Week 11: Experiment and Explanation

After immersing ourselves in certain, messy social aspects of science, it’s time to return to the some of the questions about evidence for our theories that crop up at an individual level. For the next few weeks, we’ll descend back to the individual level — or anyway the level at which sociological factors drop out.

This week we tackle two very different concepts that are central to science: experimentation and explanation. Next week, we try to leverage explanatory considerations to solve (or at least make progress on) the problem of induction. After that, we’ll address a perennial concern for philosophy of science: the realism/anti-realism debate, drawing on lots of the background you’ve been acquiring. So much for foreshadowing. To the work at hand. . . . 

Start with experiment. Do we even have a clear idea of what it is? How, for instance, does it differ from simple observation. (Of course, as we’ve already seen — both from reading French and Feyerabend —, observation isn’t nearly as simple as we’re often inclined to suppose.) How should experiments fit in with theories? Such questions are important not just theoretically, but (as O’Malley et al. argue) practically for how science is performed and funded. Their paper — published in a high-profile biology journal — examines some of the statements of funding agencies like the NSF and NIH to see how well they fit into our best understanding of how science works.

Our topic for Thursday will be explanation. What is it to explain something? I take it that we are often fairly good at offering and evaluating explanations. But once again we run into the problem of not being very good at describing what it is we’re doing. Strevens’ paper surveys some of the most popular and important accounts of scientific explanation (and the problems that they face). Though his paper doesn’t mention this specifically, you might think about the methodology of investigating these various models. How exactly are we (and should we) approach the question of whether an account of explanation is adequate?

Tuesday (11/1): Experiment & Models

• French, Ch. 6 (you might wish to review Ch. 5 as well)

• Hacking, “Experiment” (Chapter 9 of Representing and Intervening) [PDF]*

• O’Malley et al. “Philosophies of Funding” [PDF]

Questions: (respond to one)

1. What is the difference between observation and experimentation? Describe as clearly and neutrally as you can (i.e., see if you can avoid using those words to explain the difference). Is there a clean division between the two?
2. What do you think of Hacking’s view that phenomena are “created” by experiment — that they are, in a sense, artifacts of our technology? What would the consequences of this claim be, if true?
3. Say something about how your think models fit into science. You might want to think back to the Oreskes/Conway reading.
4. How do Hacking’s insights play a role in the O’Malley et al. paper? 

Thursday (11/3): Explanation

• Strevens, “Scientific Explanation” [PDF] (you may skim the sections on the IS account and the Statistical Relevance account — I won’t address these in class unless someone specifically wants to).

• French, pp. 98–99 briefly considers explanation: you might wish to look this over at this point too (it’s in Ch. 7 on Realism, which we’ll read in a few classes). 

Questions: (respond to one)

1. Read Study Exercise 2 in French (p. 88). Address the questions: Do you think it’s possible to identify the cause of the crash? and Do you think scientists face a similar sort of situation when they try to explain some phenomenon?
2. See if you can come up with a different example along the lines of the flagpole and storm examples that illustrates a problem for the DN account of explanation.
3. Between the Unification and Causal approaches to explanation, what appears to you to be the most appealing and why?
4. Address the methodological question I broached above.