Am I understanding this right? A teacher starts talking to a guy in a bar, tells him a story about how another teacher used the word “nigger,” and this results in the storyteller getting into trouble?

The intuitive sense of unfairness comes from the fact that we all understand the difference between genuinely asserting and using the same language in a way that doesn’t assert. You might overhear me utter the phrase “Ari is so handsome” as I’m in the midst of saying “Only Mrs. Kelman could think that Ari is so handsome,” for example. While the phrase itself retains its meaning in the two contexts, the sentences mean very different things.

As I recall, Frege’s general point about this is that there’s no operator that indicates what follows is being asserted. Phrases like “I’m genuinely asserting that….” are themselves subject to the same problem– they can be put in contexts where they aren’t asserted. Geach seizes on this to develop a really interesting objection to expressivism, the thesis that moral judgments are expressions of noncognitive states such as emotions, rather than statements of (moral) fact.

On the expressivist view, “lying is wrong” is doing the work of “boo lying!” But notice how “lying is wrong” appears in contexts where it’s genuinely asserted and in contexts where it isn’t. Canonical example:

1. Lying is wrong.
2. If lying is wrong, getting your little brother to lie is wrong.
3. Hence, getting your little brother to lie is wrong.

This argument looks good (by which I mean deductively valid). In order to be good, though, “lying is wrong” has to mean the same thing in (1) and (2). Wrinkle: in (1) it’s being asserted, in (2) it’s not. Even if (1) makes sense as the expression of a boo-attitude toward lying, (2) doesn’t. You’re not booing lying because you’re not saying lying is wrong, when you assert (2). So it looks like the expressivist is stuck. (Stuck in two related ways: first, it’s a problem that the expressivist hasn’t given us an account of (2), and second it’s a problem that, whatever an account of (2) would be, it won’t preserve the meaning of “lying is wrong,” and that’s needed to make sense of the validity of the argument.)

Simon Blackburn has a go at this by trying to understand (2) as something like “boo for the following conjunction of attitudes: booing lying while not booing getting little brother to lie.” Not really convincing, but a nice attempt.

There’s lots more to say, but now you’re equipped to utter “The latest O’Keefe shenanigans got me musing about old Gottlob Frege” and that will make you sound erudite.