Brian Epstein’s The Ant Trap: Rebuilding the Foundations of the Social Sciences is a wonderful contribution to metaphysics, social ontology, and philosophy of social science (and relevant for all the human sciences). I rarely find myself agreeing with Christian List, but I agree with his claim (blurbed on the back of the book) that the book is “beautifully written and packed with insights.” I predict that down the road the book will improve the practice of establishing causes in social science (about that some other time more). The book also contains lovely material on all kinds of topics (including a fantastic essay on Virchow, an excellent treatment of reduction in the sciences, fine criticism of Searle on social ontology, amongst other gems, etc.)
The key concepts in the book are "anchoring," "grounding,"* and "frame principle(s)." Here I can't do justice to Epstein's rich discussion, but if I provide his definitions, you'll get a sense of the position.
- A frame "is a structure containing...a set of possible worlds in which the grounding conditions for social facts are fixed in a particular way. Each of these possible worlds may have different grounding facts from one another." (78)
- A frame principle "gives the grounding conditions not just for the actual world, but for all possibilities" [that belong to a particular frame]. (78) This principle is the constitutive principle of the frame.
- Anchoring is "a relation between a set of facts and a frame principle. For a set of facts to anchor a frame principle is for those facts to be the metaphysical reason that the frame principle is the case." (82) (So anchoring is external to a frame.)
- Grounding is "a relation between facts."** In the context of Epstein's inquiry he uses '"grounding" to describe a relation within a frame between the more fundamental "grounding facts" (or grounding conditions) and what he calls "social facts." A social fact is a "fact that corresponds to a proposition that has any social entity as a constituent;" a social fact might "have social objects as constituents, or it might have social properties as constituents, or both." (67).
There is a "separate inquiry" within social ontology into grounding and anchoring (86). In addition, Epstein insists that "the grounding inquiry tends to precede the anchoring inquiry." (86) It is certainly the case that in much data-driven social science the anchoring inquiry is simply taken for granted, but as Epstein's wording implies, there is no firm reason why one inquiry ought to precede the other.
Epstein uses these distinctions to diagnose a lot of confusions in existing discussions about social ontology. (I was especially impressed by his ability to sort out a whole variety of conflations in discussions about 'indvidualism' in social science/social ontology.) But in what follows I want to flag one limitation of his existing (sorry for the pun) framework. I don't think the objection I am about to raise is fatal; I suspect it can be repaired, too. I quote before I comment:
In speaking about social kinds...it is useful to think of them as the categories we might use in the social sciences, but remain open-minded about the sorts of categories these might be. Maybe the social kinds are the same as the social properties. Maybe they are a subset of the social properties. Or maybe they are different things altogether. To make progress, we do not need to start with comprehensive understanding. We can just regard "social kind" as a generic way of referring to categories like these.
Social kinds--like social properties--have fixed instantiation conditions. Or, more appropriately, we might say that kinds have fixed membership conditions. The conditions under which something is a member of a social kind are the same across all times and possibilities. The reason for this is the same as the one I gave above. Social kinds serve a variety of functions: we employ them for recognizing things, classifying things in various situations, finding and correcting departures from norms, drawing inductive inferences, and accomplishing other practical matters (68; [emphasis in original]).
Here Epstein inherits what I take to be the traditional problem of analytic philosophy's thinking about concepts and kinds (a problem which tempts some folk toward pragmatism, after all). To be sure, for lots of projects, Epstein's approach is sufficient. But I think it is inappropriate if one is, say, interested in the electoral franchise over time in the UK and its (former) Colonies. Membership conditions in electoral franchises -- let's stipulate -- have evolved rather dramatically over time. Moreover, a whole host of functions and symbolisms have been added and subtracted from belonging to (or being denied membership in) the franchise in the English speaking world alone.
On Epstein's approach to social kinds, in the case of dramatic changes in franchises one is comparing different social kinds over time (as the membership conditions and the grounding relations between the frame principles and the social kinds shifts). Now this can be the right stance for some purposes; it can be really illuminating to recognize that the franchise has become a different kind. But this view of social kinds is also a self-limitation on his approach. (The issue also came up obliquely in my response to his excellent NYT piece on evolving conception of marriage.) So, what's needed is to augment Epstein's picture with a view that allows social kinds to evolve over time or that allows comparison of the 'same' social kind in different frames.