Thursday, April 28, 2011

On Evolutionary Debunking Arguments (Part Three)

(Part One, Part Two)

This post is the third in a short series of posts on the following article:

As we learned in part one, an evolutionary debunking argument (EDA) is something that attempts to undermine the warrant or justification for a particular belief by pointing out its evolutionary origins. All such arguments begin with a causal premise which specifies how evolution brings about the belief in question; they follow it up with an epistemic premise arguing that evolutionary processes do not track truth; and they thereby conclude that belief is unwarranted.

We saw in part two how such arguments are sometimes employed in disputes in normative ethics. The example given came from the work of Joshua Greene and Peter Singer. Both of these authors seemed to argue that deontological intuitions could be undermined by an EDA. This fact could be marshalled in support of utilitarian principles.

In response, it was argued that Singer and Greene’s argument is difficult to sustain since it needs to show that their preferred utilitarian principles do not draw upon other debunked intuitions. In other words, we need to be given some reason for thinking that global EDAs are not possible.

In this entry we consider whether global EDAs are possible.

1. Joyce and Street
In recent times, two authors in particular have pushed the idea of a global EDA. One of them is Richard Joyce; the other is Sharon Street. (I’ve discussed Street’s work at considerable length elsewhere on this blog, should you want more detail than you’ll be getting here). Michael Ruse should probably get an honourable mention as well.

Joyce argues that all our moral judgments can trace their origin to cultural and environmental influences affecting the hominid line. If we were, say, evolved from the social insects, we would come with a completely different set of pre-packaged moral commitments.

Joyce thinks this won’t do. On semantic grounds, Joyce maintains that moral discourse is committed to a type of absolutism, i.e. our moral discourse purports to provide us with a set of reasons for action that apply to all times, places and subjective dispositions. The contingency implied by modern evolutionary theory is diametrically opposed to this kind of absolutism. Thus we are forced to embrace a form of error theory about morality. (Joyce thinks we can still be happy with pragmatic, subjective reasons for action).

Street makes similar claims, but arrives at a different implication. She thinks that moral realists (particularly of the non-natural variety) should be deeply troubled by evolutionary history.

This history implies that many of our evaluative beliefs are directly moulded by the pressures of survival and reproduction. For example, altruism towards kin can be readily explained through evolutionary game theory. Despite this, realists must still believe that somehow these beliefs line-up with abstract moral truths. But surely this is incredible? Wouldn’t it be too much to think that the selective pressures of evolution just happened to coincide with the abstract, causally inert moral truth?

Street thinks this argument provides good reason for rejecting metaethical realism and embracing some form of antirealism (constructivism in particular). This position is not nihilistic or sceptical about moral truth. It just thinks that moral truths are not mind-independent.

Note that neither Joyce nor Street quite goes “all the way” with their debunking. Joyce still thinks it is rational to act in accordance with our subjectively perceived self-interest; and Street thinks moral truth can still exist. It might be possible to go even further with the debunking and point out that all normative beliefs (including beliefs about epistemic norms) are undermined by evolution. This is, effectively, what Alvin Plantinga does in his argument against evolutionary naturalism.

2. Responding to the Global EDA
At this stage its worth identifying the potential responses to EDAs by proponents of ethical objectivism/realism. There are three of them, and they should be unsurprising to anyone familiar with epistemological debates of this sort:

  • They can say that no evaluative beliefs are affected by the argument.
  • They can say that some evaluative beliefs are affected by the argument.
  • They can say that all evaluative beliefs are affected by the argument.

The third option seems unattractive for a variety of reasons. As noted above, if the proposed scepticism leaks into other normative domains then it’s basically impossible to rationally justify anything. The first option looks equally unappealing. Someone wishing to make this response would need to argue that evolutionary processes really do track moral truth (see here for a version of that response).

The second option is probably the most attractive but it is precariously balanced. Its defender needs to show why certain beliefs are unaffected. Basically, this requires that they show how the evaluative belief they wish to protect originates in or is supported by considerations that override evolutionary history. It is this kind of position that interests Kahane since it is maintained by the likes of Singer and Greene.

Consider once more Singer’s position. He thinks that an EDA can undermine deontological intuitions but not utilitarian ones. How can he be so sanguine? Because he thinks utilitarianism is supported by rational reflection that is not the outcome of our evolutionary past.

Does this kind of response work? Here is where the role of the reflective equilibrium (RE) in normative reasoning might be important. The RE proposes a kind of test for ethical beliefs. The test is coherentist in nature. It begins with a set of moral principles, it then tests these against a range of scenarios, and then modifies these principles in accordance with what seems reasonable, usually appealing to intuition when doing so.

Such an approach to normative reasoning might be uniquely susceptible to an EDA. Why? Because the equilibrium could be based on debunked intuitions. If that is how Singer ultimately justifies his utilitarian principle then he could be in trouble.


  1. "Someone wishing to make this response would need to argue that evolutionary processes really do track moral truth (see here for a version of that response)"

    I guess you forgot to attach the link.

  2. Yes.

    Thanks for pointing that out.

  3. The most interesting EDA are in regards to epistemic norms. In a sense they are trivial in that it's relatively easy to become convinced that evolution tracks e.g. logic. But they underlie the deep presuppositional nature of our reasoning. So I still find them disturbing.

    As regards to moral norms, I still maintain a glimmer of hope that RE would converge to some game-theoretic optimum given fairly liberal assumptions. For example, it may be possible to argue that the sense of Fairness would evolve in any sufficiently intelligent social species out of ecological considerations (biological similarity between breeding individuals, co-dependence...) and game-theory, and that this would determine the RE as some form of utilitarianism. I'm not sure how correct that is, but I think it is a (moral-realist?) position that cannot be easily written off.


  4. Interestingly enough, I made a similar argument about how game theoretic considerations could be used to support a form of non-natural moral realism in my PhD thesis. I did so on the grounds that there was some similarity between mathematically defined equilibria and the mind-independent moral entities that realists typically believe in (especially if you subscribe to Platonism, which I know you do not - incidentally, if you are not a Platonist about math and logic, should you be concerned by evolutionary debunking arguments?).

    On reflection though, I think might have confused different forms of realism when I made that argument. Non-natural realists usually believe moral entities cannot be reduced to any other type of entity, and I worry that my argument might be seen to reduce the moral to the mathematical. Still, I hold out some hope that naturalistic realists might be open to that kind of argument.

    I still prefer using game theory from within a constructivist metaethics. Depending on the liberality of your assumptions (as you point out), I think you can argue for the existence of some pretty robust moral truths (ones that hold across a number of possible worlds) like Fairness.

  5. Sounds like an interesting thesis. I do hope to read a good account of this sort one day.

    My fear with game-theoretic accounts is that they're too abstract. They might show us an abstract pattern most intelligent creatures will approach, but they (at least at the level I considered them) neglect the fact that all intelligent creatures aren't at this optimum peak, they're somewhere below that. And that, of course, includes us. They are a kind of eschatology of morality, rather than morality, in my opinion - saying where morality will end up, rather than where it is.

    Still, I'd like to read a good account in this vein.

    I do think the metaethical categories can be a bit confusing and unhelpful at times. Is the above position natural realism, or non-natural... well, as long as it's understood, it doesn't really matter now does it?

    As for my non-Platonism and EDA - the concern is that the foundation of thought, logic, remains unfounded. While one does not assume that logic exists in the world, one has to assume logic in order to talk about the world. This is not the only direction of thought that leads to this conclusion, but it's still a disturbing realization.


  6. This comment has been removed by a blog administrator.