Sunday, August 01, 2010

Malcolm Gladwell and/or Karl Weick

In his response to my criticism of Gladwell over at OrgTheory, Brayden King made an important observation that crystalized something for me. I've been aware of it for some time, but it suddenly struck me as very telling about recent work in organization studies.

Lots of completely legitimate academic articles are liberally sprinkled with “premature conclusions or misleading anecdotes.” I don’t see them as harmful as you do in either case. The point of much empirical work is to push theoretical boundaries and to get people to think. Gladwell is doing the same thing, the main difference being the intended audience.

I think it's important to keep in mind that Gladwell is in no position to "push theoretical boundaries" (whether he is able to "get people to think" is another issue, and one on which we probably also disagree). Gladwell does not master a theory and neither does his intended audience. His thinking, and that of his audience, is simply not relevantly "bounded" by theory and therefore no amount of anecdotes will push against those boundaries. They are only constrained by their imaginations, which is formed largely by the sort of imagery Gladwell excells at producing.

Now, Teppo contributed a link to a special "Meet the Person" section of JMI (mirrored at the Ross School's website in this PDF) that was published in 2006, devoted to Malcolm Gladwell. The contributions to that section also expresses a basic admiration for what Gladwell does, and in various ways laments the inability of academics to do the same—hence our reliance on popularizers who "translate" our work into more accessible and applicable language. Kathleen Sutcliffe and Timothy Wintermute, however, counter that many academics are already doing what Gladwell is doing, namely, "storytelling", and they are doing so in their academic articles, which brings us back to Brayden's point.

Sutcliffe and Wintermute offer Karl Weick's famous Mann Gulch study as an example of "research grounded in strong stories". But here's what struck me when thinking about it today: Weick is actually doing the opposite of what Gladwell does. Weick obviously couldn't popularize the story because Norman Maclean had already done that with his, let's say, "middle-brow non-fiction" book Young Men and Fire, which was Weick's only source of information about the disaster. Actually, "middle-brow" might not quite do it justice; Maclean's book is sometimes a bit overwrought in its prose—it is a very "literary" book, in my opinion. Indeed, by his own account, Weick tried to "strip Maclean's elegant prose away from the events in Mann Gulch" in order to make them available for "analysis". Where Gladwell takes the results of science and turns them into popular stories, Weick takes popular stories (even before Maclean's book, Mann Gulch was the stuff of legend at least in Montana) and turns them into scientific results. He even went so far as to recast Maclean's personal quest to understand what happened in Mann Gulch as a "methodology" and described his account, ultimately, as "data".

There are more recent examples of this. Ryan Quinn and Monica Worline's analysis of the "collective courage" of the passengers on UA93 (published in Organization Science in 2008) is based largely on Jere Longman's Among the Heroes (2002), which is a pretty standard work of popular journalism. They also reframe this material analytically, giving their study a "methodological" air and providing them with "data" for analysis. Here again, then, we have an academic text that begins with a popular account of a narratively compelling event and, if you will, "scientizes" it in order to "push the boundaries of theory". (I have written about what Quinn and Worline mean by "theorizing" before.)

So Brayden (and Sutcliffe and Wintermute) is right to say that academic work sometimes looks a bit like Gladwell's. But he is also right to say I find it harmful. Surely, it is problematic to use a popular narrative about "heroism" as "data" for a study of the social conditions of courage.

In the case of Weick's analysis of Mann Gulch, however, things are even stranger. Here Weick appears simply to spin a yarn out of his source material. His analysis depends crucially on two factual claims: the men (a) suffered from "positive illusions" about the seriousness of the fire and (b) finally panicked when those beliefs turned out to be false. Maclean's book, however, tells us that they (a) revised their view of how soon they would have the fire under control immediately after engaging with it and (b) did not panic (he says that explicitly). So "strong storytelling" comes to mean, essentially, writing fiction and passing it off as fact.*

Now that, of course, is something that Weick shares with Gladwell, as I (and many others before me) have pointed out in relation to the results of psychological research. Every time I dig into the sources of Weick's anecdotes and reports of scientific studies I find new reasons to doubt his scholarship. Like Gladwell, however, Weick will no doubt continue to be celebrated as someone who "gets us thinking", albeit, I would argue, not very critically. Perhaps we can call it, to use Gladwell's subtitle, "thinking without thinking". We just sort of "get it". In a flash. Without knowing why. Or whether it's true. Blink.


*I mentioned this briefly in my recent paper in Culture and Organization, but a more detailed critique should be coming out soon in Armstrong and Lightfoot's anthology of practical criticism, 'The Leading Journal in the Field', to be published by Mayfly.

No comments: