Almost every major work I’ve read in the past few months (years?) has invoked some variation of the twin phrases “the question is not settled” or “further research is called for.” On the one hand, I find these hedgings admirable – knowledge is as much a process as a product, and the types of questions considered are often far too broad to admit complete answers (“why do organizations exist?”, say, or “who controls the modern corporation?”). In addition, the hedging serves a couple tactical purposes – both allowing for alternatives without a loss of face and suggesting more is yet to come from the authors themselves. And yet, after reading variations on these phrases over and over, I am left to wonder: when is a question settled? When is further research uncalled for?
Another way to think about this question is to ask, what questions in Sociology are settled? On what questions would further research be considered invalid, unnecessary and wasteful? None come to mind, none of the biggies anyway. Rather, the set of questions seems always to increase – organization theorists, for example, have moved from models that considered mostly organizations in isolation (e.g. Taylor’s scientific management) to models that considered the organization and its environment, conceptualized as anything from the set of other organizations with which it interacts through exchanges (in some versions of resource dependence theory, say) to the field of all organizations in the some societally recognized area of life (i.e. Dimaggio and Powell’s 1983 concept of field). Overview texts that attempt to synthesize these various perspectives inevitably call for the inclusion of more – more variables, more empircally studies, more theories, more typologies.
When do we toss something out? When do we decide that something does not matter, or that its effects are now well-enough understood? Do we ever? Is this one of the (many) ways in which the social sciences are unlike the natural ones? It seems like the major works I can immediately conjure to mind in Physics, say, solve a problem: Einstein’s general relativity explained the orbit of Mercury. It also created new problems, problems that scientists have spent decades now working on. But the question of Mercury’s orbit is resolved, or at least, kicked up to a much higher level, general problem. Are there similar examples even in the most empirical works of social scientists?
When I read contemporary summaries of now 40-year old classics (such as Thompson’s 1967 Organizations in Action), I am struck by how these accounts reduce the classics to a set of a few propositions, and often put them in conflict with the propositions of other classics (Mizruchi and Fein’s 1999 piece on the fate of Dimaggio and Powell’s 1983 concepts of isomorphism attempts to prove just such a point). Here I am not so concerned, as Mizruchi and Fein were, with whether these works were properly remembered but rather the inevitable simplification. It seems in the social sciences we are unavoidably forgetting as much as we are discovering. For another org theory example, consider all the calls made to look back at the “old institutionalism” and its virtues (e.g. Stinchombe 1997, Greenwood and Hinings 1996). Perhaps the problem is that our intuition about the social world we live in is so good relative to our capability to formally and consciously grasp the nuances of that world that we are condemned to such remembering and forgetting.
I am not suggesting that no progress has been made – clearly, the accounts of organizational action, change, birth and death in the past 30 years feel a lot more convincing, nuanced and useful than most of the accounts from the first half of the century. And the set of questions raised in the 2nd half of the 20th century alone were worth the price of admission, so to speak. But I’m still not sure what we actually think we “know” about organizations, that is, what is settled?
One more question, I suppose, is under what conditions would these questions become settled? By the death or retirement of the proponents of competing models and their replacement with a newer generation? By the dominance of major journals by figures sympathetic to only one particular perspective? By agreement by all parties on the ‘right answer’ to a given question? As I ask these questions, I wonder if they are the wrong ones. Does the idea of ‘settled knowledge’ make much sense here?
Even if it does not, I suppose I should just let the ritual calls for more research serve their ritual purposes (with hints of Meyer and Rowan 1977, of course). What harm do they do? Still, it all feels a bit unsatisfying. So it goes.