In this week’s Wired magazine there’s an article on the way scientists think. “We’ve heard this all before,” I hear you savvy-with-the-philosophy-of-science readers say. Right. And the results reported are similar to what we’ve heard before too: scientists interpret anomalies as methodologically generated, and so removable from their data, until that is no longer an option, and a change of how one goes about interpreting the data is required (cf. Kuhn on anomalies). If Popper ever meant to describe what scientists actually do, he would have been quite wrong.
The supposed novelty of the work reported by Wired is, first, that the work is done by a psychologist (Kevin Dunbar), not a philosopher. And second, the work involves observations of scientists at work, rather than very post hoc reminiscences of what life was like when the writer once trained as a scientist.
But though the Wired article presents it as such, the second novelty is not a novelty. It’s true that some of the most famous names in philosophy of science did not engage in “in vivo” research. However, others did. Bruno Latour and Steve Woolgar’s Laboratory Life and Michael Lynch’s Art and Artefact are both extended studies of life in a lab. And interestingly, each had as its focus an issue that the Wired article omits entirely.
The Wired article paints scientists as dogmatic ignorers of the flaws in their research: the data clearly point to one’s preconceived ideas being wrong, so one ought to change the direction of one’s research. But the scientists fail to do so. But then why does science work? Because they argue it out and come to see the error of their ways. That’s the Wired story (roughly).
Lynch, and Latour and Woolgar, begin their questioning with one less assumption. How is it that when multiple interpretations of the data are always possible, any order is so often found in the data collected? The production of order: how does it happen? These authors treat the processes described by Dunbar as a way of achieving order “out of chaos” as Latour and Woolgar put it. Scientists’ well documented behavior is a solution to a problem one gets if one drops the assumption that seeing any order, getting any agreement on what the data mean, is something that does not require explanation.
One can’t help feeling that the Wired article, if not Dunbar himself, has committed that sin so despised by Kuhn: trying to understand the behavior of scientists as though they knew already what they would only come to know once they had finished their investigations.
The Psychology of Scientific Investigation
By J. D. Trout, Loyola University Chicago
(Vol. 2, April 2007)
The Paradox of Confirmation
By Branden Fitelson, University of California – Berkeley
(Vol. 1, February 2006)