Monday, August 24, 2015

A Distant Realm: Rethinking the Procreative Asymmetry

Surprisingly many philosophers seem inclined to accept
(Procreative Reasons Asymmetry): While we have strong reasons against bringing miserable lives into existence, we have no reasons (all else being equal) to bring awesome lives into existence.

I've previously argued that considerations of demandingness suffice to explain why people are not generally obliged to procreate, in a way that leaves untouched the commonsense idea that awesome lives are amongst the best things the universe can contain, and so (all else equal) it's generally a good thing to bring about more such awesome lives.

We may now add: Since we have (some) reason to bring about good outcomes, we thus have (some) reason to bring awesome lives into existence.  So PRA is false.

To illustrate with a simple case:
(Distant Realm): Suppose you learn that a new colony of awesome, happy, flourishing people will pop into existence in some distant, causally-isolated realm, unless you pluck and eat a particular apple.  What should you do?

The apple looks pretty appealing, let's suppose, so all else equal you would be inclined to eat it. But in this case, it seems to me, you shouldn't.  Preventing those good lives from coming into existence would be a real shame, so there's a real reason not to do it -- indeed you've surely got stronger reason to refrain from doing that than you do to enjoy a mere apple. It seems to me positively wrong to choose the apple in this case.

This case then brings out that (contrary to, er, just about everyone currently working on this topic) there isn't even any fundamental deontic procreative asymmetry.  All else equal, it's wrong to prevent good lives from coming into existence, just as it's wrong to bring bad lives into existence.  In everyday circumstances, we're not obliged to procreate because not all else is equal -- it would be hugely demanding, most obviously for the gestating woman, but I think there's also some plausibility to the idea that people have a moral prerogative, not easily overridden, over their genetic material, which makes it difficult for morality to demand that they create a biological child.  But these reasons are specific to human biological procreation; they do not advert to bringing people into existence in general (which would include bringing unrelated people fully-formed into a distant existence where they won't impact upon the agent's life at all).

I conclude that the widespread belief in a fundamental procreative asymmetry is a result of people's failure to recognize these contingent (even if perhaps humanly universal) confounding factors.  Even the abstract form of the standard case introduces confounders, by contrasting a putative negative duty (do anything but this: procreate [given that the kid would turn out miserable]) with a putative positive one (do precisely this: procreate [given that the kid would turn out happy]).  This is not the way to test for a fundamental normative asymmetry between good and bad possible lives.

By contrast, when we control for confounders by considering a simple case like Distant Realm, there no longer seems to be any fundamental asymmetry. (Do your intuitions agree?)  You're not obliged to procreate, but that's because your claims and interests matter, not because the possible future person's interests don't.


  1. I think I agree. I don't work in this field but thanks to the distinctions you made, I realize that all along, I've thought that bringing about awesome new lives is indeed a good thing, and would be obligatory if it were costless to the agents with the power to do it, but is actually an act of supererogation due to the costs being quite high (for contingent reasons of biology/psychology/customs).

  2. Deontological considerations are usually the issue for people who accept no consequentialist reasoning at all, and I think that's because some people think you can't have deontological obligations unless they're held as obligations to someone, and future beings don't exist. This might be resisted if some sort of B-theory of time is true, since future beings have as much reality as present beings. But the oddest thing to me about that kind of response is that most deontologists do accept some consequentialist reasoning. They just think it doesn't trump deontological considerations. My response is that once you do that, you've got prima facie obligations, at least. And that's all consequentialists believe in anyway, since all obligations are prima facie obligations until you consider the overall effects all things considered. So this is something both consequentialists and deontologists should accept, unless they think consequences have no value whatsoever.

    1. While I agree with your general point, in this particular case I think it's really more a dispute within value theory regarding how to evaluate consequences (amongst other things) -- those who accept some kind of "person-affecting" restriction don't deny that consequences matter; they instead deny that adding awesome lives makes for a better outcome. What makes things better, on their view, is just improving the lives of people who would also have existed (only worse off) in the alternative possible outcome.

  3. I think I agree with you. Before I read Bryan Caplan's writing about childrearing I was convinced that having children was not the most enjoyable thing I could do with my life. Yet I felt like I should have one or two children anyway, because I had some kind of abstract duty to the Future of Humanity to add more awesome lives. (Since reading Caplan I have updated and realized that having children will probably be more fun than I initially believed).

    On the other hand, I also have a strong sense that people are not 100% replaceable, at least in population ethics. It seems like killing one person and conceiving a new one to replace them is a bad consequence, and that it would remain bad even after all the extenuating circumstances were removed (for instance, if a hermit died and a new hermit spontaneously generated to replace her, that would be worse than if the first hermit had lived a lot longer).

    I have a lot of difficulty reconciling these two intuitions.

    1. Three main options there, I think:

      (1) Premature death is positively bad, so replacing the foregone goods (with an all new life) is still not sufficient to replace all the value that is lost here.

      (2) Global value is partly affected by timeless average utility, which is reduced by premature death & replacement (one awesome life being better than two moderately good ones in sequence).

      (3) Some kind of moderate partiality towards guaranteed-actual lives over possible-but-needn't-exist lives.


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.