Monday, March 30, 2015

Error and Blame, part 3

[Part 1 here]
[Part 2 here]

Sensemaking scholars are quite open about their standards. In her contribution on Karl Weick for a special issue of the Sociological Review on contemporary organization theory, Barbara Czarniawska suggests that "the many impressive stylists" in organization studies are likely "the result of [the field's] much bemoaned lack of clear standards." While I'm less impressed, I think she's right to suggest that the style of thinking in organization studies reflects such a lack. And I think she's right to focus on Weick as an example of what John Van Maanen has called "style as theory".

One of the most important consequences of approaching social science through its style, rather than its theories and methods in a more classical sense, is that the crucial question becomes not, "What makes our studies true?" but "What makes our studies plausible?" As I pointed out at the time, Jim March took this to (what I think is) excess, when he suggested that "The stories of novelists and social scientists are judged, in part, on whether they are credible, and it seems unlikely that the assessment of credibility is enormously different in the two cases" (March 2010, p. 69, my emphasis). It should be noted that March acknowledges in his foreword that Czarniawska has had a strong influence on his understanding of storytelling, and that he, too, selects Weick as an example to be followed. Like I say, he puts it in a somewhat over-the-top way: "How do we assess the credibility of stories told by Chekhov? Is it different from the way we assess the credibility of stories told by Weick?"

I've done a side-by-side comparison of Chekhov and Weick before to show that that's a pretty ridiculous argument, however half-serious it may be. Today I only want to deal with the specific sort of "credibility" sensemaking scholars seem to be striving for.

Czarniawska concludes her chapter on Weick by suggesting that sensemaking scholarship does not offer certainty, only plausibility. (Less charitably, we might say it offers not truth, but truthiness.) As if to anticipate the discussion we've been having here at RSL about emotional ambivalence, she quotes Weick approvingly when he explains what he is trying to accomplish:

Ambivalence, hypocrisy, inconsistency, and equivocality may be pejorative labels in times of stability, but they are markers of heightened awareness in times of transition. In times of transition people are especially sensitive to the fact that they talk reality into existence and need plausible stories to retain their successes in doing so. We all want stories that work, which is all that the managers are asking for. (Weick 2003: 381)

Like I say, sensemaking scholars are pretty open about their standards, even when there's much to "moan" about. (I'm sure they find me ridiculously pedantic.) Notice what Weick is saying here. Scholars, he tells us, are really only obligated to provide what "managers are asking for". Czarniawska puts it this way: "[Weick] turned the attention of organization scholars ... from the relevance of academia to the relevance of the field", presumably meaning the field of practice. Organisation theorists are at the service of practitioners, they are beholden to administrative standards, not academic ones. It is distressing to me to hear academics completely abandon their own autonomy like this, but it is also all too common. Even academics now use the word "academic" pejoratively.

But notice that the argument doesn't even work on its own terms. It presumes that all managers are asking for the same thing, namely, merely "plausible stories" that work in some unspecified sense. But what "works" for one manager may not work for another; and while some managers aren't asking for more than a good stories, others have an interest in knowing what actually happened. The Mann Gulch case that I've been talking about last week offers a perfect example. Weick has told us a story about what happened there and why the thirteen young men died. He claims that some firefighters (i.e., some who work "in the field") found his analysis to be more insightful than their own. But much of Weick's analysis focuses on what when wrong on the ground (with a few passing remarks about how the men were organized and trained) and he concludes (incorrectly) that the men were poorly led and finally panicked. (Maclean, who is Weick's only source, says they did not panic and that their leader performed admirably.) This is sometimes called "situated sensemaking" and studies of it have a very interesting effect on our understanding of events. We end up discovering what went wrong on the slopes of a mountain gulch, in the cockpit of a jumbo jet, or in the control room of a chemical refinery. That is, we locate the problems at the operational not political level. This is not at all trivial.

In his 1995 text book on sensemaking, Weick makes the provocative claim that the best organizational design would be to "do away with top management" altogether. Martin Kilduff has glossed this idea as an example of Weick's independence from "corporate interests and education regimes", i.e., of how irreverent he can let himself be about the "sacred cows" of higher-ups. But we're talking about the same person who, when telling the story of the death of thirteen brave young men in the line of duty, construes the policy that put them in harm's way as a "stubborn belief" they foolishly held on to for too long. Worse, he leaves out the part of the story when they in fact gave up the policy objective in question after wisely assessing the danger they were in. His analysis does, indeed, altogether "do away with top management", but only in the sense of completely ignoring all the clues Maclean leaves in his account as to the role it may have played in the death of the young men. I guess that may be "all that managers were asking for", at least some top officials in the US Forest Service in any case.

There's that quip in the Brad Pitt movie Troy, a line the Internet attributes to Franklin Roosevelt: "War is when young men die and old men talk." The least those older gents could do, I would think, is to accurately tell the story of how the young men died, and to include in their talk the policies that decided where they would be when they did.

[Continues here]


Anonymous said...

are you aware that boltanski, in "misteries and conspiracies", provides an account of how literary and social scientific "storytelling" relate to each other?

Thomas said...

No, but that book is already on my list for other reasons. Thanks for pointing it out.

Jonathan said...

Well it seems to me that confirmation bias is built in here. We can always find some literary analogy from some literary work that will support the way we think managers ought to manage--no matter what our bias is. Or we can "plausibly" interpret a literary work according to our pre-existing bias. Then what function does the work play? Is it simply a stylistically adroit way of making the same points that could have been made without the literary analogue?

Thomas said...

I think literature should be used in our thinking about management the way it is used in ordinary life. It should give us what Kenneth Burke called "equipment for living". Maybe it provides the stylistic adroitness without which a certain insight, and a corresponding managerial practice, would not be possible.

As Andrew Gelman and I argue, a story needs to be sufficiently "immutable" to challenge existing biases. But you've got a point. There are too many "degrees of freedom" in one's selection of the story. The trouble in this case, however, is that, even given that freedom, Weick has chosen to interpret his source in way that is simply wrong. Like you say, he didn't have to do that; he could just have found a different story where his theory would hold.

Andrew Gelman said...


You write that "he didn't have to do that; he could just have found a different story where his theory would hold."

But maybe that's not so easy! There aren't so many good stories out there, as we can see by the fact that the good stories we do have, are highly valued.

To say, "Why didn't Weick find a story just as good as the one he used, but which makes his point better?" is like saying, "Why couldn't the Democrats in 1992 find a candidate just like Bill Clinton, but without the sex scandals"? No such candidate was available.