Interesting paper, thanks! From a quick skim it seems to me that when asked moral questions—“is Fred morally responsible for his actions?”—most of the subjects expressed compatibilist intuitions. They only start expressing incompatibilist intuitions when asked to comment on abstract philosophical statements of a kind one would not normally encounter. So it seems to me that the data upholds my claim, with the addendum that when a layman is asked to dabble in philosophy he does indeed fall into the classic error into which professional philosophers have fallen.
They only start expressing incompatibilist intuitions when asked to comment on abstract philosophical statements of a kind one would not normally encounter.
This is the Nichols and Knobe hypothesis which argued that people in general are incompatibilists but that language which generates strong affective responses will nonetheless lead people to import moral responsibility. The hypothesis is formed by taking some vignette about free will and making the action significantly more condemnable. So people will think that someone who gives to others in a deterministic world is not responsible but someone who murders others in that world is. Every other feature of the story remains the same. The stories that generate incompatibilist intuitions aren’t different from those that generate compatibilist intuitions except in the emotional/morally condemnable content. The former aren’t more abstract or philosophical. The better interpretation of this hypothesis is that people’s actual intuitions about determinism get overrun by a desire to signal that they do not support the evil action committed in the vignette. Feel free to google Knobe’s work in intentionality for evidence that this phenomena is more general.
In any case, the Nichols and Knobe hypothesis isn’t the position of the article. The authors set out to test the hypothesis and found it false. Instead, they found that people responded to vignettes mostly consistently. About 60-67% gave incompatibilist responses, 22-9% gave compatibilist responses and mixed/inconsistent responses were returned at a high single digit rate.
Interesting paper, thanks! From a quick skim it seems to me that when asked moral questions—“is Fred morally responsible for his actions?”—most of the subjects expressed compatibilist intuitions. They only start expressing incompatibilist intuitions when asked to comment on abstract philosophical statements of a kind one would not normally encounter. So it seems to me that the data upholds my claim, with the addendum that when a layman is asked to dabble in philosophy he does indeed fall into the classic error into which professional philosophers have fallen.
This is the Nichols and Knobe hypothesis which argued that people in general are incompatibilists but that language which generates strong affective responses will nonetheless lead people to import moral responsibility. The hypothesis is formed by taking some vignette about free will and making the action significantly more condemnable. So people will think that someone who gives to others in a deterministic world is not responsible but someone who murders others in that world is. Every other feature of the story remains the same. The stories that generate incompatibilist intuitions aren’t different from those that generate compatibilist intuitions except in the emotional/morally condemnable content. The former aren’t more abstract or philosophical. The better interpretation of this hypothesis is that people’s actual intuitions about determinism get overrun by a desire to signal that they do not support the evil action committed in the vignette. Feel free to google Knobe’s work in intentionality for evidence that this phenomena is more general.
In any case, the Nichols and Knobe hypothesis isn’t the position of the article. The authors set out to test the hypothesis and found it false. Instead, they found that people responded to vignettes mostly consistently. About 60-67% gave incompatibilist responses, 22-9% gave compatibilist responses and mixed/inconsistent responses were returned at a high single digit rate.