Julia returns this week with another post on a very important topic. I hope you'll all join me in welcoming Julia back for a second round!
I’d like to thank everyone who responded to my first post – I got some helpful comments, and tips, and certainly ideas for things that I need to expand on. My second post is about another project, Humean constructivism. I became interested in this project as a result of my interest in Hume’s moral psychology and his account of virtue. This lead me to an interest in his account of agency and evaluation, which in turn led me to think about what sort of meta-ethical views could really be extracted from Hume. So, when I talk about ‘Humean’ constructivism I am not talking about what I think Hume actually believed. I think his texts support a variety of readings. Rather, I am talking about a position that is inspired by Hume’s moral philosophy and draws on certain key features of his understanding of moral psychology. Aside from this, my interest in constructivism more broadly has developed out of a real concern with the problem of normativity. Realism in the form of naturalism doesn’t seem to get at normativity, and non-naturalism, well, it just seems utterly mysterious to me. Nihilism (of the sort that Richard Joyce develops) seems really attractive in some ways, but, I would like it to be the case that nihilism is false. Nihilism is so worrisome, however, precisely because it is a view that holds our moral practices up to a rigorous, naturalistic, scrutiny. The sort of moral norm ‘exceptionalism’ non-natural realist views are committed to seems, in this light, a sort of wishful thinking. I also think expressivism has really serious problems. Basically, because I think that there are some moral claims that are true (in some robust manner), and I am very unhappy with the alternatives, I’ve been exploring constructivism.
Substantive Humean Constructivism
Hume’s brand of constructivism has generally been understood as formal, that is, as simply articulating an account of what it is to value or desire, but without making any substantive commitments about the ‘appropriate’ content of values and desires. For Hume, it is thought, that is where the story ends since Hume is non-committal about what people ought to value or desire. Sharon Street, for example, has developed a formal version of Humean constructivism. Kant, on the other hand, if one views him as a constructivist at all, is not neutral on this issue. Kantian constructivism is substantive in that it starts with an account of valuing and then extracts from that account substantive norms (such as respect for other rational beings). My project would develop the Humean picture as a substantive form of constructivism which would be committed to some substantive norms (of benevolence). This is an alternative to Street’s purely formal version of Humean constructivism. It may be that the project I have in mind is best understood as a form of constitutivism in that I believe that the norms are grounded in agency, but it is a version of agency that places at its center valuing and evaluating.
Key to distinguishing Humean moral agency from Kantian is the distinction between self-regulation of the sort endorsed by sentimentalism, and self-legislation. The Humean can accommodate a good deal of the Kantian’s intuitions about the significance of reason in moral reflection. The sentimentalist holds that self-regulation is crucial to moral agency and self-regulations of a particular sort – it involves rational reflection on one’s affective as well as one’s cognitive states and the acceptance or rejection of those states on the basis of that reflection. But it is not self-legislation since the agent does not put herself under ‘the moral law’ except in the most derivative of senses. On the Humean approach, the norms that regulate reflection are epistemic but also affective. There is enormous social pressure on agents to regulate their emotions in such a way that the emotions provide effective motivation, but fall short of being socially destructive. Effective self-regulation requires empathy and sympathy. There are varying definitions of these terms in the psychology literature, but the basic distinction I develop is between empathy as a sensitivity to the viewpoint of another (and that may involve mere emotional contagion or a more developed cognitive exercise of putting oneself in the position of someone else) and sympathy as the tendency to care about the well-being of others. Both ideas are found in Hume. On a substantive reading of Humean constructivism, it is the caring about the well-being of others that provides substance to moral norms. This is just the briefest sketch of a view that I’ve developed in some forthcoming papers that go into the importance of meta-cognitive reflection in Humean accounts of agency.
A major worry that the Humean faces is that of the contingency problem. This is a problem that plagues accounts of moral norms that appeal to human nature. This is because any particular feature of human nature arose contingently, as a product of, for example, blind evolutionary forces. This means that human nature could have been otherwise. It would seem to follow, then, that morality could have been otherwise, and this runs up against a very strong feature of moral phenomenology – the seeming necessity of moral truths. This is the problem for Humean constructivism that I’ve been thinking about most recently.
One way to go is to hold that we, at least in terms of the phenomenology, confuse metaphysical necessity or semantic necessity with psychological necessity. If I were pursuing a purely Humean project I might take this alternative more seriously than I currently do. It seems to allow for a very nice analogy between what Hume does in the first two books of the Treatise and his project in the third book. Skepticism about moral truths could be handled in a similar manner by appealing to something like natural beliefs: beliefs that cannot be abandoned in the exercise of practical reason.
However, another way to go (and is not incompatible with the first way), is to abandon the worry about contingency. Sharon Street does this in her defense of formal constructivism. I also think that there are ways of making a strong case that contingency isn’t as big a problem as it has been made out to be. For example, non-naturalists view their view to be superior precisely because there is no contingency worry, and they tend to rely heavily on drawing analogies between ethics and mathematics. But the force of those analogies can be challenged, and have been challenged. Some writers such as Alan Baker, Mark Lange, and Roy Sorensen, in writing on mathematics, have discussed the example of mathematical accidents and/or coincidences [Alan Baker, “Mathematical Accidents and the End of Explanation,” unpublished manuscript; Mark Lange, “What are Mathematical Coincidences (and Why does it Matter)?,” Mind 119 (April 2010), 307-340; Roy Sorensen, “Mathematical Coincidences,” unpublished manuscript]. For example, Baker defines a mathematical accident in the following way “A universal, true mathematical statement is accidental if it lacks a unified, non-disjunctive proof.” If a proof is purely disjunctive then it is taken to be shallow – it doesn’t provide a good explanation of the statement. Putative examples include the Goldbach Conjecture, the claim that all even numbers greater than 2 are expressible as the sum of two primes. Support for the claim is provided through “many billions of examples”. If true, it is necessarily true, but also inexplicably true. In mathematics as well, explanation seems important to having a satisfactory understanding of the truths of mathematics. This insight, it seems to me, can be exploited in meta-ethics to hold that necessity doesn’t guarantee an end to a satisfactory account: and for non-naturalism, in particular, one of the worries that I have is that appeals to necessity don’t satisfy – we still are in want of a deeper understanding. So, imagine a different sort of case, logical fatalism: there is exactly one possible world. There is no distinction between contingent truths and necessary truths. Given logical fatalism, it follows that if ‘pain is bad’ is true, it is necessarily true, yet this doesn’t settle the uneasiness or worry. We can still ask for a richer explanation of why pain is bad. So, if necessity is used to try to stop further inquiry, to make such inquiry irrelevant, it doesn’t work. It might be that there is a tendency to slip from think of something as necessary to thinking as ‘essential’ -- so if pain is necessarily bad then it is part of pain’s essence that it is bad. But Kit Fine’s work shows that this would be a mistake.
All of this makes me more likely to view the appeal to necessity as not actually being all that important, and at least one thing I can stop worrying about. Any thoughts?