Now *Open Access*: What Does it Mean to Be a Kin Majority?

My recent article for Social Science Quarterly, What Does it Mean to Be a Kin Majority? Analyzing Romanian Identity in Moldova and Russian Identity in Crimea from Below, is now open access. You can read and download the article freely on SSQ’s website.


This article investigates what kin identification means from a bottom-up perspective in two kin majority cases: Moldova and Crimea. The article is based on ∼50 fieldwork interviews conducted in both Moldova and Crimea with everyday social actors (2012–2013). Ethnic homogeneity for kin majorities is more fractured than previously considered. Respondents identified more in terms of assemblages of ethnic, cultural, political, linguistic, and territorial identities than in mutually exclusive census categories. To understand fully the relations between kin majorities, their kin-state and home-state and the impact of growing kin engagement policies, like dual citizenship, it is necessary to analyze the complexities of the lived experience of kin identification for members of kin majorities and how this relates to kin-state identification and affiliation. Understanding these complexities helps to have a more nuanced understanding of the role of ethnicity in post-Communist societies, in terms of kin-state and intrastate relations.

View on Wiley Online Library

Citation: Knott, E. (2015). What Does it Mean to Be a Kin Majority? Analyzing Romanian Identity in Moldova and Russian Identity in Crimea from Below. Social Science Quarterly, 96(3), 830-859.

“You should write an empirical paper” – an interpretive reposte


I don’t regret the methodology of my research – agency-centred, bottom-up – because it can engage with the very actors who are often left out of political science. It can also engage with actors on their own terms, in their own words, whereas surveys (I argue) largely verify researchers’ deductive ideas. However this does not mean I’m immune to concern about how I’m positioned within the political science, the dominant thrust of which remains positivist, if not highly quantitative. When it comes to the big conferences of the discipline (e.g. APSA, MPSA) in my unrepresentative experience, 90-95% of papers given feature quantitative analysis. So, where are all the people doing qualitative research?

Perhaps they’re being convinced not to do it. I was recently advised to write or co-author an “empirical paper”. I retorted that I do write empirical papers, I have data, it’s qualitative data, it’s ethnographic data, and it’s certainly data! It was gathered through blood, sweat and tears. It was transcribed, translated and coded. It was then whipped together into chapters, which I’m iteratively revising.

What they meant was, I should write a quantitative paper, replete with regressions to show I’m not anti-quantitative (which, inherently, I’m not). My argument is different: I can’t trust existing sources of quantitative data in my cases. I don’t trust existing sources of quantitative data, whether censuses or citizenship statistics. They’re gathered both in highly politicised contexts (in both my cases, Moldova and Crimea) and by politicised, interest-driven kin-states (Russia and Romania) where the regime has an interest to collect a certain kind of data to fulfil a certain kind of state-building and/or nation-building project. If the gold standard of positivist data is objectivity, then using these sources of data falls extremely short of objectivity. Rather it tells us far more about regimes than it does about social actors within these regimes .

I’m also critical of surveys. Not least because I’ve done them and found them wanting but because they work, I argue, primarily to verify and test academic assumptions about the world. They’re not good at exploring the grey zone, or informal practices, or how things actually work. Which is why, I argue, the bottom-up, agency-centred perspective helps to get at aspects of politics that usually are difficult to capture.

So I don’t regret the methods I used. I think they generated interesting data but I wish I didn’t have to explain that it was data, meaning-rich, context-rich data, that can’t be reduced to numbers and regressed.