Reply to Mahoney's Rebuttal: Hunting the Grail with Realist Enthusiasm

Alan Sica, Pennsylvania State University

The trouble with becoming antique, professionally speaking, is that one well remembers when a succession of "New New Things" were triumphantly proclaimed to have vanquished their predecessors, promising to set sociology on a higher road to greater accuracy, precision, good-heartedness, or whatever other goal the new protagonists cherished.  There are endless instances of this predictable, cyclical, somewhat disheartening phenomenon (with which some enterprising young formal modeler might amuse himself or herself).  My colleague, Duane Alwin, recently noted during a doctoral exam that what is now called "social capital" not too long ago was termed "social integration," and the great advantage for the doctoral candidate in adopting the newer term is that it allows the novice to overlook hundreds of studies which pursued the latter idea, although very similar sociological substance is at stake.  Similarly, for some time in the late sixties, it became an article of faith (sic) that Berger and Luckmann's primer on the sociology of knowledge—though entirely psychologistic in nature and noticeably conservative from a political point of view—had laid waste the longstanding edifice in that field created in the '20s by Max Scheler and Karl Mannheim, a claim which now seems peculiarly insupportable.  Its principal virtue for busy scholars was the welcome chance to substitute a short and easy text for long and difficult ones.  Within a few years it was also heard among the philosophically inclined that Althusser's so-called "structuralist" Marxism had made both the Gramscian as well as the Frankfurt school versions of same untenable; friendships ended before the dust had settled.  Roy Bhaskar's realism and Rom Harré's special brand of social psychology appeared in the late 70s as well, apparently making a shambles of more conventional ways of practicing the philosophy of social science.  Before long, from another part of the playing field entirely, rational choice theory migrated from its right-wing economic spawning ground into political science and sociology.  Meanwhile, through it all, the untouchable trinity of Marx-Durkheim-Weber continued to be taught, read, debated, and written about as if none of the competing sets of notions had ever flowered.

        Without any intention of condescending to Professor Mahoney's well-mannered, clearly drawn argument against my own, I must register some sadness that a young scholar of his obvious talents and theoretical creativity should be spending time in pursuit of a Grail that should have been sealed in its sepulchre long ago, and forgotten.  It reminds me distinctly of a seminar I took 30 years ago with Robert Leik (to whom Mahoney warmly refers), during which I realized that Professor Leik's mathematized sociology, as elegantly self-contained as it was in its own sphere, could never answer the questions about social reality which I thought most interesting, and which had been so often broached by all those theorists for whom "social action" had real meaning.  Leik was extremely good at posing questions which were amenable to his circumscribed methodology, but beyond those limits he did not care or dare to venture.  It seems Mahoney has strolled into a similar cul-de-sac.

        Mahoney draws fine distinctions between types of positivism and general theories, holding that I missed my mark in attacking the former, since it is the latter that he champions.  He may be entirely correct in the same way that I could argue—in almost perfect analogy—that the "interpretative" or hermeneutic approach to societal analysis when done under the varicolored flags of Dilthey, Weber, Schutz, Mannheim, Goffman, Garfinkel, or Gadamer, et. al. have less in common than an "outsider" might imagine, and that an attack, say, on the classical interpretive method of Dilthey means very little when evaluating more recent work in the same lineage.  Such a dispute turns on the evidence of finely tuned distinctions and a solid knowledge base which requires far more effort to secure than the ordinary "outsider" to such debates could imagine.  It is very clear that Mahoney has "done his homework" in coming to terms with certain key texts of the "general theory" camp to which he has attached himself.  And like many scholars who have found a new love or a new home, he gives much more than he gets.  He shows that demands for "clarity, rigor, and unity" (following Cameron and Morton, 2002), if pushed too hard, become unduly constraining, an argument with which, of course, I entirely agree.  But his next rhetorical move is fairly predictable: he puts me on one extreme end (formalization is bad), Cameron and Morton on the other (formalization is everything good), and then confidently posits himself between these two "extremist" positions.

        And yet if one were to draw a line and place all three parties on it, using a uniform metric for the line's entire length, Mahoney would be much closer to the formalization end than to my own; his is hardly a via media.  It would seem that his embrace of "clarity, rigor, and unity" (CRU)—"all of which enhance the ability of scholars to determine if their arguments are valid," as he puts it—begs the large question I put to him (perhaps less clearly than I could have) about the proper role of theory in sociology.  The reason he missed the point of the historical works he analyzed has nothing to do with CRU or valid arguments.  The reason that Social Origins of Dictatorship and Democracy is beloved, read, quoted, argued against, thought about, perhaps by now superseded has amazingly little to do with logic or proof or arguments.  For reasons that literary critics could explain better than I, it offers a truly persuasive narrative, in the same way that Proust is persuasive when he writes about the workings of memory or Weber was when he concocted his vision of the earnest early capitalist burgers.  No amount of "psychology of perception," formally stated, is going to dethrone Proust, any more than hundreds of counter-Weberian arguments about the origins of capitalism are going to render The Protestant Ethic useless.  If one wants ironclad proofs and steel-trapped arguments, one must look to the natural sciences or math.  And to the extent that social scientists inject into their "stories" this or that bit of mathematical or logical structure or argument, they are likely to find themselves pretty quickly leaving their intended audiences behind.  And not only because their audiences tend not to be very good at math, nor interested in its fruits. 

        Think of Coleman's Introduction to Mathematical Sociology (1964), a large book composed when he was still young.  Whereas everybody now must pretend to have read his late and even larger book, Foundations of Social Theory (1990) (in which formal mathematical expressions are not the main story), his earlier, more precisely mathematical attempt at paradigm-formation must now be judged a failure—if not intellectually, then at least as a book with a dedicated following (such as Barrington Moore's).  The reasons for this are obvious: readers interested in understanding human events like to hear about them in strong narratives; formalization lacks appeal because of its false rigidity, contrived clarity, and forced mode of expression.  All of this means that it offers weak persuasive power, since persuasion in the human realm is language-based, and not readily or frequently transmissible via formal means. (How else could one explain that George W. Bush's followers continue to believe, contrary to all available evidence, that Iraq was involved in the 9/11 attacks?  Such people are in love with Bush's soothing rhetoric, "and the data be damned.")  While it might be "nice" for those of us who adore "real proofs" to form CRU arguments at every turn, it is not going to happen; it did not for Stuart Dodd 60 years ago, and it will not for the political scientists Cameron and Morton now.  A small, unloved coterie of true believers will cling to this passionate ideal with all their might, and they will perpetually claim their imagined roles as the only "true scientists" among us.  Yet they will have no more ultimate success in persuading people of their correctness than Mother has had in convincing her children that Brussels sprouts are a "good" food sheerly because of their nutrient value. 

        Mahoney worries about "the scholarly community's current inability or unwillingness to model certain kinds of assumptions.  Furthermore, I believe that the failure to identify key underlying assumptions leaves an argument open to the legitimate criticism of lacking a high level of clarity, rigor, and unity."  I could agree with him, but it would make no difference to the way research ought to be or is likely to be carried out.  Narratives about human actions, solo or collective, have little to do with Mahoney's mantra of CRU; they have a great deal to do with analytic grace, empirical content, and persuasive rhetoric.  Without these, no book becomes a classic because it cannot speak to enough readers to bring them along for the ride.  Anyone who seriously reads Capital or Economy and Society or Division of Labor instantly recognizes that the vigorous use of language propels the arguments and dares the reader to disagree.  Which is why they are still in print long after tidy formalizations have been forgotten. 

        When Mahoney claims that, ceteris paribus, "an argument that is stated formally is better than one that is not stated formally," he is speaking a language that only fellow formalizationists could love.  For everybody else, it seems like unvarnished dogma, and with very slight likely payoff unless one is a member of that tribe.  He seems to sense this by admitting soon thereafter that "general theory encourages scholars to pursue formalization within reasonable limits."  Who is to define "reasonable"?  Mahoney says that I have "aimed my sights at the wrong target" by chastising followers of positivism, whose lineage he claims goes back to the Vienna Circle and later to the D-N model.  But how different is his "major epistemological break" from its artificially constraining predecessors?  How much would his acknowledged intellectual ancestors disagree with statements like "general theorists treat postulates as jointly sufficient for the truth of propositions" or "they advocate the use of probabilistic postulates that are inductively formulated" or "they use these postulates to derive postulates that are also probabilistic"?  One man's "major break" is not another's, and to me such claims have an eerily reminiscent tone that hails from around 1960 or so.

        When Mahoney says that I "seem to favor a 'nominalist' position in which social science concepts (e.g., state, gender, class) are deemed to be nothing more than mental constructs, lacking a true existence in reality," he is entirely wrong, perhaps because I misled him in my original comments to his article, for I surely do not hold to such a position.  There is nothing "nominal" about social class position or gender identity; they are "essentialisms" which cannot be escaped through sleight of hand or creative concept-mongering.  When Moore wrote about the ruling elites in Imperial China, or when Weber took apart the German Stände, they were not practicing nominalism.  They were portraying concretely realized historical relationships as efficiently and plausibly as they could.  Yet when Mahoney writes, "my realist position, by contrast, allows one to potentially grant the status of existence to social science concepts, including when they refer to unobservables," he is confusing the issue.  "Concepts" do not have "an existence."  They are analytic devices that interested parties agree to tolerate until something better comes along, which accounts for the unavoidably "heuristic" nature of social (or any) science.  The sentiment Mahoney expresses, to "grant the status of existence" to concepts literally makes no sense unless one wishes to practice reification on a truly heroic level. 

        More importantly, Mahoney has tied his wagon to the British tradition of Ian Hacking, Roy Bhaskar, and other "realists" (or "critical realists," to use the phrase attached specifically to Bhaskar's work), whose inspiration came not from watching sociologists work, as it were, but by studying the philosophy and history of natural sciences, especially physics.  While it is interesting to learn from Bruno Latour, Steve Woolgar, and many others that scientists in the lab behave as people rather than utterly rational automatons, by contrast the lessons one can take from the "realist" perspective are much more limited when applied to historical actors and processes, who, after all, are most definitely not bound by the norms of science that Merton spelled out decades ago.  What makes great sense to worry about in the realm of atomic physics—where, it could reasonably be argued it matters what is "real" and what is merely convention—bears little relevance to social science.  What becomes muddy in physics or cosmology is much less so here on earth, for even if nobody can agree that the "nation" is "real" or not, the fact is that humans live or die each day depending on whether that entity conveniently known as "nation" is treated as real by enough relevant social actors, particularly their armies and police.

        Put another way, in order to investigate, even superficially, (1) the meaning of "realism" as construed by Hacking, Bhaskar, and others (beyond the very limited issue of "unobservables" which somehow simultaneously exist and do not exist), and then (2) to figure out to what extent all of this might help Mahoney establish a CRU model of historical analysis, would require a very long paper.  Having read a fair amount of this material over the years, my feeling has remained that they are speaking principally to philosophers of language and not to everyday researchers, few of whom, so far as I can tell, have paid the least attention to these discussions.  Consider, for example, Bhaskar's sentiments in a recent interview: "...science was seen as a process in motion attempting to capture ever deeper and more basic strata of reality at any moment of time unknown to us and perhaps not even empirically manifest. . . And this put to the fore two notions which I think are absolutely crucial.  The notion of absence and dialectic was defined recursively, in terms of absenting constraints on absenting ills, and if constraints and ills alike are understood in terms of absence, as absenting absence on absenting absence" (Bhaskar, 2000: 2, 3).  Such argument may seem brilliantly illuminating to "ontologists"—Bhaskar's favorite self-identification—but to the comparative-historical researchers Mahoney interrogated in his article they appear thoroughly sterile.

        This is not to say that the realists have nothing to say of intellectual interest; it is more to argue that their contribution is on the meta-analytical level, and might well help scholars clarify their notions of what is and what is not "real," but empirically and conceptually, while at the same time holding limited utility for the sorts of works that Mahoney analyzed in his article.  In short, when Mahoney writes, "The burden of proof for the existence of an unobservable entity is thus grounded in explanatory criteria; unobservables that repeatedly generate successful explanations can persuasively be treated as real," he seems either to be mouthing a platitude and tautology—"That which seems real often enough becomes real"—or a fundamental error which begs the question by the use of slippery language.  What is a "successful explanation" and what does "persuasively" mean in this sentence?

        Closer to my own heart, though, is his following remark: "Sica's contention seems to be that social entities lack this mind-independent existence, which in turn motivates a post-positivist argument that sociologists should pursue a kind of interpretive or hermeneutical approach in which actors' subjective understandings are the central focus."  This neatly summarizes the different islands on which we live, Mahoney and I (and coincidentally echoes a question put to Bhaskar in the interview mentioned above: "Some philosophers argue that the realist versus anti-realist debate is one that will never be settled or achieve any genuine progress, since it is one that involves two utterly different world views and maybe two quite different sorts of ingrained philosophical temperaments" [Bhaskar, 2000: 10]).  "Social entities" are socially defined, and always have been.  Their "existence" beyond that consensus of intellectual opinion is nonsensical.  Where does "justice" live, or "happiness" or "gender equality"?  It's a perceptual matter.  I hold no brief for post-positivism, but I do think that Dilthey, writing in the 1880s, was most definitely onto something when he pointed out, thus inspiring Windelband, Weber, Simmel, and many others, including Mead, that both our perceptions of "reality" and our mode of analyzing same are grounded in the same fundamentally human mechanism: Verstehen.  Mahoney's implied model for theorizing and testing research questions seems to operate in an atmosphere which has lost this fundamental and, it seems to me, inescapable insight into all human affairs, including the operations of intellectuals as they try to make sense of an unruly world.

        I find that my annotations of Mahoney's response to my critique could easily inspire many more pages, but in an effort roughly to match the length of his statement, I will stop here, with the caveat that there is still plenty to say.  Perhaps it is worthwhile to end with a quotation from Bhaskar which Mahoney quotes approvingly, and for me indicates precisely the pointlessness of the former's "dialectical ontology" when analyzing comparative-historical sociology: "As Bhaskar (1986:52) puts it, 'the objects of scientific enquiry exist and act, for the most part, quite independently of scientists and their activity'."  For astro-physicists, surely; for sociologists, not likely.

References

Bhaskar, Roy 1986: Scientific Realism and Human Emancipation. Verso.

——— 2000. Roy Bhaskar Interviewed.  Questions by Christopher Norris.  The Philosopher's Magazine, Issue 8.
  http://www.raggedclaws.com/criticalrealism/archive/rbhaskar_rbi.html

Cameron, Charles R. and Rebecca Morton 2002: Formal Theory Meets Data.  Pp. 787–804 in Ira Katznelson and Helen V. Milner, eds. Political Science: State of the Discipline. W. W. Norton.

Coleman, James 1964: An Introduction to Mathematical Sociology. Free Press.

———1990: Foundations of Social Theory. Harvard University    Press.

Dodd, Stuart Carter 1942: Dimensions of Society: A Quantitative Systematics for the Social Sciences.  Macmillan Company.