Share this:
03/09/2006 08:00:00

Caltech Philosopher Wins Lakatos Award

PASADENA, Calif.—Those who think that philosophy is about a bunch of dead guys with names like Plato and Kant and Hume will be surprised to learn that the philosophy of science is active and vibrant these days. What's more, some of the work currently being done in the field is as relevant to our daily concerns as the question of whether a certain new cancer drug is being tested properly in clinical trials.

According to James Woodward, the Koepfli Professor of the Humanities at the California Institute of Technology, some of the most interesting work in contemporary philosophy concerns the underlying logic of causal inference. By definition, the philosophy of science is the study of scientific methodology, but new developments such as biological innovation have kept guys like Woodward busy of late.

For a recent book on the topic of causal explanation, Making Things Happen, Woodward has been named winner of this year's Lakatos Award by the London School of Economics and Political Science. The Lakatos Award is given annually for an outstanding contribution to the philosophy of science. Woodward will travel to London in May to receive the £10,000 award and deliver a lecture.

Woodward advocates what he calls an interventionist conception of causation, according to which our interest in finding causal explanations is closely related to our interest in changing or manipulating nature.

For example, research in biology and biomedicine has become more inclined toward deliberate intervention to promote better causal understanding of the rules of nature. Whereas biologists once devoted much of their resources toward observation and classification, they may now manipulate genes in order to understand an organism, its development, and its molecular composition.

Central to constructing causal explanations is providing answers to what Woodward calls "what-if-things-had-been-different questions." Explanations of this type result in "counterfactual" dependencies, and these must be understood appropriately, he writes.

As an end result, "causal and explanatory claims are informed by our interest as practical agents in changing the world," Woodward adds.

Causal explanation is not only important in scientific life, but in our everyday interactions as well. We have a vested interest in knowing whether a bogus e-mail file will cause our computer to crash, but we also want to know whether a specific drug can treat heart disease.

The reason the latter is of concern to the philosophy of science, Woodward says, is that it's necessary to know precisely what is involved in clinical trials. "You want to determine whether this new medicine improves the prospects of people with heart disease, but if you set up an experiment and the test subjects are, on balance, healthy, then you're going to be led to a mistake in your inference about whether the medicine is effective or not."

While this conundrum has been known for decades as the "problem of confounding," Woodward says that it exemplifies the kind of situation that philosophers a few hundred years ago didn't understand. This is because the preferred procedure, in the previous example, is to employ probability and statistics in randomly assigning people to the control group and the group that gets the medicine.

"You then hope that all of the other things that could cause discrepancies are approximately equal," he says.

Another especially active endeavor in Woodward's field is the creation of automated data-mining techniques. This work involves looking at statistical data in ways to properly infer causal relationships.

The Lakatos Award is made possible by a generous endowment from the Latsis Foundation, and is given in memory of the former London School of Economics and Political Science professor Imre Lakatos.

Written by Robert Tindol