Knobe or not Knobe, that is the question

Wed 02 Nov 2011 01:20 PM

A few weeks ago, I did an exercise in my intro course in which students read descriptions of two scenarios, answered some multiple choice questions individually. They then discussed their answers in groups, and we discussed them as a class.

Morton is a physicist working on a the properties of particular semiconductors. He is interested in this as a scientific problem and is only studying it because of its theoretical significance.

Yet the only obvious applications are in alternative energy. Ultimately, his research is used to develop solar technology, and the technology is used to produce power in ways that produce significantly less pollution than other methods would have done.

Marsha is a chemist working on a class of interesting synthesis problems. She is interested in this as a scientific problem and is only studying it because of its theoretical significance.

Yet the only obvious applications are military. Ultimately, her research is used to develop weapons, and the weapons are used to commit atrocities which probably would not have been committed without those weapons.

Just one of these two cases would have been enough for the topic we had read about, which was whether scientific significance can really be insulated from practical significance. I juxtaposed of the two cases, though, because the paradigm case for experimental philosophy. I was curious.

One of the things I asked was whether Morton deserves any credit for the reduction in pollution and whether Marsha deserves any blame for the deaths. Standard ethical theory suggests that the answers should be symmetrical: either both deserve credit/blame or neither do. The Knobe effect (named for Joshua Knobe) suggests that students should blame Marsha but refuse to credit Morton.

As a matter of fact, neither of those things happened. Most students answered asymmetrically. Of those, most were willing to give some credit to Morton but unwilling to blame Marsha.

I do not have anything systematic to say about this. I did not collect precise numbers, since it was a pedagogical exercise rather than an experimental one. (We discussed human subjects protections in the same class session, and I commented that I couldn't use the results in a paper even if I had recorded them.) The discussion also revealed that responses were shaped by the details of how the scenarios and questions were worded. For example, one student did not want to blame Marsha for atrocities but would have blamed her for more quotidian deaths.

Nevertheless, I wonder whether it matters that the actors in these scenarios are scientists whereas the actors in Knobe's original cases were businessmen.