This supposedly “uncontroversial” principle is false. First, however, here’s why it is not false.
“Morning Star” refers to the same thing as “Evening Star,” i.e., the planet Venus. Therefore, according to the principle, every proposition involving MS must have the same truth value as the exact same proposition with MS substituted with ES.
However, consider the proposition
(i) “‘MS’ means Venus-seen-in-the-morning.”
It’s true. Substituting “ES” for “MS,” however, renders it false. So, the principle applies to all propositions where the two identical things are used and not mentioned. “Morning Star” is merely mentioned in (i). (i) is about the term not the thing.
Now consider the example in Wikipedia:
1. Clark Kent is Superman’s secret identity; that is, they’re the same person (identical), but people don’t know this fact.
2. Lois Lane thinks that Clark Kent cannot fly.
3. Lois Lane thinks that Superman can fly.
4. Therefore Superman has a property that Clark Kent does not have, namely that Lois Lane thinks that he can fly.
5. Therefore, Superman is not identical to Clark Kent.
(2) can be rewritten as “Lois Lane thinks the proposition ‘Clark Kent cannot fly’ is true.” It should be obvious that by asserting (2) we do not say anything about a man, but only about a proposition. We learn nothing about the man, whether Clark Kent or Superman, by reading this statement; we rather learn something only about the proposition “Clark Kent cannot fly,” namely, that is it being held by Lois Lane. And who cares about that? This is a more general case of mentioning a term rather than using it.
Nor is the issue necessarily intentional properties. It may be that “Morning Star” has a property of being considered beautiful by Smith that “Evening Star” lacks. Now let “Smith thinks the MS is beautiful in the morning” be true. Then “Smith thinks the ES is beautiful in the morning” is also true. Then the MS and ES are still indiscernible.
These are not the ways to refute the indiscernibility of identicals. But another way is available.
While “MS” and “ES” refer to the same thing, they mean different things: “MS” means “Venus-when-seen-in-the morning”; “ES,” “Venus-when-seen-in-the evening.” (These definitions happen to be correct in ordinary language.) Identity of reference does not entail indiscernibility in meaning. Given these meanings, we no longer need Smith at all. E.g.,
(ii) “MS is higher above the ground than ES”
may be true, and even if actually false, is only contingently so; whereas “Venus is higher above the ground than Venus” is necessarily false. This works even if MS and ES are taken de re.
Proposition (ii) is, of course, garbage as expressed; e.g., it can be rewritten much more helpfully as “Venus appears higher from the ground in the morning than in the evening.” But indiscernibility of identicals is a philosophical principle that ought to be true on its own terms.
Again, consider a dialog like:
A: “I saw the Morning Star today.”
B: “Oh, so you saw Venus in the morning.”
“MS” can convey more information than “Venus”:
A: “I didn’t realize it was Venus. Now I’ll go to Wikipedia and read all about it!”
We can have identity in reference or indiscernibility in meaning, but those are separate things.