Current issues in causation research

This week's conference on Causality and Explanation in the Sciences in Ghent was an unusually good academic meeting (link). Participants gathered from all over Europe, as well as a few from North America, Australia, and South Africa, to debate the logic and substance of causal interpretations of the world. Among other things, it provided all participants with a very good sense of the ideas about causation that are generating the most discussion today.

A general perception that emerges from the gestalt of papers at the conference is that there are three large focus areas in current research on scientific causation. First, there is interest in specifying what causal assertions and concepts mean in scientific explanations. What are the logical, conceptual, and pragmatic issues associated with causal assertions and explanations?

Second, there is a large body of work focusing on the methods we can use to support causal inference in the sciences. Every field of science produces volumes of data about variables and events over time. What methods exist to permit inferences about causal relationships among the observed variables and entities? This includes causal modeling statistical methods, but also comparative methods deriving from Mill's methods of difference and similarity.

Third, there is a group of philosophers and scientists who are primarily interested in the ontology of causation in various parts of the sciences. How do various factors exercise causal powers in ecology, the social sciences, or complex systems? Researchers in these areas need provisional answers to questions raised by the first two groups, but their focus is on substantive causal processes rather than the logic of causal statements.

It is useful to inventory half a dozen approaches that were repeatedly cited. This survey is impressionistic but gives an idea of the current landscape.

The mechanisms approach. The idea that we can explicate causation through the idea of a mechanism has been rising in importance over the past twenty years. The idea here is that the fundamental causal concept is that of a mechanism through which X brings about or produces Y. This is argued to be key to causation from single-case studies to large statistical studies suggesting a causal relationship between two or more variables. Peter Hedstrom and other exponents of analytical sociology are recent voices for this approach for the social sciences, though expositions of this approach don't usually go into the level of detail expected by philosophers like Woodward and Cartwright. An important paper by Peter Machamer, Lindley Darden and Carl Craver, "Thinking about Mechanisms", sets the terms of current technical discussions; their view is referred to as the MDC theory. A common concern is that the approach hasn't been as clear as it should be about what precisely a mechanism is. James Mahoney made this criticism in 2001 in "Beyond Correlational Analysis" reviewing Charles Ragin, Fuzzy-Set Social Science and Peter Hedstrom and Richard Swedberg, Social Mechanisms: An Analytical Approach to Social Theory (link), and we still need a more generally recognized specification of the idea. (See an earlier post on this approach; link.)

The manipulability account. Jim Woodward is perhaps the leading exponent of the manipulability (or interventionist) account. He develops his views in detail in his recent book, Making Things Happen: A Theory of Causal Explanation. The view is an intuitively plausible one: causal claims have to do with judgments about how the world would be if we altered certain circumstances. If we observe that the concentration of sulphuric acid is increasing in the atmosphere, we might consider the increasing volume of H2SO4 released by coal power plants from 1960 to 1990. And we might speculate that there is a causal connection between these facts. A counterfactual causal statement holds that: If X (increasing emissions) had not occurred, then Y (increasing acid rain) would not have occurred. The manipulability theory adds this point: if we could remove X from the sequence, then we would alter the value of Y. And this in turn makes good sense of the ways in which we design controlled experiments.

Difference-making. Another strand of thinking about causation focuses on the explanations we are looking for when we ask about the cause of some outcome. Here philosophers note that there are vastly many conditions that are causally necessary for an event but do not count as being explanatory. Lee Harvey Oswald was alive when he fired his rifle in Dallas; but this doesn't play an explanatory role in the assassination of Kennedy. Crudely speaking, we want to know which causal factors were salient; which factors made a difference in the outcome. Michael Strevens provides a detailed and innovative explication of this set of intuitions in his recent book Depth: An Account of Scientific Explanation, where he introduces his theory of "Kairetic" explanation.

Contrastive analysis as a theory of explanation. When we seek an explanation of something, we generally have something specific in mind: why X rather than X'? And an explanation that keys off the wrong contrast will fail, even though its premises are correct. Bas van Fraassen (1980), The Scientific Image, is often cited in this context. A conference participant, Petri Ylikoski, develops a contrastive counterfactual theory in his dissertation (link). This body of work seeks to clarify pragmatic issues concerning explanation, including understanding and explanatory relevance. If we ask for an explanation for why X occurred, we are usually presupposing a question like this:

Why did X occur [rather than Y]?

• Why is John carrying his umbrella [rather than not]?
• Why is John carrying his umbrella [rather than his raincoat]?
• Why is John carrying his umbrella [rather than his assistant Harry]?

These all demand different answers:

• Because he expects rain;
• Because it is too warm for a raincoat;
• Because Harry is carrying three heavy suitcases.

Here is a much-cited review article by Nancy Cartwright on van Fraasen's work (link), and here is a discussion of contrastive explanation by Jonathan Schaffer (link).

Causal modeling theory. This topic refers to the large body of statistical theory devoted to identifying potential causal relationships among observable variables in a large data set. Hubert Blalock is a founder of this approach (Causal Inferences in Nonexperimental Research; 1964) with his statistical models for causal path analysis. (Here is a short account of the history of path analysis in genetics.) Judea Pearl has contributed a great deal to the method of structural equation modeling (SEM) in Causality: Models, Reasoning and Inference and elsewhere. Here is a handbook article in which he explains the method and its causal relevance (link). Pearl maintains a research blog on causality here. Granger causality is a specific technique for assessing causal relationships within time series data: X Granger-causes Y if variations in X and Y together do a better job of predicting Y than variations in Y by itself.

Prior foundations of philosophical theories of causation. Two older discussions of causality also received some notice in these papers: J. L. Mackie on INUS conditions and causal fields (The Cement of the Universe: A Study of Causation) and Wesley Salmon on the causal structure of the world (Scientific Explanation and the Causal Structure of the World).

Nancy Cartwright's "Causation: One Word, Many Things" provides a very good contemporary review of the varieties of approaches that are currently being taken to the idea of causation (link).

Much of the intellectual vitality of this group of philosophers is captured in the major work recently edited by Phyllis McKay Illari, Federica Russo, and John Williamson, Causality in the Sciences. The book contains a very wide range of disciplines and approaches in its treatment of the topic.

Current issues in causation research

This week's conference on Causality and Explanation in the Sciences in Ghent was an unusually good academic meeting (link). Participants gathered from all over Europe, as well as a few from North America, Australia, and South Africa, to debate the logic and substance of causal interpretations of the world. Among other things, it provided all participants with a very good sense of the ideas about causation that are generating the most discussion today.

A general perception that emerges from the gestalt of papers at the conference is that there are three large focus areas in current research on scientific causation. First, there is interest in specifying what causal assertions and concepts mean in scientific explanations. What are the logical, conceptual, and pragmatic issues associated with causal assertions and explanations?

Second, there is a large body of work focusing on the methods we can use to support causal inference in the sciences. Every field of science produces volumes of data about variables and events over time. What methods exist to permit inferences about causal relationships among the observed variables and entities? This includes causal modeling statistical methods, but also comparative methods deriving from Mill's methods of difference and similarity.

Third, there is a group of philosophers and scientists who are primarily interested in the ontology of causation in various parts of the sciences. How do various factors exercise causal powers in ecology, the social sciences, or complex systems? Researchers in these areas need provisional answers to questions raised by the first two groups, but their focus is on substantive causal processes rather than the logic of causal statements.

It is useful to inventory half a dozen approaches that were repeatedly cited. This survey is impressionistic but gives an idea of the current landscape.

The mechanisms approach. The idea that we can explicate causation through the idea of a mechanism has been rising in importance over the past twenty years. The idea here is that the fundamental causal concept is that of a mechanism through which X brings about or produces Y. This is argued to be key to causation from single-case studies to large statistical studies suggesting a causal relationship between two or more variables. Peter Hedstrom and other exponents of analytical sociology are recent voices for this approach for the social sciences, though expositions of this approach don't usually go into the level of detail expected by philosophers like Woodward and Cartwright. An important paper by Peter Machamer, Lindley Darden and Carl Craver, "Thinking about Mechanisms", sets the terms of current technical discussions; their view is referred to as the MDC theory. A common concern is that the approach hasn't been as clear as it should be about what precisely a mechanism is. James Mahoney made this criticism in 2001 in "Beyond Correlational Analysis" reviewing Charles Ragin, Fuzzy-Set Social Science and Peter Hedstrom and Richard Swedberg, Social Mechanisms: An Analytical Approach to Social Theory (link), and we still need a more generally recognized specification of the idea. (See an earlier post on this approach; link.)

The manipulability account. Jim Woodward is perhaps the leading exponent of the manipulability (or interventionist) account. He develops his views in detail in his recent book, Making Things Happen: A Theory of Causal Explanation. The view is an intuitively plausible one: causal claims have to do with judgments about how the world would be if we altered certain circumstances. If we observe that the concentration of sulphuric acid is increasing in the atmosphere, we might consider the increasing volume of H2SO4 released by coal power plants from 1960 to 1990. And we might speculate that there is a causal connection between these facts. A counterfactual causal statement holds that: If X (increasing emissions) had not occurred, then Y (increasing acid rain) would not have occurred. The manipulability theory adds this point: if we could remove X from the sequence, then we would alter the value of Y. And this in turn makes good sense of the ways in which we design controlled experiments.

Difference-making. Another strand of thinking about causation focuses on the explanations we are looking for when we ask about the cause of some outcome. Here philosophers note that there are vastly many conditions that are causally necessary for an event but do not count as being explanatory. Lee Harvey Oswald was alive when he fired his rifle in Dallas; but this doesn't play an explanatory role in the assassination of Kennedy. Crudely speaking, we want to know which causal factors were salient; which factors made a difference in the outcome. Michael Strevens provides a detailed and innovative explication of this set of intuitions in his recent book Depth: An Account of Scientific Explanation, where he introduces his theory of "Kairetic" explanation.

Contrastive analysis as a theory of explanation. When we seek an explanation of something, we generally have something specific in mind: why X rather than X'? And an explanation that keys off the wrong contrast will fail, even though its premises are correct. Bas van Fraassen (1980), The Scientific Image, is often cited in this context. A conference participant, Petri Ylikoski, develops a contrastive counterfactual theory in his dissertation (link). This body of work seeks to clarify pragmatic issues concerning explanation, including understanding and explanatory relevance. If we ask for an explanation for why X occurred, we are usually presupposing a question like this:

Why did X occur [rather than Y]?

• Why is John carrying his umbrella [rather than not]?
• Why is John carrying his umbrella [rather than his raincoat]?
• Why is John carrying his umbrella [rather than his assistant Harry]?

These all demand different answers:

• Because he expects rain;
• Because it is too warm for a raincoat;
• Because Harry is carrying three heavy suitcases.

Here is a much-cited review article by Nancy Cartwright on van Fraasen's work (link), and here is a discussion of contrastive explanation by Jonathan Schaffer (link).

Causal modeling theory. This topic refers to the large body of statistical theory devoted to identifying potential causal relationships among observable variables in a large data set. Hubert Blalock is a founder of this approach (Causal Inferences in Nonexperimental Research; 1964) with his statistical models for causal path analysis. (Here is a short account of the history of path analysis in genetics.) Judea Pearl has contributed a great deal to the method of structural equation modeling (SEM) in Causality: Models, Reasoning and Inference and elsewhere. Here is a handbook article in which he explains the method and its causal relevance (link). Pearl maintains a research blog on causality here. Granger causality is a specific technique for assessing causal relationships within time series data: X Granger-causes Y if variations in X and Y together do a better job of predicting Y than variations in Y by itself.

Prior foundations of philosophical theories of causation. Two older discussions of causality also received some notice in these papers: J. L. Mackie on INUS conditions and causal fields (The Cement of the Universe: A Study of Causation) and Wesley Salmon on the causal structure of the world (Scientific Explanation and the Causal Structure of the World).

Nancy Cartwright's "Causation: One Word, Many Things" provides a very good contemporary review of the varieties of approaches that are currently being taken to the idea of causation (link).

Much of the intellectual vitality of this group of philosophers is captured in the major work recently edited by Phyllis McKay Illari, Federica Russo, and John Williamson, Causality in the Sciences. The book contains a very wide range of disciplines and approaches in its treatment of the topic.

Current issues in causation research

This week's conference on Causality and Explanation in the Sciences in Ghent was an unusually good academic meeting (link). Participants gathered from all over Europe, as well as a few from North America, Australia, and South Africa, to debate the logic and substance of causal interpretations of the world. Among other things, it provided all participants with a very good sense of the ideas about causation that are generating the most discussion today.

A general perception that emerges from the gestalt of papers at the conference is that there are three large focus areas in current research on scientific causation. First, there is interest in specifying what causal assertions and concepts mean in scientific explanations. What are the logical, conceptual, and pragmatic issues associated with causal assertions and explanations?

Second, there is a large body of work focusing on the methods we can use to support causal inference in the sciences. Every field of science produces volumes of data about variables and events over time. What methods exist to permit inferences about causal relationships among the observed variables and entities? This includes causal modeling statistical methods, but also comparative methods deriving from Mill's methods of difference and similarity.

Third, there is a group of philosophers and scientists who are primarily interested in the ontology of causation in various parts of the sciences. How do various factors exercise causal powers in ecology, the social sciences, or complex systems? Researchers in these areas need provisional answers to questions raised by the first two groups, but their focus is on substantive causal processes rather than the logic of causal statements.

It is useful to inventory half a dozen approaches that were repeatedly cited. This survey is impressionistic but gives an idea of the current landscape.

The mechanisms approach. The idea that we can explicate causation through the idea of a mechanism has been rising in importance over the past twenty years. The idea here is that the fundamental causal concept is that of a mechanism through which X brings about or produces Y. This is argued to be key to causation from single-case studies to large statistical studies suggesting a causal relationship between two or more variables. Peter Hedstrom and other exponents of analytical sociology are recent voices for this approach for the social sciences, though expositions of this approach don't usually go into the level of detail expected by philosophers like Woodward and Cartwright. An important paper by Peter Machamer, Lindley Darden and Carl Craver, "Thinking about Mechanisms", sets the terms of current technical discussions; their view is referred to as the MDC theory. A common concern is that the approach hasn't been as clear as it should be about what precisely a mechanism is. James Mahoney made this criticism in 2001 in "Beyond Correlational Analysis" reviewing Charles Ragin, Fuzzy-Set Social Science and Peter Hedstrom and Richard Swedberg, Social Mechanisms: An Analytical Approach to Social Theory (link), and we still need a more generally recognized specification of the idea. (See an earlier post on this approach; link.)

The manipulability account. Jim Woodward is perhaps the leading exponent of the manipulability (or interventionist) account. He develops his views in detail in his recent book, Making Things Happen: A Theory of Causal Explanation. The view is an intuitively plausible one: causal claims have to do with judgments about how the world would be if we altered certain circumstances. If we observe that the concentration of sulphuric acid is increasing in the atmosphere, we might consider the increasing volume of H2SO4 released by coal power plants from 1960 to 1990. And we might speculate that there is a causal connection between these facts. A counterfactual causal statement holds that: If X (increasing emissions) had not occurred, then Y (increasing acid rain) would not have occurred. The manipulability theory adds this point: if we could remove X from the sequence, then we would alter the value of Y. And this in turn makes good sense of the ways in which we design controlled experiments.

Difference-making. Another strand of thinking about causation focuses on the explanations we are looking for when we ask about the cause of some outcome. Here philosophers note that there are vastly many conditions that are causally necessary for an event but do not count as being explanatory. Lee Harvey Oswald was alive when he fired his rifle in Dallas; but this doesn't play an explanatory role in the assassination of Kennedy. Crudely speaking, we want to know which causal factors were salient; which factors made a difference in the outcome. Michael Strevens provides a detailed and innovative explication of this set of intuitions in his recent book Depth: An Account of Scientific Explanation, where he introduces his theory of "Kairetic" explanation.

Contrastive analysis as a theory of explanation. When we seek an explanation of something, we generally have something specific in mind: why X rather than X'? And an explanation that keys off the wrong contrast will fail, even though its premises are correct. Bas van Fraassen (1980), The Scientific Image, is often cited in this context. A conference participant, Petri Ylikoski, develops a contrastive counterfactual theory in his dissertation (link). This body of work seeks to clarify pragmatic issues concerning explanation, including understanding and explanatory relevance. If we ask for an explanation for why X occurred, we are usually presupposing a question like this:

Why did X occur [rather than Y]?

• Why is John carrying his umbrella [rather than not]?
• Why is John carrying his umbrella [rather than his raincoat]?
• Why is John carrying his umbrella [rather than his assistant Harry]?

These all demand different answers:

• Because he expects rain;
• Because it is too warm for a raincoat;
• Because Harry is carrying three heavy suitcases.

Here is a much-cited review article by Nancy Cartwright on van Fraasen's work (link), and here is a discussion of contrastive explanation by Jonathan Schaffer (link).

Causal modeling theory. This topic refers to the large body of statistical theory devoted to identifying potential causal relationships among observable variables in a large data set. Hubert Blalock is a founder of this approach (Causal Inferences in Nonexperimental Research; 1964) with his statistical models for causal path analysis. (Here is a short account of the history of path analysis in genetics.) Judea Pearl has contributed a great deal to the method of structural equation modeling (SEM) in Causality: Models, Reasoning and Inference and elsewhere. Here is a handbook article in which he explains the method and its causal relevance (link). Pearl maintains a research blog on causality here. Granger causality is a specific technique for assessing causal relationships within time series data: X Granger-causes Y if variations in X and Y together do a better job of predicting Y than variations in Y by itself.

Prior foundations of philosophical theories of causation. Two older discussions of causality also received some notice in these papers: J. L. Mackie on INUS conditions and causal fields (The Cement of the Universe: A Study of Causation) and Wesley Salmon on the causal structure of the world (Scientific Explanation and the Causal Structure of the World).

Nancy Cartwright's "Causation: One Word, Many Things" provides a very good contemporary review of the varieties of approaches that are currently being taken to the idea of causation (link).

Much of the intellectual vitality of this group of philosophers is captured in the major work recently edited by Phyllis McKay Illari, Federica Russo, and John Williamson, Causality in the Sciences. The book contains a very wide range of disciplines and approaches in its treatment of the topic.

Social explanation and causal mechanisms

To explain a social outcome or regularity, we need to provide an account of why and how it came about; and this means providing a causal analysis in terms of which the explanandum appears as a result.

Having a causal theory of a realm requires having an ontology: what kinds of things exist in this realm, and how do they work? Along with others, I offer a social ontology grounded in the actions and relations of socially constituted actors, which I refer to as methodological localism (link). (This is also the ontology asserted by the programme of "analytical sociology";  link.)

This entails, basically, that we need to understand all higher-level social entities and processes as being composed of the activities and thoughts of individual agents at a local level of social interaction; we need to be attentive to the pathways of aggregation through which these local-level activities aggregate to higher-level structures; and we need to pay attention to the iterative ways in which higher-level structures shape and influence individual agents.  Social outcomes are invariably constituted by and brought into being by socially constituted, socially situated individual actors (methodological localism). Both aspects of the view are important. By referring to "social constitution" we are invoking the fact that past social arrangements have created the social actor. By referring to "social situatedness" we invoke the idea that existing social practices and rules constrain and motivate the individual actor. So this view is not reductionist, in the sense of aiming to reduce social outcomes to pre-social individual activity.

We also want to refer to supra-individual actors -- firms, agencies, organizations, social movements, states. The social sciences are radically incomplete without such constructs. But all such references are bound by a requirement of microfoundations: if we attribute intentionality to a firm, we need to be able to sketch out an account of how the individuals of the firm are led to act in ways that lead to the postulated decision-making and action (link).

So, then: what is involved in asserting that social circumstance A causally produces social circumstance B? There are, of course, numerous well developed answers to this question: statistical inference based on correlations of occurrences, conditional probabilities, and necessary-sufficient condition analysis. My view, however, is that there is a more basic meaning of causation: A caused B iff there is a sequence of causal mechanisms leading from A to B. This approach is especially suitable for the social realm because, on the one hand, there are few strong statistical regularities among social outcomes, and on the other, it is feasible to identify social mechanisms through a variety of social research methods -- comparative analysis, process tracing, case studies, and the like.

The social mechanisms approach (and the scientific realism that lies behind it) goes back at least as early as the late 1980s. An early statement of the view was presented in my Varieties Of Social Explanation: An Introduction To The Philosophy Of Social Science in 1991.  Mario Bunge and Jon Elster took similar positions. The view took a large step forward, on the theory side, with the publication of Hedstrom and Swedberg's Social Mechanisms: An Analytical Approach to Social Theory (1998), and on the empirical research side with the publication of McAdam, Tarrow, and Tilly's Dynamics of Contention (2001). There are important differences; theorists within analytical sociology largely favor methodological individualism and mechanisms grounded in rational individuals, whereas Tilly and his colleagues favor "relational" mechanisms. But in each case the model of agent-centered explanations that either require microfoundations or are plainly compatible with such a requirement.  (Here is a recent post on causal mechanisms.)

Several social scientists have anticipated this approach through their own concrete analysis of aggregation phenomena.  A good illustration is Thomas Schelling.  His work presents a large number of examples of mundane social outcomes that he explains on the basis of simple individual-level choices and an aggregation mechanism (Micromotives and Macrobehavior, Choice and Consequence). Features of organized crime, traffic patterns, segregation, and dying seminars all come in for treatment.  Schelling demonstrates in concrete terms what sorts of things we can identify as "social mechanisms" and traces them back to the circumstances of action of individuals in social situations.

The framework of social mechanisms as a basis for social explanation raises an important question about the role and scope of generalizability that we expect from a social explanation. Briefly, the mechanisms identified here show a degree of generalizability; as McAdam, Tarrow and Tilly assert, social mechanisms can be expected to recur in other circumstances and times. But the event itself is one-of-a-kind. This is a familiar feature of Tilly's way of thinking about contentious events as well: the American Civil War was a singular historical event. But a good explanation will invoke mechanisms that recur elsewhere. We shouldn't expect to find general theories of civil wars; but our explanations of particular civil wars can invoke quasi-general theories of mid-level mechanisms of conflict and escalation. (Here is a recent posting on general and specific causal claims.)

Another important methodological question for this approach to social explanation is the issue of explaining general statistical patterns in social life.  What if we want to explain something more quantitative -- say a gradually rising divorce rate or the finding that co-habitants before marriage have higher divorce rates than non-co-habitants? On the social mechanisms approach, we would want two things. First, we would like an agent-level mechanism that explains the statistic; and second, we would like to find a common cause if the phenomenon is similar in several countries.

Finally, the actor-based mechanisms approach invites an area of study which is now being referred to as "aggregation dynamics" (link, link).  We need to have theories and tools that permit us to aggregate different micro-level processes over time into meso- and macro-outcomes, taking into account the complexity of causal interactions in a dynamic process.  The tools of agent-based modeling are relevant here (link).

Social explanation and causal mechanisms

To explain a social outcome or regularity, we need to provide an account of why and how it came about; and this means providing a causal analysis in terms of which the explanandum appears as a result.

Having a causal theory of a realm requires having an ontology: what kinds of things exist in this realm, and how do they work? Along with others, I offer a social ontology grounded in the actions and relations of socially constituted actors, which I refer to as methodological localism (link). (This is also the ontology asserted by the programme of "analytical sociology";  link.)

This entails, basically, that we need to understand all higher-level social entities and processes as being composed of the activities and thoughts of individual agents at a local level of social interaction; we need to be attentive to the pathways of aggregation through which these local-level activities aggregate to higher-level structures; and we need to pay attention to the iterative ways in which higher-level structures shape and influence individual agents.  Social outcomes are invariably constituted by and brought into being by socially constituted, socially situated individual actors (methodological localism). Both aspects of the view are important. By referring to "social constitution" we are invoking the fact that past social arrangements have created the social actor. By referring to "social situatedness" we invoke the idea that existing social practices and rules constrain and motivate the individual actor. So this view is not reductionist, in the sense of aiming to reduce social outcomes to pre-social individual activity.

We also want to refer to supra-individual actors -- firms, agencies, organizations, social movements, states. The social sciences are radically incomplete without such constructs. But all such references are bound by a requirement of microfoundations: if we attribute intentionality to a firm, we need to be able to sketch out an account of how the individuals of the firm are led to act in ways that lead to the postulated decision-making and action (link).

So, then: what is involved in asserting that social circumstance A causally produces social circumstance B? There are, of course, numerous well developed answers to this question: statistical inference based on correlations of occurrences, conditional probabilities, and necessary-sufficient condition analysis. My view, however, is that there is a more basic meaning of causation: A caused B iff there is a sequence of causal mechanisms leading from A to B. This approach is especially suitable for the social realm because, on the one hand, there are few strong statistical regularities among social outcomes, and on the other, it is feasible to identify social mechanisms through a variety of social research methods -- comparative analysis, process tracing, case studies, and the like.

The social mechanisms approach (and the scientific realism that lies behind it) goes back at least as early as the late 1980s. An early statement of the view was presented in my Varieties Of Social Explanation: An Introduction To The Philosophy Of Social Science in 1991.  Mario Bunge and Jon Elster took similar positions. The view took a large step forward, on the theory side, with the publication of Hedstrom and Swedberg's Social Mechanisms: An Analytical Approach to Social Theory (1998), and on the empirical research side with the publication of McAdam, Tarrow, and Tilly's Dynamics of Contention (2001). There are important differences; theorists within analytical sociology largely favor methodological individualism and mechanisms grounded in rational individuals, whereas Tilly and his colleagues favor "relational" mechanisms. But in each case the model of agent-centered explanations that either require microfoundations or are plainly compatible with such a requirement.  (Here is a recent post on causal mechanisms.)

Several social scientists have anticipated this approach through their own concrete analysis of aggregation phenomena.  A good illustration is Thomas Schelling.  His work presents a large number of examples of mundane social outcomes that he explains on the basis of simple individual-level choices and an aggregation mechanism (Micromotives and Macrobehavior, Choice and Consequence). Features of organized crime, traffic patterns, segregation, and dying seminars all come in for treatment.  Schelling demonstrates in concrete terms what sorts of things we can identify as "social mechanisms" and traces them back to the circumstances of action of individuals in social situations.

The framework of social mechanisms as a basis for social explanation raises an important question about the role and scope of generalizability that we expect from a social explanation. Briefly, the mechanisms identified here show a degree of generalizability; as McAdam, Tarrow and Tilly assert, social mechanisms can be expected to recur in other circumstances and times. But the event itself is one-of-a-kind. This is a familiar feature of Tilly's way of thinking about contentious events as well: the American Civil War was a singular historical event. But a good explanation will invoke mechanisms that recur elsewhere. We shouldn't expect to find general theories of civil wars; but our explanations of particular civil wars can invoke quasi-general theories of mid-level mechanisms of conflict and escalation. (Here is a recent posting on general and specific causal claims.)

Another important methodological question for this approach to social explanation is the issue of explaining general statistical patterns in social life.  What if we want to explain something more quantitative -- say a gradually rising divorce rate or the finding that co-habitants before marriage have higher divorce rates than non-co-habitants? On the social mechanisms approach, we would want two things. First, we would like an agent-level mechanism that explains the statistic; and second, we would like to find a common cause if the phenomenon is similar in several countries.

Finally, the actor-based mechanisms approach invites an area of study which is now being referred to as "aggregation dynamics" (link, link).  We need to have theories and tools that permit us to aggregate different micro-level processes over time into meso- and macro-outcomes, taking into account the complexity of causal interactions in a dynamic process.  The tools of agent-based modeling are relevant here (link).

Social explanation and causal mechanisms

To explain a social outcome or regularity, we need to provide an account of why and how it came about; and this means providing a causal analysis in terms of which the explanandum appears as a result.

Having a causal theory of a realm requires having an ontology: what kinds of things exist in this realm, and how do they work? Along with others, I offer a social ontology grounded in the actions and relations of socially constituted actors, which I refer to as methodological localism (link). (This is also the ontology asserted by the programme of "analytical sociology";  link.)

This entails, basically, that we need to understand all higher-level social entities and processes as being composed of the activities and thoughts of individual agents at a local level of social interaction; we need to be attentive to the pathways of aggregation through which these local-level activities aggregate to higher-level structures; and we need to pay attention to the iterative ways in which higher-level structures shape and influence individual agents.  Social outcomes are invariably constituted by and brought into being by socially constituted, socially situated individual actors (methodological localism). Both aspects of the view are important. By referring to "social constitution" we are invoking the fact that past social arrangements have created the social actor. By referring to "social situatedness" we invoke the idea that existing social practices and rules constrain and motivate the individual actor. So this view is not reductionist, in the sense of aiming to reduce social outcomes to pre-social individual activity.

We also want to refer to supra-individual actors -- firms, agencies, organizations, social movements, states. The social sciences are radically incomplete without such constructs. But all such references are bound by a requirement of microfoundations: if we attribute intentionality to a firm, we need to be able to sketch out an account of how the individuals of the firm are led to act in ways that lead to the postulated decision-making and action (link).

So, then: what is involved in asserting that social circumstance A causally produces social circumstance B? There are, of course, numerous well developed answers to this question: statistical inference based on correlations of occurrences, conditional probabilities, and necessary-sufficient condition analysis. My view, however, is that there is a more basic meaning of causation: A caused B iff there is a sequence of causal mechanisms leading from A to B. This approach is especially suitable for the social realm because, on the one hand, there are few strong statistical regularities among social outcomes, and on the other, it is feasible to identify social mechanisms through a variety of social research methods -- comparative analysis, process tracing, case studies, and the like.

The social mechanisms approach (and the scientific realism that lies behind it) goes back at least as early as the late 1980s. An early statement of the view was presented in my Varieties Of Social Explanation: An Introduction To The Philosophy Of Social Science in 1991.  Mario Bunge and Jon Elster took similar positions. The view took a large step forward, on the theory side, with the publication of Hedstrom and Swedberg's Social Mechanisms: An Analytical Approach to Social Theory (1998), and on the empirical research side with the publication of McAdam, Tarrow, and Tilly's Dynamics of Contention (2001). There are important differences; theorists within analytical sociology largely favor methodological individualism and mechanisms grounded in rational individuals, whereas Tilly and his colleagues favor "relational" mechanisms. But in each case the model of agent-centered explanations that either require microfoundations or are plainly compatible with such a requirement.  (Here is a recent post on causal mechanisms.)

Several social scientists have anticipated this approach through their own concrete analysis of aggregation phenomena.  A good illustration is Thomas Schelling.  His work presents a large number of examples of mundane social outcomes that he explains on the basis of simple individual-level choices and an aggregation mechanism (Micromotives and Macrobehavior, Choice and Consequence). Features of organized crime, traffic patterns, segregation, and dying seminars all come in for treatment.  Schelling demonstrates in concrete terms what sorts of things we can identify as "social mechanisms" and traces them back to the circumstances of action of individuals in social situations.

The framework of social mechanisms as a basis for social explanation raises an important question about the role and scope of generalizability that we expect from a social explanation. Briefly, the mechanisms identified here show a degree of generalizability; as McAdam, Tarrow and Tilly assert, social mechanisms can be expected to recur in other circumstances and times. But the event itself is one-of-a-kind. This is a familiar feature of Tilly's way of thinking about contentious events as well: the American Civil War was a singular historical event. But a good explanation will invoke mechanisms that recur elsewhere. We shouldn't expect to find general theories of civil wars; but our explanations of particular civil wars can invoke quasi-general theories of mid-level mechanisms of conflict and escalation. (Here is a recent posting on general and specific causal claims.)

Another important methodological question for this approach to social explanation is the issue of explaining general statistical patterns in social life.  What if we want to explain something more quantitative -- say a gradually rising divorce rate or the finding that co-habitants before marriage have higher divorce rates than non-co-habitants? On the social mechanisms approach, we would want two things. First, we would like an agent-level mechanism that explains the statistic; and second, we would like to find a common cause if the phenomenon is similar in several countries.

Finally, the actor-based mechanisms approach invites an area of study which is now being referred to as "aggregation dynamics" (link, link).  We need to have theories and tools that permit us to aggregate different micro-level processes over time into meso- and macro-outcomes, taking into account the complexity of causal interactions in a dynamic process.  The tools of agent-based modeling are relevant here (link).

Spartacus, Kitty Genovese, and social explanation

What is most interesting in paying attention to social life is noticing the surprising outcomes that often materialize from a number of uncoordinated choices and actions by independent individuals. We want to understand why and how the aggregate-level social fact came to be: was it a set of features of the individual actors' preferences or decision-making?  Was it the unintended result of strategic choices by various actors?  Was it simply the path-dependent and contingent outcome of a serial interactive process?  Was it brought about by structural conditions -- power, wealth, race -- within the context of which actors made their choices?  And what were the mechanisms of constraint, aggregation, contagion, and escalation through which actions and processes at the more proximate level came together into the outcome at the distal level?

Consider a pair of examples. Kitty Genovese is attacked by an assailant over an extended period of time in a dense neighborhood in Queens, many people observe the crime, and no one intervenes.  The woman is eventually murdered.  The question here is, why was there a total absence of intervention by any individual or group in this crime?  And a parallel surprise: General Marcus Licinius Crassus, having trapped the rebellious slave army of Spartacus, announces that the slaves will be spared death if they give up Spartacus for crucifiction.  Spartacus rises to his feet to say, "I am Spartacus." And in minutes the army of men rise as well, all declaring "I am Spartacus."  Here the question is the reverse: why do these men expose themselves to death to stand in solidarity with Spartacus?  In each case, there is an occasion for action presented to a group of individuals, in which members of the group can attempt to save the life of another person.  The collective behavior is fundamentally different in the two cases.   Why so? What are the mechanisms, psychological and social, that led to non-intervention in the first case and fatal solidarity in the second?

We might form a number of hypotheses about both of these cases in order to explain the very different outcomes.  In the Kitty Genovese case, we might cite anomy and anonymity as possible causes of the lack of response by bystanders.  With a low level of civic bonds, perhaps city dwellers have such a low level of emotional involvement with each other that even the slightest effort is unjustified.  Or perhaps it is the fact that each potential responder is anonymous to the others that leads to the result; he/she can make the decision to refrain from offering aid without fear of criticism from others.  Or perhaps collective behavior is strongly influenced by the actions of the first few, with later observers mimicking earlier non-aiders.  So the outcome might have been highly different if the first or second witness had intervened; others might have followed suit.   In the case of Spartacus, we might hypothesize that the bonds of solidarity forged by a history of fighting the Roman army gave the soldiers the moral motivation to support Spartacus; or perhaps it is the publicity of the scene, or the early example of the first few supporters, that spread to the behavior of the others.  Or perhaps it is an expression of fundamental mistrust by the soldiers of the good faith of Marcus Crassus; "he will kill us anyway." With nothing to lose, the army makes its symbolic statement of rejection.  These are each social psychological hypotheses, concerning the ways that individuals choose to involve themselves in an emergency and a situation of potential sacrifice.

Each of these hypotheses represents a social mechanism that can be incorporated into a narrative explaining the aggregate outcome.  In order for such a story to be scientifically compelling, we need to have some way of using empirical evidence to evaluate whether the postulated mechanism actually works in the real world of human behavior.  Is there empirical evidence for a social psychology of mimicry?  Do we have evidence to suggest that members of a crowd are more likely to do X or Y if a few others have already done so?  Is there empirical support for a theory of solidarity as a social motivation: do combat teams, groups of deep-ground miners, or emergency room doctors develop a higher willingness to conform their behavior to the good of the group or of other individuals in the group?  These examples of social mechanisms all fall in the category of putative behavioral regularities, and it should be possible to investigate them experimentally and statistically.

The challenge of explanation for any social outcome, we might say, is that of constructing an interpretation of the states of minds of a set of actors; the constraints and opportunities within which they choose a course of action; and the interactions that are created as they act within a common environment, leading to the outcome in question.  This is what we can refer to as an aggregative explanation, and it lies at the heart of Thomas Schelling's methodology in Micromotives and Macrobehavior.  As Schelling points out, sometimes the explanation turns on specific features of agency (the actors' preferences or their modes of decision-making), and sometimes it turns on the specifics of the environment of choice (the fact that the outcome of action is a public good).  But in general, explanation proceeds by showing how agents with specific features, acting within a social and natural environment with specific characteristics, bring about specific kinds of outcomes.

This, in a nutshell, represents a simple but powerful statement of a philosophy of social explanation.  It also represents the rudiments of a social ontology: higher-level social features are composed of the actions and states of agency of a set of actors within the context of locally embodied rules, norms, and expectations.  This is what I refer to as "methodological localism":

This theory of social entities affirms that there are large social structures and facts that influence social outcomes. But it insists that these structures are only possible insofar as they are embodied in the actions and states of socially constructed individuals. The “molecule” of all social life is the socially constructed and socially situated individual, who lives, acts, and develops within a set of local social relationships, institutions, norms, and rule. (link)

Social explanations derive their force from empirical research into the nature of the actor, the nature of the locally embodied social environment, and the processes of aggregation through which actions by multiple actors coalesce into social outcomes.  The explanation satisfies us when it demonstrates the pathways through which individuals, constituted as they are found to be, located in institutions of the sort described, contribute to collective outcomes of the kinds described.

Spartacus, Kitty Genovese, and social explanation

What is most interesting in paying attention to social life is noticing the surprising outcomes that often materialize from a number of uncoordinated choices and actions by independent individuals. We want to understand why and how the aggregate-level social fact came to be: was it a set of features of the individual actors' preferences or decision-making?  Was it the unintended result of strategic choices by various actors?  Was it simply the path-dependent and contingent outcome of a serial interactive process?  Was it brought about by structural conditions -- power, wealth, race -- within the context of which actors made their choices?  And what were the mechanisms of constraint, aggregation, contagion, and escalation through which actions and processes at the more proximate level came together into the outcome at the distal level?

Consider a pair of examples. Kitty Genovese is attacked by an assailant over an extended period of time in a dense neighborhood in Queens, many people observe the crime, and no one intervenes.  The woman is eventually murdered.  The question here is, why was there a total absence of intervention by any individual or group in this crime?  And a parallel surprise: General Marcus Licinius Crassus, having trapped the rebellious slave army of Spartacus, announces that the slaves will be spared death if they give up Spartacus for crucifiction.  Spartacus rises to his feet to say, "I am Spartacus." And in minutes the army of men rise as well, all declaring "I am Spartacus."  Here the question is the reverse: why do these men expose themselves to death to stand in solidarity with Spartacus?  In each case, there is an occasion for action presented to a group of individuals, in which members of the group can attempt to save the life of another person.  The collective behavior is fundamentally different in the two cases.   Why so? What are the mechanisms, psychological and social, that led to non-intervention in the first case and fatal solidarity in the second?

We might form a number of hypotheses about both of these cases in order to explain the very different outcomes.  In the Kitty Genovese case, we might cite anomy and anonymity as possible causes of the lack of response by bystanders.  With a low level of civic bonds, perhaps city dwellers have such a low level of emotional involvement with each other that even the slightest effort is unjustified.  Or perhaps it is the fact that each potential responder is anonymous to the others that leads to the result; he/she can make the decision to refrain from offering aid without fear of criticism from others.  Or perhaps collective behavior is strongly influenced by the actions of the first few, with later observers mimicking earlier non-aiders.  So the outcome might have been highly different if the first or second witness had intervened; others might have followed suit.   In the case of Spartacus, we might hypothesize that the bonds of solidarity forged by a history of fighting the Roman army gave the soldiers the moral motivation to support Spartacus; or perhaps it is the publicity of the scene, or the early example of the first few supporters, that spread to the behavior of the others.  Or perhaps it is an expression of fundamental mistrust by the soldiers of the good faith of Marcus Crassus; "he will kill us anyway." With nothing to lose, the army makes its symbolic statement of rejection.  These are each social psychological hypotheses, concerning the ways that individuals choose to involve themselves in an emergency and a situation of potential sacrifice.

Each of these hypotheses represents a social mechanism that can be incorporated into a narrative explaining the aggregate outcome.  In order for such a story to be scientifically compelling, we need to have some way of using empirical evidence to evaluate whether the postulated mechanism actually works in the real world of human behavior.  Is there empirical evidence for a social psychology of mimicry?  Do we have evidence to suggest that members of a crowd are more likely to do X or Y if a few others have already done so?  Is there empirical support for a theory of solidarity as a social motivation: do combat teams, groups of deep-ground miners, or emergency room doctors develop a higher willingness to conform their behavior to the good of the group or of other individuals in the group?  These examples of social mechanisms all fall in the category of putative behavioral regularities, and it should be possible to investigate them experimentally and statistically.

The challenge of explanation for any social outcome, we might say, is that of constructing an interpretation of the states of minds of a set of actors; the constraints and opportunities within which they choose a course of action; and the interactions that are created as they act within a common environment, leading to the outcome in question.  This is what we can refer to as an aggregative explanation, and it lies at the heart of Thomas Schelling's methodology in Micromotives and Macrobehavior.  As Schelling points out, sometimes the explanation turns on specific features of agency (the actors' preferences or their modes of decision-making), and sometimes it turns on the specifics of the environment of choice (the fact that the outcome of action is a public good).  But in general, explanation proceeds by showing how agents with specific features, acting within a social and natural environment with specific characteristics, bring about specific kinds of outcomes.

This, in a nutshell, represents a simple but powerful statement of a philosophy of social explanation.  It also represents the rudiments of a social ontology: higher-level social features are composed of the actions and states of agency of a set of actors within the context of locally embodied rules, norms, and expectations.  This is what I refer to as "methodological localism":

This theory of social entities affirms that there are large social structures and facts that influence social outcomes. But it insists that these structures are only possible insofar as they are embodied in the actions and states of socially constructed individuals. The “molecule” of all social life is the socially constructed and socially situated individual, who lives, acts, and develops within a set of local social relationships, institutions, norms, and rule. (link)

Social explanations derive their force from empirical research into the nature of the actor, the nature of the locally embodied social environment, and the processes of aggregation through which actions by multiple actors coalesce into social outcomes.  The explanation satisfies us when it demonstrates the pathways through which individuals, constituted as they are found to be, located in institutions of the sort described, contribute to collective outcomes of the kinds described.

Spartacus, Kitty Genovese, and social explanation

What is most interesting in paying attention to social life is noticing the surprising outcomes that often materialize from a number of uncoordinated choices and actions by independent individuals. We want to understand why and how the aggregate-level social fact came to be: was it a set of features of the individual actors' preferences or decision-making?  Was it the unintended result of strategic choices by various actors?  Was it simply the path-dependent and contingent outcome of a serial interactive process?  Was it brought about by structural conditions -- power, wealth, race -- within the context of which actors made their choices?  And what were the mechanisms of constraint, aggregation, contagion, and escalation through which actions and processes at the more proximate level came together into the outcome at the distal level?

Consider a pair of examples. Kitty Genovese is attacked by an assailant over an extended period of time in a dense neighborhood in Queens, many people observe the crime, and no one intervenes.  The woman is eventually murdered.  The question here is, why was there a total absence of intervention by any individual or group in this crime?  And a parallel surprise: General Marcus Licinius Crassus, having trapped the rebellious slave army of Spartacus, announces that the slaves will be spared death if they give up Spartacus for crucifiction.  Spartacus rises to his feet to say, "I am Spartacus." And in minutes the army of men rise as well, all declaring "I am Spartacus."  Here the question is the reverse: why do these men expose themselves to death to stand in solidarity with Spartacus?  In each case, there is an occasion for action presented to a group of individuals, in which members of the group can attempt to save the life of another person.  The collective behavior is fundamentally different in the two cases.   Why so? What are the mechanisms, psychological and social, that led to non-intervention in the first case and fatal solidarity in the second?

We might form a number of hypotheses about both of these cases in order to explain the very different outcomes.  In the Kitty Genovese case, we might cite anomy and anonymity as possible causes of the lack of response by bystanders.  With a low level of civic bonds, perhaps city dwellers have such a low level of emotional involvement with each other that even the slightest effort is unjustified.  Or perhaps it is the fact that each potential responder is anonymous to the others that leads to the result; he/she can make the decision to refrain from offering aid without fear of criticism from others.  Or perhaps collective behavior is strongly influenced by the actions of the first few, with later observers mimicking earlier non-aiders.  So the outcome might have been highly different if the first or second witness had intervened; others might have followed suit.   In the case of Spartacus, we might hypothesize that the bonds of solidarity forged by a history of fighting the Roman army gave the soldiers the moral motivation to support Spartacus; or perhaps it is the publicity of the scene, or the early example of the first few supporters, that spread to the behavior of the others.  Or perhaps it is an expression of fundamental mistrust by the soldiers of the good faith of Marcus Crassus; "he will kill us anyway." With nothing to lose, the army makes its symbolic statement of rejection.  These are each social psychological hypotheses, concerning the ways that individuals choose to involve themselves in an emergency and a situation of potential sacrifice.

Each of these hypotheses represents a social mechanism that can be incorporated into a narrative explaining the aggregate outcome.  In order for such a story to be scientifically compelling, we need to have some way of using empirical evidence to evaluate whether the postulated mechanism actually works in the real world of human behavior.  Is there empirical evidence for a social psychology of mimicry?  Do we have evidence to suggest that members of a crowd are more likely to do X or Y if a few others have already done so?  Is there empirical support for a theory of solidarity as a social motivation: do combat teams, groups of deep-ground miners, or emergency room doctors develop a higher willingness to conform their behavior to the good of the group or of other individuals in the group?  These examples of social mechanisms all fall in the category of putative behavioral regularities, and it should be possible to investigate them experimentally and statistically.

The challenge of explanation for any social outcome, we might say, is that of constructing an interpretation of the states of minds of a set of actors; the constraints and opportunities within which they choose a course of action; and the interactions that are created as they act within a common environment, leading to the outcome in question.  This is what we can refer to as an aggregative explanation, and it lies at the heart of Thomas Schelling's methodology in Micromotives and Macrobehavior.  As Schelling points out, sometimes the explanation turns on specific features of agency (the actors' preferences or their modes of decision-making), and sometimes it turns on the specifics of the environment of choice (the fact that the outcome of action is a public good).  But in general, explanation proceeds by showing how agents with specific features, acting within a social and natural environment with specific characteristics, bring about specific kinds of outcomes.

This, in a nutshell, represents a simple but powerful statement of a philosophy of social explanation.  It also represents the rudiments of a social ontology: higher-level social features are composed of the actions and states of agency of a set of actors within the context of locally embodied rules, norms, and expectations.  This is what I refer to as "methodological localism":

This theory of social entities affirms that there are large social structures and facts that influence social outcomes. But it insists that these structures are only possible insofar as they are embodied in the actions and states of socially constructed individuals. The “molecule” of all social life is the socially constructed and socially situated individual, who lives, acts, and develops within a set of local social relationships, institutions, norms, and rule. (link)

Social explanations derive their force from empirical research into the nature of the actor, the nature of the locally embodied social environment, and the processes of aggregation through which actions by multiple actors coalesce into social outcomes.  The explanation satisfies us when it demonstrates the pathways through which individuals, constituted as they are found to be, located in institutions of the sort described, contribute to collective outcomes of the kinds described.

Hobbes an institutionalist?

Here is a surprising idea: of all the modern political philosophers, Thomas Hobbes comes closest to sharing the logic and worldview of modern social science. In Leviathan (1651) he sets out the problem of understanding the social world in terms that resemble a modern institutionalist and rational-choice approach to social explanation. It is a constructive approach, proceeding from reasoning about the constituents of society, to aggregative conclusions about the wholes that are constituted by these individuals. He puts forward a theory of agency -- how individuals reason and what their most basic motives are. Individuals are rational and self-concerned; they are strategic, in that they anticipate the likely behaviors of other agents; and they are risk-averse, in that they take steps to avoid attack by other agents. And he puts forward a description of two institutional settings within which social action takes place: the state of nature, where no "overawing" political institutions exist; and the sovereign state, where a single sovereign power imposes a set of laws regulating individuals' actions.

In the first institutional setting, he argues that individual competition in the context of the absence of sovereignty leads to perpetual violent competition. In the second institutional setting, he argues that individual self-striving within the context of a system of law leads to the accumulation of property and peaceful coexistence.

Here are some of Hobbes's premises about individual agents from chapter XIII of Leviathan:

From this equality of ability ariseth equality of hope in the attaining of our ends. And therefore if any two men desire the same thing, which nevertheless they cannot both enjoy, they become enemies; and in the way to their end (which is principally their own conservation, and sometimes their delectation only) endeavour to destroy or subdue one another. And from hence it comes to pass that where an invader hath no more to fear than another man's single power, if one plant, sow, build, or possess a convenient seat, others may probably be expected to come prepared with forces united to dispossess and deprive him, not only of the fruit of his labour, but also of his life or liberty. And the invader again is in the like danger of another.

So that in the nature of man, we find three principal causes of quarrel. First, competition; secondly, diffidence; thirdly, glory. The first maketh men invade for gain; the second, for safety; and the third, for reputation. The first use violence, to make themselves masters of other men's persons, wives, children, and cattle; the second, to defend them; the third, for trifles, as a word, a smile, a different opinion, and any other sign of undervalue, either direct in their persons or by reflection in their kindred, their friends, their nation, their profession, or their name.

The passions that incline men to peace are: fear of death; desire of such things as are necessary to commodious living; and a hope by their industry to obtain them. And reason suggesteth convenient articles of peace upon which men may be drawn to agreement. These articles are they which otherwise are called the laws of nature, whereof I shall speak more particularly in the two following chapters.

And these motives and forms of behavior by individuals lead to a predictable outcome for the collectivity in the state of nature: a war of all against all.

Whatsoever therefore is consequent to a time of war, where every man is enemy to every man, the same consequent to the time wherein men live without other security than what their own strength and their own invention shall furnish them withal. In such condition there is no place for industry, because the fruit thereof is uncertain: and consequently no culture of the earth; no navigation, nor use of the commodities that may be imported by sea; no commodious building; no instruments of moving and removing such things as require much force; no knowledge of the face of the earth; no account of time; no arts; no letters; no society; and which is worst of all, continual fear, and danger of violent death; and the life of man, solitary, poor, nasty, brutish, and short.

This is an institutionalist argument. It models the behavior that is expected of a certain kind of agent within a certain kind of institutional setting; and it projects the consequences of these "microfoundations" for the aggregate society. In other words, Hobbes is offering a micro- to macro-argument based on analysis of modes of agency and assumptions about a particular institutional context.

Compare this logic with a description of the logic of social explanation offered by contemporary rational-choice social theorist James Coleman in Foundations of Social Theory:

A second mode of explanation of the behavior of social systems entails examining processes internal to the system, involving its component parts, or units at a level below that of the system. The prototypical case is that in which the component parts are individuals who are members of the social system. In other cases the component parts may be institutions within the system or subgroups that are part of the system. In all cases the analysis can be seen as moving to a lower level than that of the system, explaining the behavior of the system by recourse to the behavior of its parts. This mode of explanation is not uniquely quantitative or uniquely qualitative, but may be either. (2)

So the logic of Hobbes's argument is fairly clear; and it is deeply similar to that of institutionalist rational-choice theorists. Thomas Schelling's title, Micromotives and Macrobehavior, captures the idea in three words: derive descriptions of macro-level social arrangements and behavior from premises concerning individual-level motivation and action.

It is not a profound criticism of Hobbes's philosophical analysis to quarrel with Hobbes's specific assumptions about what is possible within the state of nature. And in fact, a number of contemporary political scientists argue that it is possible for men and women to create non-political institutions within the context of what Hobbes calls the state of nature. Coordination and cooperation are indeed possible within a "state of nature"; it is possible to achieve coordination within anarchy. From a sociological point of view, this is really a friendly amendment; it simply adds a further premise about the feasibility of certain kinds of cooperation. So the "cooperation within anarchy" criticism of Hobbes is advanced as a substantive argument about the feasibility of durable social institutions that do not depend upon a central coercive authority. And it depends upon several specific assumptions about the circumstances and mechanisms through which local groups of people can establish self-enforcing forms of cooperation that overcome free-riders and predatorial behavior. It is likely enough that Hobbes would not have been persuaded by this argument; but ultimately it is an empirical question.

Several arguments against Hobbes's conclusions about the state of nature are especially valuable from this point of view. First, I find Michael Taylor's arguments in Community, Anarchy and Liberty particularly convincing -- essentially, that peasant communities have traditionally found ways of creating and sustaining cooperative institutions and relationships that persist without the force of law to stabilize them. "Contracts" backed by legal systems are not the only way of establishing coordination and cooperation among independent agents. Robert Netting provides relevant examples in Smallholders, Householders: Farm Families and the Ecology of Intensive, Sustainable Agriculture, around traditional forms of labor-sharing and seasonal cooperation. And Elinor Ostrom and her collaborators make similar arguments in their historical and sociological studies of "common property resource regimes" -- essentially, stable patterns of cooperation maintained by local voluntary enforcement rather than central legislation (Governing the Commons: The Evolution of Institutions for Collective Action). Ostrom documents dozens of important historical cases where traditional communities have managed fisheries, forests, water resources, and other common properties without having a central state to support these patterns of cooperation and coordination.

But these are empirical and theoretical refinements to a fundamentally coherent model of social explanation that is full-fledged in Hobbes's work in the mid-seventeenth century: explain aggregate (macro) social outcomes as the result of mechanisms and actions at the level of individual actors.