Methodological Unity, Revisited June 20, 2009Posted by Will Thomas in Methods.
Tags: Allan Franklin, Imre Lakatos, Jed Buchwald, Kent Staley, Peter Galison, Steven Shapin
In my recent look at historiographical language, I discussed Kent Staley’s 1999 critique of Peter Galison’s division of particle detector history pre-1970 into epistemologically discontinuous “image” and “logic” traditions. I noted that Staley might have made his point less palatable through an appeal to the methodological “unity” of science (rather than contiguousness), which Galison jumped on as contrary to the construction of coherent history. For Galison, it is the methodological divisions in science that keep it nimble, intellectually diverse, and heuristically powerful, and it is an appreciation of this disunity that allows us to make any sense of its history consistent with a detailed reading of the historical record. What, then, is the appeal of philosophical “unity”, and do historians have anything to gain by letting philosophers espousing it in the door?
If, indeed, philosophers believed philosophy could be used to reconstruct an algorithmic history of science, then surely they should be kept out at all costs, but this does not appear to be the case. Staley, for one, appears to seek conceptual clarity rather than to follow Imre Lakatos’ notion of the “rational reconstruction” of history—science may be epistemologically diverse, but underlying epistemological connections may be revealing of the sources of the strength of certain kinds of knowledge-making acts: “We might entertain the following version of the ‘unity of methods’ thesis: there are a small number of forms of argument that are shared among otherwise diverse areas of investigation, or that are employed in common during otherwise distinct historical periods of scientific endeavor.” For arguments to have power, they employ certain underlying strategic (rather than historically deterministic) forms that have a transcendental quality and that can be identified in history wherever they might occur.
The prospect of a transcendental notion of “science” that can be identified in history is contrary to post-1980s historiography, which has understood its cogency to be rooted in its historiographical insight that “science” is a product of—rather than something to be distinguished from—surrounding culture (Steven Shapin’s “lowering the tone”). But even if scientific practice emerges in culture rather than out away from it, surely we can safely acknowledge that a diverse family of knowledge-building strategies can be identified philosophically—whatever their social status—and that these strategies can be related to each other by means of an analysis of their logical structure. This is what I take Staley to mean by a “unity of methods”.
The use of such a unity to historians is in its ability to characterize historical acts. We need not share a philosophy, ideology, or terminology with historical actors in order to use a deft characterization of acts to define what kinds of acts made sense to historical actors, and the ways in which actors could communicate with each other thanks to their argumentative methods, and, just as importantly, the ways in which they talked past each other.
This is a point argued by Jed Buchwald and Allan Franklin in their essay “Beyond Disunity and Historicism” (available in full via Google Books) in their 2005 edited volume Wrong for the Right Reasons. In an explicit repudiation of “recent trends in the history of science, or ‘science studies’ in general,” they defend the philosophical characterization of historical acts: “it’s not a question of handing out report cards on the past, but rather of thoroughly understanding what was done” (my emphasis). One ought not judge the past according to present standards of good epistemology, but technical and epistemological sophistication can help identify what historical actors had in mind, and who they expected to convince, by making certain kinds of claims.
Philosophical understanding of epistemology allows historians to make useful comparisons between claims (certainly a major concern of mine in my recent article on Jay Forrester’s industrial dynamics). These can be transcendental. Per Buchwald and Franklin: “But what might count as a mistake that transcends the purely local? Is it possible to make assertions about scientific error that will not inevitably fall between the twin bulldozers of historicism and disunity? Can we compare Ptolemy with Newton without being silly?”
This last question, I think, is the key one. Philosophers and scientists clearly can, but can historians? Buchwald and Franklin think it would be rejected out of hand: “Different times, different beliefs, different sciences. No more need be said.” The argument gets a little fuzzy to me here. Transcendental philosophy allows one to determine why certain kinds of argument were ultimately dismissed. It allows us to say what philosophical, methodological, and experiential advantages Newton had that made possible the obsolescence of Ptolemy, and Copernicus for that matter.
To me, though, it’s simply the ability to compare and to understand why certain kinds of arguments and suggestions were persuasive, nevermind the comparisons across thousands of years. As I’ve mentioned before, Buchwald, in The Rise of the Wave Theory of Light was able to use methodological sophistication to help explain why Fresnel ultimately proved such an effective figure. The point is revisited here: “As a matter of historical fact, Fresnel was able to produce a quantitative account of diffraction that even determined opponents of his system of wave optics (including Biot himself, Poisson and Laplace) accepted in public.”
The fact that opponents can be persuaded indicates that there is a philosophical contiguousness between differing traditions. Coming back to Staley, this philosophical contiguousness (or unity, if you will) allows us to establish historical explanations of why historical actors were able to come to intellectual agreement. If indeed the image and logic traditions were not divided by an essential epistemological difference per Galison, we will be able to provide more powerful explanations for why particles detected in one kind of detector were accepted as real by experimenters working with another kind of detector, and why it took no great conceptual leap or the creation of intermediary languages to understand that a detector combining the benefits of traditions made pretty clear sense to the actors involved.
More broadly, there is no escape from philosophy, because it is only philosophy that allows us to characterize the differences between traditions as having a natural affinity to each other, as requiring epistemological (and sociological!) work to bring together, or as forever sundered from each other.