A Pragmatist THeory of Social Mechanisms--by Neil Gross

A Pragmatist Theory of Social Mechanisms
Neil Gross

American Sociological Review 2009 74: 358 DOI: 10.1177/000312240907400302
The online version of this article can be found at:
Published by:
On behalf of:
American Sociological Association
Additional services and information for American Sociological Review can be found at: Email Alerts: http://asr.sagepub.com/cgi/alerts
Subscriptions: http://asr.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://asr.sagepub.com/content/74/3/358.refs.html
>> Version of Record - Jun 1, 2009 What is This?
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011

A Pragmatist Theory of Social Mechanisms
Neil Gross
University of British Columbia
Some sociologists have recently argued that a major aim of sociological inquiry is to identify the mechanisms by which cause and effect relationships in the social world come about. This article argues that existing accounts of social mechanisms are problematic because they rest on either inadequately developed or questionable understandings of social action. Building on an insight increasingly common among sociological theorists—that action should be conceptualized in terms of social practices—I mobilize ideas from the tradition of classical American pragmatism to develop a more adequate theory of mechanisms. I identify three kinds of analytical problems the theory is especially well poised to address and then lay out an agenda for future research.
In recent decades, sociological positivism—the view that sociology should aim to identify uni- versal causal laws of social life—has been sub- ject to withering critique. Leaving aside the claims of postmodernists, skeptical of every effort at universalization, and humanistic soci- ologists who worry that positivism objectifies human beings, positivism has been persuasive- ly attacked on various philosophical and theo- retical grounds (see Abbott 1988, 1990; Alexander 1982–83, 1987; Seidman 1994; Steinmetz 2005; Zammito 2004). Critics point out its philosophical naiveté with regard to dis- tinctions between facts and values, observation and theory, and proof and persuasion—a prob- lem sociological positivism shares with posi- tivism as a more general philosophy of science.
Direct correspondence to Neil Gross at Department of Sociology, 6303 NW Marine Drive, Vancouver BC V6T 1Z1, Canada (ngross@interchange.ubc.ca). For their very helpful comments on earlier drafts I thank the ASR editors and reviewers as well as Gianpaolo Baiocchi, Amy Binder, Matteo Bortolini, Craig Calhoun, Charles Camic, Scott Frickel, Julian Go, Hans Joas, Erkki Kilpinen, Michèle Lamont, John Levi Martin, Robert Sampson, Mitchell Stevens, Sidney Tarrow, Stephen Turner, Josh Whitford, Christopher Winship, Matt Wray, and Stephen Vaisey. I also thank participants in workshops held at the University of British Columbia, New York University, the Radcliffe Institute for Advanced Study, and the University of Trento.
Scholars also note that in more than a century of sociological research, few universal laws have been discovered.
As criticisms mount, sociologists grasp for more adequate conceptions of the disciplinary enterprise. Moralistic and political understand- ings have attained new popularity (e.g., Burawoy 2005; Feagin and Vera 2008), but many researchers with more strictly explanatory aims have embraced the postpositivist position that sociology should center on identifying more or less general social mechanisms, or abstract causal processes, that may operate in particular settings and that may help to account for observed outcomes. Where positivism has tra- ditionally searched for laws of the form “if X then universally Y ” or “if X then universally Y becomes more likely,” social mechanisms are generally understood as intermediary process- es by which, in certain irreducible contexts, the probabilistic X→Y relationship obtains. The view that sociology should identify mechanisms underlies, for instance, Kandel and Massey’s (2002:983) attempt to discover the means “through which [the] migratory attitudes [of Mexican immigrants] spread through cultural channels to affect behavior”; Fernandez, Castilla, and Moore’s (2000) effort to deter- mine how contemporary firms leverage bene- fits in hiring by reliance on employee referrals; and Sampson, Raudenbush, and Earls’s (1997) argument that the capacity of urban neighbor- hood residents to collectively exert informal
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011 AMERICAN SOCIOLOGICAL REVIEW, 2009, VOL. 74 (June:358–379)
social control is a key mechanism mediating between structure and crime rates. Empirical work in this vein is often distinguished not only from positivism sensu stricto, but also from the sociological tradition of “correlational analysis,” which examines associations among variables but pursues explanation at a high level of gen- erality (see Bunge 1997; Mahoney 2001; Steel 2004). Both approaches, it is argued, treat causal mechanisms as black boxes (Elster 1989; Hedström and Swedberg 1998) and so fail to provide comprehensive explanations.
As more sociologists have adopted a mech- anism-centered focus, theoretical formulations of the mechanisms concept have proliferated (e.g., Hedström and Swedberg 1998; Reskin 2003; Stinchcombe 2005; Tilly 2001). There is, however, something paradoxical about many of these formulations: they owe their attrac- tiveness to a context in which sociological the- orists, applying and extending the ideas of philosophers, have helped to undermine posi- tivism. Yet they often proceed from substantive assumptions that many in the heterogeneous theory community do not consider viable. More specifically, many prominent theoretical accounts of social mechanisms are either beholden to some version of rational choice theory or essentially agnostic about the nature of social action.1
However, a majority of theorists today doubt that action typically takes the form of a ration- al calculation of means to ends, and also insist that action-theoretical assumptions necessarily factor into every account of social order and change and should therefore be fully specified. From a variety of viewpoints, contemporary theorists instead conceptualize social action as a creative enactment over time of social prac- tices. Social practices are ways of doing and thinking that are often tacit, acquire meaning from widely shared presuppositions and under- lying semiotic codes, and are tied to particular locations in the social structure and to the col- lective history of groups. Collective enactment of such practices produces and reproduces those structures and groups (e.g., Archer 2000;
1 I discuss below an exception to this generaliza- tion in Tilly’s work, which overlaps in certain respects with the perspective I develop here.
Bourdieu 1990; de Certeau 1984; Giddens 1984; Ortner 1984; Swidler 2001; see also Chaiklin and Lave 1996; Pickering 1992; Schatzki 1996, 2002; Schatzki, Knorr Cetina, and von Savigny 2001).
In this article, I show how a sophisticated theory of social action, broadly in the practice theory family—developed by the American pragmatist philosophers Charles S. Peirce, William James, George Herbert Mead, and John Dewey and elaborated most recently by Joas (1996)—can be extended into a robust theory of social mechanisms. I do not argue directly for the merits of a pragmatist theory of action; strong arguments to this effect have been advanced by others (e.g., Joas 1993, 1996; Whitford 2002). Nor do I demonstrate that my approach necessarily increases the explanatory power of every account of the operation of par- ticular mechanisms, although I identify three common analytical problems with which the theory could be especially helpful. Rather, I make a prima facie case that a great many social mechanisms, regardless of the level of analysis at which they operate, can be understood as resting on a more solid action-theoretical foun- dation than existing approaches recognize. In doing so, I offer a way to connect important strands of sociological theory with the research enterprise of “mainstream” sociology (see Calhoun and VanAntwerpen 2007) and—taking a different tack from the symbolic interaction- ists—show how the tradition of American prag- matism can provide intellectual coherence to a discipline looking to find its way in a postpos- itivist age.
Confusion abounds as to what exactly a mech- anism is. A clear definition is an essential first step toward a sociological theory of mecha- nisms. To distill such a definition, I consider five varied conceptualizations that have appeared in recent years.2
2 For a more exhaustive review of the literature on mechanisms, see Hedström (2005); Hedström and Swedberg (1998); Johnson (2002); and Mahoney (2001).
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
According to the first conceptualization— advanced by Hedström and Swedberg (1998)— a social mechanism is the structure or process S by which some event or variable I leads to or causes a change in the state of variable or event O. Where some sociologists would be content to “blackbox” S, or significant components of it, Hedström and Swedberg insist that true explanation demands fuller specification of its internal content. Such specification, in their view, should have three features. First, it should follow the principle of methodological indi- vidualism, explaining meso- and macro-level social phenomena by reference to the actions of the individuals involved. Second, it should give primacy to analytical models to be judged by their explanatory utility and parsimony, as much as by their realism.3 Third, the specifi- cation of S must not require that S be directly observable; many social mechanisms, they argue, cannot be observed. Although Hedström and Swedberg point appreciatively to work done by Coleman, Granovetter, and others, their paradigm case of an adequately specified social mechanism is Merton’s (1968) theory of the self-fulfilling prophesy, by which a false definition of a situation leads individuals to act so as to bring that situation about, as when belief in the insolvency of a bank leads to a run that causes insolvency. This theory meets their criteria because it postulates the existence of a “general belief-formation mechanism which states that the number of individuals who per- form a certain act signals to others the likely value or necessity of the act, and this signal will influence other individuals’ choice of action” (p. 21, emphasis in original).
For Reskin (2003), the specification of a social mechanism need not have all the properties demanded by Hedström and Swedberg. Laying
3 As noted below, Hedström (2005) objects to “instrumentalist” versions of rational choice theory that ignore realism altogether.
out an agenda for research on ascriptive inequality, she urges scholars to stop being concerned with models that posit motives for unequal allocations and focus instead on uncovering mechanisms by which “ascribed characteristics” are linked “to outcomes of varying desirability” (p. 7). For Reskin, as for Hedström and Swedberg, mechanisms are what happen inside the black box of social causal- ity—they are “processes that convert inputs (or independent variables) into outputs (or depen- dent variables)” (p. 7). She glosses mecha- nisms-based approaches to inequality as those concerned with the question of how inequali- ties arise in allocation.4 Unlike Hedström and Swedberg, however, Reskin argues that how questions must be answerable in terms of observable processes; in her view, this feature commends them over why questions from the standpoint of realism, for the motives of indi- viduals and groups typically cannot be seen. The only exception concerns mechanisms pos- tulated to operate at the intrapsychic level; interpersonal, societal, and organizational mechanisms must meet the observability requirement.
Stinchcombe (1998:267), building on Coleman, offers an alternative by suggesting that mechanisms are “bits of ‘sometimes true theory’ or ‘models’ that represent a causal process, that have some actual or possible empirical support separate from the larger the- ory in which it is a mechanism, and that gen- erate increased precision, power, or elegance in the large-scale theories.” Although not a methodological individualist, he argues that all social mechanisms involve processes affect- ing lower-order units of analysis—processes that in aggregate bring about the relationship X→Y for higher-order units under considera- tion. Stinchcombe, however, insists that we may be able to show that X causes Y without knowing much about the underlying, lower- order mechanisms: only when such knowledge
4 Much recent empirical work on mechanisms of inequality can be seen as carrying out Reskin’s pro- gram (e.g., Rivera 2008; Stevens 2007).
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
gives us a better understanding of the higher- order relationship—for example, of the con- texts in which the relationship is likely to obtain—will it be helpful to have a grasp of the relevant mechanisms.
“Critical realism” provides a fourth approach to mechanisms. For critical realists like Bhaskar and Collier, the search for mechanisms is the sine qua non of science. In their view, the iden- tif ication of mechanisms involves analytic movement across three ontological domains: from the empirical, where scientists access expe- rience; to the actual, where they identify the events that generate that experience; to the real, wherein lie the causal mechanisms—usually unseen—by virtue of which one event causes another. Key to critical realism’s understanding of this process is the claim that movement from the empirical to the real involves movement along a continuum from an “open” toward a “closed” system. A mechanism is “that aspect of the structure of a thing by virtue of which it has a certain [causal] power” (Collier 1994:62). Mechanisms, however, “operate [only] when suitably triggered” (p. 62), and outside the lab- oratory mechanisms almost always coexist with a host of other mechanisms, processes, and fac- tors that inhibit that triggering or otherwise interfere with the causal relationship. “Under non-experimental conditions,” in other words, “we can see only what [a] mechanism in con- junction with other factors makes it do” (p. 33, emphasis in original)—that is, we can see it operate only in an open system. Experi- mentation, by contrast, creates a closed system “to isolate one mechanism of nature from the effects of others, to see what that mechanism does on its own” (p. 33). Science proceeds by generating such isolation and thus involves nei- ther a search for covering laws nor a simple accumulation of findings. Rather, science searches for an increasingly comprehensive and deep understanding of causal mechanisms, the mechanisms that underlie mechanisms, and how the configuration of particular open systems affects the functioning of mechanisms.
Bhaskar and other critical realists devote par- ticular attention to the social sciences, which they see as studying systems especially resist-
ant to movement in the direction of closure, not least because of what they postulate to be the intrinsic capacity of human beings to work at transforming social relations. This resistance has methodological implications (see Ekström 1992). In the social sciences, explanation can only take the form of breaking events down into their component parts, identifying—by the elaboration of analytic models—the mecha- nisms that could have helped generate them, and determining, through empirically ground- ed reflection on the conditions of historical pos- sibility, whether and how those mechanisms, with others and given contingent circumstances, actually brought about the events (see Steinmetz 2004). Unlike methodological individualists, critical realists are also emergentists who argue that higher-order strata of social reality emerge out of lower-order ones, and that events within those emergent strata are caused by mecha- nisms unique to them and not reducible to lower- order mechanisms.5
A final framework is outlined by Tilly (2001), who, like most students of mechanisms, con- trasts mechanisms-based accounts with those centering on the search for covering laws. He also counterposes them with “propensity accounts” that “consider explanation to consist of reconstructing a given actor’s state at the threshold of action, with that state variously stipulated as motivation, consciousness, need, organization, or momentum” and to “systems explanations” that “consist of specifying the place of some event, structure, or process with- in a larger .|.|. set of interdependent elements” (p. 569). Mechanisms-based approaches, by contrast, “select salient features of [historical] episodes .|.|. and explain them by identifying robust mechanisms of relatively general scope” (p. 569). Tilly has a distinct understanding, how- ever, of what mechanisms consist of. They “are events that alter relations among some specified set of elements,” and they come in three vari- eties: “cognitive mechanisms operate through alterations of individual and collective percep- tion”; “relational mechanisms alter connections
5 For a discussion of social emergence, see Sawyer (2005).
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
362—–AMERICAN SOCIOLOGICAL REVIEW among people, groups, and interpersonal net-
works”; and “environmental mechanisms exert external influences on the conditions affecting [social] processes” (p. 572).
In Tilly’s view, consequently, social expla- nation should involve “pursu[ing] particular mechanisms across different settings” and examining the role of those mechanisms, includ- ing how they “concatenate” into “social process- es,” in bringing about puzzling historical episodes. In describing mechanisms as events, Tilly refers first and foremost to the different kinds of practices actors can enact together, such as pursuing “certification” of their politi- cal identities, as numerous would-be states did vis-à-vis the United Nations after World War II, or “brokerage” involving actors “establishing, severing, or realigning connections among social sites” (p. 575), which Tilly describes as a defining feature of social life in the Soviet Union. Tilly recognizes that mechanisms thus understood, while relatively general in nature, may be instantiated differently in different his- torical periods. For example, he notes that mech- anisms of competition, involving “striving among several actors within a reward-allocat- ing arena” (p. 575), are key features of the con- tentious politics waged by social movement activists, but that politics of this sort, with its unique phenomenology, emerged only in the nineteenth century. Analysts of mechanisms must therefore be attentive to time and place— in particular, to ways in which social mecha- nisms may “incorporate institutions, understandings, and practices that have accu- mulated historically” (p. 570). Tilly’s (1995a:1602) program for social research thus involves “the historically embedded search for deep causes operating in variable combinations, circumstances, and sequences with consequently variable outcomes.”
To extract a working definition of social mech- anisms from these conceptualizations, I consider the major points on which the authors agree and disagree. First the points of explicit and tacit agreement:
1. Social mechanisms are causal in that they mediate between cause and effect. In the sequence X→Y, neither X nor Y nor the causal relationship itself is a social mechanism. The
mechanism is rather the process or means by which X causes Y. This process must have a signif icant social component if the mechanism is to be considered a social one. A volcanic eruption leveling a village and destroying a community is an environmental mechanism (not in Tilly’s sense), not a social one, although it might help establish the conditions under which social mechanisms could unfold.6 It might also be connected with other social mech- anisms that incorporate and mediate environ- mental factors, such as those that help explain the geographic positioning of the village or the nature of its housing stock.
2. Social mechanisms unfold in time. Social mechanisms bring about causal effects through a temporal sequence of events or processes occurring in the social world at the micro-, meso-, or macro-level or across levels. A social fact or phenomenon that causes another social fact or phenomenon instantaneously, with no intervening processes, is unimaginable; such processes make up mechanisms and are always temporally embedded. The duration of the sequences involved may vary greatly. The sequence may be short—a matter of a few inter- actions and cognitive-affective processes—for example, when an individual in a small-group judges another with low external status char- acteristics more positively after that person demonstrates commitment to the group (Ridgeway 1982). The duration of a mecha- nism may extend over years, as for individuals in occupations involving high levels of work- place autonomy who come to value indepen- dence and self-direction (Kohn et al. 1990). Or the mechanism may unfold over centuries, as in the sequence of events by which the Protestant Reformation instilled social discipline in pop- ulations, laying the microfoundations for the rise of strong nation-states (Gorski 2003).
3. Social mechanisms are general, although in varying degrees. If a person grows up in a neighborhood with a high degree of social dis- organization, has no one exerting informal social control over her, and turns to a life of crime (Wilson 1996), a social mechanism can be said to be at play only if the process is more or less
6 For example, mechanisms relating to the result- ing high levels of anomie, as in Erikson’s (1978) classic study of Buffalo Creek, West Virginia.
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
typical of actors in similar circumstances. Every such person need not be subject to the mecha- nism, or affected by it in the same way, but a social mechanism is a causal process with some minimum level of generality. As Tilly’s analy- sis makes clear, however, mechanisms may sometimes be invoked to explain particular events (e.g., historical ones). However much the events typically studied by historical soci- ologists involve dramatic breaks from estab- lished social routines (Sewell 1996), they are explicable in terms of mechanisms to the extent that they are instances of a more general phe- nomenon, such as revolution (Skocpol 1979), or result from combinations of more general mech- anisms (see Steinmetz 2005; Tilly 1995a).
4. Because a social mechanism is an inter- mediary process, it is necessarily composed of elements analyzed at a lower order of com- plexity or aggregation than the phenomenon it helps explain. The nature of this hierarchical relationship will vary by case, but Stinchcombe speaks for most writers on social mechanisms when he argues that identifying them means peering into a layer of social reality that serves as a substratum for the phenomenon under investigation. All work on social mechanisms assumes that mechanisms are the gears in some social machinery and thus stand in a relation- ship of lesser to greater vis-à-vis the causal effect they bring about (see Johnson 2002:230).
If we let theoretical consensus be our guide, these points of agreement should be incorpo- rated into any adequate definition of social mechanisms. But such a definition should also be suff iciently broad to accommodate points of signif icant epistemological and method- ological disagreement:
1. Methodological individualism versus social ontologism. Those like Hedström and Swedberg, who believe that individual persons must be the point of departure for social analysis, take a different approach to mechanisms than do critical realists, who recognize the nonreductive reality of emergent social entities. In fact, Hedström and Swedberg (1998:12) make a case only for a “weak version” of methodological individualism. In many instances, they argue, it may be impossible for explanation to trace all the steps by which the actions of individuals aggregate to compose a supra-individual enti- ty—the demand of methodological individual- ism in its “strong version.” Insofar as this is so,
parsimony not only allows but requires “taking certain macro-level states as given and incor- porating them into the explanation” (p. 13). Generally, however, Hedström and Swedberg believe that the analysis of mechanisms should focus on processes centered on individual-level action. For critical realists, who are committed to a social ontologist position, by contrast, it is acceptable—the point about analytic hierarchy notwithstanding—to study social mechanisms without much concern for the individual-level phenomena by which they come about (e.g., Steinmetz 2005; see also Burris 2007).
2. Formal versus substantive mechanisms. Beyond the requirement that social mechanisms have a minimum level of generality, some schol- ars are concerned with causal relationships that obtain because of the form of the sociological case at hand, in roughly Simmel’s (1971) sense of the term “formal.” What matters here is that an X→Y relationship comes about because of the formal, structural characteristics of the social relations involved, as in Burt’s (2001) argument that social capital advantages accrue to actors whose network ties span the “structural holes” other actors encounter. The content of the situ- ation in which actors accrue or fail to accrue such an advantage, where content is either the actor’s subjective understanding of it or the ana- lyst’s categorization in terms of social domain or manifest or latent function, matters only indi- rectly to Burt’s argument. By contrast, Reskin’s call for the study of mechanisms generative of ascriptive inequality aims to isolate mecha- nisms operative specif ically in situations of allocation. Those who take Reskin’s view that the key mechanisms to study are substantive rather than formal typically focus on domains rich with the relevant mechanisms, whereas advocates of more formal approaches seek to identify mechanisms so abstract that they oper- ate across virtually all domains. The closer to the formal end of the continuum a conceptual- ization of mechanisms is, the less attentive it will be to variation in the working of mechanisms across time and space. Nearly all approaches, however, proceed from the recognition that in social life contingent circumstances cannot be completely explained away.
3. Analytical versus realist models. A final point of contention among those who offer con- ceptualizations of social mechanisms is episte- mological: Is the goal to produce models that
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
allow for elegant and robust predictions, whether or not the postulated mechanisms can be shown to be present and operative in reality? Or should one seek to identify mechanisms that are empir- ically observable? Hernes (1998:78), con- tributing to Hedström and Swedberg’s volume on social mechanisms, takes the former view: “A mechanism is an intellectual construct that is part of a phantom world which may mimic real life with abstract actors that impersonate humans and cast them in conceptual conditions that emulate actual circumstances” (emphasis in original). Reskin takes the latter view—without giving up a concern for robustness—as she would reject postulated mechanisms that are either unobservable or diverge from processes and sequences of events that can be observed.
Taken together, these considerations suggest the following definition: A social mechanism is a more or less general sequence or set of social events or processes analyzed at a lower order of complexity or aggregation by which—in cer- tain circumstances—some cause X tends to bring about some effect Y in the realm of human social relations. This sequence or set may or may not be analytically reducible to the actions of individuals who enact it, may underwrite for- mal or substantive causal processes, and may be observed, unobserved, or in principle unob- servable.
Recent scholarship, although helpful in shed- ding light on the term “social mechanism,” is less satisfactory when it comes to offering soci- ological theories of mechanisms—that is, gen- eral accounts, not of social causality as a philosophical concept, but of causal processes in the realm of the social. How should such processes be understood? What are their build- ing blocks? How do they vary?
With respect to such questions, conceptual work on social mechanisms tends to take one of two forms. Some work seeks to identify rela- tively abstract features of mechanisms but stops short of laying out a fully developed theory of them. Stinchcombe’s and Reskin’s contribu- tions fall into this camp. Although Stinchcombe’s work clarif ies that mechanisms bridge levels of analysis, and offers suggestions about the circumstances in which the specifi-
cation of mechanisms may be helpful, it does lit- tle to delimit the scope of possible mechanisms and provides no general account of their nature. Reskin offers a categorization of the mecha- nisms relevant to the maintenance of inequali- ty, but provides neither a reason to think her typology exhaustive nor much detail as to the workings of the mechanisms said to fall with- in each class. These omissions might stem from skepticism about the explanatory gain from general theories. They might also, however, stem from a hesitation on the part of scholars to both make strong assumptions about social action of the kind that contemporary theorists insist on and to grapple with their implications for the understanding of causal processes.
A different problem besets another strain of work. Perhaps because the idea of opening up the black box of causality to develop fully spec- ified models appeals to sociologists who value a certain kind of analytical rigor, there is often an affinity between work on mechanisms and theorization proceeding from assumptions about action thought to be highly rigorous—namely, scholarship in the rational choice theory tradi- tion. Hedström’s work provides an example: formerly a champion of rational choice theory proper, he now argues that social mechanisms should be understood through the lens of what he calls “DBO theory.” In this theory, social action results when intentional agents have something they desire (D), have a belief (B) about the world pertaining to that desire, and confront opportunities (O) that give them options for action from which they must choose. Where rational choice theory posits “an atom- ized actor equipped with unlimited cognitive abilities that allow ‘him’ to consistently choose the optimal course of action” (Hedström 2005:36), DBO theory assumes only that “the cause of an action is a constellation of desires, beliefs and opportunities in light of which the action appears reasonable” (p. 39). Moreover, while at least some rational choice approaches treat desires and beliefs as exogenous to the explanatory model, DBO theory takes serious- ly the notion that “individuals’ attitudes and beliefs are molded in interactions with others” (p. 43). Such a molding is at the core of Hedström’s conception of social mechanisms. In his view, three types of interactional mech- anisms—belief-, desire-, and opportunity-medi-
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
ated—are the building blocks for more complex social processes.
The problem with this strain of work is that relatively few in the theory community agree that rational choice theory or variants such as DBO theory offer empirically or theoretically adequate descriptions of social action. Several objections are widely shared among theorists (see Archer and Tritter 2000; Green and Shapiro 1994; Somers 1998). Rational choice theory typically conceptualizes rationality as an innate and more or less equally distributed cognitive capacity, whereas sociological theorists attend to ways in which different forms of rationality appear at different historical moments and come to be differentially distributed across social space. Rational choice approaches—especially outside the “bounded rationality” framework— assume that, in most circumstances, individu- als act rationally or at least reasonably in the light of their clear and coherent beliefs and desires. Leaving aside the question of whether most people act rationally or reasonably most of the time, many sociological theorists would fol- low Smelser (1998:4) in holding the “psycho- logical postulate” of ambiguity to have “wide applicability” in social life, and Swidler (2001) in maintaining that the logical coherence of individuals’ beliefs about the world is the excep- tion rather than the rule. Furthermore, the tem- poral phenomenology of much social action departs from that implied by rational choice approaches. While these approaches suggest an individual armed with beliefs and desires who steps out of the flow of action to face and eval- uate a choice between competing means, theo- rists note that such moments are empirically rare, tend to come about in a socially structured fashion, and often involve an inverse temporal ordering in which goals emerge and are clari- fied only after individuals tentatively embark on one means or another.
Finally, whereas rational choice approach- es—like those emphasizing the norm-directed nature of action—assume that most action is motivated, many sociological theorists argue that socially learned habit is a major proximate cause of behavior (Camic 1986). While recog- nizing that lines of habitual activity might accord with individuals’ strategic or expressive interests, theorists view most separate acts com- posing those lines as not directly motivated and see individuals’ retrospective accounts of why
they did something as post hoc rationalizations that, beyond being restricted to a prescientific “vocabulary of motives” (Mills 1940), often obscure the fact that no real motivation or choice was involved.7 Because these criticisms apply as much to DBO theory as to more conven- tional rational choice models, no sociologist who finds them convincing is likely to think Hedström’s theory of social mechanisms—or cognate theories offered by Elster and others— promising.8
In response to these concerns; in reaction to other developments in the human sciences such as existentialism, structural Marxism, and anthropological structuralism; and building on other developments including phenomenology, ethnomethodology, and work on “rule follow- ing” inspired by the later Wittgenstein, theorists in recent decades have argued that social prac- tices—not discrete actions—should be the focus of social research at the level of the individual or group. Practices are generally understood as forms of doing or ways of acting and interact- ing that appear within particular communities or groups; depend on shared presuppositions and assumptions; often have a significant cor- poreal or material dimension; and unfold in individuals’ lives as a result of active, creative, and less than fully conscious puttings into play of those presuppositions and assumptions in the context of various and intersecting sociobi- ographical and interactional exigencies. Conceptualized as such, practices are at the heart of Bourdieu’s (1990) theory of social fields, Butler’s (1990) analysis of the perfor- mativity of gender, Giddens’s (1984) theory of structuration, Knorr-Cetina’s (1999) investiga- tions of the “epistemic cultures” of science and modern society, Ortner’s (1984) efforts to reground anthropological understandings of
7 Not all theorists in the rational choice tradition are subject to these criticisms. Macy’s (1993) “back- ward-looking model of social control” posits that actors learn through experience about the general conditions under which it makes sense to participate in collective action, eliminating the need for infor- mation-intensive calculation in every instance.
8 However, a growing literature in the philosophy of social science argues that “false models” may still be extremely useful in explanation (e.g., Hindriks 2008).
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
culture, and Sewell’s (2005) contributions to historiography, among many other contribu- tions (for review, see Schatzki 1996, 2002; Schatzki et al. 2001). Nearly all specific theo- retical programs advanced under the rubric of practice theory have come in for criticism, as has the notion of social practices itself (Turner 1994), but this has not deterred a significant amount of research into the practices seen as constitutive of social life in numerous domains and historical settings.
Students of social practices have by no means ignored causality. Indeed, as Ortner notes in a seminal 1984 article, the turn toward practice among contemporary theorists, while incorpo- rating notions of the active, knowledgeable, culturally interpretive agent that can be found in earlier humanistic approaches such as sym- bolic interactionism, departs from the antide- terminism that often characterizes such approaches by seeing in patterned iterations of practice the basis for the reproduction of social structures, in particular, structures of inequali- ty. Faced with causal questions such as “Why is there not more intergenerational upward mobil- ity in contemporary capitalist societies?” researchers who take a practice approach do not hesitate to point to practices and their causal effects, as in Lareau’s (2003) claim that differ- ences in childrearing between working- and middle-class parents instill distinctive disposi- tions in their offspring that are differentially rewarded in school and on the labor market. Yet the direct production and reproduction of social structures of inequality by means of the iteration of practices is only one kind of causal effect that may interest social scientists. To the extent that it remains unclear how a variety of other causal processes build on and intersect with social practices—as it does, given that few who take a practice approach address social mechanisms—much empirical research will f ind itself deprived of sophisticated action-the- oretical foundations.
I argue that a solution to this problem can be found by developing a theory of social mecha- nisms on the basis of an approach to social action that has affinities with other strains of practice theory but is less reductive at the level of action than theories like Bourdieu’s. This approach is the one taken by the classical American pragmatists Peirce, James, Dewey,
and Mead and elaborated toward a sociological theory of action by Joas (1996).
The classical American pragmatists were philosophers, not sociological theorists per se. Yet as Joas shows, despite disagreement among them and significant interpretive disputes among contemporary scholars as to the mean- ing of pragmatism, the classical pragmatists were for the most part united in their under- standing of the basic nature of human activity vis-à-vis the social and natural worlds. Rejecting the Cartesian view that thought and action, mind and body, are ontologically distinct, the prag- matists argued that in anthropological terms, humans are problem solvers and the function of thought is to guide action in the service of solv- ing practical problems that arise in the course of life. From this claim, wide ranging and con- troversial epistemological implications followed. More important in the present context, howev- er, is the corollary claim that action, as a response to problem situations, involves an alternation between habit and creativity. The main way humans solve problems, the prag- matists held, is by enacting habits—those learned through social experience or from pre- vious individual efforts at problem solving. By habits, the pragmatists meant not rote behavior, but “acquired predisposition[s] to ways or modes of response” (Dewey 1922:42, empha- sis in original) of which actors are typically not conscious in the moment. Only when preexist- ing habits fail to solve a problem at hand does an action-situation rise to the forefront of con- sciousness as problematic. Then, the pragmatists argued, humankind’s innate capacity for cre- ativity comes into play as actors dream up pos- sible solutions, later integrating some of these into their stocks of habit for use on subsequent occasions.9
9 Space constraints prevent me from offering a more nuanced account of pragmatism or considering the implications for sociology of a pragmatist epis- temology or philosophy of science. The one point I make with regard to the latter is to reject the idea that for pragmatists any action model that “works” to yield a robust explanation will suffice. As Joas shows,
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
Blumer (1969), formulating the program of symbolic interactionism, downplayed this alter- nation between habituality and creativity, but correctly noted that meaning is also central to a pragmatist view of action. Problem situations present themselves to actors through the lens- es of the cultural environments in which they are immersed. Such environments give meaning to and help provide the content of the goals, ori- entations, identities, vocabularies of motive, and other understandings of the action situation that actors come to have. They also provide the basis for intersubjective judgments about the adequacy of problem solutions. All habits are thus enacted on the basis of culturally mediat- ed interpretations of the situation one faces (see Alexander 1988), not least interpretations of the intentions of interaction partners.
Why should sociologists take a pragmatist approach to action seriously? A full treatment of this question goes beyond the scope of this article, but I can outline some reasons why one might prefer pragmatism to both rational choice theory and practice theory approaches such as Bourdieu’s.
Pragmatism is often misunderstood as a form of utilitarianism, but there are at least five ways in which it differs from—and is superior to— rational choice theory (for discussion, see Beckert 2002; Joas 1996; Whitford 2002). First, pragmatism does not equate problem solving with the maximization of utility. To be sure, the situations humans experience as problems may involve utility maximization—for example, the need of businesses to generate revenue. But the kinds of problems of concern to pragmatists range much more widely and include all the difficulties humans or collective actors face in life, from the need to remain healthy to the need to find meaning and purpose in existence. To reduce these to the desire to maximize on a preference function is to ignore the phenome-
the epistemology of the classical pragmatists was premised on their anthropology. My view of the tra- dition draws from many texts, especially Dewey ([1910] 1978, [1920] 1982, 1922), James ([1907] 1975), and Peirce (1992, 1998). Beside classic con- tributions to symbolic interactionism, previous efforts at bringing pragmatist insights into sociology include Lewis and Smith (1980), Maines, Sugrue, and Katovich (1983), Mills (1966), Seidman (1996), and Shalin (1986). For discussion, see Gross (2007).
nological diversity involved in the experiencing of problem situations. Second, to reiterate the point about meaning, pragmatists insist that problem situations are always interpreted through cultural lenses. Even in situations of instrumental rationality, actors are enmeshed in webs of meaning that indicate the significa- tion of the ends they are trying to pursue, con- strain the choices they make by setting limits on the thinkability of means, and sustain the social relationships in which instrumentality must be embedded. Rational choice theory makes little room for culture thus understood. Third, prag- matists argue—directly against most utilitari- ans—that much action is habitual and typically involves no conscious weighing of means and ends. Fourth, pragmatists maintain that instru- mental rationality itself, when it does appear, is a kind of habit, a way that some humans can learn to respond to certain situations, and that we should be as interested in the historical processes by which the habit of rationality—in its various forms—develops and is situational- ly deployed as we should be in its effects. Finally, pragmatists suggest that means and ends are not always given prior to action, as assumed in most rational choice models, but are often emergent from action, as lines of activity are initiated that lead actors to see themselves in new ways, to value different kinds of goods, and to become attached to problem solutions they could not have imagined previously (Whitford 2002).
Thus described, pragmatism, in its under- standing of social action, sounds similar to work in the practice theory tradition. A number of commentators point to commonalities at the level of action theory between pragmatism and the thought of Bourdieu (Aboulaf ia 1999; Dalton 2004; Emirbayer and Goldberg 2005; Shusterman 1999). Bourdieu himself noted that “the affinities and convergences are quite strik- ing” and that his approach, like Dewey’s, “grant[s] a central role to the notion of habit, understood as an active and creative relation to the world, and reject[s] all the conceptual dualisms upon which nearly all post-Cartesian philosophies are based: subject and object, inter- nal and external, material and spiritual, indi- vidual and social, and so on” (Bourdieu and Wacquant 1992:122). If pragmatism and prac- tice theory, at least of the Bourdieusian variety,
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
are so similar, why should sociologists prefer the former?
Some might argue that they should not. Similar though the two approaches may be in certain respects, the claim could be made that there is one crucial difference. Practice theorists like Bourdieu routinely tie their analyses of practices to questions of social-structural pro- duction and reproduction, which have not been a major concern of scholars working in a prag- matist framework. The objection here is not simply that the work has not yet been done to link pragmatist understandings of action with accounts of meso- and macro-level phenome- na, sociology’s typical objects of explanation. As important, the lack of such linkage may lead pragmatists to ignore systematic and conse- quential patterns in the distribution of habitu- ality—by social class position, for example. To my mind, however, this argument counts in favor of pragmatism. Because approaches to prac- tice theory like Bourdieu’s aim primarily at accounting for social reproduction, they end up placing far too much emphasis on the strategic dimensions of action. Although Bourdieu does not see every individual act as motivated, he does view most lines of activity as connected to actors’ interests in leveraging themselves into favorable positions in multidimensional social hierarchies, and thus as tied to the maintenance or transformation of those hierarchies.
As critics of Bourdieu have pointed out (e.g., Alexander 1995), however, this analytical reduc- tion is as problematic in its own way as ration- al choice theory is in its. In Bourdieu’s framework, practices tend not to be seen as sub- scribed to on the basis of relatively autonomous identity commitments, or ultimate values dis- connected from broader social-structural posi- tionings, or by virtue of the sheer force of tradition or institutionalization. Yet evidence from domains as diverse as religion (Smith 2003), politics (Stryker, Owens, and White 2000), intellectual life (Gross 2008), and inti- macy (Gross 2005) suggest that factors of iden- tity, morality, or tradition can certainly underlie the adoption of a social practice by a group, as well as shape individuals’ enactments of it. Such factors must not be seen as residual or epiphe- nomenal elements but as coexisting and in some cases intersecting with strategic concerns over social positioning. In part because pragmatist understandings of action were not designed to
account for social reproduction—but also because the habituality-creativity continuum, for pragmatists, is meant to encompass rather than substitute for other forms of action, while giving pride of place to matters of identity and meaning—pragmatism is better able to accom- modate the diversity of action and practice. Although nothing in a pragmatist approach would deny that some practices are closely bound up with the reproduction of social inequality, the very thinness of the model at the meso- and macro-levels gives it a flexibility and range lacking in other approaches.
The key claim to advance in constructing a the- ory of social mechanisms on these foundations is this: Pragmatists would view social mecha- nisms as composed of chains or aggregations of actors confronting problem situations and mobi- lizing more or less habitual responses. I noted above that alternation between habit and cre- ativity is at the heart of pragmatism, and that pragmatists see this alternation as underlying— not substituting for—other action forms (Joas 1996). These characteristics of the approach, combined with the focus on meaning, yield unique leverage over the notion of mechanisms. To see why, let us follow Hedström and Swedberg at least part way and describe a social mechanism as the structure or process S by which some input I leads to outcome O. A prag- matist theory of mechanisms would hold that to understand S, we must examine the individual and collective actors A1–n involved in the I–O relationship. For each, our goal should be to understand why and how, when confronted with problem situation Pn and endowed with habits of cognition and action Hn, along with other resources, response Rn becomes the most like- ly. S will then consist of all the relations A1–n –P1–n –H1–n –R1–n that, in aggregate or sequen- tially, bring about the I–O relationship.
For example, suppose we are interested in the relationship between race and income inequality and follow Pager (2003) in consid- ering African American men and the effects of a criminal record and “negative credentialing” on the likelihood of gaining employment. Many kinds of actors, problem situations, and habit- ual responses make up this mechanism, but a
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
pragmatist approach might concentrate on understanding how, for employers trying to meet staffing needs with reliable workers, and in the context of prevailing racial-juridical cultural structures, certain habits of thought and action are employed according to which potential employees are coded in terms of trustworthiness depending on their race and history with the jus- tice system, giving rise to discriminatory allo- cation decisions. Aggregated across employers, such an A-P-H-R chain is the mechanism of negative credentialing in this case.
I hypothesize that most social mechanisms can be understood in this way—as chains or aggregations of actors, problem situations, and habitual responses—always with the possibili- ty, greater in some circumstances than others, that a novel way of responding to a problem could emerge for any of the actors involved, potentially altering the workings of the mech- anism. A pragmatist social science concerned with mechanisms would aim to uncover the nature of such chains: the types into which they may be classified, the actors involved in their operation, the habits employed by such actors and their origins, the circumstances in which the mechanisms operate, their interconnection with other mechanisms, and their causal effects.
Note the centrality of meaning in the Pager example; the mechanism is interpretive all the way down. For pragmatists, humans inhabit worlds of meaning. Pragmatism is not a form of methodological individualism; it does not require that mechanisms operating at the meso- or macro-levels be explained exclusively in terms of the actions of the individuals involved, meaning-interpretive or otherwise. It does insist, however, that the potential contribution of indi- vidual action to the operation of mechanisms be taken into account. This requires that we grasp how the relevant individuals understand the sit- uations before them and act on those under- standings, helping thereby to enact the mechanism.
In this respect, pragmatism comes close to the weak version of methodological individualism championed by Hedström and Swedberg. Hedström (2005), in particular, makes belief central to his account of social action, mobiliz- ing Weber’s stress on subjective meaning to argue that actors’ beliefs about the social world are as important as their desires and opportu- nities in explaining their actions, and hence
social mechanisms. How, then, with respect to questions of meaning, does a pragmatist approach to mechanisms differ from Hedström’s approach?
Drawing inspiration in part from Peirce’s work on semiotics (see Deledalle 2000), prag- matists would insist that meaning is not reducible to belief in Hedström’s sense of propo- sitions about the world that actors hold to be true (e.g., that a bank is or is not solvent). In the prag- matist view, consistent with other work follow- ing from the cultural or linguistic turn, such propositions, while important, become mean- ingful only insofar as they string together sym- bolic elements that acquire their individual and relational meanings in larger cultural systems and structures. The belief that a bank is or is not solvent, for example, and that its solvency has implications for whether actors should with- draw their deposits, presupposes that actors understand what a bank is, what it means to say that an institution like a bank can have sol- vency, that actors see themselves and others as oriented toward the maximization of their short- term and individual or familial monetary inter- ests, that they had enough trust in the institutional order to place their money in a bank in the first place, that they are acclimated to a system of monetary exchange, and so on. Merton’s postulated mechanism of the self-ful- filling prophesy functions in this case only because actors are positioned in cultural systems from which they derive these assumptions and orientations, and hence have the beliefs they do. A pragmatist approach, taking meaning seri- ously, would argue that mechanisms cannot be adequately understood without an analysis of such assumptions. This implies that the study of social mechanisms must be undertaken along- side a project of cultural interpretation.
Social mechanisms that affect collective actors (e.g., firms, states, or organizations) can be analyzed in the same way. Collective actors also face problem situations and respond in habit-bound, culturally mediated ways, and social mechanisms involving collective actors consist of chains or aggregations of such responses, whether or not there is explanatory value in further decomposing them into indi- vidual-level action.
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
The pragmatist theory of social mechanisms outlined here can be further developed and elab- orated—and preemptively defended—by responding to four objections it might encounter on first hearing.
The first concerns the theory’s emphasis on culture and interpretation: Will such an approach inevitably neglect the centrality of resources, and struggles over them, in social life? The prag- matist response comes into relief by compari- son with Sewell’s (1992) account of the “duality” of structure. Reformulating aspects of Giddens’s and Bourdieu’s theories, Sewell agrees with Giddens (1984) that “rules” and “resources” must factor into any understanding of social structure. He recasts Giddens’s empha- sis on rules as “generalizable procedures,” how- ever, arguing that the rules that help constitute structures “should be thought of as including all the varieties of cultural schemas that anthro- pologists have uncovered .|.|. : not only the array of binary oppositions that make up a society’s fundamental tools of thought, but also the var- ious conventions, recipes, scenarios, principles of action, and habits of speech and gestures built up with these fundamental tools” (pp. 6–7). Schemas in this sense are habits. Yet Sewell’s emphasis on cultural schemas does not entail a loss of concern for resources. He dis- tinguishes two types: “nonhuman resources are objects .|.|. naturally occurring or manufactured, that can be used to enhance or maintain power” and “human resources are physical strength, dexterity, knowledge” (p. 9). Both types, he argues, “are media of power.” Yet neither is unconnected from cultural schemas: “human resources .|.|. may be thought of as manifesta- tions and consequences of the enactment of cultural schemas,” while “what [nonhuman resources] amount to as resources is largely a consequence of the schemas that inform their use” (p. 11).
Schemas and resources are indeed interre- lated, but a pragmatist understands this rela- tionship somewhat differently from Sewell. According to pragmatists, when actors confront a problem situation they mobilize their habits, including some of the capacities described by Sewell as human resources. Yet this mobiliza- tion typically also involves putting nonhuman resources to work—for example, money. In a
pragmatist understanding, the habits an actor is endowed with will affect the ways in which she understands the significance of and uses the nonhuman resources at her disposal, while the availability of resources—an objective feature of problem situations—may help instill in her distinctive habits. Insofar as social mechanisms are decomposable into problem situations and the habits actors use to resolve them, the avail- ability of resources is, from a pragmatist view- point, a potentially important aspect of every social mechanism.
A second potential objection revolves around the notion of habit. Isn’t this concept too vague and poorly specified to make sense of causal mechanisms operating at multiple levels of analysis? This objection highlights the need to distinguish among kinds of habits. In my view there are three kinds:
1. Individual cognitive-affective habits. These are habitual ways individual actors have of understanding and responding emotionally to situations in general, resulting from their psy- chosocial experience or their biological endow- ments or propensities. Someone who is clinically depressed and sees the world through a glass half-empty is displaying a cognitive- affective habit; so too is a person continually ori- ented toward sexual conquest or the consumption of goods and services. Insofar as the tendency to employ one cognitive schema for interpreting the world rather than another also falls under this analytic rubric, cognitive-affec- tive habits are major sites of cultural-interpre- tive activity, and they refer outward to wider cultural frameworks while also revolving around internal neural processes.
2. Individual behavioral habits. These habits involve the disposition to enact specific behav- ioral responses or routines when individual actors are faced with particular kinds of prob- lem situations. They derive primarily from indi- vidual and social experience. For example, Bittner (1967:702) argues that in many social- disorder situations, “policemen not only refrain from invoking the law formally but also employ alternative sanctions,” ranging from warnings for offenders to “direct disciplining.” This is an individual behavioral habit learned on the job and through exposure to the police subculture.
3. A third category of habits consists of those that are collectively enacted: that is, ways that groups of individual actors, including those
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
who comprise collective actors of various kinds, have of working together to solve problems. I noted earlier that of the main conceptual approaches to mechanisms, Tilly’s is the one that seems unobjectionable in action-theoretical terms. Complementarities between his approach and my own are such that an example from his work can help illustrate what I have in mind by this category of habits.10 In Popular Contention in Great Britain, Tilly (1995b) argues that the best way to explain a major transformation in the British political structure between the mid- eighteenth and early-nineteenth centuries—the growing “capacity” of “ordinary people .|.|. to intervene .|.|. in national affairs” (p. 14)—is to account for the emergence of new “repertoires of contention.” By a repertoire he means “a limited set of routines that are learned, shared, and acted out through a relatively deliberate process of choice” (p. 42). By repertoires of con- tention, he means the routines by which “pairs of actors make and receive claims bearing on each other’s interests” (p. 43). A repertoire of contention can thus be understood as a set of habits or practices enacted collectively by mem- bers of a group to make political claims and attempt to resolve problems they may be facing, from political disenfranchisement to econom- ic marginalization. The idea of repertoires of contention is of considerable importance for the social analysis of politics (Tarrow 1996), but I see it as simply one type of habit that actors can enact jointly to solve problems. “Group style,” as analyzed by Eliasoph and Lichterman (2003:737)—“recurrent patterns of interaction that arise from a group’s shared assumptions about what constitutes good or adequate par- ticipation in the group setting,” such as how group members talk about personal commit- ments or politics—is another type of collective habit. So too are organizational routines and
10 Despite these complementarities, Tilly never laid out the action-theoretical foundations for his conception of practices or made clear how practices might underlie the full range of causal mechanisms of interest to sociologists. More generally, his approach to mechanisms may be too closely tied to his own interests, methodological orientation, and substantive theoretical commitments to serve as a conceptual framework for the discipline as a whole.
repertoires (Clemens 1993; Feldman and Pentland 2003).
4. Alongside this differentiated understand- ing of habit, I would argue that habits often come bundled in habit sets. These are relative- ly coherent repertoires for thinking and acting vis-à-vis a set of problems, as in Tilly’s idea that there may be specific repertoires of contention, such as arose in early-nineteenth-century Britain. To say that habit sets may be relatively coherent does not mean the cultures of practice or lifeworlds they structure and inform are seam- less cultural webs, but only that some habits tend to appear alongside others when they are dis- played by individuals or groups, and that there may be systematic relationships between them at the level of meaning and action. The distinct epistemic cultures that shape knowledge-mak- ing practices in physics and biology (Knorr- Cetina 1999) are good examples of habit sets, as are sharply defined occupational subcultures (Hughes 1971).
Unpacking the notion of habit helps explain how mechanisms operating at various levels can rest on a foundation of habit; we can add that pragmatists would see social mechanisms as varying in abstraction and clustering into an indefinite number of types. Mechanisms under- lying specific cause and effect relationships (e.g., between racial heterogeneity in cities and elite investment in social control resources) are at the low end of abstraction (Jackson and Carroll 1981). At the high end are more gener- al processes that recur across many kinds of situations, such as mechanisms of brokerage. These poles correspond to the distinction drawn above between substantive and formal mecha- nisms. While it may be easier to see how sub- stantive mechanisms conform to a pragmatist model, formal mechanisms can also be under- stood in terms of actors, problem situations, and habitual response. For example, to say that an outcome or event—such as the 1950s Mau Mau revolt in Kenya as analyzed by McAdam, Tarrow, and Tilly (2001)—involved brokerage is to say that a subset of the actors involved faced problem situations that they sought to resolve by mobilizing habits of making social connections between disparate parties. In this understanding, brokerage mechanisms revolve around a particular kind of practice (a point McAdam et al. also make); to understand the nature of the situations those actors faced, the
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
actors’ habits, and their effects is to understand the mechanism. Pragmatists would also agree with scholars such as McAdam and colleagues that there is value in cataloging the wide vari- ety of social mechanisms found in social real- ity, which presumes that mechanisms fall into distinct types and clusters. Yet the pragmatist theory is uniquely poised to understand how this presumption could be valid. According to the theory, what makes mechanisms different is precisely the configuration of actors, problem situations, habits, and patterns of aggregation of which they are composed, so that the project of cataloging them becomes one of cataloging types of A-P-H-R chains. Similar kinds of chains may recur across diverse circumstances; the extent of recurrence determines the abstract- ness of the mechanism.
This discussion raises a third objection: What about formal mechanisms that seem to operate behind actors’ backs and to involve few moments of situational interpretation? Many such postulated mechanisms concern the formal structure of social networks. The pragmatist view overlaps Emirbayer and Goodwin’s (1994:1445–46) critique of network analysis, which, they argue, neglects culture by failing “to thematize more explicitly .|.|. the inherently constructed nature of individual and collective identities .|.|. [and] the complex ways in which actors’ social identities are culturally and nor- matively, as well as societally, determined.” Glossed another way, their point can be gener- alized and squared with the theory presented here by invoking a critique of formalism going back to Durkheim’s ([1900] 1960) attack on Simmel: although the formal structure of social relations can indeed shape and constrain action, the situations in which actors act are always characterized by particularity of content, and such particularity should never be ignored in social explanation. An actor who finds himself possessing “the strength of weak ties” (Granovetter 1973), for example, does so not abstractly but in regard to specific situations such as finding a job. These situations involve configurations of objective elements and mean- ing that make possible and set the parameters for the causal effect of network structure.
To give another example, whom one knows may strongly influence the likelihood of find- inganapartmentinNewYorkCity,withspecial advantages flowing to those with many acquain-
tances rather than a few close friends, but this depends on the degree to which information about the housing market is decentralized and nontransparent, on whether landlords and prop- erty managers favor legalistic arrangements over those requiring trust between parties, whether city living is seen as so desirable that people will do so despite the difficulty and cost involved, and so on. Such factors are usually incorporated into formal models by virtue of an implied ceteris paribus clause, but insofar as they represent the conditions for the mecha- nism operating as postulated, we cannot really understand the mechanism unless we under- stand the conditions. Instead, pragmatism would suggest that mechanisms resulting from the for- mal structure of social relations are best seen as more or less obdurate features of the problem situations individual or collective actors con- front—that is, features that enable or constrain lines of activity. How actors understand and respond to the situations they face will be no less important in the context of such confronta- tions.11
A fourth possible objection follows from the third. Many mechanisms of interest to sociolo- gists, it would seem, are not formal, as I have been using the term, but center on processes of aggregation whose effects equally appear not to depend on actors’ possessing and mobilizing culturally mediated habits. For example, Schelling’s (1971) model of residential segre- gation postulates that if whites and blacks hold even mild preferences for not being outnum- bered in their neighborhoods by people from the other racial group, there will be no equilibrium in housing patterns and neighborhoods will seg- regate and resegregate over time, even if this is not desired by any individual. How can a prag- matist model accommodate such a mecha-
11 Pragmatists would have a similar attitude toward “environmental mechanisms.” These create problem situations to which actors must attempt to respond. The distinction drawn by Tilly (2001) and McAdam and colleagues (2001)—between cognitive/disposi- tional, relational, and environmental mechanisms— is thus a false trichotomy: all social mechanisms run through the nexus of habituality, creativity, and inter- pretation. This is not to deny that mechanisms may be classified as relatively more dispositional, rela- tional, or environmental.
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
nism—which is aggregative and, at the indi- vidual level, seems to involve no more than a simple decision-rule—let alone shine new light on it?
The answer to the first part of the question is that the mechanism can be respecified as an aggregation of individual actors’ efforts at prob- lem solving—the problem being to live in a neighborhood in which one is comfortable, and a key feature of the mechanism being that the kinds of situations actors confront depend on problem-solving activities enacted previously by their neighbors, which may have altered the composition of a neighborhood beyond some demographic tipping point. But what, in this case, would be the value of such a respecifica- tion? Beyond the intrinsic value of greater action-theoretical adequacy, a pragmatist approach would allow preferences for varying levels of racial homophily in one’s neighborhood and the tendency to move if those preferences are violated—the decision-rule in question—to be prof itably reconceptualized as individual behavioral habits. This would permit such pref- erences to remain latent without losing their causal power; relax information and calcula- bility assumptions and replace them with a focus on interactionally and culturally mediat- ed experiences of comfort within socially defined neighborhood boundaries; put greater emphasis on the social and historical condi- tions under which the relevant habits formed for and are enacted by the social groups in question, including those by which they came to see one another in racial or other categorical terms; and allow the possibility that under different con- ditions—for example, in societies placing more emphasis on multicultural tolerance—different habits might be in place, resulting in different aggregate-level effects.
Would such a move increase the explanato- ry power of Schelling’s model? Taking this opportunity to speak to the issue more generally, I hold it to be an empirical question whether the theory of mechanisms laid out here will give sociology more explanatory purchase. Researchers who are persuaded by the theory to reconceptualize the mechanisms they study will either f ind such reconceptualizations helpful in producing more robust explanations or they will not. Because the theory is offered as a flex- ible conceptual toolkit for comprehending social mechanisms, not a set of postulates about social
life from which a broader sociological theory could be deduced, it is impossible to predict the specific forms that explanations informed by the theory will take and equally impossible to demonstrate the theory’s explanatory value a priori. What I can do is identify three kinds of analytical problems for which a pragmatist the- ory of mechanisms seems particularly helpful.
1. The problem of specifying scope condi- tions. Alongside challenges to sociological pos- itivism in recent years have been calls from various quarters for sociologists to clarify the circumstances in which they expect their expla- nations of social phenomena to hold. Postpositivist historical sociologists, for exam- ple, point to social and cultural differences sep- arating societies at different historical junctures and the constraints they impose on the project of social-scientif ic generalization (see Adams, Clemens, and Orloff 2005). Similarly, Abbott (2001), influenced by a vision of social process- es as heavily dependent on place and time—in particular, geographic location and sequential- ity—insists that sociology not overstep certain boundaries with respect to the scope of the explanations it pursues. Inter alia, these lines of thinking demand a better specification of scope conditions, and I argue that a pragmatist theo- ry of mechanisms could be helpful in fleshing out what this means. First, the theory views social mechanisms as decomposable into prac- tices qua habits that are always located in time and space, emerging primarily from social expe- rience. Specifying scope conditions means, in part, accounting for such habits. Second, the the- ory insists that habits, and the mechanisms they compose, function as they do only in conjunc- tion with broader, historically specif ic cultural codes and repertoires. And third, reconceiving mechanisms in terms of A-P-H-R chains draws attention to their intrinsic temporality and hence to time-dependent sequential processes. To the extent that the theory thus points the way toward relatively rich specif ication of the conditions under which particular mechanisms operate, we should expect improvements in the precision of explanatory accounts.
2. The problem of accounting for behavior where cultural meanings vary widely among actors. In a pragmatist model, all social action involves cultural interpretation. But phenome- na obviously vary in the degree to which actors interpret problem situations in similar ways.
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
374—–AMERICAN SOCIOLOGICAL REVIEW For example, in the economic phenomenon
whereby lowering interest rates spurs econom- ic activity, one component of the operative mech- anism is that companies find it cheaper to borrow cash and do so to make capital investments. Differences in economic positioning aside, some variation may occur across sectors and firms in how lowered interest rates are viewed, but not much. By contrast, in the case of the mechanism by which differential resource availability in a household is linked to variation in fertility rates, different actors whose behavior might be covered by a single causal model (e.g., recent immi- grants from Mexico, poor African Americans in the inner city, and upper-middle-class suburban white professionals) may well understand child- bearing and its meaning in their lives in quite dif- ferent terms. The difference between these examples is not that the first is economic and the second familial. Rather, social and historical circumstances (e.g., processes of institutional isomorphism and the secure institutionalization of capitalist social relations) render the relevant interpretive processes more or less homoge- neous in the first case but not in the second. A pragmatist approach to mechanisms may have particular explanatory payoff in situations where interpretive homogeneity across actors is low, for here we should expect cultural differences to have a significant effect on how problem situa- tions are understood and responded to, and on how the social mechanisms thus constituted function.12 The growth of cultural sociology in recent years owes little to a pragmatist theory of action or mechanisms, but the theory I propose implies a substantially broadened disciplinary role for cultural sociology, in part because it suggests that, where meanings vary among actors, cultural interpretation may generate more explanatory specifications of mechanisms.13
12 To return to the Schelling example, this is a cir- cumstance often faced by ethnographers who study race and neighborhood change. Exemplary studies move beyond a simple preferences model to exam- ine the interplay of structural factors and complex, variegated understandings on the part of residents of what makes for a good neighborhood (e.g., Hyra 2008).
13 One might ask how the theory I propose differs from symbolic interactionism, which also claims
3. The problem of accounting for new prac- tices and mechanisms. Change and dynamism are obvious features of every society, and an important aspect of social change is the emer- gence of new practices. How do these come about? In most of the versions of practice the- ory discussed above, the answer centers on processes of social reproduction: new practices emerge as elites struggle to maintain (or chal- lengers contest) patterns of social domination in the face of exogenously generated reconfig- urations of the social field. Nothing in a prag- matist model would deny that social practices, understood as habits, may indeed have their origin in such strategic efforts, but the model would view practices emerging in this way as responses to only one kind of problem situation actors may face—that is, the need for strategic repositioning. In the pragmatist view, there are many kinds of problem situations, and how actors act to resolve the full range of these when existing habits prove inadequate is a major source of new practices, with implications for the social mechanisms with which they intersect.
A pragmatist theory of mechanisms therefore encourages researchers to examine the diversi- ty of problem-solving activities that may lie behind new practices. When accounting for social change is critical and the dynamics of social reproduction do not tell the whole story, this theory should offer greater explanatory leverage than would strains of practice theory. For example, a major change has taken place since the 1960s in how North American and European police forces deal with social protests. In the 1960s, police viewed protests as chal- lenges to social order and routinely contained them by force; today, protests have been nor- malized and police often cooperate with activist groups to ensure that protests do not get out of hand (Della Porta and Reiter 1998; McCarthy
pragmatist roots. Space limits prevent engaging this question fully, but interactionism stresses agency and contingency far more than causality. Focusing on the habit over the creativity end of the action con- tinuum, as I do, allows a pragmatist perspective to be better reconciled with the aspirations of a proba- bilistically causal social science, and to provide action-theoretical foundations for such a science that neglect neither the interpretive nor the agentic aspects of social experience.
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
and McPhail 1997). This change has important implications for contentious politics and the mechanisms surrounding it. It has lowered phys- ical risks for protesters but also rendered protests less symbolically potent. How and why did the new practices come into being? A pragmatist approach would focus on the complex prob- lem-situation to which they were a response: the need of police forces to deal with events hap- pening with increasing frequency, which were costly in terms of manpower and outside dom- inant organizational repertoires of crime-fight- ing, and that posed a high risk of loss of legitimacy. Organizational experimentation and innovation followed, with police forces near college campuses taking the lead, and the newly emergent practices diffused across the organi- zational field. Few could have predicted the specific form the response took: it arose through trial and error, in dialogue and negotiation with politicians, university administrators, and activists that altered police understandings of their goals, in an institutional context that some- times inhibited innovation, and against the back- drop of broader cultural changes in which protests came to be seen as more legitimate. A pragmatist approach would not deny that the new practices strengthened the hand of the state, contributing to processes of social reproduc- tion, but it would look skeptically on the claim that elite demands for social containment ade- quately explain their emergence. A similar explanatory logic could be pursued in many research areas where accounting for the emer- gence of new practices and habits could shed important light on operative social mecha- nisms.14
The growing interest in social mechanisms is salutary. I argue, however, that several leading conceptualizations of social mechanisms are problematic because they are constructed around inadequate understandings of social action. Championing the theory of action developed by
14 There are important similarities—and some dif- ferences—between a pragmatist approach as applied to the study of organizational change and approach- es to the topic informed by “new institutional theo- ry.” See Washington and Ventresca (2004).
the American pragmatists—an approach that has important affinities with more popular strains of practice theory—I propose an alter- native. I argue that social mechanisms—the nuts and bolts processes by which cause and effect relationships in the social world come about—are best thought of as chains or aggre- gations of problem situations and the effects that ensue as a result of the habits actors use to resolve them. This project of theoretical clari- fication aims not just at offering sociologists a better understanding of what they are doing when they identify social mechanisms, but also at reforming sociological practice. The value of the theory is ultimately an empirical question, but I offer reason to think that conceiving of mechanisms in the manner proposed may result in better specified, and very likely more robust, explanatory accounts.
These considerations suggest a clear research agenda: sociology should aim to identify the main social mechanisms by which cause and effect relationships in the social world that are of moral, political, or intellectual importance come about. This entails breaking complex social phenomena into their component parts to see how aggregations or chains of actors employing habits to resolve problem situations bring about systematic effects.
Such a project will necessarily be multi- methodological. Qualitative research—ethno- graphic, interview-based, and historical—is necessary to identify mechanisms, the habits they are composed of, and the kinds of problem situations in which those habits tend to be deployed. Cultural and historical research is needed to understand the origins of habits and habit sets, and hence of mechanisms, along with their meanings for the actors involved and the broader cultural configurations in which those meanings become possible. Quantitative research is required to establish the variable associations that lead us to inquire into cause- effect relationships in the first place, to test across a large number of cases whether one posited mechanism rather than another is pro- ducing the effect, to analyze patterns of aggre- gation, and to establish numerically how habit sets are distributed across groups and individ- uals. Finally, sociological theory is needed to establish new and fruitful conceptual vocabu- laries for thinking about problems and to iden- tify previously unrecognized social processes
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
and dynamics that have ramif ications across empirical domains. Sociology can then mark its progress by whether, for any social outcome of interest to us, we are able to identify reasonably well the often hidden social mechanisms respon- sible for it; gain some insight into how those mechanisms, or related ones, might play out under different circumstances; and, as a result, intervene effectively to bring society into greater conformity with our values and ideals.
Neil Gross is Associate Professor of Sociology at the University of British Columbia and the incoming editor of Sociological Theory. His first book, Richard Rorty: The Making of an American Philosopher, was published last year by the University of Chicago Press.
Abbott, Andrew. 1988. “Transcending General Linear Reality.” Sociological Theory 6:169–86.
———. 1990. “Positivism and Interpretation in Sociology: Lessons for Sociologists from the History of Stress Research.” Sociological Forum 5:435–58.
———. 2001. Time Matters: On Theory and Method. Chicago, IL: University of Chicago Press. Aboulafia, Mitchell. 1999. “A (neo) American in Paris: Bourdieu, Mead, and Pragmatism.” Pp. 153–74 in Bourdieu: A Critical Reader, edited by
R. Shusterman. Oxford, UK: Blackwell.
Adams, Julia, Elisabeth S. Clemens, and Ann Shola Orloff, eds. 2005.
Remaking Modernity: Politics, History, and Sociology. Durham, NC: Duke
University Press.
Alexander, Jeffrey C. 1982–83.
Theoretical Logic
in Sociology. 4 vols. Berkeley, CA: University of
California Press.
———. 1987. “The Centrality of the Classics.” Pp.

11–57 in Social Theory Today, edited by A. Giddens and J. Turner. Stanford, CA: Stanford University Press.
———. 1988. Action and Its Environments: Toward a New Synthesis. New York: Columbia University Press.
———. 1995. Fin de Siècle Social Theory: Relativism, Reduction, and the Problem of Reason. London, UK: Verso.
Archer, Margaret S. 2000. Being Human: The Problem of Agency. Cambridge, UK: Cambridge University Press.
Archer, Margaret S. and Jonathan Q. Tritter, eds. 2000. Rational Choice Theory: Resisting Colonization. London, UK: Routledge.
Beckert, Jens. 2002. Beyond the Market: The Social Foundations of Economic Efficiency. Trans. by B. Harshav. Princeton, NJ: Princeton University Press.
Bittner, Egon. 1967. “The Police on Skid-Row: A Study of Peace Keeping.” American Sociological Review 32:699–715.
Blumer, Herbert. 1969. Symbolic Interactionism: Perspective and Method. Englewood Cliffs, NJ: Prentice Hall.
Bourdieu, Pierre. 1990. The Logic of Practice. Trans. by R. Nice. Stanford, CA: Stanford University Press.
Bourdieu, Pierre and Loïc Wacquant. 1992. An Invitation to Reflexive Sociology. Chicago, IL: University of Chicago Press.
Bunge, Mario. 1997. “Mechanism and Explanation.” Philosophy of the Social Sciences 27:410–65. Burawoy, Michael. 2005. “For Public Sociology.”
American Sociological Review 70:4–28.
Burris, Val. 2007. “Fordism and Positivism in U.S.

Sociology.” Social Science History 31:93–105. Burt, Ronald. 2001. “Structural Holes versus Network Closure as Social Capital.” Pp 31–56 in Social Capital: Theory and Research, edited by N. Lin, K. Cook, and R. Burt. New York: Aldine de
Butler, Judith. 1990.
Gender Trouble: Feminism and
the Subversion of Identity. New York: Routledge. Calhoun, Craig and Jonathan VanAntwerpen. 2007. “Orthodoxy, Heterodoxy, and Hierarchy: ‘Mainstream’ Sociology and its Challengers.” Pp. 367–410 in Sociology in America: A History, edit- ed by C. Calhoun. Chicago, IL: University of
Chicago Press.
Camic, Charles. 1986. “The Matter of Habit.”

American Journal of Sociology 91:1039–87. Chaiklin, Seth and Jean Lave, eds. 1996. Understanding Practice: Perspectives on Activity and Context. Cambridge, UK: Cambridge
University Press.
Clemens, Elisabeth S. 1993. “Organizational

Repertoires and Institutional Change: Women’s Groups and the Transformation of U.S. Politics, 1890–1920.” American Journal of Sociology 98:755–98.
Collier, Andrew. 1994. Critical Realism: An Introduction to Roy Bhaskar’s Philosophy. London, UK: Verso.
Dalton, Benjamin. 2004. “Creativity, Habit, and the Social Products of Creative Habit: Revising Joas, Incorporating Bourdieu.” Sociological Theory 22:603–22.
de Certeau, Michel. 1984. The Practice of Everyday Life. Trans. by S. Rendall. Berkeley, CA: University of California Press.
Deledalle, Gérard. 2000. Charles S. Peirce’s Philosophy of Signs: Essays in Comparative Semiotics. Bloomington, IN: Indiana University Press.
Della Porta, Donatella and Herbert Reiter, eds. 1998.
Policing Protest: The Control of Mass
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
Demonstrations in Western Democracies.
Minneapolis, MN: University of Minnesota Press. Dewey, John. [1910] 1978. How We Think. In The Middle Works of John Dewey, 1899–1924, Vol. 6. Carbondale, IL: Southern Illinois University Press. ———. [1920] 1982. Reconstruction in Philosophy. In The Middle Works of John Dewey, 1899–1924, Vol. 12. Carbondale, IL: Southern Illinois
University Press.
———. 1922.
Human Nature and Conduct: An
Introduction to Social Psychology. New York:
Henry Holt and Company.
Durkheim, Émile. [1900] 1960. “Sociology and its

Scientific Field.” Pp. 354–75 in Émile Durkheim, 1858–1917, edited by K. Wolff. Columbus, OH: Ohio State University Press.
Ekström, Mats. 1992. “Causal Explanation of Social Action: The Contribution of Max Weber and of Critical Realism to a Generative View of Causal Explanation in Social Science.” Acta Sociologica 35:107–22.
Eliasoph, Nina and Paul Lichterman. 2003. “Culture in Interaction.” American Journal of Sociology 108:735–94.
Elster, Jon. 1989. Nuts and Bolts for the Social Sciences. Cambridge, UK: Cambridge University Press.
Emirbayer, Mustafa and Chad Goldberg. 2005. “Pragmatism, Bourdieu, and Collective Emotions in Contentious Politics.” Theory & Society 34:469–518.
Emirbayer, Mustafa and Jeff Goodwin. 1994. “Network Analysis, Culture, and the Problem of Agency.” American Journal of Sociology 99:1411–54.
Erikson, Kai. 1978. Everything in Its Path: Destruction of Community in the Buffalo Creek Flood. New York: Simon & Shuster.
Feagin, Joe R. and Hernán Vera. 2008. Liberation Sociology. Boulder, CO: Paradigm Publishers. Feldman, Martha S. and Brian T. Pentland. 2003.
“Reconceptualizing Organizational Routines as a Source of Flexibility and Change.” Administrative Science Quarterly 48:94–118.
Fernandez, Roberto M., Emilio J. Castilla, and Paul Moore. 2000. “Social Capital at Work: Networks and Employment at a Phone Center.” American Journal of Sociology 105:1288–1356.
Giddens, Anthony. 1984. The Constitution of Society: Outline of the Theory of Structuration. Berkeley, CA: University of California Press.
Gorski, Philip S. 2003. The Disciplinary Revolution: Calvinism and the Rise of the State in Early Modern Europe. Chicago, IL: University of Chicago Press.
Granovetter, Mark. 1973. “The Strength of Weak Ties.” American Journal of Sociology 78:1360–80. Green, Donald P. and Ian Shapiro. 1994. Pathologies
of Rational Choice Theory. New Haven, CT: Yale
University Press.
Gross, Neil. 2005. “The Detraditionalization of

Intimacy Reconsidered.” Sociological Theory
———. 2007. “Pragmatism and Phenomenology in

20th-Century American Sociology.” Pp. 183–224 in Sociology in America: A History, edited by C. Calhoun. Chicago, IL: University of Chicago Press.
———. 2008. Richard Rorty: The Making of an American Philosopher. Chicago, IL: University of Chicago Press.
Hedström, Peter. 2005. Dissecting the Social: On the Principles of Analytical Sociology. Cambridge, UK: Cambridge University Press.
Hedström, Peter and Richard Swedberg. 1998. “Social Mechanisms: An Introductory Essay.” Pp. 1–31 in Social Mechanisms: An Analytical Approach to Social Theory, edited by P. Hedström and R. Swedberg. Cambridge, UK: Cambridge University Press.
Hernes, Gudmund. 1998. “Real Virtuality.” Pp. 74–101 in Social Mechanisms: An Analytical Approach to Social Theory, edited by P. Hedström and R. Swedberg. Cambridge, UK: Cambridge University Press.
Hindriks, Frank. 2008. “False Models as Explanatory Engines.” Philosophy of the Social Sciences 38:334–60.
Hughes, Everett C. 1971. The Sociological Eye: Selected Papers. Chicago, IL: Aldine-Atherton.
Hyra, Derek S. 2008. The New Urban Renewal: The Economic Transformation of Harlem and Bronzeville. Chicago, IL: University of Chicago Press.
Jackson, Pamela Irving and Leo Carroll. 1981. “Race and the War on Crime: The Sociopolitical Determinants of Municipal Police Expenditures in 90 Non-Southern U.S. Cities.” American Sociological Review 46:290–305.
James, William. [1907] 1975. Pragmatism: A New Name for Some Old Ways of Thinking. Cambridge, MA: Harvard University Press.
Joas, Hans. 1993. Pragmatism and Social Theory. Chicago, IL: University of Chicago Press.
———. 1996. The Creativity of Action. Trans. by J. Gaines and P. Keast. Chicago, IL: University of Chicago Press.
Johnson, James. 2002. “How Conceptual Problems Migrate: Rational Choice, Interpretation, and the Hazards of Pluralism.” Annual Review of Political Science 5:223–48.
Kandel, William and Douglas S. Massey. 2002. “The Culture of Mexican Migration: A Theoretical and Empirical Analysis.” Social Forces 80:981–1004.
Knorr-Cetina, Karin. 1999. Epistemic Cultures: How the Sciences Make Knowledge. Cambridge, MA: Harvard University Press.
Kohn, Melvin L., Atsushi Naoi, Carrie Schoenbach,
Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
Carmi Schooler, and Kazimierz M. Slomczynski. 1990. “Position in the Class Structure and Psychological Functioning in the United States, Japan and Poland.” American Journal of Sociology 95:964–1008.
Lareau, Annette. 2003. Unequal Childhoods: Class, Race, and Family Life. Berkeley, CA: University of California Press.
Lewis, J. David and Richard L. Smith. 1980.
American Sociology and Pragmatism: Mead, Chicago Sociology, and Symbolic Interaction. Chicago, IL: University of Chicago Press.
Macy, Michael W. 1993. “Backward-Looking Social Control.” American Sociological Review 58:819–36.
Mahoney, James. 2001. “Review: Beyond Correlational Analysis: Recent Innovations in Theory and Method.” Sociological Forum 16:575–93.
Maines, David R., Noreen M. Sugrue, and Michael A. Katovich. 1983. “The Sociological Import of G. H. Mead’s Theory of the Past.” American Sociological Review 48:161–73.
McAdam, Doug, Sidney Tarrow, and Charles Tilly. 2001. Dynamics of Contention. Cambridge, UK: Cambridge University Press.
McCarthy, John and Clark McPhail. 1997. “The Institutionalization of Protest in the United States.” Pp. 83–110 in The Social Movement Society: Contentious Politics for a New Century, edited by D. S. Meyer and S. Tarrow. Lanham, MD: Rowman & Littlefield.
Merton, Robert K. 1968. Social Theory and Social Structure. New York: Free Press.
Mills, C. Wright. 1940. “Situated Actions and Vocabularies of Motive.” American Sociological Review 5:904–13.
———. 1966. Sociology and Pragmatism: The Higher Learning in America. New York: Oxford University Press.
Ortner, Sherry B. 1984. “Theory in Anthropology Since the Sixties.” Comparative Studies in Society and History 26:126–66.
Pager, Devah. 2003. “The Mark of a Criminal Record.” American Journal of Sociology 108:937–75.
Peirce, Charles S. 1992. The Essential Peirce. Vol. 1, 1867–1893, edited by N. Houser and C. Kloesel. Bloomington, IN: Indiana University Press.
———. 1998. The Essential Peirce. Vol. 2, 1893–1913, edited by N. Houser and C. Kloesel. Bloomington, IN: Indiana University Press.
Pickering, Andrew, ed. 1992. Science as Practice and Culture. Chicago, IL: University of Chicago Press.
Reskin, Barbara F. 2003. “Including Mechanisms in Our Models of Ascriptive Inequality.” American Sociological Review 68:1–21.
Smith, Christian. 2003. Moral, Believing Animals: Downloaded from asr.sagepub.com by Wayne Mellinger on October 20, 2011
Ridgeway, Cecilia L. 1982. “Status in Groups: The
Importance of Motivation.” American Sociological
Review 47:76–88.
Rivera, Lauren. 2008. “Cultural Reproduction in the

Labor Market: Homophily in Job Interviews.” Working paper, Department of Sociology, Harvard University, Cambridge, MA.
Sampson, Robert J., Stephen W. Raudenbush, and Felton Earls. 1997. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy.” Science 277:918–24.
Sawyer, R. Keith. 2005. Social Emergence: Societies as Complex Systems. New York: Cambridge University Press.
Schatzki, Theodore R. 1996. Social Practices: A Wittgensteinian Approach to Human Activity and the Social. Cambridge, UK: Cambridge University Press.
———. 2002. The Site of the Social: A Philosophical Exploration of the Constitution of Social Life and Change. University Park, PA: Pennsylvania State University Press.
Schatzki, Theodore R., Karin Knorr Cetina, and Eike von Savigny, eds. 2001. The Practice Turn in Contemporary Theory. London, UK: Routledge.
Schelling, Thomas C. 1971. “Dynamic Models of Segregation.” Journal of Mathematical Sociology 1:143–86.
Seidman, Steven. 1994. Contested Knowledge: Social Theory Today. Malden, MA: Blackwell.
———. 1996. “Pragmatism & Sociology: A Response to Clough, Denzin, and Richardson.” The Sociological Quarterly 37:753–59.
Sewell, William H., Jr. 1992. “A Theory of Structure: Duality, Agency, and Transformation.” American Journal of Sociology 98:1–29.
———. 1996. “Political Events as Structural Transformations: Inventing Revolution at the Bastille.” Theory & Society 25:841–81.
———. 2005. Logics of History: Social Theory and Social Transformation. Chicago, IL: University of Chicago Press.
Shalin, Dmitri N. 1986. “Pragmatism and Social Interactionism.” American Sociological Review 51:9–29.
Shusterman, Richard. 1999. “Bourdieu and Anglo- American Philosophy.” Pp. 14–28 in Bourdieu: A Critical Reader, edited by R. Shusterman. Oxford, UK: Blackwell Publishers.
Simmel, Georg. 1971. On Individuality and Social Forms: Selected Writings, edited by D. N. Levine. Chicago, IL: University of Chicago Press.
Skocpol, Theda. 1979. States and Social Revolutions: A Comparative Analysis of France, Russia, and China. New York: Cambridge University Press.
Smelser, Neil J. 1998. “The Rational and the Ambivalent in the Social Sciences.” American Sociological Review 63:1–16.
Human Personhood and Culture. New York:
Oxford University Press.
Somers, Margaret. 1998. “We’re No Angels: Realism,

Rational Choice, and Relationality in Social Science.” American Journal of Sociology 104:722–84.
Steel, Daniel. 2004. “Social Mechanisms and Causal Inference.” Philosophy of the Social Sciences 34:55–78.
Steinmetz, George. 2004. “Odious Comparisons: Incommensurability, the Case Study, and ‘Small N’s’ in Sociology.” Sociological Theory 22:371–400.
———, ed. 2005. The Politics of Method in the Human Sciences: Positivism and Its Epistemological Others. Durham, NC: Duke University Press.
Stevens, Mitchell L. 2007. Creating a Class: College Admissions and the Education of Elites. Cambridge, MA: Harvard University Press.
Stinchcombe, Arthur L. 1998. “Monopolistic Competition as a Mechanism: Corporations, Universities, and Nation-States in Competitive Fields.” Pp. 267–305 in Social Mechanisms: An Analytical Approach to Social Theory, edited by P. Hedström and R. Swedberg. Cambridge, UK: Cambridge University Press.
———. 2005. The Logic of Social Research. Chicago, IL: University of Chicago Press.
Stryker, Sheldon, Timothy J. Owens, and Robert W. White, eds. 2000. Self, Identity, and Social Movements. Minneapolis, MN: University of Minnesota Press.
Swidler, Ann. 2001. Talk of Love: How Culture Matters. Chicago, IL: University of Chicago Press.
Tarrow, Sidney. 1996. “The People’s Two Rhythms: Charles Tilly and the Study of Contentious Politics.” Comparative Studies in Society and History 38:586–600.
Tilly, Charles. 1995a. “To Explain Political Processes.” American Journal of Sociology 100:1594–1610.
———. 1995b. Popular Contention in Great Britain, 1758–1834. Cambridge, MA: Harvard University Press.
———. 2001. “The Historical Analysis of Political Processes.” Pp. 567–88 in Handbook of Sociological Theory, edited by J. Turner. New York: Springer.
Turner, Stephen P. 1994. The Social Theory of Practices: Tradition, Tacit Knowledge, and Presuppositions. Chicago, IL: University of Chicago Press.
Washington, Marvin and Marc J. Ventresca. 2004. “How Organizations Change: The Role of Institutional Support Mechanisms in the Incorporation of Higher Education Visibility Strategies, 1874–1995.” Organization Science 15:82–97.
Whitford, Josh. 2002. “Pragmatism and the Untenable Dualism of Means and Ends: Why Rational Choice Theory Does Not Deserve Paradigmatic Privilege.” Theory & Society 31:325–63.
Wilson, William Julius. 1996. When Work Disappears: The World of the New Urban Poor. New York: Knopf.
Zammito, John H. 2004. A Nice Derangement of Epistemes: Post-positivism in the Study of Science from Quine to Latour. Chicago, IL: University of Chicago Press.