on being "in the field" (and that pesky task of rationalizing one's existence)

I have been spending the past six weeks 9,499 miles away from Ann Arbor in Dili, the capital of Timor-Leste (East Timor). I have another six weeks to go, so it seemed like a good time to reflect on some of the challenges I am facing as a researcher "in the field." (I put that in quotations because I am lukewarm about the phrase, in part due to its various negative connotations. See for example this recent post on the subject by a political science PhD student doing work in sub-Saharan Africa. My post here is meant to speak to some of these issues of researcher-subject / outsider-insider dynamics that are raised by those who suggest that "fieldwork" is an outdated and potentially harmful concept. Insert the usual disclaimer about how this post is only scratching the surface of points that others have made more thoughtfully elsewhere, and you probably won't ever learn anything from my blog, ever, but hey, you are here anyway so you might as well keep reading.)

Anyway, as an American social science PhD student trying to do dissertation research in Timor, I feel subject to competing expectations:

1. Sample, collect (and eventually analyze) your data in ways that minimize the role of your own subjective biases as a researcher.
2. Convince your (future) audience that you, as a scholar, have some knowledgeable insight into a place or community that your audience does not have.

Maybe it isn't obvious how these expectations are in competition with one another. I find that it becomes especially difficult with one, particular form of qualitative data collection: the semi-structured interview. I'll clarify. First, note that we all experience things subjectively. Everyone is in on the joke that complete objectivity is impossible, so let's just get that over with. Of course, there are ways to minimize (or at least estimate and correct for) ways in which subjectivity (a.k.a. bias) may be influencing our research. With a semi-structured elite interview, this is really hard. For example, what if  we wanted to know whether having a "knowledgeable" interviewer affected an interview subject's responses? (I'll elaborate on "knowledgeable" in a minute but, for now, say someone who has been reading up quite a bit on state revenue mobilization strategies in Timor-Leste interviewing a public finance expert in Timor-Leste, just as a random example.) What if they had been interviewed by someone who was handed the same list of guiding questions for the interview, but who had no knowledge of the topic? This counterfactual is almost always nearly impossible to achieve: you can't go back and interview the same elite twice (within a practical timeframe), and even if you outsource your interviews to local research assistants, for example, they are almost certainly going to be at least as knowledgeable as you.

There is a specific reason I raise the issue of the knowledgeable interviewer. Namely, I think that this is the basic question (or doubt) that floats around in the head of many more quantitatively-oriented (or less human subject data-oriented) scholars that may be the eventual audience for the research: "What if I had gone to do these interviews in East Timor, would the subjects have told me the same thing?" In this sense, your relative knowledge about a place or topic -- a quality that is generally held in positive regard and is used as the main justification for investing in fieldwork in the first place -- may make some people trust your work less, not more.

This is of course why there is a large literature on interviewer effects (and, errr, writing this post is making me realize that I should read more of it). There are ways to address some of these issues in interview question design, subject selection, and in the analysis and presentation of your findings. However, in general, semi-structured interviews are not as popular a form of data collection in comparative politics as they once were. Seemingly "cleaner" methods such as experiments and surveys are far more popular. Other comparativists avoid these problems by foregoing the use of human subject data altogether (a decision that is often motivated by the particular research question they are pursuing). Indeed, my own work will use a combination of human and non-human subject data.  Although I am not an experimenter, I reckon that well-planned experiments do require at least several visits to the field site by the researcher, but perhaps not an extended stay. Surveys can be conducted almost entirely from a distance. (Although, I came into town with a prepared survey instrument, and after two weeks I realized about one-third of the questions I was hoping to ask would be ridiculous or irrelevant. Some of this could have been obviated by more extensive outreach and preparation before I left for Timor, but I promise that I did a fair amount of that, and there really is no substitute for getting to the country, testing out one of your questions on an acquaintance, and being met with a blank stare.) And finally, non-human-subject data is often available electronically and may not require a trip to the research site at all (although, importantly, historical data and archival records are often an exception). Ergo, the decline of extended periods of field research in comparative politics?

So, I don't know if it should be reassuring or worrisome, but I think some of the difficulties I am experiencing are due to an overall uncertain relationship between fieldwork and social science these days. Perhaps less so than before, but it seems that departments still demand comparativists with country (or regional) expertise and knowledge. Presumably this means having spent a fair amount of time there, knowing a local language, having connections with local researchers, and/or being able to teach a class on the country or region. One's value-added is one's knowledge of a region or place that other scholars don't necessarily possess. But in order to make this valuable, one has to translate that knowledge into objective, replicable, unbiased findings. Scholars who are able to conquer both of these beasts are a rare breed worthy, in my opinion, of admiration and emulation.