What causes a community to shift from one language to another is generally a significant consideration for language maintenance programs. Yet research in specific language communities seldom investigates the community's ideas about the causes of its language shift. The researcher thus runs a risk of assuming that the attitudes of one or two key informants reflect the attitudes of the entire community. This assumption might in turn influence the direction of the research and may render it irrelevant to the community's needs. The ideas voiced by members of a community experiencing language shift and the implications these ideas have for the introduction and success of any future language maintenance efforts are discussed here. Assessing the community's understanding of the causes of its language shift is a crucial first step to take before attempting language maintenance. Community members need to be empowered to see language shift as something they can influence, rather than as something beyond their control.

Speaking Arabic in Turkey

Speaking Arabic in a Muslim country might seem unremarkable to the uninitiated. In Turkey, however, the only official language is Turkish, and speaking any other mother tongue, even Arabic, is discouraged; for a time (1983-1991), some languages were prohibited outright.

Although there are no official statistics on language use or on ethnic groups in Turkey, it is clear that in the province of Hatay (in the south, bordering Syria), most people are descended from Arabic speakers. Arabic entered the area as a result of the Arab conquests in the seventh century. Prior to this, the cities were Greek-speaking; people in surrounding areas spoke Aramaic. (Trimingham, 1979) The area first came under Turkish rule for a brief time at the end of the eleventh century, when Seljuks and Turkmen began eroding Byzantine control. Crusader rule followed, and it has been alleged that from the eleventh to the thirteenth centuries even the "Christians were...Arabs, both culturally and linguistically" (although Greek was the liturgical language for many more centuries). (Courbage & Fargues, 1997) The area subsequently came under Mameluke rule (from the thirteenth to the sixteenth centuries), then under Ottoman rule (from the sixteenth to the twentieth centuries). During both periods, the people spoke Turkish. As part of Greater Syria, Hatay was still largely Arabic-speaking when it was annexed by the Turkish Republic in 1938.(1)

Until annexation, Turkish and Arabic co-existed for centuries; under republican policies, however, the use of Arabic began to decline. Less than a generation ago, a child of an Arabic-speaking family would start school unable to speak Turkish; these days, most children of Arabic families start school unable to speak much, if any, Arabic. Some Arabic speakers will deny being "Arab," a term that can be derogatory in Turkey. I will therefore refer to the community as "Arabic," reflecting the fact that the ancestral connection to the language is the common characteristic.

Despite their linguistic connections, Arabic speakers in Hatay do not see themselves as one unified community, and religion, not language, is the primary marker of identity.(2) There are three main religious divisions: Christian (mostly Greek Orthodox), Sunni Muslim (like most Turks), and Alevi (a sect peripheral to Islam, but which shares some beliefs with Shiites). Most of the Arabic community in Hatay is Alevi.

Although the Treaty of Lausanne, signed in 1923 between the new Turkish Republic and the European allies, provided for non-Muslim communities in Turkey to teach their languages, no such provisions were made for Muslim communities. The Greek Christian community in Istanbul, for example, has its own schools which continue to teach Greek to this day. When Hatay was annexed fifteen years after the Treaty was signed, however, the Arabic-speaking Christian community did not avail itself of the right to teach Arabic. Had it tried to do so, the result would have been an untenable situation, with the Christian minority permitted to teach the language of the majority, but the majority itself being prohibited from doing likewise.

Turkey's policies on language have focused on imposing homogeneity. The degree of imposition peaked in 1983, when the military government introduced a law prohibiting (to varying degrees) languages other than Turkish. For speakers of some languages (those not using a first official language of a country recognized by Turkey), the law forbade the use of those languages, even during private conversation. (Rumpf, 1989) Although the law was repealed in 1991, the Constitution still prohibits any institution from teaching a language other than Turkish as a mother tongue (Article 42.9; provisions in international treaties are ostensibly upheld even today).

Although in theory Arabic-speaking Christians have more linguistic rights than Arabic-speaking Muslims, the government's treatment of Arabic as a community language (as opposed to a "liturgical" one) has been uniform irrespective of religion; the Muslim and Christian communities can thus be regarded as a single linguistic group. Modern Standard Arabic (the literary form for any contemporary dialect of Arabic) is, theoretically, permitted as a foreign language -- for Christians and Muslims alike. In reality, literary Arabic (albeit seldom used in Hatay) is not a foreign language, nor is it taught as such.(3) The notion of teaching Arabic as a foreign language as opposed to a mother tongue has many implications, including pedagogical ones.

Community Members' Ideas about Language Shift

As part of a larger study (in progress) examining the role of code-switching in the language shift in Hatay (Myers-Scotton, 1998), a set of final interview questions was designed primarily to collect demographic information and subjective judgements on language use. The sample (an extended family and its social network, selected in order to obtain natural conversational data for the larger study) comprised 73 Arabic individuals, ranging from eleven to seventy-two years in age. Most (64) were from the city of Antakya (Antioch); the rest were from near-by villages. Most (53) were Christian; the rest were Alevi. Approximately half (39) were under 35. The participants were asked if Arabic was decreasing in Hatay. If they said "yes," they were asked why and whether they thought this was good or bad (or not important). The participants' responses suggest a sense of powerlessness on the part of speakers to affect language shift.

Eleven participants (15 percent of the sample) -- let us call them the deny-ers -- said that Arabic was not decreasing; yet some of their comments suggested an awareness of a shift:

A 63-year old Christian man said:

"No. [It's] more...the Alevi more than us of course. The Alevi are all Arab. (Interviewer: Do they speak [it]?) [They're] all Arab; that is, they don't speak anything else."

A 29-year old Christian woman said:

"No.... [There are] lots of Alevi. In general, Arabic is spoken."

A 46-year old Alevi woman living in the city said:

"No. [There are] lots who speak, lots in the villages."

In choosing to focus their answers on a group other than their own, these speakers many have implied that Arabic had been decreasing in their own group (in the case of the Christians) or in their area (in the case of the city-dwelling Alevi). Yet, because they still heard the language being spoken by others, they believed that Arabic was not decreasing.(4)

Those who said Arabic was decreasing (62 respondents, 85 percent of the sample) -- let us call them the realists -- were then asked why. Three causes far outnumbered other responses: schools forbidding Arabic or not offering it as a subject (31 respondents, 50 percent); the older generation/parents not teaching the language (29 respondents, 47 percent) and the younger generation/children not wanting to learn it (18 respondents, 29 percent). All but two realists mentioned one or more of these three causes. Other answers varied, but were similar to causes cited in other studies:

- Arabic-speaking elders dying off, reducing the opportunity to learn the language (9 = 15 percent);

- the language as unnecessary for life in Turkey (9 = 15 percent);

- increased mobility around Turkey and higher education causing people to use Turkish more frequently (9 = 15 percent);

- Arabic a difficult language to learn (6 = 10 percent);

- the government's pursuit of assimilationist policies (6 = 10 percent); and

- Turkish used to camouflage "Arab" origins (4 = 7 percent).(5)

The three main causes -- educational availability and the two generational behaviors -- far outnumbered other responses. They were considered causally interconnected, since parents mentioned pressure from teachers not to speak Arabic with their children. One woman recalled that when she was a child, Arabic was forbidden at school (as it is today) and that she told her parents she didn't want to speak Arabic anymore. She added, however: "If [I had been] in my current state of mind, I definitely would have learned Arabic."

Bringing these responses together, there were 10 members of the older generation who saw the younger generation as the cause of the language shift, 22 of the younger generation who saw the older generation as causing the shift, four who blamed their peers but considered themselves exceptions, and 12 who blamed neither generation but indicted the educational system. Not including the two who blamed both generations, we end up with approximately 46 (74 percent of the "realists") who saw the causes of language shift as being, to quote one respondent, "for reasons out of our hands."

Implications for Language Maintenance

Community members' beliefs about the causes of language shift would be immaterial if the community were happy at the decline of Arabic, but it seems as if members of this group would prefer to maintain it. Of the realists, 40 (65 percent) said that the decline was a bad thing; another 12 (19 percent) had mixed feelings (the others did not think it was important). One (Christian) deny-er linked language maintenance to religion, saying that Arabic had to continue so that the Bible would be understood when it was read. In fact, a religious motivation to maintain Arabic was shared by all three religious groups; even Orthodox Christians stated that they used Arabic as their language of religion. One Alevi convert to Catholicism said that, although he could not understand Arabic (because his parents had not taught him the language), he found its use in the Orthodox service mystical. A cultural motivation to maintain Arabic was also mentioned by a number of respondents; 11 quoted an Arabic proverb affirming the value of multilingualism: "Every language is a person."

An apparent obstacle to language maintenance efforts exists, despite religious and cultural affirmations. If community members see the causes of language shift as out of their hands, then language maintenance efforts may well seem doomed from the start, and are likely to suffer from a lack of impetus. The current lack of language maintenance efforts in Hatay is not necessarily due to indifference to the language; rather, it reflects the community's ambivalence as to whether it (can) have an effect on language shift.

Where to Start?

Since 50 percent of the realists cited the education system as a cause of language shift, perhaps language classes should be the starting point for a language maintenance program. Researchers in other parts of the world, however, have expressed concern that too much reliance is often placed on such programs. Furthermore, such programs require resources (teachers, classrooms, textbooks, etc.), and this need can lead to bureaucratic hurdles. Even with substantial state support, classroom efforts to maintain languages are not especially successful on their own. In Ireland, for example, where ample funding, resources, and incentives have increased knowledge of Irish at some levels, the number who speak it as a first language is still declining.

In Turkey, a classroom approach is impracticable for other reasons. As noted above, the Christian community could teach Arabic, but at the risk of creating resentment in the Muslim (Alevi and Sunni) communities. Arabic could be taught as a foreign language, but this would preclude a "language nest" approach to teaching Arabic (language nests are implicitly about learning a language as a native, not a foreign, tongue).(6) Furthermore, moves to teach Arabic, even as a foreign language, might be interpreted by the authorities as evidence of affiliation with Syria and thus, with the promotion of separatist goals. Moreover, smaller minorities in Turkey are less likely to pursue structured language maintenance efforts until the Kurds have unambiguously secured cultural autonomy.(7) In this climate, even reporting the results of this research to the community needs to be done in such a way so as to avoid an antagonistic reaction from authorities.

Just because an education system was a major contributor to a language shift does not mean that it is the best means by which to "reverse" that shift. The causes of a language shift are only relevant when planning a language maintenance program (insofar as they sustain the shift). The key issue to be addressed is how to inhibit (and perhaps even eradicate) language shift itself. In the case of the Arabic speakers in Hatay, the perception that the causes of the language shift are out of their hands sustains the shift by implying that community members cannot have an effect on reversing it. To initiate a language maintenance program would require addressing this sense of powerlessness. Bearing this in mind, let us re-define the concept of language shift.

Redefining Language Shift in Empowering Terms

Language shift has been defined in various ways: "the loss of linguistic skills between generations" (De Bot, 1996); a community's giving up a language "completely in favor of another one" (Fasold, 1984); or, "speech communities whose native languages are threatened because their inter-generational continuity is proceeding negatively, with fewer and fewer users (speakers, readers, writers and even understanders) or uses every generation" (Fishman, 1991).

De Vries uses Weinreich's definition of language shift as his starting point: "the change from the habitual use of one language to that of another...." (Weinreich, 1952; de Vries, 1992) However, as De Vries points out, Weinreich's definition does not address the issue of choice. This is also true of many other definitions, not least because if reference is made to the speakers, the tendency is to define language shift in terms of societal entities (generations, communities), rather than in terms of individual speakers. However, as Fishman suggests, language shift rests in the non-use of the language by people. Fasold adds: "language maintenance and shift are the long-term, collective consequences of consistent patterns of language choice."

These patterns of language choice are not selected collectively, but on an individual basis, instance by instance. (The community is a group of individuals, and a living language exists only in the minds and mouths of the individuals who speak it.) Language shift, therefore, is the cumulative effect of a group of individuals increasingly using one language over another. For language maintenance, then, every individual has a contribution to make: the proficient speaker can use the language with less proficient speakers, the semi-speaker can become more proficient, the passive bilingual can become active. By using a definition of language shift which explicitly acknowledges the role of the individual in language shift and maintenance, individuals in any community may better understand the contribution they can make to language maintenance. Encouraging speakers to ask themselves, "What can I do differently?" could be a simple starting point.

In the case of the Arabic speakers in Hatay, the belief that the causes of language shift are out of their hands must be addressed. As noted above, a number of obstacles to pursuing a structured language maintenance program exist, yet the attitudes expressed by the Arabic community suggest receptiveness to language maintenance per se. It seems, then, that the safest approach is to encourage individuals to modify their speech behavior and use Arabic in preference to Turkish as much as is practical. Grandparents and other speakers of Arabic should be encouraged to use the language more with children. The frequent claim that speaking Arabic to children will be a long-term detriment to their acquisition of Turkish needs to be refuted. Some individuals may wish to explore the legal status of the language. Even the act of conducting research on the language -- asking speakers to reflect on the language -- may prompt individuals to think consciously about their attitudes toward Arabic and to consider ways of translating those attitudes into action. Individuals empowered with the knowledge that they have the ultimate "say" over what comes out of their mouths and how they can influence the future of the language need not be dependent on the government or the education system for support. The external causes of language shift may be out of their control, but the most crucial aspect of language maintenance, the speaking of it, is in their hands.

(1). The name Hatay was given at the time of the annexation; prior to this, it was known as the Sanjak of Alexandretta. Syria still lays claim to Hatay, a point which continues to cause diplomatic tension.

(2). This reality may have its roots in the fact that both the Ottoman Empire and the Turkish Republic place(d) an emphasis on classifying citizens according to religion. The term `minority' is only applied to non-Muslims. (Turkish Daily News Online, 9/11/00)

(3). Arabic is taught as a foreign language at university level elsewhere in Turkey; it is not available at Mustafa Kemal University, Hatay's only (state) university, even though it is less than a hundred kilometers from the Syrian border.

(4). This resonates with Schmidt's finding that speakers of Aboriginal languages may only belatedly recognize that their language is under threat. (Schmidt, 1990)

(5). Fasold's list of causes includes most of those suggested by the "realists": migration (both emigration by community members and immigration by outsiders); industrialization and other economic changes; school language and other government pressures; urbanization; higher prestige for the language to which shift was occurring; and a smaller population of speakers of the language from which shift was taking place (Fasold. 1984). Hyltenstam and Stroud's list is more extensive, factoring at societal, group, and individual levels. (Hyltenstam & Stroud, 1996)

(6). The Kohanga Reo (or "language nest") movement began in New Zealand as a grassroots language revitalization strategy in the 1980s. Community elders provide a M...ori-medium environment where preschool children acquire M...ori naturally. The approach has been adapted for the Hawaiian and Saami languages (see Ole Henrik and Skutnabb-Kangas, this issue).

(7). There does seem to be movement in this direction. Most recently, the head of Turkey's National Intelligence Agency stated that permitting Kurdish language broadcasts in the Southeast was a desirable way of counteracting pro-separatist broadcasts from outside Turkey. The Nationalist movement party, MHP, a member of the ruling coalition, is yet to be persuaded. (Parsons, 2000)


I would like to thank Lyle Campbell, Susan Foster-Cohen, and Eileen Moore Quinn for helpful comments made on earlier drafts of this paper. Thanks are due also to Misçel Kocamahhul for many unpaid hours of research assistance.

References and further reading

Courbage, Y. & Fargues, P. (1997). Christians and Jews under Islam. London: I.B. Tauris.

De Bot, K. (1996). Language loss. In Contact linguistics. Goebl, H., Nelde, P.H., Stary, Z. & Wölck, W., Eds. Berlin: Walter de Gruyter. Pp 579-585.

De Vries, J. (1992). Language maintenance and shift: problems of measurement. In Maintenance and loss of minority languages. Fase, W., Jaspaert, K. & Kroon, S., Eds. Amsterdam: John Benjamins. Pp 211-222.

Fasold, R. (1984). The Sociolinguistics of Society. Oxford: Basil Blackwell.

Fishman, J.A. (1991). Reversing Language Shift. Clevedon: Multilingual Matters.

Myers-Scotton, C. (1998). A Way to a Dusty Death: the Matrix Language Turnover

The Republic of Turkey (2000, April 26). The Constitution 1982 (English translation). 

Rumpf, C. (1989). Das Sprachenverbot in der Türkei unter besonderer Berücksichtigung inhrer völkerrechtlichen Verpflichtungen. Orient 30, pp 413-27.

Trimingham, J.S. (1979). Christianity Among the Arabs in Pre-Islamic Times. London: Longman.

Weinreich, U. (1952). Languages in Contact. Cambridge: Cambridge University Press.

Article copyright Cultural Survival, Inc.


Cultural Survival helps Indigenous Peoples around the world defend their lands, languages, and cultures as they deal with issues like the one you’ve just read about.

Learn More

To read about Cultural Survival’s work around the world, click here. To read more articles on the subject use our Search function and explore 40 years of information on Indigenous issues.

Do More

For ways to take action to help Indigenous communities, click here.


We take on governments and multinational corporations—and they always have more resources than we do—but with the help of people like you, we do win. Your contribution is crucial to that effort. Click here to do your part.