But Mom, you re wrong. Crop-tops are cute! and Other Epistemic Challenges. Sally Haslanger MIT, Department of Linguistics and Philosophy

Transcription

1 DRAFT: 4/14/07 But Mom, you re wrong. Crop-tops are cute! and Other Epistemic Challenges Sally Haslanger MIT, Department of Linguistics and Philosophy A study of the science of man is inseparable from an examination of the options between which men must choose. This means that we can speak here not only of error, but of illusion. We speak of illusion when we are dealing with something of greater substance than error, [it is] error which in a sense builds a counterfeit reality of its own [Such illusions] are more than errors in this sense: they are sustained by certain practices of which they are constitutive. (Taylor 1985/1971, 54) Certainly a good deal of men s tyranny over women can be observed through data, experiments, and research Many things can be known in this way.[but it does not] show that it is unnecessary or changeable, except speculatively, because what is not there is not considered real. Women s situation cannot be truly known for what it is, in the feminist sense, without knowing that it can be other than it is. By operating as legitimating ideology, the scientific standard for verifying reality can reinforce a growing indignation, but it cannot create feminism that was not already there. Knowing objective facts does not do what consciousness does. (MacKinnon 1989, ) I. Introduction 1 In the social realm, knowledge, or what purports to be knowledge, is entangled with the reality it represents. Social institutions are constituted, at least in part, by sets of shared beliefs and conventions; in this domain, thinking can make it so. Even initially false beliefs about social phenomena, if they catch on, can cause changes in the social world that result in the beliefs becoming true (Langton 2007). When social knowledge goes wrong, it may be because it has constituted a reality and perhaps accurately represents that reality that nevertheless falls short in some way. As a result, it is sometimes suggested that an epistemology of the social realm must not simply be concerned with whether a belief is justified and true. Following Taylor (see epigraph), the suggestion might be that the social reality created by the

2 belief is an illusion. But if it is, in what sense is it an illusion? Is it an illusion about what s possible? About what s good? And is an evaluation of what knowledge, e.g., common knowledge, creates a legitimate part of epistemology? Catharine MacKinnon s work repeatedly and forcefully raises the question how an epistemology of the social should proceed in oppressive social contexts. On MacKinnon s view (1989, see also epigraph), Consciousness raising, by contrast [to scientific inquiry] inquires into an intrinsically social situation, in the mixture of thought and materiality which comprises gender in its broadest sense. (MacKinnon 1989, 83) She continues, The process is transformative as well as perceptive, since thought and thing are inextricably and reciprocally constitutive of women s oppression (MacKinnon 1989, 84). Given the interdependence of social thought and reality, a change of meaning can transform the social world. 2 This calls, however, for a new branch of epistemology: This epistemology does not at all deny that a relation exists between thought and some reality other than thought, or between human activity (mental or otherwise) and the products of that activity. Rather, it redefines the epistemological issue from being a scientific one, the relation between knowledge and objective reality, to a problem of the relation of consciousness to social being. (MacKinnon 1989, 99) Setting aside the challenge of interpreting her positive view, she is raising an epistemological problem about what should be thought in those domains where what is thought (at least partly) both determines and is determined by its object. This problem is especially pressing when this occurs at a site of injustice. My goal in this essay is to provide some resources for developing a response. II. Are Crop-Tops Cute? To make this more concrete, consider the role of fashion in schools. 3 The belief that certain girls are wearing crop-tops that expose their midriff partly constitutes the fact that it is fashionable to wear such tops and causes many other girls to do the same. Plausibly, in such situations it becomes common knowledge that, say, seventh grade girls are wearing crop-tops this spring. 4 But, one might argue, it would be better if seventh grade (roughly age 12) girls were wearing ordinary midriff covering tops 2 Haslanger

3 instead (because the crop-tops sexualize the girls who wear them, further marginalize the chubby girls, etc.). So parents who are uncomfortable with the crop-top fashion, and yet find themselves faced with a daughter who is eager to join the crowd, might suggest to her that, e.g., she shouldn t care about being fashionable, that she shouldn t let what the other girls are doing determine her choices, that she is beautiful in her track suit. However, even if the daughter is individually able to retain her self-respect without bowing to the fashion trend, it may still be true that she will be marginalized if she doesn t conform and that the fashionable girls are sexualized (Warner 2007). Bucking conventions may be a partial solution that works for some individuals. But the problem is not individual. The situation would be better if seventh grade girls are wearing crop-tops this spring wasn t part of a set of beliefs that constitute common knowledge in the school (or the broader society). With this in mind, consider the following familiar dialogue: Daughter: Can I have some money to buy a crop-top like Ashley s to wear to school? Parents: You can have a new top, but not a crop-top. Crop-tops are too revealing. Daughter: But Mom[Dad], you re just wrong. Everyone knows that crop-tops are cute; and I don t want to be a dork. Parents: I m sorry, sweetie, crop-tops are not cute, and you won t be a dork if you wear your track suit. Under the circumstances it seems that there is something right about Daughter s reply to Parents: she is saying something true about her social milieu, and their reply is not enough. And yet, aren t the parents right? One might initially assume that in this conversation there is a disagreement over the truth-value of the following claims: 1) Crop-tops are cute on seventh-grade schoolgirls. 2) Seventh grade girls who wear track suits to school are dorks. One way to unpack the truth-value reading of the disagreement is to suggest that cute and dork are evaluative predicates and those who believe (1) and (2) are wrong about the objective 3 Haslanger

4 (social/aesthetic/sartorial) value of crop-tops and track suits. But this is implausible. The patterns of social interaction at the school are what determine the extensions of cute and dork : if a girl walks like a dork, sounds like a dork, dresses like a dork, she is a dork. Where objectivist readings of statements such as (1) seem misguided, the alternative is often taken to be a subjectivist reading which renders the disagreement a matter of taste. On this reading the parents and daughter simply have different sartorial tastes, just as they might have different tastes in food or humor. In effect, the daughter is claiming that crop-tops are cute to her (or to her classmates), and the parents are claiming that they are not cute to them (or to their peers). But this fails to capture the sense in which the parents are disagreeing with the daughter and in a position to offer a critique of the fashion trends. On a broader scale, although social norms and such are at least partly constituted by the attitudes of the social group they govern, an acceptable approach must make room for meaningful critique across groups. Yet another reading of the disagreement would be to see the parent as rejecting, and urging the daughter to reject, the cute/dork dichotomy: these ways of classifying yourself and others based on a willingness to wear sexy clothing are misguided and should be avoided. Parents undertake to disrupt such classifications, as do teachers and school administrators who institute dress codes and such. Let s call this the framework reading. 5 On the framework reading (1) is true and one may be justified in believing it. But at the same time it captures and reinforces (and uttered by the right person at the right moment, might even create) a misguided distinction. Without taking a stand yet on precisely what s at issue between parents and daughter, there is, nonetheless, the makings of a puzzle. If the social reality is organized around the cute/dork dichotomy, then there are cute girls and dorky girls, and it would be a mistake not to recognize this. This is important social knowledge. So when Daughter challenges Parents, she is saying something true and (we may presume) well-justified. But at the same time it is tempting to say that the cute/dork dichotomy is an illusion. It is socially and morally problematic and because it is reified through a pattern of belief and expectation, it could be undermined by a refusing to have beliefs in its terms. More generally, the problem 4 Haslanger

5 seems to be this 6 : Daughter has social knowledge that p, and yet p is somehow illusory. What exactly is the illusion? And how can this illusion be disrupted? III. Social Reality The example of the seventh grade girl and her parents as a small instance of what s involved in navigating and negotiating the social world. The girl and her parents are members of different social groups (agewise), have different experiences, beliefs, and frameworks for understanding what actions and events mean. Both seem to have important social knowledge, but they are also deeply at odds. In the background, I believe, are important issues concerning ideology and social structure. So in the next several sub-sections I will explore some aspects of the interdependence of thought and reality in the social world so we can better understand how thought can fail us without being false. My goal is not to define the social or to give a full-blown theory of social structure, but to illuminate the example we ve been considering. I will then return to the puzzle about crop-tops and the like. a. Ideology and Hegemony In order to develop an account of social knowledge, it will be useful to think about the relationship between agents, their ideas, and social structures generally: what are social structures, and how do agents create, maintain, and change them? Let s begin with the concept of ideology. There is much disagreement over the nature of ideology, yet in the most basic sense ideologies are representations of social life that serve in some way to undergird social practices. 9 We are not simply cogs in structures and practices of subordination, we enact them. And something about how we represent the world is both a constitutive part of that enactment and keeps it going. 10 ideology and discourse refer to pretty much the same aspect of social life the idea that human individuals participate in forms of understanding, comprehension or consciousness of the relations and activities in which they are involved...this consciousness is borne through language and other systems of signs, it is transmitted between people and institutions and, perhaps most important of all, it makes a difference; that is, the way in which people comprehend and make sense of the social world has consequences for the direction and character of their action and inaction. Both discourse and ideology refer to these aspects of social life. (Purvis and Hunt 1993, 474; see also McCarthy 1990, 440) 5 Haslanger

6 Ideology in this broad sense sometimes referred to as the descriptive sense is pervasive and unavoidable. The term ideology is also sometimes used in a narrower and pejorative sense, however, to refer to representations of the relevant sort that are somehow misguided, e.g., by being contrary to the real interests of an agent or group of agents. 11 For current purposes, we can think of ideology as an element in a social system that contributes to its survival and yet that is susceptible to change through some form of cognitive critique. The belief that seventh grade girls who wear crop-tops are cute is a good candidate for a piece of ideology. It is a constitutive part of the fashion norms of seventh grade girls in the school: the belief that girls are wearing such outfits functions to set up a pattern of understandings and expectations that reinforces the pattern of behavior. Moreover, it is plausibly ideology in the pejorative sense because the behavior it sustains subordinates girls. For example, empirical research shows that under conditions of stereotype threat, e.g., in contexts where there is a background assumption that girls are worse at math than boys, anything that primes for gender identity and highly gender coded clothing has been found to be one such thing causes girls to do worse on math tests (Frederickson et al 1998; Spencer et al 1999; Cadinu et al 2005). Yet we might hope that such beliefs are susceptible to cognitive critique, perhaps even parental challenges of the sort we ve considered. We should be attentive, however, to the possibility that an ideology is not just a set of beliefs, and ideology critique is not just a matter of showing that the beliefs in question are false or unwarranted. The framework reading of the disagreement over crop-tops suggested, for example, that the dichotomy of cute/dorky itself was ideological; and the responses that have been conditioned to experience exposed midriffs as cute may be something less than full belief. Further considerations suggesting that ideology is not simply a matter of belief include: In some cases, belief seems too cognitive, or too intellectual. Ideology is concerned with the realm of the lived, or the experienced, rather than of thinking.it is precisely the spontaneous quality of common sense, its transparency, its naturalness, its refusal to examine the premises on which it is grounded, its resistance to correction, its quality of being instantly recognizable which makes common sense, at one and the 6 Haslanger

7 same time, lived, spontaneous, and unconscious. We live in common sense we do not think it. (Purvis and Hunt 1993, 479) Ideology can take the form of practical knowledge, knowledge how to do certain things. Habitual gestures and body language that are ubiquitous in human interaction are ideological. Ideologies seem to work at the level of slogans that can be interpreted differently over time and by different constituencies, e.g., American is the land of the free and home of the brave. (Fields 1982, 155-9) Beliefs have a determinate content that is not compatible with this. Beliefs may be too individualistic. Social practices are ideological, but many people who live in a culture and follow its practices don t have the beliefs that are ordinarily identified as the ideology undergirding the practices. The less belief-like form of ideology is sometimes referred to as hegemony : Ideology and hegemony are opposite ends of a continuum At one end ideology is used to refer to struggles to establish dominant meanings and to make justice claims on the basis of alternative ideologies At the other end the term hegemony is used to refer to situations where meanings are so embedded that representational and institutionalized power is invisible. (Silbey 1998, 276) Although Silbey s quote is a bit unclear on this point, the idea is that ideologies can be more or less contested, more or less hegemonic. The more hegemonic, the less conscious and less articulate they are. Ideologies Mere ideology Contested Hegemonic more conscious more articulate content more determinate Hegemony In the next section I ll develop the idea that hegemonic ideologies are less like (even tacit) theories, and are better understood as schemas. b. Social structure Ideology plays a role in constituting and reinforcing social structures. But what is a social structure? There is considerable interdisciplinary work on this topic by social historians, social 7 Haslanger

8 psychologists, and sociologists interested in subordination and critical resistance. As I am using the term here, social structure is a general category of social phenomena, including, e.g., social institutions, social practices and conventions, social roles, social hierarchies, social locations or geographies and the like. Some social structures will be formal and so the schematic element will be precise and explicit (the structure of faculty governance at any university); some will involve intricate but not fully explicit coordination (informal traffic norms); others will be informal and vague and not well coordinated (the structure of holiday gift-giving). 12 In other words, the schematic element may be more or less precise, and more or less hegemonic. William Sewell (a social historian), drawing on Anthony Giddens, argues for an account that takes structures to be both the medium and the outcome of the practices which constitute social systems. (Sewell 1992, 4, quoting Giddens 1981, 27; see also Giddens 1979). Sewell continues: Structures shape people s practices, but it is also people s practices that constitute (and reproduce) structures. In this view of things, human agency and structure, far from being opposed, in fact presuppose each other. (Sewell 1992, 4). More specifically, Giddens is known for identifying structures as rules and resources. On Sewell s account, however, the combination becomes schemas and resources in order to avoid the assumption that the cognitive element must always take the form of a rule (Sewell 1992, 8). Sewell takes schemas to include: all the variety of cultural schemas that anthropologists have uncovered in their research: not only the array of binary oppositions that make up a given society s fundamental tools of thought, but also the various conventions, recipes, scenarios, principles of action, and habits of speech and gesture built up with these fundamental tools. (Sewell 1992, 7-8). It is crucial to Sewell that these schemas are not private and personal patterns of thought, but are intersubjective and transposable in response to new circumstances. Responding to Sewell, Judith Howard (a social psychologist) points out that Sewell s (1992) use of the term schema differs from its use in social psychology. Whereas social psychologists tend to think 8 Haslanger

9 of schemas as concerned with the organization of an individual s thought, Sewell develops the notion in a way that highlights its cultural deployment. She suggests: A synthesis of these conceptions of schemas might prove remarkably useful: the stricter social cognitive models provide a sound basis for predicting how and when intra-individual schemas change, whereas the more recent sociological conceptions say more about how group interactions shape the formation and evolution of cultural schemas. (Howard 1994, 218) If we take Howard s idea seriously, we should explore the interdependence between individual schemas and their cultural counterparts. Schemas, for example, are both mental and social; they both derive from and constitute cultural, semiotic, and symbolic systems. (Howard 1994, 218). What are we to make of this? Let s take schemas to be intersubjective patterns of perception, thought and behavior. They are embodied in individuals as a shared cluster of open-ended dispositions to see things a certain way or to respond habitually in particular circumstances. Schemas encode knowledge and also provide scripts for interaction with each other and our environment. They also exist at different depths. Deep schemas are pervasive and relatively unconscious. Surface schemas are more narrow and are easier to identify and change; but their change may leave the deeper schema intact. For example, rules concerning gender differences in clothing have changed, yet the more formal the event, the more strict the gender codes. Does this suggest that in contexts where power, authority, and prestige are managed, the deep schema of women as submissive or hobbled property of men still functions? 13 On this view, schemas are one component of social structures, resources are the other. Social structures cannot be identified simply as schemas because social structures have material existence and a reality that pushes back when we come to it with the wrong or an incomplete schema. For example, the schema of two sex categories is manifested in the design and labeling of toilet facilities. If we re analyzing social structures, then in addition to the mental content or disposition, there must be an actualization of it in the world, e.g., an enactment of it, that involves something material. Resources provide the materiality of social structures. On the Giddens/Sewell account, resources are anything that can be used to enhance or maintain power. (Sewell 1992, 9) This includes human resources such as 9 Haslanger

10 physical strength, dexterity, knowledge, (Sewell 1992, 9) in addition to materials animate and inanimate in the usual sense. How do schemas and resources together constitute social structures? Sewell suggests a causal interdependence. (Sewell 1992, 13) He elaborates: A factory is not an inert pile of bricks, wood, and metal. It incorporates or actualizes schemas.the factory gate, the punching-in station, the design of the assembly line: all of these features of the factory teach and validate the rules of the capitalist labor contract In short, if resources are instantiations or embodiments of schemas, they therefore inculcate and justify the schemas as well Sets of schemas and resources may properly be said to constitute structures only when they mutually imply and sustain each other over time. (Sewell 1992, 13) So on Sewell s view a social structure exists when there is a causal, and mutually sustaining, interdependence between a shared or collective schema and an organization of resources. Sewell s claim that the two elements of structure imply and sustain each other suggests a constitutive relationship as well: the pile of bricks, wood, and metal is a punching-in station because schemas that direct employers to pay employees by the hour and employees to keep track of their hours are enacted with this tool. The schema for keeping track of hours is a punching-in schema because there is a punch-clock that the employer will use as a basis for calculating wages. Without the invention of the punch-clock, there could be no punching-in schema. There is a causal relationship, but not just a causal relationship. What else is it? 10 Haslanger

11 Consider a familiar example: a statue and the bronze of which it is composed. The bronze constitutes the statue, e.g., the figure of Joan of Arc on horseback in New York City s Riverside Park. The bronze is the statue not only by virtue of its shape, but also by virtue of having a certain history, function, interpretation, etc. Think of the bronze as resource; think of the dispositions that give rise to the statue s history, function, interpretation (roughly) as schema. The role of schema may be still more evident in the constitution of it as a memorial. The Joan of Arc statue commemorates the 500 th anniversary of Joan of Arc s birth. 14 The statue consists of the shaped bronze, and the statue in turn constitutes the memorial, understood as a further schema-structured resource [[[bronze, shape], statue], memorial]. Thus it appears that the schema/resource distinction can be applied in ways analogous to the matter/form distinction. Consider an example of a social event rather than a social object: the performance of a Bach minuet on the piano. The performance is an event that involves both the piano, the sheet music, fingers and such (as resources), and also a set of dispositions to respond to the sheet music by playing the piano keys in a certain way, plus the various ritualized gestures that make it a performance rather than a rehearsal (as schema). Considered in this light, most actions involve not only an agent with an intention and a bodily movement, but a set of dispositions to interact with things to realize the intention; think of cycling, cooking, typing. These dispositions conform to publicly accessible and socially meaningful patterns and are molded by both the social and physical context. Because often such dispositions give rise 11 Haslanger

12 to objects that trigger those very dispositions, they can be extremely resistant to change (think of the challenge of replacing the qwerty keyboard). This sort of schematic materiality of our social worlds is ubiquitous: towns, city halls, churches, universities, philosophy departments, gyms, playgrounds, homes, are schematically structured and practice-imbued material things (cf. a ghost town or a house but not a home whose schemas are lost or attenuated). The social world includes artifacts which are what they are because of what is to be done with them; it also includes schemas for action that are what they are because they direct our interaction with some part of the world. Thus at least some parts of the social/cognitive world and material world are co-constitutive. If a practice is the structured product of schema (a set of dispositions to perceive and respond in certain ways) and resources (a set of tools and material goods), it is not subjective in any of the ordinary uses of that term. Social structures are not just in our heads (just as the statue is not just in our heads); social structures are public (just as the bronze only constitutes a memorial by virtue of the collective interpretation and pattern of action in response to it); although social structures are not simply material things, they are constituted by material things. They are constructed by us in the ordinary way that artifacts are created by us. One can believe in them without accepting the idea, sometimes endorsed by social constructionists that our thought constructs, in a less ordinary way, what there is in the world (Haslanger 2003). This rough account of social structures helps to define idea of a social milieu. As we saw above, the schemas that constitute social structures are intersubjective or cultural patterns, scripts and the like, that are internalized by individuals to form the basis of our responses to socially meaningful objects, actions, and events. In many cases, perhaps even most, the dominant cultural schema will also be the one that individuals in that context have made their own. However, it is not always that simple. Individuals bear complex relations to the dominant schemas of their cultural context; they may be ignorant of or insensitive to a schema, may reject a schema, or may modify a schema for their own purposes. One may be deliberately out of sync with one s milieu, or just out of it. It is also the case that different schemas 12 Haslanger

13 vie for dominance in public space. For example, what happens when a group of people approach a closed door they want to go through? Some will employ a gallant gentleman schema and will hold the door for the ladies; others will employ a whoever gets there first holds the door schema; still others will employ a first-come, first enters, hold your own door schema. Which schema one brings to the doorway may be a matter of socialization and/or choice. For the purposes of this paper it will be useful to define an individual s (general) social milieu in terms of the social structures within which he or she operates, whether or not the public schemas in question have been internalized. Although we can choose some of the structures within which we live, it is not always a matter of choice, e.g., I am governed by the laws of the United States whether I choose to be or not. Of course, individuals do not live within only one milieu; and milieus overlap. One s workplace, place of worship, civic space, and home are structured spaces; each of these structures are inflected by race, gender, class, nationality, age, and sexuality to name a few relevant factors. So it will be important to specify an individual s milieu at a time and place and possibly in relation to specified others. In this essay I will not be able to give precise conditions that specify what milieu is operative for an individual in a given context; we ll just have to rely on clear-enough cases for now. IV. Looping and Social Kinds In his discussion of social phenomena, Ian Hacking has emphasized the phenomenon of looping. 15 On his view, the continuum between the natural and the social depends on a distinction between indifferent and interactive kinds (Hacking 1999, 32, ). Hacking describes the contrast this way: The woman refugee [as a kind of classification] can be called an interactive kind because it interacts with things of that kind, namely people, including individual women refugees, who can become aware of how they are classified and modify their behavior accordingly. (Hacking 1999, 32) The classification quark, in contrast, is an indifferent kind: Quarks are not aware that they are quarks and are not altered simply by being classified as quarks (Hacking 1999, 32). As Hacking elaborates the idea of an interactive kind it becomes clear that the interaction he has in mind happens through the awareness of the thing classified in being so-classified, though it is typically mediated by the larger 13 Haslanger

14 matrix of institutions and practices surrounding this classification (Hacking 1999, 103; also 31-2, ). For example, if a particular woman is not classified as a woman refugee,... she may be deported, or go into hiding, or marry to gain citizenship...she needs to become a woman refugee in order to stay in Canada; she learns what characteristics to establish, knows how to live her life. By living that life, she evolves, becomes a certain kind of person [a woman refugee]. And so it may make sense to say that the very individuals and their experiences are constructed within the matrix surrounding the classification women refugees. (Hacking 1999, 11) Hacking concludes that the individuals so-affected are themselves socially constructed as a certain kind of person (Hacking 1999, 11). The contrast between indifferent and interactive kinds is not a simple binary distinction, for there are several different factors that may play a role determining whether a kind is more or less indifferent or interactive. One factor is the degree to which we can have, and have had, a causal impact on members of the kind; in cases where we have had a causal impact, a further issue is whether the similarity amongst the members which forms the basis for the kind is due to our influence. For example, quarks is a highly indifferent kind because our (direct) impact on quarks is highly limited, and what impact we have had is not responsible for the similarity amongst them that qualify them as quarks. A given thing is a quark as opposed, say, to a lepton, because of its capacity for certain kinds of interaction with other particles. That quarks have or lack this capacity, is not something that is due to us. (Of course, it is compatible with quarks being indifferent that the idea of a quark is a social product of scientific research.) Hacking is especially interested in a certain kind of object construction, viz., construction works by the social context providing concepts that frame the self-understanding and intentions of the constructed agent. In cases like this, agents incorporate (often consciously) socially available classifications into their intentional agency and sense of self; but as their self-understanding evolves, the meaning of those classifications evolves with them. This forms a feedback loop (hence the term: interactive kinds ) between what we might think of as objective and subjective stances with respect to the classification. Hacking s paradigm examples concern the labeling of various mental illnesses: multiple personality disorder, autism, post-traumatic stress disorder. Individuals are diagnosed with such illnesses; 14 Haslanger

15 treatment plans are developed; their self-understanding is modified. In some cases, groups of those diagnosed develop support groups, communities, political movements. As their self-understanding and behavior changes, however, the diagnosis and patient profile must evolve to take this into account. To emphasize the importance of the agent s active awareness in this process, we might call this discursive identity construction. This is a construction of kinds of people because (at least according to Hacking) people fall into certain kinds depending on their identities, where identity is understood as a psychological notion intended to capture one s self-understanding and the intentional framework employed in action. Through being categorized as a refugee, Nadia comes to think of herself as a refugee, to act as a refugee, to live as a refugee, i.e., she becomes a certain kind of person. Hacking would have us say that she has been constructed as a refugee. We might unpack this as: her self-understanding as a refugee (and pattern of her actions conforming to this understanding) is the result of having been classified as a refugee. No doubt this is an important claim: that certain identities and ways of life come into existence and evolve in response to social and theoretical categorization (especially categorization that emerges in psychology, social work, law, public policy), has important social and political implications. (Hacking 1995, esp. Ch ). But at the same time it isn t entirely surprising that how people think about themselves is influenced by what vocabularies they are given. It is important to note, however, that relationship between schemas and resources in the constitution of social structures is, in general, loopy. Resources are formed to trigger dispositions (schemas) that are manifested in ways that, in turn, utilize and shape the resources. Cuisine is a good example. In a less-globalized world than ours, food crops were grown to support the local cuisine and the local tastes and culinary techniques evolved in ways that take advantage of the crops. In more complex and broadly social changes we can watch consumer taste develop so that certain products become must haves in a particular milieu. Trends in cuisine can become trends in production which, in turn, affect trends in labor, and this affects schemas of class and taste, etc. This loopiness can obscure the social dimension of social structures. When ideology is uncontested and hegemonic, it is insufficiently conscious to be aware of its own effects. So the causal 15 Haslanger

16 impact of hegemonic schemas on resources is typically invisible. Because the trigger for a schema is external in the world we attend to this, and social structures come to seem inevitable, natural, given. Although all ongoing social organizations incorporate contest and struggle over the constitution of their world, most aspects of social structure are taken for granted. Social actors accept a good part of their social worlds as necessary, and often as natural, as perhaps they must do to function at all in those worlds. Often invisible, and certainly uncontested, these taken-for-granted structures are thus unlikely to be the subject of justice claims and critiques, although they may be a source of disadvantage and injustice hegemony colonizes consciousness (Silbey 1998, 289) The reliance on, say, wheat in a particular cuisine may seem inevitable, natural, given. Wheat is what is available; wheat just is what we eat. But the wheat is available because of the impact of schemas on resources that establish farming practices, food distribution, etc. Given the stability of such structures, culinary taste conforms. In this context quinoa, or soy, or spelt tastes bad and has a funny texture too; so who would want to plant it? Hegemony colonizes consciousness. V. Critique The project of at least many social constructionists is to make explicit how the world we respond to, the world that triggers our schemas, is shaped by us and is not inevitable, natural, or given. In other words, the project is to make evident the role of schemas in shaping resources that fit our schemas. Once the loop is laid bare, new questions can be asked about the adequacy of the schemas, the distribution of resources, and alternative structures that might be put in place. The goal is to make explicit the hegemonies that hold us in their grip so that they can be challenged and contested. However, to argue that something is socially constructed is not necessarily to challenge its legitimacy there may be good reasons for the loop to be in place but typically the effort of unmasking is undertaken because there is a suspicion of illegitimacy or injustice. But a separate argument must be offered to establish that there is reason for more than suspicion. What this suggests, importantly, is that ideology critique is not always best thought of as a critique of belief. Hegemonic ideology may not surface as belief because, in the form of schemas, it consists in a set of dispositions to respond that can be a matter of habitual behavior, emotional reaction, 16 Haslanger

17 conceptual framing, perception, etc. To critique hegemony one must look below belief. But also, because social structures are formed by intersubjective schemas, and such schemas (in conjunction with resources) constitute social reality, social knowledge may still be a proper target of critique. This harks back to the example with which we began: In the seventh-grade milieu, crop tops are cute. And yet they aren t really cute, are they? In such cases it is possible that together with the true beliefs there are false beliefs (e.g., that the crop-tops are inherently cute?), but when such allegedly tacit beliefs are made explicit, they are often easily denied without disrupting commitment to the original claim. Although an important step in contesting ideology is to make explicit the content of hegemony, this alone is insufficient. As just noted, schemas are entrenched dispositions and often don t change in response to cognitive critique. A further, often unacknowledged, concern is that components of hegemony are polysemic, so we cannot assume that it is possible to articulate the content of hegemony. (Ewick and Silbey 1995, 212, Silbey 1998, 293). This echoes the earlier point that ideologies sometimes function more like slogans than like beliefs. The hegemonic is not simply a static body of ideas to which members of a culture are obliged to conform [it has] a protean nature in which dominant relations are preserved while their manifestations remain highly flexible. The hegemonic must continually evolve so as to recuperate alternative hegemonies. (Silberstein 1988, 127, quoted in Ewick and Silbey 1995, 212). This protean nature of hegemony can protect it from critique (Ewick and Silbey 1995, 212), but can also make room for resistance and counter-hegemony. Since power is exercised through the patterned distribution of resources and schemas, if there is resistance to this power it must also operate through the appropriation of these selfsame structures. Resistance, as much as power, is contingent upon the structural resources available to the relational participants. Counter-hegemony has to start from that which exists, which involves starting from where people are at. Such a conception of counter-hegemony requires the reworking or refashioning of elements which are constitutive of the prevailing hegemony (Hunt 1990, 316). (Ewick and Silbey 2003, 1335 (including Hunt quote)) In studies of hegemony and counter-hegemony, many humanists, legal theorists, social scientists, have focused on narrative. Narrative is important because of its power to entrench social scripts whose plot is transposable to different contexts; narratives frame the personal in cultural forms. Acts of resistance to 17 Haslanger

18 social scripts can also be narrated using the elements which are constitutive of the prevailing hegemony and become subversive stories. (Ewick and Silbey 2003) Narratives subversive or not are crucial components of the schemas we bring to social life. However, they are not the only component. For example, feminists have long noted that dualistic conceptual frameworks that oppose, e.g., reason/emotion, mind/body, nature/culture, masculine/feminine, guide and distort our thinking. It is also plausible that our thinking is guided by presumption rules that direct our reasoning in cases where evidence is slim (Ullman-Margalit 1983). Such rules are often encoded in narratives, but they are also ubiquitous in conversation and other forms of social interaction. And habits of body and mind including non-intentional behavior, body language, moods, feelings, emotions, suspicions, and the like play an important role in social life, and their interpretation and coordination depends on socializing individuals to fit (roughly) within a pattern of collective dispositions. So it would seem that ideology critique can and should take a variety of forms. For example, we can articulate the hegemonic in ways that open space for contestation and justice claims, e.g., by criticizing conceptual frameworks and offering new ones, by noting and challenging presumption rules that occlude evidence of alternatives, by pointing to the effects of social practices on consciousness. We can give voice to the counter-hegemonic by describing and recommending resistant interventions and practices. We can analyze social conditions and organization in terms that are broadly accessible so that the looping of social structures is rendered visible and so less fixed or inevitable. We can also promote norms and standards for contesting ideology that are more democratic and alert to the muting (and deafening) effects of hegemony. VI. Cuteness, or not. But how, you might reasonably ask, is all this relevant to the debate between Daughter and Parents about crop-tops? Recall the conversation: Daughter: Can I have some money to buy a crop-top like Ashley s to wear to school? Parents: You can have a new top, but not a crop-top. Crop-tops are too revealing. 18 Haslanger

19 Daughter: But Mom[Dad], you re just wrong. Everyone knows that crop-tops are cute; and I don t want to be a dork. Parents: I m sorry, sweetie, crop-tops are not cute, and you won t be a dork if you wear your track suit. The disagreement initially seemed to be about: 1) Crop-tops are cute on seventh-grade schoolgirls. One problem was that (1) seems to be both true (as determined by the standards in the middle school), but also false (as determined by the parents standards). I ve argued elsewhere (Haslanger, forthcoming) that if we employ a semantics that relativizes the truth of such claims to social milieus, then it is possible to explain how Daughter and Parents disagree about the truth of (1) and also both be saying something true (relative to their milieu). The resources of such semantics are useful. But more needs to be said, I believe, to explain how Parents rejection of Daughter s claim is warranted (given that apparently what Daughter says is true and she knows it), and how their position engages her view. Of course there are several ways one might envision the conversation between Daughter and Parents continuing, and different kinds of rationale behind Parents claim. My interest here is how we might employ the ideas about ideology and hegemony sketched above to explicate such critique. We may assume that for Daughter, (1) is articulate, but arises out of a set of unreflective responses to the pressures of her social milieu. Her responses to crop-tops are relatively hegemonic. She experiences crop-tops as cute; she imagines herself wearing one and is delighted at the thought. (She probably hasn t lived through the embarrassment of seeing a picture of herself taken a decade earlier in which she was wearing her favorite outfit which now looks appallingly outdated!) Yet also, it is very likely, she has little to say if asked why crop-tops are cute or what makes them cute. They just are. So plausibly she is ignorant of facts about her social milieu that are relevant to constituting the cuteness of crop-tops. The parents in this case are ignorant of some aspects of Daughter s social milieu, but they are more aware of the social dynamics that constitute cuteness, more aware of the effects of sexualizing twelve-year old girls, and more aware of the mercurial tendency of fashion trends. Their disagreement 19 Haslanger

20 with Daughter is not about denying that the seventh-graders have constituted cuteness in a way that includes crop-tops, but about the broader context in which they do so. They are not in the grip of the crop-top hegemony, and come to the situation with a folk sociology about the constitution of cuteness in the classroom. 16 To see the relevance of their knowledge, it will be useful to reflect for a moment on concepts and conceptual analysis. I ve argued elsewhere that in analyzing a concept, there are three different strategies one might employ: conceptual, descriptive, ameliorative. Very briefly, a conceptual project analyzes a concept, such as cute, by reflecting on the a priori intuitions we have about cuteness. A descriptive project analyzes the concept by looking at the extension of cute in a particular context and theorizing about the class of things picked out by the term (think of naturalizing projects). An ameliorative project asks why we have the concept of cute, what work it is doing for us, and considers how to define the term to best accomplish that work. In the context where crop-tops have been constituted as cute (as in many other cases), the conceptual, descriptive and ameliorative projects seem to yield different results, i.e., the different projects don t converge on what is plausibly the same concept. Parents, with their greater knowledge of social context (their folk sociology ), notice that the clothing designated as cute in the context of this seventh grade is clothing that sexualizes girls. In doing so, they are acknowledging that what cuteness is (in that context) includes crop-tops, but also find this in tension with their a priori judgments about cuteness or with their conception of how cuteness should be understood. For example, they may have the idea that cuteness is properly applied to things that are appealing in a childlike or juvenile sort of way, and sexual suggestiveness is contrary to this. Or they might have a view about how the notion of cute functions in social contexts, what work it does (e.g., valorizing those so-designated based on appearance), and reject the de facto extension as appropriate given the function. They might even hold that no concept with this particular function should be employed. When Parents are rejecting Daughter s claim, they are not simply describing how things are in their milieu, but are drawing on their deeper knowledge of her milieu to challenge her. 20 Haslanger

21 Although in this case the parents are the more knowledgeable about the social context in which cuteness is constituted, this need not be the case. Consider a lesbian teenager who refuses to wear a dress to a cousin s wedding. She argues with her parents: Teen: I m not going to wear that ridiculous dress. I ll wear my black slacks and a jacket instead. Parents: But honey, you ll look so cute in the dress, and all the other girls your age will be wearing dresses. Teen: I will not look cute in the dress, and I don t want to look cute anyway. Here, plausibly, there is a sense in which Teen would look cute in the dress (given the standards of the wedding crowd), but the teenager is the one with greater knowledge of the function of cute in the context and of background hegemony concerning gender, heterosexuality, and childlikeness. So how does this help us resolve the puzzle we started with? This is where we were stuck: a) Daughter knows that crop-tops are cute on seventh-grade schoolgirls. Yet: b) It is an illusion that crop-tops are cute on seventh-grade schoolgirls. In support of (a), the idea was that what counts as cute in a given setting is determined by the collective understandings of those in the setting, and being privy to these understandings provides important social knowledge. In support of (b) the idea is that the extension of cute in the context in question is being determined under conditions of ignorance about the function of cute and the broader social structures governing its use. The conflict emerges because given further knowledge (than Daughter has) about the extension of the term as determined in the context there is a tension with other semantic dimensions of the term, e.g., with the a priori assumptions it carries, or its proper or legitimate function. By pointing out this conflict, the critic defeats the knowledge claim; the speaker must at least step back from a claim of knowledge until there is resolution of the semantic tension. In short, if Parents can provide evidence to daughter that there is this sort of semantic tension, she cannot legitimately claim to know that crop-tops are cute after all. 21 Haslanger

22 It is important to note that resolution of such conflict is open-ended. For example, the term cute, originated as a shortened form of acute, and its first definition in the OED is: Acute, clever, keen-witted, sharp, shrewd. Like cunning, the term shifted from a focus on intelligence to appearance, from shrewdness to childishness: Cute is a good example of how a shortened form of a word can take on a life of its own, developing a sense that dissociates it from the longer word from which it was derived. Cute was originally a shortened form of acute in the sense keenly perceptive or discerning, shrewd. In this sense cute is first recorded in a dictionary published in Probably cute came to be used as a term of approbation for things demonstrating acuteness, and so it went on to develop its own sense of pretty, fetching, first recorded with reference to gals in (The American Heritage Dictionary of the English Language: Fourth Edition ) Along the way, no doubt, there were those who extended the use of the term and others who resisted, there were confusions and contestations. The direction of such semantic shifts is not, or not easily, predictable, i.e., semantic instability can be resolved in favor of either the a priori elements, the descriptive elements, or functional elements, and there are always several ways to go. However, to say that there are several ways to go is not to say that all are equally good. In the case we ve been considering Parents could argue that we should resist the tendency for cute to be applied to sexually suggestive clothing because it reinforces social structures that subordinate girls and women. Revealing the social mechanisms and processes that underlie hegemony opens space for normative considerations. As Silbey urges: Through social analysis, then, the critical scholar makes space for analysis of justice by making manifest the taken-for-granted conditions of social organization and, by implication, the possibilities of alternative social worlds. In this way, analysis hopes to disrupt hegemony s colonizing power. (Silbey 1998, 290) On this view, ideology critique does not, by itself, determine the right course of action, but rather disrupts hegemony by revealing its social basis. In other words, it makes evident the full loop between schemas and resources. In doing so, it may also reveal how power is unjustly distributed in social structures. But it does not itself provide a normative theory or a theory of justice. In an important sense, then, ideology critique is a primarily a form of epistemic critique: it provides evidence of the social basis for certain 22 Haslanger