Latour Sokal NYT

Alan Sokal writes:

I don’t know whether you saw the NYT Magazine’s fawning profile of
sociologist of science Bruno Latour about a month ago.

I wrote to the author, and later to the editor, to critique the gross lack of balance (and even of the most minimal fact-checking). No reply. So I posted my critique on my webpage.

From that linked page from Sokol:

The basic trouble with much of Latour’s writings—as with those of some other sociologists and philosophers of a “social constructivist” bent—is that (as Jean Bricmont and I [Sokal] pointed out already in 1997)

these texts are often ambiguous and can be read in at least two distinct ways: a “moderate” reading, which leads to claims that are either worth discussing or else true but trivial; and a “radical” reading, which leads to claims that are surprising but false. Unfortunately, the radical interpretation is often taken not only as the “correct” interpretation of the original text but also as a well-established fact (“X has shown that …”) . . .

numerous ambiguous texts that can be interpreted in two different ways: as an assertion that is true but relatively banal, or as one that is radical but manifestly false. And we cannot help thinking that, in many cases, these ambiguities are deliberate. Indeed, they offer a great advantage in intellectual battles: the radical interpretation can serve to attract relatively inexperienced listeners or readers; and if the absurdity of this version is exposed, the author can always defend himself by claiming to have been misunderstood, and retreat to the innocuous interpretation.

Sokal offers a specific example.

First, he quotes the NYT reporter who wrote:

When [Latour] presented his early findings at the first meeting of the newly established Society for Social Studies of Science, in 1976, many of his colleagues were taken aback by a series of black-and-white photographic slides depicting scientists on the job, as though they were chimpanzees. It was felt that scientists were the only ones who could speak with authority on behalf of science; there was something blasphemous about subjecting the discipline, supposedly the apex of modern society, to the kind of cold scrutiny that anthropologists traditionally reserved for “premodern” peoples.

Sokal responds:

In reality, it beggars belief to imagine that sociologists of science—whose entire raison d’ĂȘtre is precisely to subject the social practice of science to “cold scrutiny”—could possibly think that “scientists were the only ones who could speak with authority on behalf of science”. Did you bother to seek confirmation of this self-serving claim from anyone present at that 1976 meeting, other than Latour himself?

Sokal continues in his letter to the NYT reporter:

In the same way, you faithfully reproduce Latour’s ambiguities concerning the notion of “fact”:

It had long been taken for granted, for example, that scientific facts and entities, like cells and quarks and prions, existed “out there” in the world before they were discovered by scientists. Latour turned this notion on its head. In a series of controversial books in the 1970s and 1980s, he argued that scientific facts should instead be seen as a product of scientific inquiry. …

In your article you take for granted that Latour’s view is correct: indeed, a few paragraphs later you say that Latour showed “that scientific facts are the product of all-too-human procedures”. But, like Latour, you never explain in what sense the traditional view—that cells and quarks and prions existed “out there” in the world before they were discovered by scientists—is mistaken.

I’m with Sokal: Scientific facts are real. Their discovery, expression, and (all too often) misrepresentation are the product of human procedures, but the facts and entities exist.

As Sokal discusses, the whole thing is slippery, as can be seen even in the brief discussion excerpted above. If you give Latour’s statements a minimalist interpretation—the concepts of “cells,” “quarks,” etc. are human-constructed—there’s really no problem. Yes, the phenomena described by our concepts of cells, quarks, etc. are real and would exist even if humans had never appeared on the Earth, but one could imagine completely different ways of expressing and formulating models for these scientific facts, in forms that might look nothing like “cells” and “quarks.” Just as one can, for example, express classical mechanics with or without the concept of “force.”

And, of course, if you want to go further, there’s lots of apparent scientific facts that, it seems, are simply human-created mistakes: I’m thinking here of examples such as recent studies of ESP, himmicanes, air rage, beauty and sex ratio, etc.

So Latour’s general perspective is valuable. But Sokal argues, convincingly to me, that much of the reading of Latour, including in that news article, takes the strong view, what might be called the postmodern view, which throws the baby of replicable science out with the bathwater of contingent theories.

Sokal writes:

If Latour had really shown that scientific facts are the product of all-too-human procedures, then the critics’ charge would be unfair. But in reality Latour had not shown anything of the sort; he had simply asserted it, and many others (not cited by you) had criticized those assertions. Of course, it goes without saying that scientists’ beliefs (and assertions of alleged fact) about the external world are the product of all-too-human procedures — that is true and utterly banal. But Latour’s claims are nothing more than deliberate confusion between two senses of the word “fact” (namely, the usual one and his own idiosyncratic one). . . . muddying the distinction between facts and assertions of fact undermines our ability to think clearly about this crucial psychological/sociological/political problem.

Sokal continues with his correspondence with the New York Times (they eventually replied after he sent them several emails).

Just to be clear here, I don’t think there are any villains in this story.

Latour has a goofy view of science, and I agree with Sokal that his (Latour’s) expressions of his ideas are a bit slippery—but, hey, Latour entitled to express his views, and you gotta give him credit for being influential. Latour’s successes must in some part be a consequence of previous gaps or at least underemphasized points in discussions of science.

The author of the NYT article, Ava Kofman, found a good story and ran with it. I agree with Sokal that she missed the point—or, to put it another way, that she might well be doing a good job telling the story of Latour, she’s not doing a good job telling the story of Latour’s ideas. But, that’s not quite her job: even if, as the saying goes, Latour’s work “contains much that is original and much that is correct; unfortunately that which is correct is not original, and that which is original is not correct,” Kofman is not really writing about this; she’s writing more about Latour’s influence.

The ironic thing, though, is that Kofman’s article is following the standard template of feature stories about a scientist or academic, which is to treat him as a hero. If there’s one idea that Latour stands for, it’s that scientists are part of a social process, and it misses the point to routinely treat them as misunderstood geniuses.

Anyway, although I share Sokal’s annoyance that the author of an article on Latour missed key aspects of Latour’s ideas and then didn’t even reply to his thoughtful criticism, I can understand why the reporter wants to move on to her next project. In my experience, journalists are more forward-looking than academics: we worry about our past errors, they just move on. It’s a different style, perhaps deriving from the difference between traditional publication in bound volumes and publication in fishwrap.

Finally, perhaps there’s not much the NYT editors can do at this point. Newspapers, and for that matter scientific journals, rarely run corrections even of clear factual errors—at least, that’s been my experience. So I can’t blame them too much for following common practice.

Ultimately, this all comes down to questions of emphasis and interpretation. Latour has, for better or worse, expressed ideas that have been influential in the sociology of science; his story is interesting and worth a magazine article; writing a story with Latour as hero leads to some confusion about what is understood by others in that field. In that sense it’s not so different from a story in the sports or business pages that presents a contest from one side. That’s a journalistic convention, and that’s fine, and it’s also fine for someone such as Sokal who has a different perspective (one that I happen to agree with) to share that too.

As Sokal puts it:

The ironic thing is that Latour has spent his life decrying (and rightly so) the scientist-as-hero approach to the presenting science to the general public; but here is an article that takes an extreme version of the same approach, albeit applied to a sociologist/philosopher rather than a scientist.

A newspaper or magazine article about a thinker should not merely be a fawning and uncritical celebration of his brilliance; it should also discuss his ideas. Indeed, this article does purport to explain and discuss Latour’s ideas, not just his personal story; but it does so in a completely uncritical way, not even letting on that there might be people who have cogent critiques of his ideas. That, it seems to me, is a gross failure of balance—and more importantly, a gross abdication of the newspaper’s mission to inform its readers about important subjects. (In this case, a subject that has serious real-world consequences.) Not to mention the gross lack of elementary fact-checking that I pointed out.

Of course, one could also question whether the “hero” mode of writing is appropriate even on the sports or business pages. This mode of writing presents a contest from one side only; and it is not very often the case in sports or business that there is in fact only one side.

So, yeah, the NYT article was not so bad as feature articles go—it told an engaging story from one particular perspective—but there was an opportunity to do better. Hence Sokal’s post, and this post linking to it.

P.S. Hey, the name Bruno Latour rings a bell . . . Unfortunately, he didn’t make it out of the first round of our seminar speaker competition.

The post Latour Sokal NYT appeared first on Statistical Modeling, Causal Inference, and Social Science.