A Fluid Taxonomy of 20c. Sciences

Will (L) and Christopher pretending to work on their web project for the benefit of the AIP's weekly newsletter.

One of my ongoing concerns is the problem of writing a coherent history of 20th-century science and technology.  As I’ve been working on assembling names for my web project on notable post-1945 American physicists (on which Christopher is assisting me), I’ve been trolling through the National Academy of Sciences’ database of deceased members.  Living members are helpfully grouped by section, and we’ve simply taken members of the “physics” and “applied physical sciences” sections.  Deceased members, on the other hand, are not so grouped, so I have to look everyone up and sort out the physicists from the psychologists, anthropologists, chemists, geneticists, and so forth.

Predictably, this has resulted in some problematic category issues.  What to do with physical chemists, electrical engineers (especially those who won Nobel Prizes in physics)?; what separates an astronomer from an astrophysicist? when is mathematics physics-y enough to include mathematicians?

Another interesting problem that I’ve run into is that certain fields seem to have stopped being physics.  Ballistics research becomes more statistical than physical.  Thermodynamics, one of the great products of 19th-century physics, starts to fall solidly within chemistry.  Analytical mechanics seems to have become strictly a matter for mechanical engineers.  Sometimes the post-1945 period (which applies to lifespans rather than research) sorts these problems out, but sometimes you actually have to make a cut-off between a teacher and a student.  Early meteorology is largely an application of fluid mechanics, but around 1960, it really becomes a specialist zone with physical elements, but that would be difficult to classify as physics.  My strategy has been not to fear making semi-arbitrary decisions, and, anyway, we will be assembling a panel to review my choices and straighten me out when need be.  This may even mean going back over living members to see if those outside of the NAS sections we have chosen would fall within the criteria that include deceased members.  Messy, but inevitably so.

All of this has served to underscore a point Peter Galison made in his Ten Problems, which is the potential harmfulness of doing history within a predefined subject area.  Studying the “history of physics” lends itself to the construction of linear narratives that are, to a greater or lesser degree, teleological, since what is sought is an account of the development of things recognized as physics.  If we decide not to abandon accounts of non-trivial change (a temptation because writing these histories is problematic and scary), a likely alternative is characterizing practices and developing accounts of their continuities and discontinuities.

Galison’s Image and Logic (one of my all-time favorites) is exemplary here.  It’s a brick of a book, and I think most people just yank the general idea of a trading zone or pidgens/creoles from it.  What is less often taken away is Galison’s commitment (like Schaffer’s) to developing historical accounts, rather than simply having a historical account illustrate some truism from the current canon of epistemic claims wheeled out to satisfy the epistemic imperative.  In this respect, the far-less-mentioned notion of “intercalation” is crucial.  According to Galison’s commitment to the intercalation of traditions (he delineates theoretical, experimental, and instrumental; but the idea is far more flexible), to understand historical developments one should be able to develop a taxonomy of traditions, to understand the intellectual discontinuities between them, and continuities and transformations within them.

Galison used intercalation to great effect in his study of the history of the cloud chamber in chapter two.  Initially developed in a mimetic epistemological tradition (wherein the formation of clouds was meant to be understood by reproducing them in a laboratory setting), the cloud chamber became an “epistemic thing” (as the adherents of Hans-Jörg Rheiberger would call it) wherein cosmic rays were “seen”.  Thence it became an instrument to detect, measure, and eventually differentiate and conceptualize cosmic rays of different masses and velocities.

Now, the cloud chamber instrument and the practices for its use were modified to suit different experimental purposes, but the basic idea of the instrument remained intact.  Fine.

But Galison’s point, and my point here, is more expansive, which is that the idea of intercalation can help us develop a fluid taxonomy of 20th-century scientific practices, which, we can’t kid ourselves, is an enormous but essential—and essentially unaddressed—task.  Moving beyond the “history of a discipline” and its implied linear development, we should situate our studies around new entities.

I would propose: practices, problems, and affiliations.

Some methods of mathematical analysis are quite similar, but communities of analysts can work on different problems, which imply different affiliations than might be expected from a history of the theoretical discipline.  So, for example, nuclear physicists worked more closely with nuclear chemists than with many other groups of physicists, theoretical or otherwise, because all studied the behavior of the nucleus.

Here the Sociology of Experience and Expertise (SEE) comes in handy (and where the utility of University of Virginia professor Mike Gorman’s employment of Galison’s “trading zone” in SEE becomes obvious).  Scientists of different disciplines can communicate with each other concerning a problem—say, the behavior of the nucleus—and can allocate their expertise toward what they can agree is a robust understanding of the nucleus along interdisciplinary lines, even though no individual can have a complete understanding.  Nevertheless, other disciplines can offer knowledge that affects how, say, a nuclear physicist understands and interprets their own knowledge, and pursues further research, without the nuclear physicist having a complete understanding of what, say, nuclear chemists know and can do.

Now, the task of the historian is to set bounds on continuities and discontinuities in practices, problems, and affiliations.  Affiliations surrounding problems tend to be transient; but these affiliations can totally reconfigure the initial traditions of practice, or at least transform them in important ways.  The historian constructing an account of change should characterize those changes and trace their consequences for future affiliations.

Thus I see the semi-arbitrary decisions involved in the construction of my web project as an opportunity rather than a difficulty.  I am not especially interested in the coherence or incoherence of physics as a discipline.  What will be useful is understanding where the discontinuities within the discipline are strongest, and where affiliations within and across the boundaries of physics have substantial consequences.