Report comment

Oh, and deer. But they don’t vote.

Report comment

Report comment

It’s strange to hear the groups describe themselves like this, because currently in the US, the right is tarred as selfish and self-absorbed (while they describe themselves as individualistic) and the left is criticized as collectivist (while they call themselves unselfish) – though it was somewhat different during the Bush administration. My suspicion is that the out-of-power group is typically more individualist and the ins are more collectivist, regardless of whether left or right is in power.

Report comment

Report comment

What he cannot do here is just automatically defect from Nazism because he finds they’re doing something morally wrong. Being part of a group means that you can’t always have your way. The question them becomes whether it would be worth sacrificing a restored German Empire to save these people.

Here, there is a smooth gradient in how much divergence from your own views you are willing to tolerate among people you consider to be your in-group. Set this to zero and you have social dysfunction on the level of some parts of the French Revolution. Set it very high and you have something like an idealized fascism that requires no secret police because everyone goes along with the program sincerely.

Report comment

Report comment

Report comment

Like we could go around to the Nazis, and say “Look, you agree that helping the poor is moral, right? And caring for your aged parents? Well, now that we’ve established what morality is, we have bad news for you. You don’t have it. Moral people are much more likely to oppose you. So stop doing what you’re doing.”

I don’t think this will work Consider what would happen if it turned out that Nazis were more moral in these other ways than non-Nazis. Would you accept that as an argument for Naziism? No? Then why should it be acceptable as an argument against Naziism if the statistics go the other way?

I can think of plenty of scenarios where evil acts are associated with other good acts. For instance, consider a Nazi who is motivated to kill Jews because it gets him a promotion and that benefits his family. A selfish bastard who doesn’t care about his family would not be a Nazi for this reason. He would still want the personal benefits that the promotion brings him, but personal + family benefits are a greater motivation than personal benefits only.

It is entirely plausible–not guaranteed, maybe not even 50% likely, but certainly plausible–that the “I want to be a Nazi to benefit my family” effect associates Naziism with good things more than “I don’t mind being a Nazi because murder doesn’t bother me” associates it with bad things.

Report comment

Similar stuff to DW-Nominate is widely applied in psychology, medicine and testing, as Item Response Theory. Here, an underlying trait is derived circularly from the interaction between students and questions (or patients and questionnaires), learning both how good the students are and how hard the questions are from just looking at correctness of responses. The results are entirely relative, and if you don’t specify an orientation (such as “more correct answers should produce a positive increase in the underlying trait), the scale and location on the scale will be arbitrary, but the relationship between the items/questions and the measured traits will always “work”: the better students will all be at one end of the scale, and the harder questions will be at the same end of the scale. When you measure disparate groups over time (such as with DW-Nominate), the arbitrariness of scale and location make it impossible to just compare numbers between different groups, but under certain circumstances we can establish that there’s a linear relationship between two distinct scales, and use items or students (or bills and politicians) which turn up in both groups in order to “link” [that’s the actual technical term] the two.

To be clear, when I talk about “correctness” above, that’s only meaningful in a testing situation: if you’re measuring depression or attitude to global warming, there’s no right answer, but certain answers will (in a good test) strongly correlate with the respondent’s underlying trait being at one end of the scale or the other, or past or before a certain point on the scale.

Finally, one of the things you’re getting at is that maybe left-right isn’t the only spectrum. It is indeed a choice here to declare that there is only one underlying trait being measured, and such a choice may be valid (if these are all topology questions) or invalid (if there are also questions about cooking). You can actually experiment with allowing multiple underlying traits, and compute the likelihood of your results under 1-trait, 2-trait, etc. assumptions. Unfortunately the greatest likelihood will occur when number-of-traits ≈ number-of-questions (or bills, or whatever), so you end up having to apply Ockham’s razor or some similar rule to cut things off at a reasonable point.

So you’re almost certainly correct that reality is better described by allowing for left-rightness *and* gun-friendliness or whatever, *but* you may find (as it would appear DW-Nominate has found) that you gain very little from the additional trait, and your model of the world is accurate-enough with only one trait which turns out to correlate highly with our left-right vocabulary. The circularity of social construction of political action in context of left-right vocabulary then becomes problematic in the same way as descriptions of single malt whisky.

Report comment