I was also surprised that a plurality of people preferred dust specks to torture, given that it appears to be just a classic problem of scope insensitivity, which this site talks about repeatedly.
I was surprised as well, but I disagree that it is necessarily scope insensitivity—believing utility is continuously additive requires choosing torture. But some people take that as evidence that utility is not additive—more technically, evidence that utility is not the appropriate analysis of morality (aka picking deontology or virtue ethics or somesuch).
More specific analysis here and more generally here.
That’s only weak evidence about the correlation between non-consequentialism and dust specking. If we had 670 consequentialists, 50 deontologists, 180 virtue ethicists, and 200 others, and 40% of each chose dust specks, we’d get numbers like yours even though there wouldn’t be a correlation.
I did a crosstab, which should be more informative:
I get different totals for the number of speckers (397) and non-consequentialists (386), though. Maybe my copy of the data’s messed up? (Gnumeric complains the XLS might be corrupt.)
Anyway, I do see a correlation between specks & moral paradigm. My dust speck percentages:
41% for consequentialism (N = 560)
67% for deontology (N = 36)
47% for other/none (N = 145)
65% for virtue ethics (N = 116)
leaving out people who didn’t answer. Consequentialists chose dust specks at a lower rate than each other group (which chi-squared tests confirm is statistically significant). But 41% of our consequentialists did still choose dust specks.
[Edit: “indentation is preserved”, my arse. I am not a Markdown fan.]
Well, you cannot be totally sure. I for one would consider myself a consequentialist, but would still choose dust specks. Correlation doesn’t imply causation!
Well, I guess there are various forms of Consequentialism which would lead one to choose dust specks. That would simply depend on what you’re trying to maximize.
If you want to maximize things like pain, discomfort or the amount of dust in eyes, then yes, you would choose dustspecks.
If, on the other hand, you wanted to maximize the amount of, say, wellbeing, then the only choice available is torture.
It’s not clear to me that one can’t be a utilitarian without agreeing that utility is additive (at least, additive in that manner). Consequentialism makes way more sense to me than deontology or virtue ethics (i.e. “what’s the point of deontology or virtue ethics if it doesn’t give better results?”), but I not only remain completely unconvinced by the arguments for Torture (that I’ve seen, anyway), but also think that Eliezer’s choice of Torture contradicts some of his other posts. But this is probably not the place to have that discussion.
I think Torture vs Dust Specks is really just Eliezer being coy about prioritarianism, as an analogous issue not known by that name emerges from prioritarian maths.
Could you clarify what you mean when you say that Eliezer is “being coy about” prioritarianism?
As for me, I’d never heard of prioritarianism before; having just read the wikipedia article (which does have some style disclaimers and “citation needed”s, so perhaps is not the ideal source), I don’t think it addresses either of my objections. It does at least attempt to capture some of my intuitions about the Specks vs. Torture case.
I was surprised as well, but I disagree that it is necessarily scope insensitivity—believing utility is continuously additive requires choosing torture. But some people take that as evidence that utility is not additive—more technically, evidence that utility is not the appropriate analysis of morality (aka picking deontology or virtue ethics or somesuch).
More specific analysis here and more generally here.
In support of this, 435 people chose specks, and 430 chose virtue ethics, deontology, or other.
That’s only weak evidence about the correlation between non-consequentialism and dust specking. If we had 670 consequentialists, 50 deontologists, 180 virtue ethicists, and 200 others, and 40% of each chose dust specks, we’d get numbers like yours even though there wouldn’t be a correlation.
I did a crosstab, which should be more informative:
I get different totals for the number of speckers (397) and non-consequentialists (386), though. Maybe my copy of the data’s messed up? (Gnumeric complains the XLS might be corrupt.)
Anyway, I do see a correlation between specks & moral paradigm. My dust speck percentages:
41% for consequentialism (N = 560)
67% for deontology (N = 36)
47% for other/none (N = 145)
65% for virtue ethics (N = 116)
leaving out people who didn’t answer. Consequentialists chose dust specks at a lower rate than each other group (which chi-squared tests confirm is statistically significant). But 41% of our consequentialists did still choose dust specks.
[Edit: “indentation is preserved”, my arse. I am not a Markdown fan.]
I think we’ve found our answer, then.ETA: Really nice work from satt to prove I was jumping to conclusions here.
Well, you cannot be totally sure. I for one would consider myself a consequentialist, but would still choose dust specks. Correlation doesn’t imply causation!
Well, I guess there are various forms of Consequentialism which would lead one to choose dust specks. That would simply depend on what you’re trying to maximize.
If you want to maximize things like pain, discomfort or the amount of dust in eyes, then yes, you would choose dustspecks.
If, on the other hand, you wanted to maximize the amount of, say, wellbeing, then the only choice available is torture.
It’s not clear to me that one can’t be a utilitarian without agreeing that utility is additive (at least, additive in that manner). Consequentialism makes way more sense to me than deontology or virtue ethics (i.e. “what’s the point of deontology or virtue ethics if it doesn’t give better results?”), but I not only remain completely unconvinced by the arguments for Torture (that I’ve seen, anyway), but also think that Eliezer’s choice of Torture contradicts some of his other posts. But this is probably not the place to have that discussion.
I think Torture vs Dust Specks is really just Eliezer being coy about prioritarianism, as an analogous issue not known by that name emerges from prioritarian maths.
Could you clarify what you mean when you say that Eliezer is “being coy about” prioritarianism?
As for me, I’d never heard of prioritarianism before; having just read the wikipedia article (which does have some style disclaimers and “citation needed”s, so perhaps is not the ideal source), I don’t think it addresses either of my objections. It does at least attempt to capture some of my intuitions about the Specks vs. Torture case.