Tuesday, May 20, 2008

The Question of Conservatism: Is Value Fungible?

A common objection to utilitarianism is that it treats people as mere value-receptacles. I never thought this objection made any sense, as explained in the linked post. But G.A. Cohen does a wonderful job (in his 'A truth in conservatism') of clarifying this in terms of fungibility. He writes:
The conservative propensity is to conserve, to not destroy, and, therefore, to not replace, even (within limits) by something more valuable... Conservatism is an expensive taste, because conservatives sacrifice value in order not to sacrifice things that have value.

We all have a bias towards our actual loved ones, and so would not wish for a world in which we somehow had even better relationships with completely different people. I figure that just goes to show that love and personal attachments can distort our preferences in various (fortunate, even if not strictly rational) ways. But Cohen suggests a stronger view, elevating this bias to the status of principle: existing things of value ought to be cherished, conserved and not replaced, not in virtue of any personal connection we may have to them, but simply in virtue of their actual existence.

Hence, for example, even impartial spectators -- and not just the unfortunate teenage mother and others personally involved -- should be glad about her child's existence (once he exists; not, of course, beforehand).

Note that a Cohen-conservative may still be a maxmizing consequentialist of a sort. But what they want to maximize is the quantity of preserved or pre-existing value, rather than value tout court. The objection is thus not to consequentialism per se -- utilitarian sacrifice is unproblematic, so long as it is for the sake of a pre-existing beneficiary. The objection instead concerns the aspect of utilitarianism according to which:

the bearers of value, as opposed to the value they bear, don't count as such, but matter only because of the value that they bear, and are therefore, in a deep sense, dispensable.

This is, without question, the best objection to utilitarianism I have ever encountered. Cohen goes on to press his argument as follows:

1. A thing that has intrinsic value is worthy of being revered or cherished.
2. We do not regard something as being worthy of being revered or cherished if we have no reason to regret its destruction, as such.
3. If we care only about their value, we never have reason to regret the destruction of valuable things, as such.
4. We are right to be biased in favour of existing embodiments of value.

The third premise is analytic: utilitarians (i.e. those who care only about value) may regret the destruction of value insofar as this involves a net reduction in value; it is non-replacement, rather than destruction, that we must regret. The second premise is also very plausible. As Cohen explains:

One can say, quite properly, upon acquiring a valuable thing, "I shall value this until something better comes along", but one cannot in the same way say "I shall cherish this until something better comes along": that could happen to be a correct prediction, but it could not express a decision to cherish... So, for example, you do not cherish a Tintoretto if you happily replace it by a slightly better Picasso: you thereby treat the Tintoretto as something that has the merely instrumental value of being a vessel of aesthetic value.

So the issue really comes down to premise 1, i.e. whether all intrinsically valuable things warrant cherishing in their particularity, or whether, perhaps, only things with 'personal value', i.e. that we love or are otherwise personally attached to, call for such cherishing. 

Related posts:
- Moral Roots and Alienating Aspirations
- Rationality and Reflective Endorsement.  

Update: see also 'Cohen-Conservatism revisited'


  1. Hi Richard,

    The replaceability question is an interesting one that comes up in lots of areas. Concerning loved ones, it suggests that love is a de re attitude (R. Kraut has talked some about this, M. Bernstein too). If that's true, then a qualitative duplicate of the person one loves couldn't replace that person. Of course, such a duplicate could replace the loved one without epistemological worries. The point is rather, when asked to entertain the world in which your loved one is replaced by a qualitative duplicate, lots of people think they are worse off. I'm not sure what implications these intuitions have for utilitarianism (I'm sure it depends on the version of utilitarianism you have in mind).

  2. Richard,
    You say: "So the issue really comes down to premise 1, i.e. [a]whether all intrinsically valuable things warrant cherishing in their particularity, or [b] whether, perhaps, only things with 'personal value', i.e. that we love or are otherwise personally attached to, call for such cherishing."

    I guess I am confused about the dichotomy you are setting up.

    Taking [a] first: What do you mean by "warrant"? Do you mean: [c] Given that I cherish some intrinsically valuable thing, then I am warranted in cherishing that thing because of its particularity/ intrinsic value? Or is the idea [d] that this intrinsic value somehow calls upon me (and all other agents?) to cherish the thing (and all other intrinsically valuable things)?

    [d] seems quite implausible. A person would not be irrational for *not* cherishing some intrinsically valuable person or painting. There overwhelmingly many candidates! Surely all people will not (and perhaps should not) cherish the same intrinsically valuable things either. [c] seems much more sensible.

    Now in [b], where someone loves something, or is personally attached to it, it seems that we can appeal to more than just that love or personal attachment itself to underwrite or justify their attachment. They care about the thing *because of* its intrinsic value. We would likely do so (sometimes anyway) by appealing to [c].

    (From earlier in the post, I thought the idea might be that cherishing something somehow *creates* its intrinsic value, i.e., in cherishing it, we endow it with its intrinsic value, but I take it from what you say at the end of your post that this must not be what Cohen meant. Maybe I am missing something?)

  3. existing things of value ought to be cherished, conserved and not replaced, not in virtue of any personal connection we may have to them, but simply in virtue of their actual existence.

    Is G. A. Cohen familiar with the extensive literature on status quo bias? Is he aware that the intuitions on which he is relying have a suspect causal history?

  4. Pablo - see footnote 30: "In their interesting article on 'The Reversal Test' (Ethics, 2006), Nick Bostrom and Toby Ord scout what they call the "status quo bias". Caveats in the second paragraph or their article exclude its application against this one."

    Mike - I think Cohen's form of the objection is stronger than those others. Note, for instance, that a preference utilitarian can easily accommodate the harm of duplicate-swapping simply by allowing that people have a strong preference for their actual partner (or, perhaps, for maintaining a certain historical connection), which is thwarted by such swapping. Since making a person worse off (e.g. by thwarting our personal preferences) is a strict loss of value, utilitarianism will accommodate such concerns. The greater challenge, posed by Cohen, is that substitutions may be considered bad independently of anyone's personal concern about them.

    If Bob cherishes his Tintoretto, utilitarians will advise against destroying it, for Bob's sake. (It would make him worse off, even if we happen to think the Picasso "better".) But Cohen would further advise against destroying the Tintoretto for its own sake, because it is a thing of intrinsic value, and not just because the loss of it would cause a net harm to others who may personally care about it. Even if no-one happens to be personally attached to it, still the destruction of an intrinsically valuable item is something Cohen would see as lamentable in general -- a sentiment the utilitarian cannot fully accomodate.

    Andrew - right, [a] should be interpreted permissively, more along the lines of your [c]. The claim is that we must recognize that any intrinsically valuable thing as worthy of being cherished, but that does not oblige us to cherish it ourselves.

    Your response to [b] is spot on. I should instead have put it as follows: nothing really warrants cherishing (prior to anybody actually cherishing it, the way Cohen thinks). At best, we might say it is a happy mistake when people fall in love or otherwise become personally attached to particular embodiments of value. Personal attachments are themselves of value, most plausibly, but one that we are not strictly rational in realizing. (It is supported by state-based reasons, not object-based reasons. That is to say, it's a good state to be in, but technically unjustified, rather like a false belief that is conducive to happiness.)

  5. Thanks, Richard. Here's what Nick and Toby write in the second paragraph of their article:

    "Many ethical questions could be asked with regard to this prospect [sc., the prospect of cognitive enhancement], but we shall address only one: do we have reason to believe that the long-term consequences of human cognitive enhancement would be, on balance, good? This may not be the only morally relevant question--we leave open the possibility of deontological constraints--but it is certainly of great importance to any ethical decision making."

    As far as I can see, there is nothing in this paragraph, let alone any "caveats", that "exclude[s] its application against" Cohen's paper.

  6. Cohen probably just means that he is addressing a different "morally relevant question", since he is entirely willing to concede that violating the status quo will often be good, i.e. increase the quantity of value. Compare my earlier quote: "Conservatism is an expensive taste, because conservatives sacrifice value in order not to sacrifice things that have value."

  7. [Typo amended.]

    I am tempted to pursue this point and argue that Nick and Toby's paper can be adapted easily to deal with the claims Cohen is making, but this would actually distract attention from my original point, which transcends the specific proposal presented in that paper. The point is that we should distrust intuitions with a suspect causal history, and that cognitive psychology provides us with solid evidence that the intuitions which Cohen is relying upon have such a suspect history. It makes no difference whether these intuitions are used to support a claim stated in deontic rather than axiological language; nor does it matter whether the claim is about value bearers rather that value itself.

  8. To what extent do you think premise 1 is just analytic? I'm not sure what it might mean for a thing to have intrinsic value other than to have cherish-worthiness.

    Harder, perhaps, is how we pick out those things with intrinsic value. I'd be inclined to say that only human beings, or only living entities, can be bearers of value, and not, say, works of art, cities, colleges... but is there anything beyond sheer appeal to intuition to picking those things out?

  9. Paul - I don't think premise 1 is analytic. See my post 'Two Senses of Intrinsic Value' -- neither sense invokes 'cherishing'. So it's open to the utilitarian to claim that nothing -- including intrinsically valuable things -- warrants "cherishing" in the conservative sense. (See my response to Andrew above.)

  10. I don't find this objection as persuasive as you do. It seems subject to an iteration problem. Suppose that we switched out our loved ones for improved loved ones 1000000000000000000000000000000000000000 times, such that at the end of the process our loved ones both were producing enormous benefits to the world (basically single handedly ushering in utopia) and were also experiencing utility monster esque levels of happiness every second. It seems that Cohen's account would condemn this because it just iterated a very bad thing 1000000000000000000000000000000000000000000000000000000000000000000000000 times.
    This intuition just seems to be status quo bias in realms relating to sacred values. Similarly, one who is married might think that it would have been terrible if they hadn't married the person who they married, even if an omniscient oracle said to them that they would have married someone better.
    Finally, if we accept Parfit's reductionism about the self then our actual loved ones die and are replaced every second in some sense, so the world would have almost no value. This is implausible. Even if one holds that Parfit's reductionism is false, it seems strange that if it were true, the world would have no value.


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.