Normative theories of multiculturalism have come under attack in recent years for a number of reasons. Many critics challenge the positive arguments offered on behalf of multicultural policies, questioning whether the values of liberal democracy entail that the state ought to recognize and accommodate the distinctive concerns of minority cultures (Barry Reference Barry2001; Waldron Reference Waldron1992). Others highlight the potential costs of multicultural policies. Even if such policies do promote liberal democratic values in some respects, they compromise those same values in other respects. Familiar formulations of this second form of critique suggest that multiculturalism is “bad for women,” that it ghettoizes vulnerable minorities, and that it fosters intolerance and extremism (Barry Reference Barry2001, 103–9; Okin Reference Okin1999; Sniderman and Hagendoorn Reference Sniderman and Hagendoorn2007).

In this article, I consider a distinct kind of challenge to multiculturalism, a challenge that has been around for a while but has received some of its most forceful statements within the past decade (Appiah Reference Appiah2005, chap. 4; Barry Reference Barry2001, 7, also chap. 7; Benhabib Reference Benhabib2002; Scheffler Reference Scheffler2007). According to a growing number of critics, a fundamental problem with normative theories of multiculturalism is their reliance on a concept of culture that is empirically and morally naïve. The theory of multiculturalism is founded on an “essentialist” picture of cultures as determinate, bounded, and homogeneous, a picture that is empirically false and morally dangerous. When cultures are conceptualized as fluid, interactive, and overlapping and as internally contested and heterogeneous, they become more acceptable empirically and normatively. But then culture is no longer serviceable within a multicultural framework. Defenders of multiculturalism are left without a culture concept that allows them to make judgments about the treatment, survival, and revival of distinct minority cultures.

I shall call this challenge the dilemma of essentialism. According to the dilemma,

Whichever horn of the dilemma one opts for, the upshot is that normative defenses of multiculturalism should be dismissed.

My aim in the present article is to defend multiculturalism against this challenge by arguing that the dilemma of essentialism is false. I do this by grabbing the second horn of the dilemma and showing that it is possible to elaborate a plausible nonessentialist concept of culture that is serviceable to the normative agenda of multiculturalism. By “plausible” I mean a concept that is responsive to both empirical and normative desiderata. To be plausible, a concept should pick out as sharing a culture at least some of the groups that multiculturalists have in mind when they make their normative claims. And the concept should help us to see—in at least some cases—why culture is valuable to people. It should help us to make sense of the idea that, in a central range of cases, it is good for people when their cultures are respected and preserved.

Although nothing guarantees that there is a nonessentialist culture concept that satisfies both of these desiderata, I argue that there is such a concept. Culture, I propose, is what people share when they have shared subjection to a common formative context. A division of the world, or of particular societies, into distinct cultures is a recognition that distinct processes of socialization operate on different groups of people. Because culture on this view is the “precipitate” (to borrow Kluckhohn's (Reference Kluckhohn1950, 34) term) of a common social lineage, I shall sometimes refer to this as the “social lineage account” of culture. Although there may be some weak sense in which the social lineage account remains “essentialist”—it does say that members of a culture uniquely share a particular property, namely, the property of having been shaped by a common set of formative conditions—it is not essentialist in an objectionable manner.Footnote ¹ It is compatible with, and indeed helps to explain patterns of heterogeneity, contestation, and hybridity that commentators have rightly emphasized in pressing the critique of essentialism.

The next section takes a closer look at the problem of essentialism, explaining why it is such a serious issue for multiculturalism and how existing responses to the problem have fallen short. The article's main conceptual proposal is developed over the course of the three sections that follow. The article then considers why culture matters normatively. Although I leave this issue for the end, and only briefly sketch some considerations, it is a crucial part of the response to the dilemma of essentialism. The dilemma does not say that there could be no nonessentialist culture concept. Rather, it says that there is no such concept that is supportive of multiculturalism. To respond effectively, it is necessary to show how the social lineage account might be called upon in support of multiculturalism.

CRITIQUE OF ESSENTIALISM

Any attempt to articulate a culture concept for multiculturalism has to address the important critique of cultural “essentialism” that has been developed in recent decades by anthropologists and political theorists. In general, essentialism consists in the identification of kinds in the natural or social world through the singling out of some relevant property (or set of such properties) that are possessed by all and only the individuals who belong to that kind. The critique of essentialism consists in pointing out that the individuals belonging to the various kinds that are commonly supposed to exist do not, in fact, uniquely share a relevant property or set of properties. There is both variation within the members of the putative kind, and commonality between members and nonmembers.

Applied to culture, the critique of essentialism amounts to the argument that all of the usual features that are taken to define culture run foul of the problems of internal variation and external overlap: The relevant features are not shared by all and only the members of the groups that are generally said to share cultures. The critique is often thought to follow from, or be associated with, certain commonplace observations about human beliefs and practices. In groups of any size, beliefs and practices are heterogeneous and contested. They change and fluctuate over time. And they are formed interactively and dialogically with members of other groups, often taking on a recognizably hybrid character as a result.

To see the critique of essentialism in action, consider the familiar conception of culture as a shared framework of beliefs and values. There are actually two versions of the familiar conception, the first of which requires that all and only the members of a distinct culture hold some particular set of beliefs and values.Footnote ² If this is the proposal, however, then it is obviously a nonstarter. In virtually any of the groups that are usually thought to share a culture one can find significant variation of beliefs and values. There is unlikely to be any significant belief or value, or set of beliefs and values, that is shared by all and only the members of the group. A closely related problem is variation over time. The familiar view seems to imply that a distinct culture would cease to exist whenever its members revised or abandoned the beliefs or values that constituted it. This implication leaves insufficient space for the distinction between cultural disappearance and cultural change.Footnote ³

A different version of the familiar conception takes a cue from Clifford Geertz's (Reference Geertz1973, 12) dictum that “culture is public because meaning is.” Geertz draws on well-known accounts in the philosophy of language to argue that the meaning of words is settled publicly, by social conventions, and not by agreement of private beliefs. In the same way, he suggests, practices and established forms of social behaviors are associated with particular meanings that are settled publicly. Winking (to use Geertz's example, borrowed from Ryle) has a specific meaning in a given context, whether the winker has certain beliefs in his or her head or not. The suggestion, then, is that cultures might be defined as the “socially established structures of meaning” in terms of which people engage in social behaviors of various sorts. In this picture, the relevant framework of shared beliefs and values for identifying distinct cultures is not identified by inspecting the private beliefs and values of different individuals, but by recovering through interpretation the beliefs, values, and meanings that are embodied in particular practices and institutions. The implication is that a society can be said to contain several distinct cultures when it is home to several distinct public structures of meaning of this kind.

But once the claim is put in this way, it immediately looks vulnerable to the same challenges that undermined the first version of the familiar conception. The main problem is that meanings in a society are typically contested. For many practices and bits of social behavior, there will be several publically established meanings that people enact in their behavior (Wedeen Reference Wedeen2002, 716). There is a socially established structure of meaning only if “structure” is meant to encompass the possibility of difference and disagreement. But making this move produces a dilemma. Are lines of cultural difference demarcated by these differences of signification, or can a single culture be home to divergent understandings of the relevant meanings?

If the former, then the contours of shared cultures are going to look nothing like the contours of the groups that are typically thought of as cultures: Those groups contain a great deal of disagreement about meanings, often because at least some of their members interact in various ways with other groups and are attracted, in at least some measure, to the meanings and forms of signification that predominate in them. If, on the other hand, it is allowed that cultures can be home to divergent understandings of the relevant meanings, then the conception does not yet offer any basis for distinguishing different cultures.Footnote ⁴ Why not just say that the whole society (or the whole world) contains one single culture, albeit a culture that is subject to considerable fragmentation of meanings and values? A similar problem arises with respect to the question of disappearance vs. change. Does the culture disappear every time one or more of the meanings embodied in its practices is revised? Presumably not. But then how do we determine when a culture has disappeared and when it has been preserved?

These arguments retrace the familiar critique of essentialism. To be clear, the objection is not to the idea that there are shared beliefs and values, or to the idea that there are socially enacted, publicly established meanings. Rather, the point is that commonality and variation in these beliefs, values, and meanings do not track the cultural differences that are commonly supposed to exist. There is variation within, and overlap between, these putative cultures.

If familiar conceptions of culture fail to satisfy the empirical criterion stipulated earlier, they also do poorly by the normative criterion. Attempts to protect distinct cultures risk being oppressive to the many people within groups who do not, in fact, hold the beliefs and values that are supposedly constitutive of the group's culture (Benhabib Reference Benhabib2002, 4; Mason Reference Mason, Laden and Owen2007, 227). Moreover, if the supposedly constituting values are illiberal, then the familiar concept would end up pitting cultural preservation against liberalism. A culture would have to maintain its illiberal values and practices, or disappear altogether (Kymlicka Reference Kymlicka1989, 168–70).

One possible reaction to the critique of essentialism is that it must go wrong at some point, because people clearly do engage in actions that are “cultural.” They communicate with one another, attempting to exchange, to negotiate, to validate, to subvert, and otherwise to manipulate meanings and symbols. For these actions to be possible, it is occasionally argued, there must be some of kind of shared framework in the background that provides a common vocabulary and set of meanings. For instance, Kymlicka responds to a version of the critique of essentialism (pressed by Waldron Reference Waldron1992, Reference Waldron, Catton and Macdonald2004) by arguing that “cultural materials” are only “available” or “meaningful” to people “if they become part of the shared vocabulary of social life—i.e. embodied in the social practices, based on a shared language, that we are exposed to” (Kymlicka Reference Kymlicka1995, 103; see also Kymlicka Reference Kymlicka1989, 165; Raz Reference Raz1994, 176–77).

But this response to the critique of essentialism remains unconvincing. From the fact that people exchange meanings with one another, it does not follow that there must be some framework of shared meanings or practice hovering in the background. In reality, only a minimal amount of commonality is needed to get a fruitful discussion, or an exchange of meanings and signs, off the ground. Persons who, if the world were to be divided into discrete cultures, would clearly not be considered to share a culture can still improvise a meaningful conversation with one another, so long as there are sufficient resources or situational clues in the immediate context for interpreters to form plausible theories about what speakers are trying to convey (Davidson [1986] Reference Davidson, Ludwig and Lepore2006; Popper Reference Popper1976; Waldron Reference Waldron, Catton and Macdonald2004). This is sometimes called “cross-cultural” or “intercultural” dialogue.

Not surprisingly, many theorists of multiculturalism acknowledge the force of the critique of essentialism, and opt for the second horn of the dilemma of essentialism. Joining the chorus of critics of cultural essentialism, they announce that their defense of multiculturalism will be grounded in a nonessentialist culture concept. I am obviously sympathetic to this general strategy for responding to the dilemma, but the attempts so far to execute the strategy seem unsatisfactory. Many of the theorists adopting this approach do a good job of presenting the critique of essentialism, but then become painfully circumspect about what precisely the alternative conception is supposed to be. The effect is to reinforce the suspicion voiced by critics of multiculturalism that essentialism is continuing to play an unacknowledged role in the view's foundations.

In his early work, for instance, Kymlicka introduces a distinction between cultural “character” and the cultural “structure” or “community,” which is designed in part to guard against the critique of essentialism. The cultural structure is something that persists even while the character of the culture changes. Quebec's culture, to use Kymlicka's example, survived the dramatic transformations of the Quiet Revolution of the 1950s and 1960s. Although the example seems promising, Kymlicka never tells us very much about how to identify cultural structures. How do we know if a society contains one or several, if it is not by looking for frameworks of meaning that have distinctive characters? Kymlicka concedes that defining cultural communities is a “vexed problem” but insists that “we know that such communities exist” (Reference Kymlicka1989, 168 fn. 2). Given that challenges based on the dilemma of essentialism continue to be repeated, a fuller account is clearly needed.

The failure to elaborate an alternative to an essentialist view of culture is even more apparent in Tully's 1995 book Strange Multiplicity. The opening chapter contains a memorable statement of the critique of essentialism, directed against what Tully dubs the “billiard-ball” concept of culture. Elsewhere in the book, however, Tully routinely talks about cultures and their continuity as if his opening critique did not present weighty reasons for abandoning multiculturalist claims about culture altogether.Footnote ⁵ He claims in various places to be working with a “dialogical,” or “intercultural,” conception of culture, but it is never clear what exactly the concept of culture is that the adjectives “dialogical” and “intercultural” are meant to modify.

A third example of the simultaneous problematization of, and reliance on, the culture concept can be found in Phillips's (Reference Phillips2007) book Multiculturalism without Culture. Much of her argument proceeds squarely within a framework that is critical of multiculturalism. She rehearses the standard critique of culture as essentialist and reifying, and argues that narratives of culture are often harmful to women and others. But Phillips cannot quite bring herself to drop multiculturalism in favor of a cosmopolitan focus on individuals. In part, she has various unrelated objections to existing formulations of cosmopolitanism. But she also finds a “kernel of truth” in the multicultural claim that “majority and minority cultural groups” should be treated equally and regards this as a good reason not to abandon the discourse of multiculturalism (71–72). The puzzle about Phillips' position here is what entitles her to talk of cultural groups in this way given her explicit antiessentialism (see also 167).Footnote ⁶ What concept of culture is at work in the distinction between majority and minority groups and in the judgments about how such groups are treated?

In sum, then, the critique of cultural essentialism is a deceptively powerful one, and existing theories of multiculturalism do not tend to do a very good job of responding to it. Although many proponents of multiculturalism now recognize the problem, they end up tiptoeing verbally around it, rather than fully articulating an alternative conception of culture that avoids the pitfalls of essentialism. The remainder of the article seeks to elaborate just such a conception.

CULTURAL CONTINUITY

Normative multiculturalism assumes that it is possible to both identify distinct cultures and to distinguish between cases of cultural change (which presuppose an underlying continuity) and cultural disappearance. Without a means of identifying distinct cultures, judgments about how cultures are faring become impossible. And if there were no way to distinguish change from disappearance, there would be no way of judging the results of multicultural policies, and perhaps no reason to embark on them in the first place.

It is tempting to collapse these two problems of individuation and continuity into one, and to say that C2 is continuous with C1 just in case C2 and C1 are one and the same culture. But collapsing the two problems leads to a difficulty.Footnote ⁷ It is plausible to think that both the French culture, circa 1800, and the French-Canadian culture, circa the same date, were continuations of the French culture, circa 1600. But, by 1800, the French and the French-Canadian cultures were clearly not one and the same. And, if they are not the same as one another, then they cannot both have been the same as French culture in 1600. This suggests that there must be some different interpretation of “being a continuation of” that is weaker than “being identical to.” I begin by trying to identify this weaker notion of continuity, and then turn to individuation in the next section.

A passage in Kymlicka provides a helpful place to start. In his previously mentioned remarks on the distinction between a culture's “structure” and its “character,” Kymlicka notes that it is “right and proper that the character of a culture change as a result of the choices of its members,” but things are different when the “very survival of the culture” is jeopardized “as a result of choices made by people outside the culture” (Reference Kymlicka1995, 104; see also 1989, 167). Although not presented as a conceptual claim about the difference between continuity and disappearance, Kymlicka's suggestion points to a way of understanding that difference. Perhaps continuity involves relations of choice and disappearance their absence? When the abandonment of some relevant form of belief or practice, and its replacement with some other, is the result of the choices of members of the culture, then we can say that the culture has been continued (in changed form). When it is the result of conditions imposed from the outside, then it is better to say that the culture has ceased to exist.

The appeal of this suggestion is obvious. It aligns cultural preservation with individual choice and thus with something of clear normative significance. But, as is, the proposal suffers from the fundamental difficulty that cultures sometimes disappear as a result of the choices of their members. Choice does not always imply continuity. An example illustrates the problem with the choice-based proposal, and also points the way to a better one. The case is borrowed from John Terborgh (Reference Terborgh2002), who presents it in the course of a discussion of cultural loss:

Terborgh concludes that “here culture is being lost.” In part he means that Misael's children are losing their ancestral culture. Raised and educated in a setting where the Machigeuenga people are not in the majority, they are quickly integrated into the multiethnic, Spanish-speaking, motorized-boating population that makes up Boca Manu. At the same time, Terborgh sees Misael's choice as fairly representative. In a generation or two, as many individuals make choices like Misael's, Machiguenga culture itself will disappear.

Misael's story illustrates why the distinction between cultural continuity and loss is not merely a matter of choices being made or not made by members of the culture. Misael does seem to be making a choice of sorts. In Terborgh's recounting, there is no suggestion that he is driven by poverty or oppression to flee his ancestral village; indeed, his skill at a locally valued activity is “legendary.” Misael is moved by the prospect of a better education and better economic opportunities for his children. If we agree that Misael's culture is about to be lost, then we are forced to recognize that individual choice is insufficient to establish cultural continuity.Footnote ⁸

So how then should we understand the opposing ideas of continuity and loss? Terborgh's reflections on this conceptual question are problematic. In support of his contention that Machiguenga culture is lost to Misael's children, he points to various practices that they will not engage in, skills that they will not possess, and forms of thought and feeling that they will no longer experience. Misael's children will not share their father's facility with the bow and arrow, they will speak Spanish rather than Machiguenga, they will not “think like a Machiguenga,” and so on. But as Yu and Shepard (Reference Yu and Shepard2003) note, Terborgh seems to be relying on a “notion of culture as a static, species-like entity.” They point out that, although Machiguenga culture is now undergoing a major transformation, this has been true throughout its history. The Machiguenga culture has always been in a state of transformation or flux. It has continuously been formed and reformed through interaction with the Incas, the Spanish, the Peruvian state, Protestant missionaries, the rubber and oil companies, the panindigenous movement, and so on. In other words, Terborgh's conception of cultural loss does not deal well with the critique of essentialism. Terborgh (Reference Terborgh2003) replies by suggesting that these earlier transformations had been “incremental” rather than “fundamental,” whereas what is happening today, as Misael's children and others like them “forsake traditional life to partake in Peru's public education system,” is a fundamental break with historical continuity. But, by itself, this reply risks underestimating the possibility that a culture might survive even a fundamental change in its forms of thought and practice.

Still, I think that Terborgh's intuition about the Machiguenga is basically correct, and that his own account of their situation contains the resources needed to develop a defensible account of cultural continuity and loss. The lesson to draw from Terborgh's story is not that cultural loss occurs whenever people abandon fundamental aspects of their thought and practice but that it occurs when there is a significant disruption of the processes by which the culture is transmitted from existing members to new generations and other newcomers. The facts in Terborgh's story—especially the emphases on migration, schooling, and language—fit this “disruption-of-transmission” account of cultural loss as well as, or better than, they fit the “loss of fundamentals” account presupposed by Terborgh himself.

What does it mean for the transmission of a particular culture to be “disrupted”? The suggestion adapted from Kymlicka is that choice is the key factor, but we have seen that this suggestion fails to explain an important case. A factor that does a better job of explaining the range of relevant cases is control. If existing members of a culture control the socialization of some set of newcomers (e.g., the next generation), then to that extent there is no disruption, and the culture of the socialized group can be considered continuous with that of the group that socialized it. If, on the other hand, the existing members of the culture lose control over the socialization of a next generation, either because they choose to surrender it or because it is taken away from them, then the transmission of their culture has been disrupted. There is no group whose culture can be considered a continuation of the culture of the existing group.

By “socialization” I have in mind a broad range of different formative processes that work, in one way or another, to shape the beliefs and values of the persons who are subject to them. Socialization often occurs through participation in particular institutions and through exposure to particular practices and forms of social behavior. The family is obviously an important institution of socialization, as are the schools, the workplace, institutions of government and public administration, the media, popular culture, and even language and forms of discourse.

An institution or practice involved in socializing a new generation is “controlled” by the members of a particular cultural group if members of that group occupy most of the key positions of power and responsibility in that institution or practice, and/or if most of the other participants are incipient members of the cultural group in question, in the sense that they have already received some socialization into that group in other contexts (e.g., the family). Thus, the schools are controlled by members of culture C if the majority of teachers and key decision-makers in the schools are members of C, and/or if a significant number of the students in the schools hail from family backgrounds in which one or both of their parents belong to C.

A number of the key facts in Misael's story are facts about socialization processes. Misael's decision to move to the town makes his children subject to a number of new socializing influences. If my proposal is correct, then it is exposure to these new forms of socialization, and the disruption of the previous processes in which the beliefs and practices of members of the Machiguenga culture had been transmitted to new generations, that accounts for our sense that the culture is on the brink of disappearance.

In their original community, Misael and his fellow Machiguengas used the Machiguenga language and were continuously exposed to the concepts and forms of discourse that happened to predominate in the language at the time. Their children might have gone to school in a setting where at least some of the teachers and officials, and perhaps all of the children, come from Machiguenga backgrounds. Many of them would go on to live, and work, and worship alongside other Machiguengas. By contrast, once Misael and his family arrive in Boca Manu, they begin participating in formative practices and institutions that transcend the Machiguenga community. Of course, they bring with them whatever Machiguenga socialization they received prior to their departure. They also retain their family environment, and they may maintain some contacts with their home community and with a local diaspora. But notwithstanding these continuities, Misael's children will soon find themselves subject to a powerful set of socializing institutions and practices that are largely administered and populated by members of the non-Machiguenga majority. These include the economy, political institutions, and education system of the town (and perhaps the larger Peruvian society that the town is part of); the language (Spanish), narratives, and forms of discourse that predominate in the town; the main forms of popular culture and entertainment of the town; the commercial and neighborhood life of the town; and so on. Misael's children will go to school with non-Machiguengas and the teachers will be non-Machiguengas; when they grow older they will work alongside non-Machiguengas, and answer to bosses who are non-Machiguengas. They may marry non-Machiguengas and raise their own children in a manner that is even further removed from a Machiguenga upbringing.

Misael's story is one possible illustration of how cultural disappearance occurs, but there are other routes to the same outcome. In general, migration, economic change, rising literacy, access to new media and forms of popular culture, improvements in transportation and communication, and other broad social changes of this kind all have a tendency to expose people to new formative influences.Footnote ⁹ When, as a result of these processes, people find that traditional socialization mechanisms are no longer operative, but have been superseded by new, larger-scale processes shared with a wider population, we can say that a local culture is being lost.

Moreover, in typical cases, the weakening (if not the outright destruction) of historic cultures is consciously encouraged by nation-building policy makers, who aim to integrate all citizens into a statewide national culture. The state designates an official language, imposes a national school system, requires service in the national military, builds national transportation links, supports national broadcasting media, and designs other institutions of the state to integrate members of minority cultures into statewide processes of cultural transmission. In some cases, states have intentionally disrupted processes of familial and community socialization, e.g., by taking children from their families and communities and placing them in state-run residential schools. Measures of this kind, which are condemned by the Convention on Genocide's definition of genocide, amount to an especially blatant and brutal effort to extinguish a culture.

By contrast, Kymlicka's case of Quebec in the Quiet Revolution is a good example of cultural continuity on the view being proposed. Even though the character of Quebec society changed dramatically in those years, there was never a disruption in the basic processes of cultural transmission. The generation that revolutionized Quebec society was itself socialized mainly by the previous generation of Francophone Quebecers, and would itself go on to control the socialization of a subsequent generation of Francophone Quebecers. Despite the enormous changes in Quebec's values, practices, and institutions, there was never a rupture in the Quebecer-to-Quebecer mechanisms of cultural transmission that operated across a wide range of different areas of human life. The proposed account is also confirmed by the kinds of policies that are typically deployed in an effort to revive and protect vulnerable cultures. These policies often involve seeking control over critical transmission mechanisms, such as education and the media, or over factors that impact these mechanisms, such as access to and administration of the territory of the cultural group in question.Footnote ¹⁰

Notice, then, that we have managed to identify a conception of cultural continuity that avoids the pitfalls of essentialism. There is no reliance on the idea that certain beliefs, values, meanings, practices, institutions, and so on are somehow constitutive of the culture, or that their maintenance or abandonment is key to determining whether the culture continues or disappears. So long as one generation of a culture is controlling the socialization of a new generation or group of newcomers, there is cultural continuity, even if the later generation engages in dramatic revision of prevailing values, meanings, and practices. There is no reason to think, then, that the idea of cultural preservation is committed to the freezing of cultures in any special form or to the reification of particular ideas or traditions as somehow definitive of culture. Contrary to critics who push the dilemma of essentialism, as we shall see further hereafter, a case for multiculturalism can be grounded in a nonessentialist account of cultural preservation.

SOCIAL LINEAGE ACCOUNT

We have been exploring the conditions under which a culture can be said to continue from one moment in time to another. The key consideration, as we have seen, is an unbroken chain of intergenerational cultural transmission. But we have not said quite yet what it is, exactly, that makes something a distinct culture in the first place. In virtue of what can a person be said to share a culture with some people but not with others?

With so much attention having been devoted to the problem of continuity, however, a possible answer to this question is now staring us in the face. Members of a group that can trace back through time a lineage of cultural continuity do share something that others do not share. They share with one another a common experience of socialization that is distinct from, because historically isolated from, the experiences of socialization undergone by others. This fact suggests the following solution to the individuation problem:

Culture, on this proposal, is a kind of precipitate. At any given moment, its content consists in various beliefs, meanings, and practices, but what makes these the beliefs, meanings, and practices of a shared culture is that the people who hold them share a common social lineage.

In defining culture in relation to a set of formative conditions, the point is not that those conditions are causally related to the creation of a distinct culture. Rather, the claim is that the existence of a shared culture is constituted by the exposure by some group of people to a common and distinctive set of formative conditions. Although this suggestion may seem surprising to some readers, it is structurally parallel to the way in which many biologists understand the concept of species. For exactly the sorts of reasons we explored with regard to cultural essentialism, most biologists avoid defining species in terms of essential sets of traits. There is too much variability of traits within, and commonality of traits across, the species that are taken to exist to allow an essentialist account. Instead, the tendency amongst biologists (going back to Mayr Reference Mayr1942; see also Kitcher Reference Kitcher and Harris1999; Reference Kitcher2007; Sober Reference Sober2000, chap. 6) has been to think of species genealogically, as population lineages that are isolated reproductively from other lineages. Although the proposed account of culture is concerned with social and historical rather than reproductive isolation, the structure of the account is the same.

As with all attempts to individuate culture, culture in the social lineage account is both something shared and something distinctive. What is shared, when people share a culture, is exposure to a common set of formative influences. It need not, and generally will not, be the case that everybody sharing a culture has been subjected to identical formative influences. Each person's formative environment will typically include some elements that are idiosyncratic relative to the formative experiences of others in the group. The important point is that the environment also includes some encompassing elements, which are imposed, in common, on all members of the group. In other words, for a culture to be in existence, there has to be some set of formative institutions and practices to which all the members of the group are subject. If there were no such formative processes at work, there would be no group.

What is distinctive about a particular culture is the historical lineage of its formative institutions and practices. Those institutions and practices are, to some extent at least, isolated from the institutions and practices that work to socialize outsiders. They are controlled, that is, by a group of people different from those who control the socialization of outsiders. And the members of the controlling group may themselves have been socialized by a distinctive set of practices and institutions, which were, in turn, controlled by people who were socialized under distinctive conditions, and so on.Footnote ¹¹

As was noted at the outset, the acceptability of any proposed conception of culture depends on how well it satisfies both empirical and normative criteria. The conception should pick out cultures that accord, at least roughly, with the groups that multiculturalists have in mind when they seek protection for, and fair treatment of, distinct cultures (empirical criterion). And, in at least some range of cases, sharing a culture in the sense specified by the conception should be something that matters normatively (normative criterion). I consider the normative dimension later in this article.

As for the empirical criterion, a major advantage of the social lineage account is that it is compatible with both internal variation and external overlap of beliefs, values, and meanings. Internal variation is possible because subjection to a common set of formative influences does not imply that people will end up with a homogeneous set of beliefs or values. The most basic reason for this is that formative institutions and practices are themselves likely to be sites of difference and contestation. Their meanings, values, purposes, and so on will be disputed by the people who administer and participate in them, and, predictably, these disputes will reproduce themselves in new generations, and other newcomers who are brought into their orbit.

Moreover, it would be normal for people who share exposure to even a univocal set of institutions and practices to respond by forming a highly diverse set of beliefs and values. Just as siblings respond differently to a common upbringing, depending on their individual circumstances and characteristics, persons who participate in a common set of institutions and practices also differ in many ways. Some might respond to their upbringing by rebelling against it, others by adopting a stance of studied or ironic detachment toward it, and others by creatively improvising with some of its materials. Indeed, in any typical human group, one would expect to see a great variety of responses to a set of socializing conditions, ranging from a warm embrace of the characteristic beliefs and practices of the previous generation's culture to total alienation from them, and from navel-gazing fascination with the history and distinctiveness of those beliefs and practices to utter indifference. All these responses are consistent with the idea of a common set of formative influences.

Just as the social lineage account is compatible with internal heterogeneity, it is also compatible with external overlap. As I noted earlier, it is not necessary for the account that individuals be subjected only to those formative conditions that generate a shared culture. Individuals are exposed to a set of conditions in common with others in the culture—this is what makes it a culture—but they are also subject to the influence of various idiosyncratic pressures. These include a set of more or less individualized conditions (e.g., particular family upbringings) but also various cross-cutting groups and cultures that persons are shaped by, because nothing in the account requires that individuals be shaped by one and only one encompassing environment.

Once it is allowed that people who share a culture are also subject to various other formative cross pressures, the overlapping, interactive, hybrid character of cultures immediately becomes predictable. Individuals who participate together in a given set of institutions and practices will also bring to the table various experiences of participating in institutions and practices that unite and apply to different groupings of people. Through their consumption and reading choices, their associational memberships and religious affiliations, their family's and community's migration history, their own travel and migration, and so on, individuals will find themselves under the influence of processes that are idiosyncratic relative to the culture-generating formative influences. In many cases, their contributions to the culture may reflect these idiosyncratic influences. As various values and meanings are shaped and negotiated under conditions of multiple affiliation, they predictably acquire an overlapping, hybrid character.

Against these arguments, it might be objected that there is still a sense in which the social lineage account fails to escape the problem of essentialism. According to this objection, by understanding cultures in terms of shared socialization experiences, rather than as shared frameworks of meanings, the proposal transfers the problem up a level without eliminating it. The reasons for skepticism about shared frameworks of meanings are also reasons for skepticism about common socialization experiences. Formative influences are not neatly packaged into sets of institutions and practices that are uniformly subjected on all and only the members of some groups. Among any set of people, there will predictably be too great a diversity of formative influences to allow the identification of groups of persons who share a common formative experience.

But this objection is overstated. The objection would obviously be correct if the account assumed that members of a culture were subjected to an identical set of formative influences. But, as we have seen, this is not the assumption. It is sufficient for the account that there be some significant set of institutions and practices to which, roughly speaking, all and only the members of the group are subject. If there were no such formative processes at work, there would be no group. But to judge that there is a group sharing a culture, the relevant processes just need to reach a threshold level of significance.

Suppose, however, that the objection goes farther and challenges the idea that sets of individuals could, and sometimes do, share enough exposure to a common set of formative influences to identify them as sharing a culture. The objection is still overstated. Although individuals are subject to a hugely diverse set of formative influences, it will normally be plausible to suppose that among these influences are certain institutions and practices, as well as environmental conditions, that exert an influence on many people at once. Moreover, it is plausible to think that, in some situations, more than one of these encompassing institutions, practices, and conditions will exert an overlapping influence on one and the same group of people.

As an example of overlapping influence, consider a set of persons who live in a common territory, speak the same language, and share a set of political institutions and hence a legal and bureaucratic framework.Footnote ¹² Suppose that this framework establishes a common set of educational institutions and a common market for consumer goods, media products, and so on. With these background conditions in place, it makes sense for various forms of media (print, television, etc.) to gear themselves toward the affairs and perceived interests of those same persons and their institutions. Not surprisingly, these media are read and watched with special attention by the same set of people who are featured in them (Anderson Reference Anderson1983, chap. 3). It also makes sense for producers and sellers of consumer goods to orient their production and advertising decisions around the perceived tastes and sensibilities of that same set of people.

With these common institutions in the background, one might also expect to see the emergence of a different kind of common practice, which might be called a “dominant discursive practice.” In a group of people who live together in a common territory, speak a common language, and find themselves subject to common political, legal, bureaucratic, economic, and media processes, certain forms of speech and discourse will enjoy considerable prominence at a given point in time. It is not that everyone will speak in the same way. As I insisted earlier, one would expect to find a great variety of responses to the dominant discursive practice. Rather, the suggestion is that there will emerge a common vocabulary of stories, references, and landmarks, as well as a common set of “points of concern,” which are widely assumed to be familiar to everyone (Laitin Reference Laitin2007, chap. 3). To be sure, this common vocabulary will also reflect interactive engagement with other cultures and societies and would thus have a recognizably hybrid character. But, if the reflections mentioned earlier are correct, the institutional and environmental setting will also help to shape the character of the dominant discourse. Someone who was exposed to the dominant discourse over a period of time could be said to have been subject to a formative influence that was distinctive vis-à-vis a discourse that was shaped under different institutional and environmental conditions. The influence will be all the more distinctive, all else being equal, the farther back in time the historical lineage of the culture stretched.

Putting all this together, it is not difficult to suppose that sets of individuals do sometimes share enough exposure to a common and distinctive set of formative influences for us to identify those sets as groups and to say that the members of those groups share a culture. Of course, particular influences are bound to apply with greater or lesser strength to some individuals in the culture. Some will have very few interactions with state institutions; others will be sheltered from the media; and so on. But even people who seem to avoid many of the common formative influences can be shaped by them indirectly. A person might disdainfully resist the media and popular culture in all its forms, but still find it hard to escape their influence because of the ways in which they shape the beliefs and values of other people with whom he or she interacts. The institutions and practices have certain general social effects that make it hard for anyone in their domain to escape their influence altogether.

Of course, I have picked a particularly stark example of a set of overlapping influences, in which the various common institutions and practices (political, bureaucratic, educational, linguistic, media, entertainment, and so on) all apply to a single set of persons who share a common territory. The inclusion of territory, and of political/administrative institutions, might even suggest a kind of statist view of cultures, in which nation-states are the main species of cultures, a view that would obviously not be congenial to multiculturalism. But, even though the example has a stark and “national” character, it is suggestive of how the social lineage account might generate a range of different judgments about which cultures there are. Language is an obvious factor that need not coincide with political or administrative boundaries, as is religion. Common political and administrative institutions may have served as crucial launching pads for distinctive sets of formative conditions, but these conditions may then gain a life of their own, and survive the disappearance of the institutions that brought them into existence in the first place. In general, then, the idea of a common and distinctive set of formative institutions retains its plausibility even as we move away from the nation-state cases.

I have been arguing that individuals do sometimes share enough exposure to a common set of formative influences to warrant judging that they share a common culture. By insisting on this point, however, perhaps the social lineage account risks falling into the opposite difficulty? Whereas the earlier concern was that the account would not be able to identify any cultures, now the worry is that the account would identify too many cultures. In the United States, for instance, in addition to well-known national institutions, many institutions operate at the state or even the local level. Must each state, county, and township be regarded as having its own culture? Because families are also an obvious locus of formative influence, should we also think of each set of family members as sharing their own distinctive culture? Or, to take a particularly hard case, should we say that women and men belong to distinct cultures because they undergo different forms of socialization?

The social lineage account certainly would not want to deny that there are many local and state cultures in a country as vast and diverse as the United States, nor that families are often sites of distinctive formative influence, nor indeed that it might make sense, in some contexts, to talk of “women's culture.” As we saw earlier, the account is compatible with the view that persons can have multiple cultural affiliations and thus with the observation, to borrow an example from Scheffler (Reference Scheffler2007, 100), that a person might be culturally “Western,” American, Californian, and northern Californian, at one and the same time. Still, there are a number of reasons for thinking that the account does not imply a limitless proliferation of cultures. One check on this tendency derives from the observation that broader-level discursive practices sometimes shape and structure more local processes. Families and local institutions can often serve as conduits for broader-level social processes as much as distinctive venues of socialization in their own right. Parents and those who administer and populate local practices and institutions are part of a broader social conversation about how those processes should operate, and this conversation may be highly influential on the actual character of the processes. Rather than millions of distinct formative processes at work (each family and local community), it may instead be more accurate to think of there being a single such process (operating quite broadly) with millions of points of application.

Second, it is important to distinguish between two mechanisms by which different people (e.g., men and women) might end up receiving distinct forms of socialization. It may be that they participate in different practices and institutions, and thus receive a socialization that is colored by those practices and institutions and by the participants in them. Or it may be that they participate in a common set of practices and institutions, but those practices and institutions socialize their participants in a differentiated manner. It may be integral to the beliefs, values, and norms that are encouraged by a given set of practices and institutions that persons with different ascriptive characteristics are assigned to different roles or treated in different ways. On the social lineage account, only the first of these mechanisms implies the division of the people involved into multiple cultures. The second mechanism is compatible with the generation of a single (gendered, racialized) culture.

A third check on proliferation is that the mere presence of some institution or practice does not, by itself, betoken the existence of a distinct culture. Some institutions and practices will not have sufficient formative impact on individuals to count as establishing a culture. The “sufficiency” threshold here should be based on the normative reasons for thinking that cultures matter in the first place. The distinctive formative impact has to be great enough so that the reasons for caring about differences of culture (which are explored toward the end of the article) really do register in a morally significant way. An important implication of this last assumption is that there is a pragmatic aspect to the identification of cultures. What entities should be considered as cultures is not a freestanding fact of social reality, but is dependent, in part, on the questions that we, the inquirers, find it important to answer.Footnote ¹³

SOME RELATED CONCEPTS

To close out the exposition of the social lineage account, let us relate it to several further ideas that are often invoked in discussions of culture: race/ethnicity, identity, and societal culture.

IMPLICATIONS FOR THE NORMATIVE THEORY OF MULTICULTURALISM

The dilemma of essentialism does not rule out the possibility of formulating a nonessentialist concept of culture. What it does claim to rule out is any such concept being serviceable to multiculturalism. For multiculturalism to be a plausible normative demand, two requirements have to be met. First, it has to be possible to identify distinct minority cultures and to make judgments about how they are being treated and whether they are surviving and flourishing. And, second, it has to make sense to think of cultures as mattering to their members. If it did not matter to people how their cultures were faring, the multicultural project would have trouble getting off the ground. A successful response to the dilemma of essentialism has to show how a nonessentialist culture concept can meet these two requirements on which normative multiculturalism depends.

I hope that I have said enough about the concepts of culture and cultural preservation to establish that the social lineage account can satisfy the first of these requirements. The proposed account turns both the identification of distinct cultures and judgments about how they are faring into empirical problems, which involve determining how far the socialization experience of some group is isolated socially and historically from that of others, and how robust the transmission mechanisms are whereby new members are socialized by existing members. We should turn now to the second requirement, and explore briefly why culture, as conceptualized by the social lineage account, is something that matters normatively.

In general, there are two kinds of arguments for thinking that culture matters normatively that figure prominently in the multiculturalism literature (e.g., in Kymlicka Reference Kymlicka1989; Reference Kymlicka1995; Raz Reference Raz1994, 170–91; Taylor Reference Taylor and Gutmann1992). One has to do with the quality of the options that individuals have at their disposal. A standard claim is that the options available to people are worse when their culture is faring poorly than when it is doing well. A second kind of argument highlights a more intrinsic concern that some individuals have with their culture. The culture matters to them for its own sake, and they care about both the success of the culture and its appropriate treatment by the broader society. Our question is whether the social lineage account is compatible with one or both of these kinds of arguments for valuing culture. Does the move from an essentialist culture concept to a nonessentialist one based on social lineage undermine the case for multiculturalism, or can it help to illuminate the normative commitments of multiculturalists?

Let us begin with the options-based argument for valuing culture. The details of the reasoning are not crucial to the current discussion. In one version (the “adequacy argument”), the claim is that members of a deteriorating culture would struggle to access an adequate range of options in the dominant culture, either because they face discrimination or because they lack some of the generic capacities (e.g., language proficiency) necessary to access those options. In a second version (the “particular options argument”), the claim is that the dominant culture does not provide particular options to minority culture members that especially matter to them. Both arguments can be illustrated by returning to our earlier example of Misael. On the adequacy argument, he values his Machiguenga culture because he fears that, should it decline or disappear, he would face discrimination in Peru's dominant culture or lack sufficient fluency in Spanish or mastery of other generic skills needed to enjoy an adequate range of options. On the particular options argument, he values his culture because he expects that particular options that he values (e.g., specific Machiguenga practices and rituals) would no longer be available should his culture be eclipsed by the dominant culture.Footnote ¹⁶

Clearly, these arguments would not show every culture to be valuable. Roughly speaking, the adequacy argument would apply mainly to societal cultures, whereas the particular options argument would apply mainly in contexts where there is a fairly significant discontinuity between the threatened and dominant cultures in the options they can be expected to generate. Moreover, some cultures may be so grossly oppressive or chauvinistic as to lack any value at all. Our purpose here, however, is not to provide a comprehensive examination of arguments for the value of culture but, more narrowly, to consider whether the shift to the social lineage account makes the arguments less applicable than they would be on an essentialist view.

One reason to suspect a problem is that culture, on the social lineage account, is no longer defined in terms of any specific beliefs, practices, or options. As we saw in Section 3, a culture can survive, on the view being proposed, and yet undergo a fundamental revolution in the kinds of values, meanings, and preferences that prevail, and thus in the sorts of options and practices that are available. Given this basic feature of the social lineage account, it might be wondered how it could support an options-based argument for multiculturalism.

However, this challenge overstates the dependence of the options-based argument on an essentialist view of culture. Consider first the adequacy version of the argument. The adequacy version does not worry about the loss of any option in particular, but instead cautions that minority culture members might lack an adequate range of options in the dominant culture. Because there are many different sets of options that would constitute an adequate range (so long as there are sufficient quantity, quality, and diversity of options), this version of the argument need not rely on any problematic essentialist understanding of culture. Particular options can come and go, can be contested to greater or lesser degrees, and can overlap to greater or lesser extents with options generated by other cultures, and none of this should affect whether the options available to a member of a culture—given facts about discrimination, generic capacity, and so on—are adequate or not.

Nor is essentialism about cultures needed for the particular options version of the argument. To be sure, a nonessentialist account does not define culture in terms of the existence of particular options or practices, and so there is no necessary connection between cultural preservation and the availability of any particular options. Even if Machiguenga culture is preserved, some of the rituals and practices that Misael cares about the most may disappear. Nonetheless, the frequency with which particular preferences and values are affirmed, and particular options made available, will depend greatly on the sorts of socialization processes emphasized by the social lineage account, and this is all the argument needs. Misael can reasonably expect to have better access to the particular options he values in a Machiguenga setting than in a dominant culture setting, not because Machiguenga culture is defined in terms of those options, but because the socialization processes that make it the distinctive culture that it is will generate those options more reliably than will the dominant culture. The particular options reason for valuing culture, then, is fully compatible with the social lineage account.

An analysis of the intrinsic argument for multiculturalism produces a similar conclusion. It is true that what attracts certain people in their own culture is some specific set of essential characteristics. Abandoning the essentialist concept of culture, and replacing it with the social lineage account, will mean that cultural prosperity and preservation do not necessarily guarantee that the characteristics these people value will be present. But the social lineage account still leaves plenty of room for people to feel an intrinsic attachment to their own culture. As we saw in the earlier discussion of identity, people who are socialized as members of a particular culture will often feel a kind of attachment to the culture itself, and/or to fellow members of the culture. An identification with one another, and with the institutions and practices of their joint socialization, grows out of a history of interaction and a common set of experiences and points of reference. In addition, in the social lineage account, there is a straightforward sense in which a person's culture helped to make him or her the individual that he or she is. For people who identify with their culture, it is difficult, as a result, to distinguish disrespectful treatment of the culture from disrespectful treatment of them as individuals. In general, then, far from undermining it, the social lineage account can help to illuminate an argument for multiculturalism grounded in the intrinsic significance that some people attach to their culture.