Friday, December 7, 2007

Epistemic norms are a species of moral norms

"Don't accept the testimony of unreliable witnesses." "Avoid having contradictory beliefs." "Discard beliefs all of the justification for which has been undercut." "Accept the best available explanation that is not absurd." "If you assigned a probability to a hypothesis H, and then you received evidence E, you should now assign probability P(H|E)=P(E|H)P(H)/P(E) to the hypothesis."

But why? Well, if I don't follow these injunctions, then I am less likely to achieve knowledge, comprehensiveness, understanding, true belief, etc., and more likely to be ignorant and to believe falsely. Moreover, following these injunctions will develop habits in me that are more likely in the future to lead me to gain knowledge, comprehensiveness, understanding, true belief, etc., and to avoid ignorance and false belief.

But if that is all there is to it, then epistemic injunctions run the danger of not being norms at all. Rather, they seem to be disguised conditionals like:

  1. If you accept the testimony of unreliable witnesses, you are likely to gain false beliefs.
  2. If you don't accept the best available explanation that is not absurd, you're unlikely to gain comprehensiveness in your beliefs.
While every fact is normative in some way, these will not be normative in the relevant way, in the way of imperatives, any more than
  1. If you don't raise your right arm, at most one of your arms will be raised
is normative in that way.

This is unless knowledge, comprehensiveness, understanding, true belief, etc. are worth having, unless they are good. If they are good, then they are to be pursued, and their opposites are to be avoided. But now we see that the force of epistemic norms just comes down to the fact that, as Aquinas put it, "good is to be done and pursued, and evil is to be avoided". But the pursuit of the good and the avoidance of the bad is what morality is. Hence, the imperative force of epistemic norms--that which makes them be genuinely normative--is the same as the imperative force of moral norms. Epistemic norms just are moral norms, but moral norms concerning a particular, and non-arbitrary, subset of the goods and bads, namely the epistemic goods and bads. Likewise, there is a subset of moral norms dealing with goods and bads that come up in medical practice, and we call that "bioethics", and there is a subset of moral norms dealing with goods and bads to the agent, and we call these "norms of prudence", and so on. Non-communal epistemological norms are, in fact, a subset of the norms of prudence. Any subset of the goods and bads defines a subset of morality.

One might object that only some goods and bads fall in the purview of morality. Thus, while good is to be pursued and evil avoided, only in the case of the moral goods is this a moral injunction. But I find quite implausible the idea of identifying specifically "moral" goods. I will argue against the distinction between epistemic and moral goods in two parts. The first part of the argument will establish, among other things, that epistemic norms are a species of prudential norms. The second part will argue that prudential norms are a species of moral norms.

To help someone learn something--i.e., to help her gain certain instances of epistemic goods--for the sake of her learning is to benefit her, and can be just as much an instance of kindness or charity as relieving her pain. (Of course, not every instance of teaching is kind or charitable, just as not every relieving of pain is kind or charitable--the parallel continues to hold.) To distinguish helping others attain epistemic goods from helping others attain non-epistemic goods and to say that only the latter is moral, is to take an unacceptably narrow view of morality--indeed, I think the only major moral view that makes such a claim is hedonistic utilitarian, and its making this claim is a count against it. But if there is no difference in regard to whether we are acting in accordance with morality whether we help others achieve epistemic or non-epistemic goods, why should there be a difference in our own case? The epistemic goods in our own case are not different in kind from the epistemic goods in the case of others. If pursuit of the human good of others involves helping them achieve epistemic goods, so too the pursuit of the human good of ourselves involves helping ourselves achieve epistemic goods. But pursuit of our own human good is what prudence calls us to. Hence, epistemic norms are a species of moral norms. It is no less a part of prudence to strive for true belief than it is to surround oneself with beauty or to keep one's body healthy; it is just as much a duty of prudence to keep from false beleif as it is to avoid promoting ugliness in one's environment and disease in one's body.

Now, one might say that there is a defensible distinction between the agent's goods and the goods of others, and it is only the pursuit of the goods of others that morality is concerned with. But this is mistaken. It is an essential part in learning to be moral to realize that I am (in relevant respects) no different from anybody else, that I shouldn't make an exception for myself, that I am one of many, that if others are cut, they bleed just as I do. Utilitarianism and Kantianism recognizes this. Aquinas recognizes this in respect of charity (he thinks we owe more charity to ourselves, because we owe more to those who are closer to us, but there is no difference in the kind of duty; in charity we love people because the God whom we love loves them, and so we love ourselves in charity for the same reason that we love others in charity). And a theistic ethics that grounds our duties to people in their being in the image of God, or in God's loving them, will just as much yield duties in regard to one's own goods as duties in regard to the goods of others, since the agent is in the image of God and loved by God just as others are. And if we have duties to our friends, and our friends are in morally relevant respects "other selves", then we likewise have duties to ourselves (Aristotle would certainly endorse this). It is true that some social-contract accounts of morality do not recognize this, but so much the worse for them.

Prudential norms and prudential virtues, then, are a species of moral norms and moral virtues. And epistemic norms and epistemic virtues are a species of prudential norms and prudential virtues.


Mike said...

Alex, you write,

This is unless knowledge, comprehensiveness, understanding, true belief, etc. are worth having, unless they are good. If they are good, then they are to be pursued, and their opposites are to be avoided

But they aren't in general morally good. In fact, it many cases it is better to know less or have a false belief rather than a true one. Plantinga exploits this fact in his argument that true belief does not in general have more survival value than false belief. But you might consider acting in a way that promotes one's survival non-moral action. There are lots of other cases. For instance (falsely) believing that your chances of surviving surgery are good can improve your chances of living through it. Falsely believing that you're a better than average baseball player can make you much more helpful to other players on your team. And so on.

Alexander R Pruss said...

Knowledge, comprehensiveness, understanding and true belief are intrinsically good, though there may extrinsic bads that result from them. (Xenophon puts in Socrates' mouth a clever example showing that wisdom might be (extrinsically) bad to have. For the wise man might be kidnapped by a tyrant to act as the tyrant's advisor.)

If they're not worth having either intrinsically or extrinsically, then epistemic norms have no force at all. But surely epistemic norms do have force.

I don't know what it means to say something other than an action is "morally good".

Mike said...
This comment has been removed by the author.
Mike said...

But no case has been made here for the claim that these are intrinsic goods. And I'm a little suspicious of how intrinsic goods appear just when they're needed.

Anyway, what are you saying here?

I don't know what it means to say something other than an action is "morally good".

Are you saying you don't know what it means to say that an action is, say, morally bad? You can't mean that. Do you mean that you don't know what other sorts of goods there are apart from moral goods? But then, would you say that that if X is a good way to torture someone, that means X is a moral good?
Or maybe you're saying that only actions can be morally good. In that case you would not know what it means to say that Jones's quick recovery is good. But that an event is good is perfectly intelligible. Or, so it seems to me. So, I'm not sure what you're saying there.

chauncey said...

Hi Alex,

Are epistemic norms a species of moral norms? You argue that they are. I'm skeptical. I don't have a refutation, just a couple observations.

A simple form of your argument is something like the following:
1. Only moral norms are “categorically binding” or "genuinely authoritative".
2. Epistemic norms seem to be genuinely authoritative.
3. So, epistemic norms are a species of moral norms.

I think the brunt of your argument is born by 1. But you don't say much about that here, except for indicating that we are only really obliged to do what's really good (or what facilitates being or having the good).

I'm intrigued by the way in which epistemic and moral norms seem to be bound up together. My initial inclination is to think that they are wrapped up together by both being instances of genuine normativity. I'm more apprehensive about the priority claim, mostly because I am apprehensive about the idea that failure of epistemic duty is a failure of moral duty. (And collateral commitments, like successful discharge of epistemic responsibility is successful discharge of moral responsibility.)

I think your argument boils down to one about what makes something (e.g. a rule) genuinely normative or authoritative. In your case, it appears that the question will be why goodness is the only genuine reason to do something (or be a certain way or whatever).

A subsidiary discussion concerns what might be called the "homogeneity" of the normative. In a question: Is there only one way in which something can be genuinely authoritative?

A different issue:
How do you conceive of the claim that E norms are M norms? Is that descriptive or normative?

Alexander R Pruss said...


I think there is an additional assumption in my argument:

A. Something is worth doing or having only insofar as it is good or contributes to something good. (I shall stipulatively take the avoidance of a bad to be a good.)

This is a weaker form of what "guise of the good" theorists think: they think that not only is (A) true, but we always act in ways that we think accord with (A). (I actually think this stronger claim is true, too.)

Alexander R Pruss said...


Here's another argument. Knowledge, comprehensiveness, understanding, true belief, etc., are in fact good. That seems hard to deny. They are a part of human flourishing. So we do have moral reason to follow epistemic rules. Now the question is: Is this the only reason we have? Or is there, in addition to this, a non-moral reason, an epistemic reason? Here I'd like to use Ockham's razor: the normativity of the rules is sufficiently explained by the moral considerations, so why do we need to posit a different kind of normativity?

As for the question whether my claims are descriptive or normative, I don't actually recognize the distinction, except maybe in respect of degree. I think all facts are normative.

Alexander R Pruss said...


It seems clearly a part of human flourishing to know, comprehend, understand, truly believe, etc. Being ignorant is surely a bad. I can't really argue for this except maybe an Aristotelian framework on which these are natural tele of the human being are a part of the flourishing of the human being, and on which these are natural tele.

As for the concept of "morally good", my claim was that only actions are morally good. I was not claiming that moral goodness is the only kind of goodness I understand--that would be inconsistent with that, given the obvious fact that not only actions are good.

"A good way to torture torture" is basically "an effective way to torture". following Aquinas, I think that there are several respects to the goodness of an action: end, means, circumstantial appropriateness, likely effectiveness, etc. (my list may not be the same as Aquinas'). The action is good simpliciter if it is good in all respects. The good way to torture is good in respect of effectiveness. An instance of torture may also be good in respect of end. But it is not good simpliciter.

chauncey said...

Hi Alex,


Regarding A: yes, I agree that’s an assumption--I think I was gesturing towards that. So, what makes something good? I assume it must contribute to human flourishing. How should we understand human flourishing? Aristotle’s rational activity done well? If so, then epistemic norms seem clearly built in from the beginning. (This is manifest in your argument from E norms to prudence to M norms.) By accepting an Aristotelian conception of flourishing, then, do you beg the question about E norms being M norms? One might say: Given a sufficiently capacious conception of the moral, it’s plausible that E norms are M norms. But then it’s clear we need an argument for having a capacious conception of the moral (that there is something distinctive about moral condemnation/praise/reactions).

About normative/descriptive. Here’s a way to think about your claim. The norms of game E are really norms of game M. It is not possible to play E without also playing M; any move in E is a move in M. But these claims seem to me to be statements of rules for M and are, as such, normative claims. (They are normative in that they explicitly say what ought to be; they are distinct from putatively descriptive claims that say how things are without endorsing or repudiating them.) Something like: “We ought to treat people who fail and succeed epistemically as failing and succeeding morally”. That doesn’t mean we treat them as succeeding or failing at all of morality but rather that they are succeeding or failing at some discernible part of it.

Smaller issue. Is your claim that anything subject to E norms is thereby subject to M norms? For instance, squirrels seem like they might be subject to E norms. Do they then have a kind of ethics? If we understand ethics in Aristotle’s sense, then probably the answer is ‘Yes’, for there is probably a notion of well-functioning that applies to squirrels. One thing you can do here is restrict the claim to humans.

Alexander R Pruss said...


Good points, good questions.

Here are some scattered thoughts.

Maybe I beg the question, though so much with the notion of flourishing, as with the idea that only a good can be a reason (this isn't precise, but we don't need precision at this point). Thanks for helping me identify that assumption--I guess I took it completely for granted, and couldn't imagine someone criticizing it.

I don't know how capacious a notion of morality one needs. I kind of like the idea that the distinction is just that moral criticism can't in principle be shrugged off with "I don't care about that", while criticism that one isn't a good chess player can, but I don't think this is everything... It may also be that there is something to what Anscombe says about modern moral theories--it may be that one needs something theistic to get specifically moral criticism off the ground, a sense of sin and guilt that goes over and beyond a sense of failure.

I agree that it makes sense to talk of specifically epistemic failure, just as it makes sense to talk of specifically filial moral failure, specifically paternal moral failure, etc.

I don't know if squirrels have propositional beliefs of the sort that fall under E.