Friday, August 12, 2005

Thomas Reid's Argument against Moral Non-Cognitivism

In Thomas Reid's Essays on the Active Powers of Man, he presents a compelling argument against Hume's moral theory. As I understand Reid's criticism of Hume, this argument may have some relevance on contemporary ethics regarding whether moral claims are cognitive or non-cognitive (if unfamiliar with this issue read this). Reid argues that moral claims must have a cognitive basis. Reid's argument is essential linguistic. He compares the following two claims:
(1) “That man did well and worthily, his conduct is highly approvable”

(2) “The man’s conduct gave me a very agreeable feeling.”
If moral non-cognitivism is true, then these two statements should mean the same thing. But Reid points out that these two statements do not mean the same thing. Reid explains, “The first expresses plainly an opinion or judgment of the conduct of the man, but says nothing of the speaker. The second only testifies a fact concerning the speaker—to wit, that he had such a feeling.”

Reid extends his argument by showing that the contradictories of (1) and (2) also have different meanings, which reinforces his point that (1) and (2) have different meanings. When people hold contradictories over (1), they have a disagreement over a judgment. Whereas when people disagree concerning (2), this results in a personal affront. “[F]or, as every man must know his own feelings,” writes Reid, “to deny that a man had a feeling which affirms he had, is to charge him with falsehood.”

I believe Reid's strategy is instructive. The burden rests on non-cognitivists to find a plausible harmonization of (1) and (2), or to show that (1) is meaningless. Since I find neither of these solutions plausible, I believe Reid's criticism demonstrates a key problem with moral non-cognitivism.


At 4:00 PM, August 16, 2005, Blogger Chris said...

Isn't it question begging to assume that the two sentences actually have different meanings?

Keep in mind that just because we think sentences have different meanings doesn't mean that they actually do; to show that relies on particular views of meaning and reference. In some important sense "Hesper is bright" and "Phosphor is bright" mean the same thing, although they may not mean the same thing to a particular person. Meaning and cognitive significance are separable on many accounts.

At 9:06 PM, August 16, 2005, Blogger Johnny-Dee said...

Darn it Chris! Aren't you supposed to be on vacation from blogging? I thought I was going to get a free pass on this one...

Okay, I guess I better try to give a serious reply. You're right that it would be question begging to assume that the two sentences actually have different meanings (just as it would be question begging for the non-cognitivist to assume they have the same meaning). Reid shows, in my humble opinion, that these two sentences actually have two different meanings. He doesn't assume it. It's pretty clear that (1) is about another person and (2) is about oneself. Reid shows how (1) and (2) have different meanings also by considering their contradictories. Someone could believe (1) is true and simultaneously believe the contradictory of (2) is true (although no rational person could believe both (1) and the contradictory of (1)). The same goes for (2)--someone could believe (2) and the contradictory of (1) without any problem.

Suppose you don't take this as a knockdown argument against non-cognitivism. It surely puts "the ball in their court" to show how (1) and (2) really do mean the same thing despite the linguistic analysis that Reid has demonstrated.

At 9:43 AM, August 18, 2005, Blogger Clayton said...


I think that Reid's argument couldn't be an argument against non-cognitivism.

A cognitivist subjectivist might say that (1) and (2) have the same meaning or the same truth-conditions but it is the non-cognitivist who wants to maintain that when using moral language we do not express beliefs. One strategy for arguing for this view is to try to show that belief-stating language and moral language have importantly different properties and one way of partially carrying this out is to argue that there are no non-moral expressions that have the same meaning as moral expressions as Hare tried with his missionary/cannibals example. If (1) and (2) had the same meaning, the use of moral language would express a belief about one's own attitudes. This would establish cognitivism.

I suspect the problem here has to do with the distinction between the subject matter of what is said by an utterance and the conditions under which an utterance is made. According to the non-cog, moral language might EXPRESS certain emotions, feelings, etc., without being ABOUT such emotions, feelings, etc.

At 1:05 PM, August 18, 2005, Blogger Johnny-Dee said...

Clayton, note that (1) and (2) are de dicto claims. They are supposed to be utterances of people. So, when people say things like (1) and (2), non-cognitivists would need these utterances to mean the same thing. And, of course, for a non-cognitivist, those meanings would ammount to non-propositional expressions.

Reid's point seems to be that these utterances do not mean the same thing. And for that reason, non-cognitivism fails.

At 1:12 PM, August 18, 2005, Blogger Clayton said...


I'm not sure that the de dicto bit matters. I mean, this might just be a picky point about formulation, but typically the non-cog says that (1) is not truth-apt. On every view (2) is truth-apt. If truth-aptness is guaranteed by meaning, then the non-cog MUST say that (1) and (2) differ in meaning.

Or, in other words, how would a non-cog's view differ from a cognitivist subjectivist who thinks something like 'X is wrong' is true iff X elicits feelings of disapprobation from the speaker?

BTW, I'm happy that undetached rabbit parts is back up and running. This summer has been really slow blogwise. I think ya'll and FBC are the best group grad student blogs going.

At 3:11 PM, August 18, 2005, Blogger Clayton said...


Sorry about the double post but I think I know what is troubling me here and that is the difference between expressing an attitude and self-ascribing an attitude. One can do one without the other. If the meaning of an expression using moral vocab. is captured in terms of concepts used in ascribing an attitude, you get a cognitivist subjectivist position. If not but the force of the use of the morally loaded utterance is that of an expression of an emotion or non-cog. response, then perhaps there is an argument for non-cog. in the area.

Does that seem right? Better?

At 5:21 PM, August 19, 2005, Blogger Johnny-Dee said...

Clayton, I think I'm seeing your point more clearly now, but I still might be a bit fuzzy. (Also, I'm sure if there's a problem, it's on my side, not yours--I'm well aware of my deficiencies as a philosopher.)

At one point you ask, "how would a non-cog's view differ from a cognitivist subjectivist who thinks something like 'X is wrong' is true iff X elicits feelings of disapprobation from the speaker?" Here, I might say you've come across a weakness in my intial statement of the argument. Since I'm using anachronistic terms to characterize Reid, I would say that he's not just arguing for a cognitivist view of ethics, but also an objective, realist view of ethics. Er, at least, that's what I think is going on here.

What Reid wants to resist is the move that Hume makes where "Reason is and ought to be the slave of the passions." On this Humean view, you can have some cognitive, descriptive content about moral claims, but the moral claim itself is non-cognitive. Maybe Reid's argument, if successful, refutes this Humean view. Perhaps, I should re-think my original sweeping claims about the extent to which this argument applies.

At 3:47 PM, August 21, 2005, Blogger Clayton said...


I don't think that the argument you ascribe to Reid is terrible or anything like that, it just seemed that there were some minor points about formulation that were troublesome. Nothing that would reveal deficiencies as a philosopher.

At 1:13 PM, September 14, 2005, Anonymous Cole said...

Maybe this helps:

Someone who holds that (1) and (2) are equivalent would have to be a cognitivist. He thinks that moral judgments are belief-like mental states in the business of describing the way things are. Namely, they are in the business of describing one's own affective (non-cognitive) psychology. On this view, moral judgments are second-order beliefs about one's own affective psychology. A fortiori, then, moral judgments are beliefs--and that just is cognitivism. So, on this view, my moral judgment that murder is wrong is correct just in case I do indeed disapprove of murder. The truth-makers for the belief are found in the psychology of the believer.

A non-cognitivist, on the other hand, holds that moral judgments are not belief-like mental states. Instead, they are non-cognitive states with motivational and sentimental valence--like affections or concerns. Hence moral judgments don't describe the way the world is, they aren't descriptive representations of the way things are; instead, they are a response to how things are. So they aren't even up for evaluation as true or false, correct or incorrect. Just as it makes no sense to ask whether an itch is true or false, it makes no sense to ask whether a moral judgment is true or false. (Of course, a non-cognitivist needn't assimilate moral judgments to itches or even desires; he just needs to maintain that they don't have what it takes to qualify as belief-like, truth-apt, descriptive, informational, representational, etc. mental states)

Or maybe this helps: it's the difference between being in pain and being aware that you are in pain.

At 6:32 PM, October 18, 2005, Anonymous Anonymous said...

Very cool blog you got! I just added you to my bookmarks!

I have a great article resource you might want to check out.

At 3:03 PM, October 29, 2005, Blogger Dominicans resources said...


Free indecent blogging hosting speicalist


Post a Comment

Links to this post:

Create a Link

<< Home