Singapore Tamil

The purpose of this paper is to examine the position of Tamil as an ethnic minority and language in Malaysia and Singapore, and to draw some conclusions about the role of language planning/policy planning in the determination of linguistic outcomes, i.e. what happens as a result of (or even in spite of) the language policies in effect in the two polities. Tamils are the largest of the language groups that form the `Indian' minority in Malaysia and Singapore, constituting around 9% of the population, or 1.5 million in the former, and about 7% or 190,000, in Singapore.Within this number, people classified as Tamil-speaking amount to about 85% in Malaysia, and 65% in Singapore, or perhaps 120,000. Some people estimate only 60%, or 115,000 speakers. But in fact, with the declines in actual native speakers as evidenced by figures in the 1990 Census (see tables), what the actual Tamil population of Singapore might be is difficult to say with any accuracy. Most of the time, declaration of `Tamil' is a declaration of Tamil ethnicity, not linguistic habits. Below I will deal with the subject of the increasing number of people classified as Tamil who are not actually Tamil speakers.
In a recent compendium of articles on South Asian immigrants in Southeast Asia (Sandhu and Mani, (eds.) 1993) over half of the articles are devoted to the question of Indian communities in Malaysia---nineteen out of a total of 37, the rest being devoted to Brunei, Indonesia, Myanmar (Burma), the Philippines, Singapore and Thailand. All of them see the situation of Indians in Malaysia as somehow problematical,Contrast this with the articles on Singapore, where the future of Indians in Singapore is described as ``not without promise." (Sandhu 1993:787, op. cit.) . In fact the future of Indians in Singapore may be more secure than the languages spoken by them; it is not clear what would happen if all Singapore Indians were to become English speakers. whether it be the preferences given to Bumiputra Malaysians2 over immigrant Indians, the socio-economic conditions affecting plantation workers or the educational opportunities provided their children. In Singapore the situation is less dire, but language shift, especially among educated Tamilians, is proceeding apparently even faster than in Malaysia. I will try in this talk is to place the issue of Tamil language and language maintenance within the larger sociolinguistic milieu in Malaysia and Singapore, and see whether we can make a prognosis for the survival of Tamil, and indeed the survival of a Tamil-speaking minority, in Malaysia and Singapore in the twenty-first century.

Origins of the Tamil Community in the area.
The Tamil situation in Malaysia and Singapore must be seen in the context of an original colonial unity---after the Napoleonic Wars, Britain ceded many of its colonial `possessions' in insular Southeast Asia to the Netherlands in exchange for Dutch concessions in South Asia and South Africa. But Britain maintained a presence in the Straits of Malacca (Singapore, Penang, Malacca) as `trading posts' of the East India Company, and expanded from there into all of Malaya, and parts of Borneo. Tamils were brought to the area as indentured laborers to do agricultural work of various sorts, but eventually predominantly on rubber estates. They were drawn from two different segments of Tamil society---the workers were recruited from the most destitute landless laborers in Tamilnadu, while the overseers were recruited from educated, English-knowing graduates of Jaffna College, in what was then called Ceylon. It should be noted that the Jaffna Tamil spoken dialect is not mutually intelligible with Indian Tamil, though both share a diglossic `H' variety in Literary Tamil, and Jaffna Tamils learned to speak enough Indian Tamil to be able to communicate. These two groups were thus so different, both sociolinguistically and socio-economically that they never developed any notion of having common interests. Even today there is little intermarriage between their descendants, and it is the Sri Lanka-descended Tamils who are most urbanized and educated.In Sri Lanka itself, Jaffna Tamils have no common interests with so-called Indian Tamils, who were brought from India in the 19th century to work on tea plantations; the Sri Lankan census considers them to be different categories of people, so that despite an actual population of approximately 25% Tamils (Jaffna or Sri Lanka Tamils, Indian Tamils, and `Moors'), each group is treated differently, and sees no commonality with the other.

After World War II, Singapore joined the Federation of Malaysia but was `ejected' from it in 1965,Singapore rejected the Malayocentric view of Malaysia, since its population was predominantly Chinese in origin; in fact all of the larger cities in Malaysia, especially the coastal ones, have Chinese majority populations. In Singapore, the interpretation is that Singapore was `expelled' from Malaysia, while in Malaysia, Singapore is seen as having `withdrawn' from the Federation. so for three decades their language policies have diverged---Malaysia has moved toward a Malay-dominant policy, while Singapore enshrines Chinese, Malay and Tamil as languages given special rights (alongside English). Language policy in Malaysia is a topic that cannot be openly discussed without fear of being charged under the Sedition Act of 1948.The policy, as stated in the Constitution (Amendment) Act, 1971, is that the status of Malay as official and other languages as tolerated, ``may no longer be questioned, it being considered that such a sensitive issue should for ever be removed from the arena of public discussion." (Suffian bin Hashim, 1976:324). It is only one of those taboo issues (the place of Islam, the special status of Malays) that may not be discussed in Malaysia, for fear of disturbing certain ethnic sensibilities. Most of the writing on the topic of language policy, therefore, consists of filiopietistic articles extolling the virtues of the system, its natural fairness, its commitment to building up the national culture, and so forth. It can be described, but it cannot be criticized, so criticism of it only occurs outside the country. In Singapore, the language policy is openly discussed, and may be criticized, but rarely is, because it appears on the surface to be egalitarian, and therefore to not deserve any criticism.Singapore Tamils rarely criticize the language policy, because it seems so much fairer than Malaysia's policy; instead they lay the blame internally, at the feet of the Tamil teachers, the young people, their parents, the English language, the curriculum developers, or the kali yuga. Were they to assess the situation correctly, they would instead blame the housing policy.

My original research goal was to establish how the Tamils of Malaysia and Singapore were maintaining their language in the face of two differing policies, the former (Malaysia) emphasizing integration through Bahasa Malaysia and Islam, and the latter (Singapore), with a supposedly open, tolerant and `egalitarian' policy. Since the Tamils are known for their intense language loyalty back in their South Asian homeland, I was expecting to find that their love of the language and intense language maintenance efforts, manifested in India and Sri Lanka with strong opposition to Hindi, Sanskrit and EnglishThe current antipathy is strongest against Hindi and is known as Hindi etirppu; the opposition to Sanskrit was stronger several decades ago, and the opposition to English is mainly to English loan words being borrowed into Tamil ( angilak kalappu), not to English as an instrument or as a language per se. The opposition to Sanskrit has had the effect of ridding the written language of almost all traces of loan words from that language; in the spoken language, where no overt rules are prescribed, Hindi, Sanskrit, English, Portuguese and other loan words abound. would result in effective language maintenance in both contexts, but more so in Singapore, where Tamil actually has `rights'.

Cultural language-maintenance strategies
Tamil linguistic culture has a general approach to language maintenance, manifested in the homelands as extreme concern for hyperpuristic forms of the literary language. Sociologists of language call this corpus planning or corpus treatment; it is perceived by Tamils to be the most important kind of language maintenance, though in this day and age it may in fact have little relevance in contexts such as Malaysia and Singapore.

Language maintenance in Tamilnadu, and in contested Sri Lanka, also involves status management,1 and various measures have been undertaken to restrict the domains of Hindi, Sanskrit and English (in Tamilnadu), and Sinhala (in Sri Lanka) so that Tamil can recapture the domains of elementary and secondary education, the media, and so forth. This has been more successful in terms of keeping back Hindi and Sanskrit, but in the case of Sinhala, of course, language battles have degenerated into an overt civil war. In the case of English, which is perceived in some ways as a buffer against Hindi (and Sinhala) efforts are ambivalent, and many of those who decry angilak kalappu use English and even send their children to English-medium schools. The result is that English is still the main language of higher education in Tamilnadu; in Sri Lanka the battle to replace English with Sinhala, even in higher education, has been much more intense. In India, of course, the central government has no control over local educational policies, so no attempt to impose Hindi as a medium of instruction in Tamilnadu universities and colleges has ever been (or will ever be) attempted.

Statistics on Language Shift.
Unfortunately for my thesis, statistics on shift away from Tamil are difficult to come by, at least for Malaysia. The Malaysian census (e.g. Khoo 1983) does not enumerate its citizenry primarily by mother tongue, but by their ethnicity, so we must try to extrapolate from that. There are no tables in this most recent census enumerating how many Tamil speakers actually speak Tamil; we may read how many `Indians' speak (or are literate in) Tamil, or how many Tamil speakers also know Malay or English, but statistics that might indicate a shift from Tamil to either Malay or English are not available. Thus, knowledge of a particular language is treated as if it were immutable, i.e. one remains a mother-tongue speaker of one's ethnic-heritage language, but one may acquire knowledge of a second language. Bilingualism, in this view, is additive, never subtractive.

Statistics for Tamil school attendance are available. There seem to be 543 Tamil schools in Malaysia at last count, and 55% of Indians in Malaysia attend Tamil Schools (N.S. Rajendra, personal communication.) The remaining 45% would be either speakers of other Indian mother tongues (Punjabi, Malayalam, etc.) or are Tamils attending other kinds of schools, presumably either Malay-medium or English-medium. The census of Malaysia (cf. Khoo 1983) tends to enumerate for ethnic group (`kumpulan etnik') rather than for mother tongue, so there are few breakdowns among the Indian ethnic group as to mother tongue.

In this, we can learn from the tables that Malays are the least bilingual of all Malaysia's ethnic groups, and that Indians are the most bilingual: in Table 5.6 (1980:544) we can see that among urban Tamils aged 10 and over 69/ know English (205,459 out of 295,717) and almost half speak fluent Malay. In rural areas, the knowledge of English is 27%, while knowledge of fluent Malay is 31%. (The combined urban/rural percentage would be 44%) But we can also see from literacy tables that the percentage of Tamils literate in English has declined from 78% to 70% between 1970 and 1980, while literacy in Malay (for Tamils) has increased from 36% to 57%. But literacy appears to differ from ability to speak a language, because from Table 6.14 (1980 census) we learn that the percentage of the Indian population aged 10 and over who are able to converse in selected languages has increased: 89% of Indians could speak Tamil in 1970, increasing slightly to 90% in 1980; 26% of Indians could speak English in 1970, rising to 40% in 1980, and 50% of Indians could converse in Malay in 1970, but 86% in 1980.Note that these are figures for the Indian population at large; the Tamil population thus exceeds the general Indian group in knowledge of English by 4%; or, there are perhaps fewer people who speak English than are literate in it. This is conceivable, but this is not the usual declaration.

In Singapore, figures are more forthright. The Singapore Census of 1990 gives prominent space (Lau 1993:6) to several tables showing the increase of English or other languages at the expense of Tamil, Chinese dialects, etc. Table 6, e.g., gives the following for Indian Households:

Figure: Resident Private Households by Ethnic Group of Head of Household and Predominant Household Language Spoken, 1980 and 1990. (Source: Singapore Census, Lau 1993:6)

1980 1990
English 24.3 34.4
Malay 8.6 14.1
Tamil 52.2 43.5
Others 14.9 8.1

For Chinese households, English has increased from 10.1% to 21.4% in the same time period, Mandarin has increased from 13.5% to 30.0%, while Chinese dialects have declined from 76.2% to 48.2%. In another Table, focusing on language that children speak to their parents, the change is less dramatic, but nevertheless parallel.

Figure: Persons Aged 5 Years and Over in Resident Private Households by Ethnic Group and Language/Dialect Spoken to Parents, 1980 and 1990. (Source: Singapore Census, Lau 1993:7)

1980 1990
English 16.9 25.9
Tamil 56.8 47.1
Malay 9.4 16.7
Others 16.0 10.3

We must remember in studying figures about Tamil in Indian households that Tamils constitute only about 65% of Indian households anyway, with other languages such as Malayalam, Panjabi, and others making up the remaining original 40%. Thus a drop to 47% is only a drop from a maximum `original' 65%, but nevertheless the increase in English (and Malay!) shows that this rise is also at the expense of other Indian languages.

In addition to such primary statistical reports, we also know that there is a problematical shift away from Tamil from certain secondary indices, such as sales figures (from which is extrapolated readership) of the one daily Tamil newspaper, by public discussion of the issue among Tamils, usually the older segments of the population, but also from reports of Tamil teachers, who are faced with the problem of teaching the `mother tongue' to people whose mother tongue is in many cases actually English. `Public' discussion of this takes the form of reports by educators in such fora as Singapore, the Year in Review (Gopinathan 1991) where a discussion of `equilingualism' (Singapore's term for egalitarian bilingualism as a policy and as a proposed outcome) can be foregrounded. Here it is freely admitted that knowledge of `mother-tongue' is often weak, i.e. second to English, so that a reform proposed in 1991 might take the form of

retaining the present allocation of 27% [curriculum time] for mother-tongue and moral education [which takes place in the mother tongue]; schools will be given flexibility and latitude in the allocation of curriculum time according to the needs and abilities of their pupils, especially taking into account the language(s) they use at home[.] (Gopinathan 1992:56)

This is designed to deal with the fact that some groups, e.g. Indians, are weak in the mother-tongue, but aren't particularly worried about it, while other groups (e.g. Chinese) feel that the system penalizes knowledge of Chinese, and want instead to strengthen it.Readership of the daily tami muracu hovers around the 11,000 mark; this is only about 10% of the reported or suspected Tamil population in Singapore, but it is not much less than the number of Tamils (see tables) who speak Tamil according to one Census index. The Singapore Census (cf. e.g. 1990 Census) has many tables reporting such things as language spoken by ever-married persons to grandchildren, to spouses, by occupation, by size of household, etc. but no tables indicating non-reciprocal language use, e.g. Tamils speaking to their (grand)children in Tamil, but (grand)children responding in English. (Indeed, it would useful to know how many Chinese have become mother-tongue speakers of Mandarin in the last three decades, and only understand Chinese dialects passively, or perhaps not at all.) Thus, e.g. Table 70 (p. 137) `Students aged 5 years and over* in resident private households by language(s) spoken to household members and ethnic group of head of household' lists, under `one language spoken', a total for Tamil of 12,157 and 12,125 of Indians, by which is meant that the language spoken to household members of Indian families is Tamil 12,125 out of 12,157; in the remaining Indian families the language would be Chinese or Malay (sic!). But English in such situations is not given. Whenever English is listed in these tables (and it is frequently listed) it is always presented as something isolated from the other languages, and not something resulting from language shift from the other languages. In other words, if English is spoken, the assumption is that those speakers are not Indian; only Tamil and other Indian-language children are thus counted as Indian in these statistics. We can see statistics (e.g. Table 87, Language Spoken to Grandchildren, p. 155) that show that when Indian languages are spoken, more females speak Tamil to their grandchildren than do males, that is, of 3,858 Indian persons using Tamil to their grandchildren, 1,348 are male but 2,510 are female. But we do not see how many people whose mother tongue is Tamil speak Tamil to their grandchildren, or (what is more important) how many people whose mother tongue is Tamil speak English to their grandchildren. We must compare the figures in Tables 87 and 70 and hope that the difference between 12,157 and 3,858 is in fact an index of non-use of Tamil. The only table that perhaps indicates some of this is Table 93, which tries to correlate ethnic group of household head and predominant household language. In this table we see that of a total of 41,907 Indians, 14,363 or 34%, have English as the predominant household language; of 21,481 households that have an Indian language as the predominant household language, 18,215 or 84% (higher than the usual percentage, i.e. 65% of Tamil speakers within the Indian population) use Tamil. But there is still no index of non-reciprocal language use; the census can only recognize `predominant' language use, not isolate code-switching or one person/one language use.
In Malaysia (and in Singapore), language policy is not set by the Tamils, and Tamils are therefore in the position that Telugu speakers or Kannada speakers are in Tamilnadu:One of the great drawbacks of Indian language policy is the weakness of provisions for language groups living in linguistic states where they constitute a minority. It is fine to be a Telugu speaker in Andhra Pradesh; it is not so fine to be one in Kerala, Karnataka or Tamilnadu, and the constitutional provisions to protect such groups are noticeably without teeth. Each linguistic state, having driven out the perceived oppressor and established its own linguistic regime, turns out to be an even more ferocious oppressor of its own linguistic minority groups. they are a tiny minority, have no say in overall policy formulation, and are suffered to maintain their languages only for the purposes that the majority culture deems necessary.

Educational Policy in Malaysia
A word of background is in order here on what languages may be used in education in Malaysia. In ``National Schools" Malay is the medium of elementary education; Tamil and/or Chinese may be taught if there are fifteen students who petition for it. Otherwise, Tamil and Chinese medium ``National-type" Schools may exist, and they receive varying degrees of government support; Chinese schools tend to reject total subvention, in order to maintain more control. At the secondary level, Malay medium is the only publicly-supported schooling available. Privately-supported Chinese schools do exist, but there are none for Tamil---the Tamil community can not afford this luxury. Again, at the secondary level, Tamil and Chinese may be taught as a subject if fifteen students request it. The Malaysian constitution provides guarantees for the use of these languages in the above ``unofficial contexts", i.e. they are officially tolerated (also some use in broadcasting, the Department of Indian Studies at Universiti Malaya, and support for teacher training) but this official tolerance is thought of as unofficial since only Malay may be official.

Educational Policy in Singapore
In Singapore, the situation is more complex, with Tamil (and Malay and Mandarin Chinese) offered throughout pre-University education, but in the case of Tamil, at increasingly hyperpuristic levels. The system is referred to as a bilingual one, but in actuality it is an English-medium school system with large amounts of `mother tongue' instruction in the elementary grades. At higher levels, the amount of Tamil diminishes, and it is also decreasingly `practical' i.e. hyperpuristic to an extent not even found in Tamilnadu. The result is an increased alienation from Tamil, since the hyperpuristic version they are presented with is given no support in any other domain of Singapore life, with the exception of Tamil news broadcasts on SBC television. In other programming, and in radio, more colloquial versions of the language hold sway. Jeganathan puts it this way:

Tamil is taught as a second language for the Tamil pupils. Most of these pupils speak English with their friends and even at home with their English-educated parents. Only where there are grandparents who are not English educated might these pupils speak Tamil. ...
The Tamil language taught in schools, therefore, appears to be more for evaluation purposes ... ... This has therefore led to a concern for accuracy work ... [i.e.] accurate use of the language in the classroom [rather] than ... outside the classroom. (Jeganathan 1993:80-81).

Since Tamil has no economic value in consumer-oriented Singapore, it is seen by most Tamils as having no value at all, and many, after passing their Cambridge A-level exams (necessary for entry into higher education) never use it ever again.The question many Tamils ask themselves is, tami cooru poodumaa? `Will Tamil put rice (on the table)?'

Language-maintenance strategies that work.
The sociologist of language Heinz Kloss provides a list of language-maintenance strategies that enhance or hinder language maintenance by minority groups in immigrant societies like the United States (Kloss 1966, in Fishman 1966). One of the factors that enhances language maintenance is ``pre-immigration experience with language maintenance", particularly in dealing with linguistic suppression in the form of underground resistance to education in another language medium, self-help language schools, etc. Groups that were already ready to cope with language maintenance in their home country because of suppression there, such as the Poles under Czarist Russian rule, were more able to ``hit the ground running", as it were, perhaps because the notions that widespread community involvement was important, everyone had to participate in order to make it work, everyone had to be eternally vigilant, etc., were accepted.

Tamils who came to Malaya, the Straits Settlements, and other parts of Southeast Asia brought strategies with them that were developed in their home country, and at first these strategies seemed to work. Essentially these strategies were:

All schooling through elementary levels should be in Tamil only.

The kind of Tamil needed was whatever was being developed in India and no adaptations or compromises to local conditions were necessary, or even permissible.

English could be admitted at the higher levels and would in fact be quite useful in the new environment.

Any other local languages that were useful or necessary (e.g. Malay) could also be acquired for auxiliary use, but should not be given first or even second priority.

In the plantation economy of nineteenth century Malaya and the Straits Settlements, these strategies worked quite well. Most Tamils of the period came with the intention of returning to India at some point; British education favored Malay only, and no schools in other languages were supported by the colonial power. Tamils of the more educated classes (most, as we have noted, were from Sri Lanka) worked as clerks and overseers, and their knowledge of English was an advantage to them in this situation, since neither the Malays nor Chinese seemed to want to require English education at that time. Plantation Tamils did learn some Malay, enough to get around and do their work, but in few cases if any did it actually supplant Tamil.Only the very anciently-settled and assimilated Chitty Tamil community in Melaka had become Malay speakers; more recently arrived Tamils did not. A great cultural barrier for most Tamils, though not all, was Islam, which served to isolate and contain them.

Again a strategy brought from India (keep clear of Islam) helped maintain the linguistic isolation. Knowledge of Tamil was necessary to be a good Hindu; it would not constitute a path to Islam. The few Tamil Muslims that came at this period were indeed in a different kind of situation, and assimilation through intermarriage of Indian Muslims and Malays did occur, but this happened mostly between North Indian Muslims and Malays, not Tamils. For better or worse, Tamil Muslims tended to remain solidary with Tamil non-Muslims, and cooperated with them in language maintenance.

Though Tamils thought that the strategies delineated above would serve them well in Malaya (since they would eventually return to India where these were the priorities), these strategies have become increasingly problematical after independence and under the threat of Malaysia's very stringent language policy. It was at this point that Indians had to decide whether they would remain in Malaysia (and Singapore) or return to India. Many, having suffered the depradations of the war years, and inspired by the fact of Indian independence, returned to India. Those that remained had to decide how they would fit into the new situation, where neither their knowledge of English (their previously strong suit) or knowledge of Tamil (their weak suit) would now be particularly useful, especially in Malaysia.

It seems clear that the strategies brought from India have not been adapted, in fact may not be adaptable, to the current environment, and are not serving the cause of Tamil language maintenance. In fact, as already noted, Tamils now have a hard time even justifying maintaining Tamil, since in most cases it seems economically disadvantageous; the only reason proponents can present is the `primordial' one, i.e. `we should love Tamil because it is beautiful'.

Other factors.
But this is not the whole story. Another all-pervasive and inescapable fact about Indian communities in Southeast Asia, and especially in Malaya, is the fragmentary and disunited nature of the community (Marimuttu 1993). This is manifested in different ways:

Indians come from a number of different parts of the subcontinent, and do not all speak the same language.

Indians are settled in many different localities, often separated from and in isolation from others of their own group. (The rubber plantations are separated from each other, dotted across the landscape.)

Within any given language group, there are the usual splits involving caste, religion, and class. Even if all Tamils were concentrated in one area, there would be differences that are perceived as unbridgeable. The gulf between Sri Lanka Tamils (Rajakrishnan 1993), who acted as overseers and clerks, and laborer Tamils (from India) was vast.

This fragmentation and segmentation has remained until the present time, and underlies many of the current problems facing the Indian community in Malaysia and Singapore. As far as Tamils are concerned, it works against language maintenance in a number of important ways, and combined with the inadequate and inappropriate language maintenance strategies brought from India, is now taking its toll on the Tamil language.

Language Shift.
If language maintenance does not occur, there can be several results. One is language death; speakers become bilingual, younger speakers become dominant in another language, and the language is said to die. The speakers or the community does not die, of course, they just become a subset of speakers of another language. The end result is language shift for the population, and if the language isn't spoken elsewhere, it dies. In the case of Tamil in Malaysia, we do not speak of death because Tamil continues to live on in Tamilnadu, but the effect is the same. For the speakers who go to their death as Tamils still, it is a kind of death to see their children shift to another language.

In Malaysia and Singapore, if Tamils shift languages, there are two possible outcomes. One is that they will become Malay speakers, the other is to become English speakers. (Chinese is not a practical outcome.) In fact, few Tamils are becoming Malay speakers, except for individual Tamil Muslims who intermarry with Malays and whose offspring grow up speaking Malay. The more general outcome is that many Tamils, especially well-educated Tamils, are becoming English speakers. Less-educated Tamils, however, especially those still living in plantation communities (in Malaysia) and/or employed in menial jobs (Singapore), continue to speak Tamil, and the prognosis for their language maintenance is for the time being favorable.

In Malaysia, there are a number of reasons why English-educated Tamils are in fact switching to English as a dominant language, and there is no one reason that is more important than others. There is a tendency in the Tamil community to lay the blame for this shift at someone else's door, but neither the government's language policy, nor the Tamil community itself, nor the difficulty of maintaining a Tamil-maintenance infrastructure, nor any other reason is sufficient alone. In fact Tamil is doing fine when the conditions that enhance language maintenance pertain, and these are precisely those enumerated by Kloss for German immigrants in the US:Note that Kloss's fifteen factors contain six positive factors, and nine ambivalent factors; in the current case, factor 1 is unambiguously positive, while 2--4 are ambivalent, i.e. they can work either way. In the Malaysian case, combined with factor 1, they are positive in terms of maintenance.

Isolation and linguistic islands:
Low educational background and aspirations
Small size
Great cultural difference (including religion) between group and majority.

The plantation economy, where most of the work in rubber and palm oil tapping is performed by Tamil and other Indian workers, provides a perfect isolated environment in which Tamil can be maintained. No other language intrudes, no other language is necessary, and monolingual Tamils can live in splendid (though economically disadvantaged) isolation. As we noted, Tamil is admissible as a medium of education for elementary education in Malaysia,Malay is the medium of ``National Schools" and Chinese and Tamil are tolerated as the medium of ``National-type Schools", but English is not tolerated for state-supported education. Private schools using English do exist, and private Chinese medium secondary-schools also exist, but they do not receive any state support. and this is provided to the children of the plantation communities. Because of the segmented nature of Indian society and its perpetuation in emigration, the kind of workersThis point may not be emphasized too strongly: Indian plantation workers, mainly Tamils, came from the most destitute, impoverished and lowest-caste, including `untouchable', backgrounds. They were already socialized to be docile, servile and unquestioning of authority, and the colonial plantation capitalized on these attitudes and perpetuated them. Indian workers were praised again and again for their docility and willingness to put up with the most abject conditions, compared with the Chinese, who were rebellious, entrepreneurial, and uncooperative with the plantation system. who came to do this kind of work tend to not have much of an educational background and/or aspirations for anything more. Unlike the educated (Sri Lanka) Tamils who worked as clerks and teachers, knew English, and rose to become a professional urbanized elite, these Tamils never had educational opportunities beyond the rudimentary estate schools, and despite being able theoretically to go on to secondary education and higher education, they persist in not aspiring to do so. Their elementary education in Tamil suffices them, and since these small pockets of Tamil speakers were (until recently) always located in isolated rural areas, are perceived as no threat to Malaysian society.

Given the religious differences (Hinduism vs. Islam), plantation Tamils other than Muslim Tamils are unlikely to ever `merge' with Malay society, either linguistically or culturally. Marimuttu (in Sandhu and Mani 1993) claims that the educational system provided to the plantation Tamils does not raise them out of the cultural dead-end they are stuck in, and is not designed to do so. This system, according to Marimuttu, preserves and perpetuates the plantation system in a kind of neocolonial atmosphere. Tamil education is therefore a kind of cocoon that helps perpetuate their isolation. As such we can imagine that the Tamil language will be maintained in this environment for the foreseeable future; as long as there is rubber tapping and palm-oil cultivation, the same population is bound to continue to do that work, since Malays do not perform this work, and Chinese are primarily urbanized and in business.There is some movement out of the plantation economy into urban areas, but neither the schools nor the ``profession" of rubber-tapping provide people with salable skills in the city. Those who do leave are now being replaced to some extent by foreign (Bangladeshi and Indonesian) contract labor. Another reason for little social movement is that there has been no practical way to mechanize tapping, so there is no way to increase productivity, and wage levels; individual workers must still go to the trees and tap them.

The situation of the urbanized educated Tamilian, however, is a different one, both in Malaysia and in Singapore; this is not surprising given that Singapore has no rural environments. Here we see in operation a number of other factors that work against language maintenance. One is the pervasive segmented character of Indian culture, and Indian communities abroad. One can discern linguistic differences, caste differences, and differences of village and even `national' origin, i.e. whether Tamils came from India or Sri Lanka. Tamils (and other Indians) in the urban environment are perhaps even more segmented than are rural tapper communities, so the urge to work together on language maintenance is weak. Just like Germans of different backgrounds in the nineteenth century US, Tamils of various backgrounds do not see themselves as having any interests in common with other Tamils, or at least not enough to lay aside these differences until it is too late. Secondly the housing policy in Singapore works against language maintenance. Around 88% of Singapore's population lives in Housing Development Board (HDB) estates, and the housing policy requires a strict maintenance of the overall percentages of different ethnic groups in the general population to be present in each housing estate (77% Chinese, 14% Malay, 7% Indian). This results in Indians (thus Tamils) dispersed in housing, never to exceed 7%, i.e., with no concentration of speakers anywhere. Thus no territory in Singapore belongs to Tamil.See Ling et al. (1993) for a detailed description of this policy.

Secondly, the aforementioned language maintenance strategies brought from India turn out, in post-colonial Malaysia and Singapore, to be counterproductive. An emphasis on keeping Tamil pure of Hindi, Sanskrit and English influences is rather futile when the language of threat is Malay. And in English-dominant Singapore, only English is seen as having practical value, and only English-knowing professions have prestige. But it is the emphasis on corpus work rather than status concerns that is counterproductive. It is not the corpus of Malay (or Hindi, or Sanskrit or English) that is the problem here, it is the status of Malay within the national language policy (of Malaysia) that is a problem, but the other issue is that the status of English in this equation is also conflicted.

That is, this urban group had an original `leg-up' in colonial Malaya because of their knowledge of English, and used that advantage, and still uses it, despite obstacles from the official policy, for their own benefit. But in another sense, the status of English is a danger, since this group of Tamilians, and indeed Tamilians everywhere, have not treated the status of English as problematical.They object to mixing Tamil and English, angilak kalappu, but they do not object to anyone knowing English. They have embraced English, and continue to embrace it, as a barrier or buffer against Hindi, Sinhala, and Malay, and as a passport to a good job. The problem now is this group has relaxed its guard about English, and too much knowledge of English now means that this group now knows too little Tamil, and is in fact not committed enough to Tamil. In fact, many of my informants, though committed to Tamil, even professionally (University teaching, Ministry of Education) declared that they would not put their children in Tamil schools (in Malaysia) because Tamil schools are a dead-end professionally and socially. And in Singapore, it is hard to find children of educated Tamil Singaporeans who can carry on a conversation in Tamil.

A New Factor: Urban Squatter Settlements.
The previously described situation, in which Tamil communities either consisted of rural estate workers or urbanized middle-class Tamils is now complicated by a third factor, the urban squatter settlement populated by Tamils who have left the plantations and are now working in various low-paying jobs in urban areas such as Kuala Lumpur, Johore Baru, and Penang. They congregate in squatter settlements (Rajoo 1993, in Sandhu and Mani 1993) and send their children to Tamil medium schools.In at least one case I know of, the urban area has come to the plantation---the Kuala Lumpur megalopolis has sprawled out into Selangor State to engulf former plantation land, which has been converted into luxury single-family housing, but the Tamil school (and a squatter zone) continue to exist, cheek by jowl with the fancy housing. Such schools persist in their substandard conditions, despite their status as ``National-type" schools, which should receive state subsidies, but are provided with very little other than teachers' salaries. For whatever reasons (perhaps as a buffer against the shock of urban life?) it is that these communities still choose Tamil medium schooling, the general overall economic and cultural destitution of these groups means that Tamil medium prepares them for nothing but the substandard conditions they have always had---they work at part-time jobs, in entry-level factory jobs, as messengers and sweepers, and have the highest rate of single-parent families, alcoholism, crime, prostitution and all the other social evils of the modern urban underclass. One Tamil stated to me that it appeared that the Tamils are and will always be the ``niggers" of Malaysia; he saw no way for these Tamils to break out of this cycle and move up the socio-economic ladder. Those that manage to do so, by attendance at National Schools, will leave the Tamil language behind. In his view, Tamil will only survive in Malaysia if Tamils remain poor and at the lowest level of society.

We therefore now must contend with two language strategies employed by the Tamil ``community" in Malaysia. One continues to prefer Tamil schooling; the other abjures Tamil schooling and is economically motivated to prefer Malay and English; for the latter group Tamil may remain as a home language, but in many cases not even this happens. This is not to point the finger; this strategy, of embracing English to the detriment of Tamil, is in fact a survival mechanism engendered by the national language policy. Several elements of that policy conspire to cause this:

Admission to higher education is controlled by ethnic quotas, and seats are reserved on an ethnic basis. If certain ethnic groups do not use their seats, they are not relinquished to another group, they are simply not filled. The group that is not filling its quota is the bumiputra group. Indians and Chinese who would otherwise be qualified for these seats must go abroad for higher education.Consider a comparison with a transportation model: imagine an airline that assigned seats on its flights by ethnic quota; if certain seats are not filled because not enough members of a certain ethnic group made reservations, the flight leaves with empty seats, and members of other ethnicities are obliged to travel by some other mode of transportation.

Since it cannot be determined in advance who will be admitted and who will not, students must plan for the eventuality of expatriation in order to get higher education. Planning for expatriation means English proficiency must be high. In Singapore this strategy is almost universal; seeking education in Australia or other English-speaking countries gives people a leg up on better jobs, and many simply do not return to permanent employment in Singapore. Singapore Indians have the highest rate of emigration of all its ethnic groups.

Students who go abroad for education often do not return, but obtain jobs elsewhere.. The cost of this `brain drain' for Malaysia is immense, since, whatever else anyone cares about who gets educated, a tremendous amount of foreign exchange is leaving the country to finance this drain, and if the students do not return (and why indeed should they?) the cost of their education is lost to Malaysia.

Students who otherwise might want to return to Malaysia to work have other barriers to face. One is quotas for certain jobs; another is barriers to degree-holders from certain countries. The general atmosphere is one of not being wanted. In face of this, the strategy of planned expatriation via English is not hard to understand.

This strategy of course colludes with other strategies mentioned above, such as the predilection of (educated) Tamilians to learn English, the strategy of maintaining a puristic Tamil which has no economic value, and is therefore perceived as useless, and the strategy of non-cooperation with other similar groups.

Summary and Conclusions:
The Tamil language is still alive in Malaysia and Singapore and will probably survive into the twenty-first century, perhaps only in isolated rural pockets, or as the language of a marginalized urban underclass. When all is said and done, it is less the overt language policies (as enshrined in the Malaysian Constitution, and in Singapore educational policy) that will determine this outcome, than the socio-economic history and present conditions of the Tamil communities there. Tamil has no economic value in the area, and is therefore maintained by the socio-economically destitute only as a last vestige of primordial ethnicity. The two language policies thus seem paradoxical---the Malaysian policy penalizes Tamil maintenance, but Tamil survives in spite of it; the Singapore policy rewards Tamil maintenance, but speakers are shifting away, apparently in spite of it. In fact, however, we must see the overt policies as almost irrelevant; if there is a paradox, it is in the social conditions and social history of the South Asian communities in the area. One might almost say that Tamils brought a language policy with them from South Asia,

Since even in the developed western countries (e.g. the US) a similarly destitute urban underclass persists, and continues to maintain its own variety of English despite teachers' attempts to extirpate it, the prognosis for Tamil is unlikely to be any different in Malaysia.

Whither egalitarian language policies?

In the media and in lay discussions of language policy, one can often hear a call for ``egalitarianism" in language policy, the logic being that if right-thinking people prefer ``equal opportunity" in employment for all citizens, regardless of race, creed, ethnicity, or gender, etc., then language policies also ought to be egalitarian. But it seems clear to me that we have plenty of examples of egalitarian language policies that do not result necessarily in equal outcomes. In other words, the intent of policies may be egalitarian, but they may not in effect be equal. Or, to put it another way, if the goal is equality of outcomes, egalitarian policies may not necessarily be the effective `way to go'.

Finland's language policy seems egalitarian, at least as it is described in the literature, but it does not result in Swedish holding its own with Finnish; Swedish is losing speakers rapidly, and its speakership is aging rapidly, as young speakers switch to Finnish, or emigrate. Swedish has some historic territories (Jakobstad, other coastal areas, islands) associated with it, but this may not suffice into the future.

French in Canada, especially in Quebec, is also threatened by an ocean of English speakers, both in Canada and in the US. Attempts to control domains for French only result in an outcry from anglophone Canadians, but if Quebec were to allow equality of language, (i.e. egalitarianism) the result would be to the advantage of English. Quebecois see territorial guarantees as more important for the survival of French than egalitarianism, because they know that egalitarianism alone won't work.

Switzerland, with strict territoriality as a principle underlying rights of various language groups, seems to maintain a balance of outcomes, at least for German and French, but Italian and Romansch are definitely disadvantaged, and Romansch is losing speakers.

Other polities might be brought in as further examples, but it seems to be the case that a combination of factors, such as egalitarianism and territorial rights, may help to maintain a language, but if this is not combined with a critical mass of speakers (who knows what this is, but 7% probably doesn't qualify) and certainly other important factors (economic, demographic), egalitarian policies alone will not do much. I realize it is fashionable to address rights issues through the courts these days, especially in the US, in an attempt to get ``equal" rights enshrined in laws and constitutions etc. But juridical solutions alone will probably not suffice to actually change outcomes (though this does not stop people from trying, e.g. for minority languages in the US and for English, e.g English-only laws). Both approaches, I think, are doomed to failure, if we ignore demographics and other issues (Schiffman, forthcoming.)