I’ve been thinking about what might prove canonical for early 19th century physics, and I am imagining that Jed Buchwald’s book on the wave theory of light is going to be on there. Now, studying this or that side of a epochal debate in science (you could also talk about, say, vulcanism vs. neptunism, or catastrophism vs. uniformitarianism, vitalism vs. mechanism, whatever) is a different sort of history from something like Smith’s look at the development of the massively influential energy physics program.
As Smith so ably demonstrates, energy was very much a program (tentative definition: a directed and deliberately strategized attempt to conceptualize knowledge within a certain scheme); whereas we might discuss the epochal debates as “approaches” to subjects. If we look at competing and incommensurable approaches to a problem, it’s easy to write a history of their conflict, but only at the danger of losing the motivations of the actors involved, who may or may not be invested in the conflict (even while taking the occasional pot shot across the divide). We should probably concede that it is possible for ostensibly conflicting approaches to co-exist more-or-less peaceably. Therefore (problems of unequivocally defining “moments” in science aside) it is possibly improper to look at this or that result as proving the validity of one approach over another, or at least doing damage to opponents, because the debate might not really be within the actor’s most immediate set of concerns.
In other words, by framing our narratives in terms of an epochal debate, we impose an external set of concerns on the historical actors, which is a form of Whig history, even if we are careful to not view actors as proceeding methodically toward the “correct” conclusion. Which is not to say we shouldn’t have books about epochal debates. Indeed, they are crucial to defining the traditions in which scientists/philosophers/etc. work. We merely need to be careful about understanding what these narratives are and are not saying about what actors were up to.
All this I think is fairly second-nature now. What has been primarily accomplished is to set certain vocabularies and certain formulations of historical scenarios off-limits. “Discovery” is a big off-limits area. Most books (including Smith) trip all over themselves to note that Kuhn’s efforts to identify multiple moments of discovery of a concept like the “conservation of energy” is pointless, because it takes the principle to be a pre-existing entity that exerts a magnetic force on actors, and diverts attention away from things like efforts to build credibility for the concept, and leads to the misleading condemnation of those who didn’t “get” the discovery right away.
So, I figured, maybe we should just forget about words like “discovery”; but then I was dealing (for other reasons) with 20th century elementary particles, and found that there’s no really compelling reason to not use the term to refer to the detection of something that was not there previously. Why shouldn’t Chadwick have discovered the neutron? If something is discovered in an intellectual environment in which one would expect a discovery along those lines, then the notion of discovery can be (although not necessarily is) clear cut.
All of which gets me thinking about one of the few belabored aspects of Smith’s book, which is his clear conceptual indebtedness to early Latour and Biagioli, in the development of networks of credit and credibility. The sociological literature tends to portray this as a sort of building of alliances to achieve acceptance, but it seems to me that it serves an intellectual function as well, which is building the robustness of a concept or method, arguing consistency with other ideas, and demonstrating novel applications to various kinds of problems.
I’ve been meaning to talk about robustness for awhile now, because I think it will prove important. Building alliances not only builds more widespread recognition, it also tests one’s ideas to see if they jibe with other accepted or proposed ideas and observations. So, Thomson, by fitting his energy ideas into geological and religious ideas, builds upon the robustness of the network of theories. Kuhn, in discussing paradigm shifts, emphasized the building of discrepancies between observations as leading to the breakdown of a paradigm. But the reception literature (such as Warwick’s take on the reception of relativity) has emphasized how isolated new views are within very robust explanatory schemes, and how problematic it is to expect an entire field to abandon these schemes for new and undeveloped ones. This argues for an importance of robustness above truth value in both sociological and philosophical realms.
Anyway, my point is, we use the term robustness already. It may be time to start thinking a little more deeply about what we mean by it. End of rambling post.