Critical Realism

[John McCreery]

Elizabeth Challiinor has introduced us to the work of Margaret Archer. Hoping to learn more about Archer, I turned to Questia and found Archer's Being Human: The Problem of Agency available there. In the introduction to Being Human, I learned that Archer associates herself with a philosophical movement called Critical Realism.

Critical Realism is said to have begun with the work of Indian-British philosopher Roy Bhaskar. The Wikipedia entry on Bhaskar describes his thinking as follows,

Bhaskar's consideration of the philosophies of science and social science resulted in the development of Critical Realism, a philosophical approach that defends the critical and emancipatory potential of rational (scientific and philosophical) enquiry against both positivist, broadly defined, and 'postmodern' challenges. Its approach emphasises the importance of distinguishing between epistemological and ontological questions and the significance of objectivity properly understood for a critical project. Its conception of philosophy and social science is a socially situated, but not socially determined one, which maintains the possibility for objective critique to motivate social change, with the ultimate end being a promotion of human freedom.

To me this sounds very similar to OAC founder Keith Hart's project in The Human Economy and his constant refrain, the need to return anthropology to its Enlightenment roots, especially Kant. 

Archer herself contrasts Critical Realism with three other major movements in social science, all of which, she says, conflate the agent and society, seen by critical realists as interacting but autonomous entities. Two are reductionist, reducing society to individuals, or seeing individuals as artifacts of society. The third, which she associates with Giddens and Bourdieu, attempts to subsume both in a comprehensive larger system. 

My own first encounter with Critical Realism was The SAGE Handbook of Case-Based Methods, whose contributors discuss various methods for addressing the at least partial autonomy of agents and structures, giving full credit to the reality of both and to that of the material world in which both are situated.

I wonder if anyone else here has encountered Critical Realism and, if so, what they think of it. Or, even if they are totally new to the subject, find the description above tempting enough to join me in looking a bit deeper into what this movement may offer to anthropologists.

[Keith Hart]

Thanks for taking the care to post this thread, John. I realise that I ought to have something to say about it and in a way I do, but the term emerged at a time when I had stopped teaching social theory and I never took the trouble to get to know critical realism more than superficially. In any case, what strikes me about the position is its positioning in relation to other contemporary approaches more than any geneaology linking it to classical antecedents. My suspicion has always been that it bears some relationship to Marxism. I know that there are books on the two that would deny the possibility of reducing one to the other. But Marx, Engels and their followers sought to steer a path between critical philosophy, scientific history and economics that in Marx's case led him to flights of invention that rivalled the great novelists of his day. The better analogy (than with Kant) is with Rousseau who knew that, if you want to change the world, it makes no sense just to read off how the world is and has been. Science is conservative in that it assumes the continuity of the past in the present. That is why J-J experimented with a mixture of fiction and non-fiction genres in his great critiques of politics, education, sexuality and the self written in the 1760s. I would suggest that, far more than the structure/agency or society/individual problem that dogged 20th century social theory, the crucial issue is one of tense -- how past, present and future are constructed together -- and then one of value and fact, of judgment and empirical knowledge. But I haven't looked into the claims of this tendency closely enough to make that observation more specific.

[John McCreery]

Keith, thanks for the kind words. I posted this mainly out of curiosity to see if anyone else here had even heard of this movement. As I noted in my post, Elizabeth's pointer to Margaret Archer led me to the first reference to Critical Realism that I have encountered outside The SAGE Handbook of Case-Based Methods, whose contributors appear to be primarily sociologists, criminologists and political scientists. I bought a copy of that book because it contains a paper by Ron Breiger "On the Duality of Cases and Variables" to which I was pointed by my former student and good friend James Ennis, who studied with Breiger at Harvard and later with Pierre Bourdieu in Paris. Intellectual life in the age of the network.

Anyway, for those who might like a richer taste of where all this is leading me, I have just scanned the first two pages of the introduction to that book and marked a few passages that seem to me, with my own preoccupation with methods, of particular interest. I attach it here for you and others' reference.

The authors remark on a congeniality between Critical Realism as a philosophy and their own project, which is to get beyond the dilemma posed by classical antitheses, in particular that which contrasts scientific, causal analysis and idiographic interpretation. The former they see as too tied to the notion of universal, nomothetic laws, the latter as rendering generalization impossible. They seek a middle ground where generalization is possible but the limits of the domain covered by the generalization are also recognized. 

Why seek such middle ground? Many social and cultural phenomena are similar for different reasons; different causes have produced what seems, at least at first glance, to be the same effect. Conversely, the same cause may have quite different effects depending on context.

One might, my own example here but addressing some of your concerns, observe that inequalities in the distribution of wealth do not drive revolutions in booming economies where everyone feels like they are doing well. The same inequalities may have very different effects when an economy collapses and the difference between haves and have-nots comes to be seen as intolerable. 

Learning to think in this way is, of course, hard, trapped as we are between habits that insist on UNIVERSAL TRUTH or idiosyncratic circumstance as the only grounds of explanation.

 

[John McCreery]

P.S. Several of the authors of The SAGE Handbook of Case-Based Methods acknowledge the influence of anthropologist Fred Eggan's classic "Social Anthropology and the Method of Controlled Comparison" or sociologist Robert Merton's call for Middle Range Theory. Their aim is to build on these suggestions by asking the vital "How do you do that?" question instead of simply kowtowing to intellectual ancestors.

[Keith Hart]

It's a pity that the tone of your response lapsed in that final remark, John. As far as I am concerned a pragmatic approach to research and intellectual history are not alternatives. I have found that I think well through revisiting the classics, but I don't insist that others do so. I have taught research methods, ranging from fieldwork to statistics, in anthropology, sociology, economics and development studies all my life and I tell anyone who cares to listen that a good research proposal must connect large questions from the literature to concrete research questions through middle range hypotheses.

One might...observe that inequalities in the distribution of wealth do not drive revolutions in booming economies where everyone feels like they are doing well. The same inequalities may have very different effects when an economy collapses and the difference between haves and have-nots comes to be seen as intolerable.

Tocqueville in The Old Regime and the French Revolution attributes the latter to four causes in combination: the popular spread of Enlightenment ideas concerning freedom and equality; the rigidity of the class structure allowing for almost no mobility; an economic boom which pushed middle classes, workers and peasants to be denied social expression of their economic improvement; and a repressive rather than accommodating response of the authorities to pressure from below. Thinking about this analysis helps me to pose questions about revolution today, but of course a practical research project requires a lot more.

On the second point, I guess I would say again that it is better to read up on the history of the subject before deciding what you do as a researcher. I am lucky to have access to an ongoing seminar on the history of money in Paris which has been going for 20 years, since the euro was launched by the Treaty of Maastricht. This has generated a number of collected volumes, but two in particular bring an impressive range of examples to the comparative history of monetary crises: Volume 1 of La monnaie devoilee par ses crises (Money unveiled by its crises) edited by Bruno Theret looks at the problem on a wide comparative basis, while Volume 2 investigates German and Russian crises of the twentieth century in extraordinary depth.

Being interested in intellectual and social history is not "simply kowtowing to intellectual ancestors". It is one strategy among many and it ought to be possible to admit that without denigrating other people's preferences if they differ from our own.

[John McCreery]

@Keith

Apologies. Shouldn't have done the all-caps thing. But there is an important point that the case-based studies folks are making. Too much of everyday as well as academic argument is trapped in the frame Aristotle created, where substance is universal and forever and everything else is accident and contingent. Pointing to law like behavior between these extremes opens up all sorts of interesting possibilities.

That is, by the way, a good reason to review the literature and study the history of any topic that interests you. It is when you begin to see the patterns in the variation that you start to make real progress.

[Marie Louise Seeberg]

Hello there - I saw your post, John, and was happy about it because I've been trying out different ways of applying CR as an approach and have, till now, not really found any anthropologists who were interested in it. At best, they tend to tell me (when I present my understanding of CR to them) that "this is what all good anthropology is already doing!". I think that's true but I think CR provides a way to put into words the reason why this makes for good, sound research - anthropology or otherwise. The International Association of CR has annual conferences, I went to the one they had this year and would recommend these conferences to anyone curious to learn more. They also have their own wikispace at http://criticalrealism.wikispaces.com/.

[John McCreery]

Marie Louise, could you tell us more about your experiments? What sorts of applications have you found? I don't know about you, but I always find it easier to think concretely, with some cases to think about. Please, pretty please....

[Marie Louise Seeberg]

If you insist... I've been interested in CR ever since my PhD which I completed in 2003 and have tried out little bits of it in different projects. Here is a paragraph from a book chapter which is in press at the moment. The paper is about a man, Bienvenido, who has migrated from the Philippines to Norway as a nurse:

"How do Bienvenido’s boundary crossings relate to the intersections of systems of social inequality? My concern is with the interplay between the different social systems marked by these boundaries rather than with the boundaries as such. This necessitates an understanding of all sets of social relations as open systems that form each other’s constantly changing environments. A system is “a regularly interacting or interdependent group of items forming a unified whole” (Merriam Webster’s online dictionary). By “open system” I mean a system with porous boundaries across which elements within the system interact with elements that form part of other systems. Boundaries between social systems are always porous, since such systems are always open and interconnected with other systems (Danermark, Ekstrom, & Jakobsen, 2002). Further, I view systems not as static constructions but as dynamic entities, in continuous adaptation to changing environments."

[John McCreery]

@Marie Louise

This sounds like a great project, and I do like the way you are addressing the question of open systems with porous boundaries and thinking of them as dynamic instead of static entities. Could you say a bit more about how thinking from critical realism informs what you are doing.

As I've mentioned elsewhere my introduction to critical realism was The SAGE Handbook of Case-based Methods, whose authors are explicitly concerned with the contrast between case-based methods and classic variable-based methods, e.g., regressions. The concern with cases originates in the observation that dissimilar causes can have similar effects and similar causes can have dissimilar effects depending on other circumstances. Practically speaking, case-based approaches tend to work best with small samples and lots of variables, as opposed to standard statistical methods that work best with large samples and only a small number of variables. How these methodological concerns relate to critical realism seems to be via the idea that individuals or other social actors are at least partially autonomous and the same is true of the larger systems in which they may find themselves embedded. But that's it. That is where my grasp of the subject comes to a sudden end. Does this have anything to do with the way in which you think about critical realism and apply it in your own work?

[Marie Louise Seeberg]

@John

I haven't read the same book as you have - my main source is Sayer's Methods in social science - but let's assume CR is more or less the same family of ideas in both cases. Also, I don't really know the meaning of "regressions" - my apologies for being an ignoramus when it comes to quantitative methods.

I think a critical realist would object to the idea that individuals or other social actors are necessarily embedded in larger systems. Rather, the connections and patterns of interaction between individuals and other systems should be investigated empirically and are as likely to be just partly overlapping as to be nested in any hierarchical way. I have found the writings of Sylvia Walby very useful here. I don't know if she calls herself a critical realist, but I can't see why she shouldn't. Here is a quote from one of her papers (you'll find the paper here

http://pos.sagepub.com/cgi/content/abstract/37/4/449):

"The notion that each system has as its environment all other systems is

used to replace the rigid notion of a hierarchy of sub-systems by a much

more fluid conception of the mutual impact of systems. This means that the

phenomena that many systems-based Sociologists have treated as nested,

subordinate elements within systems are here conceptualized as separate

systems. This enables us both to keep the notion of system, and the notion

of systematic inter-relatedness, while yet not pre-specifying, in a rigid way,

the nature of these inter-connections."

My instinct would be that the clue you are looking for is in where you say that "dissimilar causes can have similar effects and similar causes can have dissimilar effects depending on other circumstances". We're not just talking about an individual and/in a system, but any number of interacting systems of which individual actors are some and more aggregated entities are others. All these systems form parts of each other's environment, and this means that context, or 'other circumstances' must be included in the equation -  necessary to analysis. I don't know if this makes any sense to you?

[Keith Hart]

Thanks very much for your contribution, Marie Louise. I have long fancied getting more acquainted with critical realism and your link gave me a chance to start reading Bhaskar's book. I note that his aim there is to give a philosophically adequate account of science, but I suspect that many of his followers use his ideas to develop a philosophically adequate social analysis of anything. I have long maintained an amateur interest in using science to understand ideas about society. I believe that scientists, especially when they study phenomena that apparently have little to do with human beings -- the stars, earthquakes, elementary particles -- express historically variable notions of society in their explanations. I take Durkheim and Mauss's essay on primitive classification as a starting point. It is probably off-topic, but I include a link to some work I did on changes in how scientists approach statistics: The social meaning of the power law.

[John McCreery]

@Marie-Louise

This makes a lot of sense to me, in principle. I say "in principle" because, as I have noted elsewhere, the next question for me is always "How?". It is one thing to agree that the world is the intersection of interacting systems and another to decide where to draw the lines that allow us to isolate a small enough chunk of the world to say something meaningful about those interactions.

The history of science teaches me that radical simplification is the key to scientific advances. Experimental method consists, above all, in excluding everything else to focus on the variables in question. Conventional statistical hypothesis testing requires random sampling in an effort to achieve the same goal. The result has been that bifurcation of the world that A.N.Whitehead describes in Science and the Modern World—the division of the world into stuff that fits the radically simplified models and the stuff that doesn't and has, historically, been brushed aside as secondary rather than primary qualities, epiphenomenal instead of essential, superstructure instead of foundation. So we now live in a world described by Gerald Weinberg in An Introduction to General Systems Analysis, in which a few things can be explained by simple, mechanical models, a few more can be explained by statistical models, and the rest, by far the larger part of the world, can be explained in neither of these ways. The usual human response when confronted by its muddle has been to tell stories—to construct narratives that trace a path from A to B while considering only a few, if any, of the other possible paths.

Recent interest in topics like chaos, complexity, and my own current hobby, social network analysis, has come from the scientific side, where realizing that current models are too simple to explain everything we want explained has led to attempts to develop more complex models, a process that I would describe by borrowing the words of Clifford Geertz, from the opening paragraph of "The Impact of the Concept of Culture on the Concept of Man":

substituting complex pictures for simple ones while striving somehow to retain the persuasive clarity that went with the simple ones.

I would argue that if the interpretive turn of the 1980s went astray, it did so by losing its focus on the second half of this proposition. Everyone got on board with substituting complex pictures for simple ones and did what humans have traditionally done with complicated stuff, telling stories about it. What they failed to do was retain the persuasive clarity that went with the simple ones. The result has been the awful muddle in which so much anthropological thinking now finds itself.

Are the best current models still too simple? Of course they are. And the people who develop them are well aware of that. It tickles me to read in John H. Miller and Scott E. Page's Complex Adaptive Systems: An Introduction to Computational Models of Social Life that that the agents in current agent-based models are either too stupid (dependent on a few elementary heuristics) or too sophisticated (constantly calculating decisions in terms of game theory) to be good models of actual human behavior. But the serious people (the ones I consider serious) haven't given up the search for persuasive clarity. They are looking, instead, for ways to improve their models, to enlarge the zone of the understood while never failing to recognize that lots, and lots, and lots remains to be explained.

Coming back, then, to Critical Realism. What I am looking for is links between the ideas developed by Bhaskar, et. al, and examples of ethnographic or theoretical practice. If, again borrowing words from Geertz, the former provide new "models of" the world in which we work, what are the "models for" anthropological practice to which they point us?