What is there to have a view on? Seems like a statement of the bleedin' obvious to me.

Rigid designation is about the evaluation of a proposition at a world, given the way that it's used. That names could mean something else, or refer to different individuals, if the language were different, doesn't make them any less rigid designators. What a word means is contingent, but given what a name means, its value, the individual it refers to, is invariant over possibilities.

So if you use modals or attitude reports, what makes the modal true can depend on different individuals if you use a non-rigid designators:

Sue thinks the president is incompetent.

Sue can think this without believing the president is any particular person: she simply believes that the unique person who is the president, whoever that is, is incompetent (she might know it is Michael, or might not). Her belief will be validated by any individual being incompetent, so long as s/he is the president. But

Sue thinks Michael is incompetent.

Sue can only believe this if she thinks some particular person, namely Michael, is incompetent. And only Michael's incompetence is relevant to her belief being true.

I agree with your take on it more or less.

Conventional rigid designator talk always struck me as hopelessly muddled. I don't understand The Great Whatever's explanation either--to wit:

"Rigid designation is about the evaluation of a proposition at a world"--not sure what "the evaluation of a proposition" is saying. Do you just mean its truth value?

"That names could mean something else, or refer to different individuals, if the language were different, doesn't make them any less rigid designators."--in those cases, though, just how are any designators rigid?

"What a word means is contingent, but given what a name means, its value, the individual it refers to, is invariant over possibilities" --its value is invariant over possibilites by virtue of what?

Do you just mean its truth value? — Terrapin Station

Yeah, the value of a proposition relative to a world is a truth value. The value of a name inside the sentence expressing that proposition relative to a world is an individual. The point is that, regardless of world of evaluation, the name will always refer to the same individual.

Formally, if you like, the semantic value of a name can be represented either simply as an individual, or as a constant function from worlds to individuals, making the world of evaluation vacuous. This contrasts with definite descriptions like 'the president,' which are represented by non-constant functions from worlds to the president in that world (which might be different individuals in different possibilities).

in those cases, though, just how are any designators rigid? — Terrapin Station

Because the issue is not what the words might have meant, but what they do mean. Given what a name does mean, its value doesn't change over possibilities. This is different from non-rigid designators like definite descriptions, which can contribute different individuals depending on the world of evaluation, even holding fixed what they mean.

its value is invariant over possibilites by virtue of what? — Terrapin Station

I'm not sure what you mean. Are you asking why names mean what they mean? That's a complicated question: the more basic descriptive claim is just that they mean a certain thing, which we can see even without knowing why they come to mean this.

For reference, the technical notion of rigid designation makes no sense outside of an intensional semantics and typically is cast in terms of possible world semantics. So it's a technical notion, but it has intuitive consequences, like what I talked about with attitude reports above. The intuitive notion of what rigid designation is can be seen from these sorts of examples, and intuitions about what makes a sentence true, independent of the formal framework.

But if you want to know concretely what rigid designation means, knowing a bit of modal logic is helpful. In modal logic, values or extensions are given relative to possible worlds: the point is just that something is rigid if the possible world with respect to which it's evaluated makes no non-trivial contribution to its value. So names are still, like in non-modal logic, behaving like individual constants, that just denote some individual.

Because the issue is not what the words might have meant, but what they do mean. Given what a name does mean, its value doesn't change over possibilities. — The Great Whatever

(I'm bypassing issues with "meaning" versus "definition" by the way)

But what a word means is simply what people use it to mean at any given time. That can shift (and it's also not universal). So it wouldn't be rigid.

But what a word means is simply what people use it to mean at any given time. That can shift (and it's also not universal). So it wouldn't be rigid. — Terrapin Station

Again, a claim that something is a rigid designator is not a claim that it must mean something or couldn't have meant something else. It's a claim about what it actually means.

Well, so it's just rigid at a particular time, to particular persons? That would make sense, at least, although I wish someone would have explained that to me long ago if that's all it's saying. (That never seemed to me to be all it was saying.)

All claims about what a word means are claims about what it means at a certain time. I'm not sure what you're getting at.

For example, if I say that 'tree' means tree, and always picks out trees relative to a world of evaluation, it's not a rebuttal to say that 'tree' might have meant turnip instead. That's just irrelevant. The point is what the word means, not what it might have meant if the language were different.

Also, the fact that a name might have referred to something else doesn't mean that it might not have been a rigid designator. It just means it might have been a rigid designator designating some other thing. The claim is that proper names as a class are semantically rigid designators, whatever they refer to.

Here's what I'm getting at:

For any term, including "Hesperus" for example:

"a" actually means m only at time T1 to person S (and of course, it can also mean m (ignoring nominalism) to person U, V, etc. at time T1, too.)

At time T2, "a" might mean l instead to person S. (Of course, it could still mean m to S at T2 (again ignoring nominalism), but it won't necessarily.)

Also, at time T1, "a" might mean k to person U.

Even at this, I was about to ask if "rigid designator" is simply saying that "a" means m to S (and U, V--whoever else) at all possible worlds at T1, but then it struck me that that's still muddled, because how the heck would we know that in possible world w, "a" would still mean m to S at T1?

I don't know what you mean by 'to person S.' Generally we talk about what a word means, not what it means to someone.

Meaning is always to someone.

OK, I don't agree with that, but I'm not sure what the point is anyway. Is the point that the meaning of words can change over time? I agree, but that's not relevant to the question of what a word means.

Meaning can change over time, and can vary per person. How would that not be relevant to what a word means? That IS what/how words mean something.

It's not relevant to what a rigid designator is. Calling something a rigid designator isn't a claim that it can't change in meaning over time. That has nothing to do with it. All words can change meaning over time, rigid designators or not.

Oy, this is remaining muddled. So what the heck is "rigid" about any of this? We're talking about reference/meanings. Well, they can change over time (so they're not rigid in that regard) and they can vary per person (so they're not rigid in that regard) and we can't know plus we're not presenting an argument for how they'd refer to the same things by the same persons in other possible worlds (so that's not rigid either). What's rigid then?

What's rigid is that their referent doesn't change with respect to world of evaluation. See the attitude report example above.

What's rigid is that their referent doesn't change with respect to world of evaluation. — The Great Whatever

What is the argument for that? How in the world would we know that someone wouldn't change how they use a word in a different possible world or a different context or whatever? (I really should read Naming and Necessity again, obviously, but my memory of it was that a lot of it was muddled nonsense.)

Again, the point is not what the word would mean in a different possible world. The point is that given what the word means as it's used now, the referent of the name is the same regardless of world of evaluation. You can see this clearly with structures that shift the world of evaluation, like counterfactual conditionals.

'If the winner were rich, I would be his friend.'

'If Michael were rich, I would be his friend.'

Suppose that in the world of utterance, Michael is the winner. Yet these sentences mean very different things. The first means that if it were the case that the person who won, whoever that might be, were rich, the speaker would be that person's friend in that world. The second says only of a single person, Michael, that this would be the case. So despite the fact that Michael is the winner (these are the same person), 'the winner' is non-rigid and can denote people other than Michael in other possible worlds (namely, whoever won). But 'Michael' always denotes the same guy, Michael, and the truth of the second sentence is not dependent on who won, whether it is Michael or not, but on Michael, whether he won or not.

One thing we should probably clear up is the possible world ontology we're using. Are we talking about possible worlds in a "realist" sense--that is in a Lewisian or MWI sense, where there literally are other worlds where counterparts of us exist? Or are we simply talking about counterfactuals and possibilites in our one world?

If the latter, by the way, and we're focusing on how a term is actually used, then it turns out that there's only one possible world--the actual world.

Aside from that, a distinction makes sense between S saying "the winner" where S does not have a specific person in mind and S saying "Michael" where S does have a specific person in mind.

However, as soon as we say something like "'Michael' always denotes the same guy," we're introducing a temporal element, and "'Michael' always denotes the same guy" is only the case if we're talking about whoever is uttering the sentence, for all T, having the same guy in mind as a referent. But we don't at all know that that would be the case.

Maybe you don't really want to introduce a temporal element--in which case we should avoid the phrase "always denotes." Maybe we just want to say that for everyone who uses "Michael" as the name of the same particular guy NOW, they NOW use that name for that same particular guy . . . which is pretty vacuous.

We can't even introduce "Michael" in a variety of sentences for this without introducing a temporal element though--the people using the name "Michael" in a particular way NOW can only entertain so many sentences with the name "Michael" at the same time.

One thing we should probably clear up is the possible world ontology we're using. Are we talking about possible worlds in a "realist" sense--that is in a Lewisian or MWI sense, where there literally are other worlds where counterparts of us exist? Or are we simply talking about counterfactuals and possibilites in our one world? — Terrapin Station

It doesn't matter. The question of rigid designation is an empirical semantic one, independent of these metaphysical claims. Rigid designation can be employed in any ontology of possible worlds.

If the latter, by the way, and we're focusing on how a term is actually used, then it turns out that there's only one possible world--the actual world. — Terrapin Station

This is false, insofar as languages have explicit mechanisms for evaluating relative to non-actual possibilities, including modal adverbs, sentential modals, conditionals and counterfactuals, attitude reports, and so on.

However, as soon as we say something like "'Michael' always denotes the same guy," we're introducing a temporal element — Terrapin Station

No. Again, this is not about the change of meanings over time. It is about which individual is denoted relative to which possible world.

is only the case if we're talking about whoever is uttering the sentence, for all T, having the same guy in mind as a referent. But we don't at all know that that would be the case. — Terrapin Station

What a name refers to doesn't depend on who someone has in mind. It depends on what the word means. If I confuse Michael and Brett, and say 'Brett is the winner,' when Brett lost, and I meant that Michael is the winner, I've literally said something false about Brett, though this was not my intention. I may have meant something true that came out wrong, and my interlocutors may be willing to accommodate me and reconstruct what I meant to say.

It doesn't matter. The question of rigid designation is an empirical semantic one, independent of these metaphysical claims. Rigid designation can be employed in any ontology of possible worlds. — The Great Whatever

But that makes no sense. If we're talking about real alternate worlds, we have no idea what our counterpart might be actually using a term to refer to in that alternate world.

This is false, insofar as languages have explicit mechanisms for evaluating relative to non-actual possibilities, including modal adverbs, sentential modals, conditionals and counterfactuals, attitude reports, and so on. — The Great Whatever

You're ignoring the "focusing on how we're using the term in the actual world" part. It's vacuous if we're not talking about real possible worlds to say that "in all possible worlds we're using the term to refer to x" if we're focusing on how we're using the term in the actual world, because with respect to how we're actually using the term, there's only ONE possibility--the actual way we're using the term. I'm not saying there aren't counterfactuals and so on. But if there's only one real world, there's only one actual way we can be using the term. There are no other worlds for usage of the term. There's just usage of the term in talk about counterfactuals and so on (though we're limiting ourselves to a single moment in time, so we can't do too much talking about counterfactuals and so on).

No. Again, this is not about the change of meanings over time. It is about which individual is denoted relative to which possible world. — The Great Whatever

What I'm referring to is the word always. That's a temporal term. If you don't want to introduce temporal considerations, you need to not use a term like that. Once you introduce temporal considerations, you can't avoid the possibility that we can start using a term differently at another time, and there's no rigidity in that.

What a name refers to doesn't depend on who someone has in mind. — The Great Whatever

Of course it does! That's ALL it depends on. That's all there is to it.

It depends on what the word means — The Great Whatever

What the word means is determined by how individuals think about it!

If I confuse Michael and Brett, and say 'Brett is the winner,' when Brett lost, and I meant that Michael is the winner, I've literally said something false about Brett — The Great Whatever

You can change your mind about who you're referring to by each name (for example if you don't know them that well or whatever and you keep forgetting who is who per what they call themselves for example). And you can also say a different sound than what you meant to say (saying "Brett" when you wanted to say "Michael"). I'm not sure which situation you have in mind there. In both cases, reference is purely a matter of what one has in mind. You can't use words incorrectly. You can use them unconventionally, and you can change your mind about how you use them, or say something you didn't mean to say, etc.

If we're talking about real alternate worlds, we have no idea what our counterpart might be actually using a term to refer to in that alternate world. — Terrapin Station

It's not about what terms they use to talk about things in their world. It's about what terms we use to talk about things in their world.

But that makes no sense. If we're talking about real alternate worlds, we have no idea what our counterpart might be actually using a term to refer to in that alternate world. — Terrapin Station

Counterparts aren't necessary for a possible world semantics, and even if they were, I don't see how it's relevant.

You're ignoring the "focusing on how we're using the term in the actual world" part. It's vacuous if we're not talking about real possible worlds to say that "in all possible worlds we're using the term to refer to x" if we're focusing on how we're using the term in the actual world, because with respect to how we're actually using the term, there's only ONE possibility--the actual way we're using the term. I'm not saying there aren't counterfactuals and so on. But if there's only one real world, there's only one actual way we can be using the term. There are no other worlds for usage of the term. There's just usage of the term in talk about counterfactuals and so on (though we're limiting ourselves to a single moment in time, so we can't do too much talking about counterfactuals and so on). — Terrapin Station

Again, the point is not how the word would be used in different possibilities, or what else it might mean. It's what, given what it actually means and how it's actually used, the word can refer to given different worlds of evaluation.

What I'm referring to is the word always. That's a temporal term. — Terrapin Station

This strikes me as a pointless quibble, but no it's not. Here 'always' means 'in any situation,' which is a perfectly ordinary use of the word. If you want to change the wording, fine. The point is that regardless of world of evaluation, the individual referred to is the same.

Of course it does! That's ALL it depends on. That's all there is to it. — Terrapin Station

It depends on the conventions of the linguistic community, which aren't reducible to any single speaker's intentions.

You can't use words incorrectly. — Terrapin Station

Yes you can, as evidenced by the fact that you can use them correctly.

It's not about what terms they use to talk about things in their world. It's about what terms we use to talk about things in their world. — Michael

Well, how many things in how many other worlds can the same person talk about at the same time though?

Here 'always' means 'in any situation,' — The Great Whatever

Similar to the above question, how many situations can you have for the same individual at the same time?

It depends on the conventions of the linguistic community, which aren't reducible to any single speaker's intentions. — The Great Whatever

Conventions refer to contingent commonaltiies among many individuals, sure. Why would we only be talking about conventions though? It's not as if conventions are right by virtue of being conventions. All of the unconventional references are equally references/meanings.

Yes you can, as evidenced by the fact that you can use them correctly. — The Great Whatever

No you can't. Obviously you can't use them incorrectly, either. You can use them unconventionally. But conventional/unconventional doesn't map to correct/incorrect.

Any number. Suppose you say 'John might be home.' This means there's a possibility he's home (say, given what we know to be true), not that he actually is. This roughly means that among all the possibilities that we can consider (say, compatible with what we know), among them are at least some in which John is at home (and maybe others in which he's not).

Well, how many things in how many other worlds can the same person talk about at the same time though? — Terrapin Station

I'm not sure what you mean.

But consider this; do you understand the difference between talking about a possible world where Barack Obama isn't the president and a possible world where there's a man called "Barack Obama" who isn't the president?

Any number. Suppose you say 'John might be home.' This means there's a possibility he's home (say, given what we know to be true), not that he actually is. This roughly means that among all the possibilities that we can consider (say, compatible with what we know), among them are at least some in which John is at home (and maybe others in which he's not). — The Great Whatever

Any number? So the same individual could consider 1,000 different propositions about John at the same time?

No, one proposition can be sensitive to any number of possibilities. That John might be home just means there's some possibility he is.

But notice if you say 'the winner might be home,' the possibilities that qualify to make this true are not necessarily those that include just one individual being home, viz. John. A possibility in which Michael is home also qualifies, so long as he's the winner.

This is because 'the winner' is a non-rigid designator: it refers to the winner in a world, whoever that is. 'John' just refers to John in every world, whether he's the winner or not.