Ethical Theory Spring 2019

Cognitivist Error Theory


Both Mackie and Ayer deny that there are any objective values. As Mackie puts it, values “are not part of the fabric of the world” (Mackie 1977, 15).

Mackie disagrees with Ayer’s analysis of moral language. Ayer thinks that moral language does not even attempt to say anything about how things really are. Rather, what it does is express the speaker’s attitudes or emotions. When you say “it’s wrong to kill the innocent,” what you are saying is “boo!,” directed at killing the innocent.

Mackie thinks that moral language does seek to describe how things really are. When you say “it’s wrong to kill the innocent,” you mean to say that something is true about killing the innocent: it’s wrong. However, Mackie believes, since “being wrong” is not part of the fabric of the world, what you say is literally false. The same is true for every other statement of values.

Ayer’s subjectivism is linguistic, that is, about language, while Mackie’s is ontological, that is, about what is real or what exists.

Ayer’s theory of moral language is non-cognitive, meaning it contends that moral language does not seek to say anything true or false. Mackie, by contrast, has a cognitivist theory of moral language: he believes that moral language does try to say things that are true.

Since Mackie thinks moral propositions are always false, he has what is called an error theory. That means just what it sounds like: moral language is systematically in error. For Ayer, by contrast, we don’t make an error in using moral language. All we mean to do is express our emotions, nothing more, and so we can’t make a mistake or be wrong.

The Argument from Relativity

Mackie thinks there are two main reasons for believing that there are no objective truths about morality. He calls these “the argument from relativity” (§8, pp. 36-38) and “the argument from queerness” (§9, pp. 38-42).

The argument from relativity turns on the alleged fact that disagreements about values persist in ways that disagreements about facts do not. To put it crudely, all the cultures of the world agree on physics and math but they continue to disagree about ethics. The inference is that there is an objective reality that confronts people with diverse backgrounds in physics and math. No matter what you start believing, your airplanes aren’t going to fly unless you design them in conformity with what we know about physics and math. To put it another way, you’re free to have an alternative understanding of physics or math, but the things you try to build based on that alternative understanding aren’t going to work; you will step on the rake of objective reality and get hit in the face. By contrast, there is no similar pressure for convergence in moral beliefs. So the argument goes.

Chris correctly noted that the argument from relativity does not logically force you to accept the subjectivist answer. You can say that others disagree with you because you have a superior source of ethical knowledge that the members of other cultures lack. Perhaps God spoke directly to you (or your people), for instance, or maybe you think your society has reached a particularly enlightened stage of history that others have not yet achieved. If so, there could still be objective truths and the explanation for continued disagreement would be that some societies simply lack the ability to understand them.

If I were Mackie I would concede that this is possible. I would try to show that it is less likely to be the correct explanation than the one I gave.

Zach said that he thought the argument from relativity exaggerates the objectivity of science and underestimates cross-cultural agreement about ethics. Every society has rules about truth telling and killing, for example. Ariel added that she thought many disagreements about ethics are superficial. Jacob anticipated Brink in saying that there is a kind of convergence in ethics that is similar to what you find in science. Not all cultures have the same views of science; the claim is only that they tend to converge on similar views, driven, presumably, by objective reality. But, Jacob said, you could make a case that ethics works the same way. Not all cultures have the same moral views, but they are all converging on rules that promote human welfare and they all move away from those that harm it. But Derick reminded us that there are some quite real disagreements: in Saudi Arabia, women are treated very differently than they are in Sweden, for instance.

The argument from queerness

The argument from queerness is really two arguments. One is metaphysical (concerning the nature of reality) and the other is epistemological (concerning the nature of knowledge).

The metaphysical argument goes like this. There are no moral facts because if there were moral facts, they would be completely different from any other facts. Why would they be different? Well, if there were moral facts, they would motivate behavior. Someone who believed that lying was wrong would be motivated not to lie, for instance. But, generally speaking, there is a gap between believing a fact and being motivated by that belief. “Giving him the stuff in the glass would kill him” is a factual statement about what the stuff in the glass would do. If you didn’t want him to die, believing this fact would lead you not to give it to him. But if did want him to die, this fact would tell you how to achieve your aim. There’s nothing about believing the fact itself that tells you what to do. That depends on what you care about.

Of course, it’s possible that beliefs about moral facts do motivate behavior. Nothing Mackie says shows that’s impossible. The most he can say here is that this would make moral facts unique (or “queer”) since factual beliefs do not ordinarily have this kind of motivational effect.1 And that, he believes, is good reason to believe that there are no moral facts.

The epistemological argument also turns on the unique qualities that moral facts would have to have if they existed. Here, the point is that moral facts would have to be known through some means other than the five senses. All the other facts are ultimately known through the senses, according to Mackie. But you can’t see, hear, taste, touch, or smell moral facts. So you would need some other way of coming to know them. That’s what makes moral facts unique (or “queer”).

Does that mean moral facts don’t exist? It’s possible that there is a way of coming to know what they are. As Emma noted, people do change their minds about moral matters and they often think that they change for the better. Mackie’s answer is that they aren’t changing their minds in the light of having discovered something about morality. If we had a special moral sense (a faculty of “intuition” as some philosophers have imagined), this might be different. But there’s no reason to believe there is such a sense. Since he thinks there is no way of knowing moral facts, Mackie believes there are no such facts.

What are the practical effects of this?

Both Ayer and Mackie expect that there would not be any practical effects of accepting their understandings of morality. In Ayer’s case, he thinks that his expressivist theory captures what everyone meant all along when they use moral concepts. So just saying so out loud wouldn’t give anyone any reason to change.

Mackie is unlike Ayer on this point. His theory is that we’re all making a pretty big mistake every time we make moral judgments. So you might expect him to say something like “since moral judgments are always false, it makes no sense to pay attention to them.” But he doesn’t want to take that step. He characterizes his theory as a second order theory rather than a first order theory. While he is maintaining that the moral rules have a subjective basis, he is not saying that what the moral rules tell us to do is whatever we like. By analogy, most of us have a subjectivist theory of law: laws are made by human beings for arbitrary reasons; this is the second order theory of law. But we don’t think that the law directs us to do whatever we want to do; that would be a first order theory of law.

Mackie’s assumption is that a subjectivist second order theory of morality would be compatible with retaining almost all of our ordinary moral beliefs. After learning that they are all false, we would just carry on as before. By analogy, people who think that colors are subjective, meaning that our perception of different colors is due to the way our perceptual systems work rather than the way things really are, still distinguish between red and green lights.

It is not obvious that either one of them has this right. If moral statements are either meaningless or false, why should you take them seriously? If you tell me that I shouldn’t lie because you don’t like it (or, if you say “because ‘boo’”), that isn’t giving me a very compelling reason not to lie. Alternatively, if you tell me that I should believe that lying is wrong even though it’s false that lying is wrong, well, I don’t know what to do with that. (In fairness, it’s false that lying is right too, at least, according to Mackie.)

In a few weeks, we’re going to read a chapter from Williams (“Subjectivism: Further Thoughts”) that will be about what he is going to call the “defusing project.” This is what he’s referring to. What is to be defused is the apparent threat that subjectivism poses to moral belief. More specifically, it’s supposed to defuse this thought: “if morality is subjective, then there is no reason to take morality seriously.”

Key Points

These are the points that you should know or have an opinion about from today’s class.

  1. How Mackie’s position differs from Ayer’s: cognitivist error theory vs. non-cognitivist expressivism.

  2. The argument from relativity.

  3. The two ways that Mackie thinks moral facts would be unique (or, as he puts it, “queer”): metaphysical (beliefs about them would motivate behavior) and epistemological (not known by the senses).


Mackie, J. L. 1977. Ethics: Inventing Right and Wrong. New York: Penguin.

  1. Brink will use “sui generis” in place of Mackie’s word “queer.” I would have just gone with “unique” myself.