H-Nationalism recently had the chance to talk with Nadav Shelef, the Harvey M. Meyerhoff Professor of Modern Israel Studies and Associate Professor of Political Science at the University of Wisconsin, Madison. His current research analyzes large collections of digitized text to track how different groups have articulated notions of homeland and uses GIS to chart the location of ethnic groups. David Prior, an Advisory Board Member at H-Nationalism, conducted the interview via email from Monday, April 20 to Monday, April 27, 2015.
DP: Nadav, your current research relies on a variety of digital tools, methods, and datasets. Could you walk us through these and explain how they link up to your research interests? What are the problems you’re working on solving and how are you going about it?
NS: I’m interested in how nationalism changes. That is, how and why do the rules change over time for deciding who is included in the nation, or for deciding the territorial scope of the homeland? My earlier work found that, in the Israeli context, evolutionary dynamics accounted for many, though not all, of the changes in how the main Israeli nationalist movements conceptualized the human scope of their nation, the borders of their homeland, and their national mission. My current projects ask how widespread such mechanisms of change are and if there is systematic variation in where or when particular mechanisms operate.
Answering these questions requires systematic cross-national data on the content and meaning of nationalism over time. Such data largely do not yet exist. Where they do, they are often based on assumptions (such as the static nature of nationalism or its uniformity within a population) that are out of step with the dominant theories of nationalism. As such, I have focused on developing better sources of data that could answer the questions I am interested in. For example, a systematic and cross-national exploration of how definitions of the territorial scope of the homeland change requires a measure of what populations believe their homelands to be. Unfortunately, most cross-national data on homelands assumes that these do not vary over time and/or use problematic proxies to identify them (such as the presence of coethnics, prior autonomy, or a territory’s legal status).
I therefore created a measure of the homeland status of territory that takes advantage of the nationalist imperative to rhetorically delineate the homeland’s boundaries. Nationalists speak of their homeland in different terms than they speak about non-homeland territory. A discourse-based measure of the homeland is theoretically consistent, but faces tremendous logistical and practical challenges – if only because nationalist discourse is in the local vernacular. To (partly) overcome this challenge, I used the domestic discourse about land on the other side of new international borders, as captured by the US government’s Foreign Broadcast Information Service (FBIS). FBIS systematically transcribed open source news broadcasts from almost every country between 1945 and 1996. As a source, FBIS has three main advantages: 1) its scope is nearly global; 2) its records are searchable by keyword; and 3) it translates all foreign news broadcasts into English. Because it provides reasonable access to how actors in (nearly) every case where a new border was drawn since 1945 talked about the land on the other side of this new border, discourse about territory captured by FBIS provides a proxy for identifying areas that the populations consider part of their homeland. The resulting data provides information about the scope of the homeland that is consistent with theories of nationalism – and is allowed to vary over time, across cases, and within them – that can be used to explore a range of questions.
Another example is posed by the assumption that the presence of co-ethnics in a particular location is relevant for a host of outcomes of interest (definitions of the homeland, irredentism, international conflict, etc.). With few exceptions the data used to explore these questions assumes that the ethnic composition of territory is static and that the relevant marker of ethnicity (religion, language, etc.) is obvious and unchanging. Often, however, neither of these assumptions is reasonable. Moreover, the most sophisticated of the projects dealing with the spatial distribution of ethnic groups is based almost exclusively on a single source: the Atlas Narodov Mira; a source which has its own well documented shortcomings. In order to build on these prior works, I am geocoding the presence of ethnic groups using ethnic maps that were created at different points in time, that use different definitions of ethnicity, and that were created by diverse actors (intelligence agencies around the world, geographers, ethnographers, etc.). The resulting data will enable scholars to consider how the changes in the ethnic composition of territory shaped outcomes of interest.
DP: So, to generalize, you are using a variety of digital tools to render a more accurate and detailed picture of ethnic groups, homelands, and borders. You are of course still working through your research, but do you have a sense of what some of its theoretical implications might be?
NS: A couple of important implications are already evident. First, the territorial dimension of nationalism (where is our homeland) is as important as the membership dimension (who we are) in shaping nationalist conflict and the two are distinct aspects of the nationalist phenomenon.
Second, it is possible to generate data that is both firmly grounded theoretically in constructivist understandings of nationalism that emphasize ideology, change, and the importance of the ways in which people themselves view the world and that can be used in positivist, quantitative scholarship. The synthesis between the two has proven to be fruitful ground, and there is much more to do.
DP: Do you anticipate that there are other areas of nationalism studies where this kind of creative development and analysis of data, especially data in tune with constructivist understandings, will advance debates? Are there specific research topics, other than of course those that you are already working on, that are particularly in need of such data?
NS: There is certainly the potential for doing so. There is a whole host of questions dealing with the ways in which multiple identities interact with each other and the ways in which they and these interactions shape outcomes of interest in the world. There is fantastic work being done using agent based modeling that is tackling the assumptions and implications of such multiple identities. Since it is now much easier to (literally) map out the relationships between people, network analysis may help us provide answers to unresolved questions about issues like the variation in the successful mobilization of ethnic or nationalist movements and the like. And issues like this are just the tip of the iceberg.
DP: Do you have a sense of how the sort of methodological innovations you’re discussing play out across disciplinary boundaries? Are some fields of study – either across the humanities and social sciences as a whole or within political science – more receptive to this kind of work than others? Have you encountered any resistance to your work?
NS: My sense is that the reception of such methodological innovation is not necessarily determined by discipline. The biggest resistance is generated by those on either “side” who minimize the possibility of bridging the gap between positivist approaches to studying ethnicity and nationalism and thickly descriptive, interpretive approaches, regardless of discipline. The former, often assuming a materialist foundation to nationalism, sometime dismiss the possibility or even utility of including constructivist-based variables in their analysis. The latter often privilege the unique aspects of individual cases, and sometimes dismiss the utility of seeking systematic relationships across cases or the ability of “numbers” to tell us anything interesting about the world. Unfortunately, too often scholars working in either tradition wall themselves off from the provocative (and destabilizing) questions asked by those in other traditions. Rather than dismissing an argument for ontological reasons, I believe we should engage with the implications raised by alternative ontological traditions. Those of us who work in the tradition of thick description need to take seriously concerns about how we would know if we were wrong, about whether our understanding of the broader phenomenon would differ if we happened to study a different case, and about causal inference. By the same token, those of us who work in the positivist tradition need to take seriously the role of socially constructed phenomena (and the process of this construction) in shaping the outcomes that we are interested in.