On Boundary Marking
On Boundary Marking
Abstract and Keywords
This chapter examines the extent to which the idea of a right is dependent on the notion of a boundary. For the concept of a right to be coherent, clear boundaries have to be drawn between different rights bearers. Demarcations also have to be made between the objects of rights, that is, the ‘things’ that rights bearers are entitled to. Such considerations lead us to ask whether language can or should be treated as a bounded entity, as would be the case with the concept of language rights.
This chapter focuses on the idea of language being amenable to boundary marking. While acknowledging that the notion of a boundary is indeed necessary when talking about rights, the chapter questions whether this is an appropriate move when it comes to language. The chapter begins by explaining why boundary marking and essentialism are inescapable features of the discourse of rights. After identifying three effects of rights discourse (selectivity, reinvention, and neutralization), the chapter concludes by observing that language neutrality is chimerical, and furthermore, that the unavoidability and hybridity of language pose significant challenges for the notion of language rights.
The Bearer of a Right, the Object of a Right
The idea of a right involves both a bearer of the right as well as the object of the right, the latter being that to which the bearer is due or that to which the bearer can lay claim. In the simplest cases, the bearer of the right is an individual and the object of the right is something that is individuated by virtue of being solely associated with the bearer. As Dershowitz (2004: 15) defines it, a right is
(p.22) This is a relatively unproblematic interpretation of a right because when both the bearer and the object of a right are conflated within an individual, these automatically acquire a boundary marking or delimitation by virtue of the fact that each individual is physically demarcated from other individuals.
Something that is due to a person by just claim, legal guarantee, or moral principle … A power, privilege, or immunity secured to a person by law … A legally enforceable claim that another will do or will not do a given act; a recognized and protected interest, the violation of which is a wrong.
This demarcation is critical because it allows us to determine when a particular bearer has had his or her rights respected or violated. For example, while everyone has the right to freedom from torture, it is extremely important not to forget that one person’s right to freedom from torture is not exchangeable with someone else’s. Each individual, depending on the circumstances that he or she finds himself or herself in, may thus experience a violation or observance of his or her particular right not to be tortured. In this way, the ability to distinguish the boundaries that individuate each rights-bearer and his or her respective object allows us to evaluate claims regarding equality of treatment: the possibility of making sensible comparisons depends on consistently distinguishing between the entities being compared. This relatively simple scenario, however, becomes considerably muddied once we start entertaining the idea of group rights, since these are rights to social goods such as culture or language, whose continued existence (and thus the possibility of continued enjoyment) depend on the activities of people acting as a collective rather than as individuals (Waldron 1993: 32).
Let us first consider the issue of the group as the bearer of rights. Each group will need to be distinguishable from every other group, so that it is possible to make sense of claims about the observance or violation of the right of a particular group. In this regard, one widely acknowledged problem with group rights is that there are often difficulties involved in identifying and defining the boundaries of a group (May 2001: 8; Waldron 1993: 32). For example, there may be conflicts between how a group is externally defined and how it is internally defined by the members themselves. And since the group itself may not be homogenous, the internal criteria employed may differ across members. Unlike groups whose existence and identities are defined independently of the individuals that make up the group, such as corporate entities, social or cultural groups are supposed to be representing the identities and activities of their members. Because members’ understanding of these identities and activities can be variable and contested, they are less likely to be resolved by definitional fiat. This then raises the question of how conflicts between individual and collective rights are to be resolved, since the privileging of some criteria for defining the group could well be at odds with the choices or preferences of some of the members. Moreover, as Edwards (2003: 555) points out, when we move from individual rights to group rights, there immediately arises the problem of differential treatment, since it is not possible to recognize the rights of all groups. As a consequence, some justifications must first be made as to why some groups and not others deserve recognition, and then, why from the set of recognized groups, rights attach to some groups and not to others.
In this regard, it is instructive to consider the United Nations’ own attempts to formulate a policy for the protection of minority groups. Duchêne (2008) demonstrates how the goal of protecting minorities by according them rights is surrounded by institutional debates about how to define the term ‘minority’, and how (p.23) such debates are influenced by the need to respect the interests of the state. States are keen to be seen as respecting the cultural distinctiveness of minorities. There is, however, a fear that too broad a definition of ‘minority’ might prove problematic for the construction of national unity by encouraging the rise of resistance or separatist movements, thus foregrounding the somewhat paradoxical relationship between minorities and states. Of particular interest from a linguistic perspective is Duchêne’s discussion of the use of language as a significant characteristic for identifying minority groups. An initial tripartite focus on race, religion, and language ultimately led to the privileging of language so that ‘linguistic minorities, although appearing later in discussions on minority protection, are progressively and almost automatically embedded in the totality of characteristics constituting a minority’ (Duchêne 2008: 86). In short, the UN’s discursive trajectory—from the Universal Declaration of Human Rights, to the International Covenant on Civil and Political Rights, and finally to the Declaration on the Rights of Persons Belonging to National or Ethnic, Religious, and Linguistic Minorities—arrived at the concept of ‘linguistic minority’ only because of the assumption that language provides a relatively unproblematic cultural characteristic for the purpose of identifying minorities compared to other cultural criteria:
And Duchêne (2008: 261) goes on to explain:
The emphasis on the language element allowed discussion on national, religious and ethnic minorities to be avoided, these being too nebulous or even politically sensitive.
The question of linguistic rights initially allowed significant ideological divisions to be avoided, and effectively applicable measures to be considered in an ‘objective’ manner: language issues were seen as easily defined and somehow a-problematic. (Duchêne 2008: 171)
This is a particularly important point, since it indicates that much policymaking is reliant on simplifying or perhaps even doggedly ignoring the complex nature of language (Blommaert 2001b; Makoni and Pennycook 2007).
Language, inasmuch as it is observable and variable, is considered as a non-problematic given, in the sense that there is no need to question it—it is evident. Discourses on language and linguistic minorities are not situated at the level of practices but at the level of object.
Having seen some of the problems involved in identifying the group as the bearer of a right, let us move on to consider the object of a group right, such as culture or language. There are equally serious difficulties involved in identifying and defining the object of the right itself, in this case, the associated culture or language, since the object of the right itself must now be delimited in such a way as to distinguish it from similar objects that may have other groups as bearers. To appreciate the difficulties involved, it is instructive to consider one widely discussed attempt to resolve this problem, namely, Kymlicka’s (1995) concept of a ‘societal culture’ (for more on Kymlicka’s work, see chapter 3), which he defines as (1995: 76)
(p.24) a culture which provides its members with meaningful ways of life across the full range of human activities, including social, education, religious, and economic life, encompassing both public and private spheres. These cultures tend to be territorially concentrated, and based on a shared language.
Kymlicka’s concept of societal culture is intended to capture a set of cultural institutions, values, and practices that is supposed to be identifiably associated with a specific group, and hence demarcatable from other forms of culture that might be associated with other groups. These other groups would presumably have their own distinctive societal cultures or have cultures that do not merit the status of ‘societal’. Kymlicka needs the concept of societal culture or something similar because, within his proposed framework for group rights, only particular cultures (those that are ‘fully developed’) qualify for self-government rights. Only societal cultures warrant or deserve the kind of political autonomy that amounts to self-government. Nonsocietal cultures do not qualify for self-government rights, but have to be satisfied instead with polyethnic rights (see the section on “Responses to Unavoidability” for an explanation of the various kinds of group rights proposed by Kymlicka).
Kymlicka’s appeal to this notion of societal cultures has been criticized for oversimplifying the vast range of cultural practices that humans engage in. The problem, as Carens (2000: 56; see also De Schutter 2007: 14) observes, is that the concept of societal culture homogenizes culture by ‘obscuring the multiplicity of our cultural inheritances and the complex ways in which they shape our contexts of choice’. For example, some cultures might not have a territorial concentration, especially if they are defined by a nomadic way of life. Others may not distinguish between the religious, economic, and educational activities presupposed by Kymlicka’s definition. Yet other cultures may not see themselves as having a shared language, since such a conception of language may be one that is imposed by external observers. Canagarajah (2007: 234), for example, points out that people in India ‘have difficulties identifying themselves in terms of one language’. Despite this, they have managed to ‘keep in tension and to dynamically negotiate competing claims such as identity/communication and inter/intra-group communication without letting them become a source of conflict and disharmony. Their language practices were based on negotiation rather than fidelity to unitary constructs’ (Canagarajah 2007: 234, italics added). All these cultures would, in one way or another, fail to qualify as Kymlicka’s ‘societal cultures’ and, as a consequence, fail to qualify for self-government rights.
An even more general problem with Kymlicka’s approach, as Benhabib (2002: 60–61) points out, is that the notion of societal culture ‘confuses social structure with social signification’:
Kymlicka has conflated institutionalized forms of collective public identities with the concept of culture. There are British, French and Algerian nations and societies that are organized as states; but there are no British, French or Algerian ‘societal cultures’ in Kymlicka’s sense. Any complex human society, at any point in time, is composed of multiple material and symbolic practices with a history … There is never a single (p.25) culture, one coherent system of beliefs, significations, symbolizations, and practices, that would extend ‘across the full range of human activities’.
Kymlicka, however, has little choice in this matter. He needs the concept of a societal culture (or something similar) because he needs to delimit the object of a (self-government) right against some other form of culture that may serve as the object of a (polyethnic) right. But it is only because of its artificiality that the concept of a societal culture is able to provide the delimitation needed to mark off the contours of one kind of culture (and one form of group right) from some other.
To summarize the discussion thus far, the idea of a right requires that clear boundaries be drawn between different rights-bearers, as well as between the objects of rights. At this point, however, we are still dealing with the question of how to justify group rights in the form of cultural (rather than language) rights. Language rights advocates make the move from the broader category of cultural rights to the narrower one of language rights by arguing that language is an especially worthy candidate as the object of a right because it is often considered a critical part of a group’s identity, and because it also influences access to socioeconomic goods.
However, given the need for boundary marking in a rights-based discourse, any language that has been enshrined as the object of a right (like the culture that it is supposed to be a part of) then acquires a solidity that can be significantly at odds with the actual experiences of its speakers. It is this that opens language rights advocates to charges of essentialism, despite the fact that some language rights advocates are acutely aware of the pitfalls of essentialism and have striven to avoid them (chapter 3). Essentialism still remains a problem, however, because of the very nature of a rights-based discourse itself. This is the focus of the following section.
The Essentialism of Rights Discourse
The reason why the discourse of rights contains essentializing tendencies is that in order for traits/practices to be the kind of thing that a group can claim entitlement to, both the group and the traits/practices, as well as the relationship between the two, have to be ontologically constituted as having a unity and consistency that allows each to be coherently identified. This happens through a significant appeal to erasure (Gal and Irvine 1995), as heterogeneity internal to the group, variability in the traits/practices, and changes in the relations between them are all downplayed or even ignored in order to fit the dictates of the discourse of rights.
It should be clarified that essentialism does not necessarily mean stasis. Essentialism is able to accommodate change in that some changes to a trait/practice may be viewed as having a relatively superficial effect that does not compromise its authenticity, while other changes may be seen as having a more fundamental effect that undermines the authenticity of the trait/practice. The point to bear in mind is that there is no objective way to decide which changes are superficial and which are not, since what counts as authentic is itself the outcome of socially negotiated (p.26) processes of authentication (Bucholtz 2003: 408), including those processes sanctioned by a rights discourse (see, in particular, the discussion of reinvention below).
The discourse of rights, despite its ‘emancipatory aura’, is neither ethically unambiguous nor neutral (Cowan, Dembour, and Wilson 2001: 11). It is crucial therefore that we analyze this discourse for the kinds of effects that it brings about (cf. Freeman 2002: 85). Focusing on the particular case of language rights, I argue that a rights discourse imposes the following three effects:
i. Selectivity: Since not all the practices associated with a social group are appropriate candidates as the objects of rights, a rights discourse exerts pressure such that a few selected practices are privileged over others (Ford 2005: 71).
ii. Reinvention: In some cases, the pressure to come up with appropriate practices can lead a group to engage in reinvention, such as modifying the practices in ways that fit the demands of a rights discourse. This may include providing the practices with the necessary authentication demanded by rights-conferring authorities, and asserting that these practices unanimously reflect the collective history of the group (Tamir 1993: 47).
iii. Neutralization: A rights discourse neutralizes the distinction between strategic and nonstrategic essentialism. Essentialism is strategic when group members or advocates acting on their behalf deliberately treat as stable and clearly defined phenomena that (they are aware) are in fact highly fluid, variable, or even conflicting (Cowan et al. 2001: 10; McElhinny 1996). Bucholtz (2003: 401) suggests that strategic essentialism is usually intended as a short-term measure, although ‘not all participants who commit themselves to an essentialist position necessarily recognize it as a temporary tactic’. Neutralization means that an essentialist claim (strategic or otherwise), once locked into the rights discourse, has no clear ‘exit strategy’, making it difficult for such a claim to work as a temporary tactic (Ford 2005: 68).
Some of these effects undoubtedly exist independently of a rights discourse, but the point I am pursuing here is that these effects are necessitated, if not exacerbated, by the demands of this discourse, where the rights tail comes to wag the cultural dog. The result is a situation where ‘rights may be constitutive of cultures and their associated identities’ (Cowan et al 2001: 11, italics in original; see also Ford 2005: 73) rather than simply protecting them.
In the following section I provide an overview of critical investigations into rights discourse, focusing on how it encourages essentialism. I then attend specifically to language rights and consider, in turn, the effects of selectivity, reinvention, and neutralization, drawing on data from Singapore, South Africa, and Sri Lanka, respectively.
Selectivity, Reinvention, Neutralization
Barry (2001: 252ff) points out that there is a tendency to defend a cultural practice on the grounds that it is part of the culture of the group whose practice it is, so that ‘simply in virtue of forming part of the group’s culture, it is essential to its (p.27) well-being’. The corollary of this tendency is that any change to the nature of the practice, particularly if such change is seen as being imposed from outside the group, can be perceived as a threat to the group’s cultural identity. Citing a variety of examples of such defenses, including the 1835 Maori massacre of the Moriori people (‘these we killed … It was in accordance with our custom’), the clubbing of seals by Canadian sealers (‘My family has gone sealing for generations … It’s a vital part of our culture’), and the killing of whales by the Chukchi (‘the right to hunt whales when it is deemed a traditional part of their culture and diet’), Barry (2001: 258) argues that
if there are sound reasons against doing something, these cannot be trumped by saying—even if it is true—that doing it is a part of your culture. The fact that your (or your ancestors) have been doing something for a long time does nothing in itself to justify your continuing to do it.
Ford (2005) delivers a similar critique in his discussion of Renee Rogers et al. v. American Airlines, Inc. (1981), where the plaintiff was a black woman seeking damages against the airline for prohibiting employees from wearing an all-braided hairstyle. Rogers’ assertion (quoted in Ford 2005: 23) was that ‘the “corn row” style has been, historically, a fashion and style adopted by Black American women, reflective of cultural, historical essence of the Black women in American society’. Ford (2005: 25, italics in original), however, points out that
What’s clear is that the assertion that cornrows are the cultural essence of black women cannot be taken as conclusive evidence that a ‘right-to-cornrows’ is an unadulterated good thing for black women. Even if we take it on faith that cornrows represent black nationalist pride as against the integrationist and assimilationist coiffure of chemically straightened hair, it’s clear that a right to cornrows would be an intervention in a long-standing debate among African-Americans about empowerment strategies and norms of identity and identification. More generally, it is by no means clear that an argument that presumes that blacks or black women have a cultural essence as blacks or as black women is a vehicle of racial empowerment. A right to group difference may be experienced as meddlesome at best and oppressive at worst even by some members of the groups that the rights regime ostensibly benefits. For the black woman who dislikes cornrows and wishes that no one—most of all black women—would wear them, the right not only hinders her and deprives her of allies, but it also adds insult to injury by proclaiming that cornrows are her cultural essence as a black woman.
Ford (2005: 26, italics in original) further observes that ‘If an all-braided style is the cultural essence of black women by law’, this would imply that nonblack women who adopted braids might be seen as ‘“white Negro” wanna-bes’:
It’s likely that a right premised on the immutable link between blacks and braids will discourage white and Asian women from wearing braids by sending the message that (p.28) the hairstyle “belongs” to another social group. Although a right to cornrows might seem only to enhance the freedom of potential cornrow wearers, it is arguably better understood as a policy of segregation through which a set of grooming styles are [sic] reserved for a particular group.
The observations by Barry (2001) and Ford (2005) are relevant to May’s (2005) assertion that when groups engage in rights claims, they do so only in order to secure cultural and linguistic autonomy without making any claims involving essentialism. There are two problems here. First, the autonomy of the rights-claiming group vis-à-vis other groups or society at large is secured on the basis that the cultural practice in question represents the group’s collective identity. Thus the autonomy of the group is privileged over the possibility that individual members within the group may in fact have a different view of the practice in question, and may reject its putative role as a representative practice.
Kymlicka (1995: 35) is well aware of the seriousness of this problem and has attempted to handle it by distinguishing between ‘internal restrictions’ and ‘external protections’. The former are intended to refer to the kinds of obligations that a group may expect of its members. In contrast, the latter refer to the kinds of demands that the group may place on the larger society. Kymlicka (1995: 37) wants to allow for (at least some kinds of) external protections, such as group rights, while rejecting internal restrictions that ‘limit the right of group members to question and revise traditional authorities and practices’. Unfortunately Kymlicka’s distinction presupposes and reifies the boundary between a group and society at large. The statement that internal restrictions should not prevent group members from rejecting traditional authorities and practices, meanwhile, does nothing to address the fact that, because a rights discourse is being invoked, there is going to be enormous pressure on the group to present a unified front. This means that dissenting members, rather than voicing a different opinion that might undermine the group’s rights claims, may feel obligated to adopt a stance of silence instead (cf. Tamir 1993: 47), such that the group’s autonomy is achieved at the expense of the autonomy of individual members.
Second, the idea that autonomy implies freedom from interference by other groups is unrealistic. If the goal is a concept of multiculturalism where different groups coexist in separate cultural cocoons, then this fails to recognize that in any plural society, some form of social interaction and cross-cultural ‘seepage’ is inevitable. Even if such cultural separatism is not the goal, entrenching practices as rights makes it difficult for groups to dynamically negotiate and accommodate changes to the practices in response to changing social conditions. This is because a rights discourse is usually intended as an attempt to enforce a legal limit on the social and political power of the state or that of a more dominant group, in order that the ‘threatened’ group can then be free to continue its practices in a maximally unfettered manner (cf. Ford 2005: 68). But rather than being conceptualized as freedom from outside interference, the idea of autonomy is better recognized as the ability to reciprocally negotiate, at a variety of levels (individual, intragroup, and intergroup), how much change to any kind of practice is acceptable, tolerable, (p.29) or even desirable. Unfortunately the often hard-fought and adversarial conditions under which rights are gained tend not to encourage this kind of reciprocal engagement (B. Turner 1993: 175–76).
The combination of privileging group autonomy over that of its individual members with the understanding of autonomy as freedom from meddling by nonmembers leads to the idea that rights can be used to curb interference from without, and so allow the rights-claiming group to freely express its cultural identity. This points clearly to an assumption of essentialism. It treats the group’s identity as sui generis, preexisting and intrinsic, thus privileging it above and beyond the diverse lived experiences of its members. It enshrines the practice as one that genuinely reflects the group’s identity only insofar as the integrity of the practice is maintained. That is, the practice is perceived to reflect the group’s ‘intrinsic’ autonomy rather than undesirable influences ‘from without’, for only in this way can it be seen as helping the group to maintain its distinctiveness.
A right seen in this way is not a device that simply reflects or protects an existing cultural practice. Rather, it transforms the practice by institutionalizing it as being especially important, and in this way elevates the selected practice over other practices that might also be said to be just as relevant to the group’s identity (cf. Freeden 1991: 43). As a consequence, the right comes to play a constitutive role in the culture of the group, since it helps to mark some practices as being more significant than others, more worthy of group attention and support. The discourse of rights thus encourages essentialism by insisting that particularly strong reasons be given for according the selected practice the status of a right. And the strongest possible reason that can be given is that the practice embodies an essential property of the group.
Having looked at how a rights discourse encourages essentialism, I now elaborate on the various effects it has on claims involving language, namely selectivity, reinvention, and neutralization, in the following subsections.
Selectivity is built into the notion of rights because it is not the case that anything can or should be accorded the status of a right. Rather, ‘the language of rights is used to refer to any demand that an individual interest should be protected or promoted … and accorded decisive moral importance’ (Waldron 1993: 32, italics added). Allowing otherwise would mean trivializing the idea of a right.
Although selectivity is clearly a separate issue from standardization, Milroy’s (2001) discussion (see chapter 4) of the ideology of a standard is still relevant here. As he observes, once a particular variety has been chosen as the standard, other varieties, by implication, are nonstandard, and consequently less prestigious. Similarly, selectivity in the case of language rights manifests itself as the selection of a particular linguistic code deemed to be of sufficient importance to represent a group’s cultural identity, with the consequence—whether intentionally or not—that other varieties are diminished in stature, or even denigrated. Speakers (p.30) who insist on using the nonselected codes may be judged less sophisticated, less respectable, or even unpatriotic. It may also be the case that clinging to these nonselected codes prevents the speakers from participating in status domains of influence, such as politics or education.
An example from Singapore will illustrate these points. In Singapore, the management of ethnic diversity is guided by the principle of multiracialism, where respect and equal treatment must be accorded each ethnic community (Benjamin 1976; see also chapter 4). This commitment to multiracialism stems from the government’s belief that the desire to maintain ethnic and cultural distinctiveness is a primordial fact, and any policy that tries to force different ‘tribes’ to submerge their differences can only lead, in the end, to insurgency. We see this in the following statement by Singapore’s first prime minister, Lee Kuan Yew (cited in Han, Fernandez, and Tan 1998: 163–65):
The Indians have their own method. So do the Malays. The Malays: Islam and also the kinship ties … I don’t think you can erase all that. That’s for hundreds of years, or thousands of years. You can’t erase it. Because I recognized it, I decided you cannot change it. Or if you tried to change it, you’d change it for the worse … In every culture, there is a desire to preserve your distinctiveness. And I think if you go against that, you will create unnecessary problems, whether it is with the Indians and their caste or with the Chinese and their clans.
Given this assumption, it is not surprising that the Malays, the Chinese, and the Indians each have an officially recognized mother tongue:1 Malay, Mandarin, and Tamil, respectively. While it is arguably the case that having the Malay language as the mother tongue for the Malays accurately reflects the sociolinguistic reality of that relatively homogenous community, this is less plausible when it comes to the more heterogeneous Chinese and Indian communities. I discuss here the case of the Chinese community (see PuruShotam  for a discussion of the Indian community).
The selection of Mandarin as the Chinese community’s official mother tongue was made in the face of a great deal of linguistic heterogeneity—until the 1980s, approximately 11 different Chinese dialects were widely spoken.2 But once Mandarin was selected, a number of language planning activities were undertaken to reflect the status of Mandarin as the Chinese community’s mother tongue, including a Speak Mandarin Campaign (SMC), first launched in 1979 and relaunched annually since. Campaign slogans include ‘hua ren hua yu’ (literally, ‘Chinese people, Chinese language’), ‘Mandarin is Chinese’, and, ‘If you are Chinese, make a statement—in Mandarin’.
(p.31) To enforce Mandarin’s preeminent status, the SMC aims not only to encourage the use of Mandarin but also to eliminate dialects. Lee Kuan Yew (1979 SMC speech) mapped out the various changes to the public environment: ‘All government officers, including those in hospitals and clinics, and especially those manning counters, will be instructed to speak Mandarin except to the old, those over 60’ (Bokhorst-Heng and Wee 2007: 326). Furthermore, ‘all Chinese taxi-drivers, bus conductors, and hawkers, can and will be required to pass an oral Mandarin test, or to attend Mandarin classes to make them adequate and competent to understand and speak Mandarin to their customers’ (Bokhorst-Heng and Wee 2007: 326). In addition, television programs and radio broadcasting eliminated the use of dialects; the censorship board stopped authorizing dialect films and videos unless they were dubbed in Mandarin; and Mandarin lessons appeared regularly on television and radio, in the newspapers, and on posters around the country. The removal of dialects from the media affected particularly the older Chinese, who spoke little or no Mandarin and were consequently deprived of access to news, entertainment, and general information. Lee Hsien Loong (then Trade and Industry and Second Defense Minister) in fact recognized this ‘sacrifice’ in his 1988 SMC speech:
At first, many, especially older folks, were unhappy that SBC was phasing out dialect programs on television. Now, Singaporeans have accepted it … I recognize what a sacrifice the older generation of dialect speaking Chinese have made, for us to achieve this transformation. I thank them for making this sacrifice, and co-operating in this effort to change the spoken language of the whole community. Their grandchildren will be grateful for what they have done.
There are a number of qualifications worth making here. One, the selection of Mandarin was not simply a top-down imposition by the government. Although early Chinese education was privately funded by Chinese businessmen of different dialect affiliations, after 1920 the medium of education shifted to Mandarin, ‘a consequence of the sweeping tide of Chinese nationalism and the increasing supply of teachers from China’ (Hill and Lian 1995: 71). In the 1950s the Chinese Chamber of Commerce argued that the Chinese language (Mandarin) deserved greater recognition in the schools and in the legislature (Hill and Lian 1995: 59). This was in vain, however, because graduates of Chinese high schools still found fewer opportunities for employment or higher education compared to those from English-speaking schools, leading eventually to a growing sense of resentment and ‘political radicalization’ (Hill and Lian 1995: 71). Clearly the government’s decision to select Mandarin was in part a response to a variety of historical factors that had already given this language a highly significant role in the Chinese community.
Two, even though there were strong justifications for the choice of Mandarin, the fact remains that a rights discourse intent on recognizing the ethnolinguistic distinctiveness of the Chinese community forced a choice as to which linguistic variety would be selected. This brings up the strongly comparative orientation (p.32) that a rights discourse encourages—the need to keep up with the Joneses, where if some group is given a particular right, then other groups must also be accorded comparable ones. In the Singaporean context, the commitment to multiracialism and the recognition of one mother tongue for each community forced the privileging of exactly one language, Mandarin, over other Chinese dialects. If the Chinese community had been allowed more than one official mother tongue, this would have been acceptable only if the other ethnic communities were also given the same number of options. But notice that even here, there is no escaping selectivity, in this case, a selection of some arbitrary number of languages.
Three, as a result of the elevation of Mandarin as the community’s mother tongue, the variegated linguistic experiences of individual speakers and their families diminished in importance. The most poignant example, as noted above, involves the elderly who were later ‘thanked’ by the state for their linguistic sacrifice—a sacrifice that they did not engage in volitionally, especially since it also created serious intergenerational communication difficulties between dialect-speaking grandparents and their (grand)children who were brought up with English and Mandarin (Mauzy and Milne 2002: 107). This reinforces the point made earlier that individual autonomy and group-internal diversity tend to be sacrificed at the altar of group rights, which insists on the presentation of a unified front.
Four, while speaking Mandarin started off as a right (albeit one primarily assigned by the government3), it quickly also became an obligation. Changes made to the public environment created a linguistic market (Bourdieu 1991) that penalized those lacking knowledge of Mandarin. In addition, dialects were characterized as vulgar, primitive, and a major cause of miscommunication; and dialect-speakers were often disparaged as being unsophisticated, uneducated, or just plain boorish (Bokhorst-Heng 1999: 250–52; Wee 2002b: 212). Thus only by speaking Mandarin (rather than dialects) was one considered to be truly manifesting one’s identity as a Chinese Singaporean.
Whether one wishes to speak nonjudgmentally of ‘nationalist imaginings’ (Anderson 1991) or adopt a more evaluative stance on their truth/falsity (Hobsbawm 1990; see May 2001: 68), the formulation of language policies is widely recognized to be influenced by ideologies of various sorts (Fishman 1973: 31–32; Spolsky 2004). The issue at hand, however, is not simply that ‘imaginings’ occur, perhaps necessarily, whenever a group of people attempts to construct a shared identity and sense of destiny. It is that a rights-based approach forces such imaginings to take on starkly defined contours so as to satisfy the demands of legality, which I refer to here as reinvention. Since claims to rights tend to take place in the context of ‘competitive struggles’ between groups (B. Turner 1993: 175–76), such (p.33) claims may be challenged, and this creates a ‘need to make a clear case to counterbalance an opposing one’ (Clifford 1988: 321, italics added). However, as Ford (2005: 71) points out, a legal demand for clarity conflicts with the flux and variability of actual lived cultural practices:
The legalism of difference discourse encourages, and rights-to-difference require, formal conceptions of social identity that easily can be asserted in courts. Courts and judges will most likely protect cultural styles that can be easily framed in terms of fixed categories, bright-line rules and quasi-scientific evidence. Courts will want experts to testify as to the content of the group culture, they will want lists of specific and concrete manifestations of the culture. Judges are likely to want the culture to be fixed and knowable and will want the protected behavior to be reflexive so as to distinguish culture from merely deviant behavior.
… The blame lies with the very project of trying to define group differences with sufficient formality as to produce a list of traits at all.
Likewise, as Clifford (1988: 338; see also Cobo 1987) tells us in his discussion of the Mashpee’s 1976 claim for tribal status in Cape Cod, legal understandings of culture contain a ‘bias toward wholeness, continuity and growth’. This is despite the fact that ‘[m]etaphors of continuity and “survival” do not account for complex historical processes of appropriation, compromise, subversion, masking, invention and revival’ (Clifford 1988: 339).
These problems are illustrated in the dilemma faced by speakers of Northern SiNdebele in South Africa (Stroud 2001). Although both Northern and Southern SiNdebele formed part of the Nguni group, only speakers of the latter had their language officially recognized under apartheid because they had accepted the offer of a homeland. This was the situation inherited by the new South African government in 1994, so that Northern SiNdebele speakers still found themselves excluded from any discussion of language rights. This led the Northern AmaNdebele National Organization (NANO) to lobby the government and the Pan South African Language Board (PANSALB) to have their language officially recognized. The call was rejected, however, because, as one PANSALB executive explained (quoted in Stroud 2001: 349, italics in original), ‘we could not promote their case until we had clarity on whether Northern SiNdebele was a separate language from Southern SiNdebele.’ Northern SiNdebele was considered a dialectal variant of Southern SiNdebele, and until proven otherwise, could not be considered sufficiently important to warrant the status of a right.
As a result, Northern SiNdebele speakers have had to accommodate a rights discourse that ‘views language as an essentially unproblematic construct—an identifiable ontological entity’ (Stroud 2001: 348). As Stroud (2001: 349) explains:
NANO found itself in the position of having to argue that Northern SiNdebele was a language, which meant a grassroots investment in developing orthography, grammar and glossaries for school. The organization also developed grassroots strategies to demand the use of SiNdebele as a medium of instruction in primary education.
(p.34) While Stroud does not himself emphasize this point, it seems clear that the activities involved in constructing an orthographic system for the language, providing a grammar and glossaries, and ensuring that the language can be used as a medium of instruction will lead to transformations in the language itself (Cooper 1989: 33; Wee 2005),4 since any language that is used as a medium of education or has its lexicon reflected in glossaries will have to undergo some degree of standardization, where there is pressure toward the elimination of otherwise informally tolerated linguistic variation (Milroy and Milroy 1999). Furthermore, the pressure to present Northern SiNdebele as a language that is equal to but distinct from Southern SiNdebele will undoubtedly also encourage various linguistic innovations that are intended to serve the purpose of marking a different group identity (cf. Bucholtz and Hall 2004: 385; King 1994; Pullum 1999: 44).
This kind of reinvention in and of itself is neither necessarily good nor bad. The point to note, though, is this: The reinvention that the Northern SiNdebele speakers are engaged in here is a direct response to the dictates of a rights discourse, which assumes that (i) only a language (as opposed to a dialect) is worthy of being granted rights status; (ii) there exist clear and unambiguous criteria for the language-dialect distinction; and (iii) if such criteria are not currently available, they can and should be created. Community resources are thus pooled specifically in order to satisfy these assumptions.
With selectivity and reinvention, there is arguably a much greater degree of agency available to rights petitioners, since they can make decisions—however constrained—regarding which language should be the object of a right, and what kinds of efforts ought to go into making a case for such a right. In contrast, neutralization draws attention to the limits of agency, since despite any intention to restrict the tenure of a right, once enshrined as such, the right becomes extremely difficult to dismantle. This indicates another difference—a sequential one—between selectivity and reinvention, on the one hand, and neutralization, on the other. The first two precede the granting of a right, whereas the third is concerned with what happens after the right has been granted.
Consider the case of Sri Lanka, where the ongoing ethnic conflict between the minority Tamils and the majority Sinhalese revolves around language issues, framed variously as ‘mother tongue versus English’ and ‘Tamil versus Sinhala’ (Canagarajah 2005: 419–20, 424). Sri Lanka’s independence from British colonial rule in 1948 was quickly followed by the demand that English be replaced by the people’s ‘own language’, even though there was no single ‘own language’ but at least two major languages: Tamil and Sinhalese (Sowell 2004: 84).
(p.35) By 1956 a new government under Solomon Bandaranaike was elected on the platform that Sinhala would serve as the official language of Sri Lanka. Crucially, Bandaranaike’s own commitment to the Sinhalase cause was strategic: he grew up unable to speak Sinhalese, only learning the language much later in life. His adoption of an extremist position on Sinhalese culture was purely driven by his desire to become prime minister (Sowell 2004: 85), and once this goal was achieved, he set about moderating his own anti-Tamil policies, ‘but this only set off howls of protest from other Sinhalese demagogues with political ambitions of their own, including a future president, J. R. Jayawardene’ (Sowell 2004: 86).
In 1990, after years of increasingly hostile anti-Tamil policies, the militant Liberation Tigers of Tamil Eelam (LTTE) managed to set up a regime that was intended to be free from Sinhala government. According to Canagarajah (2005: 424–25):
the regime insisted on Tamil Only and Pure Tamil in certain extreme terms when they established their de facto state in 1990. This policy also served to prove themselves more Tamil than the middle-class politicians … they found support for the argument that it is only in creating a separate state of homogeneous Tamil community that the mother tongue could be empowered … Their own monolingual/monocultural ideology has always been claimed by the LTTE as evidence that they were more faithful to the Tamil cause.
The result has been an emphasis on ‘pure’ Tamil and the marginalization of those Tamils for whom code-mixing might actually constitute a normal language practice. In other words, emphasizing Tamil in its purest form helped to legitimize the denigration of other codes that may be mixed or hybridized varieties, and which are considered illegitimate precisely because of their perceived linguistic impurity. As a consequence, speakers for whom such hybridized creoles (Blommaert, Collins, and Slembrouck 2005) are in fact naturalized ways of speaking may end up being penalized. Thus in the LTTE-governed regime of Jaffna, with its emphasis on pure Tamil, warnings are given that the use of English is ‘damaging traditional Tamil culture and hindering the nationalist struggle’ (Canagarajah 2005: 425). Petitions or applications to the police, the courts, and village councils are likely to be rejected if these are not made in ‘pure’ Tamil or ‘Tamil only’. Such penalization can in fact be gleaned from the following exchange (from Canagarajah 2005; 426), where a Tamil woman applying for a travel permit is observed to have unwittingly used an English borrowing (wedding) in what is otherwise supposed to be an exchange conducted in pure Tamil.
The woman’s use of wedding immediately leads to a rebuke from the government officer and ‘although it takes some time for her to realize her blunder, she corrects herself as her petition can easily be turned down for such mistakes’ (Canagarajah 2005: 426, italics added).
Officer: appa koLumpukku een pooriinkaL?
‘So why are you traveling to Colombo?
Woman: makaLinTai wedding-ikku pooren
‘I am going for my daughter’s “wedding”’
Officer: enna? unkaLukku tamiL teriyaataa? England-ilai iruntaa vantaniinkal?
Officer: enkai pooriinkaL?
‘Where are you going?’
Woman: cari, cari, kaLiyaaNa viiTTukku pooren, makan
‘Okay, okay, I am going to a wedding, son’
The insistence on Tamil only has proven more effective vis-à-vis Sinhala than English, since there is a desire for Tamil–English multilingualism among many ordinary Tamils, who associate English with socioeconomic mobility and liberal values (Canagarajah 2005: 438). Despite this, the LTTE cannot be seen to support such multilingualism since ‘the reason they won the leadership in the struggle for Tamil rights … is because they could come up with a more populist and radical slogan … to give up their language policy is political death’ (Canagarajah 2005: 441).
We therefore see evidence from both sides of the civil conflict of how essentialist positions (Sinhala only, Tamil only) are adopted by political and military leaders in order to gain popular support in the struggle for rights on behalf of the Sinhalese and Tamil communities, respectively. Furthermore, proponents of these different positions are each victims of their own success in that they are unable to moderate their respective platforms without the risk of losing their power base, since any such moderation would be vulnerable to the charge that it jeopardizes the very right that it is supposed to champion. In this way, once an essentialist position is intertwined with the discourse of rights, it then takes on a life of its own regardless of whether it was intended strategically or not. The reasons for this are fairly obvious. Neither groups (and their members) nor their cultural practices are ontologically fixed and homogenous entities, even though a rights discourse treats them as such. But precisely by doing so, a rights discourse ignores the dynamic struggle for goods, services, and privileges that results when a particular linguistic practice is given the status of a right. For example, Bandaranaike’s ‘Sinhala only’ policy gained popular support because it was seen as a response to a situation where Tamils were disproportionately represented in the upper echelons of Sri Lankan society. This policy therefore provided the rationale behind the implementation of preferences and quotas that limited the Tamils’ education and employment prospects, and included allowing Sinhalese applicants to meet lower standards for university admission than their Tamil counterparts (Sowell 2004: 86).
The link between language and socioeconomic privileges, predictably, meant that any subsequent attempt to loosen the connection would lead to protests from the Sinhalese. And, of course, there is always the issue of different cohorts internal to the group. Not all Sinhalese benefit from the ‘Sinhala only’ policy in exactly the same manner or at the same time. Thus, while some may later on be willing to moderate the policy, others will still feel that they have not been sufficiently (p.37) compensated to justify the policy’s weakening or dismantling. Likewise, the declining socioeconomic prospects for Tamils triggered a strong pro-Tamil reaction. By the same token, there was an attempt by the LTTE to create, within its own autonomous regime, a connection between the Tamil language and socioeconomic privileges. Again, predictably, any weakening of this connection becomes extremely difficult to achieve since it could lead to accusations from different cohorts within the Tamil community that the LTTE are losing their commitment to the very ideals that legitimized their ascendancy to political power.
The reason, therefore, why a rights discourse neutralizes the distinction between strategic and nonstrategic essentialism is that groups and their members are not, as Sowell (2004: 7–8) points out, ‘inert blocks of wood to be moved here and there according to someone else’s grand design’. A rights discourse has the effect of freezing group relations and privileges, while in reality, sociocultural differences and inequalities are constituted by ongoing and ever-changing processes of struggle, as individuals qua individuals as well as qua group members strategically utilize available resources to maintain, enhance, or transform their various life-chances.
Language Neutrality: A Chimera
We have seen that a rights discourse encourages essentialism by imposing sharp and rigid boundaries on cultural practices that are fundamentally at odds with their more fluid and dynamic nature, manifested via selectivity, reinvention, and neutralization.5 Such an approach has the unfortunate effect of confining and restricting ‘behavior, expression and identity to precisely the degree to which it protects them’ (Ford 2005: 90). In the case of language practices, the appeal to rights is particularly problematic because—perhaps more so than other kinds of cultural practices—language is ‘irreducibly dialectic in nature’, ‘an unstable mutual interaction of meaningful sign forms, contextualized to situations of interested human use and mediated by the fact of cultural ideology’ (Silverstein 1985: 220).
Other cultural practices are by comparison relatively stable (however, see chapter 9). For example, prohibitions against the consumption of pork or the wearing of a turban or cornrows are practices that are fairly easy to distribute in a homogenous manner across all the relevant members of a group (notwithstanding the question of individual preferences). These practices involve a high degree of agency, in that individual actors make conscious decisions about whether or not they wish to observe group restrictions concerning diet, dress, or hairstyle. Consequently changes in how these practices are observed are more easily detected, and are thus more amenable to discussions and debates regarding their (p.38) acceptability both to the group itself and to the wider society. These practices arguably involve relatively clear boundaries that make it less problematic to distinguish members from nonmembers.
In contrast, language practices are extremely variable since control over lexicogrammatical resources and how these are used is highly dependent on the gender, status, and age of individual members (Spolsky 2004: 9), as well as the kinds of activities they engage in (Levinson 1992). This means that language is never distributed uniformly across even what might be considered the same cultural group (since there will inevitably be internal sociocultural distinctions along the lines of gender, age, status, etc., and acceptable associated activities). And unless interaction between members and nonmembers is completely eliminated, some form of linguistic exchange—perhaps starting with code-switching and gradually resulting in fused lects (Auer 1999)6—is also inevitable. Such exchanges can have subtle effects on speakers’ communicative resources, effects that the speakers themselves may not be aware of nor are consciously able to control.7 This makes it even more critical to appreciate that language is possibly more unsuitable than most other cultural practices for the kind of boundary marking needed for enshrinement within a rights discourse.
Such a conclusion does not deny the critical role language plays in influencing the distribution of sociocultural and economic resources (Heller and Martin-Jones 2001: 2, 419). But it does suggest that treating language as a cultural right tends to obscure the changing interests and needs of different individuals and groups. Perhaps most significantly, the liberal ideology, which provides the grounding for the conception of rights (Ford 2005: 6; Ryan 1993: 296), will inevitably always create tensions, since expectations of autonomy and equality are simply impossible in the case of language. Neutrality is impossible in the case of language since some language will always have to be used in order to facilitate the conduct of any social activity, with the consequence that speakers of this language will automatically be advantaged in relation to those who don’t (cf. Rubio-Marín 2003: 55). Such a situation violates the liberal ideals of equality (since advantages accrue to individuals as a result of whether they happen to speak the dominant language, which is often a happenstance of one’s birthright) and autonomy (since disadvantaged individuals may have no choice but to accommodate the dominant language, possibly resulting in language shift and the consequent loss of their own language, with attendant implications for their sense of community and cultural identity).
(p.39) It is therefore imperative to recognize that language neutrality is a chimera, and consequently so is the vision of linguistic autonomy and equality that the notion of language rights promises. There is no escaping this sociolinguistic reality, and to bring this point home, I now focus on two specific properties of language: its unavoidability and its hybridity. Unavoidability refers to the situation described in chapter 1, where it is simply not possible to avoid privileging some language. This alone poses serious challenges for the liberal commitment to autonomy and equality, since it is not a normative but rather a functional fact of human interaction that some specific language will need to be used. Hybridity refers to the fact that language change and variation is inevitable as a result of language use and contact. This is particularly relevant in the context of plural societies where diverse communities and languages coexist—although it also applies in what we might want to think of as relatively homogeneous societies, and creates equally serious problems for the liberal commitment to autonomy and equality.
Responses to Unavoidability
Both Kymlicka (1995: 111) and Rubio-Marín (2003: 55) acknowledge the issue of unavoidability, the latter in her remarks that it is not possible for the state to guarantee ‘linguistic neutrality’ and the former in a succinct observation about the difference between language and religion. I reproduce Kymlicka’s comments here:
It is quite possible for a state not to have an established church. But the state cannot help but give at least partial establishment to a culture when it decides which language is to be used in public schooling, or in the provision of state services. The state can (and should) replace religious oaths in courts with secular oaths, but it cannot replace the use of English in courts with no language.
Both Kymlicka and Rubio-Marín are supportive of the notion of language rights and respond to the problem of unavoidability in similar ways—by differentiating between different kinds of rights. Kymlicka does this by introducing the different kinds of group rights briefly mentioned earlier: self-government rights, to allow a group ‘full and free development’ of its culture (1995: 27); polyethnic rights, to allow a group to ‘express [its] cultural particularity’ (1995: 31); and special-representation rights, to compensate for the underrepresentation of specific groups such as the disabled or the poor (1995: 32). Only the last is intended as a temporary right, which should be removed once specific sources of oppression have been removed. The first two are intended to be permanent, but they differ in that only self-government rights allow for the possibility of secession, while polyethnic rights, in contrast, are intended to promote integration into the larger society (Kymlicka 1995: 31).
I leave aside for the moment discussion of special representation rights (see chapter 9) since it is really the other two kinds of rights that are most directly (p.40) related to the issue of language and culture. Kymlicka’s (1995: 76) application of self-government and polyethnic rights is best illustrated by the distinction he makes between national minorities and immigrant groups, and their relationship to his highly problematic concept of a ‘societal culture’. For Kymlicka (1995: 113), national minorities have the strongest claim to language rights because they are in possession of a societal culture:
In a democratic society, the majority nation will always have its language and societal cultures supported, and will have the legislative power to protect its interests in culture-affecting decisions. The question is whether fairness requires that the same benefits and opportunities should be given to national minorities. The answer, I think, is clearly yes.
On the other hand, Kymlicka expects immigrant groups to integrate into the existing national culture because they are construed as having (mostly) come voluntarily to the new host society and, by so doing, are no longer entitled to have their original societal culture established in their new home (Carens 2000: 55). Kymlicka (1995: 114) suggests that the provision of language training may be more appropriate here.
This last observation concerning language training for nonspeakers of the dominant language leads us nicely to Rubio-Marín’s (2003) proposal, which builds this expectation that minority groups should learn the language of the dominant culture into a right and duty.8 Rubio-Marín (2003: 68) makes a distinction between noninstrumental and instrumental language rights. The former is concerned with protecting ‘one’s membership in a language community, [protecting] a certain context of choice and sense of identity provided by its culture’. In contrast, the latter is concerned with ensuring that all members of a society have the right to learn the dominant language so as to avoid ‘linguistic obstacles that may curtail the enjoyment of rights, freedoms, and opportunities that rest on the possibility of comprehensible linguistic interactions’. Rubio-Marín’s argument is that there are circumstances, such as the state’s provision of social services or schooling, where the purpose served by language is mainly instrumental or narrowly communicative in orientation, rather than identity related (2003: 57). Under such circumstances, having everyone learn the dominant language is ‘less harmful to other general interests’, such as the issue of cost-effectiveness, and the preservation of national identity through a common language (Rubio-Marín 2003: 68).
The key difference between Rubio-Marín (2003) and Kymlicka (1995) lies in whether or not self-government rights are needed. Kymlicka explicitly argues that national minorities should be accorded such rights to the point that they should be allowed to secede if they so wish, whereas Rubio-Marín does not see the need for (p.41) such rights. Consequently, while both scholars acknowledge the importance of learning the dominant language, Kymlicka’s self-government rights trump the desire to develop a unified national identity, whereas Rubio-Marín acknowledges that broader public policy considerations may override the desire to preserve a separate societal culture by highlighting cost-effectiveness and other general interests.
Unavoidability, Autonomy, Equality
In Kymlicka’s (1995) proposal, equality is itself unequally distributed, since only national minorities, in contrast to immigrant groups, qualify for self-government rights, which would allow a group to decide if it wishes to have its own language used in various social institutional settings. And even though he is aware that not all immigrants (such as refugees) would have left their countries of origin voluntarily, he seems reluctant to deal fully with the concomitant complexities that this raises, as highlighted by Carens (2000: 55). Consequently the viability of Kymlicka’s proposal rests on ‘too neat’ (Parekh 2000: 103) a distinction between immigrants and citizens, and cannot accommodate categories of individuals such as guest workers, refugees, slaves, and descendents of colonizing powers (Benhabib 2002: 63).
But this ‘neatness’, however fictive, is crucial for Kymlicka if he is also to suggest that his proposal respects the autonomy of minority groups, despite the fact that he does not think that they warrant self-government rights. This is because Kymlicka portrays immigrants as generally having voluntarily left behind their original societal cultures, and by implication, signaling their intention to become part of the new host country’s societal culture. The kind of language training Kymlicka suggests is aimed at facilitating this. In this scenario, then, autonomy is not compromised as long as we (i) are prepared to believe that every immigrant actually wants to replace his or her original culture with that of the host country, and (ii) are willing to ignore inconvenient categories such as refugees or slaves.
Rubio-Marín’s (2003) proposal in turn is, simply put, Kymlicka’s polyethnic rights writ large, in that each minority community is allowed to maintain its own language and culture because these are seen to serve primarily noninstrumental purposes. Nevertheless, in the shared public domain that the state oversees, everyone has the right and the concomitant duty to use the dominant language for instrumental purposes. Rubio-Marín’s conceptual distinction between noninstrumental and instrumental purposes is critical in seeing how her proposal deals with equality and autonomy. Each minority community has the autonomy to decide on community-specific affairs, to the extent that these do not work against the larger national interests. This includes preserving the community’s own language so long as this does not mean neglecting one’s duty and right to learn the dominant language. In this way, each community has equal status relative to every other. However, each community is also equal, relative to the state, because the public (p.42) arena of the state is envisioned as culturally neutral, bleached of any ethnic bias by virtue of the fact that the use of language here is ‘purely’ instrumental.9 Admittedly equality (of participation) in the shared public arena is achieved at the expense of autonomy by requiring all members of the society, regardless of their ethnic community affiliations, to learn the dominant language. This is problematic because the main sticking point for many language rights advocates is the fact that native speakers of the dominant language already use their language for both instrumental and noninstrumental purposes. Nonspeakers therefore carry the extra burden of having to learn the dominant language, and in Rubio-Marín’s proposal, they still do. Rubio-Marín’s proposal therefore seems to justify the very situation of linguistic inequality that most language rights advocates would consider in need of redress.
Neither Kymlicka (1995) nor Rubio-Marín (2003) provides much discussion about the complexity of language.10 Minority speakers are described as either learning the dominant language or not—they either retain their own language or not. Furthermore, each language is assumed to be clearly associated with a specific and distinct cultural community. What is missing is an attempt to come to grips with the sociodynamics of contact situations—surely a key consideration in plural societies—where, as speakers of different languages interact, hybrid varieties can emerge. The rights-based proposals of Kymlicka and Rubio-Marín are unable to accommodate hybrid varieties because their proposals presume that languages are preexisting, and hence preowned by specific communities. The status of new linguistic varieties that come about, for example, as a consequence of contact between immigrants and their new ‘host’ societies is not addressed. And addressing this is critical because it reminds us that the relationship between minority communities and the host society is not a simple dichotomy between retaining the minority language or not, or learning the dominant language or not. Hybrid varieties often emerge when different languages come into contact with each other, even when speakers of the minority language are attempting to learn the dominant language.
(p.43) Learning of the dominant language can take place both within as well as outside the confines of a formal learning environment, such as an educational institution. Let us focus first on the ‘informal learning’ that comes about as a result of interacting with members of the dominant language, or by simply being immersed in an environment where the dominant language is pervasive. In such a situation, a contact variety is more likely than not to arise, one that, while bearing some similarities to the dominant language, is sufficiently distinct to be considered a socially marked variant. Stroud’s (2004) discussion of Rinkeby Swedish provides a relevant illustration of these issues.11 As noted in chapter 1, Rinkeby Swedish is a contact variety of Swedish, used in reference to the Swedish spoken by immigrants. The invocation of Rinkeby Swedish as a characteristic of immigrant speech thus politicizes the boundary between ‘real’ Swedes and others, where this variety is presented as a clear indication of ‘contagion, inauthenticity and undesirable contact and change’ (Stroud 2004: 200). The immigrant, marked by the Rinkeby Swedish shibboleth, is therefore neither able to fully assimilate and partake of mainstream institutional resources, nor able to claim an authentic Swedish identity. In this way, public discourses around Rinkeby Swedish provide a ‘powerful but subtle means for the exclusion and stigmatization of migrants in Swedish public spaces, at the same time that the significance of speaking Swedish is resymbolized’ (Stroud 2004: 197).
All this is deeply ironic because the immigrants are being penalized for doing exactly what Kymlicka (1995) and Rubio-Marín (2003) ask of them: learning the dominant language. But even as Kymlicka and Rubio-Marín expect immigrants to learn the dominant language, they make no serious attempts to address the linguistic difficulties that learners are likely to encounter, in particular, the denigration of hybrid varieties. Rubio-Marín (2003: 63), for example, says little beyond noting that speaking the dominant language poorly can lead to stigmatization of the speaker as being illiterate. The implication is that such stigmatization can be avoided by learning the language well. But, of course, whether one is speaking a language well or poorly is, ultimately, an evaluative characterization that depends significantly on the ratification of those in power, as is the case with Rinkeby Swedish.
The widespread tendency to view hybrid varieties as linguistically deficient is also relevant when we consider language learning in formalized educational settings. Since knowledge of other languages can influence the learning of a target language, a common educational response is to ban other languages from classroom contexts, especially if these are already viewed negatively in wider society (Siegel 1999). Such a response simply assumes that the stigmatized variety, which is often the first language for the students, has no contribution to make toward learning the target language. Even worse, such a response not only dismisses the variety as useless, but by doing so, also denigrates their speakers and violates the fundamental (p.44) pedagogical principle that students’ existing (linguistic) knowledge can and should be used as a resource to help them learn more effectively (Siegel 1999).
Hybrid varieties thus pose significant problems for language rights. Being the result of intergroup contact, they cannot be easily claimed under either Kymlicka’s (1995) self-government or polyethnic rights, both of which assume that the community must have had a historical association with the language. Rubio-Marín’s (2003) proposal runs into similar problems, since hybrid varieties—by virtue of their stigmatization—cannot be said to serve the noninstrumental value of protecting a community’s identity. This is because hybrid varieties are often considered to be deficient dialectal variants of ‘proper’ or ‘fully developed’ languages, even by their own speakers. It is therefore unlikely that the speakers would want these varieties to be accorded the protection of instrumental rights, even if this was an option. For the same reason of low prestige, neither can such varieties overcome the linguistic obstacles to participation in wider society that Rubio-Marín’s instrumental rights are intended to redress. In fact, because of social prejudice, these hybrid varieties become seen as the obstacles themselves that need to be overcome. Thus, whether we are looking at language learning in formal or informal situations, these rights-based proposals are unable to come to grips with the problems posed by hybrid varieties.
Notwithstanding the fact that their stigmatized status means that most speakers would not want to have the ‘right’ to hybrid varieties, there are other problems that we should be aware of. First, recall that we are talking about a linguistic variety that arises as a result of contact, rather than one that preexists the contact situation. This is somewhat problematic for the idea of language rights, since the object of the rights discourse here is an anticipated entity, one that does not yet exist. This might not be too serious a problem since we do sometimes talk about the rights of future generations of human beings, especially in relation to environmental issues of sustainability. Since the rights-holders are imagined future entities, advocates of language rights may want to argue that the bequests to rights-holders can also be future entities. One might therefore imagine making an argument that should a new linguistic variety arise as a result of contact, then minorities should be able to lay claim to this variety. This suggestion is not without its own problems, though. Because the variety is a hybrid, then speakers of the dominant language—from which the hybrid is likely to obtain most of its lexicon—may also feel that they too (and not just the minority communities) have an equal right to it. This is a problem for the notion of language rights, which typically focuses on trying to ensure that a group’s language is accorded recognition. The assumption here is that there is an unambiguously identifiable rights-holder. The notion of language rights says little about how to handle a situation where different parties lay claim to what is ostensibly the same linguistic object, since the question then arises as to which party ‘really’ has the right to this object.
But even if rights advocates were to somehow provide guidelines for identifying which of the parties involved in a hybrid variety is the ‘real’ rights-holder, there is another problem. The hybrid variety, while stigmatized because of its association with minorities, can gain a degree of prestige once it is associated with (p.45) members of the dominant community. Recall from the Rinkeby Swedish example (see chapter 1) that native Swedes can appropriate some Rinkeby Swedish constructions as ‘trendy’ or ‘hip’, while at the same time, the constructions spoken by immigrants continue to be stigmatized (Stroud 2004: 204–6). Thus different speakers do not merely lay claim to the ‘same’ variety. Some speakers are in a better position than others to transform the variety, legitimizing it from ‘broken’ or ‘error-driven’ to ‘novel’ or ‘innovative’, and thus bestowing upon it the status of a cultural capital from which symbolic profits can be extracted (Bourdieu 1991). Addressing this requires that we attend to the issue of political power involved in defining what counts as legitimate uses of a language—including what particular linguistic constructions are being appropriated in what contexts—rather than just focusing on the concept of language simpliciter. As Duchêne (2008: 11) puts it:
Language interactions in plurilingual contexts thus bring about pragmatic, specific and contextualized ways and means. The utilization of language forms is balanced by strategies that are not homogenous and that constantly modify the normalized state of the language. Internal or external migration situations create new language forms that are progressively incorporated into the dominant language of a region, without the native speakers necessarily realizing it (cf. Grosjean and Py 1991).
And this brings us to what is perhaps the most serious problem with the notion of language rights: the ontological assumption that there is an identifiable linguistic variety that can be coherently treated as the object of a rights discourse. As mentioned in the preceding chapter, Stroud’s discussion of Rinkeby Swedish deals with a situation in Sweden where, after a period of time, there arises the general perception that there is a pan-immigrant speech that can be attributed to immigrants. When Swedes label this speech ‘Rinkeby Swedish’, they give it the appearance of ontological stability, which allows it subsequently to become the object of further public discourses and debates. But, of course, linguistic discrimination need not always occur in the context of an established variety. Perceptions that a particular speech form is ‘bad’ and its speakers ‘illiterate’ can occur even before multiple speech performances by multiple communities of immigrants have ‘congealed’ (Butler 1990: 33) to give us the impression of a ‘stable’ linguistic variety. In this way, the recognition of a linguistic variety occurs only after myriad acts of linguistic discrimination have culminated to the point where the formation of a linguistic stereotype is possible. So language rights advocates also need to address the problem of linguistic discrimination that may already precede the (subsequent) naming of a variety. An approach that advocates language rights, however, is likely to find it extremely difficult to deal with such a situation, where sociolinguistic categories and practices are not only unstable, but so incipient that no clear named varieties can yet be said to exist (cf. Clifford 1988: 338–39; Ford 2005: 71).
The kind of hybridity that is characteristic of contact situations and the problems it raises for language rights should in fact not be considered exceptional; (p.46) rather it should be treated as a characteristic of languages in general (Makoni and Pennycook 2007). Hopper (1998: 157–58; italics added) describes the matter thus: ‘there is no natural fixed structure to language … Systematicity, in this view, is an illusion produced by the partial settling or sedimentation of frequently used forms into temporary subsystems.’
In the case of linguistically diverse societies, it is therefore not possible to assume that there are at least three distinct linguistic varieties that exist in separate compartments: the dominant language A, the minority language B, and the hybrid variety C. The emergence of C (as the result of the interaction between A and B) does not necessarily leave A and B themselves untouched. Rather, we have to recognize that A, B, and C all continually undergo changes as a consequence of being in contact with each other. As such, any attempt to talk of three distinct isolable varieties is really a convenient fiction that may not really reflect the linguistic practices of the various speakers.
It should be noted that by making this observation, I am not suggesting that named varieties should be dismissed. On the contrary, I am fully aware that names of varieties, while not neatly corresponding to a coherent linguistic entity, carry ideological significance that must be accounted for. The desire of speakers to associate particular constructions with labels such as ‘Japanese’, ‘English’, or even ‘Indian English’ or ‘Marathi English’ are inevitably informed by broader ideological understandings of what these labels entail. If there exists a group of speakers for whom the labels ‘Indian English’ or ‘Marathi English’ reflect some sociolinguistic reality, this is a phenomenon that is in turn indicative of these speakers’ (growing, emerging) metalinguistic awareness that there are shared commonalities in their linguistic practices. At the very least, it is an indication of the speakers’ expectations that such commonalities do or should exist. As an in-group metalinguistic label, the name of a variety very possibly orients the linguistic practices of these speakers toward each other and away from those whom they might consider nonmembers of the group. Effects on grammatical structure can be plausibly expected as the manifestations and interpretations of linguistic forms are gradually adjusted or calibrated in the course of recurrent interactions. But it is precisely because the correlations between practices and labels are subject to changes and contestations that we should be careful about advocating the use of language rights.
I began this chapter by demonstrating that for the concept of a right to be coherent, clear boundaries have to be drawn between different rights-bearers. Demarcations also have to be made between the objects of rights, that is, the ‘things’ that rights-bearers are entitled to. Extrapolating this discussion from rights in general to the case of language rights in particular, similar considerations led us to ask whether it is likewise possible to impose boundaries between different holders of language rights and between the objects of those rights, which are, presumably, specific (p.47) languages. I argued that there are several difficulties involved in treating a language as the object of a right, since this would require homogenizing what are in fact variable practices across individuals. These difficulties were illustrated via the discussion of three effects: selectivity (where only a few selected practices are privileged over others), reinvention (where the pressure to come up with appropriate practices can lead a group to modify its practices so as to fit the dictates of a rights-based discourse), and neutralization (where, once locked into the rights-based discourse, there is no clear ‘exit strategy’, thus making it difficult for a rights-based claim to work as a temporary tactic). The chapter ended by discussing two properties of language—unavoidability and hybridity—which, taken together, pose strong challenges to the idea of language rights.
(1.) Interesting complications arise in the case of the Eurasian community, which has attempted to claim English as its mother tongue, a move rejected by the government since this is ‘essentially’ a Western language. See Wee (2002a) for a detailed discussion.
(2.) These include Cantonese, Hainanese, Hakka, Hokkien, Khek, and Teochew. Following the parlance of the Singapore government, the term ‘dialects’ here refers to Chinese languages other than Mandarin.
(5.) This is not necessarily an undesirable outcome. There are good reasons why a rights discourse ought to encourage essentialism, since it aims to protect specific attributes or conditions that are considered especially significant. But precisely because of this, we have to be careful about how broadly we are prepared to extend the notion of rights and, in particular, whether language warrants such an extension.
(6.) Auer (1999) suggests that there is a unidirectional continuum from code-switching to language mixing to fused lects. This progression is contingent, in the sense that it is not necessary for the move from one end of the continuum to continue all the way to the other. The main difference, as Auer (1999: 321) points out, is between code-switching and language mixing, on the one hand, and fused lects, on the other. The difference lies in the possibility of alternation. With code-switching and language mixing, speakers still have the option of deciding whether or not to juxtapose elements from different varieties. With fused lects, speakers no longer have such an option; the drawing together of elements from erstwhile different varieties has stabilized to the point where it constitutes a grammatical obligation or constraint.
(7.) Speakers who code-switch are not always aware that they are doing so (Blom and Gumperz 1972). Thus, requiring speakers to constantly monitor their speech for violations of ‘purity’ (as in the Sri Lanka example discussed previously) only impedes communicative fluency, which depends significantly on the use of routinized and formulaic expressions (Nattinger and DeCarrico 1992; Wray 2002).
(8.) Rights and duties can be treated as correlative (Waldron 1993: 32), where one entity’s right involves a duty (on the part of some other entity) to ensure that this right is observed. And since Rubio-Marín is especially concerned with public interaction in a shared domain, where everyone is potentially expected to interact with everyone else, then the same right and duty (to the learn the dominant language) is visited simultaneously on each and every member of the society.
(9.) Rubio-Marín (2003) also relies on making a clear distinction between the culturally inflected minority community and the culturally neutral public arena to sustain her claim that noninstrumental rights are group rights while instrumental ones are individual rights. Needless to say, the realistic existence of a culturally bleached public arena untainted by issues of culture is highly questionable. At the very least, it requires us to believe that individuals are capable of leaving behind their cultural habitus once they enter the public arena and make their decisions in purely rational terms, unaffected by their gender, class, or ethnic backgrounds—a belief already sharply criticized by Bourdieu (1977; see also Jenkins 2002).
(10.) This is not an uncommon problem. For example, De Schutter (2007: 3) observes that political philosophers tend to have an ‘outdated empirical understanding of the concept of language itself. I will call this a linguistic distinctness understanding. This notion of linguistic distinctness takes the world to be a neat patchwork of separate monolingual geographical areas almost exclusively populated by monolingual speakers. As a result, many of these philosophers, often unaware of important sociolinguistic and other research on these matters, endorse an ill-conceived notion of the concept of language on which they build their theories’.
(11.) It is important to bear in mind that even though the example of Rinkeby Swedish involves immigrants, the point about hybridity is relevant to any kind of contact involving speakers of different languages, including national minorities.