Emotivism: Philosophizing with Feelings

An attempt can be made to escape the conclusion that on emotivism, (9) “If murder is wrong, then getting your little brother to murder people is wrong.” is meaningless.

For if “murder is wrong” = “I hereby express my feeling of disapproval of murder,” then (9) means “If I express my disapproval of (a) murder, then by that very fact I express my disapproval of (b) getting your little brother to murder.”

I see two problems with this.

First, even if I express disapproval of (a), I need not do likewise for (b). Maybe I just don’t feel like expressing disapproval for (b) at this very moment. Or ever, for that matter. No reason is given that one expression, if made, must be followed by another 5 seconds later.

Second, not only that, but it seems that even if I am eager to vomit out my feelings on a random philosopher, why must one feeling follow another? Feelings by definition aren’t “logical”! It is hardly a logical mistake not to find (9), the major, to be true.

Consider the proposition, “if p, then p” or “p entails p.” If p is a proposition with a truth value, this is uncontroversial. But what if p is an act? An act of expressing disapproval? In what possible sense is it entailing itself? When and where is it entailing itself?

Thus, we have made the major somewhat meaningful at the expense of making it false.

Emotivism: “Or”

Let B = Boo!, H = Hooray!, and for the sake of simplicity, B = ~H. As before,

(9) “If murder is wrong, then getting your little brother to murder is wrong.”

Suppose we rewrite (9) as

(9′) Either ~B(murder) or B(getting your little brother to murder).

What is the meaning of (9′)? What is the function of the “or” operator? My best guess is that it proposes a choice.

(8′) B(murder).
(9′) Choose between H(murder) or B(getting your little brother to murder).
Does
(10′) B(getting your little brother to murder) follow?

Regarding (9′), I can ask, what if I like neither part of the choice? Can I reject both alternatives? To avoid this problem, we can rewrite it as “Answer: Which is the lesser of the two evils: H(murder) or B(getting your little brother to murder)?”

I’m given (8′) and (9′) as inputs at t1. I then deliberate on the choice I am required to make from t1 to t2. At t2 I make the choice. Two problems now arise.

First, I can change my mind, such that before making the choice, I booed murder, and now I no longer do. This picks up on the idea that propositions are timeless, while actions are in the here and now.

Or perhaps, B(getting your little brother to murder) is such a bad choice for whatever reason, that I am “forced” to change my mind and pick H(murder).

Second, it is not so much even a change on mind as change of heart, and there is nothing illogical about that. I felt like B(murder) and now, at t2 I just don’t anymore.

Note finally that there are many things, such as x = eating an apple, of which I neither morally approve nor disapprove. Therefore, both ~B(x) and ~H(x) can be expressed at the same time, producing another disanalogy to logic.

Emotivism: “And”

Since the meaning of the “or” operator on emotivism bears no relation to the logical “or,” and produces no happy result, perhaps the “and” operator will prove more tractable. Let’s then rewrite (9) as

(9′) H(H(murder) or B(getting your little brother to murder)) =
(9*) B(B(murder) and H(getting your little brother to murder)), according to the de Morgan’s law (assuming it applies to the specifically moral modus ponens).
Given also
(8) B(murder),
can we conclude that
(10) B(getting your little brother to murder)?

Suppose Smith somehow rejects the conclusion and says H(getting your little brother to murder). Then he will B(H(getting your little brother to murder)) or boo or disapprove of his own judgment. The emotivist concludes that Smith has failed morally if not logically.

Recall that the emotivist has imbued B and H with actual real-world meanings; we are not just pushing symbols around.

Hence, what B(H(x)) indicates is moral conflict in the heart of the person who fails to heed the moral modus ponens. That person approves of x but condemns his own approval. But B(H(x)) does not entail B(x), because in actual practice the conflict can be resolved in two ways: by approving one’s approval of x and thereby approving of x or by disapproving of x. In other words, B(H(x)) entails either B(x) or H(H(x)) = H(x).

Compare this with the logical modus ponens:

(a) p → q =
~p or q =
~(p and ~q).
(b) p.
I affirm:
(c1) ~q.
But also:
(c2) ~(~q).

If I affirm ~q, then by that very fact, I must also deny it. Which is problematic. This is because B(H(x)) is a second-level affirmation; whereas ~~q is still a first level affirmation. The denial of the logical modus ponens produces a clear contradiction at the time when the denial of the moral modus ponens does not.

Emotivism: Unholy Logic

Another point against emotivism is that while both the minor premise and the conclusion are indeed moral statements, for which an attempt can be made to rewrite them to be consistent with emotivism, the major need not be a moral proposition at all. Now

(9) “If murder is wrong, then getting your little brother to murder is wrong.”

may indeed seem to state a moral fact, namely that it is wrong to get (persuade, coerce?) another to do something wrong. But a simpler statement, such as

(9”) “If murder is wrong, then it is wrong for your little brother to murder.”

is an instance of reasoning of the type: ∀(x)[P]; pick an arbitrary x (such as little brother); then for it, P will hold. On cognitivism, this is a straightforward logical truth. As a result, just as “2 + 2 = 4″ is not a moral statement and therefore cannot reasonably be rephrased as H(2 + 2 = 4) or B(H(2 + 2 ≠ 4)) or whatever, neither can (9”) be rephrased as B(B(murder) and H(it is wrong for your little brother to murder)), etc.

In other words, it seems considerably suspicious that the emotivist rephrases both (8) and (9”) in the same way, the only difference being that (8) allegedly means a first-order approval or approval of acts, and (9”) is supposed to mean a second-order approval, namely approval of moral sensibilities themselves. At the same time, a regular garden-variety major premise that happens not to connect two moral propositions (whatever they mean for the emotivists), is not rewritten that way. Why?

Emotivism: Problem of Mind-Dependence

Consider a statement

(a) If we think that x is wrong, then x is wrong.

It well represents the essence of a mind-dependent ethical proposition. How can an emotivist back off from it? Well, according to Blackburn, Miller writes,

the higher-order attitude expressed by “It’s not the case that if we think that kicking dogs is right, then kicking dogs is right” is an attitude of approval for sensibilities which, given the belief that kicking dogs causes them pain as input, yields disapproval for kicking them as output; or disapproval of sensibilities which need some belief about our attitudes as well as the belief that kicking dogs causes them pain in order to yield disapproval of kicking dogs as output.

This is nearly incomprehensible; so, let’s go back to the definitions. To “think that x is wrong” on emotivism seems to mean to “imagine expressing disapproval of x.” Only if one actually expresses it, that is, only if one vomits his own personal feelings onto someone at some moment in time, will murder be wrong.

And this rescues the emotivist from the problem of mind- (or rather heart-) dependence how?

Color vs. Nausea

The phenomenology of color is categorical, yet the description of it is dispositional. Color, like all sensations, is first a subjective experience. But there is no “color inhering” in objects; all we have are physical properties of things like such-and-such wavelengths are absorbed and others are reflected, etc. Hence, color is best defined as the disposition to cause “this particular” subjective experience in “normal” observers under “standard” conditions. This disposition is a permanent property and persists even, say, in the dark.

Contrast this with nausea, say, upon beholding a rotting peace of meat. This, too, is a subjective experience, but I see no way to express it dispositionally. And here’s why: nausea is bodily response to disgust which itself is a spiritual response to ugliness. Finally we arrive to our subjective-real goal: ugliness is part of the rotting meat. It is still subjective, as in different people will be disgusted differently, but it is an aspect of the actual thing being seen. Ugliness is a disposition to cause “this particular” feeling (disgust and nausea) in observers with some rudimentary aesthetic sense under “standard” conditions.

Both color and ugliness then are subjective-real, but disgust and nausea are subjective-ideal.

This is in harmony with St. Thomas’ description of beauty: it is anything that when seen, pleases.

Quadriformity of Metaethical “Goods”

X is a q good if and only if (p and r):

X is a divine good ↔ (x is and is x is loved).

X is a metaphysical good (of nature) ↔ (x is and x ought to be loved).

X is a moral good (of virtue) ↔ (x ought to be and is x ought to be loved).

X is a physical good (of narrow happiness) ↔ (x ought to be and is x is loved).

Note that these cover both generation and continuance; for example, “x ought to be” can mean that x ought to come into existence or that it already exists and ought to be preserved in existence.

“Subvenience” As Regards Aristotelian Causes

Supervenience of a higher B on a lower A means that the same A entails the same B, or, alternatively, a difference in B for two things is due entirely to the difference in A in them. For example, some ethicists argue that moral propositions, such as “murder is wrong” supervene on the natural world. It’s a form of reductionism.

Let me define a relation called “subvenience” which is just like supervenience except that B is lower than A. I claim that the Aristotelian causes subvene on each other in a straightforward way.

The hierarchy is as follows: material, efficient, final, formal. Hence:

Probably the easiest subvenience to see that of the material cause on efficient. If X and Y have the same efficient cause, i.e., work or function in exactly the same way, then by that very fact they must have the same material cause, i.e., they are composed of the same stuff. Or: if any two objects are made of different matter (say, water and oil), then some of their behaviors will be different. (But not the reverse; e.g., the same metal can be used in both swords and plowshares.) How could we even say that X and Y are two different chemicals, say, if every test we ran on them produced the exact same results?

At the limit, if absolutely every behavior of X and Y were the same, then we’d have to conclude that X and Y are made of the same material.

Going a step further, if X and Y are, say, 2 identical wooden planks, then their having the same material and efficient causes does not guarantee that they will be used in the same way. For example, one plank can be made into a desk; the other, into a door. The reverse is more plausible. Or is it? Can’t a wooden desk and metal desk be equally serviceable to a person? Perhaps, and one can indeed be indifferent to what the desk he desires is made of. For example, a salesman in a furniture store can say, “We have two desks, but they are made of different stuff; one is woo…” The client interrupts: “Don’t even bother telling me; I don’t care.” By his own preference, the client demonstrates indifference.

Even if one disagrees, it seems possible for the customer to pick “at random” or based on an unrelated criterion such as which desk is shown to him first.

Again, one may buy a car which will serve him just as well whether the engine is made of steel or aluminum. For the practical purposes of an acting man, as long as, say, the gas mileage was the same, the two would engines would be completely interchangeable.

At the limit, however, if we (i.e., together as a race) could find absolutely no employment of X and Y at serving human ends wherein their efficiency or utility differed, then we wouldn’t be interested in differentiating their efficient causes.

Just as in first subvenience, we could not know if two identically working things are in fact composed of different substances, so in the second subvenience we would not care if the two identically used things worked differently. In both cases, the differences in the lower causes fade and become of no import upon the similarities of the higher causes.

Lastly, if any X and Y be designated formally the same (recall that the formal cause of X is the answer to the question “What is X?”), then their final causes are also identical, and through that, also efficient causes, and in their own turn, material causes. “What X is” includes into itself but is not limited to all the information provided by answering “What is X made of?,” “How does X work?,” and “What is X for?” Surely, we powerless to use in different ways any two things that are the same simply in every respect.

At the same time, all three answers may be the same for X and Y, yet their forms may still be different. One wooden door to the storage room could be 1 mm higher than the other yet secure the room and all that that implies equally well.

Aristotelian Subvenience: Twin Earth

This understanding allows us easily to grasp the Twin-Earth mental experiment, wherein in the actual world Earth, water is H2O, and in a possible world Twin-Earth, water, though similar to Earth water in some respects, is XYZ.

Suppose now that Earther S and Twin-Earther TS disagree whether a particular sample of a certain substance is water. What’s going on?

There are two questions here. The first concerns the formal causes of water and twin-water; it asks us whether the questions “What is water?” and “What is twin-water?” have the same answer. But we have just supposed that the material causes of water and twin-water are different. As a result, as per subvenience of matter on form, their formal causes are different, too, and if the sample is water, then it is not twin-water, and vice-versa.

The second is whether, if S uses the word “water” to refer to Earth-water, and TS uses “water” to refer to Twin-Earth-water, their disagreement is substantive or it is all semantics, i.e., they are talking at cross-purposes; and clearly, it’s the latter. S and TS are simply equivocating with respect to each other, and that’s all.

Categoricity of Moral Oughts

An example of a hypothetical proposition is: “If you feel hungry, then eat a sandwich.” The sandwich is a means to an end; and the means is only required given the end. Take away the end, and the means ceases to be imperative.

Moral propositions seem categorical rather than hypothetical. One ought not to steal “whether he likes it or not.”

To illuminate the issue, we need once again to distinguish between small and large societies.

Regarding a small society we can truly say: “It is wrong for Crusoe to kill Friday.” This law, however, is categorical vacuously, because it is contrary to Crusoe’s interests to kill Friday. There is no “Crusoe must respect Friday’s rights even if he does not feel like it,” because he will always and most pointedly feel like it, unless he’s simply horribly mistaken regarding the means to his own ends. To quote Mises again, “Every step by which an individual substitutes concerted action for isolated action results in an immediate and recognizable improvement in his conditions.” That includes the very first step of Crusoe’s deciding not to kill Friday upon meeting him.

An interesting question is whether Crusoe can kill Friday in order to eat him. Can he kill Friday and use his blood to color a flag red that he could then wave to signal passing airplanes? Wouldn’t he in so doing be using Friday according to Friday’s nature as an animal if not a rational one? Yes, but surely not to the full extent of his nature. Perhaps that is why cannibalism is taboo; for an untrained mind, thinking about this subject even purely speculatively can be dangerous in the sense of undermining his morals.