I’m looking for your help: I need references which discuss the idea of uncertainty in qualitative research. Probably in social science, but maybe history.

Here’s the point I’m trying to make.

When we tackle empirical matters – how many people have HIV, why the dinosaurs went extinct, how democratisation affects economic growth, and so on – we can never know the answers for certain. (I’m not thinking about prediction, by the way, but about description or explanation of things in the past or present.)

In quantitative social science, this idea is standard: it’s central to statistical inference. But I don’t know how much it’s been discussed in relation to qualitative research, aside of course from debates over Bayesian research. I have looked . . . but I haven’t found much.

The place of uncertainty in qualitative research is something I tried to theorise in an article in History and Theory. I argued that when we study historical texts, we often ask empirical questions, such as why Machiavelli wrote what he wrote, or what Mill meant by ‘harm’. We can’t know the answers for certain, but often we should indicate how confident we are in our findings. This reminds us that we are not telling our readers what happened: we are telling them how strong we think our evidence is.

Reporting uncertainty in qualitative research is thus subjective, whereas in quantitative research it is objective (at least, where the indication of uncertainty is part of statistical significance).

But can anyone tell me who has written about uncertainty in qualitative research, whether in social science or history?

My ideas about this issue have been greatly influenced by Gary King, Robert Keohane and Sidney Verba’s Designing Social Inquiry – see especially pp. 7-8 and 31-2 of chapter 1. Unusually, they depict uncertainty as a core feature of science. This is a crucial idea. It took me years to grasp what they were getting at, but I now agree.

However, King Keohane and Verba actually say very little about what uncertainty involves in qualitative research, as Larry Bartels notes. This is surprising, given that their book is meant to be precisely about what quantitative researchers can teach qualitative ones. When I wrote my article, I had to do much of the thinking for myself (helped by Collingwood, by Keynes, and of course by many actual examples of good and bad practice in substantive research).

I’m now interested in writing a paper about uncertainty in qualitative social science. Of course, the idea is widespread: for example, it’s implicit in any discussion of triangulation. But do you know of people who have theorised the idea and/or discussed its place in qualitative research? (Again, aside from Bayesians.) Can anyone point me to some references? I’d be very grateful – thanks!