The Great Escape

Abandoning philosophy

This post is meant to be the first in a series concerning the relationship between the history of science and the philosophy of science, paying special attention to the influential notion within the history of science that the philosophy of science has a deleterious influence on historiography.

Philosophy, in this view, injures inquiry by removing from consideration some of science’s most important non-scientific contexts; by causing historians to attempt to investigate incoherent questions rooted in philosophically defined problems (such as those relating to moments of discovery, confirmation, falsification, and proof); and by concentrating narratives on histories of disembodied ideas (vacuum versus plenum, atoms versus continuum, myth/confusion versus reason, determinism versus vitalism/free will, mind-body questions) and on the Whiggish pedigrees of disembodied theories (the theory of natural selection, the periodicity of elements, etc…), instead of on the actions and debates of scientists themselves, which the archives reveal did not turn on these preoccupations.

Sociology played a big role in “the great escape” (as I am calling it) from philosophy.  If philosophy has to do with the interaction between ideas and experience, it then has only a very narrowly defined role in the history of scientific practice.  The sensibility, I think, is captured nicely by sociologist Harry Collins in his recent overview of his career-long research program on the practice of gravitational wave physics, Gravity’s Shadow (2004).  Here he defends a “relativist” versus a “realist” (one might say sociological versus philosophical) perspective:

For argument’s sake, imagine this first sighting is of a [gravitational wave] burst, a couple of milliseconds in length, emitted by a supernova.  Thus in the half-century history of gravitational wave detection, nature would have spoken for only that instant.  Nature would have spoken for only about 0.00000000013% of this half century, and it would have been people thinking, building, calculating, hypothesizing, interpreting, spending money, writing, organizing, leading, and persuading that took up 99.99999999987% of the time.  And remember, hardly anyone would actually have “seen” the event that lasted even the milliseconds.  So whether you are a realist or a relativist, there is an awful lot for the sociologist to deal with.  (p. 15)

Collins’ demarcation of proper historical territory with the philosophers has, of course, often bled into full-scale epistemological hostility.  As John Zammito observes in his Nice Derangement of Epistemes (2004):

Despite Kuhn’s scruples, these sociologists [the 1970s Edinburgh SSK crowd] opened the way for the most “radical” impulses in the post-positivist theory of science.  They contended that nothing meaningful was left to the philosophy of science which could not better be construed by the sociology of scientific knowledge [SSK].  They insisted that philosophy of science, with its prescriptive account of science, had become thoroughly discredited by such critics as Kuhn and hence needed to be supplanted by a naturalistic and descriptive sociological account (p. 123, Zammito’s emphases).

This “descriptive” account might be thought of as relating to the problem of developing accounts that are historiographically “coherent”.  The devil here has to be Imre Lakatos, who argued for the “rational reconstruction” of the history of science according to philosophical principles—even if contra the historical record.  This argument represents a quintessential heresy in the historiography of practice.  (We’ll have to look at this heresy sympathetically at a later date).

As I have previously observed, the search for scientific theory that rings true with the historical record has fueled extensive discussion on many fronts, including Paul Feyerabend’s belief that science is defined by its methodological “anarchism” (i.e., non-formulaic practice), while Latour and Woolgar’s anthropological account, Laboratory Life (1979), emphasized the lack of an accounting for day-to-day practice in studies informed by traditional concerns.

Latour’s approach, culminating in the divorce of his school’s “actor-network theory” (ANT) from SSK, attempts to develop a history of science intelligible to outsiders (an explicit methodological choice from Laboratory Life on).  Since outsiders cannot distinguish truth from falsehood, the sociological processes of creating networks of belief and disbelief can be charted without reference to whether that (dis)belief is actually justified.  To account for the impact of reality on the contingency of historical paths (to which SSK makes no claim), nature itself is allowed to participate in these networks, but since knowledge of what nature is doing is always mediated by experts, there is, for the outsider, no way to distinguish firmly between the social and the natural, which (the claim goes) makes science-society discussions, in particular, less rigidly divided between scientific and non-scientific actors, and, thus, more coherent.

In my mind, the main trouble with this outsider phenomenological sort of approach is that it tends to exist in contradistinction to heady philosophy-dominated histories reminiscent of the history of ideas.  In ANT’s disdain for the epistemological, it offers no advice for constructing arguments for why ideas and practices made sense, outside of actors’ predisposition to hold and propagate their ideas, which, in turn, threatens to render historiography incoherent (especially by offering an irresistible tool to the historiography of exposé).  Herein, I think, are the roots of the Galison-Pickering debate over how to execute the great escape, which I mentioned a few weeks ago.

By Galison’s reckoning, Pickering (with the other SSK’ers), in adhering to a form of the old theory-laden observation thesis, was adhering to an anti-positivist philosophy of science, which supposed the dominance of theory over experiment (developed in opposition to positivist understanding of empirical evidence as standing at the root of science).  Thus, in Pickering’s Constructing Quarks (1984), the quark model exercised an overbearing influence on not only the practice of the development of programs of experiment, but on the practice of interpretating results according to the expectations of the model.

Galison’s own program to escape philosophy, first detailed in How Experiments End (1987),  also depended on a rigorous reading of the historical record of scientific practice.  This program was designed specifically to finesse a way out of positivist/anti-positivist debates by distinguishing between theoretical versus experimental types of presuppositions.  Galison used the idea that experimentation harbored specific kinds of observational presuppositions to disconnect experiment from a philosophically formulaic relationship with theorization.  This rendered science less vulnerable to the kind of exposé treatment that Pickering gave it by rooting most historical experimental decisions in more mundane and parochial concerns than ones that might be thought of as directly weighting experimental outcomes in favor of pet theories.

For Galison, linking historical contingency to anti-philosophical epistemological messiness represents not only a worthy conclusion, but a justification of an independent craft of historiography:

Experiments begin and end in a matrix of beliefs.  Some are metaphysical, others programmatic, and yet others no more general than a formal or visualizable model.  But laboratory work also exists amid practical constraints that may have little in the way of theory to support them: beliefs in instrument types, in programs of experimental inquiry, in the trained, individual judgments about very local behavior of pieces of apparatus or the tracks, pulses, and counts recorded every day.  Unraveling these factors is essentially a historical enterprise, one that follows no fixed set of rules. (p. 277)

The thesis of How Experiments End is largely a negative one: philosophically informed accounts will not account for the contingencies of scientific practice.  Galison had his own take, but the impulse to debunk the broad, philosophically-defined narrative by demonstrating its irrelevance to the material in the archives was a broad one.

Famously, these problems inspired severe fights over the foundations of knowledge, but they also had a less remarked upon chilling effect on historiography.  Entrenched in a profession firmly linked to the grand tradition of the history of ideas, historians suddenly became reticent to remark upon history at anything other than the archival level—a deeply problematic methodological tension.

I don’t feel this tension has been sufficiently recognized and resolved.  If the “historical enterprise” is one that “follows no fixed rules”, then, what, beyond history that does follow rules, defines bad historiography?  More importantly, what constitutes good historiographical craft?  We’ll look into some of the responses historians have come up with in future posts in this series.