Reading :: After Method

Originally posted: Tue, 26 Oct 2004 21:56:59

After Method: Mess in Social Science Research

by John Law

Just over a year ago -- I remember it was on a plane to SIGDOC, so it must have been mid-October -- I wrote an email to a group with whom I've been discussing research methods. The email, based on my recent readings in science and technology studies, suggested that we should stop thinking of research in terms of hierarchical categories (paradigms contain methodologies which contain methods, etc.) and start thinking of all these components as enacted practices that occupy the same level of scale. In fact, I'll be presenting on the topic at CCCC 2005, and Kristie Fleckenstein is writing an article on this same topic (almost certainly better considered than my presentation will be).

As I say, the inspiration was the work being done in science and technology studies. And one of the most active scholars in that field is John Law. So I was happy to find out that Law was writing a book along roughly the same lines that our group had been discussing. I finally got the book in the mail the other week, and it doesn't disappoint. Law really gets that "mess" in social science research -- that is, the inability to draw this research into hard boundaries with unequivocal answers -- is insoluble and ultimately undesirable. He also gets the problem that I've attempted to wrestle with elsewhere and that I've called the problem of unintegrated scope: the impulse to impose levels of scale so that one level can be construed as a master level of which the others are just effects. In After Method, he takes on this problem by describing what he calls "method assemblages." He gives two definitions:

In chapter 2 I defined this for the case of representation, as the enactment of a bundle of ramifying relations that shapes, mediates, and separates representations in-here, represented realities out-there, and invisible out-there relations, processes, and contexts necessary to in-here. In chapter 3, I offered a parallel definition appropriate to objects: that method assemblage is also the crafting of relations that shape, mediate and separate our object in-here, its relevant context out-there, and then an endless set of out-there relations, processes and all the rest that are a necessary part of the assemblage but at the same time have disappeared from it. As is obvious, the two are similar in form. But the post-structuralist philosophical tradition suggests a different vocabulary. If we use this then method assemblage becomes the enactment of presence, manifest absence, and absence as otherness. More specifically, method assemblage becomes the crafting or bundling of relations or hinterland into three parts: (a) whatever is in-here or present; (b) whatever is absent but is also manifest in its absence; and (c) whatever is absent but is Other because, while it is necessary to presence, it is not or cannot be made manifest. Note that it is the emphasis on presence that distinguishes method from any other form of assemblage. Note also that to talk of crafting is not necessarily to imply human agency or skill. The various ethnographies we have explored suggest that people, machines, traces, resources of all kinds -- and we might in other contexts extend the list to include spirits or angels or muses -- are all involved with the process of crafting. (p.84)

Clear enough? It might help to know that Law says that paradigms (the most abstract level of the research hierarchy) are embedded in craft skills (the most concrete level). That is, the hierarchy of research is an attempt to make sense out of a flat, uniscalar set of practices/enactments. And once we get that, it makes sense to think of assemblages as self-assembling elements that are entangled rather than constructed (p.42). When we hear people talking about, say, ethnography as a methodology, a method, or a paradigm, we may become agitated if we are used to thinking of these as separate levels of a hierarchy; but if we think of the whole assemblage as uniscalar, it makes a lot of sense that ethnography would slip across these categories. It makes a lot of sense that methods such as experiments can lose and gain elements until they become, say, usability testing.

I hesitate to call Law's book a Copernican revolution, but it does make sense of these incoherences that we so often see in research when people try to define and operationalize it. Despite Law's style -- which varies from lucid to obtuse and self-indulgent, alas -- this book is an important book that draws together and reinterprets many of the STS threads that have been developing since the early work of Latour and Woolgar. Buy it, read it, quote it. >

Blogged with Flock