In a recent discussion of this dialectic, it was asked what would happen if we tried to construct a Moorean open question around Ayer's construal of GOOD, a la "I approve of x, but is it good?" For Ayer, the latter question does not ask whether some object instantiates a property. What is it asking? Perhaps it is asking whether the speaker responds approvingly. Of course, this is not quite right either, because the answer to the question "Do I respond approvingly?" might be yes and accordingly "I respond approvingly" would be true and we would then (on this reading) be forced to say that it is true that the object is good. Here we have fallen into the confusion between emotivism and subjectivism. The latter has it that the meaning of "x is good" is "I approve of x", and since "I approve of x" might be true, "x is good" might be true. To confuse emotivism with subjectivism is to lose the noncognitivist dimension of emotivism. But we are susceptible to this confusion because we are imagining a speaker asking "is x good?". On an emotivist reading, this is actually a very awkward question for a speaker to ask (compare, "Hooray?"). There may be a deeper problem for emotivism here, but I don't want to explore it. Instead I consider a variant upon the open question that doesn't lead us down this road: "x is good, but I don't approve of x." Instead of an open question, we have something of a paradoxical assertion. But is it an intelligible assertion?
In discussion, this new Ayerean paradox was compared to Moore's paradox. Moore's paradox is a sentence like "Our sun is only orbited by eight planets, but I don't believe that". The puzzle that Moorean paradoxical gives rise to is generated by the fact that many assertions of the form "p, but I don't believe p" would be true. Any time I have a false belief in ~p, I could truly utter "p, but I don't believe that". Anyone living in the nineties could have uttered the previous sentence about planets and have spoken truly. And yet, no one can never appropriately utter "p but I don't believe p". How can a sentence form be such that so many of its tokens are true, and yet none of them are utterable?
The answer has something to do with the commitments that a speaker takes on in uttering assertoric sentences. By (sincerely) uttering an assertoric sentence, you commit yourself to two things being the case, both that the world is the way you describe it being, and that you are in a certain mental state, one of believing that the world is that way. It is puzzling that when I am telling you about the planets that surround the sun, I should simultaneously be committing myself to facts about my mind. And working through that puzzlement (developing theories of mental content and linguistic behavior that should show how it could be the case) has been an important project over the last century.
The comparison between the Ayerean paradoxical sentence, "x is good, but I don't approve" and the Moorean paradoxical sentence, "x is true, but I don't believe that" puzzled me at first. There seemed to be something important to Moore's paradox that wasn't captured by (what I am calling) Ayer's paradox. What's striking about Moore's paradox is that both conjuncts are often true, even though asserting the conjunction is never appropriate. No such thing could be said about the Ayer's paradox. I can show this quite cheaply by pointing out that "x is good" is never true (or false) on Ayer's account, and so a conjunction in which "x is good" appears can never have two true conjuncts. But the problem is deeper than this. Consider: expressions of emotion cannot be true or false, they can be sincere or insincere. Let us call an assertion truthy if it is either (a) an expression of an emotional state and sincere, or (b) an assertion and true. If the analogy between these two paradoxical sentences works, then it ought to be the case (it seems to me) that both conjuncts could be truthy, even though no one could ever utter it.
But here the comparison breaks down! If 'x is good' expresses approval of x, then "x is good but I don't approve of it" can never have two truthy conjuncts.
There is still something to the comparison however. In both cases a speaker utters an expression which commits her to being in a certain mental state, and then denies that she are in that mental state. To express approval of x by uttering 'x is good' commits you to being such that you approve of x, and to immediately follow by saying that you don't approve of x is to deny that which you have just committed yourself to. Likewise, to assert p commits you to beings such that you believe p, and to immediately follow by saying "I don't believe that p" you deny that to which you have just committed.
And here is where I think we get a really interesting result. Consider these two sentences:
(1) x is good but I don't approve
(2) p but I don't believe p
There are (at least) two interesting facts about (2). First, it is an instance of committing yourself to being in a mental state and withdrawing that commitment. Call this its inconstancy in assertion. Second, it is a sentence form which has innumerable true tokens, but no assertable tokens. Call this its truthiness without assertability. (recall that true assertoric sentences, along with sincere emotional expressions, are truthy.)
(2) has both inconstancy in assertion and truth without assertability. What I have shown is that if you are an emotivist, then (1) has inconstancy in assertion but doesn't have truthiness without assertability.
For an interesting comparison we can consider the status of (1) on a cognitivist moral realist (henceforward, the realist) reading of 'x is good.' The realist believes that moral properties are real, mind-indepent properties which objects and actions instantiate independently of our moral beliefs about those objects. Further, she believes that moral judgments express beliefs, beliefs that the world is such that a moral property inheres in some object or action. We might now ask, for the realist, whether (1) has truthiness without assertability and whether (1) has inconstancy in assertion. Both depend upon the realist's solution to the problem of the action-guiding character of moral judgments. Robust realist positions historically have a problem explaining why believing something to be good typically brings with it a disposition or inclination to promote it or bring it about. If moral properties are mind-independent and out there in the world, then it seems as though it should be perfectly fine to attribute such properties to object and actions without feeling motivated in any direction whatever, just as my observations that objects are red, short-lived, or in the next town don't intrinsically motivate me to do anything about those objects. But moral properties seem much more intimately tied to motivation. Of course, it's not entirely clear how seriously the realist should take this worry; the realist may insist that moral judgments don't have any essential relationship to motivation, and point to psychopaths as an example. Psychopaths, so the story goes, use moral concepts quite compentently. Accordingly, it seems to be reasonable to take them at face value when they appear to be making moral judgments. And yet they don't feel motivated to approve of or promote the things they identify as good or to disapprove of or prevent the things they identify as bad. The question then arises as to whether such judgments are genuinely moral judgments...and the shape of the ensuing dialectic is pretty familiar.
Importantly, if the realist regards themselves as needing to provide an explanation for the relationship between moral judgments and motivation, according to which belief that x is good carries along with it a defeasible, ceteris paribus inclination to approve of or promote x, then they will regard (1) as having both inconstancy in assertion and truthiness without assertability. Because judging that x has the property of goodness commits me to approving of x, I will be withdrawing a commitment I just took on by saying of an object I just called 'good' that I do not approve of it. And, because judging that x has the property of goodness commits me to approving of x, I can never appropriately assert both that x is good and that I do not approve of it. However, since GOOD is a mind-independent property that things out in the world do or do not have, and since I am not infallible, it will often be the case that I can truly assert of some object both that it is good and that I do not approve of it.
If the realist regards themselves as not needing to provide any such explanation, because they do not believe that there is any essential connection between judgments that x is good and approval of x or disposition to promote x, then they not will regard (1) as possessing inconstancy in assertion. Because asserting that x is good does not commit a speaker to approving of x, they will not be violating any such commitment when they immediately follow up on their assertion with the claim that they do not approve of x. Nor will they regard (1) as possessing truthiness without assertability. To be sure, they will regard assertions of (1) as often true. Any realist who believes that humans are morally fallible will have to think that some tokens of (1) are true. But the realist of this stripe will also think that (1) is sometimes assertable. In particular, they might say, the psychopath can often assert (1) without any problems.
No comments:
Post a Comment