I have been doing some work with Jonathan Livengood, and now David Rose, on folk judgments about causation, focusing on judgments with regard to several of the vignettes that Joshua Knobe gives in his dissertation (2006). Specifically, we have looked at the case of Lauren and Jane; on Knobe’s original version of the story, it reads:
Lauren and Jane work for the same company. They each need to use a computer for work sometimes.
Unfortunately, the computer isn’t very powerful. If two people are logged on at the same time, it usually crashes.
So the company decided to institute an official policy. It declared that Lauren would be the only one permitted to use the computer in the mornings and that Jane would be the only one permitted to use the computer in the afternoons.
As expected, Lauren logged on the computer the next day at 9:00 am.
But Jane decided to disobey the official policy. She also logged on at 9:00 am.
The computer crashed immediately.
Knobe uses this story to pump some intuitions about the role of permissibility (or more generally the normative status of an action) in judgments about causation. He continues:
Here we would attribute the crash more to Lauren’s behavior than to Jane’s behavior. But why? The two factors appear to be similar in many ways. The most plausible hypothesis seems to be that this difference in people’s attributions stems from a difference in perceived normative status — i.e., from a sense that Jane was not doing anything wrong but Lauren really ought to have refrained from using the computer. (p. 68)
Note that the discussion of people’s attributions does not derive from an empirical study using the vignette; no such data is presented. Rather, Knobe uses the story as an intuition pump. This is perfectly fine. My concern is not with the use of a thought experiment, per se; rather, as Knobe is well known as an experimental philosopher, my concern is that there is a danger of researchers mistaking the intuition pumped for a conclusion drawn from empirical work testing people’s responses to questions about who caused the outcome in the story. In fact, I’ve seen several suggestions of this in the literature discussing this case, but perhaps the most clear example is in a recent paper by Roxborough, Cumby, and Fraser currently under review (“Causation, Morality & Atypicality”). They write:
Joshua Knobe (2006) claims that moral judgments play a direct role in folk attributions of causal responsibility. Knobe makes his case using data from a study in which he presented subjects with the following vignette: [see above]. Neither Jane nor Lauren logging on alone would have crashed the computer. Both Jane and Lauren had to be logged on for the computer to crash. Nevertheless, Knob’s subjects were significantly more willing to attribute causal responsibility for the computer crash to Jane than to Lauren. The only difference between Jane and Lauren seems to lie in the moral status of their actions: Jane broke a rule while Lauren did not. Knobe thus concluded that moral judgments influence folk attributions of causal responsibility. (p. 2)
Of course, there is no study in Knobe (2006), no subjects, and no significant result. This is the danger of experimental philosophers using thought experiments.
Fortunately, in this case Knobe’s intuitions were good. Jonathan and I have run a sequence of variations on the original Lauren and Jane case. Two of those variations serve to test whether English-speakers really are more willing to attribute causal responsibility to Jane than to Lauren when Jane’s action is impermissible (compared to when there is no rule governing logging on to the computer). As our goal was somewhat larger than simply to test people’s intuitions about the basic story, it has been expanded a bit to allow us to vary the timing of when Lauren and Jane log on (more on this in a subsequent post). The results for the two conditions when Lauren and Jane log on at the same time follow Knobe’s prediction.
The two vignettes read as follows:
No permissibility information included:
Lauren and Jane both work for a company that uses a mainframe that can be accessed from terminals on different floors of its building. Though the company does not know it, the mainframe has recently become unstable, so that if more than one person is logged in at the same time, the system crashes.
One day, Lauren logged into the mainframe on the ground floor at the exact same time that Jane logged into the mainframe on the second floor. Lauren and Jane were both unaware that the other was logging in. Sure enough, the system crashed.
Permissibility information included:
Lauren and Jane both work for a company that uses a mainframe that can be accessed from terminals on different floors of its building. The mainframe has recently become unstable, so that if more than one person is logged in at the same time, the system crashes. Therefore, the company has instituted a temporary policy restricting the use of terminals so that two terminals are not used at the same time until the mainframe is repaired. The policy prohibits logging in to the mainframe from the terminal on any floor except the ground floor.
One day, Lauren logged into the mainframe on the authorized terminal on the ground floor at the exact same time that Jane logged into the mainframe on the unauthorized terminal on the second floor. Lauren and Jane were both unaware that the other was logging in. Sure enough, the system crashed.
Participants were randomly given one of the two vignettes, then asked to rank whether they agree or disagree with each of the two statements given below on a 7-point scale anchored at 1 with “strongly disagree,” at 4 with “neutral,” and at 7 with “strongly agree”:
1. Lauren caused the system to crash.
2. Jane caused the system to crash.
We collected responses for these two probes from 195 participants through the Philosophical Personality website. 38 participants were excluded because they were under 18 years of age, had taken the survey previously, or did not complete the demographic information portion of the survey. Additionally, for the sake of comparison, 14 participants were excluded because they were non-native English speakers or because they had more than minimal training in philosophy. The remaining 143 participants were 73.4% female, with an average age of 34.8 years, and ranging in age from 18 to 81 years old.
As expected, we found that participants in the “no permissibility information included” condition did not treat Lauren and Jane differently, tending to deny that either caused the system to crash (N=71; Lauren: M=2.70; Jane: M=2.70), while participants in the “permissibility information included” condition did treat them differently, tending to respond that Jane, but not Lauren, caused the system to crash (N=72; Lauren: M=2.42; Jane: M=5.21). The results are shown graphically below.
Edit: Roxborough, Cumby, and Fraser have corrected their paper (and the revised version should be online in the near future). They are hardly the only ones to have made this mistake, however. For example, Christopher Hitchcock (2007, “Three Concepts of Causation”) indicates that Knobe and Fraser performed the experiment, presumably referring to the paper that became their (2008). He writes: “[In] an experiment performed by Knobe and Fraser, subjects were presented a vignette in which two individuals, Lauren and Jane, both use the same computer…. In this scenario, subjects are more strongly inclined to judge that Jane caused the computer to crash than that Lauren did.” (page 512).
