Where social science seeks knowledge, the humanities seek understanding. While the social sciences stake their credibility on their theories and methods, the humanities stake their credibility on their style. Pure forms are hard to find, of course. Many social scientists have humanistic ambitions—roughly speaking, literary ambitions—while many humanists have grown envious of, especially, the theoretical sophistication of their peers in the sciences. For the past fifty years, the language of the social sciences (the appeal to theory and method) has been actively supported by a network of opinion leaders and funding bodies. This year, a central institution in that network, the Carnegie Foundation for the Advancement of Teaching, proposed a reorientation in a more humanistic direction. My concern, as always, is with the effect of such reorientations on the way we write the results of our research down.
(An aside: Fabio Rojas recently posted a link to Clifford Geertz's interesting first-hand account of an academic career as it develops its initial literary ambition into a pursuit of a "common language for the social sciences" [PDF].)
If the central conceit of the social sciences is that they have a shared "program of perception" (a theory) and a proven set of procedures (a method), it is the central conceit of the humanities that a good style makes such things unnecessary. The humanistic conceit is sometimes promoted within the social sciences; John Van Maanen's "Style as Theory" is probably the best example. But it's important to keep in mind that the style is here proposed precisely as a theory, and Barbara Czarniawska has, rightly, taken this proposal to have important "methodological" implications*. Here the boundaries between the social sciences and liberal arts are certainly blurred, but I think it is safe to say that this kind of rhetoric is intended to allow (Czarniawska uses the word "permit") us to use a notion of "style" to underpin both our methodological and theoretical discussions. That is, we are still dealing in theory and method, we're just using style to sell them.
The more radical proposal is to do away with theory and method, replacing both, simultaneously, with style. This, I want to suggest, means writing not as one knower to another (one social scientist to another) but as one thinker to another (one humanist to another). What is presented in the writing is not knowledge but understanding. The presentation will still consist of a series of claims, and many of these claims will be expressions of "justified, true belief" in coherent paragraphs. So, yes, there will be lots of knowledge in the text, and a humanist remains a very a knowledgeable person. But the style of the writing, not formed by theory and method, is very different. The reader is not expected to believe, but to think.
What I am making explicit here is in many ways the standard defense of a now-familiar kind of work in the social sciences. When I challenge the epistemic foundations of sensemaking research for example, I am often told that it was never meant to be "true". But it must be stressed that sensemaking research—like the kind of journalism that Malcolm Gladwell practices—depends on a reader who will will take the style of the writing as a sign of its credibility (to use James March's word), i.e., as an implicit theory and method, and who will then essentially believe, or "trust" (Czarniawska's word)*, the text. It presents the results of reading as though it were the results of data-collection, i.e., as though the reader does not have access to the sources. If the reader were being addressed not as a social scientist but as a humanist, a more careful kind of scholarship would be required. Sensemaking research is written in the voice of a humanist addressing a social scientist, the voice of someone who claims to understand something reaching out to someone who knows something (else).
I think that if we're truly going to take a turn towards "liberal learning" in business scholarship, we need to begin to write as humanists to one another. What would that mean? Well, it would mean discussing what happens in the books we read as though our readers read those books too. We would not read a novel or a work of popular non-fiction on behalf of our peers in the social sciences; we would read along with our peers in the humanities. We might say that we should address the reader as someone who has the time to read; the social scientist, by contrast, is presumably too busy (engaged in "empirical research") to read books. I can see that I'm going to have to write another post to make good on my promise to offer some practical advice for writers. More on Tuesday.
*See "Karl Weick: Concepts, style, and reflection", published in the Sociological Review.