Evolution and Normative Scepticism

Abstract

It is increasingly common to suggest that the combination of evolutionary theory and normative realism leads inevitably to a general scepticism about our ability to reliably form normative beliefs. In what follows, I argue that this is not the case. In particular, I consider several possible arguments from evolutionary theory and normative realism to normative scepticism and explain where they go wrong. I then offer a more general diagnosis of the tendency to accept such arguments and why this tendency should be resisted.

Notes

¹This condition is somewhat vague in at least one important sense. In particular, it does not specify the ease with which we could have developed faculties of this sort. For more discussion of this issue, see the following and, in particular, §3.2.

²Of course, these explanations plus our beliefs about matters of taste will have implications for the reliability of these methods.

³Of course, there may be perceived properties it does not mention. For instance, depending on one's views about the metaphysics of colour, colour properties may not be mentioned in any such evolutionary account.

⁴Similarly, when I speak of ‘normative properties’ in what follows, this should be understood in a weak, deflationary sense.

⁵And to claim that some faculty or another for the formation of normative beliefs is reliable is, implicitly, to make certain claims about matters of a normative sort. For it is to implicitly make certain claims about which normative beliefs are likely to be true and false.

⁶To be sure, there are limits to this. For instance, the development of normative faculties that are responsive to the suffering of others and to issues relating to reciprocity does have obvious value in an evolutionary context.

Thus, although an evolutionary account will show that we could have developed normative faculties that failed to be responsive to these sorts of facts, it may not show that we could have done so easily. For more on this issue, see §3.2 below.

⁷A further alternative would be to give up the project of giving a naturalistic account of the origins of our normative faculties entirely, but for the sake of this essay I want to set this option on one side.

⁸For example, Street 2006 hopes to use a sceptical argument of this sort to force us into thinking of the normative in mind-dependent terms.

⁹It is a further question whether we might be able to reduce the normative to the non-normative via making use of our normative faculties. This sort of reduction is much more plausible than the sort of reduction I am setting aside here.

¹⁰In Street 2006. See also Joyce 2006. Similar sorts of considerations are offered, with a more modest aim in mind, by Gibbard 2003. And an argument very similar to Street's lies at the centre of Plantinga's [1993] argument that naturalism is self-defeating because the best naturalistic accounts of the origins of the mind defeat our entitlement to trust our basic belief-forming faculties.

¹¹To be clear, Street offers a number of different considerations in favour of her conclusions. In what follows, I will focus on the line of argument in Street's paper that seems to me most central to her concerns and most philosophically deep. But it is possible to see Street as raising other challenges in addition to the one I focus on here.

¹²Thanks to an anonymous referee for pressing me on this point.

¹³For example, we would surely put little faith in our scientific theories if we believed that the perceptual experiences on which they are based are worthless in the manner Street argues is true with respect to our normative faculties (realistically construed).

¹⁴In other words, as Copp 2008 puts things, even if our basic normative faculties do not track the normative truth with perfect accuracy, they must at least ‘quasi-track’ it. That is, they must get us close enough to the truth so that reasoning and reflection can get us the rest of the way there.

¹⁵The point I am making here is a quite general one, but the psychological research into the pervasive role our emotional responses play in shaping our moral and normative views only strengthens this point in the human case. For given this research, there is very little plausibility in the claim that our considered normative judgments are insensitive to evolutionary-significant factors. See, for example, Haidt 2001.

¹⁶For the same reasons, it is important to recognize that Street's challenge is not based on the sort of crude evolutionary theorizing that assumes that a faculty tracks only the properties it evolved to track. Rather, Street wants to argue that, given normative realism and evolutionary theory, we could be successful at tracking the normative truth only given a wildly implausible coincidence or fluke.

¹⁷I mean this in a broad sense that includes both internalist and externalist approaches to justification so long as these approaches are anti-sceptical in character.

¹⁸Plus, potentially, other forms of evolutionary change.

¹⁹See, for example, Rozin et al. 2000.

²⁰Of course, determining the degree to which this is the case for any given normative faculty would require much closer examination of the data available to us. But the general point I am making here does not depend on the precise degree to which this or that normative faculty passes this sort of test.

²¹Thanks to Sinan Dogramaci and Matt Kotzen for discussion of this example.

²²The same is true, in the present context, even if we weaken (ii) to read:

(ii*) We have no reason to believe that the development of our normative faculties was likely to be sensitive to whether they were reliable.

For while it is true that evolutionary theory on its own gives us no reason of this sort, in the present context it would be a mistake to move from this to the claim that we have no reason to believe that the development of our normative faculties was likely to be sensitive to whether they were reliable. After all, we are assuming that we do possess a prima facie entitlement to treat our normative faculties as reliable. And so long as evolutionary theory does not provide us with some reason to regard these faculties as unlikely to be reliable, this initial entitlement will be undefeated. In which case, it will provide us with a basis for believing that their development was (directly or indirectly) sensitive to their reliability in much the manner described above. So even when (ii) is weakened to (ii*), the move from (i) to (ii*) is one the realist should reject.

²³Thanks to an anonymous referee for pressing me on this point.

²⁴On the idea that evolutionary pressures generate convergence towards strategies that embody a reasonable level of reciprocity, see, for example, Axelrod 1984 and Joyce 2006.

²⁵To be clear, I don't want to claim here that these sorts of considerations should have absolutely no force. All that I want to claim is that they fall short of providing us with grounds to completely distrust our normative faculties in the manner that authors like Street suggest is the case, at least given normative realism.

²⁶Of course, it is not at all obvious that, say, a mind-dependent or constructivist account of the nature of normative facts would really score better with respect to this criterion than the realist's account does. For a mind-dependent account of normative facts will, of course, consist in some set of basic principles about how normative facts relate to non-normative facts about our normative attitudes. And these principles will have many, if not all, of the structural features of the realist's basic claims about how the normative supervenes on the non-normative. But in order to evaluate this issue in detail, we would need to have before us a particular anti-realist theory of this sort.

²⁷See Enoch [forthcoming] for a more extended discussion of some similar ideas. Enoch also goes on to discuss the sort of explanation of the reliability of our normative faculties a realist might propose, which is the main topic of his essay, in much greater detail than I do here.

For an especially detailed discussion of how the realist might respond to this explanatory challenge, given a particular variety of realism, see Copp 2008.

²⁸Although it is not obvious that, say, a mind-dependent anti-realist's account does any better on this score.

²⁹It might be objected here that the problem with the realist's account of these issues is not simply that it is based on the conjunction of explanatorily-unrelated claims, but also that these claims require us to accept the existence of reliable correlations between explanatorily-unrelated factors. But this seems to mischaracterize the realist's position. For example, suppose the realist's account of the normative bottoms out, in part, in the claim that pain is objectively bad.

This is a claim about how two sorts of facts are correlated with one another—namely, that facts about objective badness supervene on facts about pain. But is it a claim about how two explanatorily-unrelated factors are correlated? There is no reason for the realist to accept this. After all, the realist is perfectly entitled to say that the fact that some state involves pain explains why this state is objectively bad. Thus, in accepting these sorts of claims, the realist should not be viewed as accepting correlations between explanatorily-unrelated factors. Instead, he should be viewed as claiming that there are certain explanatory relations between the non-normative and the normative which cannot be further explained. (So if the realist can be accused of accepting correlations between explanatorily-unrelated factors, this is only true in the sense that any fundamental principle in one's theory of everything will involve such correlations.)

³⁰Compare Gibbard 2003.

³¹As Hume, for example, may have believed, at least up to a point.

³²It might be doubted whether our best explanation of the development of our perceptual faculties could ever show these faculties to be completely unreliable across the board. For any such explanation would seem to completely undermine our own warrant for believing it as well. Thus, there is a lower bound on the degree to which we might fail this test. But the fact that there is such lower bound does not make the fact that we have exceeded it completely insignificant. After all, even if a professor is too kind-hearted to ever give out a grade lower than a C, this does not rob the receipt of an A from him of all value.

³³Compare Korsgaard 1996.

³⁴This paper has benefited from discussions with many individuals, including Justin Clarke-Doane, Sinan Dogramaci, Dana Evan, David Enoch, Matt Evans, Don Garrett, Matt Kotzen, Geoffrey Lee, John Morrison, Tom Nagel, Derek Parfit, Bogdan Rabanca, Kieran Setiya, Sharon Street, David Velleman, and Tim Willenken. Special thanks are also due to several anonymous referees.