Update: Readers may be amused at my reply last night in the group where this was sent out.
Two Three Four FiveSix immediate thoughts: First, I have criticisms of Haidt that don't much overlap with these. Second, Social Psychologists have been gunning for Haidt right from the start, and have really worked hard to show that MFT is wrong, because they prefer something else. (Third, ed.) Related to this is the odd criticism here that MFT isn't founded on any theory, with the insistence that it needs a theory, and that's the only way to correct things. Fourth, MFT has had a sixth foundation for years, discovered empirically, and that is not reflected in the criticism. That is a huge error, frankly. We have to keep up here. Fifth. While MFT purports to apply to mankind in general, it has been clear right from the start that it was developed on Americans and mostly reflects them, only gradually branching out to figure out how this applies to all mankind. So making a big deal that it doesn't seem to work in Korea and Turkey because it doesn't reach a conventional threshhold for "model fit." Sort of what you'd expect of a theory still being developed, isn't it? Six, the neglect of kinship, signalling, and reciprocity distinctions in MFT are excellent and quite justified.
I always think better while writing. I thought I was going to just put in two sentences and hit send, but things kept occurring to me. That crap that one learns more by listening isn't all that true. I teach myself by trying to write much better than nearly all teachers have done for me. Things occur to me along the way and I can't type fast enough. My long essay isn't likely to be much better than what I have just shot off, just more thorough with some corrections and cautions.I still haven't taken my walk to think about this, but I have reread the essay, and also the one that DOuglas2 links to in the comments, in which Haidt defended himself against similar complaints in 2011. I learned something about the field and how Haidt has developed his moral foundations ideas.
8 comments:
Perhaps he was trying the move that Saul Kripke got away with. I forget the exact wording, but he published a work that said something like 'I will not propose a theory, because you would simply poke holes in it.'
Not having read the more academic roots of the Moral Foundations Questionnaire, I'd kind of filed it with Meyers Briggs Type Index as something perhaps not very robust and repeatable but nevertheless useful in the "lie to children" sense (https://en.wikipedia.org/wiki/Lie-to-children) in order to help people be more accepting of those different from themselves.
It never occurred to me, although it certainly should have, that it was aimed at being real psychometric scales in the Guttman sense (https://en.wikipedia.org/wiki/Guttman_scale)
So there have been international studies where they attempted to validate the MFQ scales by translating the questions into a different language and administering the test to the members of the culture using that language.
A similar thing was my introduction to Guttman scaling, because we were trying to use instruments developed in Nordic countries by translating the questions to English, and it didn't work, because translation is hard and inexact. That those evaluating MFQ in this way see MFQ as the problem (rather than their replication methods) is troublesome to me.
I'm also troubled by Curry's use of the word "unprincipled". He obviously thinks he can get away with it by saying "sorry if I was misunderstood, I didn't mean that word in the meaning it has in every English dictionary, but as technical lingo". Haidt in an article linked in this critique says:
"If you start by fixating on a principle (e.g.,
that morality is justice, or empathy, or harm reduction, or
prosocial behavior) and then develop your theory in a
logical way on the basis of that principle, you will construct
an elegant and parsimonious theory, but it will crack under
the weight of empirical data"
(http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.834.1397&rep=rep1&type=pdf)
The contrast is between what critics describe as the ad-hoc selection of dimensional scales by the MFT team vs scales based upon a priori "principles", so what Curry will claim he means by "unprincipled" is proceeding from observation rather than from from theoretical deduction.
Beyond that, Curry provides himself a link to an article that Haidt et al wrote to rebut previous criticism, and I don't see that he has managed himself to overcome any of the same issues that were rebutted.
I often think that people include links hoping that their mere presence will bolster their case in the minds of their readers, and hoping even more that those same readers won't actually read what's provided in the link, because it is actually damaging to their argument.
I should note that when I wrote above
""sorry if I was misunderstood, I didn't mean that word in the meaning it has in every English dictionary, but as technical lingo"", that was not an actual quote -
I'm not at all suggesting that he said that, and even imputing that he thought that may be unfair.
I believe it is Haidt's intent to develop something that is not only "useful in getting a discussion going without rancor" but a framework that is accurate. I recall when he started he was very clear that this was new stuff that he was discovering and it would take a while to sort it out. Ten years later, I don't know if he thinks this is still new or now established.
I've just realized that I've been confusing Jonathan Haidt with the truly annoying Jonathan Chait.
Well don't do that. It will cause you great irritation in so much of life.
Back when I used to write, I walked. I would walk long after midnight along the bay front in San Diego from my street, the foot of Grape all the way out to Tom Hamms lighthouse on the old island and back with paper and pen in my pocket so I could stop under a street light and jot down the things that were so clear while out walking.
what you said in the comments! ditto.
Post a Comment