theory@orgtheory.net

I'm grateful to Teppo over at orgtheory.net for bringing up Donald Hambrick's objections to a supposed "theory fetish" cultivated by management studies. Hambrick has published an essay called "The Field of Management’s Devotion to Theory: Too Much of a Good Thing?" in the Academy of Management Review (2007, vol. 50, no. 6) and a shorter piece in BusinessWeek. This post is mainly a reposting of the comments I left over at orgtheory.net.

Hambrick points to a difference between the frequency of the word “theory” in the ASQ and the Journal of Finance. My off-the-cuff analysis of this difference is that finance is more theoretically inertial than organisation theory (or management studies) as a field. Putting the point in Kuhnian terms, I would guess that finance has a much more stable set of "symbolic generalizations" and much less problematic "metaphysical models" than we have in organization studies. His statistics, predictably, show that theory is not as much of an issue in finance as in organization theory.

But Hambrick also talks about a paper in the Journal of Marketing on the “doppelgänger brand image” (AMJ, p. 1347). I’d have a difficult time defending my point about symbolic generalisations in the case of branding “theory”; but Kuhn might still help us to understand the absence of overt references to theory with the notion of "shared values". (I'm going to have to devote a week to theorizing soon.)

Hambrick’s epidemiology example (AMJ, p. 1348) is not very well chosen, to my mind. After all, what is epidemiology but the study of “empirical patterns” of illness? What theory would the editors demand beyond the background theories about the relation of symtoms, death rates, etc., etc., to common causes? Now, what if a journal of cancer research rejected the paper? Well, then they’d be right to. Without a link to a theory about the causal mechanism, the merely disturbing pattern is not publishable in that field.

Hambrick complains that "straightforward tests of existing theories usually don't qualify" as a contribution to theory (1350). He's right about this; they don’t. But that’s because the only "straightforward" test is one that confirms the theory. So, since theories are already the received view (or at least an acceptable position), confirmations are not interesting. Another confirmation doesn't change anything. Disconfirmations, however, imply (at least the need for) modifications of a theory. That's the real connection between testing and theory; and its the good reason "simple tests" aren’t publishable.

I won't get into the very problematic idea of "facts without theory" (1348), except to say that the importance and interest of "important and interesting facts" depends entirely on your point-of-view, i.e., your theory.

Finally, Hambrick blames the theory fetish for bad writing. “Our insistence on theory,” he says, “has caused a lot of bad writing.” Nonsense! There’s lots of bad atheoretical writing out there. And lots of great theoretical writing. Theory does not cause style; it simply sets up the stylistic problem for the writer in a particular way. So does the task of "documenting and dissecting a fascinating, important phenomenon" without theory (1348).

Bad theoretical prose does have recognizable characteristics. Hambrick's epithets ("contorted" and "ponderous") are apt. But the fact that theoretical concerns often lead to a different kind of badness than empirical concerns does not prove that theorizing itself is to blame for that badness. Writing badly is the cause of bad writing.