I didn’t want to get into the statistical validity bit in the article, but I am somewhat skeptical of this. Haidt started with a theory of the sorts of moral judgments people make, and developed questions to isolate the foundations he was looking for.
The correct way to do this sort of research would be to ask a vast range of questions and see what clusters emerge, and then attempt to characterize or pin down those clusters. That would have avoided missing the liberty foundation, and might have given a broader version of the sacredness framework which includes things that liberals consider sacred.
It’s easy to ask a vast range of questions—a lot harder to get them all answered. Data isn’t free. He targeted his data acquisition to modalities he had some evidence for. I haven’t followed his work in detail, but I’d guess that he had other trial modalities that didn’t pan out.
Anyone know anything about other candidate modalities he looked at?
I didn’t mean to imply that he has totally characterized all the pattern matching involved in morality, and I doubt he has claimed that either. When confronted with new evidence—likely some squawking from libertarians—he updated his model. I’d expect him to do the same if someone came up with evidence for another moral modality.
He’s started the reductionist enterprise on morality as it exists. It’s about time someone did.
The impression I get from reading his research is that he came at it from an anthropological background (his advisor, IIRC, was an anthropologist).
My worry is that he is making the same error that early personality tests (Myers-Briggs, for instance) made; yes, they tested something, but not necessarily what they thought they were testing. Statistical tools are more powerful now, but I’m not sure they protect against this sort of error. As others have pointed out, liberals do have a strong sense of the sacred (in the environmental, and in food in particular); Haidt’s test doesn’t measure this and doesn’t have any way of detecting that this is missing.
I didn’t want to get into the statistical validity bit in the article, but I am somewhat skeptical of this. Haidt started with a theory of the sorts of moral judgments people make, and developed questions to isolate the foundations he was looking for.
The correct way to do this sort of research would be to ask a vast range of questions and see what clusters emerge, and then attempt to characterize or pin down those clusters. That would have avoided missing the liberty foundation, and might have given a broader version of the sacredness framework which includes things that liberals consider sacred.
It’s easy to ask a vast range of questions—a lot harder to get them all answered. Data isn’t free. He targeted his data acquisition to modalities he had some evidence for. I haven’t followed his work in detail, but I’d guess that he had other trial modalities that didn’t pan out.
Anyone know anything about other candidate modalities he looked at?
I didn’t mean to imply that he has totally characterized all the pattern matching involved in morality, and I doubt he has claimed that either. When confronted with new evidence—likely some squawking from libertarians—he updated his model. I’d expect him to do the same if someone came up with evidence for another moral modality.
He’s started the reductionist enterprise on morality as it exists. It’s about time someone did.
The impression I get from reading his research is that he came at it from an anthropological background (his advisor, IIRC, was an anthropologist).
My worry is that he is making the same error that early personality tests (Myers-Briggs, for instance) made; yes, they tested something, but not necessarily what they thought they were testing. Statistical tools are more powerful now, but I’m not sure they protect against this sort of error. As others have pointed out, liberals do have a strong sense of the sacred (in the environmental, and in food in particular); Haidt’s test doesn’t measure this and doesn’t have any way of detecting that this is missing.