This post is a response to this comment by Michael Bycroft on a 2009 post on Simon Schaffer’s well-known 1989 “Glass Works” paper, which brought to my attention a reply published seven years later by historian of optics Alan Shapiro: “The Gradual Acceptance of Newton’s Theory of Light and Color,” Perspectives in Science 4 (1996): 59-140.
“Glass Works” was itself a commentary on a large body of Newton scholarship, most especially Richard Westfall’s biography, Never at Rest (1980). It explicitly made use of Harry Collins’ sociology of “calibration”, which pointed to the necessity that instruments and experimental procedures gain trust before assertions based on experimental results can be accepted. Schaffer and Steven Shapin had previously used this insight in Leviathan and the Air Pump (1985) to call attention to the basis of Thomas Hobbes’ criticism of experimental philosophy as well as to the intellectual, literary, and sociological strategies Robert Boyle used to gain assent over experimental results.
Unlike Schaffer’s commentary, Shapiro assembles a synthetic history of the acceptance and replication of Newton’s important experiment showing the elongation of the light of the sun when passed through a prism, as well as his two-prism experimentum crucis, which demonstrated that white light was composed of differently refrangible rays. According to Shapiro’s lengthy account, Newton’s experiments were generally well accepted, even by intellectual opponents.
A major reason underlying this acceptance was its coherence within a strong theoretical context. According to Shapiro:
Theory provides a logical structure to claims about the natural world and — when used in conjunction with experiment — a guide to arriving at reliable conclusions about that world. Thus, theory can serve to mediate disputes about the significance of experiments or theoretical assertions and itself has persuasive power that cannot be ignored. Newton’s theory of color provided in one stroke a sophisticated theory of a broad realm of nature based on scores of experiments. It demanded serious study to understand the intricate relation of theory and experiment. The slow, gradual acceptance of Newton’s theory reflects the effort required to comprehend it. (64)
Shapiro believes that Schaffer, “by using a constructivist approach and implicitly adopting the model of a modern laboratory science,” overstates the importance of robust experimentation as a barrier to reception:
[Schaffer’s] account … must be judged a failure when weighed against the historical evidence. Applying his approach to the acceptance of Newton’s theory means focusing on Newton’s instruments, especially prisms, the difficulty of replication, the opaqueness of instrumentation and experimental procedures, the uniqueness of local practices, and Newton’s effort to establish ‘authority’ and ‘transparency’ for them. By reducing the issue of acceptance to one of power and authority, Schaffer argues that Newton established his theory by means of a virtual conspiracy among his acolytes. Newton’s power to get his theory accepted, he tells us, ‘lay in control over the social intitutions of experimental philosophy. In the 1670s, Newton exercised no such power. After 1710 his authority among London experimenters was overwhelming’ (p. 100). (60)
Aside from questions about whether Schaffer’s narrative constitutes a proper reading of the chain of the acceptance of Newton’s ideas (Shapiro argues it does not), Shapiro believes that Schaffer, on account of his constructivist methodology, neglects the crucial role of theory: “Schaffer avoids scientific theory as much as possible, since it offers a source of meaning outside of controversy, negotiations, and power” (64).
In his comment, Michael suggests that if Shapiro is right (neither of us are Newton scholars, so neither of us is really in a position to make any final judgment here), “not only does he beat Schaffer at the present-centredness game, he also beats him at the scientific practice game.” Shapiro calls attention to the importance of theory as practice, but also to the importance of not Whiggishly discussing experiment in terms perhaps more appropriate for later eras (such as Collins’s study of complex gravity wave detection experiments, which struggle to separate signal from background) — the prism experiments weren’t that hard to replicate.
(Note here that the issue of theory or ideas as a neglected aspect of practice has been explicitly discussed on this blog. See here and here.)
There may be something to this methodological criticism, but in this context I’m not really comfortable with trying to figure out who is beating whom at the game of purifying himself of recognized methodological sins, because I think there are more important issues at hand.
Before moving on to those issues in Pt. 2, I will say that my original post was titled “Schaffer Turns to Practice” precisely because “Glass Works” and “Astronomers Mark Time” (not counting Leviathan and the Air Pump, in which I tend to attribute most of the deep analysis of practice to Shapin) marked a new and unexpected turn in Schaffer’s oeuvre away from his early interest in theoretical systems in speculative natural philosophy. So, I don’t really think that plastering the “constructivist” label on Schaffer and suggesting he is naive to the importance of theory is really the way to go.
That said, Shapiro picks up instantly on some of the things I’ve noticed with Schaffer’s oeuvre beginning in the 1990s, which is that he transplanted his thinking about natural philosophical cosmology onto his studies of practice, translating implications of diverging conclusions within a natural philosophical system to an “insultography” that accompanied basic failures to come to agreement about facts, conclusions, and appropriate practice. As Shapiro notes, an increasing historiographical recognition that the correctness of scientific conclusions was rarely obvious has resulted in an emphasis on the “study [of] rhetoric, experimental replication, and ‘negotiations'” (60).
The point that isn’t sufficiently fleshed out in Shapiro’s piece is that the goals of Schaffer’s (and, for that matter, Shapin’s) project of commentary are decidedly different from Shapiro’s emphasis on synthesis. Over his entire oeuvre, Schaffer has never written a synthetic history of anything. In his interview series with Alan Macfarlane (I forget precisely where), he explicitly says that his output has mainly been commentary on his students’ work. I would argue that even Leviathan and the Air Pump should be read mainly as a commentary on a very specific set of 17th-century debates.
The goal of this commentary is to use a certain theoretical vantage point to identify points in the historical record that might not otherwise be subject to scrutiny, and to elucidate what happens there. So, in Leviathan and the Air Pump or in “Glass Works” the point of the exercise is to elucidate what strategies have been historically used in specifically those instances where agreement cannot be secured. Disputes over instrumentation is simply one point where these strategies can be pinpointed.
For my part, I don’t think Schaffer is deploying a particularly “modern” concept of experimentation here, nor do I think we can say that the absence of theory in Schaffer’s account is necessarily a problem with his commentary. Further, I wouldn’t take Schaffer’s account to imply the Newton’s success was all about “power”. I do not think that commentary should be judged by the standards of completeness and coherence demanded by synthesis.
Are we, then, to simply accept a programmatic difference between Schaffer and Shapiro? I urge that we should not.
For one thing, if we accept that commentaries have no responsibility to make explicit claims about what their conclusions imply for conclusions to be drawn in more synthetic accounts, we have no good way of evaluating the significance of the practices and strategies that are the subject of commentary.
I would argue that the project of commentary has legitimized itself not by making clear what the implications of commentaries are for syntheses, but by participating in what we might call a historiographical cult of invisibility. Schaffer, for instance, tends to make the case for the significance of his latter-day insultography mainly because polemics have traditionally been considered to be exterior, and thus invisible to a proper history of science. Thus insultography is taken to be important mainly because it needs to be restored to visibility in portraits of historical scientific practice.
(For more on this, see my “Schaffer on the Hustings” series, Pt. 1, Pt. 2, and Pt. 3.
Fine. But we have to understand that this argument for significance is historiographical, and provides little clue to historical significance. One often gets the impression in reading insultography that a Post-It note has been slapped on the historical record saying, “By the way, in addition to the story we already have, some insults were thrown around as well, which reflected the cultural milieu of the debates.” We are to take interest in this, because it is also implied: “This is a major revelation in our understanding of historical socio-epistemology, which symmetrically takes into account all actors’ perspectives (even though alternative actors are mainly of interest not in and of themselves, but because they object to the subject at hand).”
(This argument about the selective use of the symmetry principle can be read as an agreement with points in the latter part of Michael’s comment.)
A simple lesson to take away is that historiography cannot long live on commentary alone, and Shapiro hits hard on the importance of evaluating historical significance as a means of securing correct conclusions about historical developments. But, in addition to that, he also makes a very cogent case for the historiographical project of synthesis measured against the project of commentary. These are both points I will address in Pt. 2 of this post.