Monday, April 26, 2010

Hot off the RAND Press: Insurgencies are Complicated

Ignore the glib title above (I'm trying to make up for some of Gulliver's alleged turn to sincerity) and check out RAND's latest major study How Insurgencies End. Written by Ben Connable, a retired USMC Major, and Martin C. Libicki, it tackles an impressively broad swathe of insurgencies stretching back to the 1930s, and asks a bunch of interesting and highly policy relevant questions. I've barely begun reading it, but two things strike me.

First, the list of ways governments lose insurgencies (see p. 152) maps pretty well onto Afghanistan 2001-2009 (possibly 2010):

  • Governments ignore the insurgency until it develops into a credible threat.
  • Governments fail to address root causes.
  • Governments address root causes half-heartedly or too late, stoking discontent.
  • Governments fail to identify major shifts in strategic momentum.
  • Governments fail to extend credible control into rural areas.
  • Governments become dependent on a fickle sponsor.
Which is pretty depressing. More heartening, the study also argues that "No insurgency ending is inevitable." Let's hope so.

Second, their emphasis on the role of political/social/economic discontent is a healthy reminder to keep our priorities in line even as the Marines in Helmand fight to retain (or regain at this point?) the momentum, and the Kandahar offensive slowly builds:
The kind of grassroots support necessary to build and sustain an insurgency is fed on social, economic and political discontent. If a government successfully addresses root causes, it is possible to defeat an insurgency without defeating the insurgents themselves.
I do note with concern that the study consistently mislabels the Ugandan Allied Democratic Forces as the 'Arab Deterrent Force' (who were, um, Syrian, and in Lebanon), and claims that it lasted from 1986-2000 (the ADF in Uganda fought from 1996 until, arguably*, they were chased out of their sanctuary by a joint UN-Congolese operation in 2005). I suspect this is a small blip on their data, and not indicative of larger data or coding problems, but I'd suggest keeping an eye out for any other issues.

*Arguable, because the Uganda Army claims to have been hunting them down as recently as 2007.

28 comments:

  1. I haven't read the RAND report in much detail, but I thought I noticed, and an SWJ post seemed to confirm, that the dataset does not include civil wars. I thought that an interesting omission. I think it might affect results, too. Rather than simply the FM 3-24 emphasis on grievance as a cause of insurgency, perhaps expanding the dataset to include civil wars might emphasize more greatly the role of greed in civil war/insurgency. Is my thinking simply not robust - always quite possible - or am I spotting something perhaps important?

    ADTS

    ReplyDelete
  2. Well, they list all the cases they included, and all the cases they excluded, and there are plenty of what we generally call 'civil wars' in there - southern Sudan, Angola, Lebanon, China, Mozambique, Cambodia...it's a long list. They're also explicit about their inclusion criteria:

    "1. They involved fighting between agents of (or claimants to) a state and organized, nonstate groups that sought to take con- trol of a government, take power in a region, or use violence to change government policies.
    2. The conflict killed or has killed at least 1,000 people over its course, with a yearly average of at least 100.
    3. At least 100 people were killed on both sides (including civilians attacked by rebels)."

    But let me turn this back on you: what do you think the distinction is between insurgencies and civil wars?

    ReplyDelete
  3. MK-

    I'd just classify a civil war as a Phase III of a Mao's protracted war. In other words, a civil war is a type of insurgency. In a "simple" insurgency with two actors, host nation and one insurgent group- no external actors, this phase would begin once the guerilla massed enough military power, resources, land, etc to fight the government in the open. The difficulty is determining when it starts/ends.

    ReplyDelete
  4. That's interesting, Mike - I suspect many would have reversed the hierarchy, and imagined insurgency as equivalent to guerilla war, and therefore a subset of civil war in general. I'm not arguing that either way is 'correct' - I can see argument going both ways.

    That said, I'd query whether in cases where federations or empires break up and you already have full administrative structures and uniformed conventional forces on both sides, whether it makes sense to characterize it as a phase of Maoist guerilla war. I'm thinking here of something like Croatia seceding from the FRY.

    In any case, the RAND study does include examples of relatively symmetrically conventional fights - Eritrean secession, and the Sri Lankan civil war, for example.

    ReplyDelete
  5. MK-

    I try to keep the terms as general and broad as possibility to simplify the complex topic into a single model. As no two insurgencies are the same, no single case study will fit inside the model, but it will offer a general understanding. Of course, there are many that will disagree with me, and that's okay. IMO, the real need for research and debate is on how and when the government should react.

    I like to think guerilla equals insurgent. Basically, in the beginning, a group has an idea that they dislike the control of the governing authority. They seek to use violence to force certain changes, overthrow the goverment, or secede.

    In order to do this, they have to grow their base and mobilize in order to get people, guns, and money. 99% of the time, the gov't stops them early on or their ideas fail to take root with the local populace.

    IRT Croatia, I'd call it a secession.

    ReplyDelete
  6. MK

    "imagined insurgency as equivalent to guerilla war, and therefore a subset of civil war in general."

    That's about my definition, too, although I think we're all probably (a) sympathetic to coding subjectivity and (b) the fact that most cases probably are not that hard to code (if I read part of the report correctly, they farmed out individual cases to RAND SMEs - easy to do when you're as big as RAND, I guess).

    The US Civil War was, so far as I know, waged predominantly along conventional lines. We still consider it to be a civil war, though.

    I'd argue that excluding North Vietnam from the analysis (so, pre-1964), the Vietnam War was one of insurgency rather than civil war.

    The American Revolution fits Mike F's commment, though, nicely: it was not a civil war - a war between two sides within a present or past polity - and was not predominantly guerrilla per se, but perhaps fits Mao's Phase Three. How would we code that conflict? Insurgency? Civil War?

    Mike F, I've been thinking a lot about case studies of late: what constitutes one, and what constitutes a good one? You seemed (very) down on an SWJ post of late about quantitative dissertations. My own view is good case studies are like pornography: I know them when I see them. I'm probably off-topic here, because we're talking about coding cases, rather than exploring cases through small-N analysis, but I just thought I'd note how surprised I was at what seemed like pretty strong disdain.

    ADTS

    ReplyDelete
  7. ADTS-

    Actually, I think we're on the same page particularly with your first post.

    I was more frustated at the process rather than personal disdain for the authors or the academia. I was honestly hoping to find a great dissertation that could advance my thinking. In some cases, I'd email the authors. In other cases, we got to meet personally. The typical response that I got was, "Dude, you're right, but by the end of school, I was brain dead, and I was just hoping to finish the paper and get it approved." I definitely have sympathy with that, but I think we can do better.

    Social Sciences is an imperfect science, and the field of small wars has been rather small until recently. But, it seems like we're trying to force too much quantitative analysis to explain human nature. I took the question back to some old math and econ professors. They reminded me that the great thinkers in the field (Keynes, Nash) would try to simplify using assumptions that did not ever/rarely occured in real life. For example, perfect communication and fair games. MicroEconomics is based off those assumptions.

    On the other hand, I was working on a study that will take a long time and is still FOUO. We were coding 30 years of daily OPSUMs/INTSUMs in a conflict. With each report, I as the coder had to make decisions that could effect the entire project. Literally, I could shape the data to prove pop-centric COIN or CT approach. And, in most cases, the decisions were not clear cut. That's a lot of power and responsibility so I saw how data can be manipulated if one had an agenda particularly if it shapes policy. Luckily, the team working on this dataset is impartial. They are some of the idealistic non-partisan people that just want to find real answers.

    So, I've been trying a different approach of quan/qual analysis with broad definitions. We'll see if works. I actually sent SWJ a short piece on the Lifecycle of Small Wars today. I'm not sure if it's detailed enough for publication or not, but I thought it was a good time to get it out there. To keep myself sane, I watch Glenn Beck once a week and question whether or not I'm generalizing like he does :).

    As far as case studies, that is a tough one. Really good ones have a lot of drama like a good movie or Shakesperean play- violence, competing actors, betrayal, But, it also has to be well-documented. I'll give some more thought and provide a better response.

    ReplyDelete
  8. Mike F:

    Your response is fine. :)

    If I can ask, do they have a second coder working on the study you're working on, to ensure intercoder reliability (ie, if you and he/she agree on how a case should be coded, everything is cool, but if there's disagreement, a third-party is brought in)?

    I'm not sure we're trying to force too much quantitative analysis to explain human nature. I think that the attempts to explain insurgency and COIN quantitatively have flaws - forget about the Lyall and Wilson piece everyone keeps banging on, Fearon and Laitin's "Ethnicity, Insurgency, and Civil Wars" (I think the second or third top-downloaded article over the last X years from the American Political Science Review) was thoroughly based by Nicholas Sambanis - but that's the nature of the social sciences. Nothing is perfect. In fact, I think one could argue that's a virtue, not a vice of quantitative analysis: easier to falsify, and easier where to pick where it's wrong.

    As for what constitutes good case studies, the only recurring criteria that I could see are (and not all have to be met, except the first): absolutely no more information than is absolutely necessary; good knowledge of the empirics about which one is writing; some linkage to an existing literature or body of theory; some convincing explication of causal mechanisms. Again, though, it's an "I know it when I see it" type of thing.

    ADTS

    ReplyDelete
  9. ADTS-

    I don't disagree with you. I'm just trying to take a different approach. Here's one study that I would find fascinating.

    Compare and contrast the Civil Rights movements of MLK and Malcolm X.

    - One advocated violence and aimed towards secession. The other advocated non-violence and aimed towards equal rights.
    - One was from the South. One was from the North.
    - One was Christian. The other was Muslim.
    - Both leaders were African-American and both were murdered.
    - One was succesful. The other was not.

    To tackle this problem, I'd want an interdisciplinary team consisting of an anthropologist, sociologist, economist, pyschologist, and historian. Each one would tackle the problem quant/qual through their perspectives disciplines and methodologies. The final product would be hollistic and solid provided the team had good cohesion.

    WRT to the US Civil War, using the period from 1860-1865 makes it difficult to apply Mao, but that was the Phase III. The resistance actually started back in the 1770's. The Union almost fell apart in between the Articles of Confederation and the Constitution, and South Carolina tried to succeed during Andrew Jackson's presidency.

    WRT to the study I worked on, they have 3 coders that meet at least once a week with a panel of 10 experts and hash out the parameters to try and ensure a good product.

    ReplyDelete
  10. Mike F:

    Interesting approach, because I'm much more used to thinking in terms of, "What is the independent variable?" "What is the dependent variable?" and "What is the causal relationship or research question we're trying to answer?" and "How can we design the research or try and control variables to focus on precisely the one or ones [variables or causal relationships] we're trying to examine?" I "sort of" see you doing that with your Malcolm X/MLK paired comparison and sort of don't. I wonder how your hypothetical "team" would cohere - whether they'd decide to have an edited volume with each contributing a chapter, or actually produce a truly jointly written product, and whether the coherence would be a function of personality or disciplinary approaches (biases?).

    Your US history may well be better than mine, although I'd look at a recent article in Political Science Quarterly (Europe's Structural Idol - this Spring?) which, I think, argues that the Union was actually pretty solid between the Articles of Confederation and the Constitution; but then again, perhaps it's not a very well-done case study. :) That South Carolina tried to secede (something I know nothing about) would seem incontrovertible.

    ADTS

    ReplyDelete
  11. Great discussion, guys! Let me throw in on a couple of points.

    I'm so glad you guys highlight the coding issue. I actually come from a hard sciences background, and you'd be amazed the number of judgment calls intrinsic even there. Once you start digging into the philosophy of science, you realize that while the scientific method is far and away the best method we've got (and fairly reliable at that), it's also far from perfect.

    As for Sambanis' dissection of Fearon and Laitin - it's kind of ironic, because Doyle and and Sambanis' Making War and Building Peace was rife with coding errors, IMO.

    I also completely agree with the multi-disciplinary approach you suggest, Mike: academia has imposed a series of false divisions on the messiness of real life in this respect, and it's time to do away with them.

    Mike - I actually think your comment way back at 6:33PM argues for not eliding the difference between insurgency and civil war: the particular collective action problems faced by a group at various points in their evolution. I think Weinstein takes this a little too far (and oversimplifies a little too much), but marrying Weinstein with Kalyvas with PE approaches might start to get us somewhere. And then you've got to work social identity issues in, too. Hard to do while keeping the variables for a quantitative study manageable, let alone a methodology for studying them all.

    ADTS - I'll just say that when all is said and done, I tend to think controlled case comparisons complemented with some limited quant analysis is better than the other way round. I find it provides deeper and more nuanced insight, and avoids fetishizing either mode of analysis.

    ReplyDelete
  12. ADTS-

    I started out as an Economist, and I'm moving towards just specializing in Wicked Problems. This is one way to tackle a tough problem- bring in people who view the world differently.

    For starters, I'd have an economist and/or political scientist research the independent/dependent variables. I'd want to know:

    - What was the difference in violent attacks on MLK's group and Malcolm's X's group by aggravated whites?
    - What was the difference in violent attacks on Southern Blacks and Northern Blacks by aggravated whites?
    - What was the difference in arrests between the groups by local law enforcement?
    - What was the amount of reconnaissance, surveillance and harrassment by the FBI towards the different groups?
    - Etc, etc...

    Next, the anthropologist conduct detailed ethnographies, and the sociologists conduct their surveys. Then, the historian will be interviewing various groups to try and get the story right. Finally, the psychologists will examine the impact of the childhoods and time in jail on the leaders. We'd then sit down with all of the evidence and build a narrative describing the findings.

    Will this work better than a political scientist using the classic methodology? Maybe, maybe not, but I think the final version would be more comprehensive than a simple quantitative study that could lead to a better understanding on the topic.

    Back to American history, I doubt that I'm a better expert on the subject.

    ReplyDelete
  13. MK:

    I've read (see if you can google Mike Reay, and maybe with uchicago thrown in, or on the uchicago website, although I think he's at Reed College now)- in a study of economists that, in part, looked at whether they considered themselves scientists - that "scientists" (eg, natural/biological-physical) *themselves* do not consider there to be an objective "science" or objective scientists.

    As for arbitrary divisions between the social sciences, I couldn't agree more. I may have said this in this discussion, but on AM's board, I mentioned that I think we're seeing a merging of the social sciences. Two political scientists have won Nobels (including one this year), at least one psychologist has won and second one would have won had he not died before the Prize was awarded, behavioral finance and economics is burgeoning, etc, etc.

    I haven't read Weinstein the whole way through, and sadly, he doesn't follow the convention of an introduction, theory chapter, large-n, case studies, conclusion, so that it's relatively easy to skim. I assume by PE you mean political economy? I'm not sure I have a good sense of the synthetic methodological approach you're speaking of, but it may be telling that in his Perspectives on Politics review, Tarrow spent a considerable amount of time comparing the two works side-by-side (which I didn't think was warranted) and made some of the same points as you, I think (eg, social identity).

    As for controlled cases with limited quant versus the reverse, I think it - and this may seem funny given the methodological discussion we're having - doesn't matter. What determines your research method is your research question. I can think of a book off the top of my head that has case studies when I think it really doesn't need them, and that they were put there to satisfy convention and reviewers. I'm not a quant; I'm just saying if a quant approach is called for and that's all that's called for, then use the quant approach and only the quant approach and be done with it.

    Mike F:

    After my last email, I was thinking of fact-finding commissions (9/11, Challenger, etc) and how they function - do they bring in different people with different backgrounds, who determines what gets written and how, etc. I agree your approach would be more comprehensive; my only quibble is the role of the historian, although I think it morphs into something greater. Who's to say he/she can "get the story right"? My suspicion is each discipline and each practitioner of each discipline will think they have, if not a monopoly on truth, then a chunk of it. History is good - and arguably bad - for being atheoretical, and I'm not sure that there is a historical method per se. My reading (and listening to speakers) is (very) limited on the topic and I'm sure there's more to what they do and that some if not many would criticize or disagree with me on this point. But leaving that aside, even if the group is cooperative, each member has a story to tell, arguably one that could be termed comprehensive even if told strictly from the confines of his/her own discipline, and relegating that story to the historian seems something they'd be disinclined to do. If the anthropologist can write the ethnographies, who's to say that's not sufficient to tell "the story?" Ditto for the psychologist, sociologist, etc. Each could construct a "history" using the tools of their own discipline. I'm not anti-historian - I'm just pondering their roles.*

    ADTS

    *For a decent discussion of this, see perhaps "The Historic Turn in the Human Sciences," although most of it is written by non-historian social scientists, IIRC.

    ReplyDelete
  14. Mike F:

    Now I'm wondering *if* anthropologists, etc., *were* brought in on the 9/11 Commission. I know the head of the staff IIRC was a diplomatic historian (Zelikow) - maybe historians are best at synthesizing and writing a cohesive and coherent whole out of disparate pieces, and imposing quality control (ie, different people were assigned to write different parts of the 9/11 Commission)? I know there has been at least one ethnography of the Directorate of Intelligence, but it wasn't done in the context of 9/11, I think. Robert Jervis's new book "Why Intelligence Fails" has his post-mortem of the DI's predecessor WRT the Iranian Revolution and the DI's failure to predict it, but he is not an anthropologist or an organizational behavior specialist (even if he is a very smart man). I think the Challenger commission was far more eclectic and had at least one natural scientist and one sociologist. It'd be interesting to see the sources of personnel selection and the effects of personnel selection.

    ADTS

    ReplyDelete
  15. Mike F (and apologies to others - eg, MK - for hogging the board)

    I got the Political Science Quarterly article's argument backward: the interregnum between the end of the Revolution and the adoption of the Constitution were, rather, fairly rocky times, and definitely different times (ie, peoples' "nations" were their *states*).

    Just wanted to make the correction.

    ADTS

    ReplyDelete
  16. No need to apologize, ADTS - fire away.

    Don't know whether I'd characterize it as ethnography, but Jervis has also done some non-public work on organizational culture contributing to more recent intel failures (I can't recall if it was re 9/11, or the NIE on Iraq).

    ReplyDelete
  17. I didn't know Jervis had done non-public work on the topics at hand. I thought it was damn impressive - not that I'm an expert on FOIA, or security classifications and consultant/employee security/counterintelligence vetting - that he was able to get his Iran work publishable alone.

    I'm guessing Jervis's work probably does not qualify as ethnography, at least one that an anthropology would recognize as such. I don't think Jervis was trained in how to "do" ethnography, and I doubt he ever took the time to formally learn the technique(s) (why would or should he? he's been tremendous beyond belief with the tools he possesses). And parts of his Journal of Strategic Studies piece on the WMD NIE read like classic political science, eg, mentioning selection on the dependent variable and failure to address competing hypotheses.

    He may have been chosen for the Iran gig because of his emphasis on individual-level decision-making, although I suspect "any political scientist will do - they're all interchangeable, more or less" operated to some degree. Granted, one who specializes in international politics might be more appropriate than one who studies Congress. But I might think that superior to one who studies international politics, and who also has superior acumen in political psychology, nonetheless might be less optimal than someone who studies bureaucracy and public policy formation. Maybe - and this gets to Mike F's point - it would have made sense to have both a political psychologist (or just psychologist) *and* an organizational behavior specialist, and, assuming both of the above were either political scientists or public policy scholars, a sociologist and an ethnographer. And it's worth noting that Rob Johnston's ethnography of the IC seems to have received solid reviews on the Center for the Study of Intelligence website.

    ADTS

    ReplyDelete
  18. Great discussion overall. I appreciate it.

    MK-

    If you have a chance, could you provide a brief explanation on why there have to be judgment calls in hard sciences? I suppose some of that is my naiveté, but I always thought the hard sciences were black and white in terms of measurement. Maybe the arguments are in forecasting causality like with the great debate on climate change. Anyways, I was just curious. I think the post at 6:33 was from ADTS, but if not, it would not be the first time that I’ve confused myself.

    ADTS-

    A year ago in the Emerson piece, I wrote,

    “Conceptual blocks confound the most informed as the scientific method and unproved theorems cloud the framing. Specialization in learning separates emotion and utility as mutually exclusive. Is the function of my heart not intertwined with my brain?”

    I think this stills holds true. Sometimes, our specialization blinds us to seeing the problem or asking the right research questions. I’ve seen this in combat with some military leaders too set in how they thought war “should” be to realize the reality of what was going on. In academia, I’ve ran across a few brilliant professors that were so entrenched in their field that they lost sight of reality. In the medical field, particularly with TBI, you have to find a doctor who will look at the problem holistically. Otherwise, you get fed large doses of meds. On the other hand, those same specialists, when working in a large diverse group, can find the proverbial needle in the haystack that others will miss precisely because of their degree of insight within their field.

    Committees and collaborations are one way of tackling complex issues. The differing personalities and egos will spark lively debate that must be appropriately moderated. This idea is not mine. You see it within Small Wars Journal, inter-agency groups, and intelligence fusion cells. The most striking example that I’ve read about was Google bringing in musicians from the San Francisco Symphony to help them get another look at a tough problem.

    When trying to solve issues in Iraq, I constantly asked my own men, “What are we missing here?” That was my primary CCIR that we would challenge ourselves to consider.

    ReplyDelete
  19. How interesting! Over at zenpundit I kind mentioned that I didn't know - as a chemistry major that went on to medical school - how to approach studies in the social sciences.

    My first instinct is to ask - what is the correct question? Focusing on variables sounds a bit like worrying more about the method than first principles, but I don't know!

    ADTS - there is a great link over at zenpundit to "Statistics as the new Grammar." It's fantastic and deals just with the problems discussed here - how to we interpret large sets of data. Again, I want, as a physician, to go first to the data collection, and then wonder, "are we asking the correct question?"

    Basically I was questioning an OECD model regarding social mobility and its validity. I was surprised that in the chapter linked in the discussion, I couldn't quickly see how the data was generated, but I might have read too quickly and missed it.

    Isn't it fun to talk cross-discipline and how much we learn from it!

    MikeF: I like asking what "we are we missing here?" too. It's fundamental.

    ReplyDelete
  20. The judgement call in the hard sciences, partly, is likely in the measurements and means of measurements. Is that it MK?

    Still, you have a method to follow and you may collect data that can falsify your conclusion. Social science models seem strange to me because in science you can collect data that invalidates your hypothesis. That is the essence of a scientific study. How does that work in modeling?

    ReplyDelete
  21. correction: "falsify, or invalidate, your hypothesis"

    - Madhu

    ReplyDelete
  22. Madhu,

    While sitting in the Birmingham jail, MLK wrote, “in any nonviolent campaign there are four basic steps: collection of the facts to determine whether injustices exist; negotiation; self purification; and direct action.” I think this philosophy and methodology lends the best approach. The more that I consider, the more that I understand the amount of humility required to try and tackle the big questions.

    We have theories not laws, and I try to remember that Machiavelli was not a political scientist, Adam Smith was not an economist, and, in the very beginning, Albert Einstein was hardly a physicist. They were men trying to describe or influence the world around them.
    In one of my favorite prefaces, Maynard Smith’s Evolution and the Theory of Games, Smith admits that his insecurities and his doubt that game theory can ever be applied to human nature. While describing these constraints, he offers the ESS model that eventually relates game theory from nature to human interaction- competition between competing actors with limited resources.

    So that’s what we have- hypothesize, test, disprove, start over…reaching greater understanding until we find some truth. Or as Emerson put it, “until the time comes that the Creator allows us to see his methods.”

    ReplyDelete
  23. Mike F:

    "Sometimes, our specialization blinds us to seeing the problem or asking the right research questions."

    I couldn't agree more. I was emailing an acquaintance on a topic he's working on, and wrote, "This journal's *truly* interdisciplinary" because, well, most aren't. The only thing I'd say, though, is there, of course, is *some* value to specialization, not that you're arguing otherwise. I don't want to get into the Israel-Palestine debate or something controversial, but to use an example, I wonder if Mearsheimer and Walt would be on stronger ground if they studied *American* politics rather than *international* politics. In other words, when they talk about Congress and lobbies, they're talking about stuff they really haven't studied for much of their careers (Walt *does* discuss this a bit in his first book). At the same time - and in conjunction with what you write in the Emerson piece - perhaps such a provocative question could be asked only by two people *not* consumed by the debates between those who *do* study Congress and lobbies. I'd add to your list, although only the first (so far as I know) was trying to change the world: who was Marx? Weber? or for a far more recent (and relatively speaking, lesser) example, Robert Axelrod (a political scientist or an economist?). I imagine most would say the former, but I can't help but imagine he's a Nobel contender. My point is simply that people - at least some, as you've already pointed out - can't be pinned down too easily. Getting back to Mearsheimer and Walt, both are steeped, I think, in the idea of policy relevance: you can glean that in what they wrote and said prior to the Israel lobby conflagration; I bring this up only because of your noting those who not only studied but sought to "describe or influence the world around them."

    Madhu:

    At one end of the spectrum, I've seen the approach of "First you pick your question [derived from existing theories], then you pick your cases [which would be "data" in small-n qualitative research]." At the other end of the spectrum, I've known very informally and very casually an eminent sociologist who told me, "First I see what data I can come up and play around with, and then I see what theories I can link it back to" (which practically constituted a stunning admission to me at the time). I told that to another acquaintance who's a sociologist, and she said either, "That's what I feel like I can do" (because she knows the theory[ies] and data so well)" or simply "That's like what I do." It may be telling that the former person was originally a quant, although he moved into qualitative research and the latter is a quant, but I think in actuality, a lot of people start with their cases or their (numerical) data and then work their way to theory. There are standard datasets for quantitative research (Correlates of War, POLITY IV, Penn World Tables), and for qualitative researchers, they arguably "create their own" datasets by building case studies, although whether one can falsify a case study is, I think, a matter of ongoing contention and perhaps always will be.

    I don't think I'm the autodidact you are, Madhu, but yes, it's fun to learn other disciplines (at least to a degree - I stopped wanting to be an economist around the time when we started the income versus substitution effect in Intro to Micro, and I stopped wanting to be a doctor if not before, then certainly after Immunology. :))

    ADTS

    ReplyDelete
  24. MikeF:

    MLK is always good to read....

    ADTS:

    You seem pretty darn well, and widely, read to me, ADTS - at least from your comments here and elsewhere.

    I'll let you all in on a little something: I'm likely in very a different phase of my career than many of you (got my promotions, etc), and I specifically left my last institution so that I could focus more on teaching and teaching related-research. Thus, I am being introduced to qualitative literature for the first time and the discussions around here help me out. A lot.

    So, I'm not so much an autodidact as I've learned how to scramble after what I need to move from A to B to C, etc. There just isn't time or money in some academic hospitals, so this, my online friends, is some version of an academic seminar for me.

    ReplyDelete
  25. Thanks, Madhu. :) Hope/glad the transition is going well.

    ADTS

    ReplyDelete
  26. I think at this point, I've exhausted everything that I can possibly add to this discussion. Y'all given me a bunch to consider.

    Madhu- good luck. In the next month or so, I've got to make a career decision too. If I transition back to private citizen, I'm considering teaching high school for a bit. I'd like to practice some of these big ideas with a younger audience. I'm curious to see what can happen when we pose the "why" questions to those who are not yet specialized. That, and I want to coach football.

    ReplyDelete
  27. Funny! - I also have a big decision coming up in the next month or so (although I don't really want to discuss it online). As before, hoping/glad your transition is going well, Madhu, and best to you Mike as you go through your decision-making process. :)

    ADTS

    ReplyDelete
  28. Thanks for the good wishes, all, and good luck on your respective transitions too.

    (I'd like to add, also, that any group of students to have you as a teacher MikeF, would be pretty darn lucky, judging from the clarity of your comments.

    Given my college town upbringing, I had a certain percentage of transitioning-to-real-life-hippies as teachers. It was, uh, interesting.

    I kid: my teachers and professors have been wonderful for the most part. Still, the hippies did dopey stuff like play protest songs for us. And then I was all, "so if your generation is so awesome how come you're here and not solving world poverty or something?"

    We were mean kids.)

    ReplyDelete