In lieu of an abstract, here is a brief excerpt of the content:

Social Forces 83.2 (2004) 491-501



[Access article in PDF]

Why "Unobservables" Cannot Save General Theory:

A Reply to Mahoney*

Pennsylvania State University

In James Mahoney's article, "Revisiting General Theory in Historical Sociology," published in Social Forces, volume 83, number 2 (pp. 459-90), certain foundational claims are made, including these, all of them knowingly inflammatory:

The aspect of a general theory that is general is its use of an abstract causal mechanism that exists outside space and time. (Mahoney 2004b:459)

A general theory is a postulate about a foundational cause that features two components: a causal agent and a causal mechanism. (Mahoney 2004b:459)

These mechanisms are empirically underspecified, exist outside specific spatial and temporal boundaries, and cannot be directly observed . . . they are original movers or "ultimate causes" . . . For example, instrumental rationality — the causal mechanism of rational choice theory — is an unobservable property devoid of precise empirical content and specific time/place referents . . . The positing of omnitemporal and unobserved mechanisms that serve as primitive causes has been discussed primarily in the natural sciences [In the penultimate version of the article, Mahoney's supporting citations included the following in note 2: "A lesson from the natural sciences is that one should not evaluate general theories based on the empirical plausibility of their causal mechanisms" (Mahoney 2004a).], but the same practice applies to the social sciences. Taken together, a causal agent and a causal mechanism represent the "hard core" of a general theory, or that part of the theory that is shared by all scholars who use it. This hard core is not usually directly tested. . . . a general theory is not itself a testable hypothesis. (Mahoney 2004b:459-60)

Resources [in "power theory"] are potentially unobservable bundles of ideas and materials. (Mahoney 2004b:461)

When a general theory is used, the starting postulate assumes a given causal agent and mechanism. Since the causal mechanism refers to an abstract property, [End Page 491] subsequent postulates are "bridging assumptions" that add empirical content to this mechanism (Hempel 1966:77-82).

(Mahoney 2004b:464)

Insofar as one is committed to scientific realism (which is the case for most proponents of general theory, including mid-twentieth-century philosophers such as Karl Popper, Carl Hempel, and Ernest Nagel), one treats all postulates — even those that make reference to unobservable entities — as reflecting an independent reality (emphases added). (Mahoney 2004b:482)

Most tellingly, in a paragraph of the penultimate version of the article (dated March 10, 2004) which was deleted in the final iteration, perhaps in response to my already written critique of it, these remarks appeared in order to summarize the importance of "outcome explanation":

Scholars using this strategy test the postulates because they contain the hypothesized explanation for the outcome. Notably, however, the initial postulate refers to an abstract causal mechanism that cannot itself be tested . . . In the end, nevertheless, the analyst must make a leap of faith in the existence of the causal mechanism, since this beginning postulate is never directly tested, and since the final proposition (i.e., the outcome of interest) is also not tested.
(Mahoney 2004a:16, emphases added)

Thomas Aquinas would have appreciated these remarks, as would any of his brighter, scholastic colleagues in the thirteenth century. In fact, they might well have written them, since they were "realists" and not "nominalists" (even without the tutelage of the twentieth century "neorealists," G.E. Moore, Bertrand Russell, Samuel Alexander, et al.). There is a touchingly medieval quality to the statements quoted above, and even though I cannot debate each point here, they do invite oblique examination.

Whenever I read an article of this general type, certain images come to mind, as if I had been asked to write a screenplay which might capture its unintended meaning. To wit: while their dedicated, officious assistants flutter about, I picture James Mason (playing Karl Popper) and Rod Serling (Carl Hempel) — both of whom serve as ceremonial godfathers for the article's main thrust — gently lording over a summer conference at some gemütliches chalet in Austria around...

pdf

Share