contexts; these can be internal or external or both. However, the study also reported families from Ukraine who were unsuccessful in promoting their heritage language due to the lack of ethnic resources.

 In multilingual societies, code-switching and code-mixing can involve a complex language mixing (Benchiba 2008, Kheir 2019, Zabrodskaja 2021). Soomro and Larina (2023) work on address terms by multilingual speakers in Pakistan display complex use of codeswitching terms for addressing. This complex language mixing has linguistic items from two or three languages and identifying the dominant language in such a discourse can be oftentimes a difficult task. Researchers have used different terminologies for analysing and identifying this complex mixing. Auer (2019) defines this process as multi-layered switching where there is an intermixture of multiple languages. Similarly, Myers-Scotton (1992) uses the terms code alternating and code shifting but relies upon the degree of ethnic language in code-switched discourse in ascertaining the extent of language shift or maintenance. Myers-Scotton (1992) provides an elaborative ‘Markedness Model’ which can help identify and analyse code switches as more or less marked or unmarked code choices. Wei (1994) analysed the bilingual behaviours, specifically code-switching strategies used by individual speakers for different interactions, and language choices of three generations of a Chinese speech community in Tyneside, north-east of England and found that the older generations and parents communicated in Chinese most of the times while occasionally shifting to English mostly to indicate turn-allocation and repair initiators. In contrast, the members of the younger generation communicated in English when interacting with their peers. However, they sometimes switched and borrowed chunks and phrases from Chinese language mainly depicting insertion sequences.

Kuang (2002) focused on one Malaysian Chinese family who could speak their mother tongue Hokkien along with other languages, such as Cantonese, Mandarin, English, and Malay. She argued by providing numerous utterances of various family members that code switching of multiple languages occurs within a single discourse for many reasons –like home language for making clarification, Hokkien for showing emphasis and English for displaying authority. Hence, in a multilingual context, code-switching of multiple languages often results in mixed discourse.

Blackledge and Creese (2010) discussed a case of bilingual practices of a teacher and students using Gujarati and English in a classroom in the Great Britain. Code switching between vernacular British English and Gujarati/English mixing register resulted in mixing of discourse. Mesthrie’s (1992) study showed how the language (Gujarati) is subject to heavy lexical insertion from the society’s dominant language, English. Blackledge & Creese (2010) considered this to be a mixed register characteristic of informal in-group usage. It might be labelled as colloquial British Gujarati; however, since its grammar is largely that of Gujarati as a matrix language with English vocabulary embedded in it. In other contexts, this variety itself is subject to code alternation with English – that is syntactic switching between English (with only English lexemes) and colloquial British Gujarati. This usage may be said to involve switching as well as mixing. In terms of Mesthrie’s (2001) overview of code-switching, “unmixing” is relatively complex and difficult to identify. Unmixing involves demarking the lexical, phrasal, and syntactic boundaries. In the above scenario, unmixing can be possible in the direction of English; however, unmixing towards Gujarati shorn of its British English lexis is less possible in colloquial contexts, warranting the possibility of it being a mixed code/register. However, it would be interesting to see what kind of Gujarati is used in formal and religious speech making. Mesthrie’s (2001) experience in Cape Town suggests that in this domain speech makers are indeed able to unmix Sanskritised rather than Anglophile register.

Sindhis in Karachi are not only bilinguals but also multilingual speakers of Urdu, English, and Sindhi. In urban areas like Karachi, they have frequent contact with the mainstream language (Urdu) and medium of instruction (English) language and tend to code-switch, code-mix, and insert lexical borrowings to facilitate communication in any of these languages. The present study in view of the linguistic patterns used in daily life aims to explore the choice of language and dominant language in the discourse of young and older Sindhis to determine whether such circumstances result in language shift or not.

3. Data and methodology

A multiple qualitative case study was conducted to focus on young Sindhi speakers, their language patterns, and reasons for changing patterns within their family. A multiple case study is used to focus on the cases that involve similar issues and problems (Creswell 2015). As this study focuses on young Sindhis’ language patterns and reasons for such patterns, multiple cases were selected to compare the information provided by participants. The target population for this study was young native Sindhi speakers (aged between 18 and 24 years) who had migrated from their hometowns for educational purposes and were exposed to a multi-ethnic and multilingual context, where they socialised with different language groups in an academic institute. Hence, they needed to learn and use a language of wider communication. Data were collected using a purposive sampling from twenty participants. Purposive sampling is conducted to get detailed information about the problem and helps in understanding the case effectively; it relies on the selection of participants based on an objective or criteria for enrolling participants in a research study (Creswell 2015). The participants who were selected a) had been residing in Karachi for more than 5 years, b) were aged between 18 and 25 years, c) and were enrolled in an undergraduate degree program at an academic institute and spoke three languages: Sindhi, Urdu, and English. It was not difficult to include these participants in this study as the academic institute which had been selected for the study, had students from diverse ethnic and educational backgrounds.

 Data were collected in two phases. First phase data for language patterns and self-reported proficiency in different domains was collected from twenty young Sindhi speakers using a sociolinguistic profile, which contains information about language proficiency and language use in different domains (Fishman 1991). For second phase, after analysing the responses in the sociolinguistic profile, observation and interview were conducted. The participants who were selected and included in the study for the second stage were those whose parents reside in Karachi and who permitted observation of their daily discourse. Therefore, after analysing their choice of language, interviews and observations were conducted with them and their families. After formal consent and frequency analysis of the initial data, two younger Sindhi speakers along with their families (two families) voluntarily became research assistants to help observe and record the discourse in real settings. Within the families, data were collected from 02 male parents (Older Sindhis aged 40 years and above), 02 female parents (Older Sindhis aged 40 years and above) and 02 younger Sindhi speakers. The discourse was recorded after the formal consent in the home and education domains for twelve hours (eight hours in the home domain and four hours in the education domain).

 Semi-structured interviews were conducted with the members of the families to elicit naturally occurring multilingual speech. Informal discussions with the six participants were recorded for twelve hours (eight hours with older participants and four hours with younger participants) in the mother tongue (Sindhi) and when needed to accommodate language preferences of younger participants, researchers switched to Urdu and English. The interviews sought to explore reasons for changing language patterns and mixed discourse in the participants’ daily life.

 The data have been analysed using frequency analysis for language patterns and thematic analysis for observations and interviews. The observed data were reported in the form of extracts. These extracts were read and reread to do coding, develop broader categories, and ultimately placing together the segments of the text that co-related with a theme (Nesteruk 2010). Once the data are compiled, for triangulation, the analysed data are shown to the participants for verification (Patton 2002). The younger Sindhi participants confirmed the data and its accuracy in the extract; hence confirming the credibility of the data. For ethical purposes, a consent was signed by the participants before observations and interviews, and it was assured that the data would be used for research purpose, and their identity would not be disclosed.

4. Findings

4.1. Dominant language choices of Sindhi youth

A sociolinguistic profile provides information about the four skills of language and language use in different contexts (Abbasi & David 2020a). Table 1 shows the language patterns of the speakers.

Table 1. Language Proficiency of Young Sindhis

Domain	Sindhi	Urdu & English	Mixed
Home	30%	20%	50%
Other Domains Neighborhood, market, university, media etc.	5%%	90%	5%

Table 01 shows that 30% of the participants use the Sindhi language, while 20% switch to Urdu and English. However, 50% of the participants reported that they used a mixed variety like Urdu-Sindhi, Sindhi English, Urdu-English or a combination of the three in their home domains with their parents and siblings. While responding to the question related to language use in other domains (neighbourhood, market, university media etc), almost 90% (18 participants) reported they used Urdu in the ‘other domains’, such as in a market, classroom, social gatherings, workplace, canteen, restaurant, and social media. In fact, on social media, the medium of conversation with family and friends was Urdu.

 Participants also self-reported their proficiency in Sindhi reading and writing as poor. As for the spoken discourse, 25% of the participants said they used the Sindhi language in their home, 12.5% said they could partially communicate in Sindhi, while 12.5% could only understand it (i.e., they were passive or receptive bilinguals). Moreover, remaining 50% of the participants said that they could communicate more fluently in Urdu and English than in Sindhi. No participant reported reading a Sindhi newspaper and listening to a Sindhi television channel. Almost (90%) all of them rated their speaking and writing proficiency as good in English and Urdu. From this frequency analysis, it appears that the young Sindhis who reside in the city have low proficiency in writing and reading in Sindhi.

4.2. Dominant language discourse among Sindhi speakers

 The second aim of the study was to determine the language choices of the Sindhi community in different settings. Data in extracts 1–3 demonstrate a mixed discourse of younger and older Sindhis. Codes are used for different family members in the extracts. G2FAs and G2FBs mean father parents’ discourse in the extracts 1–3; G2MAs and G2MBs refer to mother parents’ discourse in extracts 1–3. Moreover, GY2As and GY2Bs are codes for younger Sindhis’ discourse, and the ‘friend’ word denotes a Sindhi speaking friend of younger Sindhis in extract 1–3. The data have been transcribed using the conventions of Roman transcription and symbols in Sindhi and Urdu as used by Pakistani linguists. Hence, to differentiate and identify the discourse patterns, different types of fonts have been used. Italics have been used for Sindhi, Bold for Urdu, and Roman for English in Extract 1–3.

4.2.1. Discourse among younger and older Sindhis

The generational code-switching patterns observed in the data follow the patterns of prior studies (Myers-Scotton 1993). The overall analysis of extract 1 shows that the older generation uses Sindhi, and that within a single clause, Sindhi is the matrix language where the verb and object are from the Sindhi language, while the other lexical items are from English or Urdu. As in the first utterance in extract 1 line 1 Sindhi matrix (as shown by verb and auxiliary plus pronoun) become one clause. However, English lexical insertion of “federal public” or “education-related: examination test” lexical items are added to the utterance. In this example the older respondent used English as the embedded language.

Extract 1

Extract 1 Communication and Language between Older and Younger generation while discussing about Education. Speaker G2FAs: Male Older Sindhi G2FYAs: Male Younger Sindhi [English word Times New Roman, Urdu bold and Sindhi Italicised] 1 G2FAs: Chha tvn federal public Jv examination test DynDy? [Will you give Federal public examination test?] 2 G2YAs: Test tu mushkil nahi hota, sochna paRay ga. [Test is not difficult; I should think about it] 3 G2FAs: Test ta ASan Aa, bus current affairs jyy knowledge ho [test is easy, but one should have knowledge of current affairs] 4 G2YAs: Bilkul; yes [Of course, yes] 5 G2FAs: PaRa tayyari institute Khaa karna [But do preparation from an institute] 6 G2YAs: tekh, indeed. [Okay, indeed] 7 G2FAs: test mei~ common sense Aa, English and composition; you will do it. [Test is about common sense] 8 G2YAs: Pass karna issue nahi; bus material mil jaye; toh it will be easy. [Passing the exam is not difficult, but if one gets the material so it will be easy]

Older generation (G2FAs) also uses Sindhi as a matrix language in the utterance signalled by active system morphemes ta and the auxiliary verb followed by the lexical insertion of the English word test. As indicated in Extract 1 (Line 3, G2FAs), the utterance starts with an English subject test in which a Sindhi verb ta (auxiliary verb ‘is’), and a Sindhi object ASan Aa ‘easy’ are used, and the second clause starts with an Urdu lexeme bus ‘but’ followed by an English noun phrase current affairs, a Sindhi verb jyy ‘is’, an English object knowledge and ends with an Urdu lexeme ho ‘should’. Urdu is syntactically active for this clause with heavy lexical insertion from English (almost relexification – meaning that the grammatical sequence is Urdu, with the slots being filled by English words). Hence, a single clause contains an Urdu lexeme, an English phrase, a Sindhi auxiliary verb, an English phrase, and an Urdu lexeme in the utterance. The sentence structure has complex intermixing of Sindhi, Urdu and English words which makes the identification of a lexeme, phrase, and clause difficult, whereas in simple intermixing there is often lexical insertion of loan-words from the two languages (English and Urdu). Hence, the older generation i.e., a father parent in this utterance more frequently uses a mixed variety with lexical borrowings from English and Urdu in a natural Sindhi matrix sentence to facilitate the young Sindhi speakers to better understand the utterance due to their exposure to a range of multilingual contexts.

 Comparatively, the younger Sindhi speakers shift to Urdu and English (the matrix languages) by reducing the use of Sindhi lexical items and phrases in their utterances. As in the first utterance in extract 1 line 2, a younger Sindhi speaker (G2Yas) says “Test tu mushkil nahi hota, sochna paRay ga” ‘Test is not difficult,  I should think about it’. There is an English lexical insertion which is a polyvalent part of colloquial Urdu. The younger Sindhi speaker uses Urdu as the matrix language clause in their utterance because Urdu is used as a common lingua franca in academic institutes.

 The first Sindhi family mostly used Urdu syntax as a matrix with some English lexical insertions (test, yes, pass, interview, and issue). As an example, in Line 8 (by G2YAs), the first two clauses signalled by semi-colons have Urdu as a matrix language. It is noticeable that the educational-related, lexical vocabulary is mostly taken from the English language (pass, issue, material). The last clause resumes the Urdu syntax (with a discourse connector toh which means ‘so’), but the insertion of an English “S” clause tails off the Urdu syntax. In Muysken’s (2000) words, this is code alternation. Moreover, this complies with Myers-Scotton’s (1993) concept of an island – i.e., a stretch of exceptional material in the embedded language, in this case necessitated by having started the clause with the pronoun “it”. (Pronouns frequently impose a constraint against switching (Poplack 1980). In contrast, a discourse connector like toh or even its equivalent in English, (‘so’), which is syntactically external to the clause imposes no such constraint, and in this case, the speaker does indeed switch to another language.

4.2.2. Discourse of young Sindhis

 In Extract 2 G2YBs uses a mixed code with his Sindhi speaking friend in the university where Urdu, English and Sindhi lack grammatical integration and there are clear patterns of an alternational hybrid code. There is no clear distinction at the clausal level about the dominance of one language rather three languages are  inter-mixed forming a new pattern which is composed of translanguaging of Urdu, Sindhi, and English vocabulary, phrases, and clauses. As in Line 1 Extract 2 the utterance by a young Sindhi speaker began with an English phrase followed by an Urdu verb, intermixed with an Urdu and Sindhi adjective which is used to emphasize feelings and ends with an English phrase. Therefore, a simple SVO utterance has become complex, where Urdu is intermixed with Sindhi for emphasis. Similar utterances are quite common among the young, as shown in the same extract. These utterances show the pattern that within a single multilingual utterance there is a frequent shift towards English and Urdu lexical items rather than Sindhi lexical items. Moreover, in complex multilingual utterances, Urdu and English phrases are more often used, and there is a shift away from the Sindhi language as observed in the extract 2.

Extract 2

G2YBs: Male Younger Sindhi Friend: Male Younger Sindhi [English word Times New Roman, Urdu bold and Sindhi Italicised] 1 G2YBs University mein life buhat Dukeh aa [Life is so difficult in university] 2 Friend Bilkulsahe galh aa [Of course, true it is] 3 G2YBs Teacher sakht Aa number dadahmuskil se gain kran pawanda [Teacher is strict, gaining the marks is a tough task] 4 Friend: Yahan toh system he aisa ha dost, wrna Paisa kaya tahan [The system is like this friend otherwise give bribe] 5 G2YBs..hiteh dunya fastAa…asan Keh fish catch krne hoge but ustaad kehte habajoh na. [Life is too fast and we have to catch the fish but teacher say don’t run] 6 Friend Dost fikar na Kar; Allah malik aa; beya semester beh guzre wenda [Friend do not worry, God is with us, and this semester too shall pass] 7 G2YBs hum society per impact kehro tha deyo … yeh dekhna pare ga long term mei~…. [What impact are we creating on the society? we have to see that] 8 Friend Dost fikarna kr, jobs available ha asan sabneyn jee laye bus mehnat kayo ta sabasanjo aa [Friend, do not be concerned, jobs are available we just need to work hard] 9 G2YBs Hazir bhai… Halaon compensatory class mei~ [Okay brother, let’s go to the compensatory class] 10 Friend Halaon tha compensatory watahar [Okay lets go to take the compensatory]

4.2.3. Dominant language variety among older and younger Sindhis

In Extract 3, different utterances are provided by the speakers [Speaker G2FBs a father; speaker G2YBs, a nephew; and Speaker G2MBs, a mother]. Although the father (Speaker G2FBs) uses Sindhi, but English and Urdu lexical items are frequently inserted. However, Sindhi is dominant in the mixed register. The nephew from the younger generation (speaker G2YBs) uses Urdu with a mix of English words. Hence, the older Sindhis (both father and mother) use a mixed variety while communicating with the young Sindhi speakers perhaps to facilitate conversation, while Speaker G2YBs (a nephew) shifts to Urdu with English words inserted in the discourse. These patterns show that the nephew (a young Sindhi speaker) in this conversation shifts away from his mother tongue (Sindhi) and prefers to use English and Urdu by adding clauses and lexical items from the two languages rather than Sindhi lexical items when communicating with older native Sindhi speakers.

Extract 3

Speaker A: Male Older Sindhi Speaker B: Male Younger Sindhi Speaker C: Female Older Sindhi [English word Times New Roman, Urdu bold and Sindhi Italicised] 1 Speaker G2FBs: Hammad Rvzv Aa? 2               [Hammad are you fasting?] 3 Speaker G2YBs: Nahi.                                                                  [No] 4 Speaker G2MBs: Hammad kese roza rakhe ga?                         [How will Hammad fast?] 5 Speaker G2FBs: jeya zamanorakhta ha in the holy month         [As the society fasts in the holy month] 6 Speaker G2YBs: par yaha~ bohat garmi he, university bhi jana hota he [But it’s too hot here, we must go university as well] 7 Speaker G2MBs: phir kya sara din AC me guzaro gay? [Then you will spent the whole day in AC?] 8 Speaker G2FBs: ThyKh poth; University ha [Okay son, University is on] 9 Speaker G2YBs: asal me yaha~ bohat garmi hoti he; phir aap ko university me classes bhi leni hoti he; it becomes hectic. [Actually, it’s too hot here and then you have to take classes in university as well, so it becomes hectic] 10 Speaker G2FBs: par beta water to nahi milay Ga University me? [But son you won’t get the water in university] 13 Speaker G2MBs:                                                                        [tum kya Khaao gi? [What will you eat?] 14 Speaker G2FBs: gosHt and naan* {*a cultural food item}                           [beef and naan] 15 Speaker G2YBs: nahi uncle; I have already eaten Chines    [No uncle, I have already eaten                                                                                                   Chinese] 16 Speaker G2FBs: Thiik     [Okay]

Table 2 shows the analysis of the data in the form of frequencies. A detailed analysis has already been provided in Extracts 1-3 above.

Table 2. Frequencies of Syntactic Language Patterns 

Clause content	Speaker 1 (older) Male	Speaker 2 (older) Female	Speaker 3 (younger) Male	Speaker 4 (younger) Female
Sindhi clause without mixing	20	16	04	03
Mixed clause showing alternation of syntax **	-	-	02	02
Mixed clause with Sindhi structure (and lexical insertions)	13	08	04	07
Mixed clause with Urdu structure (and lexical insertions)	05	02	23	21
Mixed clause with English structure (and lexical insertions)	01	-	12	07
Urdu clause without mixing	03	-	16	09
English clause without mixing	-	-	20	22
Other/unclear/unsure	-	-	-	-

Differences in code-switching patterns between 2 older generation and 2 younger generation members were illustrated by dominant codes per clause in table 2. The data in the table showed that the older Sindhis frequently used Sindhi clauses. Similarly, the young Sindhi speakers frequently used mixed clauses with Urdu and English structure. This could be due to exposure to foreign and local languages in the urban context. Moreover, the older Sindhis hardly uttered complete English clauses and merely used Urdu clauses in their discourse. Comparatively, the older Sindhi female speaker used more Sindhi sentences than Urdu and English in the utterance. Moreover, the younger Sindhis used mixed structures with English and Urdu with greater frequency as compared to older Sindhis.

 Overall, the analysis of the interactional practices shows that the older Sindhi participants use a mixed code, but Sindhi lexical items dominate. In contrast, the younger Sindhi participants produce a mixed discourse which includes the three languages Urdu, English, and Sindhi, though their Sindhi discourse tends to have fewer lexical Sindhi items. This use of a mixed code is facilitated by older Sindhis and Sindhi friends of younger Sindhis. Thus, in the urban areas, young Sindhi speakers are producing a mixed code. A new code of communication is identified, and it is a hybrid code that involves mixing of Urdu, Sindhi, and English. However, the hybrid code is so complex that it is difficult to identify lexical and phrasal level utterance of a single language, as already highlighted in the extracts (See Extract 1-3). Another trend identified is that the younger participants are shifting away from their mother tongue and in the hybrid translingual code Sindhi lexical, phrasal, and clausal choices are minimal as compared to Urdu and English. Perhaps over time, young Sindhis living in a multilingual city like Karachi, will become more proficient in the dominant languages and the mother tongue might over time disappear from their spoken discourse resulting in language shift.

4.3. Reasons for changing language patterns among Sindhis in Karachi

These findings emerged from the interview data and provided reasons for adopting new language patterns and mixed discourse and shift to more dominant languages.

4.3.1. Migration to the city and significance of Urdu

The Sindhis migrated to the city due to educational and economic reasons. The parents reported that as they migrated to Karachi, “they had to leave their ancestral homes, villages, and their parents (grandparents of younger Sindhis) and adopted urban lifestyle”. G2Fas (Father: Older generation) says: “mujhe achha laga, taliim aur bachhay bhi khush [I feel good for my children’s education, and they were happy too). While narrating the significance of the Urdu language in the urban area, the participants (G2MBs & G2FBs) emphasized that the basic education system was in Urdu; and even in English schools Urdu was emphasised as G2FBs said “aur bachhay bhi Sindhi nahi boltay thay balakay urdu ki taraf ail hogaye thay, taliim bhi unki urdu me ho rahi thi, grammar school thay kuchh, English Grammar school lekin un me ziada bol chaal urdu me hi thi [And kids do not speak Sindhi and they shifted to Urdu, because basic education was provided in Urdu and English in their school and most of the communication was in Urdu in the school]. In short, the younger generation adopted Urdu and English in different language use contexts and shifted away from the habitual use of their language.

4.3.2. Intergenerational language transfer

Most of the participants while discussing the importance of intergenerational maintenance of the ethnic language admitted they facilitated the transfer of the dominant language as parents were working professionals and grandparents resided in villages, so they did not get many opportunities to transmit heritage language like telling folktales in Sindhi, reading stories or grandparents communicating in Sindhi. The parents also reported that there were few language resources like G2Mas said “jab bchon ko khelana hota, video ya cartoon dekhana hote toh who Sindhi mein available nahe the pehle” (when they want to engage or show videos to their child, it was not available in the local heritage language)”. Similarly, G2Fas said “there were no local language centres in Karachi to teach the Sindhi language”. Hence, Urdu was the language frequently used in society and education centres, and therefore parents facilitated this shift to Urdu. Another reason for encouraging the use of Urdu was exogamous marriages (Abbasi & David 2020b). In such cases a neutral language was adopted. One of the participants said:

“bas easy lag raha tha, bachho~ ne urdu bolna shuru kiya, unki maa tou Punjabi thie~, me~ sindhi tha aur zaahir he bachho~ ne jab school me paRhna shuru kiya, gali me ya mohallay me ziada tar urdu hi bolne lagay tou urdu ko hi unho~ ne apna liya. Tou ham ne bhi ziyada munasib nahi samjha ke zor daal do~ ke sindhi siikhe~ ya zor daale~ ke Punjabi siikhe~. Urdu ko hi un ki mother tongue rehne diya”.

[It felt like kids started speaking in Urdu and their mother was Punjabi although I was a Sindhi speaker. Of course, the kids started communicating in Urdu in school and neighbourhood because of the presence of the Urdu community, so the kids adopted the Urdu language as their preferred language and we did not emphasise the learning of Sindhi or Punjabi, so they made Urdu as their mother tongue].

The younger participants stated that their siblings and parents talked to them in Urdu; so, they also used Urdu.

In mixed marriages and in an urban setting (Abbasi & David 2020b, David 2008) where other languages are more powerful, children from such exogamous marriages will tend to use a language of wider communication, such as Urdu.

4.3.3. Status and value of the Sindhi language among the Sindhi community

The language patterns reflected the status of languages among the two families who were shifting to Urdu and English; yet it was important to explore the Sindhi families’ attitudes towards the Sindhi language. G2Fas narrated: “hamare bche Sindhi nahe bol rehe ha, lekin phr bhe society me~ Sindhi bole jaate he, offices, shadio~ aur ab toh Karachi me~ buhat ziada Sindhi rehte he, 1990s mein itne nahe the” (Our children cannot speak the Sindhi language, but it is still spoken in society, offices, and social events. Today, there are more Sindhis in Karachi than in 1990s). Similarly, G2MBs said: “hamare bache Sindhi bolne ke koshish krte he, lekin university mein nahe bole jate esleye I English aur Urdu per focus krte he; lekin phr bIwoh apne culture se hum ahang he” (Our children try to speak Sindhi, since it is not spoken in educational institutes, they focus on English and Urdu but still they are aware of the Sindhi cultural values).

When asked about the Sindhi culture, most young Sindhis were aware of the cultural cuisine, weddings, rituals, and events. Some of them even participated in the Sindhi cultural events in their academic institute. While visiting the house of the participants, the younger participants showed their traditional Sindhi cap (Topi) and a Sindhi cultural shawl (Ajrak). Hence, the younger Sindhis were proud of their cultural values and participated in social events. However, to assimilate to city life they had shifted to Urdu and English.

5. Discussion

 The present study analysed the language patterns and composition of the language used by twenty selected Sindhi respondents in Karachi, Pakistan. It investigated reasons for the dominant language use in their mixed discourse, with a view to exploring the phenomenon of language shift in the Sindhi community.

 Daily life language practices are the strongest indicator of language maintenance (Fishman 1991, Nesteruk 2010); however, these trends are changing in this study. The language patterns showed that only 30% speak Sindhi at home, while 20% use Urdu and 50% used mixed discourse (borrowing words from Urdu and English). Similarly, 90% reported that they used Urdu in social domains. Fishman (1991) reported that home is the place where heritage language can be spoken frequently and preserved, however, this study showed that the language patterns were changing, and younger Sindhis relied on the dominant languages (Urdu and English) rather than their heritage language. Nesteruk’s (2010) study showed that educational and community centres, educational resources, role of mothers, society and grandparents were highly significant in promoting a heritage language. In this study, it was reported by the parents that there were no significant language resources and language centres available, hence the reading and writing skills in Sindhi were not developed. Mothers were professionals and spent less time with children, while grandparents resided in distant villages. Therefore, as reported in this study the young Sindhi participants in Karachi did not read and used social media in their heritage language. Hence, their syntactic patterns were bound to change as reported in the observation and interviews. In this study, Myers-Scotton’s (1993) intergenerational code-switching patterns have been followed; the older Sindhis maintained their heritage language and facilitated language shift among younger Sindhis by assuming a positive attitude towards the dominant languages in society and using a mixed discourse.

The language used by the younger Sindhi speakers reflected a complex phenomenon of polylanguaging in which three languages are used simultaneously within a single utterance, making the lexical, phrasal, and clausal boundary difficult to identify. Auer (2019) termed this phenomenon as translanguaging (complex language mixing which has been observed in the discourse as shown in  extracts 1–3). The phenomenon has also been denoted as alternational mixing (Muysken 2000) where there is no grammatical integration and insertional mixing (Muysken 2000) where words from one language are inserted into another at lexical level, however, phrasal, and clausal boundaries are identical. This is like what Blackledge and Creese (2010) reported in their study of classroom discourse, where Gujarati and English mixing reproduced a new code as the linguistic boundaries were not identical. Gujarati was subject to lexical insertions from societal language (English). Similar findings were reported among Bhojpuris in Mesthrie (1992). The same findings have been reported in this study as well in the discourse of young Sindhis which have lexical and phrasal insertions in a single utterance from Urdu and English to the extent that the lexical boundaries are blurred, and a new code has been formed of a mixed variety of three languages.

 Similar findings were reported by Wei (2008) among Singaporian Chinese where such mixing happens and results in multilingual code between Hokkien, Teochew, Mandarin, Malay, Cantonese, English, and Singlish. Meeuwis and Bloomaert (1998) call this phenomenon as ‘multilayered switching’, while Haque (2011), associated such language practices among three Indian families in Europe speaking three to five languages with ‘truncated multilingualism’ where participants shifted from one language to another according to specific domains using limited lexical chunks from different languages in their discourse.

 Similarly, when inquired about the status, value, and intergenerational transmission of the Sindhi language, the participants reported that internal migration, educational needs, and social mobility are some of the reasons for adopting Urdu language. The older Sindhis admitted that they facilitated the shift of the Sindhi language to Urdu, so that their younger Sindhis could learn to live in the larger Urdu speaking society. However, they had introduced their children to their heritage culture, and were aware of the Sindhi traditions and values.

Hence, the present study explores a situation where the language patterns are shifting, and discourse is built upon a hybrid of three languages which has no clear boundaries.

Such extensive code mixing and switching using more than two languages can be designated as translanguaging. As patterns identified in this study show a merging of the three languages, but the extent of use or non-use of Sindhi suggests a shift away from the ethnic language, at least for members of the speech community who reside in Karachi. Lexical shifting in many utterances appears to be the first step towards language shift, and perhaps with time, language attrition and finally language loss.

6. Conclusion

The findings of this study reveal that the older Sindhis in Karachi are facilitating the language change process. Older Sindhis are not promoting the use of the Sindhi language in the home domain and are using mixed discourse. This results in mixing of the three languages especially Sindhi-Urdu-English lexical and phrasal insertions. This shift is attributed to lack of resources, language centres and the social settings where Urdu and English predominate.

The younger Sindhi respondents are exposed to multilingual practises in a multilingual setting; they frequently switch from one language to the other. The older and younger participants communicate in a mixed variety with lexical, phrasal, and syntactic insertions from Urdu and English. As a result, a new code, with minimal use of Sindhi lexical items has emerged. When comparing the discourse of the older with younger Sindhi speakers, the younger respondents are tilted towards the mainstream languages Urdu and English.

 Such a process would ultimately lead to language shift and language loss among the younger Sindhis, and the language may not be transferred to coming generations. Therefore, significant steps must be taken by the community members and Sindhi language experts to produce language resources in the Sindhi language using print media and social media. Books should be published in the Sindhi language that introduce young children to Sindhi alphabets, it is also worthwhile to open the Sindhi language centres in Karachi that can offer to teach vocabulary and writing skills to children. These centres should be open in different areas of the city.

 

¹ Mesthrie, R. 2001. Concise Encyclopedia of Sociolinguistics. Elsevier Science, Regional Sales Office, Customer Support Department, PO Box 945, New York, NY 10159-0945.

About the authors

Muhammad Hassan Abbasi

Dawood University of Engineering and Technology Karachi

Author for correspondence.
Email: hassan.abbasi@duet.edu.pk
ORCID iD: 0000-0002-1232-0963

Lecturer in English at Dawood University of Engineering and Technology, Karachi, Pakistan. He is a PhD Fellow at the Institute of English Language and Literature, University of Sindh. His area of expertise includes language attitudes, language shift and maintenance, indigenous languages, language planning, family language policy, language and gender, TESOL and digital education.

Pakistan

Maya Khemlani David

Asia Europe Institute, University of Malaya

Email: mayadavid@yahoo.com
ORCID iD: 0000-0002-5569-3591

Ph.D., is an Honorary Professor at the Asia Europe Institute, University of Malaya, Kuala Lumpur, Malaysia. She is a sociolinguist who was awarded the Linguapax Prize for her studies on language shift in Malaysia and neighbouring regions. She has published widely and made plenary presentations in several countries.

Kuala Lumpur, Malaysia

Ameer Ali

College Education Department, Government of Sindh

Email: ameer7037@gmail.com
ORCID iD: 0000-0001-9438-4903

Lecturer in College Education Department, Government of Sindh, Pakistan. He is a solidarity member of the Foundation for Endangered Languages. His areas of interest include sociolinguistics, forensic linguistics, discourse studies, language shift and maintenance, language and ageism.

Pakistan

References