Writing, Theory, and the History of Science

Let me use this post to follow up on "The Science (Studies) Wars: Daston v. Jasanoff" by linking it up with a few different conversations that have taken place at Princeton over the past couple weeks.

First up was a discussion with Jim Endersby on 23 February about a pre-circulated paper called "Of Fleas and Whales: Turning History of Science Inside Out." Much of the colloquium was devoted to what exactly Endersby meant by turning our histories "inside out," since his emphasis shifted between aims like reaching wider audiences, tackling larger-scale problems, integrating with history departments, and saving the humanities and democracy itself.

While Endersby's concerns certainly resonate with the Daston/Jasanoff debate I summarized last week, he didn't put his paper into dialogue with them. Had he done so, I think he'd have aligned himself with Daston - though where Daston looks ahead from a history/history of science marriage to further liaisons ("Philosophy, anyone?"), Endersby wants to see more linkages with general historians. It became pretty clear in the Q&A that one's sense of the relationship between the two fields depends on institutional context: e.g. it's friendly at Princeton, but not-so-friendly, apparently, across the pond at Sussex.

Another phase of the conversation hinged on the relative importance of technical detail in histories of science (this should sound familiar from here and here). Ultimately, Endersby sided with the partisans of the less-technical, in the interest of maintaining the interest of non-expert readers and general historians. While in theory one should be able to balance the two, in practice we have to come up with techniques for sustaining attention without "selling our souls, as far as the scientific content."

This sounded like a concern about writing strategies for balancing faith to historical episodes against entertainment of present-day readers (the theoretical assumptions underlying this position weren't really up for debate), which linked it up to the second event I want to mention: a paper, by Aaron Sachs (Cornell), called "The Middle Landscapes of Arcadian America: Reconsidering Hudson Valley Culture, 1840-60."

This paper is part of a larger project on death (and life) in the American landscape tradition. One of Sachs's big interests, here and elsewhere, is in the writing of history - why and how we do it, and whether we shouldn't think of it as, to quote from the title of one of his recent articles, "creative nonfiction - or maybe even poetry."

While Sachs doesn't identify primarily as a historian of science, many AmericanScience readers will be familiar with his first book (on Humboldt's American legacy) and he shares many of Endersby's concerns about balancing "technical" material with more writerly interests. In Sachs's case, this interest takes the form of interspersing the historical narrative with bits of memoir, a tactic aimed at both readers (to leaven the book with present emotion and presentist concern) and writers (to jar them out of constraints of genre and chronology).

I'm still figuring out for myself how this decision maps onto Endersby's worries about writing the history of science for general audiences. It strikes me that both prompt us to rethink (or at least clarify why, how, and against whom we wield) our accusatory claim of "Whiggishness." Interestingly, neither discussion dwelled for very long on the theoretical implications of these decisions: implicit, to my mind, at both events was a sense that matters of writing were matters of craft, while matters of theory were something else.

Let me try to connect this issue of writing/theory back to the Daston/Jasanoff debate by way of a third event: a midday colloquium and then an afternoon seminar in Princeton's Sociology Department with the inimitable John Levi Martin, whose work in field theory has been on my mind lately and whose Social Structures awaits my next spare moment for a thorough read.

If the Endersby and Sachs events were all about writing, Martin's was all theory. His current book project, from which he circulated a chapter on "causality and persons," is aimed at what he sees as a dominant, and profoundly inappropriate, theory of causality at the heart of sociological explanation. He calls this theory simple counterfactualism (SCF), and, while I want to avoid getting into too much detail on an incredibly rich paper and discussion, let me suggest that part of what was at issue was what "theory" (scare-quotes intentional) is for in sociology, and how its use might be improved.

Because theory operates so differently (so much less explicitly, though no less fundamentally) in history, and because the problem of faulty analogizing was part of our discussion, I hesitate to share the musings to which I was spurred, as the lone historian in the room, about its implications for my own disciplines. I'll just close with one:

At least where I've been trained, questions of writing seem to be discussed separately from questions of theory. This is interesting, of course, since the two are bound up in a number of important ways, the histori(ographi)cal importance of Hayden White's work certainly not least among them. Writing, in my experience, is still taken up as a matter of rhetoric: making arguments more persuasive, clarifying, structuring chapters, &c.; theory, on the other hand, is treated as a cognitive tool taken up to grapple with historical materials (often not explicitly, but operative nonetheless).

What's the takeaway for historians of American science? I suggest that we can heed Endersby's call to make our work more amenable to American historians, take seriously Sachs's urging to flout the constraints of genre and temporality, and think more critically, with Martin, about the theory underpinning our assumptions, all at once. That is, one "resolution" of the Daston/Jasanoff face-off over the future of science studies would be to emphasize the common cause of writing.

There's no reason why this couldn't also be a turn to theory. Such an emphasis would mark a return not only to White, but also to Raymond Williams, whose "keywords" offer an intriguing and somewhat untapped resource from which (cultural) historians of science could begin to recast the old structure/agency debates in a way that gives credence to the structural (or field-theoretic) forcing of that powerful constellation of human ideas and activities ("science") on which we attend.

In a subsequent post, one of us (Dan? maybe me?) will address the rolling-out of Errol Morris's "Ashtray" series in the New York Times this week. It's a story, which Dan already previewed, in which history and philosophy were twined together at Princeton in the 1970s, with Kuhn's "paradigms" hardening to orthodoxy and Saul Kripke delivering “Naming and Necessity" while looking through a water glass. The series promises to be pretty ridiculous - many of Morris's anecdotes feel fabricated, and he leans on a strong over-reading of Kuhn's Structure and its general acceptance - and we'll try to take it's pulse in a few days...


It’s unclear to me what Martin’s take-away point would be for historians. The quest for rigor in social sciences has pushed methodologists and statisticians towards a view of causality and causal inference so far removed from everyday notions that Martin’s criticism comes off more as preaching to the choir (sociologists) than general proscriptions for the “human sciences” like history. That said, much of what he was arguing against in the paper arose in its strictest form first in historical sociology, which cast the comparative-historical program as a sort of natural experiment. Because you couldn’t re-run the history of America five hundred times and build a probability distribution for the American Revolution, historical sociologists reasoned we might look at other countries which varied differently to understand the counterfactual cases. The result was a strange rebirth of Millsian-style causality (necessary/sufficient causation), which subsequently made its way into hardcore statistical modeling (but only recently).

In my happier days in HoS, historical causation in terms of necessary and sufficient causes never once came up. In fact, does causality, period, ever really come up in professional history? Perhaps, this is what is at stake when someone makes the claim: “To what extent is your 19th century story really an 18th century story?” or some variation thereof. Clearly, historians strive to get things “right”, but in my experience what’s “right” depends more on correctly framing your case in a conventional manner (one part “institutional history”, one part “transnational flows”, two parts “history from the ground up”), than actually identifying the three necessary and sufficient causes of some historical event. If the former is true, then historical rhetoric and its expression in writing would seem far removed from more theoretical/epistemological questions of causality that Martin puzzles through.

ps. If you want to get a real taste of the causal kool-aid, check out the blog of statistician/political scientist par excellence Andy Gelman: http://www.stat.columbia.edu/~gelman/blog/

DR: Thanks for the thoughts, and the link to Gelman - as you know, his name came up a lot at the seminar, so I'm happy to have his musings.

As to whether causality as such ever comes up for historians, the answer is a dull "sometimes." Besides a relatively recent focus issue of Isis on counterfactuals in history, there's something of a division between those who are obsessed with rooting out the intentions of actors to produce given effects before those effects are gone into in an historical narrative, and those with alternative approaches to the question of intention, cause, and effects.

That is (and this is to put it somewhat dumbly for the sake of space): it comes up when people of the former type demand clarity on the causal claims underlying the events being narrated, or when people of the latter type raise it in the more theoretically-inflected manner of Martin and others.

These latter tend to be interested in the issue as it plays out in other disciplines (social sciences, but also English Lit), while the somewhat more orthodox former group tend to assume that history "has" a theory of causality, albeit an assumed one, that basically lines up with our commonsense notions of actor-produced effects.

Make sense? Maybe not. More on this soon...

Note: Only a member of this blog may post a comment.

back to top