«POLITICS & SOCIETY 10.1177/0032329204267290 BENT FLYVBJERG A Perestroikan Straw Man Answers Back: David Laitin and Phronetic Political Science BENT ...»
POLITICS & SOCIETY
A Perestroikan Straw Man Answers Back:
David Laitin and Phronetic Political Science
This article addresses three main issues. First, it argues that David Laitin, in a misguided critique of Bent Flyvbjerg’s book Making Social Science Matter for being a
surrogate manifesto for Perestroika, misrepresents the book in the extreme. Second, the article argues that Laitin’s claim that political science may become normal, predictive science in the natural science sense is unfounded; the claim is a dead end that perestroikans try to get beyond. Finally, the article proposes that political scientists substitute phronesis for episteme and thereby avoid the trap of emulating natural science. By doing so, political scientists may arrive at social science that is strong where natural science is weak: in the reflexive analysis and discussion of values and interests aimed at praxis, which is the prerequisite for an enlightened political, economic, and cultural development in any society.
Keywords: phronesis; Perestroika; David Laitin; philosophy of social science;
context; judgment I am grateful to David Laitin and Stephen White for pointing out that my book Making Social Science Matter (MSSM) captures many of the core themes in a perestroikan political science.1 I share the basic intent and argument of perestroikans and would be delighted if the book might help advance Perestroika in political science. However, where White provides a balanced review of the book, in the hands of Laitin I feel like the proverbial straw man.
I will address three main issues in what follows. First, I will show that Laitin misrepresents my work in the extreme. Second, I will assess Laitin’s proposed The author would like to thank Erik Albæk, Irene Christiansen, John Dryzek, Ido Oren, Tim Richardson, Sanford Schram, Georg Sørensen, Stephen White, Alan Wolfe, and the editors of Politics & Society for valuable comments on an earlier draft of this article.
POLITICS & SOCIETY, Vol. 32 No. 3, September 2004 389-416 DOI: 10.1177/0032329204267290 © 2004 Sage Publications 390 POLITICS & SOCIETY alternative to the methodology he claims I present in MSSM, his tripartite method, and “scientific frame.” Third, I will outline what I call phronetic social and political science, a methodology for the analysis of values and interests aimed at praxis.
LAITIN’S MISREPRESENTATIONSDavid Laitin’s main move in developing both his critique and his alternative is to distort my distinction in MSSM between phronetic and epistemic social science. Laitin equates phronetic disciplines with qualitative and narrative methods, whereas epistemic disciplines have formal modeling and statistics at their core, according to Laitin. He thus invokes the dualisms of qualitative versus quantitative methods, case study research versus large samples, and narrative versus formal modeling. This makes Laitin’s job easy in attacking soft political science and promoting his own hard methodology. But the dualisms Laitin calls upon are rhetorical devices that misrepresent what I say in MSSM.
Below I will present a number of examples to document this. For reasons of space, not all examples are included. However, because Laitin’s misrepresentations and misuses of MSSM are so far-reaching, I want to refute the most important examples in some detail.
Dominance or Balance for Qualitative Methods?
Laitin states, as one of the main assertions in the
to his article, Bent Flyvbjerg makes the best case for a renewed dominance for qualitative and case study work throughout the social sciences. (Laitin, 163).2 In fact, I do not make the case for dominance for qualitative and case study work. I make the case for balance and integration in several highly visible places in MSSM. The following is verbatim what I say about the case study and large
[My assessment of case study research] should not be interpreted as a rejection of research which focuses on large random samples or entire populations; for example, questionnaire surveys. This type of research is essential for the development of social science; for example, in understanding the degree to which certain phenomena are present in a given group or how they vary across cases. The advantage of large samples is breadth, while their problem is one of depth. For the case study, the situation is the reverse. Both approaches are necessary for a sound development of social science. (Flyvbjerg, 87, from the summary of chap.
6 on case study research, “The Power of Example”; see also 75, 79, and 83 for other examples.) BENT FLYVBJERG 391 Laitin continues his misrepresentations by criticizing perestroikans for having an “abhorrence of all things mathematical” (163). Following this line of argument he asserts, again mistakenly, Social scientists, [Flyvbjerg] concludes, should not even try to imitate the scientific method with fancy statistics and impressive regressions. (165) In MSSM I argue that social science probably cannot become scientific in the natural science sense, but I do not argue against statistics and regressions, as Laitin claims I do. In line with this misrepresentation, Laitin incorrectly presents my research on city politics and planning in Aalborg—which I use as an example of
phronetic social science in MSSM—as if it were entirely qualitative (Laitin, 170Here is what I actually write in MSSM:
In answering the question of who wins and who loses in the Aalborg Project, I carried out environmental and social impact audits using statistical and other quantitative analyses.
This was necessary for relating process to outcome in the project. Here as elsewhere, the sharp separation often seen in the literature between qualitative and quantitative methods is a spurious one. (Flyvbjerg, 196) At a more fundamental level, Laitin misrepresents my conception of phronesis and narrative. Laitin (165, 169, 175) writes that I see phronesis as narrative. This
is false as is the following conclusion:
Flyvbjerg’s attempt to separate out phronesis (as a kind of narrative) from its statistical and formal complements is radically incomplete. (Laitin, 175) Compare this with what I actually write about phronesis and quantitative methods
In my interpretation, phronetic social science is opposed to an either/or and stands for a both/and on the question of qualitative versus quantitative methods. Phronetic social science is problem-driven and not methodology-driven, in the sense that it employs those methods which for a given problematic best help answer the four value-rational questions [which stand at the core of phronetic social science; see below]. More often than not, a combination of qualitative and quantitative methods will do the task and do it best.
(Flyvbjerg, 196) Thus I do not separate phronesis from statistics or other social science tools. I argue for their integration. Nor do I see narrative and quantitative methods as opposites or as methods that stand outside each other. I integrate these as well, as documented by my empirical work, because it makes for better social and political science. In short, there is no factual basis for David Laitin’s claim that I make a 392 POLITICS & SOCIETY case for dominance for qualitative and case study work throughout the social sciences.
A Brief Example of Narrative Combining Qualitative and Quantitative Methods Let me give a brief example of how I work with qualitative and quantitative methods and with narrative. Laitin (170) says that it was my “deep understanding of all backroom deals”—gained through in-depth qualitative research—that helped make planning in Aalborg more democratic. This is incorrect. What triggered change was my relating the backroom deals, once I had uncovered them, with street-level, everyday outcomes—established through statistical and other quantitative analyses. For instance, through statistical analyses of large samples and time-series data on traffic accidents, I established that the backroom deals had transformed a planned and politically approved reduction in traffic accidents of 30% to 40% into a statistically significant increase in accidents of about the same size.
I then integrated these qualitative and quantitative analyses into one narrative together with legal and ethical considerations on democracy. When published, this narrative created considerable commotion among politicians and the public, because it made it uncomfortably clear that more people were killed and injured in traffic because city officials had allowed the local chamber of industry and commerce an illegitimate influence on planning outcomes during the backroom deals.
After this, it became impossible for officials to continue to practice the backroom setup for policy and planning. In this case it was the combination in one narrative of uncovering relations of power through qualitative analyses and linking power with outcomes through quantitative studies that helped make for more democracy.
This example is typical of my work and shows I do not conclude against statistics and regressions, as Laitin claims. I see choice of method as dependent on the research problem at hand. Sometimes, quantitative methods will best help answer the problem; sometimes, qualitative methods may do the job alone; and most often—if the problem is of practical-political import and the researchers want to make a difference with their research—a combination of quantitative and qualitative methods will be most effective. MSSM explicitly reflects this position, as does my book Rationality and Power: Democracy in Practice, which is my main
example of phronesis in MSSM.4 My most recent book, Megaprojects and Risk:
An Anatomy of Ambition, and many of my articles also make extensive use of both quantitative and qualitative methods.5 If Laitin wanted to criticize a purely qualitative researcher and use this to undermine Perestroika, as seems to have been his strategy, however misguided, he picked the wrong person.
BENT FLYVBJERG 393 Social versus Natural Science On the issue of social versus natural science, Laitin misrepresents MSSM like
The social world, [Flyvbjerg] argues, is sufficiently different from the natural world that any hopes for a Galilean conquest over the unknown in social science will forever remain unrealized. (Laitin, 164) Compare Laitin’s statement with what I actually write and emphasize in MSSM,
again in several highly visible places:
It is worth reiterating that [the] plea for the importance of context [to social phenomena] is not an ultimate proof that social science can never be explanatory and predictive. It only makes it probable that this is so. (Flyvbjerg, 46, emphasis in original; from the first paragraph in the conclusions to chap. 4, “Context Counts”; see also 4, 62, and 76) Laitin (168) further claims I do not define context in MSSM. In fact I use two chapters (chaps. 3 and 4) to establish that context in social science is human beings’ everyday background skills.
And the examples of misrepresentation go on. I will limit myself to only three more brief instances. First, Laitin (170-71) writes that I do not explain the influence of power in Aalborg. In fact, I present a historical explanation in terms of the longue durée in Rationality and Power similar to the type of explanation Robert Putnam and his associates use to explain power in Italian local government in Making Democracy Work, and I explicitly refer to Rationality and Power in MSSM (144-45) for the full story on this.6 Second, Laitin (168) says I do not provide any philosophy of science cites to support my claims regarding prediction as a criterion for science. In fact, the cites can be found on pages 38 and 175 in MSSM. Third, Laitin (167) claims I “admit” that Hubert and Stuart Dreyfus used a standard scientific procedure to make a discovery concerning human cognition.
Laitin even tries to lend credence to this claim by inserting a bogus endnote referring to where I am supposed to admit this (Laitin, 182, note 8). The problem for Laitin is that the two sources he quotes as evidence, both of which I authored or coauthored, do not even touch upon this issue, let alone “admit” to it. Moreover, one of the sources is in Danish, and I doubt Laitin has read this. Finally, it was not the Dreyfus brothers who made the discovery Laitin says they did, nor did they conduct the experiment or author the study that reported the experiment, as Laitin (166-67) wrongly claims. Laitin is as mistaken in the details of his article as he is on the larger issues.
394 POLITICS & SOCIETY Possible Explanations of Laitin’s Misrepresentations The examples presented above show that Laitin’s misrepresentations regarding MSSM are extensive and concern the core of the book’s argument. The examples document that Laitin has made up facts and results that are not in the book and presented them as if they were. The examples further show that Laitin has changed or omitted other facts and results so the research reported in MSSM is presented in Laitin’s article in a highly inaccurate and biased manner.
Why would Laitin make himself guilty of such misrepresentations? I see two possible explanations; Laitin may have made his misrepresentations in error or deliberately. First, error would explain the misrepresentations if, for instance, Laitin had not really read MSSM but only sampled it carelessly. If you are a hegemon, as Laitin says he is, and know your methodology is correct, you do not have to read critiques carefully, and over time you are likely to develop a tin ear for such critiques. Error would also be a plausible explanation if Laitin had operated with overly simple distinctions in his analysis of MSSM that would make him all too ready to script MSSM and me into the preselected role of narrative, qualitative villain. My problem with this type of explanation is I do not see a serious and experienced scholar make such mistakes. Making errors to a degree that would explain the misrepresentations documented above would imply a recklessness on Laitin’s part that would be no less problematic and unethical than deliberate misrepresentation; it would violate basic scholarly canons of reasonable handling of information and debate.