Against social ontology

At first I thought the vogue for “social ontology” was just a pale reflection of the (ultimately Quine-inspired) revival of ontology over the past few decades in analytical philosophy more generally. But then just in the past ten or fifteen years, social ontology rather dramatically took on a life of its own. John Searle proclaimed, for instance, that “where the social sciences are concerned, social ontology is prior to methodology and theory.” (Imagine someone saying such things about physics — since Descartes, anyway!) This burst of enthusiasm reached its apogee in about 2015, I think it fair to say, with the launch of the Journal of Social Ontology and with Brian Epstein’s The Ant Trap: Rebuilding the Foundations of the Social Sciences, perhaps the most brazen attempt in recent memory to reclaim the Platonic mantle of philosopher king. Historians, economists, sociologists, and assorted others working in the trenches were told that (unbeknownst to them) their disciplines were in crisis, and could only be rescued if they jettisoned their “foundations” and accepted Epstein’s application of “the sophisticated toolkit of metaphysics” to understand what they were even talking about. (p. 9). That theories or questions of actual social science were left out of the picture (and the book) didn’t bother philosophers, who mostly reviewed The Ant Trap glowingly.

Now it’s certainly true that, as Ladyman and Ross very sternly and thoroughly pointed out, philosophers have also been doing this sort of thing with the ontology of the physical and biological sciences more generally. But they’re more careful there. They evidently think that the poor benighted social sciences need especially conscientious bossing around by wise philosophers. Philosophers of physics or biology take it for granted that they need to know something about the subject they are providing with the ontology the scientists themselves thoughtlessly omitted, and mostly accept that their ontological supplements can’t conflict with what the scientists think they’ve tentatively established. It would be considered childishly anthropomorphic to think that the philosopher in her armchair could think up a better set of basic concepts from scratch for, say, chemistry than the chemist in her lab. 

Epstein’s equally anthropomorphic efforts, though applauded by philosophers, did not go over so well with social scientists. His book was thoroughly eviscerated by Robert Sugden in the Journal of Economic Literature, who pointed out its complete irrelevance to anything in economics, and ended his review with a famous quotation from Neurath on the dispensability of “foundations.”

Philosophers have responded by backing away gradually from the extremes of Searle’s and Epstein’s philosopher-king gambit. In 2019 Richard Lauer, for instance, criticized what he somewhat facetiously called “OM!” (“ontology matters!”) views such as Epstein’s and Searle’s. He distinguished what he called a “realist” interpretation of ontology from a “pragmatic” interpretation — actually referring to Carnap’s ESO to indicate that he has something like Carnap’s internal/external distinction in mind — and argued that the realist version is otiose, since it requires “ontological commitment” which goes beyond, and is unnecessary for, “empirical adequacy.” That sounds promising, but Lauer hadn’t actually succeeded in fully extracting himself from the ontological goo of our times. He went along with the idea (perhaps only for rhetorical reasons, who knows) that we should perhaps regard social ontology as prior in some sense to actual research in social science, and assumes without argument that social ontology can lead to greater empirical success or adequacy in social investigations. He ends his paper by saying “So, social ontology may be prior to social scientific methodology, but perhaps not social ontology as traditionally construed.” (p. 187) — which is not exactly a ringing rejection of ontology.

His paper occasioned another special issue in the same journal two years later, in 2021; the contributors take various positions along the lines of “social ontology de-dramatized,” but none want to rid themselves of ontology completely, even Harold Kincaid, who openly says that “Much of the current social ontology literature is on my view largely an unfortunate and sad extension of unfruitful analytic metaphysics.” (p. 41) While the Searle-Epstein gambit is quietly disowned, lip service is still paid to the supposed conduciveness of ontology to empirical success.

And that’s where we seem to be stuck now. Francesco Guala and Frank Hindricks have just published a programmatic paper called “The Nature and Significance of Social Ontology” in Synthese that exemplifies this stuckness, at least on the surface. They offer an “ecumenical” approach to social ontology, meaning that it should not be dictated solely by philosophers in armchairs (the Searle/Epstein approach), nor just by social scientists, nor by common sense and ordinary language, but by a mishmash of all three.

Social ontology should clarify what the social world is made of and how social entities are related to non-social ones, including psychological and physical entities. However, it should also reflect on the extent to which different images of reality – those provided by various branches of science as well as those that inform our everyday activities – are compatible. In Wilfred Sellars’s (1963) terms, it should in part be concerned with how the manifest image of the social world relates to the scientific image. (p. 145)

Sellars, however, sought a “synoptic view” of “stereoscopic vision” that could find a place for the manifest image within the scientific image; in articulating this desideratum he was stating a general form of what Huw Price would later call “placement problems” — problems about how to situate our “M-worlds” (morality, modality, meaning, and the mental, among other such “manifest image” things) in the naturalistic world view science conveys to us. Price’s solution was to distinguish what he called “object naturalism” (ontological naturalism) from a more modest, non-ontological “subject naturalism”; in other words he (explicitly) classified these distinct forms of naturalism as responding to Carnapian “external” questions and “internal” questions, respectively. (Also I had argued in a long-ago paper that Carnap’s program of explication offered a concrete proposal to implement precisely the kind of integration of the manifest and scientific images that Sellars was looking for; see esp. p. 346.)

Price argues very convincingly that the “placement problems” are insoluble if we insist on “object naturalism,” i.e. if we insist on making them ontological placement problems. Turning placement problems into ontological ones forces us either to abolish the “M-worlds” (as Quine more or less proposed) or to deny the sciences any cognitive authority whatever, an equally unappetizing option. But that’s where Guala and Hindricks seem to be headed — perhaps without even being aware of it.

For they combine the insistence on ontology (and thus object naturalism) with an “ecumenical” political correctness that allows neither physics itself, nor folk physics, nor metaphysics to have the final say on what there really is, but only some sort of colloidal slurry of the three. (How this tripartite compromise is to be arrived at they give no hint.) Perhaps they don’t mean their prescription to apply to physics, but nothing in their paper tells us (a) why their three-part ontological compromise should apply only to social entities; or even (b) how to distinguish a “social” entity from other kinds.

But Guala and Hindricks have no business invoking Sellars and his goal of a “synoptic view,” since for them it’s not a “placement problem” at all any more. They’ve shrugged off the problem of “placing” our “M-worlds” within the scientific image by telling us that all we have to do is add a bit of folk physics and a bit of metaphysics to the actual physics, and that way we’ll all get along — the tension between the manifest image and the scientific image will disappear! What that amounts to is that the scientific image shouldn’t be taken seriously at all (at least in the special case of the social sciences, though they don’t say why only in that case or how that case is even to be identified); it can be added to or subtracted from at will, without any cost in coherence or truth or empirical accuracy — we are essentially still back with Epstein, then, despite the emollient rhetoric of “ecumenical” harmony. The metaphysical wolf, now in sheep’s clothing, is back at the door.

What’s wrong with that? Well, we’ve seen that social ontology can only have an obstructive role in addressing philosophical problems about how to understand the social sciences, or to “place” them within, integrate them with, the physical, biological, and cognitive sciences. The invocation of ontology puts the “placement problems” raised by Sellars and Price completely out of reach. But social ontology also addresses no actual substantive problems within social science, or among the social sciences.

For instance. It seems obvious at a common sense level that the different social sciences are all addressing pretty much the same thing (the same empirical phenomena) from different perspectives, with different tools. But they largely refuse to acknowledge each other, and this mutual isolation is clearly getting worse. I was lucky to be living in Hyde Park back when Gary Becker and Jim Coleman were holding their joint evening economics-and-sociology seminars at the University of Chicago. Whatever you may think of their own work, both of them were at least seriously trying to bridge this gap! Who (of that stature, at least) is even attempting such a thing now? Also, a generation ago economic history was a thriving subdiscipline within history; the papers in the Journal of Economic History and Economic History Review were written mostly by historians, people who may have known a bit of economics, but mainly knew archives and Quellenkritik and wanted to do justice to the facts from those sources. Those journals and similar ones have almost entirely been taken over by economists (this was already under way in 1985, as Robert Solow bitterly complained); historians have turned their backs on economic history almost completely. I have myself (jointly with an economic historian) tried to address this particular rupture in some detail. I challenge those who think that social ontology can somehow assist or support empirical research to suggest how it could possibly even be relevant to that discussion, or to the kinds of solutions envisaged in that paper. Or for that matter, how it could possibly have any relevance whatever to this larger problem of growing mutual isolation among the social sciences.

Of course it may well be that this problem is itself largely an institutional one, arising mostly from the different incentive structures within different disciplines, and especially the rise of economics as the all-conquering dominant gorilla among them, due to its prevalence in B-schools and undergraduate business programs. Then it would be incumbent on philosophers to point that out and encourage social scientists to study this process empirically, rather than going on about whether dimes are physical or social entities, or enriching their social ontology with Margaret Thatcher’s (presumably “common-sense” or ordinary-language) contention that “there is no such thing as society.” They will never get the attention of social scientists with that (or any) sort of ontological inquiry. The Ant Trap is the worst sort of vandalism; insofar as social scientists took note of it at all, it has very likely convinced a generation of them that philosophy does not and cannot have anything at all to contribute to solving the problems they care about.

Philosophers may well shrug and say they don’t care whether social scientists (or any scientists, or any people) pay any attention to what they write. In that case, their increasing irrelevance is richly deserved.

Comment (I will respond, however old the post!):