History-Philosophy Relations, Pt. 2: The Weltphilosophie of Historical Epistemology February 16, 2013Posted by Will Thomas in Uncategorized.
Tags: Alfred Nordmann, Allan Franklin, Andrew Pickering, Augustine Brannigan, David Bloor, David Gooding, Friedrich Nietzsche, Gaston Bachelard, Georges Canguilhem, Gerald Holton, Hans-Jörg Rheinberger, Harry Collins, Ian Hacking, Kent Staley, Larry Laudan, Lorraine Daston, Martin Kusch, Michel Foucault, Peter Galison, Simon Schaffer, Steve Woolgar, Thomas Kuhn, Trevor Pinch
The program of “historical epistemology” represents one of the more ambitious and thoughtful projects espoused by historians of science in recent years. The self-conscious efforts of people like Hans-Jörg Rheinberger, Lorraine Daston, and Peter Galison to renew interest in epistemological questions among historians is laudable. And their point that epistemology is something that is invented rather than transcendental—and thus historically variable in its content—is surely a correct observation, at least from a historiographical standpoint.
That said, I have never been fully comfortable with the history produced by historical epistemology. To date, the program has received the most intensive scrutiny from philosophers. A good example is Martin Kusch’s 2010 paper, “Hacking’s Historical Epistemology: A Critique of Styles of Reasoning”.* My own interest in the subject has less to do with the integrity of historical epistemology as epistemology (a subject I am happy to leave to philosophers), as it does with its Weltphilosophie and its conception of the history-philosophy relationship.
Historical epistemology is similar to social constructionism in that it rejects the utility of traditional philosophy of science on account of its failure to conform to the historical reality of science. Also, like social constructionists, proponents of historical epistemology are apt to privilege accounts of scientific practice exhibiting internal-external hybridity. For example, they might stress the existence of a moral imperative (e.g., self-abnegation) that guides notions of proper scientific practice, but is also prominent in a surrounding culture.
At the same time, though, proponents of historical epistemology are less apt to invoke the professional, political, or cultural implications of a scientific result in accounting for the existence of that result—typically, they place a higher emphasis on the real force of professional integrity. Relatedly, they are less apt than social constructionists to suppose that theorization and the interpretation of observations are prone to radical contingency and “negotiation”. For them scientific epistemology may be historicizable, but it nevertheless exhibits substantial homogeneity and stability. In this limited sense, the proponents tend to have a more “internalist” conception of science than a social constructionist would. (However, that said, some social constructionists—I would point to David Bloor, Harry Collins, and early Trevor Pinch—often focus on the internal in their analyses as well, and much of their presentation and even interpretation is indistinguishable from what would be produced by historians working in an internalist vein.)
A major component of historical epistemology (generally associated with the Max Planck Institute for the History of Science in Berlin) is the self-conscious continuation of the historiography of implicit (rather than intellectual) concepts, which draws a line through Friedrich Nietzsche’s (1844-1900) historicization of morals, Gaston Bachelard’s (1884-1962) psychoanalytical histories of science, Georges Canguilhem’s (1904-1995) inquiry into the “normal” and the “pathological”, and, of course, Michel Foucault’s (1926-1984) archaeology of “epistemes”.
The content’s not really relevant, but here’s a cool old French interview with an older Bachelard:
Historical epistemology often focuses its attention on concepts with some particular “epistemic” connotation. Daston, notably, began her work with an investigation (following and alongside Ian Hacking) on the probable, before moving on to other concepts such as objectivity, observation, and the rational. Sometimes these studies discuss historical scientists’ and philosophers’ explicit treatment of such concepts, but more often they are read into practices and “epistemic objects”.
Galison’s work has rather different roots. In the 1980s, Galison was part of a more general set of historians and philosophers inquiring into the epistemology of experiment, which also included (again) Ian Hacking (esp. Representing and Intervening, 1983), David Gooding, and Allan Franklin (esp. The Neglect of Experiment, 1986), among others. In How Experiments End (1987), Galison argued expressly against theory-centric philosophical accounts of experimental interpretation, which either supposed that theories emerged inductively from observations, or that observations were theory-laden. Instead, Galison supposed that interpretations were generally based upon specialist interpretive practices fostered by experimenters, and constrained in certain ways that could only be uncovered through intensive historical research.
At that time, Galison’s view contrasted to that of social-constructionist Andrew Pickering. Galison objected to Pickering’s presentation of the discovery of the J/ψ particle in Constructing Quarks (1984), which suggested that the experiment had been “tuned” to affirm the discovery. By his view, the particle was detected through practices and standards specific to experimenters, in which the notion of such tuning made little sense. More generally, Galison posited, contra Pickering, that experimentation was a more tightly constrained activity than constructionists allowed. I would suggest one way of interpreting the disagreement was that where Pickering posited, as a matter of socio-epistemic principle, that constraints were not necessarily constraining, Galison’s position derived from the point that, as a matter of historical fact, they are nevertheless not so flexible.
Another key element of Galison’s thinking on specialist interpretive practices is that they can be heterogeneous, where different experimenters possess different predilections for what kinds of evidence they find “persuasive”. As I point out in my recent “Strategies of Detection” paper, Galison’s original articulation of this point in a 1982 paper on the discovery of the muon was developed for the purpose of problematizing the concept of a moment of discovery. This came amid a flurry of philosophical and sociological meditation on the problem, including the reprinting of Thomas Kuhn’s 1962 paper “The Historical Structure of Scientific Discovery” in his 1977 The Essential Tension volume, Steve Woolgar’s 1976 paper “Writing an Intellectual History of Scientific Development: The Use of Discovery Accounts”, and Augustine Brannigan’s 1981 book The Social Basis of Scientific Discoveries. The issue would show up in Simon Schaffer’s 1986 piece, “Discovery and the End of Natural Philosophy” as well. (Michael Bycroft recently called my attention to a 1980 piece by Larry Laudan on the prior abandonment of consideration of the “logic of discovery”, which seems to be part of a related but distinct philosophical discourse.)
In How Experiments End, Galison developed a socio-epistemological hybrid model of discovery as an expanding “circle” or overlapping “circles” of “belief”. In papers leading up to 1997’s Image and Logic, Galison developed this model of discovery into both a history and epistemology of a scientific enterprise characterized by its “disunity”. According to Galison, scientific figures are personally persuaded by evidence conforming to one or another epistemological ideal type. For much of the twentieth century, particle detection was divided into “image” and “logic” traditions of instrumentation, which were associated with ideal forms of evidence—the demonstrative image versus the rigorous logical proof. However, the division between ideals does not cripple the advance of science. Rather, scientists collaborate through “trading zones” where stripped-down languages permit theorists and experimenters, scientists and engineers, and, of course, experimenters abiding by different ideals to fruitfully collaborate. Further, the tandem hybridization of practices and ideals provides an important source of dynamism and change in the scientific enterprise.
It is Galison’s interest in epistemological ideals that brings his work into alignment with that of the concept historians. Like concepts, epistemological ideals boast “mesoscopic” periodizations, which historians can elucidate. Further, they bear strongly upon a highly personal, aestheticized, non-negotiable conception of what is “known”. This Weltphilosophie of individuals continually striving to produce results that conform to personal ideals seems to owe a great deal to what in the 1970s Gerald Holton called the “thematic” element of scientific theorization, which was effectively non-rational, and stood apart from science’s “empirical and analytical content”.
How are we meant to interpret this ideal-centric conception of scientific work and thought? Is it merely meant to suggest that the aesthetic has a role to play, or is the point actually more forceful than that? The recent and strongest—and, I would stress, most radical—expression of this Weltphilosophie appears in Daston and Galison’s 2007 book Objectivity, wherein the concept of “objectivity” is wedded to Galison’s epistemological idealism, and is so elevated to emotionally charged, identity-defining status of an “epistemic virtue” which informs how images are produced in characteristic ways in certain periods across a large number of sciences.
Within this Weltphilosophie science is essentially reduced to an image-producing enterprise apparently devoid of argumentation (or, at least, argumentation is confined to those who adhere to the virtue of “structural” objectivity). In one sense, Daston and Galison seem to cast the epistemological integrity of knowledge as a phenomenon that emerges from essentially non-rational ideal-seeking behaviors. Evidently aware that this would not really be a sustainable portrait of scientific thought, they hedge their claims: epistemic virtues accumulate rather than displace each other, and so overlap in scientific practice. Further, objectivity is but one of many epistemic virtues and concepts that are to be chronicled.
To my mind, this hedging one’s way to reality causes long-term problems for the historical-epistemological Weltphilosophie. On the one hand, concepts and epistemic virtues are taken to be so absolutely central to scientists’ sense of “self” that they pursue them all the time. On the other hand, the multiplicity of virtues seems to suggest that scientists must navigate a multi-dimensional terrain of virtues, any one or more of which might be at play at any given time, meaning that scientists must deploy them in complex ways. Is this deployment haphazard and effectively schizophrenic, or does is it tend to follow rational principles? If it is rational, does this not open the door back up to old-fashioned philosophers of science to describe those principles?
I would propose that it does open such a door, and, in doing so, it suggests that the historical-epistemological Weltphilosophie, as it is currently conceived, is, in spite of being an attempt to reconcile epistemology with the historical record, in danger of becoming the basic elements of an abstract metaphysics that is simply asserted, without a satisfactory account of its mechanisms, to generate the complexities of observed reality.**
Galison’s particular conception of the relations between epistemology and history has been questioned by historically minded philosophers. After Image and Logic appeared, a special issue of Perspectives on Science was dedicated to the book, and two of the essays suggested problems with the Weltphilosophie Galison evidently followed there. In “Establishing Commensurability: Intercalation, Global Meaning and the Unity of Science” (pdf), philosopher Alfred Nordmann suggested that the apparent epistemic incommensurability between ideal-oriented traditions was overcome not only in local trading zones and in spite of scientists’ disparate ideals, but though the “idea of a unified science,” which functioned “as a regulative ideal” that actively prompted scientists to come to terms with each others’ perspectives.
Moreover, in his paper, “Golden Events and Statistics: What’s Wrong with Galison’s Image/Logic Distinction” (pdf), the philosopher Kent Staley more radically suggested that behind Galison’s “image” and “logic” ideals stood a single, statistically defined logic of argumentation, which hinted that science might possess an underlying argumentative “unity” after all.
Now, Staley is not an armchair philosopher. He has written the detailed historical account, Evidence for the Top Quark: Objectivity and Bias in Collaborative Experimentation (2004). Further, Allan Franklin (like Galison, a student of the special epistemology of experiment), has championed Staley’s views in his 2002 book, Selectivity and Discord: Two Problems of Experiment. These are points of view that proponents of historical epistemology, and historians more generally, can fruitfully reckon with.
In Pt. 3, I will examine the tensions between Staley’s and Galison’s points of view, and I will relate them to fundamentally differing conceptions of the history-philosophy relationship, and also to my own history-centered (rather than epistemology-centered) perspective in my “Strategies of Detection” paper.
*Addendum (17 February 2013): A better example might be Kusch’s “Reflexivity, Relativism, Microhistory: Three Desiderata for Historical Epistemologies” (2011)
**Stray observation: the troubled issue of reducing complex phenomena to the interplay of basic metaphysical principles seems to go back a long way—it came up in my post on David Hume’s criticism of the “Hobbist” philosophy of sentiments—and this may be an issue worth further reflection, from either or both a philosophical and historical point of view.