So I have been real busy preparing for next weeks ethis conference at Felician College in New Jersey. I will post the virtual presentation later this week if anyone is interested. Part of the argument that I make is that moral truths are analytic. Below are some thoughts on a different argument to that effect.
One common relativist argument starts from the observation that there is widespread disagreement on matters of morality. This is taken to be some kind of evidence that there is no real fact of the matter here. Where there are facts, like in math, we do not see this kind of disagreement. It is not as though we expect to find some tribe in some remote part of the world that thinks 2+2=5. Now, this is a bad argument, as I think is well known. The fact that there is disagreement about something does not imply that there is no fact of the matter about that thing. If anything it implies that we do not know the truth (that is, if both parties to the dispute know all the relevant facts and are rational).
But even so, it is often pointed out that there is less disagreement than there seems (James Rachels is one classic source for this kind of point) So, to take one kind of example, the Eskimo who leaves their child out on the ice to die does not think that they are commiting a murder. They think that they are performing an action that is justified. What we argue about when we disagree with the Eskimo is about whether the killing is justified or not. Both parties make a distinction between justified and unjustified killings. Both parties think that murder is wrong; the Eskimo denies that putting the child out on the ice is a murder (it is justified by the lack of resources, the length of the winter, and the belief that it is better to die quickly when you are less likely to notice than to have a long drawn out death with a lot of conscious suffering), we assert that it is a murder (it is not justified).
‘murder’ just means ‘unjustified killing’, ‘lie’ just means ‘unjustified falsehood/untruth’ and so ‘murder is wrong’ and ‘lying is wrong’ are analytic truths and mean ‘unjustified killing is unjustified’ and ‘unjustified falsehoods are unjustified’. All rational people know that these are definitional truths just like they know that bachelors are unmarried males or that brothers are male siblings. So the claim that the moral truths are analytic and that what people argue about is whether or not some action is a (say) murder or not, actually better captures what people do when they argue about morality.
Richard Brown 
Very Lockean (Essay IV.3.19-20)!
Is it just me, or does this seem really revisionist? It works for “murder,” but it doesn’t seem to work in too many other cases. If the man hiding a Jewish family in his basement tells the SS that he is not hiding anyone, I think most competent speakers would agree that he is lying, and yet most competent moral agents would agree that he is justified in doing so. This seems to be true for most terms aside from “murder,” which just looks like a special case (in English–I don’t think it holds for Russian or German, for example). And, in any case, do you want to be committed to a position of having to find two terms (one justified, the other not) for each morally contestable action?
Furthermore, it doesn’t seem like this will cover all cases of moral disagreement. For example, many people will agree that “interrogation techniques” are justified, whereas “torture” is not. On the other hand, there are plenty of people who think that torture is (in some cases) justified. Now, if you say that moral debates are about whether or not actions of a particular kind are justified, you’re covering most moral debates. But if you say that some moral debates are about whether or not certain kinds of interrogation techniques constitute murder, then you’ve broken up moral debates on the issue into two different categories. This seems excessive given that, once again, you can keep all the debates in one category by saying that all moral debates are about whether an action is justified or unjustified (or right or wrong; or permissible or impermissible; etc).
Finally, there seems to be a difference between knowing that “all bachelors are unmarried men” and knowing that “unjustified killing is unjustified.” The first is true by virtue of the meaning of words; the second is true by virtue of grammatical structure. In a similar way: everyone knows that “unoriginal poetry is unoriginal,” and debates along these lines will focus on whether a given poem is original, not whether it is a poem.
Hey Brandon,
yeah, I am very influenced by book four of the Essay…I have talked about it before…
Hi Roman,
Thanks for the comment!
“If the man hiding a Jewish family in his basement tells the SS that he is not hiding anyone, I think most competent speakers would agree that he is lying, and yet most competent moral agents would agree that he is justified in doing so.”
Yeah, on my view this person has not lied because they have told a justified untruth. This may be a bit revisionist, since the average person does not distinguish between lies and justified untruths in the way that they do with murder and justified killing…but at this point the issue becomes merely verbal. Everyone agree that what the person did is justified and so moral, what is at issue is whether we call it a lie or not. I think we have better reason to not call it a lie.
“And, in any case, do you want to be committed to a position of having to find two terms (one justified, the other not) for each morally contestable action?”
That seems right to me…what is a counter-example?
The torture example is a good one. I think that ‘torture’ means inflicting intense pain or suffering in order to coerece or for sadistic pleasure. That such an action is wrong by definition seems right to me. This is because it can be read as saying ‘unjustified infliction of pain is unjustified’. The debate then become one of what counts as justified infliction of pain. If the person knows where the bomb is a we can save a million lives if we knew where it was, would we then be justified in inflicting pain? Of course this depends on what normative theory one adopts. So, I agree with you that there will be a limited range of analytic truths that are uncontestable (these will be the ones that all normative theories agree on) and the rest will be up for grabs.
In regards to your last point. ‘all bachelors are unmarried males’ is analytically true because ‘bachelor’ means ‘unmarried male’. SO the sentence really says ‘all unmarried males are unmarried males’. This is true in virtue of its form (as you say…but I also think it depends on the world in a way), this is exactly the same in the moral case.
Richard, maybe I am just missing your point?
“Everyone agree that what the person did is justified and so moral, what is at issue is whether we call it a lie or not. I think we have better reason to not call it a lie.”
I can’t see any reason at all to not call it a lie. What we call a particular action seems to me utterly irrelevant–for moral purposes–once there is already agreement on whether or not it is justified. Which is why it seems that the important moral debates are about whether particular actions (or particular action-types) are justified (and about what normative theory to apply), not about what to call them. As I read you, you seem to want to say that there is general agreement about what sorts of action-types or actions are justified/unjustified, and thus the disagreements are really about how we should describe particular actions, but that seems to me to obscure the nature of moral debate.
Take abortion. One side insists that “terminating a fetus is murder” because the fetus is “alive” and a “person,” and the harm of killing a person outweighs considerations of the mother’s wishes and well-being. The other side may insist that a fetus is not a person, and so terminating a fetus is not murder, and therefore can be outweighed by the mother’s right to have control of her body. So the public face of the debate is a terminological one: “pro-life” vs “pro-choice,” where everyone involved agrees that “choice is good” and “murder is bad.” But I doubt that this is where the real moral debate is–the debate is moral, not terminological. References to “murder” and “choice” and “personhood” are red herrings, meant to appeal to people’s intuitions about the words in order to throw them off the underlying issues. Calling abortion “murder” is a strategy designed to derail the debate, and to avoid the more fundamental issue of whether ending the life of an entity that could develop into a human being can be justified.
Maybe I am wrong on this, but I worry more that I am missing your point, because I get the sense that you want to reduce moral debates to terminological debates, instead of letting them be properly moral debates, i.e., debates about whether or not something is justified. And it still seems to me that the latter types of debate will cover all moral debates, whereas the former (terminological) debates will cover only some (unless, of course, we revise our language so that, e.g., only unjustified falsehoods will be called lies). But what I don’t understand is why we would want to go to the trouble of revising language in order to make terminological debates easier, when instead we could stick to core issues of whether the rational norms speak in favor of or in opposition to a particular action?
To put it another way: how we describe a particular action is important. The question is: once we have a complete description of an action, do we have an analytic truth about whether or not that action is justified? Or is there further room for debate? That is, is morality entirely reducible to our descriptive language, or does it exceed that language?
Ok, sorry, let me try something else. You say: “I think that ‘torture’ means inflicting intense pain or suffering in order to coerece or for sadistic pleasure. That such an action is wrong by definition seems right to me. That isn’t quite right, because you agree that there can be reasonable moral debate about whether inflicting pain in order to coerce can, in some cases, be justified (the ticking bomb scenarios, maybe). And, you’re right (I think) in the implication that we can’t define torture without reference to purposes (i.e., just as “inflicting intense pain”, because dentists might do that, but they don’t normally torture people). So here are four types of actions: (a) inflicting (nonconsensual–we have to include this to eliminate masochism) pain in order to gain pleasure from it; (b) inflicting pain in order to coerce someone into revealing information; (c) inflicting pain in order to extract retribution for a past crime.
We can agree that all of these constitute torture. But that doesn’t really help us solve any moral dilemma. Moral agents will generally agree that (a) is always wrong; but (b) and (c) are open to debate–(b) is open to debate in particular types of cases, whereas (c) is open to debate in principle. So it seems like the real debate is about whether (b) and (c) are wrong, not about whether or not torture is wrong.
My concern about revisionism is this: we could say that instances of (b) and (c) constitute torture when they are unjustified, but something else–like “interrogation” or “punishment”–when they are justified. But–aside from the fact that this seems not to add much to the moral debate–it seems to conveniently rule out the problem of dirty hands. The problem being this: you could (on some theory) be justified in inflicting pain on someone in order to obtain information, but your action of inflicting pain would still be wrong in itself, though it would be right overall.
Similarly, though there are certainly cases of justified falsehood-telling, I think there are moral benefits to retaining the word “lying” for these cases: even if your action is justified, it is still of a type that is, overall, wrong.
Hey Roman, thanks for the very useful comments and questions! I appreciate it.
I think you may be misinterpreting what I am trying to say, rather than just missing the point as you suggest. I agree with you that what is at issue in normative debates is whether or not some action is justified or not. An action is justified or not relative to some particular normative theory. So, a utilitarian will think an action is justified if it maximizes utility, a Kantian if it passes the test of the categorical imperative, a relativist might think that the action is justified relative to some societies beliefs, but not to another’s, etc. Calling abortion murder is claiming that abortion is an unjustified killing. The other side responds by trying to show that the killing is justified. Both sides agree that murder is wrong. Pro choice-ers do not think of themselves as being pro murder…Though you are right that the other side throws the word around to inflame passions…it does have an emotive component to its meaning…
I am not claiming that once we have a complete description we have an analytic truth. I claim that what is in dispute is not whether, say, ‘stealing is wrong’ is true; that it is true follows from the meaning of ‘steal’ and ‘wrong’. When we call a certain action ‘stealing’ what we mean is that it is an unjustified property transfer. We then have a debate about whether or not the action was justified. If it was, then it is not stealing. If it wasn’t then it is stealing. So, what we are arguing about is whether or not to apply the morally perjorative term ‘stealing’ to the action…whether or not it is in fact an unjustified action or not.
As for the torture stuff…that is a good example. I take yor point that if we include (b) and (c) in the defintition of torture then ‘torture is wrong’ will have the problems that you talk about. But, I am inclined to think that this is wrong. It’s my fault…I should not have said that that is what ‘torture’ meant. I take it that Moore’s open question argument showed that we cannot define moral terms with non-moral ones (though this doesn’t say anything about what the nature of moral properties is…I did not argue for that here, but I can point you to some places where I do if you are interested…). So, if ‘torture’ is a moral term, it has got to mean something like ‘inflicting pain without justification’. Then (a), as you say, is unquestionably an instance of torture. But, as you say, there is room for debate about (b) and (c)…but I see this as debate about whether (b) and (c) count as torture or not (that is, whether they are justified or not). I think this depends on whether one is a utilitarian, or a Kantian and which of these is right.
“we could say that instances of (b) and (c) constitute torture when they are unjustified, but something else–like “interrogation” or “punishment”–when they are justified.”
Yeah, what you say is the general idea. They count as torture when they lack justification, and something else when they don’t. What it adds to the debate is that it clarifies what the debate is about. Everyone agrees that torture is wrong (even Bush thinks that!). What is at issue is whether or not our tactics count as torture or not (that is, are they justified or not, or are we justified in using them or not).
I have to say, that I don’t agree with you last comment at all. To say that something is right is to say that it should be done. To say that something is good is to recomend it to others. So if an action is ‘overall’ wrong then one should not perform that action. To say that an action is wrong is to say that it is unjustified, so calling something justified and say that it is wrong contradict each other. These concepts are all inter-related.
We both agree that the action is justified and so that the agent is doing the moral thing. Do we call what they are doing lying? I say no, you say yes. I guess that have something like the following view in mind. We always have a duty to tell the truth, and so anytime we knowingly violate that duty with intent to decieve we have lied. On some occasions, though, our duty to tell the truth can be over-ridden by a duty to, say, save a life (as in the famous man-at-door-with-gun case is supposed to show). In such a case what we do is to lie, it is a willful untruth with intent to decieve, but in this instance it is moral to lie. One reason to do things my way instead of yours is that it is hard to see what kind of a theory of motivation you can give on a view like the one that you suggest. On my view a person should do what ius right because they see that ‘what is right should be done’ as a tautology. Everyone knows that a morally wrong action should not be done. That’s just what it means to say that it is morally wrong.
You wouldn’t want to say the same thing about murder or stealing, would you? You say that the repo man steals the car? That the soldier murders enemy combatants?
Ok, I guess I am just being very dense here, because I still really don’t see what you are getting at. “In such a case what we do is to lie, it is a willful untruth with intent to decieve, but in this instance it is moral to lie.” I fully agree with this: there are cases where it might be morally justified to tell a lie; that doesn’t make the action less of a lie, but it does make it less blameworthy, and maybe even meritorious. I wouldn’t say that the repo man steals the car, since (if he is a legit repo man) presumably the car does not rightly belong to the present owner. I wouldn’t say that the soldier murders the enemy (but again, I still suspect you are being influenced by a contingent feature of this particular word in this particular language), but I would say that the soldier kills the enemy, which is also pretty damn bad and exactly the reason that wars are bad things. And I just don’t think that “stealing is wrong” is an analytic truth–it is a moral heuristic that we use, and that generally serves us well, but that isn’t always right. Here’s an example (sorry, I’ve been watching Doctor Who): Aliens are planning to destroy the Earth. For that purpose, they have brought a special talisman that will activate their planet-eating machine. The Doctor, being the nice guy that he is, sneaks into their layer and takes off with the talisman, thereby saving the Earth. Now it is pretty clear that he’s done a good thing, and looking for a word other than “stealing” to account for his action is both misleading and unnecessary, since there are lots of other ways to describe his action (at least on a coarse grained account), such as “Saving the Earth.”
“They count as torture when they lack justification, and something else when they don’t. What it adds to the debate is that it clarifies what the debate is about. Everyone agrees that torture is wrong (even Bush thinks that!). What is at issue is whether or not our tactics count as torture or not”
See, this is where I disagree. I think asking whether a particular activity counts as torture doesn’t clarify the debate; it confuses it, by appealing to our simplified moral heuristics and thus nullifying moral debate before it starts. And Bush is a perfect example of this. He doesn’t think torture is wrong (from what I can tell); he just says that torture is wrong, because saying otherwise would sound really really bad for him. His goal is not to make a moral argument; his goal is to influence public opinion. Which is why he will say that “torture is wrong” but “waterboarding is not torture.” That is: he relies on the strategy you seem to be proposing in order to evade moral debate, not to clarify it. And this is why the people we should listen to on the issue are not the politicians, but the CIA generals who will get up in front of Congress and say something like this: “Yes, waterboarding is torture. That is precisely why we need to be allowed to do it in extreme cases. Something that doesn’t constitute torture–like telling terrorists that their mothers hate them–isn’t going to get the job done sometimes.” And this is where real moral debate, rather than just pointing to the strongest intuition, starts.
“it is hard to see what kind of a theory of motivation you can give on a view like the one that you suggest. On my view a person should do what ius right because they see that ‘what is right should be done’ as a tautology. Everyone knows that a morally wrong action should not be done. That’s just what it means to say that it is morally wrong.”
I don’t really see why your approach would fare any better or worse than mine on the issue of motivation, to which this discussion seems largely irrelevant. “X is wrong, therefore I shouldn’t do it” may be an analytic claim, but “X is wrong, therefore I am motivated not to do it” certainly isn’t. (“X is wrong, therefore I should be motivated not to do it,” on the other hand, might be.) Having a motivational state is not just a case of accepting the truth of a proposition.
“An action is justified or not relative to some particular normative theory. So, a utilitarian will think an action is justified if it maximizes utility, a Kantian if it passes the test of the categorical imperative, a relativist might think that the action is justified relative to some societies beliefs, but not to another’s, etc.”
To be honest, I find this dubious. While some actions will clearly be right or wrong relative to a theory, I am skeptical about the enterprise of developing a moral theory coarse grained enough to count as a theory (rather than an infinite set of particular judgments) yet fine grained enough to account for every action. And the fact that non-theoretical judgments about particular cases remain the major means by which we test theories seems to suggest that, really, particular normative theories are justified or not relative to whether or not they account for why particular actions are or are not justified.
“To say that something is good is to recomend it to others. So if an action is ‘overall’ wrong then one should not perform that action.” I’m not sure we disagree here. I didn’t say that the action is overall wrong, but the type of action. As in: killing people is overall wrong. There might be circumstances under which killing someone is, all things considered, the best course of action. That doesn’t mean that killing stops being wrong in those cases; it just means that those cases feature demands that override the overall wrongness of killing.
In any case, I do get the sense that I am misinterpreting you. And for some reason the follow-up discussion isn’t making it clearer at all. This would be a great time for a third-party, who can see both of our points, to jump in and explain what is going on, wouldn’t it? And sorry for going on so long. I was trying to be thorough, but I have a feeling it isn’t helping. Maybe it’s me.
I think, while many people might see stealing as “unjustified taking”, if
1) “what is justified” is a point of dispute
and
2) we are aiming for understanding
then one should try to use words one knows the other side will understand – i.e ones that are not implicitly by definition good or bad (question begging).
For example when Hillary said she ‘misspoke’ as opposed to ‘intentionally lieing’ all it seems to do is move the focus from ‘did she intentionally lie’ to ‘what does mispoke mean?’ the latter seems to be just a tool of persuasion via confusion.
I think roman suggests in an earlier post talks about inflicting pain as wrong in itself but potentially right overall. I’d read this as “inflicting pain is bad, all things being equal, but not in this case”.
Roman, thanks again for the detailed comment! I am going to have to think about it a bit, and I’ll get back to you…Relying on third paties is not very reliable, so I may try to sort out what is going on here myself…
Ok. I’ll try the same. Love the blog, especially the anti-zombyism.
USING WORDS WITH NON-STANDARD, CONFUSING DEFINITIONS
To lie is to affirm a statement that one believes to be false. Telling a lie is not the same thing as stating a falsehood (justified or unjustified). For example, suppose that I believe that I accidentally burned down my neighbor’s house. Now, as it turns out, I didn’t. Instead his house burned down because a fire was started by an electrical problem. If my neighbor asks me whether I accidentally burned down his house and I reply, “No, I didn’t” then I have told a lie, because I affirmed a statement which I believe to be false. The fact that what I said really is true is surprising, but I am lying nevertheless. Lying is a matter of how what you says stacks up against what you believe (or disbelieve). (Lying is inherently epistemic.)
To claim that ´lie´just means ‘unjustified falsehood/untruth´strikes me as both confused and confusing (R. Brown post #0); especially since it’s commonplace to think that it is morally justifiable to tell lies under certain conditions. Hence, I agree with Roman (post #4): “I can’t see any reason at all not to call it a lie”.
ONTIC / EPISTEMIC CONFUSION
“If the person knows where the bomb is we can save a million lives if we knew where it was, would we then be justified in inflicting pain? Of course this depends on what normative theory one adopts.” (R. Brown post #3)
“An action is justified or not relative to some particular normative theory” (R. Brown post #6)
One’s belief (or, far better, non-cognitive attitude) concerning the proposal of using torture to gain the desired information might be a function of what normative theory one adopts, but if there really are moral truths and moral facts, then they are ontic, not epistemic. Facts makes theories correct or mistaken, not the other way around. Also, it’s extremely far-fetched to think that most people have adopted some particular moral theory and act in accordance with it. Virtually everyone just “wings it”.
NON-COGNITIVIST CONCERNS
As far as I can figure it, there are no moral truths at all; and hence no analytic moral truths. Morality is about what to do, not what is true. It’s about how to act, not what are facts. (Notice the catchy rhyming slogans.)
In any case, Roman certainly has a good point when he says that the substantive moral problems (or moral disputes) involve the issue of whether certain acts, policies or proposals are good or bad, justified or unjustified, permissible or impermissible. And no so-called analytic moral truth is going to be of any help in resolving such disputes. (This is related to Roman’s final paragraph of post #4).
As a non-cognitivist I’m curious about why you think that it’s TRUE that torturing people for sexual gratification is wrong. What sort of truthmaker do you think that this claim has? Don’t get me wrong, I too disapprove of the policy of torturing people for sexual gratification; and I SAY that’s it wrong as a means of influencing people’s behavior, by letting them know that I will express very strong disapproval (and will even advocate that they be legally punished) if they are convicted of performing such acts. Nevertheless, I don’t believe that moral claims express either truths or falsehoods.
If there are moral facts, then what are their components and how are the components related to each other?
SHIT!
I just typed out a long response to Richard and Roman and I accidently erased it! DAMN IT!!!! I will have to give it another try later, I don’t have the stamina to redo it right now…
[…] Responding to Roman In the comments to my post Some Moral Truths are Analytic I got some very useful and detailed comments. Sadly, I was caught-up in the zombie wars and did not […]
[…] of lying. Everyone recognizes a duty to tell the truth and that lying is wrong (indeed, as I argue ‘lying is wrong’ is analytic) but we think there are some circumstances where one can be excused from this duty and so can tell […]
[…] Some Moral Truths are Analytic […]