Winner of the 2016 Lakatos Award for outstanding contribution to the philosophy of science Winner of the 2016 Joseph B. Gittler Award for outstanding contribution to the philosophy of social science We live in a world of crowds and corporations, artworks and artifacts, legislatures and languages, money and markets. These are all social objects - they are made, at least in part, by people and by communities. But what exactly are these things? How are they made, and what is the role of people in making them?
In The Ant Trap, Brian Epstein rewrites our understanding of the nature of the social world and the foundations of the social sciences. Epstein explains and challenges the three prevailing traditions about how the social world is made. One tradition takes the social world to be built out of people, much as traffic is built out of cars. A second tradition also takes people to be the building blocks of the social world, but focuses on thoughts and attitudes we have toward one another. And a third tradition takes the social world to be a collective projection onto the physical world. Epstein shows that these share critical flaws. Most fundamentally, all three traditions overestimate the role of people in building the social they are overly anthropocentric.
Epstein starts from scratch, bringing the resources of contemporary metaphysics to bear. In the place of traditional theories, he introduces a model based on a new distinction between the grounds and the anchors of social facts. Epstein illustrates the model with a study of the nature of law, and shows how to interpret the prevailing traditions about the social world. Then he turns to social groups, and to what it means for a group to take an action or have an intention. Contrary to the overwhelming consensus, these often depend on more than the actions and intentions of group members.
thought the distinction between grounding & anchoring was cool and models were intuitive! also find the idea that social sciences are 'too anthropocentric' compelling but still thinking about that
The Ant Trap is incredibly ambitious and rewarding. Though there is really too much packed into this book of just under 300 pages to give an adequate review, a simple example will explain the upshot of the book. Recently, a man moved to Washington, D.C., taking up residence on Pennsylvania Avenue. When he signs his name to certain pieces of paper it seems to make a difference to what many other people do. What makes it the case that this man is the president of the United States? Epstein answers this in terms of two sets of conditions. First and straightforwardly, we could reply that Trump was elected in in 2016; voters in such and such states voted for him; the electoral college voted in such and such a way, etc; and so on. Epstein calls these facts the grounds of Trump's presidency.
But Epstein goes one step further, arguing that these conditions - as necessary and sufficient conditions for Trump to be president - must be anchored in some way. Anchoring is a matter of establishing the grounds that decide when some social fact is the case. What anchors the fact that Trump is president? At the very least we must point to the ratification of the US Constitution. By dividing anchors and grounds Epstein allows for more nuance concerning the creation and structure of the social world than many other theorists.
Particularly he is concerned with rejecting both ontological (or grounding) individualism and anchor individualism, the idea that social facts are grounded or anchored, respectively, by facts about individuals. Both positions have many adherents in philosophy and the social sciences but Epstein raises important questions about both. About the former Epstein asks us to take note of the fact that the grounding of social facts often depends upon facts about physical objects, so the financial solvency of Starbucks (his example) depends upon facts about the machines and buildings used by the company, conditions that may be altered independently of human attitudes (in the case of a power-spike that destroys Starbucks' facilities).
Epstein is more reticent about anchor individualism. Though he does note, "In particular, we will need to understand how functional roles and environmental facts can partly anchor a set of frame principles" (p. 278). One way to understand this is in terms of material conditions for certain social facts. For example, that certain objects are money is a type of social fact but it is clear that not any type of object can be money. Facts about the objects partly anchor their status as money. In order to serve this purpose the objects that are constituted as currency must at least be scarce and a store of value: neither air nor fresh apple pies could plausibly function as currency.
Future work should seek to compare and contrast Epstein's book with Guala's 'Understanding Institutions: The Science and Philosophy of Living Together,' as this latter books has much to say about questions about the material conditions or substrates of institutions like money. The biggest drawback of Epstein's book is his silence about anchoring. This is no doubt entirely understandable since it raises questions of immense complexity. To understand this one should realize that the most important book about anchoring is probably Hegel's Phenomenology of Spirit, a book about how normative attitudes change over time. Future work will inevitably have to confront Epstein's approach to social ontology with an Hegelian understanding of normativity. Finally, it remains to be seen whether Epstein's book can benefit working social scientists, hopefully future work will attempt to employ his distinction between anchors and grounds to better model social phenomena.
This is a really nice book. The exposition of the central theories is great, sympathetic and insightful. The metaphysical structure Epstein erects is great. The one question I'm left with is: do our models in the social sciences really suffer because of a bad ontology? There are some references to other work re: applications of Epstein's improved ontology at the very end, so that's the next place to look.
Before reading this, I thought that having the right metaphysics of the facts some science is concerned with would be nice, but not really essential--you can do the science successfully with all kinds of different metaphysical views. For example, the science I'm most interested in is linguistics. I didn't particularly care whether internalism or externalism about meaning facts was true--you could just investigate meaning experimentally in the way that is done in semantics and pragmatics, and whatever metaphysics best fit with the practice of linguists would be the right one.
In contrast, this book presents a strong case that a bad metaphysics is one that doesn't leave room for all the sorts of facts that play a role in explaining some complex phenomenon (Epstein makes good use of Virchow's cell theory of organisms to help illuminate this--assuming that all living things are made up of exclusively of cells misses a bunch of important stuff and leads to a deformed ontology). Having a bad metaphysical theory will make doing actual science harder--Epstein is most concerned with individualism in the social sciences. Assuming that all we need to do explanatory work in the social sciences are individuals is as impoverished, Epstein argues, as Virchow's cell theory of living things.
I found this a really stimulating way of thinking about debates about internalism and externalism in the metaphysics of meaning. Why only use the resources of the individualist cognitive sciences to explain meaning, when pretty clearly meaning is a social phenomenon? (Putnam says that semantics is a social science in "Is Semantics Possible?") And Epstein points out that the grounding conditions for different kinds of social facts will likely be very different--which should also be kept in mind when thinking about meaning facts. How the meaning of, say, "hardness" (to take an example from Mark Wilson) is grounded will be different than the way the meaning of a discourse marker (like, "well...", or "y'know") is grounded. That will have consequences for how we experimentally investigate the meaning of these expressions--studying participant judgments or speech may be all there is to investigating the meaning of discourse markers, but probably not all there is when investigating the meaning of an expression like "hardness", which will be dependent on scientific practices of various kinds.