Friday, February 11, 2011

Mind the Gap

I often read papers that begin with a "gap" in a literature. These papers try to render their research question "interesting" by focusing on phenomena that have not yet been studied, or have not been studied closely enough, even though they are clearly relevant to existing bodies of theory. The approach, then, has a certain intuitive plausibility: interesting phenomena that have not been studied closely should, we would think, be studied closely. But what the authors of these papers forget is that it is not the research question that is publishable or not but the research result. In fact, I find that papers that begin with a gap in the literature often fail to articulate a clear thesis in the introduction (or elsewhere, for that matter). Instead of "showing that" something is the case, these papers tell us that they will "explore how" something happens. They construct the "interest" of their paper independently of their result. They make it seem as though any description of the hitherto under-examined object will count as a "contribution". They imagine the gap is simply "empty" and ready to be filled any which way.

In their recent critique of the "gap-spotting" approach, J├Ârgen Sandberg and Mats Alveson suggested that researchers should instead use their results to challenge assumptions in the literature. This is the right approach, and it can in fact easily be combined with our eye for gaps in the literature. All you have to do is realize that if there is some object that is already relevant to a theory, and that object has not been seriously studied by researchers, then the object is already 'covered' by the assumptions of the theory. That is, even in the absence of detailed study, the theory makes assumptions about your object (a particular kind of organization, say, or a particular aspect of organizational life). Since the object has not been studied, however, it is reasonable to hypothesize that the theory's assumptions about it are not wholly accurate. That's really the question your research should be addressing: how do our assumptions about objects in the gap hold up when we actually look at them?

The mistake that gap-spotters make is not considering that an object we have not yet examined closely may turn out to be completely uninteresting once we do. So you have to ask yourself in advance (as part of articulating your research question) what does the theory expect you to find out about the object you have chosen to study. If you find that those expectations are wholly warranted then you don't have much in the way of a publishable result, and do notice why that's entirely reasonable: a theory covers all cases of a particular kind—a potentially infinite number of cases; but it is based on research into particular cases—a finite number. The theory makes a number of assumptions and then claims that "all things being equal" (ceteris paribus) the as-yet-un-examined cases will illustrate the same theoretical truths as the already examined ones. You want to show that those assumptions need to be modified; if you can't, you have simply legitimized the lack of detailed study of the object in the gap. Past scholarship was simply right not to take an interest.

The gap is not just a source of fresh material that is somehow pre-approved for publication because it has not yet been studied. In an important sense, there is no gap. If the object is interesting at all, the theory has already implicitly (and sometimes explicitly) filled in the gap with its assumptions. Your theory of organization already claims to understand the organization you are studying, even if no researcher has yet looked at it. The "gap in the literature" is full of ideas that have been produced by generalizing the results of previous studies. So you have to think the gap through on your field's behalf. You have to make the implicit assumptions about your object explicit; you have to acknowledge that your peers already know a great deal about what you may have thought of as your personal object. You have to mind the gap, not just mine it for raw data.

* * *

[Update at 7:03 am] This post took 49 minutes to write. It consists of 692 words composed into four paragraphs. This confirms some basic quantities quite nicely. I normally say that a trained writer who is at the top of his form (as I arguably am) should be able to produce a prose paragraph of about 200 words in under 30 minutes. My paragraphs aren't quite that long, but they have been written in under 15 minutes each. They are also a bit chattier than really publishable prose would be (though not altogether sloppy, I hope you'll grant.) As regular readers know, my training regimen consists in sitting down every other morning (Monday, Wednesday, Friday) between six and seven to write a post on a topic I've decided on before going to bed. There's no secret, no trick. You just have to practice.


Jonathan said...

It's as though you breathed in four or five paragraph essays.

Thomas said...

Thanks, Jonathan. That's very kind. It reminds me of this exchange:

Robert Duncan : . . .Inspiration. You can't learn that, nobody can teach it to you, you either feel it or you don't.
Allen Ginsberg: I think you can teach inspiration.
RD: Teach inspiration?
AG: Taking it literally, inspiration being a matter of breath, you can teach breathing.
RD: Oh, breathing, right. And you can teach vowels.
AG: And if you can teach breathing then you can teach a certain body looseness and mind freshening-- (Allen Verbatim, p. 109)

Alex Vozny said...

At first I thought to disagree: every existing hypothesis has to be tested to become a theory, so testing the hypothesis (even to prove that it was correct) is a necessarily step (often this is a required routine that we do not want to but have to do, to support our more interesting claims).

On the other hand, this is exactly what I hate in most of the papers others (and often me too) write. Routine work.

So yes, in terms of looking for an interesting topic, we have to look for contradictions rather than for confirmations.

The other common problem is also that we find interesting (unexpected) results starting from a completely unrelated motivation. Serendipity?

Thomas said...

I like to think that much of what I post here has that character. At first you feel like disagreeing, but then, when you think about it, what I'm saying makes a lot of sense.

I want to emphasize that this post is about a way of presenting your results in writing, not about a way of generating them. There is no shame in rationally reconstructing an interesting problem around a solution that emerged from looking for something completely different.

All I'm saying is that what we sometimes thinking of as "gaps", blank spaces on the map, are really filled in by alot of presumptions, and investigation will often challenge those presumptions. That's normally what makes it interesting.

Alex Vozny said...

Agree with the writing, still would like to understand more about initializing the research with question - still need to look for gaps? It's a pity I have no access to Sandberg's paper.