Last week I went to a rather interesting little conference in Budapest organized by Ádám Tamás Tuboly at the Hungarian Academy of Sciences.   Given its focus on “sociological” aspects of logical empiricism, most of the papers were focussed on Philipp Frank (about whom I learned a lot) and Neurath (about whom I learned even more, though I knew a lot more about him than Frank to begin with). I was a little surprised at the neglect of Richard von Mises, an outsider I’ve always found very attractive, especially in this connection, and especially of Felix Kaufmann. Neither, admittedly, belongs to either of the two notorious “parties” of the Left or Right Vienna Circles, so both are somewhat lonely eccentrics on the fringe. But then so is Wittgenstein (though of course he’s a much bigger name than either Mises or Kaufmann), to whom Martin Kusch devoted a superb paper focussing on the intellectual context of Wittgenstein’s many remarks on color and color perception, showing in detail how some, at least, of Wittgenstein’s ideas were formed in response to the experimental psychology he encountered in Cambridge when he was a student there before the First World War.

Kusch’s paper gave a new impetus to my long-term preoccupation with the cultural/institutional relativity not just of higher-level concepts but basic (even perceptual) common sense. The idea of such relativity flies in the face of some widespread and rarely questioned prejudices. In much of cognitive science, also in the philosophy of cognitive science, there is a fundamental assumption that a certain basic cognitive set-up is universal — perhaps hard-wired — and therefore culture-neutral. In Dennett’s writings, for instance (as in many other places), we often encounter the idea of “the” manifest image. To be fair, this goes way back, and even Sellars (who coined the term “manifest image”) uses the definite article — in fact he effectively attributes its use to Wittgenstein as well; we hardly needed Kusch’s painstaking investigation to appreciate how mistaken Sellars was about that. What Kusch’s paper helps with, though, is to understand just how deeply into what is usually regarded as a matter of basic perceptual universals Wittgenstein’s preoccupation with the cultural anchoring of cognition penetrates (and how well his critique holds up today).

It has many ramifications. Experimental philosophy, for instance, which I am otherwise rather partial to, has now largely imported that mistake from cognitive science, without realizing it. When I draw attention to this in conversation with experimental philosophers, they tend to brush such objections aside as a minor detail, easily remediable by slight changes in experimental design. They don’t take it seriously. But then, of course, the ones I’ve talked to (not an unrepresentative sample) are Americans. And this blindness toward cultural specificity strikes me as yet another manifestation of the worldwide Americanization of philosophy over the past half-century. The idea that there can’t be fundamentally different ways of articulating the conceptual universe is deeply, deeply American – the most fundamental and unquestioned article of American cultural faith is the rejection of culture (which of course in the context of American history and social development makes total sense, but this isn’t a reason to take it seriously as doctrine!). And this basic, not fully articulate or self-conscious article of unquestioned belief lies somewhere near the bottom not just of American pragmatism, but more specifically of its more recent heirs such as Quine, Rorty, or (most explicitly) Davidson.

Lots of people outside the US share Gereon Wolters’s dissatisfaction with the American dominance of academia worldwide and object that the deck is stacked against non-English-speakers. I worry more about the corruption of the humanities by the uncritical importation of basic, unspoken, indeed hardly conscious American assumptions. The best-known example of this was the worldwide degradation of literary (and to some extent historical) studies by “French theory” (which any French person will tell you was a purely American phenomenon), but the subtler sort of thing I’m talking about here is also very much at work. What I find scary about it is not so much that it’s happening, but that it’s never discussed under the heading of creeping Americanization. It’s taken at face value.

Budapest is an ideal place for such worries.  On previous visits, I’ve stayed on the Pest side; this time, as a guest of the Academy, I was up in the castle, and what a different perspective one gets!  Not only that it’s largely 17th and 18th-century up there (and largely 19th-century down below), but the many steep, narrow little streets and the stairways down to the river give one a better appreciation of how the whole thing fits together, and how the city developed over the centuries.  I wish I could have stayed longer.

What didn’t surprise me at all, at this conference, was the relative neglect of Carnap. There is nothing about him, on the surface, that obviously invites attempts to apply his ideas, early or late, to social science. However, I took the opportunity in my own talk not only to do that, but to try to explain my social-science preoccupations to philosophers, rather than (as usual) trying to explain to social scientists why certain philosophical considerations might be worth their attention. I sketched the program (based on an existing social-scientific research tradition) of my 2009 Economic History Review paper with Sheilagh Ogilvie, and then tried to show how that program is not only consistent with logical empiricism — understood as the later, Carnapian continuation of ideas from the Left Vienna Circle — but actually realizes many of their ambitions.

This is highly counter-intuitive for many people, because the method we describe in that 2009 paper does not sound like Hempel at all. Its starting point is pure Droysen-style historicism, immersion in documents or (for sociology or anthropology) in participant observation, to obtain an “intuitive grasp” of how the society “works.” But then after that starting point, unlike historicism (as understood, say, by the hermeneutic tradition), this method triangulates on the intuitive understanding thus gained by reconstructing the practice in which the discourse we’ve been immersed in has meaning, including especially as many quantitative parameters of that practice (basic stuff such as life expectancy, age at first marriage, family size, composition of households, social and economic stratification within villages, etc) as the data allow. Such empirical data enrich and improve the intuitive grasp, making it more explicit and articulate, to the point where it then becomes the basis for a new round of systematic empirical inquiry, and so on.

This notion of iterative empirical improvement of an “intuitive grasp” is rejected by both sides – not just by Hempel (an intuitive grasp is not a theory), but also by the hermeneutic tradition, which mostly focusses on continuing, iterative immersion in texts or documents themselves rather than a dialectical interchange between intuitive understanding and data (let alone quantitative data).

It would take up too much space here to engage adequately with the hermeneutic side, but concerning the logical empiricist side I once again made the point, in my talk, that for Carnap (unlike Hempel, in this respect) there is no minimum standard of precision for explications.  He did not require them to be articulated in some target language.  On the contrary, he recognized that formal and fully explicit languages were the outcome of a long and continuous evolutionary process; nowhere along the line could one say, “okay look, here’s the point where it’s definitely science and no longer just a tightening up of folk categories” — there is a continuum. There is no hard boundary on one side of which is formal language, adequate for explication, and on the other side of which things are just too vague and ambiguous to qualify. Carnap’s criterion for something to be an explication was not that it meet some minimum standard of precision but that it be more precise, more explicit than its explicandum.

And by that standard, it seems, an improved, empirically enriched intuitive understanding surely qualifies. It may not yet even be at the point where it is fully expressible even in ordinary language, except by means of — yes, sorry — “thick description.” That’s just ordinary language that’s a bit vaguer than ordinary ordinary language. But it’s moving in the right direction, isn’t it? By measuring itself against hard facts it becomes more explicit, acquires sharper edges. And that’s what counts.