Theodore Schatzki's A New Societist Social Ontology


A New Societist Social Ontology
Theodore R. Schatzki



Philosophy of the Social Sciences 2003 33: 174 DOI: 10.1177/0048393103033002002
The online version of this article can be found at:
http://pos.sagepub.com/content/33/2/174
Published by:
http://www.sagepublications.com
Additional services and information for Philosophy of the Social Sciences can be found at: Email Alerts: http://pos.sagepub.com/cgi/alerts
Subscriptions: http://pos.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://pos.sagepub.com/content/33/2/174.refs.html
>> Version of Record - Jun 1, 2003 What is This?
page1image8160
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011

1PS0cHh.1IAa1Lt7zOR7kS/TiO0/I0PC4AH8L3NYE9E3OW1F03TS2OH51CE6IS8EO0TCISITASLOSCIAENL COENST/OJLuOneG2Y003
A New Societist Social Ontology
THEODORE R. SCHATZKI
University of Kentucky
This article delineates a new type of social ontology—site ontology—and defends a particular version of that type. The first section establishes the distinc- tiveness of site ontologies over both individualist ontologies and previous societist ones. The second section then shows how site ontologies elude two per- vasive criticisms, that of incompleteness directed at individualism and that of reification leveled at societism. The third section defends a particular site ontol- ogy, one that depicts the social as a mesh of human practices and material arrangements. The article concludes by outlining what is involved in giving site- ontological analyses of social things.
Keywords: socialontology;sociality;individualism;wholism;socialfacts;Heidegger; social practices
Social ontology concerns the nature of the social. It addresses such questions as, What is sociality? In and of what does the social consist? and What are the basic structures of social life? The broad front of individualism has historically played a dominant role in this domain. Not only has a weighty line of theorists defended individualist accounts of social phenomena, but those defending alternatives have had to define themselves against individualism. This dominance allows a division of social ontologies into two camps: individualism and nonindividualism.
The basic thesis of ontological individualism is relatively straight- forward. It is that social facts and phenomena are nothing but con- structions out of, or constructions of, individual people and—on some versions—their relations. Nonindividualist ontologies contest this thesis. About all they agree on, however, is that something about the social is not susceptible to individualist analysis—that is to say, that the social encompasses something in addition to the stuff of indi- viduals into which individualism seeks to resolve social entities. Con-
Received 23 August 2000
Philosophy of the Social Sciences, Vol. 33 No. 2, June 2003 174-202 DOI: 10.1177/0048393103251680
© 2003 Sage Publications
174
page2image17824
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 175
siderable disagreement reigns among these societist1 ontologies about the identity of this irreducible phenomenon.
Three prominent societist ontologies in the 20th century were wholism,2 Durkheimian sociology, and structuralism. Wholism focuses on such large-scale phenomena as societies and economic sys- tems and argues that these phenomena are wholes. This means that such entities have a nature and are subject to dynamic principles that differ from those of and pertaining to the individuals populating them. On this line of thinking, social wholes are what there is to the social that is impervious to individualist analysis.
The second prominent societist ontology, that of Durkheim, con- tends that social facts are irreducible to “individualist” ones. They are categorically distinct from, and subject to types of explanation not applicable to, the latter. Social facts, consequently, are what of the social that escapes individualism. Indeed, Durkheim’s categorical distinction between social and individualist facts entails that the social in toto (the sum of social facts) eo ipso evades individualism. Social facts, moreover, are not the same as wholes. Being a whole is not a defining feature of them, and the class of social facts is much broader than that of social wholes. Alongside economic organization and financial systems, it embraces, among other things, language, customs, population movements, and crowd phenomena.
A third prominent form of societist ontology, structuralism, main- tains that the possibilities (in Lévi-Straussian versions) or the compo- sition (in, e.g., structural Marxist versions) of such social phenomena as economies and political formations is governed by, or made up of, abstract, non-spatial-temporal structures. These structures form a determining and/or constitutive dimension of social life that cannot be decomposed into the stuff into which individualism seeks to resolve social phenomena. Like wholism and unlike Durkheimian social theory, moreover, structuralism recognizes the stuff of individ- uals as a constituent of the social: social phenomena consist in the activities of individuals governed by abstract structures. Parallel to Durkheim and divergent from wholists, however, many structuralists contend that abstract structures govern and compose not just large-scale social, economic, and political formations but also such clearly nonwhole phenomena as myths, religions, and languages.
This article is concerned both to delineate a new type of societist ontology that has appeared in recent decades and to defend a particu- lar version of that approach. I call this new type “site” ontologies. Site
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
176 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
ontologies contend that social life, by which I mean human coexis- tence, is inherently tied to a type of context in which it occurs. The contexts involved, sites, are contexts of which some of what exists or occurs within them are inherently parts. Site ontologies maintain that social phenomena can only be analyzed by examining the sites where human coexistence transpires. It is in highlighting this type of context that this approach differentiates itself from societist ontologies that emphasize wholes, sui generis facts, or abstract structures.
THE IDEA OF A SITE ONTOLOGY
This first section seeks simply to establish site ontologies as a dis- tinctive brand of societist ontology. I begin by explicating the notion of a site.
As indicated, X’s site is a type of context. It is where X takes place, in a special sense of “where” to be delimited shortly. From the begin- ning, it is important to resist the spatial connotations of the expression “where.” Spatial sites are only one genre of sites. To delimit, more- over, the type of site with which site ontologies work, it is helpful to distinguish three senses of “where” something takes place, that is, three sorts of site.
Something’s site is, first, the location where it is or takes place. All phenomena and events that exist or occur, either amid a broader set of phenomena or in a wider region, have locations in those sets or regions. Something’s spatial location, for instance, is its location in space. Something’s temporal location, moreover, is where it takes place in clock time or in the course of a meeting, the progress of daily events, or the sweep of the century. Human activities have locations in time and objective space. They also possess teleological and activity- place space locations. The teleological location of, say, a token act of requesting a credit report is where that act fits into the hierarchies of ends, purposes, and tasks that imbue the banking practices of which it is a moment. Its location in activity-place space is where it occurs in the activity-place space of the bank office, that is, the particular activ- ity place(s) in the layout of places and paths in that setting where it is performed.
Something’s site is, second, the broader region or set of phenomena in which it exists or takes place. In this sense, space is where spatial phenomena occur, time is where temporal phenomena occur, and teleological hierarchies and activity-place spaces, among other
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 177
things, are where banking activities occur. Notice that anything that occurs amid multiple sets of phenomena, or in multiple regions, has multiple sites in the first two senses of “where.” Requesting a credit report occurs, for example, in space, in various temporal manifolds, in the banking practices of a given country, in the history of banking in that country, and so on.
Something’s site is, third, that realm or set of phenomena (if any) of which it is intrinsically a part. X’s site is thus, third, that set of phenom- ena or realm (if any) as part of which X is or occurs. This type of where shares with the second the sense of wider scene; it also shares with both the first and second the intuition that where something is is the place it can be found. An example of a site in this third sense is abso- lute space, which is the site of spatial locations. Not all sites of the first two types, however, are instances of the third. On neither an absolut- ist nor a relationist account of space, for example, is space a site of the third type where human activities occur. For qua the particular activi- ties they are, human activities do not occur as part of space—they do not constitute this space, which, if they did, would also constitute them. Similarly, neither the activity-place spaces of the material set- tings found at banks nor the history of banking in a given country forms a site of the third sort where such banking activities as request- ing credit reports occur. Rather, banking practices are this third type of site, for requesting reports and other banking activities occur as part of banking practices. Requesting and other activities help make up those practices, which in turn help constitute them. It is sites of this third type that I shall refer to as “sites.” Henceforth, something’s site is that realm or set of phenomena (if any) of which it is intrinsically a part.3
To advocate a site ontology is to claim that the character and trans- formation of social life are inherently tied to the site of the social. Site ontologies represent a distinct approach to social ontology. In individ- ualist hands, the issue of the nature and constitution of the social metamorphoses into two paramount issues: What is social action? and What is a social phenomenon? Social actions are identified as a particular kind of individual action (e.g., ones oriented to others’ behavior, ones composing social phenomena), and social phenomena are analyzed as constructions of, or as constructions out of (including amalgamations of), the actions, mental states, and relations of indi- viduals. I should point out that this formulation is designed to cap- ture both (1) the “constructivist” type of individualism that was prominent in the 1950s (and earlier), according to which social forma-
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
178 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
tions consist in/of features of and relations among individuals (cf. Hayek 1952; Schutz in Grathoff 1978), and (2) the “institutional” type of individualism that has more recently (again) become prominent, according to which social phenomena are instituted, that is, consti- tuted in their being and thereby made to be the case, by people’s pos- session of certain thoughts, beliefs, and attitudes or by their perfor- mance of particular actions (cf. Gilbert 1989; Searle 1995). Although individualisms, even of the same general sort, can diverge signifi- cantly, there is no need for the present article to attend to the differences.
In societist hands, meanwhile, investigations of the nature of the social center on that component of it that resists reduction to individu- alist stuff: whole societies, social facts, or abstract structures. The social is identified either as the irreducible component itself or as an amalgam of individuals and this component (e.g., either as the sum of societies, social facts, or structures or as people, say, acting within the institutions or subsystems prescribed by the society totality or occu- pying slots in social structures and carrying out the activities appor- tioned to these slots). Formations such as governments, economies, and armies are then analyzed as particular institutions, subsystems, sets of social facts, or activity arrays governed by particular structures.
Site ontologies proceed differently. Addressing the nature of the social involves identifying the type of site where social life exists and develops. Since a site, as noted, is a kind of context, the focus is on a special type of context, not wholes, sui generis realities, or abstract structures. Like all accounts of the social, however, site accounts underwrite analyses of social formations such as governments and economies (see the fourth section)—for their accounts of the site of social life specify material out of which these formations are com- posed. Like wholist and structural, and unlike Durkheimian, accounts, finally, site accounts acknowledge individuals as constitu- ents of such formations: an economic system, for example, is com- posed of a nexus of people’s actions taking place in specific contexts.
It might be objected that I have not sufficiently distinguished site ontologies from either individualist or other societist ones, for ontologies of these other sorts recognize contexts and the relevance of contexts to social phenomena. These ontologies do, indeed, recognize these. The contexts they acknowledge, however, are not sites.
Individualism maintains that social phenomena are constructions of or out of the actions and mental states of (and maybe the relations
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 179
among) individuals. Many of a person’s actions, mental states, and relations, however, depend on those of others. Hence, the individual- ist matters that institute social phenomena or in which social phe- nomena consist exist in a context, namely, the encompassing tissue of interrelated individuals. According to individualism, consequently, an individual life is contextualized within other lives, and configura- tions, as well as mental and “actional” features of configurations, of lives are contextualized within further such configurations.
It turns out, however, that for individualists neither an individual life nor a configuration of lives is intrinsically part of any collection or further configuration of lives that forms its context. In order for A to be intrinsically part of C, A’s identity (being) must be inherently tied to C. Individualists do not argue, however, that the identities of the individualist stuff that constitute social phenomena (e.g., actions and mental states) are tied inherently to the actions and mental states of other people. Which actions a person performs, and which mental states she is in, might be related to the properties of other people, but the connection is contingent (and typically causal). When individual- ists do argue that the identities of social phenomena–constituting actions and mental states are inherently tied to further phenomena (and many do not argue this), the phenomena concerned are further properties of the individuals who perform the actions or are in the states (e.g., purposes and intentions; perceptions, other mental states, and actions). As indicated, furthermore, individualists do acknowl- edge that actions form configurations, for instance, the configurations in which such social phenomena as economies and governments con- sist. But the identities of these actions as actions do not derive from those configurations, for instance, from something that characterizes economic or government actions in common. According to individu- alism, an action configuration is always an aggregate, each of whose component actions is only contingently dependent on features of the actors forming its contexts. Hence, the contexts that individualism acknowledges are not sites: individuals are not intrinsically part of them.
Hegelian, Durkheimian, and Saussurian ontologies also acknowl- edge contexts. These contexts are composed of those components or dimensions of social life that are not amenable to individualist analy- sis. These components and dimensions qualify as contexts of individ- ual lives because they determine the actions and mental states of mul- tiple individuals, sometimes causally and sometimes via enablement
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
180 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
and constraint. Although societist contexts differ greatly from the tex- ture(s) of individuals that individualism promotes, they do not amount to sites. The burden of these societisms has always been to defend a difference in being between individualist stuff and some dis- tinct, irreducible component of the social. As a result, mental states and actions and so forth cannot be intrinsically part of these phenom- ena—inseparable from them, perhaps, but not an inherent compo- nent of the whole societies, social facts, or abstract structures that embed and determine them.
Before concluding this section, I might explain the absence of interactionism from my discussion. In my opinion, the founder of interactionism, G. H. Mead, offers the outlines of a societist ontology. The societist character of Mead’s (1934, 261, 270) ideas lies above all in his claim that society is composed of “institutions,” of “organized forms of group activity,” each of which is an organized, end-related set of responses (to given situations) that are differentiated according to roles. What makes this claim societist is, inter alia, that these responses are not actions of particular individuals but instead consti- tute a structure into which “anyone” can be slotted. Mead’s position also bears affinities to site ontologies because he further contends that the identities of actions and mental states derive from the institutions in which human organisms participate. I do not pursue these affini- ties here, for two reasons. First, Mead does not elaborate his ontology sufficiently to allow definitive comparison with alternatives. Second, later interactionists inspired by Mead’s formulations have tended toward individualism.
The symbolic interactionism of Harold Blumer nicely exemplifies this subsequent individualist streak. Blumer (1969, 70-71) claims, for example, that society is a totality of “joint actions,” each of which is a set of actions performed by actors who fit together their behavior in the knowledge that their actions are part of a collective activity. He writes that
these large societal organizations . . . [are] arrangements of people who are interlinked in their own actions. . . . At any one point the partici- pants are confronted by the organized activities of other people into which they have to fit their own acts. The concatenation of these actions . . . constitutes the organization of the given molar unit. (P. 58)
For Blumer, the social world is a sum of coordinated actions that are performed by individuals who take account of what others do.
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 181
Also exemplifying this penchant is Barry Barnes’s (1995, 80) thesis that social phenomena and institutions consist of interactions. This proposition might sound societist, but on Barnes’s account interac- tions are events that occur to people: they consist in individuals acting toward one another on the basis of their knowledge, preferences, desires, and commitments. Barnes acknowledges that individuals are constituted in interactions (p. 70), but what he means is that their knowledge, preferences, desires, and the like change as a result of the other-oriented actions they perform. Barnes’s individualist interactionism is summed up in his thesis (p. 75) that the force of norms is the power of other people, as this is exerted on specific indi- viduals through interaction.4
In sum, site ontologies forge a path between individualism and hitherto dominant societisms. They join cause with societism against individualism in contextualizing the actions, mental states, and rela- tions of individuals within wider social vistas. Conversely, they join individualism against traditional societisms in espying a continuity of being between this individualist stuff and the wider vista forming its context. Because of this, site ontologies are able to elude two perva- sive criticisms, that of incompleteness directed at individualism and that of reification leveled at societism. I want now to explore this feat in greater detail.
SKIRTING INCOMPLETENESS AND REIFICATION
One prominent form taken by the charge that individualism is incomplete questions the integrity of the reduction base, that is to say, whether the individualist stuff that allegedly constitutes social phe- nomena itself presupposes something that cannot be understood as just more individualist stuff. Maurice Mandelbaum (1955) develops a famous argument to this effect in his article “Societal Facts.” Here is a different argument, directed at one prominent set of individualisms.
In virtue of what is it the case that among the things someone is intentionally doing on a given occasion is A-ing? For instance, in vir- tue of what is it the case that among the things someone, who is wav- ing his hand, is intentionally doing is calling another person hither? To narrow this issue, suppose that the performance of any action, A, consists in the performance, under certain conditions, of some basic action, B. In virtue of what, then, does a performance of basic action B in conditions C amount to the performance of action A?5
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
182 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
One prominent type of individualist theory replies that B amounts to A in virtue of the actor’s intentions, desires, and beliefs. It might be said, for instance, that the actor wanted to A (or wanted to carry out project or plan, P, the execution of which involved A-ing). The actor, furthermore, must have had certain pertinent beliefs, for instance, that the circumstances were appropriate for A-ing or that carrying out plan P required A-ing. Some accounts also claim that the actor must have formed the intention to A. All told, that John’s waving consti- tutes a call for Mary to come hither depends on his possessing certain mental states and these states joining together to determine his action. (I write “determine” instead of something more precise so as to encompass a variety of accounts of this “determination.”)
Of course, that John’s waving is calling might also have something to do with the circumstances in which it occurs. Maybe John’s father told him to call Mary hither, and John always obeys his father. Maybe Mary called John hither, and John, always the control freak, responded by calling Mary hither. As suggested, individualism can accommodate these cases. This is not so clear, however, when that to which someone responds is not another’s action but a social state of affairs, say, a declaration of war (heard on the radio) or the victory of the home team (witnessed at the arena). In these cases, the actor does not respond (though he could have) to the president’s announcement or the players’ actions. He responds, instead, to a social state of affairs: the declaration of war or the team’s victory. Further argument is needed to demonstrate that these states of affairs are either instituted without remainder by the thoughts, attitudes, or actions of specific individuals or consist in nothing but the actions and mental states of specific individuals. Accordingly, the constitution of actions might not be amenable to individualism when what the performance of basic actions constitutes is an action that responds to a social state of affairs.
What, however, enables a wave to be a call? After all, it is not the case that a person, by virtue, say, of her desires and beliefs, can per- form whatever action she pleases by performing this or that basic action. Only certain actions can intelligibly be carried out through the performance of particular basic actions, and which these are is not typically up to the actors involved (cases of explicit conventions offer exceptions). What if, furthermore, John lives in a social world where the only action waving can amount to is calling? In such a world, men- tal states, actions, and circumstances would not be immediately perti- nent to his waving amounting to a calling. Considerations such as
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 183
these indicate that even if individualist analyses marshalling beliefs, desires, and actions capture part of the institutional story, something is missing, namely, that which is responsible for the facts (1) that only certain actions, A(1)-(n), are intelligibly constituted by a given basic action, B, and (2) that different basic actions, B(1)-(n), can amount to the same action, A. This further factor is an essential part of the com- plete story about why the performance of B constitutes the perfor- mance of A.
Different accounts have been offered of the further factor. Many theorists invoke norms or rules of intelligibility. I (Schatzki 1996, chap. 2) have argued against this alternative elsewhere. Others cite beliefs about what amounts to what in given circumstances. Brief comments on this alternative are offered below. My own Wittgensteinian predilections point toward the claim that which nonbasic actions are performed when someone performs particular basic ones are tied to the understandings of action that are carried in the practices the actor carries on. Another term for understandings in the present context is possible intelligibilities, for part of what it is to understand action A is for a variety of basic actions, B(1)-(n), to be intelligible, in their circumstances, as As. In any case, waving can amount to calling only within practices in which basic actions are understood to constitute callings and waving, in particular, are intelli- gible as such. Apart from such understandings and practices, waving cannot amount to calling (explicit agreements aside). The relevant understandings are those carried in the practices the actors concerned carry on, since it is as informed by these understandings that their lives proceed as they do.
These understandings, or possible intelligibilities, cannot be con- strued as possessions of specific people, say, John and Mary. They are, instead, something carried in the practices that John and Mary carry on. This means that as a matter of fact, though not necessarily of necessity, action understandings are established, acquired, sustained, and transformed through the actions that compose practices. The intelligibility of B amounting to A is, in other words, an affair—a fea- ture, if you will—of certain practices, of the entire open-ended arrays of actions that constitute those practices. Of course, John and Mary, in participating in those practices, each have an understanding of A. Their understandings, however, are versions of that understanding of A that is carried in the practice, and the action understandings that organize a practice are distinct from the versions participants have— they are features of the overall practice and cannot be fractured into
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
184 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
the participants’ sometimes divergent versions. (Note, participants in any but the simplest practice possess versions of different subsets of the understandings carried there.) I would, furthermore, hazard the opinion that participants’ own understandings presuppose the nonpersonal practice understanding, of which they are versions.
Incidentally, what qualifies two participants’ understandings of A as versions of the same understanding of A is that either participant’s judgments of which basic actions, B(1)-(n), constitute A-ings are intel- ligible to the other (provided they share knowledge of both the action circumstances and the mental states as well as histories of the actors involved). When participants in a practice possess versions of the same understanding, they can be said to share understanding. Shared understanding, it should be pointed out, does not exclude disagree- ments about whether a given B-ing amounts to an A-ing. Participants typically diverge in their knowledge of actors’ circumstances, states of mind, and histories, and these differences underpin divergent judgments. Even when, moreover, they share knowledge of these matters, what is known might underdetermine whether the B-ing was an A-ing, again leaving open the possibility of divergence. And, of course, there are the proverbial unclear borderline cases arising from the “fuzzy” character of concepts and understanding.
The understandings carried in practices are social, therefore, in two senses: (1) in that multiple people carry on the practices involved and possess versions of the understandings carried therein and (2) in that both the intelligibilities and the practices carrying them are, in Charles Taylor’s (1971) phrase, “out there” in public space accessible in principle to anyone. It might be added, parenthetically, that belief is a poor concept for capturing this complicated situation. It is not the case that actors collectively—and perhaps independently—believe that B-ing is intelligible as A-ing, or that some practice carries the belief that B-ing is intelligible as A-ing and actors believe this (or something approximating it). B(1) . . . (n) are intelligible as As. These intelligibilities, moreover, are out there in the practice, carried in the ways people act and react to one another, regardless of whatever beliefs they might have about A, B(1) . . . (n), and circumstances.
In sum, the performance of basic action B amounts to the execution of action A only given public, practice-borne understandings that As can be performed via Bs. Some subset of the participants in the prac- tice is attuned into this public understanding, which does not pre- clude dissension over particular cases. So, one key category of indi- vidualist stuff, action, presupposes something social.
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 185
The above incompleteness argument contends that the stuff of individuals presupposes social phenomena. The flip side of this argu- ment is the claim that any attempt to resolve a social phenomenon into this stuff leaves as residue the social phenomena that the individ- ualist stuff concerned presupposes.6 This is a second form of the incompleteness argument. Consider John’s interaction with Mary. The social relation that this interaction instantiates (e.g., domination, companionship) depends on the identity of the interaction. This iden- tity rests, in turn, on what John and Mary are doing when waving, moving their legs, uttering words, and so on. As discussed, what they are doing depends, among other things, on social understandings and practices. When, therefore, the complete story is told, the social relation(s) their interaction realizes depends on something social. An individualist analysis of this social affair is incomplete.
Arguments such as these illustrate why individualism has never successfully shaken the suspicion that its analyses are incomplete. This incompleteness, in turn, makes societist alternatives appealing. By citing phenomena that are (allegedly) irreducible to the stuff of individuals, and by arguing that these phenomena are a dimension of the social, societism in effect theorizes the residue that escapes indi- vidualism. Site ontologies are of a piece with other societist ontologies in this regard. They characterize the nonindividualist dimension of the social as the site where individuals act and relate. I emphasize that site ontologies need not abandon the individualist intuition that social life concerns interrelated individuals and can be defined as something like human coexistence (furthermore, that such ontologies need not abandon the individualist insight that it is the actions of indi- viduals, singly or compounded, that are the principal causal force at work in social affairs). Site ontologies contend, however, that human coexistence transpires as part of wider context that cannot be treated as just more features of interrelated individuals.
Although site ontologies concur with other societisms that some- thing about the social escapes individualism, they avoid a problem that affects their brethren, namely, reification. The variety of nonindividualist phenomena that societisms trumpet renders a quick demonstration of the scope of this problem impossible. To give the accusation teeth, however, I shall discuss a prominent case.
Durkheim (1964) famously contends that social facts, which he also calls collective representations, are a sui generis order of facts over against facts about individuals. More specifically, social facts are ways of thinking, feeling, or acting that have two features distin-
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
186 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
guishing them from facts about individuals: their externality to and exercise of coercive power over individuals. A language, for instance, is not contained in individual actions and states of mind. To say such and such, moreover, a person must utter certain words and not others. Facts about individuals are, incidentally, above all psychological facts about individuals separately.
Durkheim is unclear whether social facts are external to individu- als taken one by one or external to all pertinent individuals as a collec- tion. Many of his phrases and examples suggest the former. If his posi- tion is to distinguish itself from individualism, however, he must affirm the latter. The clearest indication that he does this is his thesis that the substratum of a social fact is a group. What this means is that the ways of thinking, feeling, and acting he labels social facts are fea- tures of groups of individuals. Groups, moreover, are not just collec- tions. Durkheim draws his position into the orbit of wholism by sug- gesting that a group of individuals is a totality distinct from the sum of individuals in it. Accordingly, social facts are properties of something that is distinct from collections of individuals. In this sense, these facts are external to such collections. In defending, consequently, the sui generis nature of social facts, Durkheim opposes facts about groups of individuals to sums of facts about individuals.
An ambiguity remains, however. Does Durkheim think that a social fact is distinct from a sum of facts about individuals separately or a sum of facts about individuals and their relations? He speaks of social facts arising from the “synthesis” and “association” of individ- uals. He also indicates (Durkheim 1964, xlvii) that a group is a “total- ity formed by the union” of individuals. It is at least plausible that such a totality is distinct from the collection of the individuals com- posing it. It cannot, however, be distinct from the complex of interre- lated individuals involved—for a group cannot be anything different or apart from the complex involved, that is to say, from a specific (pos- sibly open) set of individuals in their relations to one another. If, then, groups are the substrata of social facts, social facts are properties of complexes of interrelated individuals. And, as such, it is not obvious that social facts are distinct from collections of facts about individuals and their relations. In any event, in defending the sui generis nature of social facts, Durkheim opposes facts about interrelated individuals to sums of facts about separate individuals.
It is misleading, consequently, to speak of the sui generis nature of social facts. Facts about interrelated individuals are distinct from sums of facts about individuals. But facts about group, or complex, A
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 187
are also distinct from those about group, or complex, B. And it is hard to know whether facts about interrelated individuals are more or less distinct from facts about individuals than is one collection of facts about interrelated individuals from another such collection. Once the individual—social opposition becomes one between sums and groups of individuals, that is to say, between individuals aggregated and interrelated, the distinction loses its clear categorical character.
As I have been portraying matters, the externality mark of the dis- tinctiveness of social facts is a by-product of the specification of the substratum of that set of facts (social facts) whose distinctiveness from another set of facts is at issue. There remains the causal criterion. If social facts are facts about groups, is the coercive power of such facts something other than the coercive power that group individuals exert over one another? Durkheim needs to answer this question affirma- tively to maintain the distinctiveness of social facts. No forthright demonstration is to be found, however, in his text. His examples con- cern the coercive power of social facts over a particular individual (referred to as “I”). They are compatible, therefore, with the coercive power of social facts being nothing but the coercive power of group individuals over one another.
My diagnosis of the situation is as follows. Durkheim seeks to delimit the domain of the social. He wants, moreover, this domain to be distinct from that of the psychological, since only so will sociology have its own subject matter different from those of psychology and biology. Accordingly, he isolates a category of facts that boast certain properties distinguishing them from psychological facts and calls these facts “social facts.” Problems arise when he feels himself called on to specify the substratum of these facts, that is to say, when he feels obliged to specify of what configuration or dimension of the stuff, of which all contrahents agree the world is composed, these facts are properties. His specification of this substratum (groups of individu- als) merges—rightly, in my opinion—the social and the individual. It reveals the dubiousness of the initial presumption that the social is distinct from the psychological. Far from the psychological and the social forming two distinct levels of reality, the substance of the social is groups, that is, complexes of interrelated individuals and thus something partly constituted by the psychological. This result under- mines, however, the sui generis character of the social. It shows that social facts are nominalist entities and that the clear distinctiveness of the social can be upheld only so long as these facts are treated as a dis- tinct level of reality, that is to say, only so long as they are reified. Once
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
188 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
a substratum is provided for them, and their reification is correspond- ingly cancelled, the distinctiveness of the social disappears.
This example illustrates the thesis that societist ontologies suffer from a penchant toward reification. Such ontologies often populate the social with phenomena and principles that are better thought of as the constructions of these ontologies. Not all societist ontologies, however, succumb to this tendency. Site ontologies hold that human coexistence is intrinsically part of a context that cannot be construed as more individualist stuff. As will emerge in the next section, the most prominent such ontologies conceive of this context as, centrally, a set of practices. A set of practices is not a substantialized thought construction as are, in different ways, whole societies, a sui generis level of social facts, and abstract structures. For such a set embraces the organized activities as part of which people’s actions and relations occur. Practices are something at once different from but ontologically continuous with this individualist stuff. The most viable sort of societist ontology, in other words, is the one closest to individualism.
To conclude this section, I consider an objection to the general sort of ontological theory pursued in this article. According to René König (1961), among others, Durkheim should not be read as offering an account of the nature of social facts. He should instead be interpreted as offering methodological directives—for example, treat social facts as things, explain social facts by other social facts—that carry no onto- logical assumptions, that is, that assume nothing about the character of social facts other than that they are not facts about individuals (though individuals experience them) and that they should be uncov- ered by the methods of positive science. This is not the place to con- sider the merits of this suggestion as an interpretation of Durkheim. It points, however, toward a wider thesis, namely, that social research- ers are best advised simply to implement methodological strategies for investigating social affairs and to avoid ontology altogether—for ontologies are nothing but unnecessary and empirically unconfirmable presumptions.
Taken by itself, an injunction such as explain social facts by other social facts might carry no implications for the nature of social facts. Following it, moreover, might demand nothing more ontologically robust than the use of an explicit, or inchoate and unformulated, crite- rion with which social facts can be distinguished from other sorts. The types of ontology discussed in the current article, however, have implications for method. The choice and use of particular methods, the inferences that are made from observations and measurements to
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 189
statements about social matters, the formulation of these statements, what of the social is thought to be directly experiencable, observable, and measurable—these matters will vary depending on whether a researcher is an individualist or believes in social structures, wholist social entities, or social facts distinct from facts about individuals (and maybe their relations). An investigator might proceed oblivious to this dependency and simply carry on research. How he proceeds, however, will implicate stands on these issues that collectively affirm at least some general type of ontology (e.g., individualism). It follows that discussions such as the current article offer a number of benefits to social research, including those of making implicit ontologies explicit, of charting the different ontologies available to social research, of clarifying the differences among these alternatives, of dis- cussing the implications of wielding (even unknowingly) particular ontologies: in short, the advantage of ontological self-consciousness and choice when studying the social world.
The methodologist objector might now reply,
OK, a variety of ontologies might inform social research. There is still no reason to argue for and against particular ontologies à la the current article. Justification in social science is empirical validity, and ontologies cannot be empirically tested. At best, they can be indirectly tested. Let researchers utilize different ontologies. Whatever ontology informs the most successful research program shall be accepted.
There are at least two problems with this response. First, unanimity does not exist in social science about what counts as a “successful” research program. All the issues that philosophers have discussed under the rubric of “the pragmatics of scientific explanation” are rele- vant here. Second, and more specifically, what counts as success often reflects ontology. What counts as a successful research program for an Althusserian political or Luhmannian sociological economist differs from what passes muster for an orthodox microeconomist—in part because of different ontologies. So empirical validity is not ontologi- cally innocent. And because ontology is tied up with social research, there is room and need in the overall enterprise of social research for ontological disputation.
In the end, I believe, one should adopt a neo-Quinian picture of social investigation in which (1) ontologies are part of the conceptual armature of social investigation and (2) arguments about ontological issues are part of the overall enterprise of social research, another part of which is the methodic gathering of data.
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
190 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
VARIOUS SITE ONTOLOGIES
Many site ontologies are Heideggarian in inspiration. They make, in Heideggarian terms, a clearing (Lichtung), that is to say, an abstract space central to the constitution of the social. Because spaces are emi- nently cut out to be that where and as part of which entities exist and events occur, space ontologies are preeminently suited to be site ones. The spaces that Heideggarian ontologies envision, moreover, are spaces of being-intelligibility; that is to say, they are spaces in which things are (meaningful) as such and such. It is within these spaces, for example, that what people are (intelligible as) doing is this and that, and what (who) they are is that and this. Accordingly, these spaces of being-intelligibility are the site where human beings carry on their lives. Heideggarian ontologies also portray social life as transpiring within such spaces. They thereby diverge from individualism, since actions of and relations among individuals exist only within some- thing nonindividualist. They also thereby differ from other societist ontologies because spaces of being-intelligibility are quite different beasts from whole societies, sui generis levels of facts, and abstract structures.
Perhaps the most prominent type of Heideggarian site ontology analyzes the social as nexuses of practice that carry spaces of intelligi- bility. According to Charles Taylor (1971), for instance, social reality is practices. In any given practice, moreover, actions and language (along with desires, feelings, and emotions) articulate a “semantic space” that envelops everything that is either part of the practice or encountered there. Practices, in other words, maintain fields of intelli- gibility in whose terms they themselves proceed. It follows that the individuals who carry on practices, as well as the relations they main- tain while doing so, are embedded in the fields of meaning that the practices concerned sustain. In this way, practices form the site where humans live and relate. A different version of this sort of ontology is found in Spinosa, Flores, and Dreyfus’s (1997) Disclosing New Worlds. The authors define a world as any organized set of practices that pro- duces a relatively self-contained web of meanings. Any world con- tains a specific range of activity, a specific range of human identities, a specific array of equipment, and an embracing style that coordinates its activities. All human coexistence transpires as part of particular such worlds.
A neo-Heideggarian site ontology closely related to the two just discussed is found in Pierre Bourdieu’s account of social life, in partic-
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 191
ular, in his notion of a field of activity.7 Examples of such fields are those of agriculture, cooking, politics, marriage, and education. A field is a domain of activity marked by particular objective condi- tions, certain overall goods to be pursued (stakes), a range of capitals that are drawn on when pursuing them, the layouts of the settings where the field’s activities transpire, and a space of actual and possi- ble activities and meanings. Both individual lives and human related- ness occur within, and combine features of, specific fields. According to Bourdieu, moreover, different fields are organized homologously and compose wholes whose scope resembles the wholistic wholes that social thought has traditionally labeled “societies.” For Bourdieu, the social is a compound of practice fields that embraces an array of action-meaning spaces.
A final example is Wittgensteinian. Collins and Kusch (1998) con- ceptualize the social as a set of societies, each of which is a conglomer- ation of forms of life. A form of life (e.g., sports, or football and cricket) is composed of a unique set of central or essential actions, along with a unique set of concepts coordinated therewith. The authors also con- tend that all actions are social institutions. Vis-à-vis individual actions, this means that a piece of behavior (bodily movement) is an action only if the group to which the actor belongs treats it as such (assuming they witness it). This is another way of saying that behav- ior B amounts to action A if it instantiates the concept of A that helps compose a form of life that the actor’s group carries on (about this concept of institution, see Barnes 1983; Bloor 1997). An action, conse- quently, is always part of some form(s) of life or other. Indeed, Collins and Kusch define actions as “things one can intentionally do in a soci- ety, that get their sense from taking place in that society” (p. 7). Because, accordingly, actions instantiate types that help compose forms of life, every interaction transpires within this or that form(s) of life. Social life, consequently, occurs within a wider setting of which it is intrinsically a part: the forms of life that compose some society. Forms of life are the site where social life transpires.
The ontology I advocate treats the site of social existence as a mesh of practices and arrangements. The differences between this ontology and the foregoing ones, to which it bears differing and sometimes great resemblance, will be outlined later. By “practices” I mean orga- nized human activities. Examples are religious practices, political practices, economic practices, baseball practices, and cooking prac- tices. Each is an open-ended set of actions linked by pools of under- standings (pertaining to action), a collection of rules (explicit formula-
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
192 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
tions), and a “teleoaffective structure” (a range of normativized, hierarchically ordered ends, projects, and tasks, to varying degrees allied with normativized emotions). Bank loan practices, for instance, are a set of such actions as filling out loan application forms, transfer- ring money, waiting in line, securing credit reports, calling people on the phone, and meeting in offices that are linked by (1) understand- ings, among other things, of how to carry out and recognize these actions in particular circumstances; (2) rules that pertain to the grant- ing of loans, the transferal of information, administrative procedure, and the like; and (3) a range of task-project combinations that are enjoinedly or acceptably carried out for the sake of such prescribed or acceptable ends as making profit, earning interest, gaining access to capital, and securing promotion. To say that particular actions are linked by these phenomena is to say that these actions express the same understandings, observe the same rules, or pursue ends and execute projects that are elements of the same normative teleological order, namely, those end project–task combinations that are accept- able or prescribed for participants in the practice in question.
I emphasize that these practices are not aggregates of the proper- ties of specific individuals, for instance, the actions, understandings, rules, desires, and emotions of specific people. The actions that com- pose a practice are, to be sure, ones specific individuals perform. But the organization of a practice is not a set of properties of specific indi- viduals. A practice is organized by an array of intelligibilities, rules, ends, projects, and ways things matter. This array is distinct from, and differentially incorporated into, the minds of participants. As broached in the previous section, the phenomena that organize prac- tices are incorporated into minds in the sense of being contained in people’s “mental states,”8 for instance, their understandings, desires, beliefs, and emotions. (For example, for X to be intelligible as Y is for X to be understood as such; in order for something to be an end, it must be the object of desire.) Nonetheless, the organization of a practice remains distinct from its incorporation: the end of making profit that helps organize bank loan practices, for example, is distinct both from any given participant’s desire and from any sum of participants’ desires for profit.9
Because intelligibilities and ends and so forth are contained in mental states, the organization of a practice can also be portrayed as a normativized array of mental states. This formulation emphasizes the continuity in being between practices and individuals. At the same time, it violates individualist sensibilities in envisioning something
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 193
that might be called “objective mind.” The mental states that organize a practice, qua components of the practice’s organization, are not the states of particular individuals. Qua organizing phenomena, they are distinct from whatever mental states actual participants possess. Par- ticipants’ states are, at best, versions of these organizing states, and many of their mental states are not even this. For example, the desire to earn profit that helps organize bank loan practices is distinct from both the desire of any given bank employee and the desires of any sum of employees to do so. That is to say, that this desire organizes the practice is distinct from either this or that or any specific set of actual participants desiring profit. As a set, moreover, the array of states that organizes the practice is expressed in the entire open manifold of actions composing the practice. It is an affair (a feature) of the mani- fold and not of any subset of practitioners. Of course, through the training and learning people undergo in catching on to the practice, they come to possess versions of many of its organizing states. More- over, any pair of participants possesses versions of some subset of these states in common and in this sense share mentality. But the states organizing the practice remain distinct from versions of themselves.
Another reason the mentality that organizes a practice is distinct from that of its participants is that a teleological structure contains a range of ends and projects, that is, desires, namely, all those that are acceptable or prescribed in the practice. This range is broader than any sum of participants’ pertinent actual desires.
Note that this ontology is individualist in spirit in so far as it spurns entities that resist resolution into entities of sorts acceptable to indi- vidualism. It differs from individualism, however, because the partic- ular entities of these sorts it cites are the properties of neither specific individuals nor specific sets thereof.
Because the organization of a practice is expressed only in the (open) set of actions that composes the practice, and is not contained in the sum of minds of those participating in it, practices resemble Durkheim’s social facts, which are facts about complexes of individu- als (as opposed to collections of psychological facts about individuals separately). Despite this convergence in ontological status, three major differences distinguish practices from Durkheim’s collective ways of thinking, acting, and feelings. First, practices do not enjoy causal power over individuals. To the extent, for example, that people become participants in practices by being “molded” as such, it is the actions of other participants, not the practices themselves, that
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
194 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
accomplish this. (As suggested, my ontology converges with individ- ualism in deeming the activity of individuals, singly or compounded, the principal causal force in social life.) Practices, in addition, do not cause the actions of their participants. At best, the versions that partic- ipants possess of the mental states that organize practices help deter- mine what they do. A second difference is that practices do not require a substratum of the sort that Durkheim supplies for his social facts. There is no need to specify the configuration of individualist matters of which they are properties. For the concept of a practice does not presuppose that the social and psychological (i.e., individual) levels are distinct: practices are composed of both the actions of individuals and a range of mental states, of which individuals have versions. A final difference is that of all the phenomena Durkheim calls social facts, I accord practices alone the property of being irreducible to indi- vidualist stuff. Practices, moreover, number among the phenomena through which other Durkheimian social facts such as financial sys- tems, language, and crowd phenomena are to be analyzed.
In sum, although my site ontology converges in many regards with individualism, the central component of sites, namely, practices, defies individualist analysis. Practice organization is “out there” in the practice; only versions of different parts of it are “in here” in differ- ent individuals’ minds.
How are practices part of the site where social life transpires? Social life (human coexistence) can most generally be construed as the hanging together of human lives. Two key sorts of link through which lives hang together are (1) chains of action encompassing the acts of different people and (2) commonalities, as well as meshings, of the ends, projects, and emotions people pursue and suffer. Many instances of these two sorts of link are elements of the nexuses of action or teleological structures that characterize various particular practices. For instance, what bank employees and customers do at a bank (e.g., ask for identification, complete forms, hand over cash, wait in line) and, thus, the chains of action their acts form there are largely, though not exclusively, moments of banking practices: these actions and chains occur as part of these practices. Similarly, the ends and pro- jects these employees and customers pursue (e.g., serving customers’ needs, making loans, withdrawing money, earning interest) fall mostly within the teleological structure of these banking practices: they pursue these ends and projects in participating in those practices. Accordingly, the commonalities and meshings of their ends and pro- jects are a feature of the practices. In ways such as these, banking prac-
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 195
tices form that wider expanse, as part of which the links, through which the lives of bank employees and customers hang together, occur. Practices are one dimension of the site of social life.
The second dimension of this site is material arrangements of peo- ple, artifacts, organisms, and things. An arrangement of such entities is a configuration of them in which they have meaning and identity. I just mentioned that two types of link through which lives hang together are chains of action and commonalities, as well as meshings, of goals, projects, and emotions. A third prominent type (there are more) is arrangements that encompass, as well as physical connec- tions (e.g., telephone lines) among, both specific entities in and the layouts of material settings. The lives of bank employees hang together, for instance, through the layouts of individual offices, the organization of offices, passageways, and other rooms at the bank, as well as the telecommunication links between different locales. These layouts, organizations, and links are all arrangements of artifacts, organisms, and things. Hence, because human coexistence also tran- spires through material arrangements such as these, these arrange- ments are a second dimension of the site of social life.
Links among the lives of bank employees and their customers are features of banking practices and the material arrangements at banks. These practices and arrangements are meshed in the senses that the practices transpire at these arrangements and are molded in various ways by them, just as arrangements form the settings of practices and are, to varying extents, set up and altered within them. It follows that the site of the sociality that transpires at the bank is not merely a col- lection but rather a specific mesh of practices and arrangements. This is, furthermore, always the case: the site of any aspect of social life is some mesh of human practices and material arrangements. What is more, practices, arrangements, and practice-arrangement meshes are intricately interlaced, forming, among other things, nets of meshes and confederations of nets. And at their widest expanse, they form an immensely complex overall nexus of practices and arrangements that connect, overlap, and interpenetrate in a labyrinthine, unsystematic, and contingent fashion. This immensely complex nexus is the overall site of human coexistence at any moment or durée of time.
Networks of practices are what escape individualist accounts of social life. They are what actions, mental states, and relations among individuals presuppose and what must supplement these phenom- ena in analyzing either sociality in general or the composition of specific social phenomena. For instance, that intelligibility of B-acts as
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
196 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
A-acts that enables the performance of basic action B to amount to the performance of action A is part of the organization of the practices in which A-ings appear. Almost anyone who participates in these prac- tices is bound to grasp this intelligibility (though, as discussed, this does not exclude disagreements about whether a specific B-ing did, in fact, amount to an A-ing).
Meshes of practices and arrangements also differ considerably from the phenomena that previous societisms have judged immune to individualist analysis. Meshes are not wholes since neither they nor individual practices are subject to dynamic principles or causal forces that are different from those bearing on individuals. What’s more, be- cause practices, arrangements, and meshes thereof are spatial-temporal phenomena, they are phenomena of a fundamentally different sort than the abstract structures of structuralism. Even the organization of a practice is a spatial-temporal phenomenon. Differences between practices and Durkheimian social facts, finally, have already been aired. In sum, the site of social life is not a reification. It is composed of organized manifolds of concrete human activity meshed with arrangements of human beings, artifacts, organisms, and things. This ontology dances between the individualist and extant societist variet- ies, combining and refusing the virtues and sins of each.
The just outlined ontology is also superior to the other exemplars of the site genre discussed above.10 One reason for this is that it grants both practices and arrangements immense, though not equal, onto- logical, causal, and prefigurative (future-organizing) significance in social life. This feature distinguishes it from Taylor’s ontology. Like many other theorists, Taylor simply overlooks the significance of material entities. He construes practices alone as the context where social life transpires and ignores the significance of arrangements of entities for the interrelatedness of human lives.
Bourdieu (1990) and the team of Spinosa, Flores, and Dreyfus (1997) treat practices and arrangements alike as significant dimen- sions of the site of social life. They acknowledge the role of arrange- ments by making the equipmental layouts of settings components of these sites. Their ontologies suffer, however, from a tendency toward overunification. As described, Bourdieu ties objective conditions, stakes, capitals, layouts, and spaces of action and meaning into tight, self-propagating bundles called “fields.” He further contends that fields are homologously organized and also compose larger units bearing great resemblance to societies as traditionally conceived. Spinosa, Flores, and Dreyfus presume, meanwhile, that the overall
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 197
mesh of practices and arrangements (to use my terminology) takes the form of a set of worlds, each of which exhibits a unified style. These unifying visions overlook the ragged, fine character of social life. More specifically, they overlook, among other things, how prac- tices in different fields or worlds interweave (for instance, mutual assistance practices at work and child care practices at home); how arrangements reach across fields and worlds, often independently of the practices found there (e.g., how the telecommunications links between bank buildings are continuous with wider telecommunica- tions networks connecting myriad locales); how local socialities can accommodate multiple styles and how single styles can traverse local socialities (e.g., how highway traffic can accommodate multiple driv- ing styles and hospital affairs different healing and nurturing styles; how these healing and nurturing styles can be manifested in the dif- ferent socialities of family, friendship, public meetings, and hospital affairs); and how the packages that practices and arrangements form are contingent and unstable (e.g., how a given setting can be the scene of extremely different practices, for instance, a public square as the scene of entertainment and political practices). There is no space at present to justify the intuition of difference, complexity, and meta- morphosis that inform these criticisms, but I might mention that the works of Foucault partly fuel it.
Finally, my ontology differs from that of Collins and Kusch (1998) in its conception of the “enveloping” phenomenon in which social life transpires: practices (organized action manifolds) versus forms of life (nets of action types and concepts).11 These two phenomena converge in that each is a concept-imbued phenomenon of action. Collins and Kusch evince little concern, however, for either the organization of or the relations that characterize the action bundles that compose the site of human coexistence. All they mention in this regard is that roles dif- fer in the actions they permit. On my account, however, the nexuses of activity, as moments of which individual actions transpire, are com- plex affairs intricately ordered normatively and teleologically.
EPISTEMOLOGICAL IMPLICATIONS
I conclude this article by sketching the tasks involved in develop- ing a site-ontological analysis of social events and phenomena. The principal tasks are fourfold: (1) delimiting whatever activity episodes (if any) compose an event or phenomenon of interest (this is espe-
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
198 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
cially crucial vis-à-vis social events, for example, interactions); (2) uncovering the practice-arrangement mesh, net of meshes, and/or confederation of nets (a) of which these activity episodes or the phe- nomenon of interest is a component or, conversely, (b) that composes the phenomenon in question (as in the case of, e.g., government); (3) uncovering the further meshes, nets, and confederations to which this mesh, net, or confederation is connected, intentionally or uninten- tionally; and (4) tracing the chains of human and nonhuman action that, in circulating within, passing through, and linking all these meshes, nets, or confederations (a) render the latter harmonious, competitive, or conflictual and (b) either lead to and spiral away from the social event of interest or maintain or transform the target social phenomenon. I stress that the order of 1 through 4 does not indicate temporal, or even logical, priority. As examples of phenomena ana- lyzable as events, meshes, nets, and confederations, consider loan transactions at a bank, the branch office of a bank, the bank corpora- tion, and the banking industry.
Suppose the investigative aim is to understand loan transactions at, or the structure and proceedings more generally of, a branch office of a bank. Two crucial tasks are delimiting the pertinent actions (either by conceptual specification or through empirical inquiry) and uncov- ering the practice-arrangement mesh that composes the office. Prac- tices are uncovered by identifying the ends, projects, rules, and actions that compose them, whereas arrangements are uncovered both by attending to the setups where specific practices occur and by tracing the physical connections that link these setups to further ones.12 The bank office is composed of a complex of practices such as accounting, advertising, counter transacting, waiting in lines, and borrowing/loaning funds, which occur amid a complicated material order that encompasses rooms, technological arrays, queue-control devices, and the like. Prominent among the nets and confederations to which the office mesh is linked are the bank corporation and the banking industry. Whereas the bank corporation is composed of a net of practice-arrangement meshes, in particular, the interrelated meshes composing its branches and central headquarters, the bank- ing industry is composed of an interrelated confederation of corpo- rate nets.
Although understanding either loan transactions at or the pro- ceedings and structure of the branch office probably does not require a detailed overview of the corporate net and industry confederation, it does require some grasp of the linkage between the branch and the
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 199
other meshes composing the corporation. This linkage, like all link- ages among meshes, nets, and confederations, is composed of over- laps and interactions among the meshes concerned. Meshes overlap when they share teleological orders, rules, actions, or arrangements. Meshes interact, most importantly, when chains of action pass between and through them. Overlaps and interactions among meshes are such that the linked meshes, nets, and confederations are harmo- nious (mutually sustaining), competitive, or in conflict (destructive). For example, linkages between a given branch office and other meshes in the corporate net are typically harmonious, though branch meshes sometimes compete with one another (say, for increased cor- porate funds). Conversely, linkages between the corporation and the nets composing other bank corporations are usually competitive, though they can be harmonious (as when the banking industry lob- bies the government). In any event, understanding the proceedings at and structure of the branch office requires ascertaining, first, the prac- tice-arrangement mesh that composes the office; second, the har- mony and competition between this and the other meshes composing the corporation; and maybe third, the competitive (and harmonious) linkages between the corporation and other nets (as well as meshes, e.g., bank robberies).
The above discussion abstracts from many facets of social investi- gation, for example, the selective attention to linkages that follows from investigative interest in particular proceedings at or features of the branch office and not others (loan policy, say, versus internal orga- nization). In this vein, someone interested not in proceedings at a bank office but in the workings of the corporation or banking industry requires a much thicker grasp of the linkages both between the office and the corporation and among the corporations. In every case, fur- thermore, a description of the corporation (net) or the industry (con- federation) is in effect a specification of the state of the practice- arrangement meshes of which the net or confederation is ultimately composed—though most of the details of these meshes are obviously left out of the description, and the practices, arrangements, meshes, and linkages are designated with singular expressions. This means that someone investigating the corporation or the industry (or the office for that matter) must mediate between detailed grasps of partic- ular interconnected practice-arrangement meshes and summary overviews of more inclusive complexes.
Finally, understanding the perpetuation and transformation of either office proceedings or the office (corporation or banking indus-
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
200 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
try) requires a grasp of the chains of action that circulate within, pass through, and link the office and so forth to other social formations, both within the corporation and industry and without. Again, carry- ing out this task typically requires the investigator to gain an over- view of the myriad chains that crisscross some space-time swath of human coexistence and to identify particularly consequential ones. I believe that there cannot be a general theory of action chains and that overviews and salient selection are the most investigators can achieve, but this is an issue for a different article.
NOTES
1. I marshal this unfortunate neologism as the positive label for nonindividualist ontologies because of the political connotations of the expression that is morphologi- cally more satisfying in this context, namely, “socialist.”
2. In the literature, the expression “holism” is regularly used to designate accounts that advocate social wholes. I use “wholism” in its place because “holism,” in many areas of contemporary philosophy, connotes inner interconnectedness, and what are so joined need not form a whole.
3. Other versions of this sort of context include Deleuze’s (1988) notion of the out- side and the conception found in Wittgenstein’s (1958) texts of the context in which lan- guage and action have meaning. Although, as will be discussed in the third section, several prominent site ontologies are interpretations of Heidegger’s clearing of being, I hesitate, because of the ontological difference, to cite the clearing as an example of a site ontology.
4. I should add that it is not clear to me how Barnes’s account of interaction jibes with the site-supporting concept of institution he develops elsewhere (e.g., Barnes 1983). See the discussion of Collins and Kusch in the third section.
5. Notice that I am not addressing the general nature of what Goldman (1970) calls “level generation.” I am asking only about, so to speak, the first level of generation: the first intentional action constituted by a given basic action. What further actions that action, in turn, might constitute are not at issue, although some considerations in the text bear on these further generations.
6. Mandelbaum’s (1955) argument illustrates this flip side as well—indeed, it shows that the two claims are sides of the same coin—since it is in order to show that an individualist analysis of social facts is incomplete that he argues that the specification of individualist facts presupposes social ones.
7. I might mention that a Derridean (as opposed to naturalized) Heideggarian site ontology is found in Laclau and Mouffe (1985). The authors identify the social as the sum total of discourses coalescing within what they call the “field of discursivity.”
8. For reasons that cannot be aired here, I prefer to call understandings, desires, beliefs, and emotions “life conditions” rather than “mental states.” Doing so helps, among other things, overcome the Cartesian intuitions widely associated with such conditions. For the sake of exposition, however, I will refer to such conditions as “men- tal states.” For discussion of this issue, see Schatzki (1996, chap. 2).
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
Schatzki / A NEW SOCIETIST SOCIAL ONTOLOGY 201
9. The organization of a practice, especially its teleoaffective structure, might seem to possess the same ontological status as the rule-resource structures that Giddens (1979) attributes to practices. One major difference is that Giddens’s structures both govern and are reproduced in human activity, whereas teleoaffective structures are simply recurring effects of what actors do in conjunction with the “practical intelligibil- ity” that determines this. Another difference lies in our divergent accounts of under- standing/intelligibility. For discussion of these points, see Schatzki (1997).
10. There is no space to discuss how my ontology departs from these other exem- plars in conceptualizing the space of intelligibility sustained in a practice as a space, not of world intelligibility alone (the intelligibility of entities) but of world and practical intelligibility (what makes sense to actors to do) together.
11. It might appear that like Taylor, Collins and Kusch (1998) ignore objects. This is not so, however, because they are deeply concerned with the roles that machines and other artifacts can play in forms of life. Their remarks on society are too brief, moreover, to know whether they follow other site ontologists in overunifying social life.
12. These operations proceed via observation, dialogue, participant observation, reading, and the like. They are also subject to important hermeneutic-interpretive con- siderations and strictures that cannot be examined here.
REFERENCES
Barnes, B. 1983. Social life as bootstrapped induction. Sociology 4:524-45.
———. 1995.
The elements of social theory. Princeton, NJ: Princeton University Press. Bloor, D. 1997. Wittgenstein: Rules and institutions. London: Routledge.
Blumer, H. 1969.
Symbolic interactionism. Englewood Cliffs, NJ: Prentice Hall. Bourdieu, P. 1990. The logic of practice. Translated by R. Nice. Stanford, CA: Stanford
University Press.
Collins, H. M., and M. Kusch. 1998.
The shape of actions: What humans and machines can do.
Cambridge, MA: MIT Press.
Deleuze, G. 1988.
Foucault. Translated by S. Hand. Minneapolis: University of Minne-
sota Press.
Durkheim, E. 1964.
The rules of sociological method. Translated by S. Solovay and J.
Mueller. New York: Free Press.
Giddens, A. 1979.
Central problems in social theory. Berkeley: University of California
Press.
Gilbert, M. 1989.
On social facts. Princeton, NJ: Princeton University Press.
Goldman, A. 1970.
A theory of human action. Princeton, NJ: Princeton University Press. Grathoff, R. 1978. The theory of social action. Bloomington: Indiana University Press. Hayek, F. A. 1952. The counter-revolution in science: Studies on the abuse of reason. Glencoe,
IL: Free Press.
König, R. 1961. Einführung. In
Die Regeln der soziologischen Methode, by E. Durkheim.
Neuwied, Germany: Luchterhand.
Laclau, E., and C. Mouffe 1985.
Hegemony and societist strategy: Toward a radical demo-
cratic politics. London: Verso.
Mandelbaum, M. 1955. Societal facts.
British Journal of Sociology 6:305-17. Mead, G. H. 1934. Mind, self, and society. Chicago: University of Chicago Press.
Downloaded from pos.sagepub.com by Wayne Mellinger on October 20, 2011
202 PHILOSOPHY OF THE SOCIAL SCIENCES / June 2003
Schatzki, T. 1996. Social practices: A Wittgensteinian approach to human activity and the social. New York: Cambridge University Press.
———. 1997. Practices and actions: A Wittgensteinian critique of Bourdieu and Giddens. Philosophy of the Social Sciences 27:283-308.
Searle, J. 1995. The construction of social reality. New York: Free Press.
Spinosa, C., F. Flores, and H. Dreyfus. 1997.
Disclosing new worlds: Entrepreneurship, dem-
ocratic action, and the cultivation of solidarity. Cambridge, MA: MIT Press.
Taylor, C. 1971. Interpretation and the sciences of man.
Review of Metaphysics 25:3-51. Wittgenstein, L. 1958. Philosophical investigations. Translated by G. E. M. Anscombe.
New York: Macmillan.
Theodore R. Schatzki is professor and chair of philosophy at the University of Kentucky. He is the author of Social Practices (Cambridge University Press, 1996) and The Site of the Social (Penn State, 2002) and coeditor of The Practice Turn in Contemporary Theory (Routledge, 2001). 

Comments