Qualitative Transparency Deliberations

[Nancy Hirschmann The University of Pennsylvania]

I confess that I feel a little out of place contributing to this website, because in one sense DA-RT has nothing to do with political theory. The fact is that transparency is a problem for quantitative political scientists, because it depends on data that is not itself explained in the course of the argument. By contrast, the whole process of writing a political theory paper is to display to the reader how we achieved the interpretation that we are advocating. Political theory is about making an argument, persuading readers, and generally doing so through arguments made in the paper itself and by quoting political theory texts, whether primary (e.g. Rousseau’s Discourse on the Origins of Inequality) or secondary sources (E.g. Judith Shklar’s Men and Citizens: A Study of Rousseau’s Social Theory). The argument will succeed or fail based on our ability to persuade the reader of the validity of our argument. The entire text of the article displays the workings of our own minds as we do the work. No quantitative article, no matter how much data is posted, is as transparent as that. As for Data Access, the texts that we read are available to anyone. Even if we use obscure archival material, much of that is increasingly online; and even material that is not available online can be accessed by anyone who can go to those same archives and read those materials.

When this point is missed, well-intentioned efforts to figure out ways to “fit us in” to DA-RT under the guise of “inclusiveness” may lead to misguided results. Some, for instance, have suggested that “active citation” would be appropriate; and while it may, in some cases, provide theorists with the opportunity to add lengthy footnotes in which longer textual passages could be cited from original texts, in the vast majority of cases it would run up against issues of copyright infringement; certainly, many canonical texts are on the web, but they do not provide definitive and in many cases accurate renditions of texts, even those originally in English: punctuation, italicization, capitalization, at times even spelling of words in their twentieth-century incarnations rather than their original spelling present inaccuracies. Differences among translations in different editions of texts originally in Greek, French, Italian, etc. are vast and different authors rely on different translations (including their own). And even if particular editions of various texts come to be accepted as “authoritative”, they are almost always copyright protected. Active citation in most cases simply makes it even more difficult than it already is for political theory to appear in the discipline-wide journals, and may have a chilling effect on submissions for no good intellectual reason.

But this specific problem points to a macro problem, which concerns me even about this very website: the acceptance of DA-RT as a given for the entire discipline of political science, giving way to earnest discussions of how to make it work for the different subfields. This point, also raised by Jeff Isaacs in his editorial essay in Perspectives on Politics, obscures a broader conversation that has been going on all year, but has not been listened to. So let me state it clearly: DA-RT may be highly applicable to quantitative political science, even if it poses some serious ethical and professional problems, particularly for junior colleagues who hope to gain multiple publications out of data sets that have taken them years of graduate and post-graduate work to accrue. But it is simply not applicable to large swathes of the discipline, such as political theory, most interpretive work, and considerable subsets of qualitative empirical work.

So am I saying “leave political theory out of this”? Yes and no. Yes, in the sense that this problem is irrelevant to us: we have always been transparent, it is the essence of the work we do, and anyone can read the same texts our work is about. But no, in this more complicated sense: journal editors need to move away from considering political theory as an “exception” to the “norms” of political science as defined by the principles of DA-RT, which are designed primarily with quantitative data in mind, and are being bent and twisted to apply to other forms of data. Rather, editors have a responsibility to recognize and acknowledge, publicly, that they understand that political science research has many different sorts of norms that are appropriate to their respective subfields, and work should be evaluated by and in terms of those respective norms by individuals in those subfields.