Hi all,
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!
-dd
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?
Hi Julia (if I may),
This is super interesting stuff! My initial reaction is that the way your constructivism (as I’m understanding it) places affect and valuing at the center of agency makes it look very promising in a way that is relatively unique to views of this sort.
I was wondering if you would be willing to say a bit more about how such a view as yours will get the substantivity which sets it apart from a view like Street’s. So, for example, how does a substantive norm like benevolence fall out of the requirements of agency, regardless of an agent’s broader psychology?
Preston, thanks very much for the comment. You have asked *the* tough question, but here goes…I have a view of agency in which a key component is evaluation. I think that early sentimentalists, and Hume, had such a conception because they held that meta-cognitive processes were crucial to agency — and a particular sort of meta-cognitive process to moral agency. As we move through the word we are assessing things, people, and ourselves. Hume had the idea that in order to do this properly, so as to rule out personal biases and prejudices, we needed to assess from the proper perspective (the general point of view). His arguments for this were not that great, really — he argued that we needed to do this so that we could effectively communicate our assessments with each other. This seems true, but not enough. It seems to me that as we move through the world making assessment, and some of those assessments involve other people and ourselves, we are engaging in behavior that underlies our ‘critical practices.’ Those practices will vary between normative domains. In the case of morality, Hume believed that what guided these practices was a norm of sympathetic engagement with others. He did not mean that a morally good person needed to actually feel sympathy with others, rather, a morally good person recognizes that sympathy is called for and that recognition has an impact on her judgements. Now, at this point I thought I might make use of the idea of practical necessity, and draw analogies between Hume’s strategy for dealing with skepticism in the early books of the Treatise, and what he could have done in the Third book: there are some beliefs, and maybe some attitudes, that are crucial to effective practical deliberation in different contexts. At a very deep level, skepticism is a problem, just not a practical problem. The same could be said for morality — be can be skeptical regarding a robustly real basis for moral norms or reasons, yet committed to using them or being guided by them. I confess that I am not really happy with this because it strikes me as not deep enough. But I’m still thinking about it.
Another possible avenue is to make an argument that the general point of view just is the right corrective standard in virtue of the fact that keeps critical practices consistent. This would require much more argument, of course.
Hi Julia,
thanks for these interesting posts. I would like, if I may, to pick up on something you say above in your reply to Preston. Namely, the following remarks:
“In the case of morality, Hume believed that what guided these practices was a norm of sympathetic engagement with others. He did not mean that a morally good person needed to actually feel sympathy with others, rather, a morally good person recognizes that sympathy is called for and that recognition has an impact on her judgements.”
As I was reading that bit, the question I was asking was the following: what is the status, the origin, and/or the nature of the norm of sympathetic engagement on this type of Humean view? Is it that we, or the overwhelming majority of us, have a basic attitude, or calm passion, that has as its content the approval of sympathetic engagement with others?
But then you directly go on to write that “at this point I thought I might make use of the idea of practical necessity, and draw analogies between Hume’s strategy for dealing with skepticism in the early books of the Treatise, and what he could have done in the Third book: there are some beliefs, and maybe some attitudes, that are crucial to effective practical deliberation in different contexts.”
This sounds a little like a line of argument Michael Smith is pursuing in his recent work, according to which being robustly able to fully exercise both (i) sensitivity to evidence in our beliefs and (ii) instrumental rationality in our actions requires having certain “coherence-inducing desires” that are stronger than any rationality-undermining desires might be. Smith then argues (for example in “Agents and Patients”) that the content of those “coherence-inducing” desires matches up with what are widely recognized as moral duties. I was wondering if you had a similar type of argument in mind.
And, if so, then my further question would be how you think of the need (so to speak) to be “effective”, as you put it, in our practical deliberation. Is that something your account assumes that we all have a basic or innate desire for, or are you thinking in Korsgaardian terms, whereby we are all, in some respect, practically committed to thinking of ourselves as effective practical deliberators?
Again, very interesting posts!
Thanks Sven, those are all great questions. Hume himself thought that sympathy was universal, though it came in varying strengths and degrees amongst individuals. But I don’t want to go that way, at least as far as grounding the norms. I *may* rethink this a bit, since one significant worry I had for this approach was that it seemed *very* subject to the contingency worry, and I’m thinking contingency isn’t as troubling as I first thought. But, even if I decide to return to this later I would want to develop a deeper analysis and not simply appeal to the universality of sympathy. Further, sympathy is not actually universal — there may be some very extreme sorts of psychopaths who lack it, for example. So, the project that I see myself engaged in is like Michael Smith’s project in that I think that there are certain commitments we have in virtue of desires for effective agency: I talk about effective agency because I believe that someone can be an agent but lack crucial skills that undermine her effectiveness: for example, someone might lack perspective taking skills which make it very difficult to determine the sorts of things that other people desire, or find disturbing. They are unable to perform what I call ‘proxy’ evaluations, or evaluations from another’s perspective. These are crucial to successful practical deliberation. And, along Korsgaardian lines, as you suggest we are practically committed to effective agency.
Many thanks for your reply, Julia! It makes me want to ask a follow-up question, namely this one. And this is, in a way, a sharpening of one of the questions I asked above: What, I am curious to know, is the status and nature of the practical commitment to efficient agency that you have in mind?
It seems to me (and this depends on how I myself think of Hume’s view) that depending on what you say about this question, this could either take you away from a Humean view and take you into a more broadly Kantian direction, or else firmly establish your view as being of a clearly Humean kind.
The latter possibility, as I think of things, would be the case in which you’d say that we have a certain feeling or sentiment (which may be calm or violent) that makes us partial to the idea of being efficient agents. Perhaps it is part of our basic feelings or sentiments that we like to think of ourselves as effective agents. On that basis you might then develop an argument, which would remain distinctively Humean, for the requirement to development sensitivity to others’ perspectives.
If you said, in contrast, that our commitment to effective agency does not arise out of any feeling or sentiment, and that it is either (i) a basic direction of our “will” – or that (ii) it is perhaps something that thinking of ourselves as agents forces us to accept – then the type of constructivism you’re developing would seem to push in the direction of a more Kantian type of view. Kant himself says that some desires arise within a “higher faculty of desire”, or “practical reason”, and have feelings and sentiments as their effects rather than as their causes. If the desire for effective agency were of this kind, then the view would look much more Kantian than Humean in nature.
Anyway, thanks again.
Hi Sven, thanks for the follow-up — that’s really helpful to me, actually. I don’t want to be wedded to all that Hume said, but my picture is the first you mention. The reason is that I don’t think that we can actually get much substantive just out of thinking of ourselves as agents. Having said that, there is a lot about Kantian constructivism that I like, and one goal for this project is to sharpen what I see to be the difference between a Humean (broadly speaking) and a Kantian perspective. What I like about the Kantian perspective is that it promises normativity that is free of the contingency that I discussed in the main post. Now, however, I am thinking that contingency really might not be such a problem. But another related virtue of the Kantian approach — or, at least, as I envision it could be developed — is that it avoids relativity. And I just don’t see that happening with the view that I am proposing. The relativity may be at a fairly high level (what social beings are committed to, for example), but it is still there. And, I take it, the Kantian account can avoid this precisely because there is no reliance on norms that have emerged from desires that we have as social beings. I view this as a major advantage for the Kantian approach…if it were to work in other ways.
Thanks for this second reply, Julia!
Thanks, Julia!
Hi Julia,
What an interesting set of ideas. Thanks to your post I got around to re-reading Korsgaard’s stimulating paper on the general point of view and that provoked a(n at best) half baked question: Will your constructivism aim to ground norms of virtue and vice, moral right and wrong, or both?
On Korsgaard’s reading, it looks like the general point of view is normative for love/sympathy but that this because love/sympathy constitutively hones in on quality of character (virtue and vice). You might argue, along somewhat similar lines, that our need for emotional regulation pressures us to form *character* judgments that are disciplined by the general point of view.
I guess that thought leaves me with several questions:
(1) Do you plan to run your constructivist account to ground virtue and vice judgments? If so, how would this square with your independent account of virtue and vice? It seems like they might conflict and that you would be left thinking of the constructivist account as something that yields a practical justification for having false views about virtue in some cases?
(2) Do you plan to run the constructivist account to ground judgments of moral right and wrong? If so, will those be partially determined by character judgments?
Hi Brad, thanks for those really interesting questions. My initial idea was to think of it as grounding very basic norms used to justify further judgments of both virtue/vice and right/wrong — so I didn’t view the two as posing a tension. I actually view the Humean approach as supporting norms central to consequentialism — those of benevolence. Then we work up from there. I don’t view the virtues of ignorance material as being incompatible with this. This is because if those traits are really virtues, on my theory the justification is instrumental. Were you thinking of the virtues of ignorance when you raised question #1?
I was thinking more about the context-relative aspect of your view, e.g. the Mutor’s example from your book (I am commenting on an APA paper on your work and that has that aspect of your view at the front of my mind).
I guess I was thinking that, in effect, your account of virtue makes a trait’s being a virtue independent of facts about contingent human nature (which may well fix what it is actually virtuous for us to do), while the humean constructivist story probably would tell us that we should disapprove of, or despise, people whose virtue traits would be vices in us (e.g. Mutors).
If that is right then the constructivist story could help explain away, and practically justify, intuitons about the Mutors being vicious, but it would not show that those intuitions are true. And this would be an example of how the two accounts of virtue would come apart.
Thanks Brad, that’s a good point. To mitigate the relativism what I am going to try to do is make the contingency more general — with respect to social creatures, not simply human beings. But also, since the constructivist is trying to extract very basic norms, it is those that would be used in any justification — and any account of what makes a trait a virtue. Even in the Mutors case. So, I don’t think that they do come apart. However, I may rethink what I say about that case in the future.
Thanks for the response, Julia. Makes sense, and makes me more interested to see the details when they are in print!