Markers of ethnic identity: focus on the

Malaysian Tamil community

IBTISAM M. H NAJI and MAYA KHEMLANI DAVID

Abstract

If members of a community feel that their ethnic language is threatened,

would they resort to other markers of ethnic identity like clothing, food,
celebration of religious and cultural festivals, etc.? This paper examines
the use of other markers of identity among the Malaysian Tamil community.
As women were the catalysts of language shift (David and Naji 2000) a
secondary aim of the paper was to determine if they were more inclined to
use these other markers of ethnic identity.

Background and aim of the study

The issue of the link between language and ethnicity is controversial. While
sociolinguists like Omar, in a study on Kuala Lumpurians (1991: 98) who
have shifted to English, demonstrate that there is no extrinsic correlation
between language and ethnicity, others like Fishman (1989: 5) assert that
there is.
Views of the language ethnicity connection therefore appear to vary.
While for some communities ethnic identity and language maintenance are
closely connected, for others the ethnic language may not form an impor-
tant part of their identity. David (1998) explains how the Sindhi language is
no longer a marker of ethnic identity for the Malaysian Sindhi community
and discusses other markers of identity.
This research attempts to investigate whether imminent language shift
has prompted the generation of Malaysian Tamils who are initiating the
move away from the ethnic language, Tamil, to move to other markers of
ethnic identity.

01652516/03/01610091 Intl. J. Soc. Lang. 161 (2003), pp. 91102

Walter de Gruyter

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
92 I. M. H. Naji and M. K David
Background to the setting

Malaysia is a multiethnic, multilingual country with a population of about

18 million people and at least a hundred languages. Malaysia comprises
two areas, Peninsular Malaysia (previously known as Malaya) and Sabah
and Sarawak. Malaysias multiracial population consists of Malays and
other indigenous people (61%), Chinese (28%), Indians (8%), and various
other ethnic groups (Khoo 1991: 40).
The Tamil community of Malaysia falls under the umbrella of Indians.
They make up the largest Indian community in Malaysia, as 85.9 percent
of the Indians in Peninsular Malaysia are Tamils. The other 14.1 percent
include Malayalees (3.2%), Telegus (2.43%), other Indians (6.77%), Sri
Lankan Tamils (1.6%), and Pakistanis (0.91%) (Khoo 1991).
A majority of the members of the Tamil community have traditionally
been rubber tappers and are members of the lower socioeconomic group in
Malaya. In fact, the Tamils were brought to Malaya by the British colonial
government to work in the rubber estates. More recently, many members
of this ethnic group have shifted to urban areas, and even in the urban
areas they occupy low-status jobs in factories. As for religious beliefs,
Malaysian Tamils practice a number of different faiths, which include
Hinduism, Christianity, and Islam. The majority are, however, Hindus.
This study focuses on identity markers of the Tamil Hindu community.

Language shift among the Malaysian Tamil community

In an earlier questionnaire study David and Naji (2000) showed that Tamil
is the dominant (i.e. most often used daily) language for most of the older-
age respondents but more than a quarter of the younger-age respondents
used a mix of English, Tamil, and Malay. More importantly, half of the
middle-age respondents had English as their dominant language. This
difference could be attributed to the fact that this middle age group had
had English as the medium of instruction. Both male and female respon-
dents stated they used more English than Tamil and tend to use a mix of
three languages. Based on these findings it could be inferred that language
shift is slowly but surely taking place in favor of either English or a mix of
three languages, especially for the middle age group.
Even though in the home domain the majority of the respondents across
all age groups often use Tamil, more of the younger respondents have
shifted away from the use of their ethnic language. While a large majority
(67%) of the respondents above sixty use Tamil with children, only 41.7
percent of the middle age group use Tamil. More men (66.7%) than women

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
 Ethnic identity in the Tamil community 93
(49.1%) use Tamil in the home domain, and the women tend to use English
or a mix of languages.
With regard to the language most often used with family members, the
vast majority of participants across all age groups (79.588.9%) use Tamil
with their mother. However, there were significant differences between
participants across age groups regarding the language used with the father.
More of the younger respondents communicate with their fathers in
English.
It could be concluded that the use of Tamil decreases with age. In other
words, there is a language shift among the middle age group in favor of
English or a mix of three languages, namely, Tamil, English, and Malay. In
terms of gender, more males (52%) than females (35%) tend to use Tamil
with their spouse. In short, this inclination to use English is more obvious
among females and the middle and younger age group, but by and large,
language shift at home is still fairly limited.
Tamil is also the language most often used with Tamil ethnic friends.
However, those above sixty reported higher levels of using the ethnic
language (Tamil) compared with the middle and younger group (55.6%,
35.7%, and 35.9% respectively). Again, the middle and younger age groups
used English and Tamil. This analysis gives the impression that there is a
shift of language across age with ethnic friends. It should be pointed out
that more males than females (54.5% and 28.1% respectively) use Tamil
with ethnic friends.
Both age and gender resulted in statistically significant differences (at the
0.05 level) among participants in some domains, for instance outside
the home. Broadly speaking, the younger generation and females have a
tendency to shift from Tamil to more English and a greater mix of the three
languages, English, Malay, and Tamil. This could perhaps be because the
younger and middle age group are more exposed to English and Malay.
In contrast to the language maintenance at home, Tamil is rarely used
outside the home. The majority of participants across all ages and genders
reported that they use either English or a mixture of three languages,
namely, English, Tamil, and Malay, when not at home. Forty-five percent
of the respondents in the middle age group do not use the ethnic language,
compared to 36 percent of the younger and 33 percent of the oldest age
groups. It is clear that there is a real shift outside the home from the ethnic
language in favor of either English or a mix of three languages.
It appears, then, that a shift from Tamil might occur in the future.
According to the criterion established for language shift, this is determined
by the movement away from the use of the mother tongue by respondents
from different age groups. In the light of this criterion the level of language
shift in certain domains, such as at home, with father and mother, and
with paternal and maternal grandparents, is relatively modest. By

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
94 I. M. H. Naji and M. K David
contrast, there is a startling shift with interlocutors like children
and spouse, ethnic friends, and in other domains such as outside the
home. These findings point to a language shift, which may not be radical
at the moment. The results support Gals (1979) thesis that younger
community members play the role of shift agent and are catalysts of com-
munity language use. The results also converge with studies by Milroy in
Belfast (1982) and Gal in Oberwat (1979), who argue that women are
the catalysts in such linguistic change. This is perhaps because the new
language they shift to is seen as a language of power.

Methodology and statistical treatment

In the face of imminent shift the researchers wanted to determine if

younger and female community members would latch onto other markers
of ethnic identity. The primary data for this analysis come from an eleven-
item questionnaire given to 90 respondents. The questionnaire elicited the
following details: (a) demographic data; (b) Indian food and traditional
Indian clothes; (c) lifestyle at home; (d) celebration of Indian festivals; and
(e) marriage partner. The data generated from the questionnaire was
processed using the Statistical Package for Social Sciences (SPSS-win).
Descriptive and inferential statistics are used in the analysis. Of the
descriptive statistics, simple frequency distributions and percentages were
applied to present the responses. For the inferential statistics, chi-square
(x2) was used.

Analysis of data

Demographic profile of the respondents

The respondents were 90 members of the Tamil community. Thirty-nine

(43.3%) respondents were youth whose ages ranged from 2024; the 3460
age group consisted of 42 participants (46.7%), while nine (10.0%) were in
the above-sixty category (Figure 1).
Thirty-three (36.7%) of the respondents were men and 57, or 63.3
percent, were women. The number of females was almost double that of
the male respondents. All the respondents reported that Tamil was their
mother tongue. A large majority of the respondents (86) were born in
Malaysia. Only two respondents were born in India and another two in
Singapore. The vast majority of the respondents (85, or 94.4%) had never
lived in India.

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
 Ethnic identity in the Tamil community 95

Figure 1. Respondents by age

Definition of ethnicity

For Fishman (1989: 216) ethnicity is concerned with origins and cultural
behavior. Ethnicity pertains to peopleness, that is, actions, views, or
attributions pertaining to and belonging to a people. A groups actions and
views are manifested through a number of symbols. These include food,
clothes, religion, customs, culture, and language.
The research investigated the use of a number of markers of identity.
These included the following:
a. dietary habits,
b. use of traditional clothes,
c. celebration of Indian festivals,
d. lifestyle,
e. choice of marriage partner.

Consuming Indian food

Respondents were asked to indicate how often they ate Indian food.
Table 1 shows that the vast majority (88.1100%) of respondents across all

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
96 I. M. H. Naji and M. K David
Table 1. The frequency of eating Indian food/wearing Indian clothes by age

Frequency Age group Chi-sq.

2024 3460 above 60
Indian food 1. all /most of 89.8 88.1 100.0
the time
5.0
2. half time 5.1 11.9 00.0
3. once in a while 5.1 00.0 00.0
Indian clothing 1. all/most of 46.2 47.6 33.3
the time
2.3
2. half time 35.9 33.3 44.4
3. once in a while 17.9 19.0 22.2

Table 2. Eating Indian food/wearing Indian clothing by gender

Frequency Age group Chi-sq.

male female
Indian food 1. all/most of the time 94.0 87.7
2. half time 3.0 10.5 1.8
3. once in a while 3.0 1.8
Indian clothing 1. all/most of the time 48.5 43.8
2. half time 33.3 36.8 .345
3. once in a while 18.2 19.3

the age groups frequently ate Indian food. This can be understood in the
light of the availability of Indian food in Malaysia: in schools, universities,
and work places, and at home. Although the young generation (2024)
have tasted Western and fast food, they still eat Indian food.
The effect of gender is shown in Table 2. The figures show that the
responses were independent of the influence of gender. Slightly more males
(94.0%) than females (87.7%) take Indian food. Chi-square (1.8) did not
reach the level of significance (0.05).

Wearing traditional clothes

With regard to wearing traditional clothes, Table 1 shows that across the
three age groups there is a shift away from the use of traditional Indian
clothes. Less than half of the respondents across all three age groups
(46.2, 47.6, and 33.3%, respectively) said that Tamil women wear
traditional clothes all or most of the time. About one-third of the
respondents from the young, middle, and old age groups (35.9, 33.3, and

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
 Ethnic identity in the Tamil community 97
44.4% respectively) reported that women wear traditional clothes half
of the time.
As the younger group is more exposed to other cultures, to the Western
media, and to interaction with other ethnic groups, it was expected that
more respondents in the younger age group might prefer to wear Western
clothes. However, the data indicated that more of the younger generation
(46.2%) are keen on wearing traditional clothes all/most of the time, as
compared to 33 percent of the older generation. This finding could reflect
the younger generations desire to maintain their ethnic identity (in the
face of losing another marker of identity, i.e. language).
Concerning the frequency of the use of Indian clothes, gender did
not result in statistically significant differences between males and females
(x2 = 0.345). About one-third of both the males and females (33.3
and 36.8% respectively) reported that the women in their families wear
traditional Indian clothes half of the time.
The Tamil community continues eating Indian food, but they shift from
the wearing of traditional Indian clothes. However, the younger respon-
dents reported higher levels of wearing traditional Indian clothes, a finding
that signifies a need on their part to emphasize their cultural identity.

Celebrating Indian festivals

The participants were also asked to determine the importance of celebrat-

ing Indian festivals regularly. There is a general agreement between the
three age groups with regard to the importance of celebrating annual
Indian festivals (Table 3). However, there were also remarkable differ-
ences. Again an overwhelming majority of the younger group (97.5%)
attached special importance to celebrating annual Indian festivals, while
only 73.8 percent of the middle age group and 66.6 percent of the old group
felt the same way. This could probably be because the younger generation,
who have shifted to English with their fathers in the home domain, feel a
greater need to emphasize their cultural identity in other ways. Chi-square

Table 3. Celebrating Indian festivals by age and gender

Age group Chi-sq. Gender group Chi-sq.

2024 3460 above 60 male female
1. Extremely important/ 97.5 73.8 66.6 81.8 84.2
Important 13.5* .63
2. Somewhat important 2.6 19.0 22.2 15.2 10.5
3. Not important 0.0 7.1 11.1 3.0 5.3

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
98 I. M. H. Naji and M. K David
(13.5) indicates large and statistically significant differences (at the 0.05
level) between the three age groups with regard to the importance of
celebrating annual Indian festivals.
In short, although it is generally assumed that the older generation
maintains the cultural values of a community, this study shows that the
younger age group emphasizes the celebrating of annual Indian festivals
more than the middle and older age group.
With regard to gender, the vast majority of participants of both sexes
(81.884.2%) reported that it is extremely important for them to
celebrate annual Indian festivals regularly. Chi-square (0.63) did not reach
the level of significance.

Lifestyle at home

This study was also concerned with the magnitude of change in the respon-
dents lifestyle; whether a traditional lifestyle has remained intact. The
figures in Table 4 reveal that nearly half of the respondents in the young
and old age groups live what they consider an Indian lifestyle. In addition,
a majority (73.8%) of the respondents in the middle age group view their
lifestyle as mostly Indian.
It appears again that the shift to using English with their fathers and the
use of a mixed discourse outside the home, indicating a shift away from the
ethnic language, Tamil, do not mean that the community has lost its sense
of ethnic distinctiveness. They still practice an Indian lifestyle. The fact that
the majority of respondents in the middle age group, who are the ones who
have most shifted to English, viewed their lifestyle as Indian does indicate
that other markers of ethnicity are replacing the language marker.
Although the respondents from the middle age group reported higher levels
of living an Indian lifestyle, chi-square value, though relatively high (8.5),
did not reach the level of significance of 0.05. Again, the responses were not
associated with gender. Though more males than females were inclined to

Table 4. Lifestyle at home by age and gender

Age group Chi-sq. Gender group Chi-sq.

2024 3460 Above 60 male female
1. Indian only 30.8 16.7 33.3 30.3 21.1
2. Mostly Indian 48.7 73.8 44.4 8.5 57.6 61.4 1.5
3. Equal I + M + Cha 20.5 7.1 22.2 12.1 15.8
4. Mostly Chinese 0.00 2.4 0.00 0.00 01.8

a. I = Indian, M = Malay, Ch = Chinese.

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
 Ethnic identity in the Tamil community 99
Table 5. Marriage partner

Age group Extent of importancea

Ex.I. + I. S.I. L.I. + N.I. Chi-sq.
From ethnic group 2024 97.5 2.6 0.00
3450 71.4 11.9 16.7 16.8
above 60 55.5 33.3 11.9
From your caste 2024 76.9 15.4 7.7
3450 35.7 16.7 47.7 23.4
above 60 22.2 33.3 44.4
Speaks your native 2024 43.6 33.3 17.9
language 3450 38.1 26.2 35.8 9.3
above 60 55.5 22.2 22.2

a. Ex.I. = extremely important; I = important; S.I. = somewhat important; L.I. = less

important; N.I. = not important.

live only an Indian lifestyle (30.3 and 21.1% respectively) chi-square (1.5)
was far below the 0.05 level of significance.

Marriage partner

Table 5 displays the data concerning the marriage partner by age. Fifty-
five and a half percent of the respondents above sixty were of the view that
the marriage partner should come from the same ethnic group. Again,
ironically, this is more important for the younger respondents. The
younger the respondents, the more they were concerned that the life/
marriage partner came from the same community. The differences
between different age groups are statistically significant at the 0.05 level
(chi-square = 16.8). The direction of the significance was in favor of the
young age group. The younger group even wanted the partner to be from
the same caste (chi-square is 23.4). This finding is not consistent with the
traditional notion that older members of a community are more intolerant
of exogamous marriages.
Table 6 shows that there is no difference between males and females as
far as the ethnicity of the marriage partner is concerned. A vast majority of
both sexes feel that it is extremely important or important for them
that the partner comes from the same ethnic group. This number falls when
it comes to caste and language. Males and females (36.4 and 26.4% respec-
tively) reported that it is less important or not important at all that the
partner has to be from the same caste. Similarly, 27.2 percent of males
and 26.3 percent of females reported that it is less important or not
important at all that the partner has to speak their native language. The

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
100 I. M. H. Naji and M. K David
Table 6. Marriage partner: gender

Age group Extent of importancea

Ex.I. + I. S.I. L.I. + N.I. Chi-Sq.
From ethnic group male 81.8 9.1 9.1
female 80.7 10.5 8.8 .801
From your caste male 45.5 18.2 36.4
female 56.1 17.5 26.4 2.7
Speaks your native male 45.5 24.2 27.2
language female 40.4 31.6 26.3 1.5

a. Ex.I. = extremely important; I. = important; S.I. = somewhat important; L.I. = less

important; N.I. = not important

differences between the respondents from different age groups are not
large or statistically significant (chi-square = 9.3).

Discussion

Tamil is the dominant language for three-quarters of the older age group
and slightly more than half of the younger generation. Almost 30 percent
of the younger group used a mix of English, Tamil, and Malay. Further-
more, for half of the middle age group, English had become their dominant
language. In the home domain, although the majority of the participants
across all age groups often use Tamil, the number of respondents across
all age groups who use Tamil in the home falls with younger respondents.
There were significant differences between respondents across age groups
regarding the language used with the father (and a relatively high percent-
age, 28.2%, of the younger group use English with their father). It should
also be pointed out that more men (66.7%) than women (49.1%) use Tamil
in the home domain, and that the women tend to use English or a mix of
languages. Outside the home there is a real shift away from the ethnic
language in favor of either English or a mix of three languages.
The use of Tamil decreases with age. The inclination to use English is
more obvious among females and the middle and younger age group. Both
age and gender resulted in statistically significant differences (at the 0.05
level) among respondents in some domains, for instance outside the home.
Broadly speaking, the younger respondents and the women have a
tendency to shift from Tamil to more English and also to a greater mix of
the three languages, English, Malay, and Tamil.
With this background of imminent language shift among the young and
middle age groups and between men and women, it was hypothesized that

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
 Ethnic identity in the Tamil community 101
these respondents would maintain other symbols of ethnicity as they were
losing their ethnic language.
When it came to dietary habits almost all the respondents across
the three age groups frequently eat Indian food. The data indicated that
more of the younger generation is keen on wearing traditional clothes
all/most of the time as compared to one-third of the older generation.
This finding could reflect the younger generations desire to maintain
their ethnic identity in the face of losing another marker of identity,
their language.
Again, an overwhelming majority of the younger group (97.5%)
attached special importance to celebrating annual Indian festivals, while
only 73.8 percent of the middle age group and 66.6 percent of the old
group felt the same way. Chi-square (13.5) indicates large and statistically
significant differences (at the 0.05 level) between the three age groups with
regard to the importance of celebrating annual Indian festivals.
A majority (73.8%) of respondents in the middle age group who were
shifting to English perceived their lifestyle as being mostly Indian. Further-
more, the younger the respondents, the more they were concerned that
the marriage partner be from the same community. The differences
between different age groups are statistically significant at the 0.05 level
(chi-square = 16.8).
In relation to all these markers of ethnic identity, there were no statisti-
cally significant differences between the male and female respondents. This
implies, therefore, that although the women may have initiated the shift to
English, there were no marked differences between the sexes in the mainte-
nance of other markers of identity and ethnicity. It must be emphasized
that those who shifted most away from the ethnic language held on more
strongly to other markers of ethnicity, like the wearing of Indian clothes,
celebrating of Indian festivals, and choice of life partner from within the
community.

Conclusion

Language is but one aspect of ethnic and cultural distinctiveness. Ethnicity

can still be maintained by the use of other symbols. For Malaysian Tamils,
ethnic identity (food and Indian lifestyle) remain strong although the
ethnic language is not widely used outside the home domain. For the
middle and younger groups, who appear to be shifting away most from
their ethnic language, other markers of identity like wearing Indian clothes,
celebrating Indian festivals, and contracting endogamous marriages have
become more important. These symbols of ethnicity are less important to

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM
102 I. M. H. Naji and M. K David
older community members. However, although women appear to be the
prime catalysts of language change, gender did not result in statistically
significant differences in the upholding of other markers of identity.

University of Malaya

References

David, M. K. (1996). Language shift among the Sindhis of Malaysia. Unpublished Ph.D.
thesis, University of Malaya.
(1998). Language shift, cultural maintenance and ethnic identity: a study of a minority
community. The Sindhis of Malaysia. International Journal of the Sociology of Language
130, 6776.
; and Naji, I. (2000). Do minorities have to abandon their languages? A case study of
Malaysian Tamil. International Scope Review 2(3), 115.
; and Noor, F. (1999). Language maintenance or language shift in the Portuguese
Settlement of Malaccca? Migracijske teme 4, 465482.
Fishman, J.A. (1989). Language and Ethnicity in Minority Sociolinguistic Perspective.
Clevedon: Multilingual Matters.
Gal, S. (1979). Language Shift: Social Determinants of Linguistic Change in Bilingual
Austria. New York: Academic Press.
Khoo, S.G. (1991). Population Census, vol. I. Kuala Lumpur: Department of Statistics.
Milroy, L.M. (1982). Social network and linguistics focussing. In Sociolinguistic Variation in
Speech Communities, Suzanne Romaine (ed.). London: Edward Arnold.
Nik Saffiah (1996). New Straits Times, 21 June 2000.
Omar, Asmah (1991). The Linguistic Scenery in Malaysia. Kuala Lumpur: Dewan Bahasa
dan Pustaka.

Brought to you by | Rice University

Authenticated
Download Date | 6/5/15 5:16 PM