> It is very easy for anyone to only care about their explicit moral principles (Freedom! Security! Equality! Social Justice!), durably ignore all of their implicit moral beliefs, and then assume that their opposition are assholes who deny their values.
>
> Whereas the reality is that the other side is often doing that very same thing. And this leads to very bad attractors.
That’s not what Haidt found it its "moral foundations" work. He found that, yes, liberals focus on only one (well, two) moral principle and ignore others, but the error is *not* symmetrical, conservatives having a wider set of principles.
For the sake of honesty: I usually pretty quickly discard such contextless graphs.
This graph can result from _conservatives caring equally about these values_, _all conservatives putting 3 to every moral question whatsoever_, or _all conservatives putting 5 to 2 of these 5 axes, but just different axes_.
Even with the examples above, I am engaging with the subject matter too directly.
It can just be selection bias at the population level (stats are hard) or at the indicators level (no box plot).
More fundamentally, values are much more complex than 5 axes. There are too many degrees of freedom in picking these 5 ones, and in how one may frame the questions around them.
To be clear, if you have an article or an essay giving the context, I would likely read it. Explaining the set-up, discussing how reliable it is, showing what other experiments Haidt ran, why he ran these specific ones, what were his actual policy opinions, etc.
I should have prefaced my post with an epistemic status note. I’m not saying "Science Has Proven You Wrong", I’m saying more "my-superficial-reading-of-your-post" is very much not compatible with "my-superficial-reading-of-moral-foundations-theory", without any solid opinion on which one of the two is closer to truth (I only know of moral foundations theory through Bryan Caplan’s blog posts about it).
More "have you considered existing literature on that topic ?" than "you’re wrong and stupid !"
I misunderstood you (but I still thank you for your answer!)
I haven't done a literature review of the field, but I had already read about Moral Foundations theory.
In general, I am not very interested in questionnaire approaches. I think they clearly measure something, I simply do not think it's very related to the things I care about.
(For instance, this is linked to my skepticism of evaluating the alignment of LLM models through questionnaires.)
> It is very easy for anyone to only care about their explicit moral principles (Freedom! Security! Equality! Social Justice!), durably ignore all of their implicit moral beliefs, and then assume that their opposition are assholes who deny their values.
>
> Whereas the reality is that the other side is often doing that very same thing. And this leads to very bad attractors.
That’s not what Haidt found it its "moral foundations" work. He found that, yes, liberals focus on only one (well, two) moral principle and ignore others, but the error is *not* symmetrical, conservatives having a wider set of principles.
See for example https://en.wikipedia.org/wiki/Moral_foundations_theory#/media/File:Haidt-political_morality.png
Thanks for your answer!
For the sake of honesty: I usually pretty quickly discard such contextless graphs.
This graph can result from _conservatives caring equally about these values_, _all conservatives putting 3 to every moral question whatsoever_, or _all conservatives putting 5 to 2 of these 5 axes, but just different axes_.
Even with the examples above, I am engaging with the subject matter too directly.
It can just be selection bias at the population level (stats are hard) or at the indicators level (no box plot).
More fundamentally, values are much more complex than 5 axes. There are too many degrees of freedom in picking these 5 ones, and in how one may frame the questions around them.
To be clear, if you have an article or an essay giving the context, I would likely read it. Explaining the set-up, discussing how reliable it is, showing what other experiments Haidt ran, why he ran these specific ones, what were his actual policy opinions, etc.
This would be great.
I should have prefaced my post with an epistemic status note. I’m not saying "Science Has Proven You Wrong", I’m saying more "my-superficial-reading-of-your-post" is very much not compatible with "my-superficial-reading-of-moral-foundations-theory", without any solid opinion on which one of the two is closer to truth (I only know of moral foundations theory through Bryan Caplan’s blog posts about it).
More "have you considered existing literature on that topic ?" than "you’re wrong and stupid !"
As far as I can tell, that graph is table 2 of https://web.archive.org/web/20170809220202/http://www-bcf.usc.edu/~jessegra/papers/GNHIKD.2011.Mapping%20the%20moral%20domain.JPSP.pdf
I misunderstood you (but I still thank you for your answer!)
I haven't done a literature review of the field, but I had already read about Moral Foundations theory.
In general, I am not very interested in questionnaire approaches. I think they clearly measure something, I simply do not think it's very related to the things I care about.
(For instance, this is linked to my skepticism of evaluating the alignment of LLM models through questionnaires.)