The Multi-Dimensional Moral Matrix

Recent events compel to revisit Jonathan Haidt's concept of the moral matrix. I've covered this a few times before, but it's worth trying to draw some different ideas together.

The concept is that a person's moral world view is governed by how much they value five key attributes : care, fairness, loyalty, authority, and purity. Liberals value care and fairness most of all whereas conservatives prefer loyalty, authority and purity.

This system is fine as long as we restrict its use to understanding people's ideologies and not morality itself. The actual moral outcomes of these different viewpoints are almost entirely context-dependent. For example, authoritarianism has a great deal to offer in an emergency while "fairness" (or more accurately reciprocity) doesn't really come into it. If a ship's heading towards an iceberg, you want the captain to immediately overrule any objections and steer it away, not sit down for a protracted discussion about who's in charge. Likewise, "care" is great during a humanitarian catastrophe but pretty useless if someone's holding a gun to your head.

Purity, however, is perhaps where things come unstuck. Certainly this is something that people do value to different degrees, but in what sense that can be said to be anything related to morality is something I haven't been able to determine (suggestions welcome*). It highlights very clearly that this system is about world views, not morality itself.

* The only thing I can think of is from an online psychology course, reminding the audience that people have great difficulty defining why incest is so icky whilst being of the very firm opinion that it is.

The moral matrix (though perhaps it's more of an ideology or political matrix) is useful in reminding us that people can have perfectly decent, honourable intentions but reach different conclusions due to their different innate values. But it's at best highly incomplete when it comes to morality proper. If, for example, you genuinely thought that your freedom and democracy were threatened, would you not think that taking direct violent action was morally justified ? Democracy was, after all, born in blood. Yet while you might sincerely believe that your action was moral, whether you actually were acting morally or not depends firmly on context.

There were two recent comments on social media that made my blood boil. One was that we should "condemn ALL the violence", including that of Black Lives Matter protestors, which is plainly ludicrous - and wasn't helped by the same person linking to a site called "hang the censors" without any apparent embarrassment (I've explained the asymmetry of the issues before so I won't do it again). The second, by another*, was that the recent events in America were not an attempted coup** but a "protest gone sideways". Yes, because it's all so easy for a protest to go sideways if you come armed with guns and pipe bombs and maps of the building you're planning to accidentally storm in a fit of pique. The only interpretation of this that makes any sense is if the rioters were actually crabs.

* A much more respectable fellow, I think, but completely wrong in this particular instance.
** Albeit with piss-poor prior planning, as befits the stupidity of its instigator.

I've mentioned before that if we're constructing a truly moral matrix, respect for the truth has to be included in there somehow. These comments reminded me of two other key factors.

First, a moral person should be able to overcome their own bias. If, according to your own standards, data indicates a viewpoint contrary to your existing one, then your belief should shift towards it and away from your existing stance. Note the careful wording; there is no conflict here with the fact that data doesn't speak for itself and interpretation is something that happens only within your head. And you should follow a similar progression not only with regards to specific conclusions, but how you evaluate information as well. When you learn about different fallacies, you should then take account of them.

Second, a moral person needs a degree of analytic intelligence, to be able to draw the correct conclusions (at least according to their own standards) from the data. If you can't see how the presence of carrying an explosive device into the heart of democracy is the action of a terrorist, you might simply come up short here - a badly organised attempted coup is still, FFS(!), an attempted coup. There might be other reasons, but simple stupidity is a perfectly reasonable explanation at this point. It's true that calling people stupid won't help persuade them, but that's not the goal here.

The "moral matrix" has some real use in reminding us that people genuinely think differently from one another and that those viewpoints are not innately moral or immoral. A true moral matrix, however, would have to be rather different. It would have to include compassion, certainly, and fairness also - but likely greatly expanded beyond the simple idea of reciprocity. Loyalty, for instance, is only sensible when someone acts fairly; being loyal to someone actively trying to harm you is not moral but daft. Likewise authority : having trust in an expert in an emergency is very different from wanting a strongman in power in all situations. One is fair, the other is not. Purity, as far as I can tell, would not come into it at all, while both analytical and critical intelligence would surely deserve inclusion.

So the real moral matrix would perhaps not be a set of simple sliders, but a complex, multi-dimensional parameter space. Such a matrix would have to include true malevolence, which we might, as a starting point, define as wanting to harm others who have never harmed anyone else nor had any intentions to do so, in full knowledge of the situation (this is something that neither Plato nor Epictetus believed was possible, but I'm convinced otherwise). Hypocrisy might feature in there as well. Designing such a system would be a formidable challenge that's quite beyond me.

But for all its complexity, all its shades of multidimensional grey, there are extremes which any halfwit ought to be able to identify as right or wrong. If we can't agree on those, we've reached a "fire is hot" problem, the limits of Aumann's Agreement Theorem beyond which one of us must simply be irrational and not worth reasoning with. 

We should hardly rush to jump to such a conclusion. Just as every person who checks their horoscope is hardly likely to be a Flat Earther or otherwise anti-science, so most opinions are at most warning signs. But those extremes, those positions between which the truth does not lie somewhere in the middle, most assuredly do exist. Just as someone who is a professed Flat Earther is by definition wholly irrational, so someone supporting the failed coup is not someone I ever want to interact with. Their conclusion is so manifestly wrong that we cannot possibly have any meaningful interaction, so I don't propose to try.