Modern Moral Philosophy: Lecture Four (week 16) Sensibility Theory
1. Introduction
(a) previously we have discussed IMR, ethical naturalism, error theory, and expressivism. (Phew!) Now it is time for a very influential type of moral realism, but one which proves to be quite slippery.
2. Mapping the Territory
(b) Perhaps you don¹t like any of the positions we have so far discussed. IMR seems to posit queer properties, ethical naturalism iswell fishy (more on this below). Error theory makes a good point, but there is a worry that it dismisses all sorts of realism with only an argument against IMR (aha!, keep that thought) and you don¹t like expressivism because you like Jackson and Pettit¹s argument, or you have done some reading about the Frege-Geach problem, or you still want to hold to the idea of there being an ethical reality and that our judgements are really truth-apt. Or perhaps you like expressivism, and need convincing otherwise.
3. Prelude
(c) You will have encountered this whole section last year in PL303 Ethics (I hope):
We saw in week 14 a particular definition of objective¹ based on mind-independence:
O1: Something (a property, say) is objective if its existence is independent of what anyone does or could believe, desire, etc. about it.
We saw that combining this with the idea or prescription might render the supposed moral properties queer. But there are other meanings of objective¹. For example:
O2: Something (a certain subject matter such as ethics or aesthetics, say) is objective if it admits of correct and incorrect answers, if statements of the type of subject matter in question can be true or false.
(d) A subject matter (ethics, say) can be objective in both senses. For example, if the type of moral realism from last week (IMR) is correct, then we have mind-independent moral properties, and judgements are correct or incorrect in so far as they represent such properties correctly (let¹s say).
But, these two senses of objective¹ are different. It is arguable that a position can admit of O2 but not O1. It¹s all to do with how we (metaphysically) conceive the properties that make judgements true. To illustrate, consider these two definitions of subjective¹:
S1: Something (a property, say) is subjective if its existence is somehow dependent on what humans, or best humans (or etc.) do or could believe, desire, etc.
S2: Something (a certain subject matter such as ethics or aesthetics, say) is subjective if it does not admit of correct and incorrect answers.
Example of S2: ice-cream. (Well, possibly. What about dirt ice-cream?) Notice, this isn¹t a matter of reporting or describing the tastes that people have. It is a fact that I prefer chocolate to strawberry, and an answer to the question Which ice-cream flavour does Simon prefer?¹ admits of correct and incorrect answers. If the example is an apt one (and it might not be), we are thinking of questions such as Which is the best ice-cream flavour?¹. So it is clear that S2 stands opposed to O2.
In contrast with S2, S1 does admit of correct and incorrect answers. The definition says that there are properties (ethical properties, aesthetic properties, scientific properties, properties of the best tasting ice-cream flavour¹), even though their existence is dependent on what humans (or best humans) think exists. So, we don¹t have a situation where anything goes¹. There are certain standards, and depending on the details of the theory, they could be very exacting standards about who counts as a good judge. Anyway, ignoring that, we can see that S1, unlike S2, admits of correct and incorrect answers or judgements. And, that S1 stands opposed to O1 because of how each definition conceives of the properties: mind-independence versus mind-dependence. So, and here¹s the crunch idea, notice that O2 and S1 do not stand opposed. It is arguable that a position could be both objective and subjective, in certain senses of those terms. And such a position would be realist, but not a version of IMR.
(e) Let¹s imagine what such a theory might look like in more detail. (Assume that IMR is MR1.) Here are three examples:
MR2: Ethical matters admit of correct and incorrect answers (and hence we can speak of moral properties). What determines what is morally correct and incorrect is what George Bush (or, alternatively, what Cliff Richard, or what SK) thinks.
(I know, hilarious example isn¹t it.) We¹ve got both the elements we want, O2 and S1. There really are ethical properties, and their existence is dependent on what a named individual thinks. (Notice, then, that this type of position is subjective in a third sense - S3: Matters depend on what a single named individual subject thinks.) Now, this position, as a position about what¹s ethically right and wrong is clearly implausible. A contrast with football referees and cricket umpires. So, here¹s a better example:
MR3: Ethical matters admit of correct and incorrect answers (and hence, we can speak of moral properties). What determines what is morally correct and incorrect is what a certain society, or the majority in a certain society thinks.
Hmmm. But, in the end, this might be no better that MR2. After all, we¹ve known societies (or majorities, or influential minorities) to get things morally incorrect in the past. (Think about the objective pull¹ mentioned in week 13.) What is missing here is the fact that we don¹t want to name a certain section of people as being the arbiters (or stipulators) of moral values; we want some good reason to say that so-and-so should occupy such a position. We then get two (broad) choices:
MR4: Ethical matters admit of correct and incorrect answers (and hence we can speak of moral properties). What determines what is morally correct and incorrect is what the best judges judge to be right. Best judges are characterized in morally neutral terms, but they should have access to best information, think about things coolly and rationally, etc.
MR5: Ditto..but.. Best judges are characterized in moral terms, so they need to be kind, courageous, etc.
Now there are two sorts of problem here. First, the choice between MR4 and MR5:
(i) Why choose MR5? Why think that one is a good judge of what is morally right just because one can act well? (Well, why not? Seems borne out by experience.)
Secondly, and this is the important one, choosing either MR4 or MR5:
(ii) MR5 is inherently moral, but MR4 is also normative (e.g. best information). Surely what it is to be kind, what information is best or what it is to be rational will be controversial ideas. And this takes us all the way back to the first week on IMR and all the material about what makes moral matters tainted¹.
4. Sensibility Theory
(f) It took us a whole to get there, but we now have everything in place.
(g) Popular late-70s-present (McDowell, Wiggins, McNaughton and Dancy). Sensibility¹ because it is based on what we sense about the world, and our ethical sensibilities contribute in some way to there being ethical reality. Often called simply cognitivism¹. It is a type of ethical nonnaturalism
(h) The main idea is that we want to be realists, but do not wish to be IMRealists. So, ethical properties are the products (and existing products) of human senisbilties and reasoning abilities coming into contact with the world. Ns get to be a certain E because we value them in a certain way. For those of you who are taking the Des to K module, this is all very much like Lockean secondary properties (on a certain reading), which McDowell discusses. So, just as one might characterize redness as:
x is red iff x is such as to appear red to a normal/good/best perceiver in good/etc. conditions
One might do the same for goodness. This looks awkwardly circular. But people hold out that it is not viciously so. The idea is that we specify or understand what something is by how it appears to us (or the best people) to be, precisely because we wish to link redness with (some) humans¹ experience of redness. Note the caveats regarding best perceivers¹ and the like. This is because some people are colour-blind and lighting conditions are sometimes imperfect. So, we still think that there are correct and incorrect judgements, it¹s just that what a thing is depends on (certain) human responses. Note that McDowell is partly motivated to argue for the possibility of this position because of Mackie assuming that IMR is the only possible realist position. Why make that assumption?
(i) There are problems. See above in (e). If one is tempted by the objective pull, this still looks open to some relativistic-like objection. Why assume that the truth is to be characterized by what you supposedly wise people say?¹
(j) How do sensibility theory and IMR differ? Sensible IMRealists do not discount human desires, emotions and the like (recall the slighting example from week 13). So, what¹s the difference? Think about colour again. There is a distinction between colour and ethics. It is plausible to claim that one cannot consciously change one¹s colour responses. Not so with ethical responses. Or, at least, one can hold fast to them and provide reasons why one should (one has some sort of freedom to change, if one wished). If ethical responses were purely automatic then sensibility theory would be like IMR. Of course, often our initial responses to ethical situations are automatic (hence the label initial¹). But, if we wish to we can change them, perhaps change them over a period of time. One can change one¹s personality. And, one can do it through one¹s use of reasoning. (A theory can tell some complicated psycho-philosophical story here about moral development.) So the difference between colour and ethics, for McDowell, must centre on the fact that in the latter one is free to question and change one¹s responses. (I think this is why McD is so keen to emphasize the disanology and what meriting¹ is supposed to indicate in his (1985).) Hence, there is a difference between IMR and sensibility theory: the human responses count towards something having value, they don¹t play only an (at most) epistemic role as they do according to IMR.
(k) But this creates a big question for sensibility theorists, given by noncogs. and others. (Blackburn Ruling Passions ch. 4 §§2-4 is good on this.) We have a choice between description and not. We can either say that such and such a state of affairs (including responses to it) is ethically good. This looks like naturalistic-IMR where we ask what it is which makes such a state of affairs ethically good and makes certain responses the right ones to give. Or we can say that this state of affairs is ethically good (and these responses are correct) because we choose them (and similarly for bad and disapproval). That is, we explain how the ethical gets off the ground¹. But this looks like noncog. We have a nonethical world and then our noncognitive responses to it. See last week, 4(j).
5. So why prefer Sensibility Theory to Expressivism?
(l) Aside from the problems that expressivism has, there is a further problem which some think it has and which shows sensibility theory in a good light. Note that earlier on I said that sensibility theorists think, in some (hard to define) sense, that Ns get to be Es because we value them in a certain way. But expressivists say the same thing. What is the difference? The difference is whether, essentially, one should see us as describing stuff already there (in some fashion) or expressing attitudes.
(m) This takes us to a complex, subtle argument called the disentangling objection, which incorporates the shapelessness thesis (and takes us back to Moore). In effect, the argument is simple (deceptively so) and I will present it only in a simple form. See Miller for more commentary.
(n) Can expressivists home in on the ethical? Can they distinguish ethical responses from aesthetical ones? Indeed, can they distinguish responses towards just things from responses towards kind things? (Both are hoorahed after all.) It is a working assumption of the debate that one has to be able to do this, since we do it all the time. So, here is the argument. There has to be a reason (and not just pure a whim) why we distinguish the kind things from the just things. There has to be something about the kind things that all and only all of them share. Can expressivism make sense of this? Well, on pain of giving up their position, they cannot say that the distinguishing feature is the property of kindness, so only two other sorts of feature seem available to them. First, the attitudes that one expresses towards kind things. Secondly, the natural, nonmoral features of the kind things. But, the problem with basing it on attitudes is whether our attitudes are fine-grained enough to achieve this. It is highly debatable whether they are. And who says that exactly the same type of attitude is always associated with the same sort of feature? So, secondly, what about features? This introduces the shapelessness objection. There are many, many types of kind action (many, many ways in which actions get to be kind) and many of them do not share many or any of the same kind-making feature. So, it is implausible to think that there is a nonmoral feature that kindness can be reduced to, or even a small group. (Anyone recall Moore at this point and the OQ argument?) Perhaps kindness is some large disjunctive nonmoral property? But perhaps whatever nonmoral characterization of kindness one can give, perhaps one¹s concept will always outrun it. For example, dresses and honesty. So, perhaps the thing that all and only all the kind actions have in common is something that it is best to call the property of kindness. This is our starting point. Sure, there is a lot of problems with trying to articulate what this property is, particularly if one wants to avoid IMR. But they are the right sorts of problem to be thinking about. Expressivism simply doesn¹t seem to explain in the right way our starting intuitions.
(o) There is a lot more to say here, particular about the concept of outrunning and about whether this whole objection begs the question against noncognitivism by assuming that there is an infinite number of ways in which actions get to be kind. But, that¹s enough for this week. If you want to read about this further, see Miller.