Reading Recommendation: Abbott’s Methods of Discovery

Well, on the road to my specialization exams, which will be held in a couple of weeks, I continue to take reading detours that I can’t help.

The latest trip: I just finished reading Andrew Abbott’s Methods of Discovery: Heuristics for the Social Sciences and was blown away. Although Abbott writes that the book is intended for undergraduates, it can be read with profit by graduate students as well - for my own part, at any rate, I found it one of the more accessible and exciting discussions of ‘how to do social science’ that I’ve read in a long while.

Certainly, the book expresses a kind of sensibility that may not be to everyone’s taste. It’s certainly not for someone who believes that sociology (or any of the individual social sciences) should be devoted to the discovery of The One True Idea or The One Proven Method. Abbott finds such thinking stultifying and unnecessarily restrictive, when what the discipline needs to do is to excite the intellectual imagination and encourage its free movement. In order to do so, Abbott argues, we have to recognize that good social science draws upon a range of heuristics, which refer to the kinds of intellectual and classificatory operations and rhetorical moves that social scientists undertake in order to generate good research ideas and productive research agendas. Thus, scientists draw analogies, problematize the obvious, reversal causal sequences, engage in cross-disciplinary borrowing of methods, change contexts, identify latent functions, freeze or quicken time, and so on. While stated in these plain terms it may seem like Abbott is doing nothing more than rehearsing some familiar literary and rhetorical devices, he weaves in a number of well-known works from a variety of disciplines into his discussion, making a persuasive case for the importance of heuristics in carrying out meaningful, interesting, productive social scientific work.

He has a long chapter on fractal heuristics, in a kind of summary version of his earlier book, Chaos of Disciplines. I’ve been meaning to write my take on that book (also essential reading) as well, but I’m still in the process of digesting its insights for a paper I’m trying to write. Fractal heuristics refer to the kinds of internal divisions that re-emerge within intellectual positions or camps that have managed to set themselves up in opposition to other camps. The irony of fractals is that these internal divisions recapitulate the basic division, such that, for example, the division between positivism and interpretivism reappears within each camp. As a result, we see some positivists incorporating interpretive work into their research and findings, thus allowing for a contrast to emerge within positivism between those who allow for interpretivist elements and those who continue to have no truck with such material, and so on.

Abbott also does something really interesting with the menu of research methods that’s paraded in front of every grad student: are you going to do ethnography? institutional analysis? formal modeling? small-N? large-N? etc. Instead of offering up the familiar continuum between ‘qualitative’ studies on one end (e.g., ethnography) and ‘quantitative’ analysis on the other, Abbott uses Charles Morris’s categories of pragmatic, semantic, and syntactic relations to create a kind of 3-D space for arranging different research methods and approaches. The three dimensions intersect where the concrete categories of ‘everyday common sense’ resides, and the different research methods and strategies are classified by the extent to which they abstract from common sense and the dimension along which they do so. Semantic approaches, for instance, emphasize the assimilation of particulars into general patterns of signification, while pragmatic approaches focus on identifying and isolating specific causes for social phenomena and events. It’s really a very neat reconstruction, and the imagery that one gets is of the different research methodologies moving and orbiting one another, like galaxies, within a shared universe. This allows for a much more nuanced understanding of how near or far different methodologies may be relative to one another, suggesting possibilities for explanatory combinations that are more complex than just what the familiar quantitative/qualitative divide offers.

My only question is whether this book explains how social science is done or how it ought to be read. Surely a few of the works that Abbott discuses reflect metodological approaches that may be more apparent in retrospect than were consciously held or deployed. Nevertheless, even as a primer in how to read, the book teaches a lot. It encourages one to become almost playful with the possibilities of multiple observational and data-gathering approaches. The question, from a lowly grad student’s perspective, is how to recover this sense of the imagination within the conventionalized routines of institutionalized research.