Language and Agency

Annu. Rev. Anthropol. 2001. 30:109–37 c 2001 by Annual Reviews. All rights reserved Copyright ° LANGUAGE AND AGENCY Laura M. Ahearn Rutgers University, New Brunswick, New Jersey 08901; e-mail: [email protected] Key Words practice, grammar, dialogic, gender, literacy ■ Abstract This review describes and critiques some of the many ways agency has been conceptualized in the academy over the past few decades, focusing in particular on practice theorists such as Giddens, Bourdieu, de Certeau, Sahlins, and Ortner. For scholars interested in agency, it demonstrates the importance of looking closely at language and argues that the issues surrounding linguistic form and agency are relevant to anthropologists with widely divergent research agendas. Linguistic anthropologists have made significant contributions to the understanding of agency as it emerges in discourse, and the final sections of this essay describe some of the most promising research in the study of language and gender, literacy practices, and the dialogic construction of meaning and agency. WHY AGENCY NOW? The term agency, variously defined, has become ubiquitous within anthropology and other disciplines. This essay describes and critiques some of the many ways agency has been conceptualized in the academy over the past few decades. While I propose a skeletal definition for the term, my purpose is not to dictate how scholars should define agency, or even to insist that they should use the term at all. Rather, my purpose is to survey the scholarship on agency and to suggest how important it is for scholars interested in agency to look closely at language and linguistic form. I argue that the issues surrounding language and agency are relevant to anthropologists with widely divergent research agendas because most anthropologists—whether archaeological, biological, cultural, or linguistic—are concerned, in one form or another, with what people say and do. Linguistic anthropologists have made significant contributions to the understanding of agency as it emerges in discourse, and in the final sections of this essay, I describe some of the most promising research in the study of language and gender, literacy practices, and the dialogic construction of meaning and agency. Before turning to definitional issues, it is worthwhile to reflect for a moment on our own intellectual practice and ask ourselves why so many scholars in so many fields are currently interested in the concept of agency. Messer-Davidow poses this question directly, asking, “Why agency now?” (1995, p. 23). While there are undoubtedly many answers to this question, one is that there is a clear connection 0084-6570/01/1021-0109$14.00 109 110 AHEARN between the emergence of interest in approaches that foreground practice on the one hand, and the social movements of the 1960s and 1970s on the other (Ortner 1984, p. 160). In addition, the social upheavals in central and eastern Europe in the late 1980s and early 1990s led many scholars to articulate more clearly their ideas about human agency and social structures (Sztompka 1991). As a result of witnessing or participating in actions aimed at transforming society, then, many academics have begun to investigate how practices can either reproduce or transform the very structures that shape them. I believe it is no coincidence that the recent agentive1 turn, an outgrowth of the trends Ortner identified in 1984, follows on the heels not only of the social movements of the past few decades but also of postmodern and poststructuralist critiques within the academy that have called into question impersonal master narratives that leave no room for tensions, contradictions, or oppositional actions on the part of individuals and collectivities. It is because questions about agency are so central to contemporary political and theoretical debates that the concept arouses so much interest—and why it is therefore so crucial to define clearly. DEFINITIONAL STARTING POINTS In most scholarly endeavors, defining terms is half the battle. This is especially true for a word like language, which is so commonplace that researchers often mistakenly assume its meaning is self-evident (Williams 1977, pp. 21–44). Precise definitions are equally essential for words such as agency that have taken on new meanings on entering academic discourse. As a starting point for this review, I discuss language as a form of social action, which is the approach to language that many linguistic anthropologists take, and then present a provisional definition of agency. Language as Social Action Whereas most linguists follow de Saussure (1986) and Chomsky (1965, 1986) in studying language as a set of formal structures set apart from everyday interactions (“langue” rather than “parole,” and “competence” rather than “performance”), most linguistic anthropologists regard language as a form of social action, a cultural resource, and a set of sociocultural practices (Schieffelin 1990, p. 16). People do things with words (Austin 1962, Searle 1969, cf. Butler 1997). Brenneis & Macaulay (1996), Duranti (1997), and Hanks (1996) present persuasive and thorough explications of this approach to language. Linguistic anthropologists consider language, whether spoken or written, to be inextricably embedded in networks of sociocultural relations. When scholars treat language, culture, and society as mutually constituted, one of their main responsibilities then becomes to study how 1 There is no unanimity in the choice of an adjectival form for agency. While other writers use agential or agentic, I prefer agentive. LANGUAGE AND AGENCY 111 discourse both shapes and is shaped by sociocultural factors and power dynamics (Urban 1991). There are no neutral words, Bakhtin (1981, p. 293) reminds us: “All words have the ‘taste’ of a profession, a genre, a tendency, a party, a particular work, a particular person, a generation, an age group, the day and hour. Each word tastes of the context and contexts in which it has lived its socially charged life. . . .” Unequal power relations can result in—and be the result of—symbolic violence (symbolic power, symbolic domination), which, Bourdieu (1991, p. 170) maintains, occurs when individuals mistakenly consider a standard dialect or style of speaking to be truly superior to the way they themselves speak, rather than an arbitrary difference afforded social significance. Language and power are therefore commonly intertwined. Note that in this view, language is not defined as a conduit that merely conveys information (Reddy 1979), and it is not a transparent vehicle carrying only referential meaning (Goodwin 1990, p. 4). In order to understand how linguistic anthropologists approach language, we have to set aside this vehicular metaphor— unless, that is, we say that linguistic anthropologists view language as a vehicle that people themselves are continually in the process of building together. According to this approach to language, meanings are co-constructed by participants, emergent from particular social interactions. Scholars have proposed various strategies for understanding how this works. Early work in the ethnography of communication encouraged researchers to look for patterns in actual speech (Gumperz & Hymes 1964, Bauman & Sherzer 1989). Scholars grounded in fields as diverse as ethnomethodology, sociolinguistics, the sociology of language, linguistic anthropology, and conversation analysis have contributed to an understanding of how meanings emerge in conversations by focusing on the microprocesses of linguistic interactions (Garfinkel 1967; Goffman 1974, 1981; Ochs et al 1996; Sacks 1992). The appropriate unit of analysis for many scholars who treat language as social action is not the sentence, the individual, or even the conversation but rather speech acts (Austin 1962, Searle 1969), speech events (Jakobson 1960, Hymes 1972), participant structures (Philips 1972), participation frameworks (Goffman 1981), participant frameworks and situated activities (Goodwin 1990), or communities of practice (Eckert & McConnell-Ginet 1992). Within these contexts, as meanings are coconstructed, social reality is also constructed. In the approach advocated here, then, language does not merely reflect an already existing social reality; it also helps to create that reality (Gumperz & Levinson 1996, Hill & Mannheim 1992, Lucy 1992, Sapir 1949, Spender 1980, Whorf 1956, Williams 1977). With such a dialogic, coconstructed view of language as a form of social action, linguistic anthropologists face the challenge of interpreting fluid, often ambiguous linguistic data with important sociocultural implications. How can this task best be accomplished? Both text and context must be taken into consideration, and they must be understood to be intrinsically interwoven (Duranti & Goodwin 1992). We must acknowledge the inevitability of a certain degree of interpretive indeterminacy while also recognizing that indeterminacy is not limitless (Derrida 112 AHEARN 1972). Elsewhere I have argued that we should espouse what I call a practice theory of meaning constraint (Ahearn 1998, 2001). According to this perspective, we must shift our focus away from searching for definitive interpretations and instead concentrate on looking for constraints on the kinds of meanings that might emerge from an event such as a song performance or a text such as a love letter. Meanings might be infinite in number, but they are very tightly bounded. As Eco (1990, p. 42) notes, “If it is very difficult to decide whether a given interpretation is a good one, it is, however, always possible to decide whether it is a bad one.” Appadurai (1991, p. 472) takes a similar view of language, calling for a new “theory of reception” that incorporates an understanding of intertextuality and situatedness. In advocating a practice theory rather than a theory of reception, however, I emphasize how individuals, including scholars, actively construct and constrain—rather than passively receive—interpretations that are both socially mediated and intertextually situated within a bounded universe of discourse. From the foregoing discussion of language, it should be clear that as linguistic anthropologists increasingly treat language as a form of social action, the task of developing a theoretically sophisticated understanding of agency becomes ever more urgent. We turn therefore to the challenge of defining the concept. A Provisional Definition of Agency Jean and John Comaroff have called agency “that abstraction greatly underspecified, often misused, much fetishized these days by social scientists” (1997, p. 37; cited in Ortner 2001, p. 1). While this assessment may be a bit harsh, it is true that scholars often fail to recognize that the particular ways in which they conceive of agency have implications for the understanding of personhood, causality, action, and intention. Agency therefore deserves “deeper consideration and more extensive theoretical elaboration” (Dobres & Robb 2000, p. 3). Let me propose, then, a provisional definition of the concept: Agency refers to the socioculturally mediated capacity to act. According to this bare bones definition, all action is socioculturally mediated, both in its production and in its interpretation. Although this definition provides us with a starting point, it leaves many details unspecified. The following are some questions to ponder—questions that may be answered in different ways by different scholars. Must all agency be human? Can nonhuman primates (Small 1993), machines (Pickering 1995), technologies (Dobres 2000), spirits (Keane 1997, pp. 64–66), or signs (Colapietro 1989, pp. 95–97; Peirce 1955) exercise agency? Must agency be individual, leading to charges of unwarranted assumptions regarding Western atomic individualism (Ortner 1996)? Or can agency also be supraindividual—the property, perhaps, of families, faculties, or labor unions? Conversely, can agency be subindividual—the property of “dividuals” (Daniel 1984, p. 42; Marriott 1976; McElhinny 1998, p. 181), as when someone feels torn within herself or himself? What does it mean to be an agent of someone else? Must LANGUAGE AND AGENCY 113 agency be conscious, intentional, or effective? What does it mean for an act to be conscious, intentional, or effective? We might begin to answer some of these questions by considering, as Karp (1986) does, what distinguishes an “actor” from an “agent.” In Karp’s view, an actor refers to a person whose action is rule-governed or rule-oriented, whereas an agent refers to a person engaged in the exercise of power in the sense of the ability to bring about effects and to (re)constitute the world (Karp 1986, p. 137). Actor and agent should be considered two different aspects of the same person, according to Karp, or two different perspectives on the actions of any given individual. Ortner (2001) proposes differentiating among various types of agency, such as “agency of power” and “agency of intention,” though she is careful to note that any such distinction is purely heuristic because types of agency are often inseparable in practice. Some scholars, such as Wertsch et al (1993), advocate a nonindividualistic notion of agency. Drawing on Vygotsky (1978, 1987) and paraphrasing Bateson (1972), they argue that agency “extends beyond the skin” because it is frequently a property of groups and involves “mediational means” such as language and tools (Wertsch et al 1993, p. 352). It is especially important for anthropologists to ask themselves how conceptions of agency may differ from society to society, and how these conceptions might be related to notions of personhood and causality (Jackson & Karp 1990, Skinner et al 1998). Pickering suggests that “within different cultures human beings and the material world might exhibit capacities for action quite different from those we customarily attribute to them” (1995, p. 245). Desjarlais presents an illustration of this within the United States itself in his study of a homeless shelter in Boston, in which he argues that the forms of agency he observed emerged out of a specific sociocultural context. Agency was not ontologically prior to that context but arose from the social, political, and cultural dynamics of a specific place and time (Desjarlais 1997, p. 204). In my own work, I have maintained that it is important to ask how people themselves conceive of their own actions and whether they attribute responsibility for events to individuals, to fate, to deities, or to other animate or inanimate forces. In the case of Junigau, Nepal, people’s conceptions of their own and others’ actions are changing rapidly, demonstrating the need for anthropologists to ask not only what agency means for themselves as theorists, but what it means for the people with whom they work, and how those meanings may shift over time (Ahearn 2000b, 2001). PROBLEMS IN DEFINING AGENCY Several uses of agency that are common in the literature are, in my opinion, of questionable use to anthropologists (though perhaps not to scholars in other disciplines). In the following overview, which draws from several fields but does not purport to be a full delineation of the debates within any given discipline, examples are discussed in which agency is defined too simplistically, too narrowly, or too opaquely. 114 AHEARN Agency as Free Will One of the most common tendencies in discussions of agency is the treatment of it as a synonym for free will. This is especially evident in what philosophers call action theory. Within action theory, philosophers attempt to distinguish an “action” from an “event.” Davidson (1980[1971], p. 43) begins his famous essay, “Agency,” with the following question: “What events in the life of a person reveal agency; what are his deeds and his doings in contrast to mere happenings in his history; what is the mark that distinguishes his actions?” Twenty years later, Segal (1991, p. 3) echoes Davidson as he explains philosophical action theory: “Hitting a ball is an action, falling down a flight of stairs is not. A theory of action seeks, among other things, to explain the distinctions we make.” In attempting to explain human agency, action theorists and other philosophers generally argue that agency requires some sort of concomitant mental state, such as “intention” (Davidson 1980[1971], p. 46), “presence of the self” (Segal 1991, p. 113), a “rational point of view” and a “domain of intentional control” (Rovane 1998, p. 85), or “motivation, responsibility, and expectations of recognition or reward” (Mann 1994, p. 14). The main weakness in treating agency as a synonym for free will is that such an approach ignores or only gives lip service to the social nature of agency and the pervasive influence of culture on human intentions, beliefs, and actions. Even Taylor (1985, pp. 1–44), a philosopher whose writings on language and agency are extremely thought provoking, locates agency inside the mental processes of particular individuals when he connects agency with “second-order desires,” “strong evaluation,” and “a vocabulary of worth.” Similarly, Ludwig Wittgenstein, the famous philosopher of language to whom linguistic anthropologists increasingly look for inspiration, fails to theorize adequately the sociocultural nature of language and action. While Wittgenstein (1958) recognizes the degree to which language and social forms are intertwined, he leaves the details of this interrelationship unexplained. Giddens (1979, p. 50) notes this shortcoming in Wittgenstein’s work on language and action, stating, “Wittgensteinian philosophy has not led towards any sort of concern with social change, with power relations, or with conflict in society. Other strands in the philosophy of action have operated at an even further distance from such issues, focusing attention almost exclusively upon the nature of reasons or intentions in human activity.” Traces of this tendency to equate agency with socially unfettered free will can be found in many other disciplines, including anthropology, psychology, political science, and history. Sometimes scholars contend that only certain individuals “have agency,” while others have little or none. Some historians, for example, locate agency solely in the power of individual “Great Men.” A recent debate surrounds the publication of comparative political scientist Daniel J. Goldhagen’s (1996) book, which argues that ordinary Germans played an active, agentive role in the Holocaust. Moses (1998) states, “Having raised the question of the perpetrators’ choice, Goldhagen must convince the reader that they were not ‘just following orders,’ that is, that these actors possessed agency” (p. 205). According to Moses, LANGUAGE AND AGENCY 115 however, Goldhagen’s implicit methodological underpinnings are contradictory. On one hand he espouses rational choice models2 to stress that ordinary Germans had “agency,” which he equates with free will, and yet on the other hand he relies on behaviorism to account for the prevalence of antisemitism in German society (Moses 1998, p. 209). Some scholars, especially those studying colonialism and postcolonialism, have been moving away from approaches that treat agency as a synonym for free will as exercised by completely autonomous individuals (e.g., Cooper 1994, Cooper & Stoler 1997, Pieters 2000, Pomper 1996, Scott 1988, Sewell 1992). Historian Lalu (2000, pp. 50–51) offers an observation that applies equally well to historians, philosophers, anthropologists, and all other scholars interested in agency: “[T]he question of agency, it seems, may be posed in ways other than in terms of the autonomous subject or authorial subject. . . . [We] may have to think of the ways in which agency is constituted by the norms, practices, institutions, and discourses through which it is made available.” Such a linguistically and socioculturally mediated conception of agency is discussed further below. Equating Agency with Resistance Another misguided approach to agency is to consider it a synonym for resistance. This approach is characteristic of the work of some anthropologists, many scholars in subaltern studies, and feminist theorists in a number of fields. Feminism has always addressed issues of agency, if only implicitly (Mann 1994, p. 14), but recently the term has been cropping up with increasing frequency (Andermahr 1997, Davies 1991, Dissanayake 1996, Gardiner 1995, Goddard 2000, Kumar 1994, McNay 2000). Fraser (1992, pp. 16–17) explains that agency has become a problem in recent feminist theory because of two equally important goals. On the one hand, feminists have sought to establish the seriousness of their struggle by demonstrating the pervasiveness and systematicity of male dominance. This has led to the development of theories that emphasize the constraining power of gender structures and norms, while downplaying the resisting capacities of individuals and groups. On the other hand, feminists have also sought to inspire women’s activism by rediscovering lost or socially invisible traditions of resistance in the past and present. In some scholars’ work (both feminist and nonfeminist), instead of a balance between these two countervailing tendencies, there is an overemphasis on resistance (Abu-Lughod 1990). According to many feminist theorists, in order to demonstrate agency, a person must resist the patriarchal status quo (e.g., Goddard 2000, p. 3). While one can certainly understand the impulse behind equating agency with resistance, agency should not be reduced to it. Oppositional agency is only one of many forms of agency. Many scholars interested in other forms of social and economic oppression also equate agency with actions that resist domination (Pruyn 1999; Scott 1985, 2 See Burns (1994) for a critique of rational choice models of agency. 116 AHEARN 1990). As useful as many of these studies are, I take to heart Abu-Lughod’s (1990) caution against the “romance of resistance” and second Ortner’s (1995) conclusion that there is no such thing as pure resistance; motivations are always complex and contradictory (Ahearn 2000a, Gamburd 2000, Jeffery & Jeffery 1996, Jeffery & Basu 1998). I find MacLeod’s work very helpful in conceptualizing both women’s and men’s agency. She notes that women, “even as subordinate players, always play an active part that goes beyond the dichotomy of victimization/acceptance, a dichotomy that flattens out a complex and ambiguous agency in which women accept, accommodate, ignore, resist, or protest—sometimes all at the same time” (MacLeod 1992, p. 534). Such a nuanced understanding of the multiplicity of motivations behind all human actions should be at the core of our definition of agency. The Absence of Agency? Another approach to agency that presents challenges to scholars is that of Foucault (1977, 1978). On one level, Foucault can be read as stating that omnipresent impersonal discourses so thoroughly pervade society that no room is left for anything that might be regarded as agency, oppositional or otherwise. In The History of Sexuality, Volume I, for example, Foucault (1978, pp. 93, 95) writes, Power is everywhere; not because it embraces everything, but because it comes from everywhere. And “Power,” insofar as it is permanent, repetitious, inert, and self-reproducing, is simply the over-all effect that emerges from all these mobilities, the concatenation that rests on each of them and seeks in turn to arrest their movement . . . there is no power that is exercised without a series of aims and objectives. But this does not mean that it results from the choice or decision of an individual subject . . . There have been numerous critiques of Foucault’s definition of power, many of them focusing on the problematic implications it has for human agency (Bartky 1995, Hoy 1986). Even though Foucault states that “[w]here there is power, there is resistance,” he continues on to say, “and yet, or rather consequently, this resistance is never in a position of exteriority in relation to power” (Foucault 1978, p. 95). The problem is that in History of Sexuality, Volume I, Foucault never explains how power is enforced or personified, and the processes of resistance remain similarly opaque. Nor, despite the centrality of Foucault’s work to scholars of colonialism, does he examine colonial politics in any detail in that volume (Stoler 1995). Many scholars agree with Said (1983, p. 246), who argues that “[t]he disturbing circularity of Foucault’s theory of power is a form of theoretical overtotalization. . . .” Others, however, have maintained that Foucault’s definition of power does not eliminate the possibility for agency, however defined. O’Hara (1992, p. 66), drawing largely on Foucault’s later work, argues that Foucault proposes a model of agency that is “a matter of plurality, mobility, and conflict.” According to Halperin (1995, pp. 16–17), Foucault’s notion of power is not a substance but a relation, a LANGUAGE AND AGENCY 117 dynamic situation; it produces not only constraints on, but also possibilities for, action. Nevertheless, even if Foucault’s formulations do leave room for agency, his focus is more on pervasive discourses than on the actions of particular human beings. PRACTICE THEORY Consider Marx’s famous words in “The Eighteenth Brumaire of Louis Bonaparte”: Men make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly found, given and transmitted from the past. The tradition of all the dead generations weighs like a nightmare on the brain of the living (Marx 1978[1852], p. 595). How can we reconcile the fact that, as Marx noted almost a century and a half ago, individuals appear to create society even as they are created by it? Berger & Luckmann turn this question into a trilogy of paradoxical statements in their famous book, The Social Construction of Reality: “Society is a human product. Society is an objective reality. Man is a social product” (1966, p. 61; emphasis in the original). The most promising approach for understanding these seemingly contradictory statements is practice theory, which Ortner (1989, p. 11; 1984, 1996) defines as “a theory of the relationship between the structures of society and culture on the one hand and the nature of human action on the other.” The emphasis in practice theory is on the social influences on agency; human actions are central, but they are never considered in isolation from the social structures that shape them. Structuration Theory Giddens is perhaps the central figure in the debate about agency and structure and is considered one of the founders of practice theory (Giddens 1979, 1984, Archer 1988, Burns & Dietz 1994, Karp 1986). Explicitly drawing on the insights of ethnomethodologists such as Garfinkel and interactionist sociologists such as Goffman, Giddens attempts to breathe life into social structures and bring social structures into contact with human actions (Giddens 1979, p. 57, 68, 83; Bryant & Jary 1991, Sewell 1992). Unlike scholars who treat agency as a synonym for free will or resistance, Giddens consistently links agency to structure through his discussion of rules and resources. Central to Giddens’ theory of structuration is the understanding that people’s actions are shaped (in both constraining and enabling ways) by the very social structures that those actions then serve to reinforce or reconfigure. Given this recursive loop consisting of actions influenced by social structures and social structures (re)created by actions, the question of how social change can occur is crucial and is taken up below in the context of other practice theorists. 118 AHEARN Some sociologists prefer to use the term practice or praxis (drawing on and redefining the Marxist term) in addition to, or instead of, agency (Giddens 1979, p. 56; Sztompka 1994). Sztompka, for example, distinguishes the two terms in the following manner: “Agency and praxis are two sides of the incessant social functioning; agency actualizes in praxis, and praxis reshapes agency, which actualizes itself in changed praxis” (Sztompka 1994, p. 276). Thus, agency can be considered the socioculturally mediated capacity to act, while praxis (or practice) can be considered the action itself. Agency and the Habitus Aside from Giddens, the most influential theorist within practice theory is Bourdieu, a professor of sociology who has conducted ethnographic fieldwork in Algeria. Bourdieu borrows and redefines the term habitus, first used in anthropology by Marcel Mauss to refer to a habitual condition, particularly of the body (Farnell 2000, p. 399). Bourdieu’s definition refers to a generative process that produces practices and representations that are conditioned by the “structuring structures” from which they emerge. These practices and their outcomes—whether intended or unintended—then reproduce or reconfigure the habitus (Bourdieu 1977, p. 78). The recursive nature of this process mirrors that found in Giddens’ theory of structuration. The habitus generates an infinite but bounded number of possible actions, thoughts, and perceptions, each one of which is imbued with the culturally constructed meanings and values embodied by the habitus. These actions, thoughts, and perceptions in turn then recreate and/or challenge the culturally constructed meanings and values. With this analysis of agency, Bourdieu moves us far from the concept of free will. Although he defines the habitus as “an endless capacity to engender products,” Bourdieu emphasizes dispositions in order to preclude any assumption of absolute free will on the part of actors, repeatedly pointing out how far removed his concept of the habitus is from a creation of unpredictable novelty. What prevents the creation of unpredictably novel sociocultural products are the (pre)dispositions the habitus embodies in its many forms and structures. Of the infinite thoughts, meanings, and practices that the habitus can produce at any given historical moment, there is only a minimal probability that any will ever be thought or practiced because individuals are predisposed to think and act in a manner that reproduces the existing system of inequalities. As necessary and helpful as his reminders are of the constraints on individuals’ actions and thoughts, Bourdieu, like Giddens, faces the dilemma of explaining how social reproduction becomes social transformation (Sewell 1992). Bourdieu emphasizes the reproductive tendencies of the habitus, which, because it is sturdy and well-rooted, located in the physical environments containing actors, and embodied mentally and physically within the actors themselves, can be applied in new as well as familiar situations to reinforce the status quo. Despite the theoretical possibility of social transformation resulting from actions generated by the habitus, Bourdieu’s framework leaves little room for resistance or social change. LANGUAGE AND AGENCY 119 The microprocesses of resistance are taken up in the Practice of Everyday Life, written by another theorist commonly associated with practice theory, historian Michel de Certeau. De Certeau encourages other scholars to attend to the actions of ordinary people, especially when they engage in “la perruque” (literally, “the wig”), a French idiomatic expression that refers to the work one does for oneself in the guise of work done for an employer (de Certeau 1984, p. 25). De Certeau uses the trope of la perruque to describe how individuals use strategies and tactics to carve out a semi-independent domain of practice within the constraints placed on them by the powerful. Although de Certeau, Bourdieu, and Giddens offer us theories with significant explanatory power in regard to the persistence of deeply embedded relations of inequality, they give insufficient attention to the question of how any habitus or structure can produce actions that fundamentally change it. In an attempt to understand more fully how social change occurs, let us look at the work of other practice theorists working within anthropology. Anthropological Contributions to Practice Theory In his Historical Metaphors and Mythical Realities, Sahlins sets for himself the task of understanding how an attempt at social reproduction can become social transformation (Sahlins 1981, Obeyesekere 1992). Sahlins, unlike Bourdieu, attends closely to the processes of social transformation and emphasizes the importance of history in his historical and ethnographic account of the transformation that Hawaiian society underwent in the wake of Captain Cook’s arrival and his subsequent murder. Noting (perhaps too perfunctorily) that such transformations can occur even without intercultural collisions, Sahlins nevertheless focuses on how these cross-cultural contacts may facilitate unprecedented change. When individuals bring their cultural understandings, as derived from structural principles (what Bourdieu would call their habitus), to bear on new situations, the dynamics of practice [what Sahlins calls “the structure of the conjuncture” (1981, p. 35)] can cause unintended outcomes. What starts as an attempt to reproduce social structure may end in social transformation. By interweaving history and structure in this manner, Sahlins not only highlights the importance of agency and its often unintended consequences, he also emphasizes the temporality of agency and throws into question the concept of resistance as conscious activity. Nevertheless, because Sahlins’ work, like Bourdieu’s, evinces traces of its structuralist roots, the processes of social reproduction/transformation he posits are rather mechanistic, and his “permanent dialectic of structure and practice” (Sahlins 1981, p. 54) has little room in it for tensions inherent within social structure itself. Addressing this very issue, Ortner (1989) builds on the theories of both Sahlins and Bourdieu in High Religion: A Cultural and Political History of Sherpa Buddhism. In her elucidation of the terms practice, structure, actor, and history, Ortner sets out the four cornerstones on which her ethnography is built, thereby sidestepping the dualistic, mechanistic formulations of Bourdieu and Sahlins. Practice for 120 AHEARN Ortner entails the recognition of asymmetry and domination in particular historical and cultural settings, along with an awareness of the cultural schemas and constraints within which individuals act. Departing from the claims of both Bourdieu and Sahlins, Ortner emphasizes the existence of inherent structural contradictions that keep a simple reproduction of the hegemonic social order from being a foregone conclusion. As Williams (1977, p. 113) notes, “The reality of any hegemony, in the extended political and cultural sense, is that, while by definition it is always dominant, it is never either total or exclusive.” Because of the tensions and contradictions inherent in the habitus, actors are neither free agents nor completely socially determined products. Instead, Ortner (1989, p. 198) suggests that they are “loosely structured.” The central question for practice theorists, then, is determining how such loosely structured actors manage at times to transform the systems that produce them. Such loose structuring can occur linguistically as well as socioculturally. Speakers of a given language are constrained to some degree by the grammatical structures of their particular language, but they are still capable of producing an infinite number of grammatically well-formed utterances within those constraints. Moreover, languages, like cultures, change over time through drift and contact despite their supposedly self-reproducing structures (DeGraff 1999, Lightfoot 1999, Sapir 1933[1949], de Saussure 1986). It is therefore helpful to look closely at language (both its grammatical structures and its patterns of use) in order to gain a more thorough understanding of how people reproduce and transform both language and culture. The following section describes some of the grammatical constraints, either universal across all languages or particular to a smaller set of languages, that may predispose people to conceptualize agency and subjecthood in certain ways. GRAMMATICAL AGENTS Any discussion of agency and language must consider how grammatical categories in different languages distinguish among types of subjects, for such categories, “to the extent that they are obligatory and habitual, and relatively inaccessible to the average speaker’s consciousness, will form a privileged location for transmitting and reproducing cultural and social categories” (Hill & Mannheim 1992, p. 387). Although each language has its own set of linguistic resources that can be used to exercise, attribute, or deny agency, there are also some features that can be found in every language (Comrie 1981). According to Dixon (1994, p. 6), for example, all languages work in terms of three basic relations–S, A, and O—defined as follows: S—Subject of an intransitive verb (e.g., Sita went to Kathmandu); A—Agent, or subject, of a transitive verb (e.g., Parvati loves Shiva); and O—Object of a transitive verb (e.g., Maya ate rice). LANGUAGE AND AGENCY 121 Semantically, there are various roles the subject of a sentence can take, such as the following (cf. Duranti 1994, pp. 122–123; Keenan 1984): Agent Pabi read the book. Actor Perceiver Instrument Patient/Undergoer Shiva danced. Tika heard the news. The stone broke the window. The old woman died. These semantic roles can be treated in various ways syntactically. Defining the linguistic subject in a way that applies to all languages turns out to be a challenging and controversial topic over which linguists differ (Comrie 1981, pp. 98–101). In the majority of languages, including most of the languages of Europe, the subjects of transitive and intransitive verbs are treated the same way syntactically, while the object of a transitive verb is treated differently. This pattern is known as accusativity (Dixon 1994, pp. 16–17).3 In about a quarter of the world’s languages, however, a complementary pattern obtains in which the subject of an intransitive verb and the object of a transitive verb are treated the same way syntactically, while the subject of a transitive verb is treated differently. This pattern is known as ergativity (Bittner & Hale 1996, Dixon 1994, Plank 1979). In ergative languages, there is usually a grammatical marker that distinguishes Agents (of transitive verbs) from Subjects (of intransitive verbs) and Objects (of transitive verbs). Consider the following examples in Samoan, taken from Duranti (1994, p. 122), in which the ergative marker e is present only in (a), before the Agent of the transitive verb, and not before the Subject of the intransitive verb in (b): (a) ’ua fa’atau e le tama le suka. TA4 buy ERG ART boy ART sugar The boy has bought the sugar. (b) ’ua alu le tama ’i le maketi. TA go ART boy to ART market The boy has gone to the market. 3 Languages in which the subjects of transitive and intransitive verbs are treated the same way syntactically while the transitive object is treated differently are also called “nominativeaccusative.” Languages in which the subjects of intransitive verbs and the objects of transitive verbs are treated the same way syntactically are also called “ergative-absolutive.” I follow Dixon (1994) in shortening these terms to “accusativity” and “ergativity,” respectively, in order to emphasize which case is being treated uniquely; with accusativity, Objects are placed in the accusative case and are treated differently from Subjects and Agents, whereas with ergativity, Agents are placed in the ergative case and are treated differently from Subjects and Objects. 4 The abbreviations used in the interlinear glosses have the following meanings: TA refers to a marker of verb tense or aspect; ERG refers to an ergative marker; ART refers to an article (Duranti 1994, pp. 177–78). 122 AHEARN Some languages have “split” grammatical systems in which speakers follow an accusative pattern in some cases and an ergative pattern in other cases. In standard Nepali, for example, the ergative marker le is obligatorily used with the Agents of transitive verbs in the past tense only—not in the present or future tense. In the dialect of Nepali spoken in the village of Junigau, Nepal, however, people use the ergative marker le in nonobligatory ways in the present and future tenses when they want to place emphasis on the Agent, as can be seen in the following example taken from a Junigau woman’s narrative of marriage (Ahearn 2001): (c) mai le pani mān garchhu. I ERG too respect do I, too, respect [my husband]. A related sort of split appears in languages that have grammatical systems in which the subjects of some intransitive verbs are categorized with transitive subjects, while the subjects of other intransitive verbs are still considered intransitive. In Guaranı́, for example, when “I” is used with more agentive intransitive verbs, such as “go” and “get up,” it is placed in the same (ergative) case as when “I” is used with the transitive verb “bring” (Mithun 1991, p. 511). “I” is placed in a different case in Guaranı́ when used with less agentive intransitive verbs, such as “to be”— the same case that is used for the direct object pronoun “me.” In these languages, attributions of agency are built right into their semantic and syntactic structures. Let me emphasize, however, that in none of these cases is it possible to draw a simplistic connection between the presence of ergative case markings and “more” or “less” agency.5 Nevertheless, ergative languages present researchers with a valuable tool they can use to explore notions of subjectivity and action in other cultures. While languages may encode agency differently in their grammatical categories, there are some universal patterns that can be discerned regarding the types of nouns most likely to appear in the Agent position. Drawing on linguistic data from Chinook and Dyirbal, both of which are split ergative systems that use an ergative pattern of case-marking for certain types of noun phrases and an accusative pattern for other types of noun phrases, Silverstein (1976, pp. 116–122) proposes an animacy hierarchy that predicts where on the spectrum of noun phrases the split between ergativity and accusativity will occur. Dixon generalizes from Silverstein’s animacy hierarchy, a revised version of which is shown in Figure 1, arguing that in all languages, the items toward the right of the spectrum are more likely to be in the Agent function, and the items to the left of the spectrum are more likely to be in the Object position. Dixon (1994, p. 84) summarizes this important linguistic universal as follows: 5 It is not useful, in my opinion, to talk of having “more,” “less,” or even “no” agency. As I hope I have demonstrated in this essay, agency is not a quantity that can be measured. Rather, researchers should focus on delineating different kinds of agency, or different ways in which agency is socioculturally mediated in particular times and places. LANGUAGE AND AGENCY 123 Figure 1 The animacy hierarchy (based on Dixon 1994, p. 85; Foley 1999, p. 210; revised from Silverstein 1976, p. 122). Put very roughly, a speaker will think in terms of doing things to other people to a much greater extent than in terms of things being done to him. In the speaker’s view of the world, as it impinges on him and as he describes it in language, he will be the quintessential agent.6 In other words, from the universal grammatical principles underlying all languages, we know that the most salient person in a linguistic interaction is the speaker, “I” (Foley 1999, p. 210). The second most salient person is the addressee, “you.” Both “I” and “you” are more salient, and therefore more likely to be found in the Agent position, than the absent participants in the interaction, ranked in the following order: third person pronouns, proper nouns, common nouns referring to humans, common nouns referring to animate nonhumans, and common nouns referring to inanimate objects. While there have been some challenges and revisions to this model (cf. Dixon 1994, pp. 83–94), the implications of a possibly universal tendency regarding the attribution of linguistic agency are worth considering. Note carefully, however, that we are talking about grammatical, not social, definitions of agency here. There are times when the grammatical and social categories of Agent will overlap, but this remains to be determined in each particular context (cf. Duranti 1994, p. 124). How can the grammatical details regarding Agent, Subject, and Object in particular languages be relevant to scholars interested in the social aspects of agency? 6 Dixon’s use of the masculine generic demonstrates yet another example of how grammatical categories predispose speakers to attribute agency more often to certain kinds of subjects. See McConnell-Ginet (1979), Silverstein (1985), and Waugh (1982) for analyses of markedness in the use of masculine and feminine pronouns. 124 AHEARN As DuBois (1987) notes, ergativity originates in discourse itself, in other words, in naturally occurring conversations. Derbyshire (1987, p. 319), for example, reports that in many Amazonian languages, when a noun phrase describing a highly ranked person is the subject in a transitive clause, the accusative pattern is followed, whereas when a noun phrase describing the higher ranked person is the object (a more marked, or unexpected, occurrence), the ergative pattern is followed. In English, LaFrance (1992) has shown that when subjects are asked to supply plausible scenarios of events that might have preceded and followed a set of sentences alternating male and female subjects and objects, they demonstrate a linguistic bias against women that she calls “the disappearing agent effect.” Her findings indicate that if a sentence is phrased such that a female is described as doing something or feeling something, especially with respect to a male, then she fades from causal view, but when she is on the receiving end of someone else’s actions, then the subject or source of these events, rather than she herself, is highlighted (LaFrance 1992, p. 341). Although these responses were elicited rather than taken from naturally occurring conversations, ethnographically informed investigations of this phenomenon demonstrate exactly how these linguistic usages reflect, reinforce, and sometimes reconfigure agency and status hierarchies in the society. Duranti’s From Grammar to Politics: Linguistic Anthropology in a Western Samoan Village (1994) provides just such an ethnographically rich example of how attention to linguistic forms can shed light on human agency.7 Duranti maintains that the Samoans’ use of ergative markers reveals how they attribute agency, especially in cases of praise or blame. Powerful individuals are more likely to use the ergative marker when they want to accuse someone of a malicious act, whereas less powerful individuals try to resist such accusations by suggesting alternative linguistic definitions of events. Thus, Duranti’s “grammar of praising and blaming” demonstrates how agency is expressed in, and shaped by, the linguistic forms that a socially and linguistically embedded speaker uses. AGENCY IN LINGUISTIC ANTHROPOLOGY Linguistic anthropologists seek to understand how grammatical categories “loosely structure” speakers and therefore look carefully at how speech both shapes and reflects social and cultural realities. For these reasons, they are well situated to contribute to the scholarship on agency. Indeed, long before Giddens first popularized the term agency, linguistic anthropologists (and some scholars in related fields such as discourse analysis, ethnomethodology, and sociolinguistics) were writing about language as a form of social action. For years, linguistic anthropologists have examined specific speech events in order to illuminate how people think about their own and others’ actions. By analyzing grammatical markers, pronoun use, 7 See also Duranti & Ochs (1990). LANGUAGE AND AGENCY 125 turn taking, narrative structures, dispute resolution, overlapping utterances, and other linguistic features, linguistic anthropologists have looked to language for concrete examples of effective (and ineffective) social action. In the sections that follow, I present several paths linguistic anthropologists have taken to show how culture in all its forms emerges from everyday linguistic agency that is itself shaped by sociocultural formations. There are several bodies of literature not dealt with here, for although they make contributions to the study of language and agency, they are summarized well elsewhere. The burgeoning area of language ideology research, for example, is summed up in Kroskrity (2000) and Schieffelin et al (1998). Language change, creolization, and bilingualism are also not treated here (cf. DeGraff 1999, Lightfoot 1999, Romaine 1995). Nevertheless, while the following sections touch on only some of the many areas relevant to language and agency that have been explored by linguistic anthropologists, they illustrate the important contributions linguistic anthropology has to make to social theory as a whole. Language and Gender One of the areas within linguistic anthropology most centrally concerned with questions of agency is the field of language and gender. The scholarship in this area generally avoids relying on a definition of agency as resistance, which can be found in much of the gender literature in other fields, and instead draws on more nuanced understandings from within linguistics and sociology of language as social action. While many of the language and gender scholars do not use the term agency in their work, they explore the relationship between linguistic practices and social structures in ways that contribute to our understanding of the concept of agency. Dozens of articles demonstrating how gender as a social construct emerges from particular linguistic interactions are contained within a handful of indispensable anthologies (Bergvall et al 1996, Bucholtz et al 1994, 1999, Hall & Bucholtz 1995, Hall et al 1992, Livia & Hall 1997, Philips et al 1987, Roman et al 1994, Tannen 1993). The interested reader can find within these volumes studies that describe, for example, how phone sex workers exercise ambiguous agency by using traditionally “powerless,” stereotypically feminine speech to become economically independent (Hall 1995); how preschoolers’ dispute resolution practices reflect and shape their developing gendered identities (Sheldon 1993); and how gay men produce coming-out narratives full of references to personal agency regarding the learning of distinctively gay ways of talking (Leap 1999). In an insightful essay that traces the intersections between practice theory and feminist theory, McElhinny (1998) identifies four scholars in the field of language and gender—Goodwin, Ochs, and the writing team of Eckert & McConnellGinet—who have made important contributions to both theoretical realms, and she urges that they be added to the canon. Goodwin, deservedly well known for her meticulously researched book, He-Said-She-Said: Talk as Social Organization 126 AHEARN among Black Children, focuses on “situated activities” rather than on whole societies or particular individuals because such a unit of analysis enables her to demonstrate how stereotypes about women’s speech become untenable when moving from one activity to another (Goodwin 1990, p. 9). Goodwin discovered that the girls on Maple Street did indeed use talk for different purposes than boys did when interacting among themselves, but when interacting with boys, they frequently took on the boys’ speech patterns, at times even outperforming them in verbal contests. She concludes, “This analysis has examined ways in which aspects of gender are manifested in speech activities, but more important, I have investigated how speech events can themselves provide for social organization, shaping alignment and social identities of participants to the present interaction” (Goodwin 1990, p. 286; emphasis in the original). Goodwin’s work calls attention to the many different ways that agency can be exercised linguistically and to the importance of looking closely at linguistic as well as sociocultural contexts when attempting to understand social dynamics and social change. Ochs’ work treating language socialization as a lifelong activity also provides us with important insight into the microprocesses of social change and continuity (Ochs 1988, 1992, 1996; McElhinny 1998, p. 168). Because people are constantly learning new ways to speak and act for particular sociocultural contexts, a close examination of this learning process in children and adults can shed light on the slippage between social reproduction and social transformation (to which practice theorists such as Bourdieu and Giddens allude but fail to elucidate). Citing Bourdieu and Giddens, Ochs notes, “This focus on language practices as resources for socializing social and cultural competence links language socialization research to post-structural sociological paradigms that portray social structures as outcomes of social practices . . .” (Ochs 1996, p. 408; emphasis in the original). Ochs, often in collaboration with Schieffelin, looks closely at indexicality, honorific pronoun use, word order, case-markings, and other grammatical features to investigate how linguistic practices encode and socialize information about society and culture (Ochs & Schieffelin 1983, 1984, Schieffelin & Ochs 1986). In her research on Kaluli children’s language socialization, Schieffelin (1990, p. 239) concludes that, “The microethnographic methods used in this study enable one to specify and interpret the words, interactions, relationships, and contexts in which cultural meanings are displayed to young children and reproduced by them. . . . This study shows how language is a resource for social theory.” This latter statement, while true, is too modest. Linguistic anthropologists do not merely provide social theorists with the “resource” of linguistic data; they also contribute unique insights to the process of theory building. In focusing on language acquisition and socialization, Ochs & Schieffelin contribute to our understanding of the microprocesses of social reproduction, thereby helping us identify the potential slippages between social reproduction and social transformation. Eckert & McConnell-Ginet also advance our understanding of linguistic and social practices in their individual and joint research on gendered social categories LANGUAGE AND AGENCY 127 in a Detroit area high school (Eckert 1989, Eckert & McConnell-Ginet 1992, 1995; cf. McElhinny 1998, p. 171ff.). They write, “Language is a primary tool people use in constituting themselves and others as ‘kinds’ of people in terms of which attributes, activities, and participation in social practice can be regulated” (Eckert & McConnell-Ginet 1995, p. 470). Perhaps their most significant contribution to practice theory is their emphasis on a “community of practice” (cf. Lave & Wenger 1991), which they define as “an aggregate of people who come together around mutual engagement in an endeavor. Ways of doing things, ways of talking, beliefs, values, power relations—in short, practices—emerge in the course of this mutual endeavor” (Eckert & McConnell-Ginet 1992, p. 464). The concept of communities of practice offers scholars a processual yet structural unit that can easily be viewed as both constitutive of, and constituted by, its participants. In placing linguistic and social practices within the contexts of communities of practice, Eckert and McConnell-Ginet contribute to a more nuanced view of the varying ways in which agency is socioculturally constrained and enabled. Literacy Practices Language, of course, can be written as well as spoken. Another field of scholarship well situated to make significant contributions to our understanding of language and agency, therefore, is the study of literacy practices. Within linguistic anthropology and related disciplines in recent years there has been a theoretical debate, summarized nicely by Besnier (1995), Collins (1995), and Street (2001), regarding how literacy should be defined and studied. On one side of the issue are scholars like Goody & Watt (1963), who were early proponents of what Street (1984, 1993, 2001) has called the “autonomous” model of literacy. Goody and other supporters of the autonomous model maintain that the advent of literacy in a society will cause the same social and psychological effects, no matter which society is being studied. These scholars “conceptualise literacy in technical terms, treating it as independent of social context, an autonomous variable whose consequences for society and cognition can be derived from its intrinsic character” (Street 1984, p. 5). Another proponent of the autonomous model, Ong (1982, pp. 14–15), asserts boldly that “without writing, human consciousness cannot achieve its fuller potentials, cannot produce other beautiful and powerful creations. In this sense, orality needs to produce and is destined to produce writing.” Ong, Goody, and others who espouse the autonomous model see a “Great Divide” separating “oral” societies from “literate” ones—a gap similar to the one turn-of-thecentury anthropologists used to claim existed between “primitive” and “civilized” societies. On the opposing side of the issue are those scholars engaging in what Street (2001, p.10) calls New Literacy Studies. Researchers such as Street himself (1984, 1993, 2001), Basso (1989[1974]), Baynham (1995), Besnier (1995), and Finnegan (1988) favor an “ideological” model for studying literacies, an approach that has 128 AHEARN benefited from, and contributed to, practice theory. Besnier (1995, p. 5) describes the goals of this approach as follows: “Rather than seeking an overarching and context-free characterization of the cognitive and social consequences of literacy, proponents of the ideological model focus on the activities, events, and ideological constructs associated with particular manifestations of literacy.” This approach examines the specific ramifications of the advent of literacy in each society, claiming that there are no universal attributes of literate societies and maintaining that it is impossible for literacy skills to be acquired neutrally. Most anthropologists agree with Baynham (1995, p. 71) that it is important to understand literacy as a form of social practice (or agency), and to investigate the way it interacts with ideologies and institutions to shape and define the possibilities and life paths of individuals. My own work on Nepali love letters derives inspiration from the work of Barton, Besnier, Street, and others who have explored the manifestations of various literacies in their social contexts (Barton & Hall 2000, Barton & Hamilton 1998, Barton et al 2000). In Junigau, Nepal, newly literate young women are applying their literacy skills in a novel form of courtship: love letters that echo the development discourses and changing notions of agency that can be found elsewhere in the society (Ahearn 2000b, 2001). This scholarship is an example of what Besnier (1995, p. 9) calls “event-centered studies” of literacy. He defines such studies as ethnographic investigations into the ways that literacy derives its meaning from the broader context in which it is practiced, and the ways that aspects of the situation acquire meaning from acts of reading and writing. Because cultural meanings are often constituted through literacy practices as well as through verbal interactions, scholars interested in the role of different types of agency (oral and literate) in the reproduction and transformation of cultural meanings can benefit from the work of researchers in this field. Dialogic Approaches Many linguistic anthropologists interested in agency (including some mentioned in previous sections) are taking a dialogic approach following Bakhtin (1981, 1984, 1993) and, in a few cases, the Soviet psychologist Vygotsky (1978, 1987, Holland et al 1998, Wertsch et al 1993). In a statement that summarizes the approach to language and agency espoused by these scholars, Bakhtin (1984, p. 183) notes that “Language lives only in the dialogic interaction of those who make use of it.” Mannheim & Tedlock (1995, p. 4) explain that dialogue, which etymologically refers to talk (logos) that goes back and forth (dia), can involve any or all of the following: straightforward verbal exchange, a social field across which multiple voices and multiple cultural logics contend, or a text that is multivocal and egalitarian rather than univocal and authoritarian. In all cases, however, the traditional relationship between structure and action, in which action is treated as a reflection of a prior structure, is rejected in favor of one in which structure emerges through situated action. Words or texts are socially situated by, not created by, individuals (Mannheim & Tedlock 1995, p. 5). LANGUAGE AND AGENCY 129 Locating language, culture, and agency in the interstices between people, rather than within individuals themselves, requires a different way of thinking about and studying linguistic and cultural interactions. While it may appear that a dialogic approach precludes the possibility of studying individuals, in their edited volume, The Dialogic Emergence of Culture, Tedlock & Mannheim (1995) provide numerous examples of how scholars can study the words and actions of particular people while also situating those individuals within sociocultural fields that are always fluid and in process. In Hill’s contribution to the volume, for example, she draws on Bakhtin’s notion of heteroglossia to analyze how the narrator Don Gabriel, a native Mexicano speaker, tells of his son’s murder. The narrative contains dysfluencies when he speaks in Spanish of profit motives for the murder, but it remains elegantly fluent when he speaks in Mexicano about the loss of his son. “The narrative reveals a veritable kaleidoscope of ‘emotional selves,’ which are all art, distributed in fragments across the rhetorical systems of the narrative,” writes Hill (1995, p. 139). Among these selves is one that positions itself squarely within the “domain of ongoing ideological resistance to a capitalist ideology,” an ideology indexed by the use of Spanish. By locating multiple socially embedded voices among and within individuals in this narrative, Hill demonstrates the usefulness for linguistic and cultural anthropologists alike of a dialogic approach to the study of language and agency. Another contribution to the Tedlock & Mannheim volume, a reprint of a 1983 paper by McDermott & Tylbor (1995[1983]), looks at how one short classroom interaction among several students and a teacher produces an outcome not predictable from analysis of the words alone. Rosa, a first-grade student who cannot read, constantly calls out for a turn at reading aloud—and yet on close examination, Rosa, her classmates, and the teacher all seem to be colluding through the use of subtle gestures and timing cues in order not to give Rosa a chance to read aloud (McDermott & Tylbor 1995[1983], p. 223). This collusional approach to understanding language and agency “refers to how members of any social order must constantly help each other posit a particular state of affairs, even when such a state would be in no way at hand without everyone so proceeding” (McDermott & Tylbor 1995[1983], p. 219). Just as meanings and outcomes are coconstructed, so is agency. Basso (1996) explores the dynamics of “place-making” among the Western Apache, whose storytelling activities dialogically produce and reproduce historical knowledge and moral wisdom. Place-making through storytelling, Basso claims (1996, p. 7), is a way of constructing social traditions and identities, a way of “doing human history.” Drawing on Bakhtin’s (1981, p. 7) notion of chronotopes, which are places where time and space have fused to create culturally and historically charged locations, Basso describes how historical tales themselves have agency and shape the moral judgments that Apaches make about themselves and other people (Basso 1996, p. 62). The landscape itself also exercises agency in this process, as the historically and morally significant places serve to remind Apaches of the stories associated with them. Through telling stories associated with particular places, the Western Apache coconstruct a spatial, temporal, and cultural 130 AHEARN world that then serves to shape their future conduct. This is truly an instance of agency extending “beyond the skin” (Bateson 1972, Wertsch et al 1993). Another important work that takes a dialogic approach to understanding language and agency is Hill & Irvine’s anthology, Responsibility and Evidence in Oral Discourse (1993). Hill & Irvine write that the connection between knowledge and agency is of central importance to an approach that emphasizes dialogicality and the social construction of meaning. Interpreting events, establishing facts, conveying opinion, and constituting interpretations as knowledge are all activities involving socially situated participants, who are agents in the construction of knowledge and agents when they act on what they have come to know, believe, suspect, or opine (Hill & Irvine 1993, p. 2). As one example of such activities, Besnier’s contribution to this volume demonstrates how residents on Nukulaelae Atoll in Polynesia take advantage of the multifunctionality of reported speech in order to inject a greater or lesser amount of affect into an utterance, thereby manipulating the audience’s perception of the quoted individual (Besnier 1993). Besnier argues that the meanings of Nukulaelae utterances—like those of everyone else—cannot be understood without locating the speakers in temporally specific sociocultural fields of relationality. Linguistic agency is molded by these sociocultural fields, which it then proceeds to recreate or reconfigure. CONCLUSION To conclude, let me reiterate two points I have attempted to make throughout this review. First, scholars who choose to use the term agency should define it carefully. The provisional definition I offered at the outset of this essay—that agency refers to the socioculturally mediated capacity to act—leaves a great deal unspecified. For example, where is agency located? Must agency be human, individual, collective, intentional, or conscious? Some studies of agency reinforce received notions about western atomic individualism, while others deny agency to individuals, attributing it instead only to discourses or social forces. It is absolutely crucial that theorists consider the assumptions about personhood, desire, and intentionality that might unwittingly be built into their analyses. No matter how agency is defined—and it can be defined in any number of ways—implications for social theory abound. Scholars using the term must define it clearly, both for themselves and for their readers. For anthropologists in particular, it is important to avoid treating agency as a synonym for free will or resistance. One fruitful direction for future research may be to begin to distinguish among types of agency—oppositional agency, complicit agency, agency of power, agency of intention, etc.—while also recognizing that multiple types are exercised in any given action. By doing this, we might gain a more thorough understanding of the “complex and ambiguous agency” (MacLeod 1992) that always surrounds us. Second, I hope I have demonstrated how focusing on linguistic interactions can provide important clues for scholars interested in the micro- and macro-processes LANGUAGE AND AGENCY 131 of agency. Because language and culture are so tightly interwoven, neither should be studied in isolation from the other, especially when a researcher seeks to understand a concept as complex as agency. While practice theory offers several promising avenues that treat agency and structure as mutually constitutive, I maintain that attending closely to linguistic structures and practices can shed even more light on practice theorists’ main dilemma: how social reproduction becomes social transformation. Because grammatical categories within particular languages construct the roles of Subject, Agent, and Object differently, researchers can benefit from examining such categories carefully when listening to how people attribute responsibility, credit, or blame for an event. Three areas in which scholars are skillfully combining a close examination of language with a concern for broader social issues are the fields of language and gender, literacy practices, and dialogic approaches to language. Such nuanced treatments of language and action serve as excellent models for the development of a more sophisticated understanding of agency. ACKNOWLEDGMENTS Many people read and commented on earlier drafts of this essay, including my wonderful in-house editor, Rick Black; my eloquent and loyal critic, Peter Laipson; and the following generous and knowledgeable colleagues: John Adams, Anne Blackburn, Marcia-Anne Dobres, Michele Gamburd, Karl Heider, Louise Jennings, Alice Kasakoff, Ann Kingsolver, Bruce Mannheim, and Gail Wagner. I am grateful to Sherry Ortner for introducing me to practice theory, to Bruce Mannheim for pointing out to me the connection between language and agency, and to Tom Fricke for helping me understand how agency works in Nepal. I also benefited enormously from the lively discussions of the Agency Reading Group at the University of South Carolina and from the research assistance of Gail Davis, Changyong Liao, and Jennifer Whetstone. This review is a product of collective rather than individual agency, but only I am to blame for any weaknesses that remain. Visit the Annual Reviews home page at www.AnnualReviews.org LITERATURE CITED Abu-Lughod L. 1990. The romance of resistance: tracing transformations of power through Bedouin women. Am. Ethnol. 17(1):41–55 Ahearn LM. 1998. “A twisted rope binds my waist”: locating constraints on meaning in a Tij songfest. J. Ling. Anthropol. 8(1):60–86 Ahearn LM. 2000a. Agency. J. Ling. Anthropol. 9(1):9–12 Ahearn LM. 2000b. True traces: love letters and social transformation in Nepal. See Barton & Hall 2000, pp. 199–207 Ahearn LM. 2001. Invitations to Love: Literacy, Love Letters, and Social Change in Nepal. Ann Arbor: Univ. Michigan Press. Forthcoming Andermahr S. 1997. Agency. In A Concise Glossary of Feminist Theory, ed. S Andermahr, pp. 11–12. New York: Arnold Appadurai A. 1991. Afterword. In Gender, 132 AHEARN Genre, and Power in South Asian Expressive Tradition, ed. A Appadurai, FJ Korom, MA Mills , pp. 467–76. Philadelphia: Univ. Penn. Press Archer MS. 1988. Culture and Agency: The Place of Culture in Social Theory. Cambridge, UK: Cambridge Univ. Press Austin JL. 1962. How to Do Things with Words. London: Oxford Univ. Press Bakhtin MM. 1981. The Dialogic Imagination: Four Essays, ed. M Holquist. Transl. C Emerson, M Holquist. Austin: Univ. Texas Press (From Russian) Bakhtin MM. 1984. Problems of Dostoevsky’s Poetics. Theory and History of Literature, Vol. 8. Transl. C Emerson. Minneapolis: Univ. Minn. Press (From Russian) Bakhtin MM. 1993. Toward a Philosophy of the Act, ed. V Liapunov, M Holquist. Transl. V Liapunov. Austin: Univ. Texas Press (From Russian) Bartky SL. 1995. Subjects and agents: the question for feminism. See Gardiner 1995, pp. 194–207 Barton D, Hall N, eds. 2000. Letter Writing as a Social Practice. Amsterdam: Benjamins Barton D, Hamilton M, eds. 1998. Local Literacies: Reading and Writing in One Community. London: Routledge Barton D, Hamilton M, Ivanič R, eds. 2000. Situated Literacies: Reading and Writing in Context. London: Routledge Basso K. 1989(1974). The ethnography of writing. See Bauman & Sherzer 1989, pp. 425–32 Basso K. 1996. Wisdom Sits in Places: Landscape and Language Among the Western Apache. Albuquerque: Univ. New Mexico Press Bateson G. 1972. Steps to an Ecology of Mind: A Revolutionary Approach to Man’s Understanding of Himself. New York: Ballantine Bauman R, Sherzer J, eds. 1989(1974). Explorations in the Ethnography of Speaking. New York: Cambridge Univ. Press. 2nd ed. Baynham M. 1995. Literacy Practices: Investigating Literacy in Social Contexts. London: Longman Berger PL, Luckmann T. 1966. The Social Con- struction of Reality: A Treatise in the Sociology of Knowledge. New York: Doubleday Bergvall VL, Bing JM, Freed AF, eds. 1996. Rethinking Language and Gender Research: Theory and Practice. London: Longman Besnier N. 1993. Reported speech and affect on Nukulaelae Atoll. See Hill & Irvine 1993, pp. 161–81 Besnier N. 1995. Literacy, Emotion, and Authority: Reading and Writing on a Polynesian Atoll. New York: Cambridge Univ. Press Bittner M, Hale K. 1996. Ergativity: toward a theory of a heterogeneous class. Ling. Inq. 27(4):531–604 Bourdieu P. 1977. Outline of a Theory of Practice. Cambridge, UK: Cambridge Univ. Press Bourdieu P. 1991. Language and Symbolic Power. Transl. G Raymond, M. Adamson. Cambridge, MA: Harvard Univ. Press (From French) Brenneis D, Macaulay RKS, eds. 1996. The Matrix of Language: Contemporary Linguistic Anthropology. Boulder, CO: Westview Bryant GA, Jary D, eds. 1991. Giddens’ Theory of Structuration: A Critical Appreciation. New York: Routledge Bucholtz M, Liang AC, Sutton LA, eds. 1999. Reinventing Identities: The Gendered Self in Discourse. New York: Oxford Univ. Press Bucholtz M, Liang AC, Sutton LA, Hines C, eds. 1994. Cultural Performances: Proceedings of the Third Berkeley Women and Language Conference. Berkeley, CA: Berkeley Women Language Group Burns TR. 1994. Two conceptions of human agency: rational choice theory and the social theory of action. See Sztompka 1994, pp. 197–249 Burns TR, Dietz T. 1994. Introduction to the series. See Sztompka 1994, pp. vii–xiv Butler J. 1997. Excitable Speech: A Politics of the Performative. New York: Routledge Chomsky N. 1965. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press Chomsky N. 1986. Knowledge of Language: Its Nature, Origin and Use. New York: Praeger Colapietro VM. 1989. Peirce’s Approach to the LANGUAGE AND AGENCY Self: A Semiotic Perspective on Human Subjectivity. Albany: SUNY Press Collins J. 1995. Literacy and literacies. Annu. Rev. Anthropol. 24:75–93 Comaroff J, Comaroff JL. 1997. Of Revelation and Revolution, Vol. 2. Chicago, IL: Univ. Chicago Press Comrie B. 1981. Language Universals and Linguistic Typology: Syntax and Morphology. Oxford, UK: Blackwell Cooper F. 1994. Conflict and connection: rethinking colonial African history. Am. Hist. Rev. 99:1515–45 Cooper F, Stoler AL, eds. 1997. Tensions of Empire: Colonial Cultures in a Bourgeois World. Berkeley: Univ. Calif. Press Daniel EV. 1984. Fluid Signs: Being a Person the Tamil Way. Berkeley: Univ. Calif. Press Davidson D. 1980(1971). Agency. In Essays on Actions and Events, ed. D Davidson, pp. 43– 61. Oxford: Clarendon Press Davies B. 1991. The concept of agency: a feminist poststructuralist analysis. Soc. Anal. 30:42–53 de Certeau M. 1984. The Practice of Everyday Life. Transl. S Rendall. Berkeley: Univ. Calif. Press (From French) DeGraff M, ed. 1999. Language Creation and Language Change: Creolization, Diachrony, and Development. Cambridge, MA: MIT Press Derbyshire DC. 1987. Morphosyntactic areal characteristics of Amazonian languages. Int. J. Am. Ling. 53:311–26 de Saussure F. 1986. Course in General Linguistics, ed. C Bally, A Sechehaye, with collab. A Riedlinger. Transl., annot. R Harris. La Salle, IL: Open Court (From French) Derrida J. 1972. Signature, event, context. Glyph 1:172–97 Desjarlais R. 1997. Shelter Blues: Sanity and Selfhood Among the Homeless. Philadelphia: Univ. Penn. Press Dissanayake W, ed. 1996. Narratives of Agency: Self-Making in China, India, and Japan. Minneapolis: Univ. Minn. Press Dixon RMW. 1994. Ergativity. New York: Cambridge Univ. Press 133 Dobres M-A. 2000. Technology and Social Agency: Outlining a Practice Framework for Archaeology. Oxford: Blackwell Dobres M-A, Robb J, eds. 2000. Agency in Archaeology. New York: Routledge DuBois J. 1987. The discourse basis of ergativity. Language 63:805–55 Duranti A. 1994. From Grammar to Politics: Linguistic Anthropology in a Western Samoan Village. Berkeley: Univ. Calif. Press Duranti A. 1997. Linguistic Anthropology. Cambridge, UK: Cambridge Univ. Press Duranti A, Goodwin C, eds. 1992. Rethinking Context: Language as an Interactive Phenomenon. Cambridge, UK: Cambridge Univ. Press Duranti A, Ochs E. 1990. Genitive constructions and agency in Samoan discourse. Stud. Lang. 14(1):1–23 Eckert P. 1989. Jocks and Burnouts: Social Categories and Identity in High School. New York: Teach. College Eckert P, McConnell-Ginet S. 1992. Think practically and look locally: language and gender as community-based practice. Annu. Rev. Anthropol. 21:461–90 Eckert P, McConnell-Ginet S. 1995. Constructing meaning, constructing selves: snapshots of language, gender, and class from Belten High. See Hall & Bucholtz 1995, pp. 469– 507 Eco U. 1990. The Limits of Interpretation. Bloomington: Indiana Univ. Press Farnell B. 2000. Getting out of the habitus: an alternative model of dynamically embodied social action. J. R. Anthropol. Inst. 6:397– 418 Finnegan R. 1988. Literacy and Orality: Studies in the Technology of Communication. Oxford, UK: Blackwell Foley WA. 1999. Information structure. In Concise Encyclopedia of Grammatical Categories, ed. K Brown, J Miller, consulting ed. RE Asher, pp. 204–13. Amsterdam: Elsevier Foucault M. 1977. Discipline and Punish: The Birth of the Prison. New York: Pantheon Foucault M. 1978. The History of Sexuality, Vols. 1, 2. New York: Pantheon 134 AHEARN Fraser N. 1992. Introduction: revaluing French feminism. In Revaluing French Feminism: Critical Essays on Difference, Agency, and Culture, ed. N Fraser, SL Bartky, pp. 1–24. Bloomington: Indiana Univ. Press Gamburd MR. 2000. The Kitchen Spoon’s Handle: Transnationalism and Sri Lanka’s Migrant Housemaids. Ithaca, NY: Cornell Univ. Press Gardiner JK, ed. 1995. Provoking Agents: Gender and Agency in Theory and Practice. Urbana: Univ. Illinois Press Garfinkel H. 1967. Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall Giddens A. 1979. Central Problems in Social Theory: Action, Structure and Contradiction in Social Analysis. Berkeley: Univ. Calif. Press Giddens A. 1984. The Constitution of Society: Outline of the Theory of Structuration. Berkeley: Univ. Calif. Press Goddard VA, ed. 2000. Gender, Agency and Change: Anthropological Perspectives. New York: Routledge Goffman E. 1974. Frame Analysis: An Essay on the Organization of Experience. New York: Harper Row Goffman E. 1981. Forms of Talk. Philadelphia: Univ. Penn. Press Goldhagen DJ. 1996. Hitler’s Willing Executioners: Ordinary Germans and the Holocaust. New York: Knopf Goodwin MH. 1990. He-Said-She-Said: Talk as Social Organization Among Black Children. Bloominton: Indiana Univ. Press Goody J, Watt I. 1963. The consequences of literacy. Comp. Stud. Hist. Soc. 5:306–26, 332– 45 Gumperz JJ, Hymes D, eds. 1964. The Ethnography of Communication. (Am. Anthropol. 66(6), Special Publication) Washington, DC: Am. Anthropol. Assoc. Gumperz JJ, Levinson SC, eds. 1996. Rethinking Linguistic Relativity. Cambridge, UK: Cambridge Univ. Press Hall K. 1995. Lip service on the fantasy lines. See Hall & Bucholtz 1995, pp. 183–216 Hall K, Bucholtz M, eds. 1995. Gender Ar- ticulated: Language and the Socially Constructed Self. New York: Rutgers Hall K, Bucholtz M, Moonwomon B, eds. 1992. Locating Power: Proceedings of the Second Berkeley Women and Language Conference. Berkeley, CA: Berkeley Women Language Group Halperin DM. 1995. Saint Foucault: Towards a Gay Hagiography. New York: Oxford Univ. Press Hanks WF. 1996. Language and Communicative Practices. Boulder, CO: Westview Hill JH. 1995. The voices of Don Gabriel: responsibility and self in a modern Mexicano narrative. See Tedlock & Mannheim 1995, pp. 97–147 Hill JH, Irvine JT, eds. 1993. Responsibility and Evidence in Oral Discourse. Cambridge, UK: Cambridge Univ. Press Hill JH, Mannheim B. 1992. Language and world view. Annu. Rev. Anthropol. 21:381– 406 Holland D, Lachicotte W Jr, Skinner D, Cain C. 1998. Identity and Agency in Cultural Worlds. Cambridge, MA: Harvard Univ. Press Hoy DC, ed. 1986. Foucault: A Critical Reader. Oxford: Blackwell Hymes D. 1972. Models of the interaction of language and social life. In Directions in Sociolinguistics: The Ethnography of Communication, ed. JJ Gumperz, D Hymes, pp. 35– 71. New York: Holt, Rinehart, Winston Jackson M, Karp I, eds. 1990. Personhood and Agency: The Experience of Self and Other in African Cultures. Washington, DC: Smithsonian Inst. Jakobson R. 1960. Closing statement: linguistics and poetics. In Style in Language, ed. T Sebeok, pp. 398–429. Cambridge, MA: MIT Press Jeffery P, Basu A, eds. 1998. Appropriating Gender: Women’s Activism and Politicized Religion in South Asia. New York: Routledge Jeffery P, Jeffery R. 1996. Don’t Marry Me to a Plowman!: Women’s Everyday Lives in Rural North India. Boulder, CO: Westview Karp I. 1986. Agency and social theory: a LANGUAGE AND AGENCY review of Anthony Giddens. Am. Ethnol. 13(1):131–37 Keane W. 1997. Religious language. Annu. Rev. Anthropol. 26:47–71 Keenan EL. 1984. Semantic correlates of the ergative/absolutive distinction. Linguistics 22:197–223 Kroskrity PV, ed. 2000. Regimes of Language: Ideologies, Polities, and Identities. Santa Fe, NM: School Am. Res. Press Kumar N, ed. 1994. Women as Subjects: South Asian Histories. Charlottesville: Univ. Virginia Press LaFrance M. 1992. When agents disappear: how gender affects the implicit causality of interpersonal verbs. See Hall et al. 1992, pp. 338–43 Lalu P. 2000. The grammar of domination and the subjection of agency: colonial texts and modes of evidence. Hist. Theory 39(4):45–68 Lave J, Wenger E. 1991. Situated Learning: Legitimate Peripheral Participation. Cambridge, UK: Cambridge Univ. Press Leap W. 1999. Language, socialization, and silence in gay adolescence. See Bucholtz et al. 1999, pp. 259–72 Lightfoot D. 1999. The Development of Language: Acquisition, Change, and Evolution. Malden, MA: Blackwell Livia A, Hall K, eds. 1997. Queerly Phrased: Language, Gender, and Sexuality. Oxford, UK: Oxford Univ. Press Lucy JA. 1992. Language Diversity and Cognitive Development: A Reformulation of the Linguistic Relativity Hypothesis. Cambridge, UK: Cambridge Univ. Press MacLeod AE. 1992. Hegemonic relations and gender resistance: the new veiling as accommodating protest in Cairo. Signs 17(3):533– 57 Mann PS. 1994. Micro-Politics: Agency in a Postfeminist Era. Minneapolis: Univ. Minn. Press Mannheim B, Tedlock D. 1995. Introduction. See Tedlock & Mannheim 1995, pp. 1–32 Marriott McK. 1976. Hindu transactions: diversity without dualism. In Transaction and Meaning: Directions in the Anthropology 135 of Exchange and Symbolic Behavior, ed. B Kapferer, pp. 109–42. Philadelphia: Inst. Study Hum. Issues Marx K. 1978(1852). The eighteenth brumaire of Louis Bonaparte. In The Marx-Engels Reader, ed. RC Tucker, pp. 594–617, 2nd ed. McConnell-Ginet S. 1979. Prototypes, pronouns, and persons. In Ethnolinguistics: Boas, Sapir, and Whorf Revisited, ed. M Mathiot, pp. 63–83. The Hague: Moulton McDermott RP, Tylbor H. 1995(1983). On the necessity of collusion in conversation. See Tedlock & Mannheim 1995, pp. 218–36 McElhinny B. 1998. Genealogies of gender theory: practice theory and feminism in sociocultural and linguistic anthropology. Soc. Anal. 42(3):164–89 McNay L. 2000. Gender and Agency: Reconfiguring the Subject in Feminist and Social Theory. Oxford, UK: Blackwell Messer-Davidow E. 1995. Acting otherwise. See Gardiner 1995, pp. 23–51 Mithun M. 1991. Active/agentive case marking and its motivations. Language 67(3):510–46 Moses AD. 1998. Structure and agency in the Holocaust: Daniel J. Goldhagen and his critics. Hist. Theory 37(2):194–222 Obeyesekere G. 1992. The Apotheosis of Captain Cook: European Myth-Making in the Pacific. Princeton, NJ: Princeton Univ. Press Ochs E. 1988. Culture and Language Development: Language Acquisition and Language Socialization in a Samoan Village. Cambridge, UK: Cambridge Univ. Press Ochs E. 1992. Indexing gender. See Duranti & Goodwin 1992, pp. 335–58 Ochs E. 1996. Linguistic resources for socializing humanity. See Gumperz & Levinson 1996, pp. 406–37 Ochs E, Schegloff EA, Thompson SA, eds. 1996. Interaction and Grammar. Cambridge UK: Cambridge Univ. Press Ochs E, Schieffelin BB. 1983. Acquiring Conversational Competence. Boston: Routledge & Kegan Paul Ochs E, Schieffelin BB. 1984. Language acquisition and socialization: three developmental stories. In Culture Theory: Essays on Mind, 136 AHEARN Self, and Emotion, ed. RA Shweder, RA Levine, pp. 276–320. Cambridge, UK: Cambridge Univ. Press O’Hara DT. 1992. Radical Parody: American Culture and Critical Agency After Foucault. New York: Columbia Univ. Press Ong WJ. 1982. Orality and Literacy: The Technologizing of the Word. London: Methuen Ortner SB. 1984. Theory in anthropology since the sixties. Comp. Stud. Soc. Hist. 26(1):126– 66 Ortner SB. 1989. High Religion: A Cultural and Political History of Sherpa Buddhism. Princeton, NJ: Princeton Univ. Press Ortner SB. 1995. Resistance and the problem of ethnographic refusal. Comp. Stud. Soc. Hist. 37(1):173–93 Ortner SB. 1996. Toward a feminist, minority, postcolonial, subaltern, etc., theory of practice. In Making Gender: The Politics and Erotics of Culture, pp. 1–20. Boston, MA: Beacon Ortner SB. 2001. Specifying agency: the Comaroffs and their critics. Interventions. Forthcoming Peirce CS. 1955. Philosophical Writings of Peirce. New York: Dover Philips SU. 1972. Participant structures and communicative competence: Warm Springs children in community and classroom. In Functions of Language in the Classroom, ed. CB Cazden, VP John, D Hymes, pp. 370–94. New York: Columbia Teach. Philips SU, Steele S, Tanz C, eds. 1987. Language, Gender, and Sex in Comparative Perspective. Cambridge, UK: Cambridge Univ. Press Pickering A. 1995. The Mangle of Practice: Time, Agency, and Science. Chicago: Univ. Chicago Press Pieters J. 2000. New historicism: postmodern historiography between narrativism and heterology. Hist. Theory 39(1):21–38 Plank F, ed. 1979. Ergativity: Towards a Theory of Grammatical Relations. London: Acad. Press Pomper P. 1996. Historians and individual agency. Hist. Theory 35(3):281–308 Pruyn M. 1999. Discourse Wars in GothamWest: A Latino Immigrant Urban Tale of Resistance and Agency. Boulder, CO: Westview Reddy M. 1979. The conduit metaphor—a case of frame conflict in our language about language. In Metaphor and Thought, ed. A Ortony, pp. 284–324. Cambridge, UK: Cambridge Univ. Press Romaine S. 1995. Bilingualism. Oxford, UK: Blackwell. 2nd ed. Roman C, Juhasz S, Miller C, eds. 1994. The Women and Language Debate: A Sourcebook. New York: Routledge Rovane C. 1998. The Bounds of Agency: An Essay in Revisionary Metaphysics. Princeton, NJ: Princeton Univ. Press Sacks H. 1992. Lectures on Conversation, Vols. 1,2. Cambridge, UK: Blackwell Sahlins M. 1981. Historical Metaphors and Mythical Realities: Structure in the Early History of the Sandwich Islands Kingdom. Ann Arbor: Univ. Mich. Press Said EW. 1983. The World, the Text, and the Critic. Cambridge, MA: Harvard Univ. Press Sapir E. 1933(1949). Language. See Sapir 1949, pp. 7–32 Sapir E. 1949. Selected Writings of Edward Sapir, ed. D Mandelbaum. Berkeley: Univ. Calif. Press Schieffelin BB. 1990. The Give and Take of Everyday Life: Language Socialization of Kaluli Children. Cambridge, UK: Cambridge Univ. Press Schieffelin BB, Ochs E. 1986. Language Socialization Across Cultures. Cambridge, UK: Cambridge Univ. Press Schieffelin BB, Woolard K, Kroskrity PV, eds. 1998. Language Ideologies. Oxford, UK: Oxford Univ. Press Scott JC. 1985. Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven, CT: Yale Univ. Press Scott JC. 1990. Domination and the Arts of Resistance: Hidden Transcripts. New Haven, CT: Yale Univ. Press Scott JW. 1988. Gender and the Politics of History. New York: Columbia Univ. Press LANGUAGE AND AGENCY Searle JR. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge, UK: Cambridge Univ. Press Segal JM. 1991. Agency and Alienation: A Theory of Human Presence. Savage, MD: Rowman Littlefield Sewell WH. Jr. 1992. A theory of structure: duality, agency, and transformation. Am. J. Sociol. 98(1):1–29 Sheldon A. 1993. Pickle fights: gendered talk in preschool disputes. See Tannen 1993, pp. 83–109 Silverstein M. 1976. Hierarchy of features and ergativity. In Grammatical Categories in Australian Languages, ed. RMW Dixon, pp. 112–71. Canberra: Aust. Inst. Aborig. Stud. Silverstein M. 1985. Language and the culture of gender: at the intersection of structure, usage, and ideology. In Semiotic Mediation: Sociocultural and Psychological Perspectives, ed. E Mertz, RJ Parmentier, pp. 219–59. Orlando, FL: Acad. Skinner D, Pach A III, Holland D. 1998. Selves in Time and Place: Identities, Experience, and History in Nepal. Lanham, MD: Rowman Littlefield Small MF. 1993. Female Choices: Sexual Behavior of Female Primates. Cornell, NY: Cornell Univ. Press Spender D. 1980. Man Made Language. Boston: Routledge Kegan Paul Stoler AL. 1995. Race and the Education of Desire: Foucault’s History of Sexuality and the Colonial Order of Things. Durham, NC: Duke Univ. Press Street B. 1984. Literacy in Theory and Practice. Cambridge, UK: Cambridge Univ. Press Street B, ed. 1993. Cross-cultural approaches to literacy. Cambridge, UK: Cambridge Univ. Press Street B, ed. 2001. Literacy and Development: Ethnographic Perspectives. New York: Routledge 137 Sztompka P. 1991. Society in Action: The Theory of Social Becoming. Chicago: Univ. Chicago Press Sztompka P, ed. 1994. Agency and Structure: Reorienting Social Theory. International Studies in Global Change, Vol. 4. Langhorne, PA: Gordon Breach Tannen D, ed. 1993. Gender and Conversational Interaction. New York: Oxford Univ. Press Taylor C. 1985. Human Agency and Language: Philosophical Papers, Vol. 1. Cambridge, UK: Cambridge Univ. Press Tedlock D, Mannheim B. 1995. The Dialogic Emergence of Culture. Urbana, IL: Univ. Ill. Press Urban G. 1991. A Discourse-Centered Approach to Culture: Native South American Myths and Rituals. Austin, TX: Univ. Texas Press Vygotsky LS. 1978. Mind in Society: The Development of Higher Psychological Processes, ed. M Cole, S Scribner, E Souberman. Cambridge, MA: Harvard Univ. Press Vygotsky LS. 1987. Thinking and Speech, ed., transl. N Minick. New York: Plenum (From Russian) Waugh LR. 1982. Marked and unmarked: a choice between unequals in semiotic structure. Semiotica 38:299–318 Wertsch JV, Tulviste P, Hagstrom F. 1993. A sociocultural approach to agency. In Contexts for Learning: Sociocultural Dynamics in Children’s Development, ed. EA Forman, N Minick, CA Stone, pp. 336–56. New York: Oxford Univ. Press Whorf BL. 1956. Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf. Cambridge, MA: MIT Press Williams R. 1977. Marxism and Literature. New York: Oxford Univ. Press Wittgenstein L. 1958. Philosophical Investigations. Transl. GEM Anscombe. Oxford: Blackwell. 2nd ed.