Non-action in times of catastrophe

Ivan Ermakoff's 2008 book Ruling Oneself Out: A Theory of Collective Abdications is dense, rigorous, and important. It treats two historical episodes in close detail -- the passing of Hitler’s enabling bill by the German Reichstag in the Kroll Opera House in March 1933 (“Law for the Relief of the People and of the Reich”) and the decision by the National Assembly of the French Third Republic in the Grand Casino of Vichy to transfer constitutional authority to Marshal Pétain in July 1940. These legislative actions were momentous; "the enabling bill granted Hitler the right to legally discard the constitutional framework of the Weimar Republic," and the results in France were similar for Pétain's government. In both events major political parties and groups acquiesced in the creation of authoritarian legislation that predictably led to dictatorship in their countries and repression of their own parties and groups. Given that these two events largely set the terms for the course of the twentieth century, this study is of great importance.

The central sociological category of interest to Ermakoff here is "abdication" -- essentially an active decision by a group not to continue to oppose a social or political process with which it disagrees. Events are made by the actions of the actors, individual and collective. But Ermakoff demonstrates that sometimes events are made by non-action as well -- deliberate choices by actors to cease their activity in resistance to a process of concern.

Abdication is different from surrender. It is surrender that legitimizes one’s surrender. It implies a statement of irrelevance. When the act is collective, the statement is about the group that makes the decision. The group dismisses itself. It surrenders its fate and agrees to do so, thereby justifying its subservience. This broad characterization sets the problem. Why would a group legitimize its own subservience and, in doing so, abdicate its capacity for self-preservation? (xi)

Based on a great deal of archival research as well as an apparently limitless knowledge of the secondary literature, the book sheds great light on the actions and inactions of the individual and collective actors involved in these enormously important episodes of twentieth-century history. As a result it provides a singular contribution to the theories and methods of contentious politics as well as comparative historical sociology.

The book is historical; but even more deeply, it is a sustained contribution to an actor-centered theory of collective behavior. Ermakoff wants to understand, at the level of the actors involved, what were the dynamics of decision making and action that led to abdication by experienced politicians in the face of anti-constitutional demands by Hitler and Pétain. Ermakoff believes that the obvious theories -- coercion, ideological sympathy, and the structure of existing political conflicts -- are inadequate. Instead, he proposes a dynamic theory of belief formation and decision making at the level of the actor through which the political actors arrive at the position they will adopt based on their observations and inferences of the behavior of others with regard to this choice. Individuals retain agency in their choices in momentous circumstances: "Individuals fluctuate because (1) they are concerned about the behavioral stance of those whom they define as peers and (2) they do not know where these peers stand. Individual oscillations are the seismograph of collective perceptions. Their uncertainty fluctuates with the degree of irresolution imputed to the group." Ermakoff wants to understand the dynamic processes through which individuals arrive at a decision -- to abdicate or to remain visibly and actively in opposition to the threatened action -- and how these decisions relate to judgments made by individuals about the likely actions of other actors.

Ermakoff's key theoretical concept is alignment: the idea that the individual actor is seeking to align his or her actions with those of members of a relevant group (what Ermakoff calls a "reference group"). "By alignment I mean the act of making oneself indistinguishable from others. As a collective phenomenon, alignment describes the process whereby the members of a group facing the same decision align their behavior with one another’s." For individuals within a group it is a problem of coordination in circumstances of imperfect knowledge about the intentions of other actors. And Ermakoff observes that alignment can come about through several different kinds of mechanisms (sequential alignment, local knowledge, and tacit coordination), leading to substantially different dynamics of collective behavior.

Jon Elster and other researchers in the field of contentious politics and collective action refer to this kind of situation as an "assurance game" (link): an opportunity for collective action which a significant number of affected individuals would join if they were confident that sufficient others would do so as well in order to give the action a reasonable likelihood of success.

Though Ermakoff does not directly suggest this possibility, it would appear that the kinds of decision-making processes within groups involved here are amenable to treatment using agent-based models. It would seem straightforward to model the behavior of a group of actors with different "thresholds" and different ways of gaining information about the likely behavior of other actors, and then aggregating their choices through an iterative computational process.

Much of the substance of the book goes into evaluating three common explanations of acquiescence: coercion, miscalculation, and ideological collusion. And Ermakoff argues that the only way to evaluate these hypotheses is to gain quite a bit of evidence about the basis of decision-making for many of the actors. A meso-level analysis won't distinguish the hypotheses; we need to connect the dots for individual decision-makers. Did they defer because of coercion? Because of miscalculation? Or perhaps because they were not so adamantly opposed to the fascist ideology as they professed? But significantly, Ermakoff finds that individual-level information fails to support any of these three factors as being decisive.

Part III moves from description of the cases to an effort at formulating a formal decision model that would serve to explain the processes of alignment and abdication described in the first half of the book. This part of the book has something in common with formal research in game theory and the foundations of collective action theory. Ermakoff undertakes to provide an abstract mid-level description of the processes and mechanisms through which individuals arrive at a decision about a collective action, illustrating some of the parameters and mechanisms that are central for the emergence of abdication as a coordination solution. This part of the book is a substantial addition to the literature on the theory of collective action and mobilization. It falls within the domain of theories of the mechanics of contentious politics along the lines of McAdam, Tarrow, and Tilly's Dynamics of Contention; but it differs from the treatment offered by McAdam, Tarrow, and Tilly by moving closer to the mechanics of the individual actor. The level of analysis is closer to that offered by Mancur Olsen, Russell Hardin, or Jon Elster in describing the logic of collective action.

Consider the logic of the "abdication game" that Ermakoff presents (47):

The "authoritarian challenger" is the leader who wishes to extend his powers beyond what is currently constitutionally permissible. The "target actors" are the groups and parties who have a say in modification of the constitution and legislative framework, and in this scenario these actors are assumed to have a blocking power in the legislative or constitutional process. If they remain unified in opposition the constitutional demands will be refused and either the status quo or an attempt at an unconstitutional seizure of power will occur. If they acquiesce, the authoritarian challenger will immediately undertake a transformation of the state that gives him unlimited executive power.

There is a difficult and important question that arises from reading Ermakoff's book. It is the question of our own politics in 2020. We have a president who has open contempt for law and political morality, who does not even pretend to represent all the people or to respect the rights of all of us; and who is entirely willing to call upon the darkest motivations of his followers. And we have a party of the right that has abandoned even the pretense of maintaining integrity, independent moral judgment, and a willingness to call the president to account for his misdeeds. How different is that environment from that of 1933 in Berlin? Is the current refusal of the Republican Party to honestly judge the president's behavior anything other than an act of abdication -- shameful, abject, and self-interested abdication?

It seems quite possible that the dilemmas created by authoritarian demands and less-than-determined defenders of constitutional principles will be in our future as well. This book was published in 2008, at the beginning of what appeared to be a new epoch in American politics -- more democratic, more progressive, more concerned about ordinary citizens. The topic of abdication would have been very distant from our political discourse. Today as we approach 2020, the threat of an authoritarian, anti-democratic populism has become an everyday reality for American society.

One other aspect of the book bears mention, though only loosely related to the theory of collective action and abdication that is the primary content of the book. Ermakoff's discussion of the challenges that come along with defining "events" is excellent (chapter 1). He correctly observes that an event is a nominal construct, amenable to definition and selection by different observers depending on their theoretical and political interests.

Events are nominal constructs. Their referents are bundles of actions and decisions that analysts and commentators abstract from the flow of historical time. This abstraction is based on a variety of criteria—temporal contiguity, causal density, and significance for subsequent happenings—routinely mobilized by synthetic judgments about the past. Because events are temporal constructs, their temporal boundaries can never be taken for granted. They take on different values depending on whether we derive these boundaries from the subjective statements left by contemporary actors (Bearman et al. 1999) or construct them in light of an analytical relevance criterion derived from the problem at hand (Sewell 1996, 877).

Ermakoff returns to this theme at the end of the book in chapter 11. The approach here taken towards "events" is indicative of one of the virtues of Ermakoff's book (as well as the work of many of the comparative historical sociologists who have influenced him): respect for the contingency, plasticity, and fluidity of historical processes. We have noted elsewhere (link, link, link) Andrew Abbott's insistence on the fluidity of the social world. There is some of that sensibility in Ermakoff's book as well. None of the processes and sequences that Ermakoff describes are presented as deterministic causal chains; instead, choice and contingency remain part of the story at every level.

(Levitsky and Ziblatt's How Democracies Die provides a stark companion piece for Ermakoff's historical treatment of the ascendancy of authoritarianism.)

Graphing the English-speaking university curriculum

Here is a fascinating and ambitious "big data" project that aims at probing and mapping the structure of the disaggregated university curriculum in the United States and other English-speaking countries. The project is called the Open Syllabus Project and is hosted at Columbia University with a team including Joe Kraganis (project director), David McClure (Stanford University Library), Dnnis Tenen, Jonathan Stray, Alex Gil, and Ted Byfield, along with others.  The project has collected over a million syllabi that are openly accessible on university websites; it has then created a database of the books and articles included in these syllabi. There are currently 934K items in the database of texts.

The project provides two basic tools. First, it allows users to search the database by title and see its frequency and co-references. It is also possible to filter by field, institution, state, and country; for each title it is possible to find the other titles most commonly associated with it in the filtered field of syllabi.

One immediate interest in this project is to see the top 10 or twenty titles overall and in various fields. Overall The Elements of Style, the Republic, the Communist Manifesto, Biology, and Frankenstein are the top five titles. In sociology the top five include Aristotle's Ethics, JS Mill's Utilitarianism, Plato's Republic, Descartes' Meditations, and Aristotle's Nicomachean Ethics. In politics the top five include Huntington's Clash of Civilizations, Hobbes' Leviathan, Tocqueville's Democracy in America, Aristotle's Politics, and Machiavelli's Prince.

These listings give a somewhat musty feel to the collective curriculum -- the most common titles are "classics" at least 150 years old. (Sam Huntington's Civilizations book and Ken Waltz's Theory of International Politics are the only 20th century works of political science in the politics field top ten. Ranks 11-20 are more contemporary.) This might give the impression that the typical syllabus is dominated by the classics -- not very encouraging for anyone who thinks that new ideas are needed to solve contemporary problems. But that impression is probably a statistical illusion. If the canon is relatively small (perhaps 100 titles) and the more contemporary and topical literature is very large (10,000 titles) then it is likely enough that items in the canon will show up more frequently, if only because there are lots of introductory courses in which those titles are used. There will be a broad spread of the more contemporary titles over the large number of syllabi, with the result that only a few will break through into the top rank.

Here is an example -- my own Varieties of Social Explanation.

VSE appears in 62 syllabi in the database, and it is most commonly paired with Kuhn, Elster, Taylor, Lijphart, Fearon, and Winch. The syllabi including VSE break down by discipline like this: sociology (23), politics (5), economics (5), psychology (4), philosophy (1), other (28). There are 200 titles with which it is paired in at least four syllabi. Some of these would fit handily into a philosophy of social science course; others are more distant.

It is also interesting to use the search tool to gauge the prevalence of various topics within the curriculum. Here is the search for titles that include the word "Power"; quite evidently the topic of power is highly important in the university curriculum. Foucault heads the list, with 1,774 citations. 8th on the list is Steven Lukes' Power: A Radical View. (The high rank for Lukes' book doesn't surprise me; the second-most trafficked post on Understanding Society is an entry on Lukes' theories of power; link.)

Intriguing as the search function is, the most interesting tool currently available through the project is an interactive network graph of the 10,000 most frequently cited titles.

The methodology is not fully explained on the website, but it appears that items are assigned a metric to other titles based on the frequency of cross-reference between them. It results in an intriguing map of knowledge across the galaxy of disciplines. It is possible to zoom in on any region; below is a snapshot within the social science cluster.

It is possible to pick out the classics in the field -- Durkheim, Weber, Mead, Dewey, Goffman, Bourdieu, Freire -- in this cluster of titles. But this should not be equated with importance or influence. One's eye may be drawn to the large central references in this graph -- Thomas Kuhn's The Structure of Scientific Revolutions, for example, and these may appear as stars around which other items revolve. But equally interesting are the titles on the periphery  and in the interstices -- titles which occur with a measurable frequency within the corpus of syllabi but which have relatively fewer affinities with other titles. And in fact many specialists in the various fields would probably find the most innovative works in their discipline on the periphery rather than in the core of these clusters.

This project is a great example of a big-data problem in the sociology of knowledge. The very hard question now is what to make of it; what light does it shed on the structure of university instruction and knowledge conveyance? What analytical questions need to be posed to this data set?

The public sphere

The current issue of Social Science History is devoted to a series of articles in honor of Charles Tilly (link), around the general theme of the "public sphere" (the theme of the Social Science History Association annual meeting in 2007). Tilly was an active presence in the Social Science History Association, and this issue recognizes Tilly's originality and influence.  The volume contains contributions by several distinguished historical sociologists, including Tilly, Andreas Koller, Craig Calhoun, Andrew Abbott, and Elisabeth Clemens.

The concept of the public sphere isn't a subject to which Tilly gave a lot of explicit attention; in fact, there is very little research on the social reality of the public sphere within comparative historical sociology quite generally.  The most directly relevant discussion of some of these topics in Tilly's work probably occurs in his 2007 book, Democracy.  But the topic is ripe for consideration by comparative and historical sociologists, and for this reason the current SSH issue is a welcome start.

Andreas Koller formulates the general research question about the public sphere in his introductory essay in these terms:

Despite its central relevance for the members of modern societies for determining the course of their own history through reasoned debate and public choice, the study of the public sphere is not an integrated research field. ... This introduction seeks to provide an overview of analytic and historical dimensions that enables one to decipher a number of discussions that are spread out over many disciplines and often proceed in multiple disciplinary terminologies.  Such a quest for an integrative framework is a necessary condition for well-defined comparative and historical research. (262-63)

Before a comparative historical sociologist could begin to investigate a phenomenon such as the "public sphere," it is necessary to have a preliminary conception of what we are talking about.  As Koller points out, most discussions of the concept begin with the ur-text in the study of the public sphere: Jurgen Habermas's 1962 book, The Structural Transformation of the Public Sphere: An Inquiry into a Category of Bourgeois Society.  Here are some preliminary descriptions of the public sphere offered by several of the contributors to this special issue of SSH:

Andreas Koller: [The public sphere] refers to a public of speaker(s) and audience that organizes itself and determines its own future by the force of the better argument, and it refers to its object, the public good. (Koller, 263)

 Charles Tilly: No one has so far developed crisp measures of the public sphere's expansion and consolidation in one regime or another.  In that regard, comparative-historical research faces gigantic conceptual, technical, and empirical challenges.  But surely one indicator worth tracing is change in the frequency and character of gatherings in which people make collective claims on others, including public authorities. (Tilly, 292)

Craig Calhoun: It is instructive to situate the idea of the public sphere in this context.  This gives the influential account of Jurgen Habermas its central pathos: the public sphere arises as part of civil society, incorporating adults who have gained maturity and intellectual autonomy in another of its parts, the family.  It is oriented to forming rational-critical opinion on matters of universal interest to citizens, and through this to informing state policy. But it is debased and corrupted when the state-society division collapses amid bureaucratization, organized interest-group politics, and mass society in the twentieth century. (Calhoun, 302)

Craig Calhoun: The notion of the political public sphere centered on the idea that private persons might come together through reasoned communication to consider public issues and inform public policy. (Calhoun, 303)

Andrew Abbott: In this article I take this last as my definition of public spheres: public spheres are symbolic spaces within which a group's normative affairs are discussed in some sense for themselves.... So I shall take public spheres as an empirical possibility while making no detailed claims about their characteristics. (Abbott, 338)

Elisabeth Clemens: At the core of the philosophical legacy of the Enlightenment lies a vision of rational individuals governing themselves through collective deliberation.  By means of critical discourse, self-interested or private individuals reflect on common concerns and discover the nature of the public good, justice, and truth. (Clemens, 374)

What these snippets have in common is the idea of a public consisting of deliberative individuals engaging in debate over policies and legislation, in relation to conflicting ideas of the public good. (There is an evident connection between this definition and Rousseau's theories of the general will; link.) There is the idea here that a collectivity can arrive at a publicly shared conception of its good, through open and public debate.  And, in common with theorists of deliberative democracy, there is the idea that public debate can transform individual citizens' conceptions of themselves and the public good.  So debate is not merely expressive of current opinions and preferences; it is potentially transformative.

These ideas are expressed in the language of political philosophy and the theory of democracy.  But the subject matter becomes an object of study for sociologists when we realize that each aspect of the definition refers to a social reality that is highly variable across time, space, and culture.  So it is an empirical question, to consider to what extent Qing China, pre-revolutionary Iran, or medieval England had social realities that corresponded to any of these categories: the public intellectual, the engaged citizen, public debate, or public policy.  Was there a Qing public? Were matters of common concern debated openly or publicly, or were they decided behind closed doors by an Imperial bureaucracy?  Did subjects of the Qing state regard themselves as public individuals?

Tilly's essay in this volume focuses on one aspect of this sociological topic: to what extent did new forms of public debate and agitation begin to emerge in England in the eighteenth and nineteenth centuries?  He treats this as an empirical matter:

My collaborators and I gathered the evidence to examine how the development of British capitalism, transformation of the British state, and popular political struggle itself shaped changes in the ways that ordinary Britons made collecdtive claims -- changes in their repertoires of contention. (292)

He finds that there was a marked increase in the frequency of contentious gatherings, which he attributes to a rise in mass-based organizations and a mass-based public media.  (Much of this research is presented in his more extensive contribution to Social Movements and Networks: Relational Approaches to Collective Action, "Contentious Connections in Great Britain, Britain, 1828-34", with Lesley Wood.)

So this is one empirical approach that a comparative historical sociologist could take to the problem of trying to assess the scope and growth of the public sphere.  A very different approach is offered by Andrew Abbott in his contribution to the SSH volume, "Pragmatic Sociology and the Public Sphere: The Case of Charles Richmond Henderson."  Abbott moves from meso to micro in this piece.  He too counts things in order to assess the scope of the public sphere.  But in this case, what is counted is appearances in the Chicago Tribune.  Abbott attempts to gauge Charles Richmond Henderson's prominence in the public sphere by counting and classifying Henderson's presence in Chicago's leading newspaper.  How frequently does Henderson's name appear in the Tribune, in comparison to other prominent professors?

Henderson was chaplain and professor of sociology at the University of Chicago between 1892 and 1915, and it turns out that he had a remarkably high level of visibility in the Tribune.  He was a prominent public figure.  Abbott attempts to make sense of the public persona of Anderson through a brief intellectual and professional biography of the man; and he tries to arrive at some judgments about the causes and impact of his prominence.  "Over his quarter century at the University of Chicago, Henderson became one of Chicago's and even America's most visible reform figures" (342).  And much of his prominence was deliberate: Henderson sought out opportunities for bringing his convictions to the attention of a broader public than the university.  Clubs and conferences were a frequent venue; Henderson was deeply interested in bringing his ideas to the public through these venues.  And Abbott makes the important point that Chicago consisted, not of one public, but of an archipelago of publics: business elites, religious communities, immigrant communities, professional groups, ... (351).

Both of these empirical studies of certain aspects of the public sphere are intriguing and engaging.  Taken as a whole, the SSH volume provides a diverse palette of work, and it plainly does no more than scratch the surface of the kinds of sociological research that are suggested by the topic of the public sphere.  Abbott, Calhoun, Tilly, and the other contributors give an intriguing sense of the kinds of investigations that can be undertaken; there is much work to do in this area.  Consider this variety of questions that need to be posed about the public sphere from a comparative sociology point of view:

• How did the public sphere evolve in England between 1600 and 1900?
• How does the public sphere differ in France, Germany, and England?
• Did China have a public sphere in the late Imperial period?
• What are some important differences in repertoire and performance within the public spheres of different countries and periods?
• How do intellectuals participate in the public sphere in different times and places?
• What role does organized public protest play in the public sphere?

(The photo above captures two historical ends of the idea of the public sphere: the polis and the protest. It captures a major protest in Athens following the 2009 financial crisis of the Greek state.)

Spatial patterns in the US

Here are four interesting graphics representing different kinds of activity in the United States.  The top panel represents population concentrations across the United State.  The second image is air traffic across the country, and the third image is internet traffic across the country.  The final image is a photograph of the United States from space at night, showing the concentration of lights across the country.  Basically the images correspond to where people live; where they travel; and where they exchange data.  Unsurprisingly, the maps line up very well.

The most interesting question to consider is not the structure of the networks represented by air travel and internet activity.  The nodes of both air travel and internet traffic line up exactly with the cities and metropolitan areas represented in the population map, and they align well with the concentration of lighted areas in the bottom frame as well.  The patterns of both air travel and internet traffic take the form of a swath extending from the dense eastern corridor of the US (Boston to Washington) to California (San Francisco to San Diego), with Chicago standing out as a significant node in the middle of the country.  This is entirely obvious and predictable; travel and communication follow population density.

Rather, the more interesting question is this: to what extent is there evidence of internet and air traffic at a node that is disproportionately greater than would be expected given the population of that node?  This is the kind of question that drives Saskia Sassen's classification of cities as local, regional, national,  and global (The Global City: New York, London, Tokyo and Global Networks, Linked Cities).  (Here is an earlier post on Sassen's work.)

Fundamentally, a city that originates 10 times the volume of internet traffic compared to other cities of similar size is worth looking at in detail.  Its activity level suggest an exceptional concentration of organizations and people that are unusually integrated into global networks of communication and data exchange.  It is a location for knowledge-intensive activity: high-end services, banking, universities; and it is a place with intensive relations to other nodes.  Likewise, a city that originates 5 times as many air-travel passengers as comparably sized cities elsewhere is likely to be a specialized location for business and high-end service activity (or else a population of very dedicated tourists).  In other words, the most interesting feature that these data sets might show us about the economic roles of US cities is not visible in the graphics presented here.

So it would be very interesting to be able to "divide" the activity levels represented by the middle two graphs by the populations represented by the top graph, so we could see which cities in the US are "super data exchangers" and "super travel generators".  And this would give us an indication of the degree of high-end, knowledge-intensive activity that is concentrated in the place -- thereby providing a measure of its importance in the national and global economy.  It is possible, for example, that Ann Arbor or Madison would show up as spikes of internet activity relative to their relatively small populations; and Raleigh-Durham might show up as a spike of air travel relative to population, reflecting an unusual concentration of high-end service businesses in this region.  The above-average data and air-traffic nodes are perhaps the dynamic centers of 21st-century economic activity.

Likewise, it would be interesting to identify the cities that have lower-than-expected levels of travel or internet activity; this would suggest local economies that are somewhat more self-contained and less integrated into the national economy than other locations.  And it would be interesting to see if there are significant pairings among locations for either kind of transaction; for example, is the volume of data exchange between Los Angeles and New York significantly greater than that between New York and Chicago or Boston?  And would this serve as an indicator of the degree of economic and business integration between these specific locations?

These views of the United States are interesting because they allow us to see the country as an inter-connected system of places and activities. They serve as something like a dynamic CT scan of the brain: certain connections between places "light up", providing an indication of systemic activities that warrant further investigation.

(The middle panel was published in the Harvard Alumni Magazine (link).  The population map comes from Urbanomnibus (link).)

Are social networks fundamental?

There are several natural starting points when we begin thinking seriously about the social world and how it works. For example, we can begin with individual agents and try to understand social patterns as the expression of common features of reasoning and motivation by stylized agents. This is roughly the strategy underway in rational choice theory, neoclassical economics, game theory, and methodological individualism. Or we might begin with an account of group attributes -- race, class, gender, ethnicity, religion. This is roughly the way in which Durkheim, Giddens, and Du Bois begin -- with a kind of macro-social set of categories in terms of which we attempt to understand social structure and behavior.

The concept of a social network doesn't fit neatly into either category. It is larger than a collection of individuals, in that we have to specify a set of relationships among individuals in order to define a social network. But it is much more concrete and agent-based than the super-categories of race, class, or gender turn out to be. So my question here is a fundamental one: Is the concept of a social network one of a very small number of concepts that must be invoked in virtually every kind of social explanation? As such, is the concept of a social network, and the associated concepts of concrete social relationships it brings with it, a fundamental component of any satisfactory social ontology?  And does the concept of a social network define a crucial space between the micro and the macro?  (A good recent effort to link social networks theory to an important area of social science research is Mario Diani and Doug McAdam, Social Movements and Networks: Relational Approaches to Collective Action.)

A couple of points are pretty obvious. One is that social networks do in fact constitute a key causal mechanism underlying many social processes. We can explain important features of social and political life by identifying the concrete social networks that exist within the population: the transmission of ideas, knowledge, and styles through a population; the selection of important leaders in government and industry; the effective reach of the state; the course of mobilization within a community around an important issue; and the effectiveness of a terrorist group, to name a few examples. A second point is that networks have specific features of topology and functioning that have causal consequences that are largely independent from the personal characteristics of the people who constitute it. For example, information may travel more quickly through a network of people containing many midsized nodes than one containing just a few mega-hubs. And this structural fact may suffice to explain some social outcomes: for example, this rebellion succeeded (because of rapid transmission of information) whereas that one petered out (because of ineffective communications).

Consider two very different examples of group behavior: synchronized cheering in a stadium and the spread of boycotts in Alabama in the early 1960s. The first case involves no social network at all. Cheerleaders stationed around the field initiate the chant as the noise moves to their part of the stadium, and many fans respond when called. Fan behavior is explained by the fan's observation of the behavior of other fans and the motions of the cheerleader. The boycotts had a different dynamic. Organizations emerged which set about to mobilize support for the strategy of boycott. Some of this effort took the form of public calls to action. But a larger part of the mobilization occurred through the workings of extended networks of engaged people -- ministers, union activists, student organizations, and civil rights groups. And the effectiveness and pattern of dissemination of the call to action depended critically on the scope and structure of each of these networks of networks -- networks among leaders of diverse organizations and subordinate networks clustered around each leader. (Doug McAdam describes these processes in detail in Political Process and the Development of Black Insurgency, 1930-1970.)

These examples seem to lead to a couple of observations. One is that social networks are not critical for every form of social action. But the exceptions are pretty simple cases of spontaneous coordination. And second, the example of civil rights mobilization illustrates very clearly why we should expect that social networks are usually crucial. The reason is straightforward: almost all social outcomes require a degree of coordination, communication, and mobilization. A social network is not the only way of bringing these factors about -- cheerleaders and television stations can do it too. But the causal importance of social networks is likely to be great in many cases. And for this reason it seems justified to conclude that social networks are in fact fundamental to social explanation.  Likewise, it appears correct to say that they function as bridging mechanisms from micro to macro, in that they help to convey the actions of local agents onto larger social outcomes (and back!).

(Several earlier posts are relevant to this topic: agent-based modeling, transnational protest, ethnic strife.)

Microstructure of strife

Let's work backwards in thinking about sustained inter-group violence, and begin by considering some of the street-level incidents that constitute a period of violence against or between groups. What factors are necessary to the occurrence of inter-group violence in a region? And how can an understanding of these factors contribute to better strategies of conflict reduction and prevention?

I'm thinking here particularly of ethnic and sectarian violence, including examples like these -- periods of violence in Northern Ireland, upsurges of the Intifada, stone-throwing against vehicles of another group, violent ethnic cleansing in Bosnia, violent settler resistance to resettlement in Israel, ultra-orthodox attacks on more secular Jews in Jerusalem, or Hindu violence against Muslim communities in India. A group of teenagers throw rocks at visible members of another religious group. A cell of young men place a fire bomb in a department store in a Protestant area. A mob rages through a Muslim neighborhood, attacking innocent households. A gang of toughs pressures a minority family to move from the majority neighborhood with threats and beatings. What motivates the participants to involve themselves in these violent actions? And what social factors are necessary in order to turn a few violent individuals into a major violent inter-group event?

Quite a few earlier postings are relevant to various aspects of these questions (thread). And, as has been frequently mentioned here, Charles Tilly's theories of contentious politics are crucial here (Dynamics of Contention, Contentious Performances). Here my interest is in line with the idea of promoting peace: if we understand the dynamics of contention better, perhaps we can do a better job of designing institutions and policies that minimize the occurrence of inter-group strife.

We can begin to analyze these examples by providing some analytical questions: what are the contentious social groups?Are acts of violence spontaneous or orchestrated? Are there contentious organizations providing a degree of stimulus and coordination to the violent acts? Are participants "professionals" or ordinary members of civil society? What is the nature of the grievances that motivate typical participants and stimulate the incidents of violence? What role do media play in the etiology of the outburst of violence? (For example, it is now well understood how radio broadcasts were used to spread ethnic killings in Rwanda.)

Studies of contentious politics can perhaps be summarized along the lines of a small handful of causal components:

• motivation and mobilization of followers;
• actions and reach of organizations;
• availability of resources and opportunities; and
• existence of social networks.

Who are the followers and what motivates them? What are the organizations that are working towards mobilization for acts of violence? What are the motivations and tactics of leaders of these actions? The role of political entrepreneurs and their private political interests appear to be important factors in the occurrence of ethnic and religious strife. (Atul Kohli offers this kind of analysis of Hindu violence against Muslims in India in Democracy and Discontent: India's Growing Crisis of Governability.) And finally, what networks of communication and mutual support exist among individuals, leaders, and organizations? (Mario Diani and Doug McAdam have a very interesting recent collection on the latter factor; Social Movements and Networks: Relational Approaches to Collective Action.)

This broad analysis of the components of contention is useful for peace studies because it suggests a number of avenues of strategy and tactics for reducing inter-group violence. Violence requires followers; so reducing the motivations and grievances that ordinary people have to join a violent social group is obviously a positive step. (This pertains to the line of thought expressed in an earlier posting about the relationship between justice and peace.) Violent movements usually require organizations to coordinate and stimulate attacks; so governments and security services can work to disrupt or contain violent organizations. (The multi-decade struggle in the United States against the Ku Klux Klan is an example.) And, symmetrically, people interested in peace can support organizations in the same terrain that reject violence -- thus reducing the appeal of violent organizations. Once the centrality of social networks is recognized in the mechanisms of stimulating, spreading, and escalating violence, security agencies can themselves undertake to map out the networks of violence that exist and disrupt them. And the crucial role that resources play in violent mobilization -- access to funds, weapons, or media, for example -- suggests a strategy of resource denial to the forces of order. The state and other agencies can work to reduce the availability of necessary resources to violent organizations.

It seems apparent that if we are to succeed in reducing social conflict and violence, we need to have a good understanding of the social mechanisms through which these conditions arise. And fortunately, there is a very rich literature on social contention that can be incorporated into the study of the structural conditions of peace.

Social entrepreneurs

Social entrepreneurs are people who want to bring about non-routine projects, collaborations, or organizations where they didn't previously exist in order to solve a perceived social problem. This is very different from working within an existing organization and using its official resources to bring about a particular result. An example might be a community activist who conceives of a storefront operation providing financial advice to local homeowners facing foreclosure. In order to bring this about, he or she needs to secure financial commitments from some potential partners, recruit expert volunteers from others, and gain cooperation from community neighbors to trust and use the service. And, finally, the project needs to deliver a reasonable percentage of its promised benefits -- some number of troubled mortgage holders need to be successful in retaining their homes thanks to the service.

There are examples of highly successful social entrepreneurs everywhere -- think of Geoffrey Canada and the creation of the Harlem Children's Zone (link), Matt and Jessica Flannery and the creation of Kiva, a successful peer-to-peer micro-lending program (link), Father William Cunningham and Eleanor Josaitis and the founding of FocusHope in Detroit (link), or Millard and Linda Fuller, founders of Habitat for Humanity (link). And we don't have to turn to the high-profile cases to find great examples of skillful social entrepreneurship -- every city is home to community leaders and innovators who have created coalitions or organizations that have significantly contributed to solutions for important social problems.

It is interesting to notice that the idea of social entrepreneurship brings in the social in two different ways. Social entrepreneurs are involved in addressing social problems; so the object of the entrepreneurial activity is to enhance some aspect of the public good or to solve a social problem, rather than achieving a private "return on investment". But second, social entrepreneurship depends on putting together voluntary coalitions among a number of independent social actors in a sustained common effort. A purely private and self-financed effort to provide meals for homeless teenagers would be a laudable effort; but I'm not sure I would call it "social entrepreneurship." It doesn't confront the key problem of locating potential partners, engaging them in the project, and sustaining the collaboration over the long term. So a social entrepreneur involves social cooperation at the input side and public goods on the output side.

The question here is a simple one: what are some of the factors that permit some people to be successful in efforts at social entrepreneurship and others strikingly not? There are also other groups of people -- those who would never consider the idea of creating a new social project in any circumstances and those who look at social causes as a moment for personal gain, for example. But just consider the well-intentioned actors who undertake to put together a successful coalition for the public good -- what distinguishes between those who succeed and those who fail?

A couple of related concepts in the social sciences are immediately relevant to this question -- in particular, the concept of social capital and the idea of a social network. Putting the point very intuitively -- success in a project of social entrepreneurship is more likely if the agent has more social relationships of trust and confidence with other influential players to bring to bear in the effort. And likewise, he/she is more likely to succeed in some rough proportion to the depth and weight of the social networks within which the agent is located. (Robert Putnam's Bowling Alone : The Collapse and Revival of American Community is one place where the idea of social capital is applied to contemporary society.)

Crudely, the individual's capacity to make things happen is multiplied by the strength of the organizations he/she can call upon in support of the effort. This point segues to the point about social networks: if the agent is in a position to call upon the support and collaboration of other resource holders as a result of personal and professional relationships, then he/she is more likely to succeed in putting into place the elements needed for establishment of the project.

Let's apply these points to our community activist. Suppose this individual has developed strong relationships with regional foundations, businesses, semi-governmental organizations, and elected officials. And suppose that these relationships have resulted in a deep reservoir of trust in the activist's perseverance and integrity. Finally, suppose the agent has a well-developed plan for establishing and maintaining the center with high credibility of success. He/she has a good "business plan" for the social project.

It is believable that this actor is much more likely to succeed than his twin with very weak personal and organizational relationships. Our organizer will have a good chance of persuading these potential partners to make the financial commitments and commitments of engagement needed to put the financial counseling center in place. And if the center is successful over several years, it is also likely that the agent's fund of trust and confidence will deepen -- making the next project even more achievable.

This part of the story concerns the social location of the actor -- the networks and organizations the agent can turn to in seeking support for the project. What are the features of personality and character that support and deepen these elements of social capital?

Here are several. Trust and credibility are essential. This is true because the success of the enterprise requires the continuing cooperation of voluntary partners. The agent needs to inspire trust in the partners. And inspiring trust isn't just a fact about current behavior or a persuasive smile -- the agent needs to have a track record of behavior that validates that trust as well.

Second, credible competence is crucial as well. The agent needs to be able to make the case that the project can succeed and that he/she has the skills needed to carry it out. This too depends heavily on one's history -- the best evidence of future competence is demonstrated past competence.

This also implies a third characteristic -- a need to be able to communicate clearly and credibly. Simply being a good communicator isn't enough, but failing to communicate well certainly weakens the likelihood of success in bringing together a durable partnership.

And we can't overlook the characteristics of originality and imagination. Solving difficult social problems requires new ideas and innovative approaches. So the successful social entrepreneur probably needs to have a focused imagination that permits him/her to conceive of new potential solutions -- or at least recognize them when other people invent them.

One wonders, finally, if there is a dimension of empathy and emotional intelligence that plays an important differentiating role among different social actors in this kind of activity. How important is it for the social entrepreneur to be able to understand and express the human importance of the issue? And how important are these skills in sustaining coalitions once they have begun? It seems believable that this is very important indeed.

It appears that social entrepreneurship is ready for some serious academic study, along the lines of studies of private entrepreneurship over the past several decades. Here is a report of a preliminary study of Ashoka fellows in social entrepreneurship (link). And the Center for Social Innovation at Stanford Graduate School of Business offers some interesting examples as well (link).

The reality of society

We sometimes speak of "global society", we refer to "French society"; and we also think of face-to-face organizations and neighborhoods as small societies or social groups. There is an important conceptual point in the background in these common ways of speaking: what are the features of interaction or relationship that must obtain in order for a group of people to constitute a "society" or a "social group"?

There are a couple of points that are fairly obvious. These ensembles of individuals are not social groups:

• all the people in the state whose last name begins with "J"
• all the people in the world who enjoy spicy food
• all the people in the world
• the set of people who live within 100 miles of their state or provincial capital city
  

We would probably say that these aren't social groups or societies for several reasons:

• these ensembles bring together very heterogeneous and disassociated individuals
• these individuals don't interact significantly and persistently with each other
  
• the individuals in each case lack a common identity
• the individuals in these groups do not share a single set of values or mores
  
• the populations described here do not possess a dense set of social networks that link almost all members of the group together
• there is not a set of social structures that serve to coordinate and orient the behavior of all or most of the members of these ensembles
  

The fundamental point is that it would seem that the members of a society, as opposed to a random assembly of individuals, must have some strands of connection with each other.

So we might try this out: a society is a set of individuals --

• who share a broad identity with each other, in at least the minimal sense that they regard themselves as members of the same society.
  
• who share some set of values and ideas -- perhaps non-uniform but overlapping
• who are related to each other through economic, political, or social interactions and networks of connections
  
• who are subject to a common set of social institutions.

But these criteria are debatable. Does the first criterion above threaten to rule out Canada and Spain, because there are Quebecois and Basque separatist groups within these countries? Are the people who choose to live in the isolated compound of the Yearning for Zion ranch a part of United States society, given their extreme efforts to avoid any relationships with the larger society? Is a Facebook group of "friends" a society, given that the members are generally geographically and socially dispersed?

Most fundamentally, the criteria for defining an assembly of people as a "society" can't be too restrictive because a "society" is a looser assembly than some other kinds of social groupings -- religious organizations, social movements, or labor unions, for example. In each of these latter instances there is a high degree of coherence, solidarity, and shared identity and values across members of the group. Societies, on the other hand, embody diversity and difference across persons: multiple values, multiple social networks, multiple group identities. So somehow our definition of society needs to fall intermediate between the random assemblages of persons listed first, and the intentional communities mentioned above.

We might say, then, that a society is knit together by only an overlapping but non-comprehensive set of relationships, values, and identities. Individuals share values and identities with some other individuals; this defines one aspect of the "social-connectedness" graph of a society. And individuals interact with other individuals through economic, political, or cultural transactions; this defines another aspect of a social-connectedness graph. Everyone in a society is related through a set of network relationships to many other people in society; but there is no set of network relationships that encompasses everyone. And I suppose that it is possible that, when we have drawn out a massively complex graph of networks and relationships within the population, that there may be some groups that exist in "islands" within the larger social map, with relationships with each other but not with outsiders.

Are there discrete social mechanisms?

McAdam, Tarrow and Tilly direct our attention to the level of the concrete social mechanisms that recur in many instances of social contention (Dynamics of Contention). They specifically refer to escalation, radicalization, brokerage, and repression as examples of social mechanisms that produce the same effects in the same circumstances, and that concatenate into historical processes and events. To this list I would add my own examples -- free-rider problems, norm diffusion, and communications networks.

I wholeheartedly endorse the idea that social explanations need to proceed on the basis of an analysis of underlying social mechanisms. But can this program be carried out in a Mendeleev sort of way -- try to discover a "table of elements" of causal mechanisms that aggregate into "molecules" of social contention?

The closer I look at the argument, the more concerned I become about the discreteness and elementality of the items MTT offer as examples. Take brokerage -- isn't this really an umbrella term that encompasses a number of different kinds of negotiation and alliance-formation? So brokerage isn't analogous to "expansion of ice during freezing" -- a clear example of a physical causal mechanism that is homogeneous across physical settings. Brokerage is rather a "family-resemblance" term that captures a number of different instances of collective behavior and agency.

If we find this line of thought somewhat persuasive, it suggests that we need to locate the causal connectedness among social settings at an even deeper micro-level. It is the situation of "agents with interests, identities, networks, allies, and repertoires" that constitutes the causal nexus of social causation on contention -- not a set of frozen mid-level groups of behaviors such as brokerage or radicalization. Instead, these mid-level concepts are descriptive terms that allow us to single out some broadly similar components of social contention.

Or in another vocabulary: the level at which we find real causal connections in the social world is the level of the socially situated and socially constituted individual in interaction with other individuals -- the perspective of methodological localism (Levels of the Social). This doesn't undermine causal realism -- but it does undermine the idea that there are meso-level "causal mechanisms" such as brokerage that really recur across instances.

New forms of collective behavior?

Personal electronic communication and the Internet -- have these new technologies changed the game for collective action? Here I am thinking of email and instant messaging, but also cell phones and other personal communications devices, as well as the powerful capacity for dissemination of ideas over the web -- has this dense new network of communication and coordination fundamentally changed the ability of groups to pursue their political or social goals?

There is no doubt that these technologies are relevant to collective action. Communication, coordination, and assurance are crucial features of successful collective action -- and these are precisely the qualities that current technologies offer. Moreover, the ability for a party or movement to disseminate its programs, ideas, and promises to potential followers is crucial for its ability to gather support; and this is what the Web offers better than any prior form of communications technology.

A couple of data points are relevant.

• The City of New York has recently subpoenaed the software and records of TXTmob from an MIT graduate student (story). TXTmob is a software tool created more or less on the fly before the party conventions in 2004 to permit demonstrators to use text messages to assemble and disperse quickly and effectively.
• Will.i.am's music video of Barak Obama's "Yes We Can" speeches has been viewed by eight million people since posting on YouTube -- generating funds, votes, and passion for the candidate.
• Cell phone photos and videos have made their way out of Tibet and Burma documenting the crackdowns that have occurred in those places -- allowing passionate groups of people outside the area to bring their protests to bear.

So what is genuinely new in this list? Covert cameras and travelers have existed for a long time. Cell phones were available in Gdansk and Teheran during street protests there in the 1970s. And newspapers, magazines, and television and radio have disseminated ideas widely. So, again, is there any reason to think that current communications technologies have changed anything fundamental -- either the nature of popular mobilization or the balance of power between the powerful and the numerous?

Two factors are important enough to significantly change the nature of struggles between the powerful and the popular. First is the capacity for coordination among a large group that is created by cell phones and IM devices. A "flash mob" can form and dissolve in minutes. This can make their actions and demonstrations more effective and more difficult to repress. And there is a secondary benefit for the organization -- rapid multi-sided communication can help to maintain solidarity and commitment within the group.

Second, the low cost and broad distribution of web-based communication gives a new advantage to the numerous but poor. Swift Boaters required hundreds of thousands of dollars to disseminate their attack ads against candidate Kerry -- whereas a six-minute video can reach millions of people on YouTube for free. This tips the balance of power away from the deep pockets towards the creative activist group.

So it seems reasonable to judge that these communications technologies are indeed a significant new element in the field of play of collective action. Groups can self-organize more effectively; they can coordinate their actions; and they can share and reinforce the urgency of their commitments through the use of cell phones, text messages, web pages, and dissemination points such as YouTube.

All this has implications for popular politics within a law-governed democracy. It is less clear that these technologies offer as much leverage for the powerless within an authoritarian state. Combine a powerful authoritarian state's ability to monitor communications with a perfect readiness to repress activists and dissidents and to control the technology -- and you get a situation in which these tools of communication are much less useful for an opposition.

Chaos and coordination in social life

Much social behavior is chaotic, in that it simply emerges from the independent choices of numerous agents during a period of time. It is analogous to Brownian motion -- particles in a liquid moving in random motions as a result of innumerable bumps and pushes at the molecular level.

However, there are also many patterns that become visible in social behavior -- examples of what I would like to call "coordinated social action": stock market panic selling, holiday travel, rumors, style, riots, pickpockets in train stations. And we can identify many causes of coordination of individual behavior into larger patterns: commands, regulations, institutions, customs, conventions, collective plans, shared beliefs about social behavior, common sources of information, and common changes in the environment of choice, for example.

What I mean by "coordination" here is the opposite of chaos -- something analogous to the coherence of photons associated with the laser effect. In a laser a set of photons are stimulated to fire coherently with each other, resulting in a beam of light that possesses focus and parallel propagation that is different in kind from the scattered diffusion of photons from an incandescent bulb. "Coordinated" social action is a set of actions that possess synchronicity or regularity in their occurrence, resulting in an observable regularity of behavior over time and space. A crucial problem for social inquiry is to provide an explanation of the mechanisms that underlie the instances of coordinated social action that we can identify.

Examples of coordinated social action can easily be offered, and specific mechanisms can be identified that produce these forms of social coordination. An army moves in concert across a landscape (command). People drive on the right in North America (regulation). People send their children to school (institution). People greet each other with a polite "good morning" (custom). Villagers come together to fish as a group in the morning (convention). People discuss a spontaneous demonstration in front of the mayor's office on Wednesday, and many appear (collective plans). Drivers choose Route 3 rather than Route 1 because they expect a lot of traffic on Route 1 (shared beliefs). People buy a large number of batteries and chocolate, anticipating an approaching hurricane (common environmental change).

These are all mechanisms that create a degree of coordination or synchronization of behavior among independent agents. There seem to be several large categories of mechanisms here: hierarchical coordination (command, regulation); common response coordination (each individually responds to the same signal); communications and network coordination (individuals exchange messages to secure coordination); and strategic coordination (each intends to behave in a way that will be desirable given his/her expectation of actions by others). Might we try out the thought that all forms of regularities of social behavior derive from one or another of these forms of coordination? This thought is probably somewhat too strong a claim. For example, there are probably social regularities that derive from our biology and evolutionary histories -- limitations of memory, bonds of intra-group loyalty, kin altruism. But the impulse is a sound one: when we are able to observe patterns of social behavior, there must be a cause of those regularities that works its way through influence on the individual actors who constitute the domain of action. And there are only so many mechanisms that might serve.

These sorts of regularities and mechanisms constitute part of the regularity of social life, but perhaps only part. It may be that they don't capture other kinds of more "structural" regularities -- for example, "racial discrimination increases health disparities," "feudal political systems are slow to respond to external aggression", "capitalist market systems are more innovaative than planned economies". But there is an important aspect of social explanation that centers exactly on this question: what are the social mechanisms that bring a degree of coherence and coordination among the actions of a population of independent actors?

Is network analysis inconsistent with agent-centered explanation?

Quite a few researchers who study dynamic social processes are making use of some of the tools of network analysis. And it is sometimes maintained that this approach is inconsistent with an agent-centered approach to social processes. Some of these researchers take the view that "it's not what is in the heads of various actors, but rather their relationships in networks that provide the causal underpinnings of social change." And they sometimes maintain that the actor's psychological states can't even be identified in isolation from his/her social relationships. So, once again, explanation cannot rest upon facts about individuals alone. And this sort of finding is thought to cast doubt on methodological individualism in particular, and agent-centered explanatory strategies more generally. (Chuck Tilly and co-authors sometimes take a view along these lines; for example, Doug McAdam, Sidney Tarrow and Charles Tilly, Dynamics of Contention.)

There is something right about the intuition that we can't ground social explanations on assumptions that are too narrowly confined to features of individual psychology. Individuals are socially constucted and socially developed, and our explanations of social processes need
to reflect this fact. This is why I prefer the phrase "methodological localism" to "methodological individualism." But both ontologies are agent-centered. So the question remains: does the causal salience of social networks demonstrate that agent-centered accounts are inherently incomplete -- or even worse, inherently unworkable (because we can't even specify the individual agent's powers and motives independently of his/her networks)?

I don't think so, for several reasons. First, what is a network but a set of socially constructed agents in concrete relations with each other -- communication, coordination, power, subordination, and recognition? The facts about the network are exhausted by a description of the social beliefs of the relevant actors and their material relations to each other.

Second, it is certainly true that an agent's possibilities for exercising power are a function of facts beyond his/her own psychological characteristics. So Albert, the peasant activist in the tiny Breton village, is much more empowered than his psychological twin across the border in Normandy, by the fact that he alone has strong relationships with leaders in both the Catholic Church and the wine-growers' guild. His social networks permit him to amplify the scope of action and effect he may attempt. What this means is that Albert's social networks are a causal component in his ability to wield influence. In this sense it is reasonable as well to attribute causal status to the network and to characterize this standing as being independent of Albert as an individual.

But it remains true that all of the causal powers associated with the network depend on the states of agency of the many persons who make it up. We therefore need to be able to provide an agent-centered account of the network's causal powers, distributed over the many agents who make it up. We must have "microfoundations" for the claim that the network exercises social influence. If the actors who constitute nodes within the network didn't have the right mental frameworks, motivational dispositions, or bodies of knowledge, then they would not in fact behave in a way that was sustaining of the network's social-causal properties.

So, it seems inescapable that, when we say that "Albert's power as a peasant activist depends upon the social fact that he is part of such-and-so networks" -- that we have only uncovered another field of research where more agent-centered research is needed. The network's social-causal properties must themselves disaggregate onto a set of facts about the agents who constitute the network. The current causal properties of the network and the agents who make it up are the complex and iterative result of many inter-related actions and alliances of prior generations of agents.

And this in turn demonstrates that network analysis is by no means inconsistent with an agent-centered approach to social explanation.

(See "Levels of the social" for more on this subject.)

Agent-based modeling as social explanation

Logical positivism favored a theory of scientific explanation that focused on subsumption under general laws. We explain an outcome by identifying one or more general laws, a set of boundary conditions, and a derivation of the outcome from these statements. A second and competing theory of scientific explanation can be called "causal realism." On this approach, we explain an outcome by identifying the causal processes and mechanisms that give rise to it. And we explain a pattern of outcomes by identifying common causal mechanisms that tend to produce outcomes of this sort in circumstances like these. (If we observe that patterns of reciprocity tend to break down as villages become towns, we may identify the causal mechanism at work as the erosion of the face-to-face relationships that are a necessary condition for reciprocity.)

But there are other approaches we might take to social explanation and prediction. And one particularly promising avenue of approach is "agent-based simulation." Here the basic idea is that we want to explain how a certain kind of social process unfolds. We can take our lead from the general insight that social processes depend on microfoundations at the level of socially situated individuals. Social outcomes are the aggregate result of intentional, strategic interactions among large numbers of agents. And we can attempt to implement a computer simulation that represents the decision-making processes and the structural constraints that characterize a large number of interacting agents.

Thomas Schelling's writings give the clearest exposition to the logic of this approach Micromotives and Macrobehavior. Schelling demonstrates in a large number of convincing cases, how we can explain large and complex social outcomes, as the aggregate consequence of behavior by purposive agents pursuing their goals within constraints. He offers a simple model of residential segregation, for example, by modeling the consequences of assuming that blue residents prefer neighborhoods that are at least 50% blue, and red residents prefer neighborhoods at least 25% red. The consequence -- a randomly distributed residential patterns becomes highly segregated in an extended series of iterations of individual moves.

It is possible to model various kinds of social situations by attributing a range of sets of preferences and beliefs across a hypothetical set of agents -- and then run their interactions forward over a period of time. SimCity is a "toy" version of this idea -- what happens when a region is developed by a set of players with a given range of goals and resources? By running the simulation multiple times it is possible to investigate whether there are patterned outcomes that recur across numerous timelines -- or, sometimes, whether there are multiple equilibria that can result, depending on more or less random events early in the simulation.

Robert Axelrod's repeated prisoners' dilemma tournaments represent another such example of agent-based simulations. (Axelrod demonstrates that reciprocity, or tit-for-tat, is the winning strategy for a population of agents who are engaged in a continuing series of prisoners' dilemma games with each other.) The most ambitious examples of this kind of modeling (and predicting and explaining) are to be found in the Santa Fe Institute's research paradigm involving agent-based modeling and the modeling of complex systems. Interdisciplinary researchers at the University of Michigan pursue this approach to explanation at the Center for the Study of Complex Systems. (Mathematician John Casti describes a number of these sorts of experiments and simulations in Would-Be Worlds: How Simulation is Changing the Frontiers of Science and other books.)

This approach to social analysis is profoundly different from the "subsumption under theoretical principles" approach, the covering-law model of explanation. It doesn't work on the assumption that there are laws or governing regularities pertaining to the social outcomes or complex systems at all. Instead, it attempts to derive descriptions of the outcomes as the aggregate result of the purposive and interactive actions of the many individuals who make up the social interaction over time. It is analogous to the simulation of swarms of insects, birds, or fish, in which we attribute very basic "navigational" rules to the individual organisms, and then run forward the behavior of the group as the compound of the interactive decisions made by the individuals. (Here is a brief account of studies of swarming behavior.)

How would this model of the explanation of group behavior be applied to real problems of social explanation? Consider one example: an effort to tease out the relationships between transportation networks and habitation patterns. We might begin with a compact urban population of a certain size. We might then postulate several things:

• The preferences that each individual has concerning housing costs, transportation time and expense, and social and environmental environmental amenities.
• The postulation of a new light rail system extending through the urban center into lightly populated farm land northeast and southwest
• The postulation of a set of prices and amenities associated with possible housing sites throughout the region to a distance of 25 miles
• The postulation of a rate of relocation for urban dwellers and a rate of immigration of new residents

Now run this set of assumptions forward through multiple generations, with individuals choosing location based on their preferences, and observe the patterns of habitation that result.

This description of a simulation of urban-suburban residential distribution over time falls within the field of economic geography. It has a lot in common with the nineteenth-century von Thunen's Isolated State analysis of a city's reach into the farm land surrounding it. (Click here for an interesting description of von Thunen's method written in 1920.) What agent-based modeling adds to the analysis is the ability to use plentiful computational power to run models forward that include thousands of hypothetical agents; and to do this repeatedly so that it is possible to observe whether there are groups of patterns that result in different iterations. The results are then the aggregate consequence of the assumptions we make about large numbers of social agents -- rather than being the expression of some set of general laws about "urbanization".

And, most importantly, some of the results of the agent-based modeling and modeling of complexity performed by scholars associated with the Santa Fe Institute demonstrate the understandable novelty that can emerge from this kind of simulation. So an important theme of novelty and contingency is confirmed by this approach to social analysis.

There are powerful software packages that can provide a platform for implementing agent-based simulations; for example, NetLogo. Here is a screen shot from an implementation called "comsumer behavior" by Yudi Limbar Yasik. The simulation has been configured to allow the user to adjust the parameters of agents' behavior; the software then runs forward in time through a number of iterations. The graphs provide aggregate information about the results.