Professor Sokal's Bad Joke

by Stanley Fish

[Published in The New York Times, May 21, 1996.]

When the editors of Social Text accepted an essay purporting to link developments in quantum mechanics with the formulations of postmodern thought, they could not have anticipated that on the day of its publication the author, Alan Sokal, a physicist at New York University, would be announcing in the pages of another journal, Lingua Franca, that the whole thing had been an elaborate hoax.

He had made it all up, he said, and gloated that his "prank" proved that sociologists and humanists who spoke of science as a "social construction" didn't know what they were talking about. Acknowledging the ethical issues raised by his deception, Professor Sokal declared it justified by the importance of the truths he was defending from postmodernist attack: "There is a world; its properties are not merely social constructions; facts and evidence do matter. What sane person would contend otherwise?"

Exactly! Professor Sokal's question should alert us to the improbability of the scenario he conjures up: Scholars with impeccable credentials making statements no sane person could credit. The truth is that none of his targets would ever make such statements.

What sociologists of science say is that of course the world is real and independent of our observations but that accounts of the world are produced by observers and are therefore relative to their capacities, education, training, etc. It is not the world or its properties but the vocabularies in whose terms we know them that are socially constructed -- fashioned by human beings -- which is why our understanding of those properties is continually changing.

Distinguishing fact from fiction is surely the business of science, but the means of doing so are not perspicuous in nature -- for if they were, there would be no work to be done. Consequently, the history of science is a record of controversies about what counts as evidence and how facts are to be established.

Those who concern themselves with this history neither dispute the accomplishments of science nor deny the existence or power of scientific procedure. They just maintain and demonstrate that the nature of scientific procedure is a question continually debated in its own precincts. What results is an incredibly complex and rich story, full of honor for scientists, and this is the story sociologists of science are trying to tell and get right.

Why then does Professor Sokal attack them? The answer lies in two misunderstandings. First, Professor Sokal takes "socially constructed" to mean "not real," whereas for workers in the field "socially constructed" is a compliment paid to a fact or a procedure that has emerged from the welter of disciplinary competition into a real and productive life where it can be cited, invoked and perhaps challenged. It is no contradiction to say that something is socially constructed and also real.

Perhaps a humble example from the world of baseball will help make the point. Consider the following little catechism:

Are there balls and strikes in the world? Yes.

Are there balls and strikes in nature (if by nature you understand physical reality independent of human actors)? No.

Are balls and strikes socially constructed? Yes.

Are balls and strikes real? Yes.

Do some people get $3.5 million either for producing balls and strikes or for preventing their production? Yes.

So balls and strikes are both socially constructed and real, socially constructed and consequential. The facts about ball and strikes are also real but they can change, as they would, for example, if baseball's rule makers were to vote tomorrow that from now on it's four strikes and you're out.

But that's just the point, someone might object. "Sure the facts of baseball, a human institution that didn't exist until the 19th century, are socially constructed. But scientists are concerned with facts that were there before anyone looked through a microscope. And besides, even if scientific accounts of facts can change, they don't change by majority vote."

This appears to make sense, but the distinction between baseball and science is not finally so firm. On the baseball side, the social construction of the game assumes and depends on a set of established scientific facts. That is why the pitcher's mound is not 400 feet from the plate. Both the shape in which we have the game and the shapes in which we couldn't have it are strongly related to the world's properties.

On the science side, although scientists don't take formal votes to decide what facts will be considered credible, neither do they present their competing accounts to nature and receive from her an immediate and legible verdict. Rather they hazard hypotheses that are then tested by other workers in the field in the context of evidentiary rules, which may themselves be altered in the process. Verdicts are then given by publications and research centers whose judgments and monies will determine the way the game goes for a while.

Both science and baseball then are mixtures of adventuresome inventiveness and reliance on established norms and mechanisms of validation, and the facts yielded by both will be social constructions and be real.

Baseball and science may be both social constructions, but not all social constructions are the same. First, there is the difference in purpose -- to refine physical skills and entertain, on the one hand, and to solve problems of a theoretical and practical kind, on the other. From this difference flow all the other differences, in the nature of the skills involved, the quality of the attention required, the measurements of accomplishment, the system of reward, and on and on.

Even if two activities are alike social constructions, if you want to take the measure of either, it is the differences you must keep in mind.

This is what Professor Sokal does not do, and this is his second mistake. He thinks that the sociology of science is in competition with mainstream science -- wants either to replace it or debunk it -- and he doesn't understand that it is a distinct enterprise, with objects of study, criteria, procedures and goals all of its own.

Sociologists of science aren't trying to do science; they are trying to come up with a rich and powerful explanation of what it means to do it. Their question is, "What are the conditions that make scientific accomplishments possible?" and answers to that question are not intended to be either substitutes for scientific work or arguments against it.

When Professor Sokal declares that "theorizing about 'the social construction of reality' won't help us find an effective treatment for AIDS," he is at once right and wrong. He is right that sociologists will never do the job assigned properly to scientists. He is wrong to imply that the failure of the sociology of science to do something it never set out to do is a mark against it.

My point is finally a simple one: A research project that takes the practice of science as an object of study is not a threat to that practice because, committed as it is to its own goals and protocols, it doesn't reach into, and therefore doesn't pose a danger to, the goals and protocols it studies. Just as the criteria of an enterprise will be internal to its own history, so will the threat to its integrity be internal, posed not by presumptuous outsiders but by insiders who decide not to play by the rules or to put the rules in the service of a devious purpose.

This means that it is Alan Sokal, not his targets, who threatens to undermine the intellectual standards he vows to protect. Remember, science is above all a communal effort. No scientist (and for that matter, no sociologist or literary critic) begins his task by inventing anew the facts he will assume, the models he will regard as exemplary and the standards he tries to be faithful to.

They are all given by the tradition of inquiry he has joined, and for the most part he must take them on faith. And he must take on faith, too, the reports offered to him by colleagues, all of whom are in the same position, unable to start from scratch and therefore dependent on the information they receive from fellow researchers. (Indeed, some professional physicists who take Professor Sokal on faith report finding his arguments plausible.)

The large word for all this is "trust," and in his "A Social History of Truth," Steven Shapin poses the relevant (rhetorical) question: "How could coordinated activity of any kind be possible if people could not rely upon others' undertakings?"

Alan Sokal put forward his own undertakings as reliable, and he took care, as he boasts, to surround his deception with all the marks of authenticity, including dozens of "real" footnotes and an introductory section that enlists a roster of the century's greatest scientists in support of a line of argument he says he never believed in. He carefully packaged his deception so as not to be detected except by someone who began with a deep and corrosive attitude of suspicion that may now be in full flower in the offices of learned journals because of what he has done.

In a 1989 report published in The Proceedings of the National Academy of Science, fraud is said to go "beyond error to erode the foundation of trust on which science is built." That is Professor Sokal's legacy, one likely to be longer lasting than the brief fame he now enjoys for having successfully pretended to be himself.

Stanley Fish is professor of English and law at Duke University and executive director of the Duke University Press, which publishes the journal Social Text. His most recent book is "Professional Correctness."