“Emotivism” is best understood not as an ethical theory per se, but as a metaethical theory. That is, it’s a theory about ethical theories, or at least a theory about moral language.
In order to explain emotivism, it is necessary to first provide some background about language.
Consider the different types of sentences:
1. Declarative sentences: These sentences make some cognitive claim that in principle is true or false.
* Walruses speak with a German accent.
* Loretta is wider than Marco’s tuba.
* There are at least three prime numbers between 5 and 500.
2. Interrogatory sentences: In other words, questions.
* Is my shoe on fire?
* How many pieces did Tyrone eat?
* Are wombats indigenous to Australia?
3. Imperative sentences: In other words, commands or requests.
* Give Mrs. McGillicuddy her teeth.
* Place Peg A in Slot B.
* Keep off the grass.
4. Exclamatory sentences: These sentences directly express emotion.
* Oh my God!
Note that it is only declarative sentences that can be true or false. They are the only ones that have that kind of cognitive content.
So, “I am in pain” is true or false. That is not the case with “Are you in pain?” “Let me know if you are in pain,” or “Ouch!”
Even though “I am in pain” and “Ouch!” convey pretty much the same information, they do not do so in the same way. “Ouch!” is more akin to a wince or some physical manifestation of pain than to a statement attributing pain to oneself.
OK, keep this in mind, and let’s talk about moral philosophy.
Many philosophers have long been puzzled by the status of moral claims. It isn’t just that there’s some uncertainty about the truth or falsity of “Abortion is wrong,” “You ought to treat others the way you want to be treated,” or “Rapists deserve the death penalty,” but it’s not clear what it would even mean for them to be true or false, what would constitute evidence or reasons making them true or false.
Are normative statements like that equivalent to or deducible from descriptive statements? How so, how do you get an “ought” from an “is”?
Roughly speaking, to say a sentence is true generally means it corresponds to reality. So the sentence “Hans went fishing” is true if and only if the person designated by the name “Hans” went fishing. Truth is a relation between a sentence and whatever aspect of the world it’s supposed to correspond to. If it fails to correspond-if Hans didn’t go fishing but snuck off to a strip club instead-then it’s false.
It’s a little trickier with non-empirical claims like about mathematics or pure logic, but in those cases you can give an account of truth based on what follows from the definitions of the relevant concepts.
But what does “Slavery is wrong” correspond to, or fail to correspond to? What are the truth conditions for a sentence like that?
OK, now, finally, let us turn to emotivism.
Emotivism is a 20th century metaethical theory put forward by philosophers including A.J. Ayer and Charles L. Stevenson, though philosophers from earlier centuries including David Hume had expressed somewhat related ideas.
Emotivists claimed that where moral philosophers had gone wrong was in assuming that moral judgments, or normative claims in general, are declarative sentences, when in fact they are not. Thus the reason it’s always been such a challenge to determine the truth or falsity of moral statements is precisely because they are not the type of sentence that even can be true or false.
In grammatical structure, moral judgments certainly look like declarative sentences. (Compare “Racism is bad” and “Guyana’s climate is tropical.”) They seem to straightforwardly attribute properties to things.
But, the emotivists say, that’s why people have been fooled for so long. They may superficially have the structure of declarative sentences, but in fact they do not function as such.
Instead, they are imperative sentences or exclamatory sentences. “Abortion is wrong” is not to be understood as attributing some property of wrongness to the act of abortion. Depending on the context and the way it is stated, it is better translated as “Don’t have an abortion” or “Boo, hiss! Abortion! Boo, hiss!”
Basically moral sentences are sentences that express approval or disapproval in a way that cannot be true or false. An imperative sentence like “Don’t have an abortion” or an exclamatory sentence like “Boo, hiss! Abortion! Boo, hiss!” are neither true nor false.
Or to put it another way, moral sentences, according to the emotivists, are “non-cognitive.”
Does this mean emotivists are relativists or subjectivists? Is “slavery is wrong” just a statement about oneself and not about slavery, equivalent to “I disapprove of slavery”?
Sort of, but not quite. Because note that “I disapprove of slavery” is a declarative sentence. It’s a declarative sentence attributing a certain property to oneself rather than attributing a property to slavery, but it’s still a declarative sentence, and as such is still cognitive, is still true or false.
What the emotivists are saying moral language does is evince or express approval or disapproval, not attribute an attitude to oneself like that.
If that distinction is obscure, think of it this way: Remember the difference between “I am in pain” and “Ouch!” The first is a declarative sentence that is true or false, the second is an exclamatory sentence that cannot be true or false. For a subjectivist, when you say “Slavery is wrong,” it’s equivalent to “I disapprove of slavery,” which is analogous to “I am in pain.” For an emotivist, when you say “Slavery is wrong,” it’s equivalent to “Boooo slavery!” which is analogous to “Ouch!”
In the ensuing decades, much has been written pro and con about this metaethical theory. One point that seems to be pretty well established on the con side is that emotivism probably fails as a complete description of how real people use moral language and what they intend by it.
But a discussion of the merits of emotivism goes beyond the scope of this article.