A while ago now, a friend alerted me to Errol Morris’s series The Ashtray for the NY Times’ Opinionator blog. Morris was writing about his time in graduate school, where he studied philosophy of science (!) with Thomas Kuhn (!!). If you haven’t read this series yet, you should really just go read it, now. I can wait.
At the center is an argument Morris had with Kuhn about the notion of incommensurability. The term means…well, this is part of the problem, it’s not clear precisely what the term means, but one of the consequences of paradigm shifts, according to Kuhn, is that the new body of knowledge created by one becomes incommensurable with what it replaces.
This captures the insight that, for example, the paradigm shift from alchemy to chemistry didn’t just entail a shift in the interpretation of particular observations but a wholesale revision of the methods, theory, and even many of the “facts” that supported the prior discipline (say, about the specific heat of dung from different animals). This is what Kuhn refers to as a paradigm shift or scientific revolution, and the major contribution of his work was to point out that progress in science isn’t necessarily a straightforward accumulation of knowledge, with each generation standing on the shoulders of the giants that preceded them. Progress is punctuated by destructive phases, in which much of what was thought to be known previously is thrown out, old problems must be solved anew, and things that were treated as central issues in the field are simply ignored.
There are many smaller and more ambiguous examples of paradigm shifts than the shift from alchemy to chemistry, and in arguing that they are characteristic of much scientific activity, and trying to address them systematically, Kuhn made what turn out to be some fairly sloppy generalizations. The most dire of these, from Morris’s perspective, involves his defense of a form of relativism whose corrosive influence on a variety of intellectual fields and the culture at large Morris draws out very convincingly (see also Ehrenreich & McIntosh on the “New Creationism” or, if you are feeling very brave, wade into the extended back-and-forth between physicist Alan Sokal and philosopher Bruno Latour). Morris makes a connection between the postmodernist trope that various aspects of reality are actually socially constructed and the cynical assault on meaning that is playing out on a number of fronts in our culture at the moment:
Please remember: This is not an empty intellectual exercise. It is not a matter of indifference whether it was God or natural selection that produced the complexity of life on earth. Nor whether there is such a thing as global warming. The devaluation of scientific truth cannot be laid on Kuhn’s doorstep, but he shares some responsibility for it.
And yet none of this, not even the fact that Kuhn at some point threw an ashtray the head of the man who would go on to direct “Vernon, Florida” and “The Fog of War” really dampens my enthusiasm for the occasional re-reading of The Structure of Scientific Revolutions. Because what Kuhn gets so right in this book is the very human way in which science makes “progress.”
Maybe I feel this way because I have always worked in fields — cognitive psychology and cognitive neuroscience — in which nothing approximating what Kuhn describes as “normal science” ever appears to be going on. In so many of the arguments in these fields, choosing sides means aligning yourself with a particular approach to collecting and analyzing data, which in turn means endorsing different ideas about what even counts as data and often disagreeing about what the goals of the enterprise are, in terms of what there even is to explain in the first place. In short, any of the things you would expect a mature field to have consensus on as a prerequisite to making progress are often up for grabs. Declaring yourself a “nativist” or a “localist” or a “dynamicist” does imply, in a very real way, that you are doing something fundamentally different from people on the opposite sides of those debates.
Sometimes this clash of perspectives is productive, because the competition between rival camps spurs creativity in the development of new techniques and arguments. Ideally, scientists would be partisan but fair, and each round of criticism would send them back to the lab to develop tests to deal with problems pointed out by their rivals, do a fair accounting of the results, and present a newer, sharper theory as a result. This is why Popper put such an emphasis on theories being falsifiable: he was working from a kind of free-market view in which competition produces progress. This dogma was the background against which Kuhn’s argument plays out, and the examples he gives of how science fails to fit this description feel familiar to anyone who works in science. In short, there are many occasions when competition produces worse, not better science. Winning an argument is different from finding the truth.
In my case, I spent about a quarter of my graduate career working on what amounts to a hatchet job on a rival model of reading. The authors had published a high profile paper claiming that the model covered a very wide range of phenomena. They also made the model available so you could test it out yourself. So, I took the model and began seeing how general its coverage of the various phenomena was. In many cases I discovered that the model could only account for the particular experiments the authors reported, and not other studies in which the same or similar effects could be observed. Further, the authors made some rather odd choices of which experiments to simulate — rather than the most recent, or most highly cited papers, they often simulated the results of more obscure studies. The model often failed when tested against more obvious choices of target studies.
I began to suspect that the modelers had stacked the deck when evaluating the model. They were behaving, I felt, l like in-house accountants putting the best face on the quarterly report to protect stock prices. And I was behaving like a forensic accountant, gathering evidence not so much to get to the truth, but to really nail them. This was probably the least productive period of my career. Errol Morris compared his graduate school experience to “an extended visit with a bear in a cave.” This was more like wallowing in filth with pigs.
So, what Kuhn got right was the rough and ready texture of much scientific work, and the fact that it is a social and even a cultural activity, rather than a rational machine for generating knowledge. His mistake may have been to try and systematize this behavior at all, and to paint himself into a corner from which he was forced to deny that anything like progress was actually possible. Morris’s objections to relativism are trenchant, and worth keeping in mind: the truth is real, and science, as a broad cultural endeavor, has moved us incrementally closer to it over the centuries. At the same time, the world Kuhn describes, which has a lot in common with the worlds explored by Morris the filmmaker — filled with obsessives, fantasists and the narrative chaos they engender — is ultimately much more like the world many scientists inhabit than we might like to admit.
Image from chriscardinal via Flickr