Wednesday, September 28, 2005

Red Pill: Ethics & Rationality

Is it irrational to be selfish or evil? You might think not, as people typically separate ethics and rationality, but I think this is a mistake. From genocidal tyrants to inconsiderate neighbours, wrongdoers are not just mean and nasty, they’re also making intellectual errors. They fail to draw the conclusions and perform the actions that they rationally should.

This might sound surprising. It’s widely assumed that the only form of rationality is instrumental rationality – that is, taking effective means to achieving your desired ends, whatever they might be. If you want to be rich, and could achieve this by exploiting other people who you don’t care about, then the “rational” thing to do is exploit them – or so economists and game theorists would have us believe. They assume that there is no possible basis for assessing ultimate goals. The fact that you care more about money than people might make you mean, but it’s no failure of rationality on your part.

Although one can see why this claim might appeal to economists (ha, sorry, cheap shot), we need not accept such an impoverished conception of rationality. We can go beyond mere means-ends reasoning, and assess a set of values for internal consistency and coherence. That is, we can assess a set of ‘ends’ as being more or less rational to desire. We are not limited to merely assessing the efficiency of various ‘means’ to achieving them.

For example, we tend to consider it irrational for an agent to disregard their future interests. Suppose Larry only cares about what will happen to him during the present year, and has no concern whatsoever for how he fares after that. He might get into huge debt, seeking immediate gratification without regard for the costs he’ll suffer later. Wouldn’t you consider such behaviour to be irrational? But note that Larry exhibits no flaw in his means-ends reasoning: he’s successfully achieving precisely what he wants; the problem is that he wants the wrong things. Larry might not care about his long-term interests, but he ought to, and a more reasonable person would.

We can bring out the inconsistency in Larry’s desires by noticing that he draws arbitrary distinctions. He cares about what happens to him on New Year’s Eve of this year, but not what happens the day after. But there is no relevant difference between these two cases that would justify his taking such different attitudes towards them. If he cares about one then he rationally ought to care about the other, for they are similar in all relevant respects.

So we see that rationality requires us to treat like cases alike, and not to draw arbitrary distinctions. We can apply this to morality by examining the distinctions we draw between people that we think do or don’t “matter”. All of us think at least some people matter: at a minimum, our selves, friends, and family. But sometimes we disregard others’ interests. The most rationally coherent value set would contain a general principle explaining why we care about the welfare of the first group of people but not the others. Otherwise we are just like Larry, inconsistently caring about some cases but not others, when there is no principled basis for distinguishing between the two.

The question now arises: is there such any such principle? We might illuminate this by considering a visual metaphor. Imagine yourself at the centre of the universe (too easy!), with everybody else arranged around you based on how similar you are according to the relevant criteria, whatever that might turn out to be. So all the “like cases” are clumped around close to you, with people differing more and more as they get more distant.

Now imagine that you hold a powerful light above your head – representing your moral consideration – so that the light reaches everyone that matters to you. The light is clearly going to touch many other people too, since they will be relevantly similar to yourself or people you care about, and thus will be positioned nearby and within the light’s reach. What this shows is that you rationally ought to think that the welfare of these others matters too. You should take their interests into moral consideration.

That’s not to say that you must care about all people equally. The fact that someone is your close friend gives you reason to care more about them than a mere acquaintance. There are some relevant differences, and they will eventually add up. Returning to the ‘light’ metaphor, we can note that the light will get progressively dimmer the further away it extends. Those that are most important according to the relevant criteria will be near you in the centre, thus receiving the most light – the strongest weight in your moral consideration. This will get weaker as minor relevant differences build up into more and more significant distinctions as people get “further away” from you in our imagined ordering.

Perhaps you ought to have most concern for friends and family, a bit less for acquaintances and neighbours, and again for fellow citizens, then foreigners, etc. At each step you can identify some relevant differences, but none so significant as to justify a clear-cut distinction of saying that the first person matters and the second one doesn’t matter at all. To draw such a distinction would be arbitrary, and so leave you open to rational criticism for your inconsistency.

We’re now in a position to see why immorality is irrational. Ethics is essentially a matter of taking others’ interests into account. It is wrong to cause undue harm to another person. Is it also irrational? Given that we already take some people’s interests into account, consistency requires that we expand our sphere of moral consideration to encompass others that are relevantly similar. In order to avoid drawing arbitrary distinctions, we must recognize that other people matter too. In other words, yes, we are rationally required to be ethical. Contrary to common assumptions, the evil man is not just nasty, he’s downright irrational.


  1. Two comments:

    1. It is nice to see someone else criticize economic notions of rationality as impoverished. For the record -- and this is something that every philosopher should be shouting everyday to any economist they can find -- the word "rationality" was first used in philosophy 2300 years ago to describe thinking which could be justified with reasons.

    Only in the last half century was the word stolen by economists and applied to a specific, culture-dependent and very narrow (as you say) form of reasoning. Indeed, one of the first economists to apply the word in this way was Oscar Lange who in 1945 defined rationality as the "maximization of some quantity". Could there be a more autistic definition of the concept? Is it any wonder that the rest of us question mainstream economics?

    It is high time for philosophers (and others with a humane orientation) to take the word back from the economists!

    2. You say "rationality requires us to treat like cases alike". I don't see that this is necessarily the case. I reply with the reply of former Californian Governor (and now Mayor of Oakland) Jerry Brown when questioned on an inconsistency between two different statements he had made: "That was then, this is now."

  2. > was the word stolen by economists

    If all the externalities were internalized then being economically selfish would produce a fairly optimal result. It is jsut a matter of ensuring that.

  3. Hume gives the example of the 'sensible knave'. This is a person who is immoral but smart enough to make sure that their immoral deeds never come back to haunt them. Hence, they can act as selfishly as they please without fear of reprobation.

    Also, think of sociopaths. They might be immoral--or at least fail to fully comprehend moral distinctions--and yet be perfectly rational.

  4. "sociopaths ... perfectly rational" - you're just illustrating that it depends how you define "rational". We all know that sociopaths have a screw loose : that's the whole point of this discussion : disregard for the social consequences of one's actions is not rational. The same can be said of "rational" economists.... oh I get it... you're being ironic.

  5. You're right. While I was writing I was aware that I was, at least somewhat, begging the question. It depends on how you define rationality. To my mind, an at best prima facie definition might be "not believing, explicitly or implicitly, any contradictions." Being a sociopath does not commit you to contradictions, so the sociopath can indeed be perfectly rational, even if they are immoral.

  6. Yes, my whole point was that we need to go beyond the limited conception of rationality one finds in Hume et al.

    I think that merely avoiding contradictions is fairly clearly inadequate for full rationality. For one thing, this is a purely epistemic notion, and so leaves practical rationality completely open. Compare:

    A: I want an ice-cream, but all I've got is this useless lump of coal.
    B: I'll trade you an ice-cream for the coal?
    A: No.
    B: Why not? Do you want the coal?
    A: No, I don't want the coal, I told you, it's useless.
    B: Then you don't want the icecream?
    A: Yes I do! I very much want the icecream!
    B: So, you agree you would benefit from the trade? It's what you rationality ought to do?
    A: Yeah, I guess so.
    B: So let's trade then!
    A: No. Go away.

    Here A is clearly irrational. He's not epistemically irrational though, his beliefs (even his normative beliefs) are all in order. He just fails to act on them appropriately. We could come up with myriad other examples of practical irrationality.

    You might then respond by adopting a Humean or instrumentalist conception of practical rationality as acting so as to achieve your goals (whatever they may be). That would be a great improvement on the purely epistemic account, but my post argued that it was still inadequate because of cases like future-discounting Larry. The rest of my post then explains why amoralists (including sensible knaves and sociopaths) are open to rational criticism: they exhibit a form of loose inconsistency, for drawing arbitrary distinctions and failing to treat like cases alike.

    I think an adequate account of rationality will need to encompass these broader concerns that I've raised.

    Kofi - Presumably the point of saying something like "That was then, this is now." is to emphasize that there are relevant differences between the cases. If it really were rationally unobjectionable to arbitrarily treat like cases un-alike, then there would be no need to say such things. The fact that people appeal to such differences when challenged thus supports my position. We usually hold inconsistent treatment to be rationally suspect.

  7. Richard,

    How do you define what distinctions are arbitrary and what are justifiable?

    If you have 24 hours to live is it irrational to accept that your life will end tommorow? How about deciding that you wouldn't mind if you're life ended tommorow. You could thus disregard your future interests consistent with your planned suicide or just indifference to any other future considerations.

  8. *shrug*, I don't have an algorithm, if that's what you're asking, but I think it isn't too difficult to decide in most cases. For example, there's presumably nothing wrong with believing, and accepting, that you have 24 hours to live if this is in fact the truth (say you injested an incurable poison, or whatever).

    Suicide is a rather different case. Nevertheless, if one's life is so awful that they want it to end, there is a sense in which suicide shows a proper regard for your future welfare. If your suicide is rational, this is because your future would have been unbearable. Presumably the act of dying isn't altogether pleasant either, but, being rational, you're willing to pay this small cost in the present to avoid the enormous burdens that would otherwise befall your future self.

    So there's no "disregarding" of future interests going on here, or at least there shouldn't be. The rational suicidal person takes their future interests into account just like everybody else. If they didn't, then I say that would be irrational of them.

  9. What if I don't actually care about my friends and family at all, and want good things for them merely for the instrumental benefit it brings me? In that case there is a relevant difference between them and others, in that my friends and family really do make a large impact on my life, and therefore my own well-being, than others do. So I'm free to be perfectly amoral without making any irrational distinctions. All distinctions are based on instrumental value to myself.

    Maybe you'd object that I seem to treat my friends and family better than one would expect from someone who merely values them instrumentally? I guess it's hard to calculate just what one would expect - some economists might say that acting merely for one's own self-interest really does result in a lot of seeming altruism.

  10. Hi Kenny!

    My argument here is merely directed at those people who already have some intrinsic concern for at least some other people (e.g. friends and family). I'm not sure that any pure egoists really exist. But, supposing one did, how would I respond?

    I'd clearly have to go beyond my present argument, as you rightly note. In fact what I'd do is appeal to the arguments I make here, to the effect that some of the most valuable goods in life depend upon a non-calculating reciprocity, so that prudence itself recommends that the egoist change his character and develop an intrinsic concern for others.

  11. I'm not sure "matters to me" isn't a consistent principle. We can point out to Larry that he ought to worry about Jan 1st; and that it's likely his views will change come midnight. And this has force because we expect he'll be around and dealing with much the same things then; if he happens to be aware of an impending asteroid, however, it's simply not arbitrary, and we see the rationality.
    It's not quite the same with people. Someone who only cares for themself can point to a non-arbitrary boundary we do understand- self/others. The same could be said for a boundary between family/non-family; tribe/outsiders, etc etc. Some of these boundaries will seem more or less arbitrary; but others might be quite explanable- at least in non-moral terms- and perhaps defensible.
    The same might be said for a "principle" dividing "people I care for " from "people I don't care for." I don't think we can assume this is arbitrary, any more than we can assume Larry's love of Dostoyevsky and distaste for Tolstoy are arbitrary.
    From the other side, what could be more arbitrary than "people within the ambit of Larry's shining light"? Perhaps his batteries are dim; perhaps he's using too small a bulb; what happens in a power black-out?

  12. There are certainly some relevant differences between, e.g., family and non-family. This can justify us showing more concern for family than others. But I think this only justifies a difference of degree, and not of kind. There is no clear-cut distinction to be made here, such that your brother matters absolutely whereas your neighbours matter not at all. That seems much too strong.

    As for "people I care for", you take this group to be determined prior to ethical evaluation, whereas I think it comes afterwards. Caring for someone, in my sense, is to desire their well-being, i.e. to judge that their interests matter. The question is, what relevant differences justify you caring for one person but not another? (Note that I'm not assuming this is arbitrary, it all depends on whether there are any relevant differences in fact to justify your difference in attitudes.)

    I take it your closing questions were tongue-in-cheek. The light's "ambit" is merely metaphorical, of course. What it stands for is treating like cases alike, which is the very opposite of arbitrariness.

  13. Hi Richard-
    Doesn't there have to be a distinction between caring about someone's well-being and an ethical evaluation? I care deeply about my own and my family's well-being, but I don't believe our interests matter morally more than those of, say, Joe Hutchinson of Beckenham, whom I've never met, and of whose existance I've just learnt.
    Putting this another way, to ask us to care for everyone equally is unrealistic and somehow debases the intense caring we do for those we specifically love; conversely, basing morality or ethics on who we happen (arbitrarily or not) to care for, would make ethics or morality hopelessly relative. (This is certainly what you're grappling with- and it's a big question, to be sure).
    On the other hand, I absolutely agree that empathy does seem to be the root of most of our moral feelings and intuitions.
    (My own feeling is that it's a deep part of being human. The truely immoral person lacks or denies empathy completely, and I'd argue (contentiously to be sure!) that they a/ fail in some way to be true to their human nature and b/ miss most of what is richest and most rewarding in life. Is that irrational? It might be. Or they could have been acting rationally on bad information- by listening to economists, no doubt! :-)

  14. (I used to live in Beckenham! Knew of some Hutchinsons too. Don't recall ever meeting Joe, though.)

    I agree with your latter points -- indeed, that was more or less my response to Kenny, above.

    "Doesn't there have to be a distinction between caring about someone's well-being and an ethical evaluation?"

    Yeah, good point. I was using 'care' in a sense which just meant the latter notion, but perhaps that's misleading. The word normally refers to a more emotional connection, and that does seem rather different from what I'm talking about here.

    That does put me in something of a dilemma. On the one hand, it does seem that emotional connections could provide a principled basis for differential treatment. But then I'd have to argue that rationality requires us to have emotional connections with a broader circle of people, and that seems a rather tall ask.

    Perhaps I should tackle the first horn, and deny that contingent emotional connections could appropriately play too large a role in our ethical evaluations. I'm not too sure how to do that though. Any suggestions?

  15. One of the first moral intuitions children express is that of fairness (all too often in the negative "that's not fair!"). A lot of the universality of moral judgements seems to flow from this.
    Some of it comes from empathy- and imagination- being able to think "there but for the hand of fate go I".
    Some of it simply has to do with applying rules equally; there's a sort of "game theory" rationality to denying anyone special status with regard to the rules if we accept that we are as likely to be disadvantaged as anyone else- but I don't think that gets far- sorry- and I must get back to "work" work.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.