Political polarization?

Is the American electorate "polarized" with regard to sets of political issues? McCarty, Rosenthal, and Poole accept the common view that we have in fact become more polarized in our politics over the past twenty years, and they offer an interesting theory of what is causing this polarization in Polarized America: The Dance of Ideology and Unequal Riches (Walras-Pareto Lectures). This theory was discussed in an earlier post. Delia Baldassarri and Peter Bearman take a different perspective, however, in a 2007 article, "Dynamics of Political Polarization" (AMERICAN SOCIOLOGICAL REVIEW, 2007, VOL. 72 (October:784–811)). Here is how Baldassarri and Bearman frame their research:

In this article we provide a parsimonious account for two puzzling empirical outcomes. The first is the simultaneous presence and absence of political polarization—the fact that attitudes rarely polarize, even though people believe polarization to be common. The second is the simultaneous presence and absence of social polarization—the fact that while individuals experience attitude homogeneity in their interpersonal networks, these networks retain attitude heterogeneity overall. We do this by investigating the joint effects of personal influence on attitudes and social relations. (784)

Baldassarri and Bearman quote a range of studies that find that the mass of the US population is not polarized in the bulk of its political attitudes, and that it has not increased in polarization in the past decade. "The evidence suggests that, aside from a small set of takeoff issues, 'the policy preferences of different social groupings generally move in parallel with each other'” (784). They resolve part of the paradox by distinguishing between activist opinions and public opinion:

In the same vein, Fiorina and colleagues (2005) dispute “The Myth of a Polarized America” and suggest that the “culture war” commonly conjured up in the media is a fictive construction. According to their analysis, there is no popular polarization, but simply partisan polarization—“those who affiliate with a party are more likely to affiliate with the ‘correct' party today than they were in earlier periods” (p. 25). It is the political elite and a small number of party activists that are polarized.

This all seems a little paradoxical, so it's worth looking at the assumptions these two groups of researchers are making about "polarization".

To start, what is meant by polarization with respect to a given issue -- say gay marriage? Essentially the concept is a characteristic of a population's distribution across an attitudinal scale from strongly support to strongly oppose with respect to the issue in question. A population is homogeneous if the distribution of scores has a single peak and a small standard deviation, and is polarized if it has two (or more) peaks. Here is a diagram representing the results of their agent-based model of attitude diffusion. Each issue eventually shows a pronounced degree of polarization after several hundred iterations, with about half the population distributed around a positive attitude and the other half distributed around a negative attitude. Presumably we can define increase in polarization as a shift apart of the two peaks (kurtosis) and perhaps a decrease in the deviation around the peaks.


Theoretically a population could be segmented into three distinct groups -- perhaps one-third who cluster around the zero point of indifference and two extreme groups on the left and right.

The most original part of their work here is an effort to model the emergence of issue polarization based on a theory of how social interactions in networks and small groups influence individuals' attitudes. They offer a sociological theory of inter-personal influence to explain how attitude diffusion occurs within a population, and they report the results of network simulations to illustrate the consequences of this theory. They argue that this model explains how members of society can perceive polarity while actually embodying a high degree of homogeneity.

In more general terms, we show that simple mechanisms of social interaction and personal influence can lead to both social segregation and ideological polarization. (785)

Our goal has been to deploy a model of inter-personal influence sensitive to dynamics of political discussion, where actors hold multiple opinions on diverse issues, interact with others relative to the intensity and orientation of their political preferences, and through evolving discussion networks shape their own and others' political contexts. In the model, opinion change depends on two factors: the selection of interaction partners, which determines the aggregate structure of the discussion network, and the process of interpersonal influence, which determines the dynamics of opinion change. In the next section, we organize the description of the model around these two elements. Table 1 summarizes the simulation algorithm. (788)

The simulations are very interesting. The authors specify assumptions about the structure of interactions; they specify how an individual's attitude is affected by the interaction; and they creat an initial distribution of attitudes for the 100 actors in the simulation. They then run the set of actors and interactions through 500 iterations and observe the resulting patterns of distribution of attitudes.

The cases resolve into two large groups: non-takeoff, where polarization does not emerge and takeoff, where polarization does occur. The first group is much more common, validating the prior finding that public opinion is not becoming more polarized. The "takeoff" group is much less common but important. For some initial distributions of attitudes and interaction pathways the population does develop IMO two sharply divided sub-groups. These two diagrams illustrate these two possibilities.



In the second figure the population is moving strongly towards polarization around the issue, whereas the first figure represents a population with no pattern of polarization.

Several things are striking about this work. First is the degree to which it presents a picture of public opinion that seems highly counterintuitive in 2012. The first half of their paradox seems even more compelling today than five years ago -- the American public does seem to be very divided in its opinions about social and moral issues. The second striking thing is perhaps an omission in the foundations of their theory of attitude formation. Their model works through 1-1 interactions. But it seems evident that a lot of attitude formation is happening through exposure to the media -- television, radio, Internet, social media. There doesn't appear to be an obvious way to incorporate these powerful influences into their model. And yet these may be much more influential than 1-1 interactions.

This research is of interest for two important reasons. First, it is a sustained effort to account for how issue separation occurs in real social groups. And second, it provides an excellent and detailed example of a microfoundational approach to an important social process, using a variety of agent-based modeling techniques to work out the consequences of the theory of social influence with which they begin. The models allow Baldassarri and Bearman to carefully probe the assumptions of the theory of individual-level attitude dynamics that they postulate. So the work is both substantively and methodologically rewarding. It is analytical sociology at its best.

Meso powers and causal mechanisms

I've argued in earlier posts for several ideas:

1. Meso-level social structures have causal powers.
2. Meso-level social structures must have microfoundations.
3. Causal relations are carried out by causal mechanisms.

Are these claims consistent with each other?  An earlier post argues for the consistency of (1) and (2).  Here I will address the relation between (1) and (3): what kinds of causal mechanisms are available to mediate the causal powers of meso-level structures? Are there meso-level mechanisms as well as meso-level causes?  Or are the mechanisms that mediate meso causal powers themselves necessarily located at the micro level?

It is evident that there are micro-level mechanisms for meso-level causal powers. The mechanisms of a run on the bank or a self-fulfilling prophecy are both micro-level; they show how a meso-level event disaggregates onto the beliefs and actions of the individuals involved, leading to collective behavior that results in the meso-level effect. These are This is exactly the kinds of mechanism that analytical sociologists, including Thomas Schelling, want to be the exclusive kind of causation in the social realm. So the hard question is this: Are there meso-level causal mechanisms?

I addressed one aspect of this question in an earlier post, where I considered whether causal mechanisms can be complex and multi-staged.  There I focused on the idea that most theories of causal mechanisms require that there be a reasonably strong regularity between input and output.  This implies that a multi-step causal process will probably not qualify as a mechanism, because the intervening causal steps will imply a low association between initial condition and output condition.  (See the post for more details.)  And this in turn implies that a social mechanism needs to consist of one step: the instigating condition occurs and then the output condition occurs. The diagram illustrates the point. Scenario I represents a single-step causal mechanism; scenario II represents a three-step causal process composed of three component mechanisms.  Scenario II does not qualify as a mechanism itself, because the compound probability of D given A is only 50%.

So this would seem to answer the question for the case in which the putative meso-level causal mechanism involves multiple steps of causation. For example: {[Distracting work environment] => [low productivity] => [low profits] => [downsizing]} is a multiple-step causal process entirely at the meso level; but it doesn't establish a meso-level mechanism {[Distracting work environment] => [downsizing]} because the intervening steps imply new contingencies and a subsequent low conditional probability between antecedent and consequent.

There is another possible case to consider as well, however: the case in which a given meso-level structure or occurrence itself gives rise with high probability to a given meso-level result.  Are there such examples?

It would appear that there are.  Social isolation (referred to in an earlier post) is a meso-level characteristic of a population.  And Alford Young argues that decreasing social isolation causes rising inter-group hostility, another meso-level characteristic of a population.  This seems like a clear and uncontroversial example of a meso-meso causal link that is also a meso-level causal mechanism. It looks like Scenario I above.  Moreover, this mechanism can be employed in explanations of more complex social events like race rebellions.  It would seem that there are similar examples in the work of Robert Sampson on neighborhood effects (link).

And here is a third possibility. A meso-level structure S's causal powers may result from the causal powers of its component systems Ti, and the causal mechanisms that embody the causal powers of S may be found at the level of mechanisms at the level of Ti.  For example, a certain type of police organization S may have a high level of corruption C among its street officers.  So S has the causal power to bring about C. The mechanisms that establish this power may reside at a lower level of organization: the processes for giving assignments and the processes for checking incident reports. The diagram might look something like this.

The dashed line indicates the causal power. The solid arc lines on the left indicate "composition". The solid arrows indicate meso-meso causal mechanisms.  And the dotted lines indicate the aggregate consequences of "poor orientation to goals" and "weak incentives for conformance", which is an elevated level of officer corruption. So the meso-level power depends on sub-meso level mechanisms and actor-level mechanisms.

So the conclusion I'd like to tentatively offer is this: Meso-level causal powers are created as a consequence of specific causal mechanisms triggered by the properties and changes of the meso-level structure. Those mechanisms may be at the actor level; they may be at a sub-meso level; or they may be at the meso level.

Organizational failure as a meso cause

A recurring issue in the past few months here has been the validity of meso-level causal explanations of social phenomena. It is self-evident that we attribute causal powers to meso entities in ordinary interactions with the social world. We assent to statements like these; they make sense to us.

• Reorganization of the city's service departments led to less corruption in Chicago.
• Poor oversight and a culture of hyper-masculinity encourages sexual harassment in the workplace.
• Divided command and control of military forces leads to ineffective response to attack.
• Mining culture is conducive to social resistance.

We can gain some clarity on the role played by appeals to meso-level factors by considering a concrete example in detail. Military failure is a particularly interesting example to consider. Warfare proceeds through complex interlocking social organizations; it depends on information collection and interpretation; it requires the coordination of sometimes independent decision-makers; it involves deliberate antagonists striving deliberately to interfere with the strategies of the other; and it often leads to striking and almost incomprehensible failures. Eliot Cohen and John Gooch's Military Misfortunes: The Anatomy of Failure in War is a highly interesting study of military failure that makes substantial use of organizational sociology and the sociology of failure more broadly, so it provides a valuable example to consider.

Here are a few framing ideas that guide Cohen and Gooch in their analysis and selection of cases.

True military "misfortunes" -- as we define them -- can never be justly laid at the door of any one commander. They are failures of the organization, not of the individual. The other thing the failures we shall examine have in common is their apparently puzzling nature. Although something has clearly gone wrong, it is hard to see what; rather, it seems that fortune -- evenly balanced between both sides at the outset -- has turned against one side and favored the other. These are the occasions when it seems that the outcome of the battle depended at least as much on one side's mishandling of the situation as on the other's skill in exploiting a position of superiority ... The causes of organizational failure in the military world are not easy to discern. (3)

From the start, then, Cohen and Gooch are setting their focus on a meso-level factor -- features of organizations and their interrelations within a complex temporally extended period of activity.  They note that historians often start with the commander -- the familiar explanation of failures based on "operator error" -- as the culprit.  But as they argue in the cases they consider, this effort is as misguided in the case of military disaster as it is in other kinds of failure.  Much more fundamental are the organizational failures and system interactions that led to the misfortune. Take Field Marshal Douglas Haig, whose obstinate commitment to aggressive offense in the situation of trench warfare has been bitterly criticized as block-headed:

Not only was the high command confronted by a novel environment; it was also imprisoned in a system that made it well-nigh impossible to meet the challenges of trench warfare. The submissive obedience of Haig's subordinates, which Forester took for blinkered ignorance and whole-hearted support, was in reality the unavoidable consequence of the way in which the army high command functioned as an organization under its commander in chief. (13)

So why are organizations so central to the explanation of military failure?

Wherever people come together to carry out purposeful activity, organizations spring into being. The more complex and demanding the task, the more ordered and integrated the organization. ... Men form organizations, but they also work with systems. Whenever technological components are linked together in order to carry out a particular scientific or technological activity, the possibility exists that the normal sequence of events the system has been designed to carry out may go awry when failures in two or more components interact in an unexpected way. (21, 22)

And here is the crucial point: organizations and complexes of organizations (systems) have characteristics that sometimes produce features of coordinated action that are both unexpected and undesired.  A certain way of training officers may have been created in order to facilitate unity in combat; but it may also create a mentality of subordinacy that makes it difficult for officers to take appropriate independent action.  A certain system for managing the flow of materiel to the fighting forces may work fine in the leisurely circumstances of peace but quickly overload under the exigencies of war.  Weapon systems designed for central Europe may prove unreliable in North Africa.

Eliot and Gooch place organizational learning and information processing at the core of their theories of military failure.  They identify three kinds of failure: "failure to learn, failure to anticipate, and failure to adapt" (26). As a failure to learn, they cite the US Army's failure to learn from the French experience in Vietnam before designing its own strategies in the Vietnam War.  And they emphasize the unexpected interactions that can occur between different components of a complex organization like the military.  They recommend a "layered" analysis: "We look for the interactions between these organizations, as well as assess how well they performed their proper tasks and missions" (52).

The cases they consider correspond to this classification of failure.  They examine failure to learn in the case of American antisubmarine warfare in 1942; failure to anticipate in the case of the Israel Defense Forces' failure on the Golan Heights, 1973; and failure to adapt in the British disaster at Gallipoli, 1915.  Their example of aggregate failure, involving all three failures, is the defeat of the American Eighth Army in Korea, 1950.  And they reserve the grand prize, catastrophic failure, for the collapse of the French army and air force in 1940.

Each of these cases illustrates the authors' central thesis: that organizational failures are at the heart of many or most large military failures.  The example of the failure of the American antisubmarine campaign in 1942 off the east coast of the United States is particularly clear.  German submarines preyed at will on American shipping, placing a large question mark over the ability of the Americans to continue to resupply Allied forces in Europe.  The failure of American antisubmarine warfare was perplexing because the British navy had already developed proven and effective methods of ASW, and the American navy was aware of those methods.  Unhappily, Eliot and Gooch argue, the US Navy did not succeed in learning from this detailed wartime experience.

The factors that Eliot and Gooch cite include: insufficient patrol boats early in the campaign: insufficient training for pilots and patrol vessel crews; crucial failures on operational intelligence ("collection, organization, interpretation and dissemination of many different kinds of data"; 75); and, most crucially, organizational failures.

A prompt and accurate intelligence assessment would mean nothing if the analysts could not communicate that assessment directly to commanders on the scene, if those commanders did not have operational control over the various air and naval assets they required to protect shipping and sink U-boats, if they saw no reason to heed that intelligence, or if they had no firm notion of what to do about it. The working out of correct standard tactics ... could have no impact if destroyer skippers did not know or would not apply them. Moreover, as the U-boats changed their tactics and equipment ..., the antisubmarine forces needed to adopt compensating tactical changes and technological innovation. (76)

This contrasts with the British case:

The British system worked because it had developed an organizational structure that enabled the Royal Navy and RAF to make use of all of the intelligence at their disposal, to analyze it swiftly and accurately, and to disseminate it immediately to those who needed to have it. (77)

So why did the US Navy not adopt the British system of organization?  Here too the authors find organizational causes:

If the United States Navy had thought seriously about adapting its organization to the challenges of ASW in a fashion similar to that chosen by the British, it would have required major changes in how existing organizations operated, and in no case would this have been more true than that of intelligence. (89)

So in this case, Eliot and Gooch find that the failure of the US Navy to conduct effective antisubmarine warfare off the coast of the United States resides centrally in features of the organization of the Navy itself, and in collateral organizations like the intelligence service.

This is a good example of an effort to explain concrete and complex historical outcomes using the theoretical resources of organizational sociology.  And the language of causation is woven throughout the narrative.  The authors make a credible case for the view that the organizational features they identify caused (or contributed to the cause) of the large instances of military failure they identify.

Are there meso-level social causes?

Social structures and other social "things" are ontologically peculiar in some ways. Most especially, they are abstract, distributed, and non-material. We can't put a culturally dominant food aversion or a group prejudice in a box and weigh it. And yet many of us want to say that social structures are "real", not merely theoretical constructs.

One important aspect of something's being real is that it has causal powers: the specific properties of the thing bring about differences in the world on the behavior of other things. This is a version of the interventionist theory of causation associated with Jim Woodward: change something about C and you bring about a change in E (link).

So let's consider this question: Can meso-level social structures have meso-level effects?

Of course meso-level structures have effects -- on individuals. The fact that there are laws and enforcement mechanisms governing highway speed has some effect on drivers' behavior. The question here is whether it is legitimate to postulate causal powers for structures whose effects are realized in other meso-level structures. And I want to explore the affirmative answer to the question: it is legitimate and coherent to assert meso-meso causal interactions, and we sometimes have empirical evidence to support such assertions.

(It would be possible, of course, to take the view that social structures are epiphenomenal and have no causal properties whatsoever. On that approach, what seems to be the effect of the legal system on individual behavior is really just the aggregate effect of the many individuals involved in the legal system. I don't find this view at all compelling, however.)

My question is relevant to two groups of sociological theorists, each of whom thinks the answer is trivial and obvious -- but in opposite directions. The new methodological individualists, represented by analytical sociology, think the answer is trivially "no", because social causation proceeds always and exclusively through actions and interactions of individuals (link). This is the fundamental idea underlying Coleman's Boat as a model of the relationship between macro and micro.  And a range of anti-individualists -- Giddens, Elder-Vass, Archer -- believe it is self evident from everyday experience that causal structures do have causal powers, and that it is a waste of time to defend the notion (link). It is obvious.

My position is a precarious one. On the one hand I advocate an actor-centered approach to sociology and the social sciences. I defend the idea that social claims need microfoundations in a specific (weak) sense. And on the other hand I believe that structures have a degree of stability that permits us to couch causal claims in terms of those structures directly, rather than needing to supplement those claims with disaggregated foundations at the level of the individual. So I argue for the idea that we can sometimes regard causal powers of social entities as "relatively autonomous" from individual-level facts.

By meso-level structures I mean to refer to things like these:

• National Science Foundation
• Nuclear Regulatory Commission
• IBM corporation
• AFL-CIO
• German paramilitary organizations 1930
• German ideology of cultural despair 1910
• Islamic norms of Zakat
• ...

In each case there are numerous actors assigned to roles, governed by rules defining their activities, and leading to a certain kind of functioning in the broader social environment.

Generically I would define a meso-level structure as --

A composite of individuals and roles that incorporates a set of rules and norms for internal and external actors, and that possesses procedures of inculcation and enforcement through which internal and external actors are brought to comply with the rules and norms (to some degree).

I would define a normative system as ...

a set of rules, norms, and expectations embodied in a population of actors and meeting a threshold level of success in coordinating and constraining behavior.

We have a number of sociological concepts that capture social items at this level: organization, bureaucracy, institution, normative community, social network, communications system, legal system, civil war, military coup, advocacy group.

It is evident that social entities often incorporate elements of several of these kinds of things. Organizations and structures often incorporate or depend upon normative systems, and normative systems often generate organizations and institutions that convey their impact to the young and adult actors.
What about other mid-level social nouns -- ethnic group, electorate, financial crisis, ...? These strike me as being compounds of a miscellaneous set of social things -- there are bits of organizations, normative systems, affinity groups, and social networks in each of them. The concept of assemblage seems to fit these nouns well.

The "meso" qualifier is a bit more difficult to specify. It is intended to focus our attention on mid-level social arrangements, between actors and global institutions like the US state, global Islam, and the world trading system. The intuitive idea is straightforward. These are the smaller-scale, lower-level social arrangements or units of which macro structures are composed. Bert Leuridan makes an effort to offer a more specific definition based on causal roles.

So I have a simple but important question in mind here. Is it ever legitimate to assert something like this:

• Meso structure X produced changes in meso structure Y,

without being obliged to demonstrate the individual-level pathways through which this effect is thought to have come about? Is a type 4 causal claim ever supportable (link)?

The question I am posing is related to the idea of the methodological individualism associated with James Coleman. Basically the idea propounded by Coleman and more recently by the analytical sociologists is that all social properties, including causal powers, work through the activities of individuals, and we need ideally to replace claims that appear to attribute causal powers to structures with theories that disaggregate these powers onto the patterned activities of individuals. This is a reductionist theory.

Other theorists, notably Dave Elder-Vass, want to assert that social structures have "emergent" properties and powers (link). An emergent property according to E-V, is one that is possessed by the aggregate but not by the composing units. On this account, there are causal properties of structures that cannot be represented as the aggregate effect of individual actors.

My own approach depends on a line of reasoning long familiar in the special sciences. It is anti-reductionist, in that it denies that we need to derive higher-level properties from lower-level properties. It accepts the compositional ontology: social structures are composed of individual actors. But it asserts explanatory autonomy for theoretical statements about mid-level mechanisms (link, link, link).

One particularly direct way of supporting the idea that structures have meso effects is to establish correlations at that level for a few examples. But this isn't the only way we establish causation in other areas of the sciences. We do experiments ("remove X and observe whether Y persists"), we analyze the outcomes of "natural" experiments, we do comparative studies, and we engage in process tracing of particular cases. We even engage in theoretical analysis to try to determine what causal powers a certain entity ought to be expected to have given its constitution.

So it seems that there is ample room for sociologists to assert and investigate the causal properties of social structures. And given appropriate attention to the principle of microfoundations, we have a social ontology that supports the legitimacy of such claims as well.

A related question is whether there are "mechanisms" that operate at the meso level, or whether all social mechanisms must operate at the individual level (as Hedstrom and the AS world believe).  I will return to this question.

The social world as morphogenesis

Critical realism has progressed far since Roy Bhaskar's early writings on the subject in A Realist Theory of Science.  One of the most important thinkers to have introduced new ideas into the debate is Margaret Archer. Several books in the mid-1990s represented genuinely original contributions to issues about the nature of social ontology and methodology, including especially Realist Social Theory: The Morphogenetic Approach and Culture and Agency: The Place of Culture in Social Theory.

Archer's work addresses several topics of interest to me, including especially the agent-structure dichotomy. This is key to the twin concerns I have for "actor-centered social science" and "autonomous meso-level explanations".  Anthony Giddens offers one way of thinking about the relationship between agents and structures (link).  Archer takes issue with the most fundamental aspect of Giddens's view -- his argument that agents and structures are conceptually inseparable. Archer argues instead for a form of "dualism" about agents and structures -- that each pole needs to be treated separately and in its own terms.   (Chapter 5 provides a detailed discussion of both Bhaskar and Giddens on levels of the social.) She acknowledges, of course, that social structures depend on the individuals who make them up; but she doesn't believe that this basic fact tells us anything about how to analyze or explain facts about either agents or structures.  Here are the opening paragraphs of Realist Social Theory.

Social reality is unlike any other because of its human constitution. It is different from natural reality whose defining feature is self-subsistence: for its existence does not depend upon us, a fact which is not compromised by our human ability to intervene in the world of nature and change it. Society is more different still from transcendental reality, where divinity is both self-subsistent and unalterable at our behest; qualities which are not contravened by responsiveness to human intercession. The nascent 'social sciences' had to confront this entity, society, and deal conceptually with its three unique characteristics.

Firstly, that it is inseparable from its human components because the very existence of society depends in some way upon our activities. Secondly, that society is characteristically transformable; it has not immutable form or even preferred state.  It is like nothing but itself, and what precisely it is like at any time depends upon human doings and their consequences.  Thirdly, however, neither are we immutable as social agents, for what we are and what we do as social beings are also affected by the society in which we live and by our very efforts to transform it. (1)

Archer argues that the two primary approaches that theorists have taken -- methodological individualism and methodological holism -- are fundamentally inadequate.  They represent what she calls upward and downward conflation.  In the first case, "society" disappears and is replaced by some notion of aggregated individual action; in the second case "agents" disappear and the human individuals do no more than act out the imperatives of social norms and structures.  She associates the first view with J.S. Mill and Max Weber and the second view with Durkheim.  On her view, agents and structures are distinct, and neither is primary over the other.

She refers to her view as the "morphogenetic" approach.  Here is how she explains this concept:

The 'morpho' element is an acknowledgement that society has no pre-set form or preferred state: the 'genetic' part is a recognition that it takes its shape from, and is formed by, agents, originating from the intended and unintended consequences of their activities. (5)

Morphogenesis applies at all levels, from "the capitalist system" to "the firm" to "the actor" to personal identity and motivation.  And she believes that properties at various levels -- micro and macro -- have a degree of autonomy from each other, which she refers to as "emergence":

I want to maintain that 'micro' and 'macro' are relational terms meaning that a given stratum can be 'micro' to another and 'macro' to a third etc. What justifies the differentiation of strata and thus use of the terms 'micro' and 'macro' to characterize their relationship is the existence of emergent properties pertaining to the latter but not to the former, even if they were elaborated from it. (9)

Later in the book she amplifies this idea:

Autonomy is also temporal (and temporary) in the joint senses that such structural properties were neither the creation of contemporary actors nor are ontologically reducible to 'material existents' (raw resources) and dependent upon current acts of human instantiation (rule governed) for all their current effects. (138)

And this is where her theory exhibits its "realism": she asserts that the properties we identify at various levels or "strata" are real and causally powerful.

Thus in the course of this book, frequent references will be made to 'the societal'. Each time, this has a concrete referent - particular emergent properties belonging to a specific society at a given time. Both the referent and the properties are real, they have full ontological status, but what do they have to do with 'the big'? The society in question may be small, tribal and work on a face-to-face basis. Nor do they have anything to do with what is, relatively, 'the biggest' at some point in time. We may well wish to refer to certain societal properties of Britain (the 'macro' unit for a particular investigation) which is an acknowledged part of bigger entities, like Europe, developed societies, or the English speaking world. We would do so if we wanted to explain, for example, the role of the 'Falklands factor' in recent elections and in so doing we would also incidentally be acknowledging that people who go in for it take their nationalism far from 'impersonally', and that the 'site' of neo-colonialism may be far distant. (10)

 And she offers a different way of thinking about "micro" and "macro" -- not from small to large, but from interactional and local to systemic (11).

This all adds up to a social realism that is militant in affirming the reality of social properties as emergent properties.

Conversely social realism which accentuates the importance of emergent properties at the level of both agency and structure, but considers these as proper to the strata in question and therefore distinct from each other and irreducible to one another, replaces the terms of the traditional debate with entirely new ones. Irreducibility means that the different strata are separable by definition precisely because of the properties and powers which only belong to each of them and whose emergence from one another justifies their differentiation as strata at all. (14)

So what is Archer's central notion, the idea of morphogenesis?  It is the idea that processes of change occur for agents and social structures in interlocking and temporally complex ways.  Agents are formed within a set of social structures -- norms, language communities, power relationships.  The genesis of the agent occurs within the context of these structures.  On a larger time scale, the structures themselves change as a result of the activities and choices of the historically situated individuals who make them up.  She summarizes this ontology as a set of cycles with different time frames: structural conditioning => social interaction => structural elaboration (16).  This notion leads Archer to a conception of the social and the actor that reflects a fundamentally historical understanding of social processes. Formation and transformation are the central metaphors (154).

A final comment about Archer's philosophy of social science is relevant here.  She provides a theory which is abstract and philosophical -- ontology, debates about emergence and reduction, epistemology.  But she does so in consideration of her own detailed research on the history and development of an important aspect of social reality, educational institutions.  So her abstract philosophical ideas are grounded not only in philosophy but also in historical and sociological research.

Perhaps I'm over-interpreting, but it seems to me that Archer's realist theory of morphogenesis is highly complementary to the ideas about methodological localism that have been argued for here.  The idea that actors are socially constituted and socially situated (methodological localism) is a different way of expressing her point that actors are constituted by surrounding social structures. The idea that structures are themselves adapted and changed by active individuals doing things within them corresponds to her "social interaction" and "structural elaboration" phases of morphogenesis. The methodological insight that seems to come along with morphogenesis -- the idea that it is valuable to move both upwards towards more comprehensive social structures and downwards towards more refined understanding of action and interaction -- is certainly a part of the view associated with methodological localism and actor-centered sociology. Her view of the inherent "transformability" of society (1) parallels my own view of the heterogeneity and contingency of social arrangements.  Finally, her notion that social ontology must be addressed before we can make much progress on issues of methodology and explanation seems right to me as well.

Giddens on agents and structures

Anthony Giddens is one of the theorists whose ideas are most often invoked when the idea of social-structural explanation is in play. His 1979 collection of essays, Central Problems in Social Theory: Action, Structure, and Contradiction in Social Analysis, is a classic statement of some of his views.  Here is how he frames his core concern in a key essay, "Agency, Structure":

The principal issue with which I shall be concerned in this paper is that of connecting a notion of human action with structural explanation in social analysis. The making of such a connection, I shall argue, demands the following: a theory of the human agent, or of the subject; an account of the conditions and consequences of action; and an interpretation of 'structure' as somehow embroiled in both those conditions and consequences. (49)

Giddens refers some of these issues back to the tradition of American pragmatism, and the theories of George Herber Mead in particular (link).

Within more orthodox sociological traditions, symbolic interactionism has placed most emphasis upon regarding social life as an active accomplishment of purposive, knowledgeable actors; and it has also been associated with a definite 'theory of the subject', as formulated in Mead's account of the social origins of reflexive consciousness. (50)

Giddens faults this tradition for not being able to conceptualize the social-structural context with sufficient precision.  He finds, for example, that Durkheim's efforts to provide theoretical resources for describing the "external or objective" character of society were inadequate (51).  Generally his view here is that theorists have failed in their conceptualizations of structures and agents:

Parson's actors are cultural dopes, but Althusser's agents are structural dopes of even more stunning mediocrity. (52)

The problem is that neither individualists nor structuralists have succeeded in expressing the inherent interdependence of the two poles.  Give primacy to structures and the agents are "dopes" -- robots controlled by structural conditions.  Give primacy to individuals, and structures and institutions seem to disappear.  His own view is that the two poles of structure and agency must be considered from within a common formulation:

I shall argue here that, in social theory, the notions of action and structure presuppose one another; but that recognition of this dependence, which is a dialectical relation, necessitates a reworking both of a series of concepts linked to each of these terms, and of the terms themselves. (53)

Giddens observes that action necessarily implies a temporal framework.

'Action' or agency, as I use it, thus does not refer to a series of discrete acts combined together, but to a continuous flow of conduct.  We may define action ... as involving a 'stream of actual or contemplated causal interventions of corporeal beings in the ongoing process of events-in-the-world'. (55)

Actions take place in contexts; and the contexts include crucially the actions of other people and the constraints and opportunities created by social structures.  Giddens adds another component of action: the forms of knowledge that actors have on the basis of which they tailor their interventions.

So what about "structure"?  Giddens prefers to talk about "structuration" -- the temporally extended processes through which social constraints evolve and take hold.

I want to suggest that structure, system and structuration, appropriately conceptualised, are all necessary terms in social theory. (62)

What is a "structure"? Here is one effort at definition provided by Giddens:

The term 'social structure' thus tends to include two elements, not clearly distinguished from one another: the patterning of interaction, as implying relations between actors or groups; and the continuity of interaction in time. (62)

He refers to these two aspects as "syntagmatic" and "paradigmatic" dimensions of social structures.  And he summarizes the concept of structure in these terms: "the rules (and resources) that, in social reproduction, 'bind' time" (63). "Structures can be identified as sets or matrices of rule-resource properties" (63-64).  Here is a summary table:

source: Giddens, Central Problems in Social Theory, p. 66.

The activity of structuration is key to Giddens's theorizing agents and structures, because it represents the link between the two.

The concept of structuration involves that of the duality of structure, which relates to the fundamentally recursive character of social life, and expresses the mutual dependence of structure and agency. (69)

Knowledge plays a key role in structuration; it provides the basis on which agents both understand and transform the rules around them. Agents, in other words, are reflexive cognitive actors.

 Institutions involve signification, domination and legitimation (106). In "Institutions, Reproduction, Socialisation" Giddens talks about structures in terms of means of mediation and transformation within a collection of active participants.

I want to suggest that each of the three aspects of structure I have distinguished can be understood as ordered in terms of the mediations and transformations which they make possible in the temporal-spatial constitution of social systems.... Writing and other media of communication ... bind much greater distances in time and space. (103)

A distinction that comes up a lot in Giddens's work in this collection is that between synchronic and diachronic states of affairs.  We can think of a structure as a snapshot at a moment in time of a set of relations, beliefs, rules, and opportunities.  This would be a synchronic description of the structure; it is a static approach to a social structure. Or we can look at a structure as being in a process of generation, reproduction, and transformation; this would be a diachronic and dynamic way of thinking about structures.  Giddens's affinity to the idea of structuration suggests that he is especially interested in the dynamic questions -- the ways in which actors, roles, and rules interact over time, leading to changes in the snapshot.

What Giddens's treatment here doesn't adequately express, in my reading, is what we think structures and institutions really are.  Are they complexes of patterned activities by numbers of actors?  Are they ensembles of social practices? Is the IBM corporation simply a large set of social interactions?  Or is there an abiding abstract social reality -- division of labor and authority; segmentation of responsibilities; interlocking productive activities -- that can be identified as a social entity? What about the capitalist economic structure; is this a stable social entity, or is it simply an ensemble of patterns of relations of meaning and power?  It seems that these examples are the kinds of thing that Giddens wants to refer to as a "system"; but it's not clear.

So the concept of social structure still seems underdeveloped here. What about the idea that agent and structure are inseparable? This I understand in a fairly direct way: agents are always located in a web of social relationships that define them and define the opportunities they confront.  And structures are always constituted by individuals thinking, acting, and interacting in specific ways.  So we literally cannot separate agents and structures; they are mutually constitutive.

(Here is an intriguing dynamic network simulation of the spread of HIV infection based on data from the National Longitudinal Study of Adolescent Health.  "Small differences in reported behavior can potentially explain the large racial disparities in HIV infection observed in the United States.")

Does "culture" require microfoundations?

I have consistently argued for a philosophy of social science that emphasizes the actor and the availability of microfoundations. I argue for an actor-centered sociology. But I have lately been arguing as well for the idea that it is legitimate for social scientists to treat claims about the causal properties of meso-level social structures as being relatively autonomous from their microfoundations.

This approach doesn't satisfy all comers.  Some don't like meso-level causal properties at all, and others don't like the idea that meso-level properties are in any way dependent on the level of individuals and their actions and agency.

In particular, some readers would prefer a meso-autonomist strategy that dispenses with individuals altogether; one according to which we can identify certain causal factors that do not need microfoundations at the level of the individual at all. Candidates for such factors often fall under the large umbrella of culture: symbols, meanings, practices, rituals, traditions, grammars, and the like. I would say, however, that these items too require microfoundations. Cultural items are sometimes thought to be supra-individual and independent from the concrete individuals who live within their scope. And it is true that culture exercises a specific kind of independence.  But no less than any other social characteristic can cultural features evade their embodiment in individual actors and institutions.

If we take the view that the obligations of zakat (charity) are a profound part of Muslim identity and that this element of Islam explains certain social outcomes, then I want to know how these elements of identity are conveyed to children and practitioners at the local level. What are the concrete social mechanisms of inculcation and communication through which a Bangladeshi child comes to internalize a full Muslim identity, including adherence to the norms of zakat? To what extent are there important differences within Bangladeshi society in the forms of identity present in Muslims -- urban-rural, male-female, rich-poor? And equally interestingly – in what ways do those processes give rise to a Muslim identity in Bangladesh that is somewhat different from that in Indonesia, Morocco, or Saudi Arabia?

Identities, cultures, and systems of meaning are no less embodied in the states of mind of actors than are the calculating features of rationality that underlie a market society. So the fault of methodological individualists in this sphere is not that they fail to recognize the inherent autonomy of systems of cultural meaning; it is rather that they adhere to a theory of the actor that does not give sufficient attention to the variations and contingencies that characterize actors in various social and historical contexts. Ideas about the independence of cultural items from the level of individuals are suggestive and interesting, and I think they need to be fully confronted by an actor-centered sociology. But I do not believe they are incompatible with an actor-centered sociology.

Take the independence of a code of behavior from the specific individuals who are subject to the code. It is true that one individual cannot influence the code, which is embodied in the thoughts and actions of countless others. But the reality of the code at any given time is in fact entirely dependent on those thoughts and actions (and artifacts created by previous actors). Moreover, the individual's embodiment of the code of behavior is in turn caused by a series of interactions through childhood and adulthood within a social setting.

It is certainly true that facts about culture make a difference in meso- and macro-level outcomes. A collective farm that was populated by actors who embodied Chairman Mao’s ideal of “socialist man” would have functioning characteristics very different from those observed -- no “easy riders,” lots of earnest Stakhanovites.  So standard organizational analysis of the tendencies towards low productivity in collective agriculture is dependent on something like a purposive agent theory of the actor.  Different kinds of actors give rise to different kinds of organizations.

This does not mean, however, that we could not have reasonably good understandings of “organizations” under differently realized structures of agency.  This seems to be part of the work that Andreas Glaeser (2011) is doing in Political Epistemics: The Secret Police, the Opposition, and the End of East German Socialism. Glaeser tries to understand how organizations like the Stasi functioned in a setting in which participants’ understandings and motivations were changing rapidly.

So my answer to my own question is, yes.  Cultural entities and characteristics do require microfoundations, and it is in fact a fruitful avenue of sociological and ethnographic investigation to discover the concrete social mechanisms and pathways through which these entities come to be embodied in various populations in the ways that they are.

Does the microfoundations principle imply reductionism?

My philosophy of social science has always and consistently maintained the idea that social facts depend on the activities and beliefs of individuals. There is no social "stuff" that exists independently from individual actors. I have encapsulated that idea in the form of the "microfoundations" principle: any claim about the characteristics or causal powers of social entities must be compatible with there being microfoundations for those properties and powers at the level of the actor.

At the same time, I also believe that there is an appropriate domain for social science: the exploration of the features and powers of the social world. I don't believe that methodology should force the sociologist to become a psychologist or to shift his/her attention to the micro level.

Are these two premises compatible? Or does the microfoundations principle actually entail reductionism? Does it imply that explanations couched at the level of social vocabulary are incomplete and derivative, and that the real explanation must be found at the level of the micro-activities of individuals?

I attempt to resolve this apparent dilemma by distinguishing between strong and weak versions of the microfoundations principle: "social explanations must provide microfoundations for their assertions about social properties and powers" versus "social explanations must be compatible with there being microfoundations for their assertions about social powers and properties." The weak version reflects an appropriate stipulation based on what we know about the ontology of the social world, whereas the strong version is a kind of explanatory reductionism that is unjustified.

My position, then, is that sociology is a special science in Fodor's sense, and that sociologists both can and do treat their domain as relatively autonomous.

Several commentators allege that my commitment to microfoundations -- which is unwavering -- vitiates my ability to claim relative explanatory autonomy for the meso level. Some don't like my distinction between weak and strong microfoundations, and others think that commitment to MF means explanations have to proceed through explicit discoveries of the MF pathways.

My position is intended to exactly parallel physicalism in cognitive science: we are committed to the idea that all cognitive processes are somehow or other embodied and carried out by the central nervous system. But we are not obliged to actually perform that reduction in offering a hypothesis and explanation at the level of cognitive systems.

Even more prosaically: we believe that the properties of metals depend upon the quantum properties of subatomic particles. Does anyone seriously believe that civil engineers aren't giving real explanations of bridge failures when they refer to properties like tensile strength, compression indices, and mechanisms like metal fatigue? We can observe and measure the metal's properties without being forced to provide a quantum mechanical deduction.

One observer writes that "Little's examples actually confirm that meso-level mechanisms work only through micro-level processes." Yes, and I likewise confirm that cognitive processes work only through neural events and material properties work only through quantum physics. But I don't accept that this demonstrates that the higher level cannot be treated as having real causal properties. It does have those properties; and we simply reaffirm the point that somehow or other those properties are embodied in the lower level elements. This isn't a new idea; it was contained in Jerry Fodor's "Special Sciences" article years ago. If the argument is generally a bad one then we are forced to undo a lot of work in cognitive science. If it is generally compelling but inapplicable to social entities then we need to know why that is so in this special case of a special science.

To be clear, I too believe that there is a burden of proof that must be met in asserting a causal power or disposition for a social entity -- something like "the entity demonstrates an empirical regularity in behaving in such and such a way" or "we have good theoretical reasons for believing that X social arrangements will have Y effects." And some macro concepts are likely cast at too high a level to admit of such regularities. That is why I favor "meso" social entities as the bearers of social powers. As new institutionalists demonstrate all the time, one property regime elicits very different collective behavior from its highly similar cousin. And this gives the relevant causal stability criterion. Good examples include Robert Ellickson's new-institutionalist treatment of Shasta County and liability norms and Charles Perrow's treatment of the operating characteristics of technology organizations. In each case the microfoundations are easy to provide. What is more challenging is to show how these social causal properties interact in cases to create outcomes we want to explain.

The best reason I am aware of to doubt stable causal powers for social entities is founded on the point that organizations and institutions are too plastic to possess enduring causal properties over time. I've made this argument myself on occasion. But researchers like Kathleen Thelen in The Evolution of Institutions demonstrate that there are in fact some institutional complexes that do possess the requisite stability.

So I continue to believe both things: that statements about social entities and powers must be compatible there being microfoundations for these properties and powers; and that it is theoretically possible that some social structures have properties and powers that are relatively autonomous, in the sense that we can allude to those properties and powers in explanations without being obliged to demonstrate their microfoundations.

Macro causes of European fascism

Michael Mann's book Fascists makes use of causal claims at a range of levels, from the macro to the micro, to explain the emergence of European fascism.  Here is a passage that highlights four macro-level causes of fascism:

The interwar period in Europe was the setting that threw up most of the self-avowed fascists and saw them at their high tide. My definition is intended firstly as “European-epochal,” to use Eatwell’s (2001) term (cf. Kallis 2000: 96), applying primarily to that period and place – though perhaps with some resonance elsewhere. The period and the continent contained four major crises: the consequences of a devastating “world,” but in fact largely European, war between mass citizen armies, severe class conflict exacerbated by the Great Depression, a political crisis arising from an attempted rapid transition by many countries toward a democratic nation-state, and a cultural sense of civilizational contradiction and decay. Fascism itself recognized the importance of all four sources of social power by explicitly claiming to offer solutions to all four crises. And all four played a more specific role in weakening the capacity of elites to continue ruling in old ways. (23)

So what are the causal ideas expressed here?

The factors Mann singles out here are decidedly macro-level:

• war
• class conflict and economic depression
• rapid transition to democratic nation-states
• cultural impressions of decay

These are high-level social conditions involving military power, economic power, political power, and cultural realities. Perhaps not surprisingly, these factors correspond to the main legs of Mann's own theory of social power: "My earlier work identified four primary 'sources of social power' in human societies: ideological, economic, military, and political" (5).

So the causal factors identified here are clearly at the macro level.  The outcome Mann identifies is equally macro-level: the advent of fascist movements and governments in a handful of major European states.  So the basic claim here is a macro-macro causal claim.

The causal claims expressed in the paragraph can be summarized in this way:

• Factors F1, F2, F3, F4 each played a causal role in the rise of fascism
• Factors F1, F2, F3, F4 each weakened (caused) the capacity of elites to continue to rule

What is the meaning of the idea that "F1 played a causal role in the rise of fascism"? Most simply, it is the notion that the factor occupies a position in the full causal diagram or causal narrative of the rise of fascism, beginning at some point in time.  The action of hops in the process of brewing beer plays a causal role: many events and processes must occur in a timed sequence, but the activity of the hops is one necessary part of the overall process.

And how would an investigator piece together the causal narrative of a complex happening?  It would appear that the method of "process tracing" is the most direct way of piecing together a causal narrative.  This requires going through one or more empirical cases and probing the events that occurred to attempt to assess whether and how they played a causal role in the production of the outcome. This is exactly the form that Mann's investigation of the various fascisms of Europe takes; he examines the histories and tries to discern the causal sequences that are contained in them. (George Alexander and Andrew Bennett consider some of the challenges of this methodology in Case Studies and Theory Development in the Social Sciences.)

To say that a condition is a cause of a given outcome expresses as well the idea that the condition is either necessary or sufficient for the outcome; the presence or absence of the condition makes a difference for the occurrence of the outcome. The appearance of the cuckoo is neither necessary nor sufficient for the chiming of the clock; so the cuckoo is not a cause of the chiming. It would appear, then, that Mann is also committed to claims like this:

• If war and depression had not occurred then fascism would not have prevailed in Italy 
• If Spain's democracy had been more solid and well established, then fascism would not have prevailed in Spain

Other causal ideas are suggested by the paragraph, even if not explicit:

• War has the causal power to stimulate powerful social movements in combatant countries.
• Widespread economic depression has the causal power to stimulate class antagonism.
• Ideologies have the causal power to stimulate mobilization of adherents.
• Ideologies of cultural decay have the power to weaken the capacity of elites to govern.

How do these ideas about causal powers flesh out in detail?  How does "war" possess a causal power? War encompasses a complex set of circumstances and interlocking organizations: mobilization and demobilization of mass armies, disruption of civilian production, massive damage to people and property, unusual stresses on governments, etc. And each of these circumstances in turn has consequences which ultimately influence the circumstances in which the mass home population finds itself.  Those home circumstances in turn play into the factors that are known to stimulate and amplify social movements -- popular grievances about government, economic deprivation, a general environment of uncertainty, and the availability of entrepreneurial leaders and organizations prepared to take advantage of these conditions. So war has a causal power that is embodied by the social, economic, demographic, and political circumstances that accompany it; and that power is expressed through influence on the home population.

One way of encapsulating this kind of story about the causal powers of a structural circumstance is to say that the circumstance conditions and motivates actions by many individuals in ways that lead to a certain class of outcomes.  So the causal power story is also a Coleman's Boat kind of story; it is a specification of the microfoundations of the causal power in question.  However, once we have satisfied ourselves about the microfoundations, we are not compelled to retrace our steps through the individual level in order to move the argument from Italy to Spain. We can rely on the idea that war has a given set of causal powers on macro outcomes in the next case in which we observe war and disorder.

Actor-centered sociology

I've advocated many times here for the advantages of what I've referred to as "actor-centered" sociology. Let's see here whether it is possible to say fairly specifically what that means. Here is an elliptical description of three aspects of what I mean by "actor-centered sociology":

First, it reflects a view of social ontology: Social things are composed, constituted, and propertied by the activities and interactions of individual actors -- perhaps 2, perhaps 300M. Second, it puts forward a constraint on theorizing: Our social theories need to be compatible with the ontology. The way I put the point is this: social theories, hypotheses, and assertions need microfoundations. Third, "actor-centered sociology" represents a heuristic about where to focus at least some of our research energy and attention: at the ordinary processes and relations through which social processes take place, the ordinary people who bring them about, and the ordinary processes through which the effects of action and interaction aggregate to higher levels of social organization.

(a) This means that sociological theory need to recognize and incorporate the idea that all social facts and structures supervene on the activities and interactions of socially constructed individual actors. It is meta-theoretically improper to bring forward hypotheses about social structures that cannot be appropriately related to the actions and interactions of individuals. Or in other words, it means that claims about social structures require microfoundations.

(b) The meta-theory of actor-centered sociology requires that all social theories, at whatever level, require a theory of the actor. Economics and ethnomethodology differ in the level of specificity they offer for their theories of the actor; but both have such a theory.  They both put forward fundamental ideas about how actors think and the mental processes that influence their actions.

(c) Actor-centered sociology suggests that careful study of local social mechanisms and behaviors is a worthwhile exercise for sociological research.  Ethnomethodology and the careful, place-based investigations offered by Goffman and Garfinkel move from the wings to the stage itself.

(d) It appears to imply that we may be able to provide an explanation of at least some higher-level social facts by showing how they emerge as a result of the workings of actors and their structured interactions. This is the aggregation-dynamics methodology (link).  Or in terms discussed elsewhere here, it is the micro-to-macro link of Coleman's boat (link).

(e) The actor-based sociology approach seems to imply that the regularities that may exist at the level of macro-social phenomena are bound to be weak and exception-laden. Heterogeneity within and across actors -- across history and across social settings -- seems to imply multiple sets of attainable aggregate outcomes.  Would fascist organizations flourish in Italy after World War I? The answer is indeterminate.  There were numerous groups of social actors with important differences in their states of agency, and these groups in turn were influenced by organizations of varying characteristics. So it would be impossible to say in advance with confidence either that fascism was likely to emerge or that it was unlikely to emerge (link).

(f) The actor-centered approach suggests that we can do better sociology by being more attentive to subtle differences in agency in specific groups and times. George Steinmetz's careful attention to the processes of formation through which colonial administrators took shape in nineteenth-century Germany illustrates the value of paying attention to the historical particulars of various groups of actors, and the historically specific circumstances in which their frames of agency were created (link). It implies that context and historical processes are crucial to sociological explanation.

(g) The actor-centered approach highlights the importance of careful analysis of the mechanisms of communication and interaction through which individuals influence each other and through which their actions aggregate to higher level social outcomes and structures.  Social networks, competitive markets, mass communications systems, and civic associations all represent important inter-actor linkages that have massively important consequences for aggregate social outcomes.

(h) Finally, the actor-centered approach has some of the advantages of the spotlight in a three-ring circus. The idea of actor-centered sociology points the spotlight to the parts of the arena where the action is happening: to the formation of the actor, to the concrete setting of the actor, to the interactions that occur among actors, to the aggregative processes that lead to larger outcomes, and to the causal properties that those larger structures come to have.

One thing that is somewhat troubling for anyone who has been reading this blog over time is that there seems to be a glaring inconsistency in two lines of thought emphasized repeatedly here: first, that social facts require microfoundations; and second, that meso-structures can have autonomous causal properties. Are these two ideas consistent?

In particular, one might interpret the imperative of actor-centered sociology as a particularly restrictive view of social causation: from configurations of actors to meso-level social facts.  So all the causal "action" is happening at the level of the actors, not the structures.  Dave Elder-Vass attempts to avoid this implication by arguing for emergent social causal properties (link); I've approached the problem by talking about relatively autonomous causal properties at the meso-level (link).  I continue to think the latter view works reasonably well.  In a post on "University as a causal structure," for example, I think a plausible case is made for both ideas: the tenure system is causally effective in constraining individual faculty members' behavior as well as being causally effective in influencing other structural features of the university; and every aspect of this system has microfoundations in the form of the structured circumstances of action and culturation through which the bureaucratic agents in the system behave. Or in other words: it is consistent to maintain both parts of the dilemma, actor-centered sociology and relatively autonomous meso-level social causation (link).

George Herbert Mead on the self

Sociologists sometimes come back to George Herbert Mead as a founder who still has something important to contribute to contemporary theory. This is especially true in ethnomethodology and symbolic interactionism, but it comes up in current lively discussions of pragmatism and action as well. So what can we learn from reading Mead today?

Mead writes and thinks in a way that is both scientific and philosophical. His contributions are to the field of "social psychology," and he locates himself within a discourse that includes Watsonian behaviorism and William James's introspectionism. But much of his prose seems very familiar to me as a philosopher. You can hear the reverberations of earlier philosophical debates in his writing -- Cartesianism, Hegelianism, Dilthey's hermeneutics -- and his style of argumentation also feels philosophical. (I never read Mead during my training as a philosopher, though the pragmatist spirit surely infused the Harvard philosophy department, with its intellectual affiliations to James and Peirce.)

Let's take a quick tour through some of the topics in Mind, Self, and Society: From the Standpoint of a Social Behaviorist (Works of George Herbert Mead, Vol. 1). The title is entirely descriptive; the core issue is how to characterize the "me" -- the personal, the conscious individual, the intentional actor, and to theorize about how the self is related to the social world. Mead's fundamental view is that the tradition of philosophy has gotten the relationship backwards; philosophers have built the social from the individual, but actually the self is in some important way the sum of its social relations.

The difference between the social and the individual theories of the development of mind, self, and the social process of experience or behavior is analogous to the difference between the evolutionary and the contract theories of the state as held in the past by both rationalists and empiricists. The latter theory takes individuals and their individual experiencing—individual minds and selves—as logically prior to the social process in which they are involved, and explains the existence of that social process in terms of them; whereas the former takes the social process of experience or behavior as logically prior to the individuals and their individual experiencing which are involved in it, and explains the existence in terms of that social process. (222).

Mead favors the "social first" approach. This doesn't rest on some kind of spooky Durkheimianism about irreducible social wholes, but rather the point that individuals always take shape within the ambit of a set of social relationships, language practices, and normative cues.

Our contention is that mind can never find expression, and could never have come into existence at all, except in terms of a social environment; that an organized set or pattern of social relations and interactions (especially those of communication by means of gestures functioning as significant symbols and thus creating a universe of discourse) is necessarily presupposed by it and involved in its nature. (222)

Mead's theory postulates that the self is built up out of imitative practices, gestures, and conversations over time. The individual forms a reflective conception of his / her self that derives from example and engagement with specific other actors within his / her social space. Here is how he puts his theoretical stance in the first few pages:

I have been presenting the self and the mind in terms of a social process, as the importation of the conversation of gestures into the conduct of the individual organism, so that the individual organism takes these organized attitudes of the others called out by its own attitude, in the form of its gestures, and in reacting to that response calls out other organized attitudes in the others in the community to which the individual belongs. This process can be characterized in a certain sense in terms of the “I” and the “me,” the “me” being that group of organized attitudes to which the individual responds as an "I". (185)

One thing that makes Mead's position here more distinctive is the way that it fits into his broader theory of symbolic manipulation. His ideas about rationality rotate around the human being's ability to use and manipulate symbols. This is what reflective thought involves, according to Mead: to assign symbols to features of he world, and then to choose actions based on reasoning about the relationships among those symbols.

Here is another clear statement about the self and the social:

The mind is simply the interplay of such gestures in the form of significant symbols. We must remember that the gesture is there only in its relationship to the response, to the attitude. (188)

This insistence on the primacy of social relationships for defining the self might imply a problem for the first human self; but actually the development of sociality presumably parallels exactly the development of language and action. We aren't forced to begin in a social contract, state of nature point of view.

So what do action and intention look like on Mead's approach? He asks the question, what role does thought play in action? He concludes that it does play a role; but that the role is not entirely inside the head. His example turns the rational actor model on its head. Rather than deriving outcomes from the bare calculating actor, he understands the actor's deliberations in terms of the values and attitudes of his/her social environment. Speaking of a hypothetical policy maker who identifies strongly with his/her community, he writes:

He is successful to the degree that the final “me” reflects the attitude of all in the community. What I am pointing out is that what occurs takes place not simply in his own mind, but rather that his mind is the expression in his own conduct of this social situation, this great co-operative community process which is going on. (187)

And here is a nice description of purposive action.

In the type of temporary inhibition of action which signifies thinking, or in which reflection arises, we have presented in the experience of the individual, tentatively and in advance and for his selection among them, the different possibilities or alternatives of future action open to him within the given social situation—the different or alternative ways of completing the given social act wherein he is implicated, or which he has already initiated. (90)

This is a reasonable statement of the situation of purposive deliberation: the person entertains in thought the various behaviors he/she can undertake and the possible consequences of those behaviors. The person then chooses a behavior in consideration of which of those consequences is most favored. So this passage conforms loosely to the desire-belief-outcome model and provides an explication of an aspect of consciousness and reflexivity. What is perhaps somewhat more surprising, however, is that Mead's position here seems mildly inconsistent with the earlier expressed ideas of the self as a reflection of the social world in which the biological individual abides. This passage suggests more of a traditional individual-rationality approach to action.

Another common thread in Mead's various discussions of action and behavior is his use of the idea of "habit".  Mead places "habit" as an alternative to "intelligent conduct":

It is the entrance of the alternative possibilities of future response into the determination of present conduct in any given environmental situation, and their operation, through the mechanism of the central nervous system, as part of the factors or conditions determining present behavior, which decisively contrasts intelligent conduct or behavior with reflex, instinctive, and habitual conduct or behavior—delayed reaction with immediate reaction. That which takes place in present organic behavior is always in some sense an emergent from the past, and never could have been precisely predicted in advance—never could have been predicted on the basis of a knowledge, however complete, of the past, and of the conditions in the past which are relevant to its emergence; and in the case of organic behavior which is intelligently controlled, this element of spontaneity is especially prominent by virtue of the present influence exercised over such behavior by the possible future results or consequences which it may have. (98)

He turns to the concept of habit to explain language and to describe ordinary actions in the world.

One of the most interesting currents in sociology today is the new pragmatism -- I'm thinking of work by Neil Gross and Hans Joas in particular.  Several earlier posts have focused on their efforts to provide a new theory of the actor that draws upon pragmatism (link, link).  Mead's theory of the self provides some of the intellectual foundations of this approach; but it doesn't tell the whole story.  In particular, the provocative ideas that are foundational in the new pragmatism --

• "focus on the action rather than the actor", 
• "action is a flow of improvisational adaptations", and 
• "action is relational rather than individual" 

-- seem not to originate in Mead.  Mead's central contributions (in a Twitter-sized bite) seems to be that the self is constituted and created by its social context; and there is a large component of "habit" in ordinary social action.

Are mechanisms complex?

Source: D. Little, "Causal explanation in the social sciences" (link)

How can we distinguish between causal mechanisms and extended causal processes? Is the difference merely a pragmatic one, or is there some reason to expect that mechanisms should be compact and unitary in their workings? Is the children's story leading from the "want of the nail" to the loss of the kingdom a description of an extended mechanism or a contingent causal process?

My preferred definition of a social causal mechanism runs along these lines (“Causal Mechanisms in the Social Realm” (link)):

A causal mechanism is (i) a particular configuration of conditions and processes that (ii) always or normally leads from one set of conditions to an outcome (iii) through the properties and powers of the events and entities in the domain of concern. 

This captures the core idea presented in the Machamer-Darden-Craver (MDC) definition of a causal mechanism (link):

Mechanisms are entities and activities organized such that they are productive of regular changes from start or set-up to finish or termination conditions. (3)

There is also an ontological side of the concept of a mechanism -- the idea that there is a substrate that makes the mechanism work. By referring to a nexus between I and O as a "mechanism" we presume that there is some underlying ontology that makes the observed regularity a "necessary" one: given how the world works, the input I brings about events that lead to output O. In evolutionary biology it is the specifics of an ecology conjoined with natural selection. In the social world it is the empirical situation of the actor and the social and natural environment in which he/she acts.

So mechanisms reflect regularities of input and output. In this respect they correspond to pocket-sized social regularities: observed and sometimes theoretically grounded conveyances from one set of circumstances to another set of circumstances.  Take free riding as a mechanism arising within circumstances of collective action:

When a group of individuals confront a potential gain in public goods that can be attained only through effective and non-enforcible collective action, enough individuals will choose to be free riders to ensure the good is not achieved at the level desired by all members of the group.

This states a regularity (conditioned by ceteris paribus clauses): groups of independent individuals are commonly incapable of effective collective action. And it is grounded in a theory of the actor; rational individuals who pay attention to private costs and benefits but not public costs and benefits can be predicted to engage in free riding.

Now consider the mechanism described in social psychology as "stereotype threat" (link):

When subjects are exposed to signs of negative stereotypes of their group with respect to a given kind of performance, the average performance of the group declines.

This is a mechanism that can be identified in a number of different settings, both observational and experimental; and it can be combined with other mechanisms to bring about complex results. The substrate here is a set of hypothetical cognitive structures through which individuals process tasks and influence each other. 

Now consider an instance of concatenation. Suppose we are interested in military mistakes -- weighty decisions that look in hindsight to be surprisingly poor given the facts available to the decision makers at the time. Our theory of the case may involve three separate mechanisms that interfere with good reasoning: stereotype threat, inordinate hierarchicalism, and the effects of agenda setting. These are independent social cognitive mechanisms that impair group decision making. And our theory of the case may attempt to document the workings of each on the eventual outcome and the ways they aggregated to the observed decision. 

Are mechanisms thought to be simple, or can we consider composite mechanisms -- mechanisms composed of two or more simpler mechanisms? Our definition above required that a mechanism should link I to O with a sufficiently high probability to count as "likely". This puts a practical limit on the degree to which simple mechanisms can be composed into composite mechanisms. Take the sequential case: iMj (prob=.90) and jNk (prob=.90). Then let V be the sequential composite mechanism "M then N". Then we have iVk (prob=.81). The probability of the final endstate given the initial starting condition drops with each additional mechanism that we insert into the composite mechanism.  So eventually concatenation will bring the probability of an antecedent leading to a consequence below the threshold of likelihood required by the definition of a mechanism. 

McAdam, Tarrow and Tilly refer to a concatenation of mechanisms in a concrete instance as a process, not a higher-level mechanism.  The reason for this, it would seem, is that processes are highly contingent in their workings precisely because they incorporate multiple mechanisms in series and parallel, all of whose causal properties are probabilistic.  So there is no reason to expect that processes describe reliable associations between beginnings and endings.

So this implies that mechanisms should be conceived at a fairly low level of compositionality: to preserve the likelihood of association between antecedent and consequent, we need to identify fairly proximate mechanisms with predictable effects.  This doesn't mean that a mechanism has little or no internal structure; rather, it implies that the internal structure of a mechanism fits together in such a way as to bring about a strong correlation between cause and effect.  The mechanism of stereotype threat mentioned above presumably corresponds to a complex set of functionings within the human cognitive system. The net effect, however, is a strong correlation between cause (expressing a stereotype about performance to an individual) and effect (suppressing the level of performance of the individual).