A cool way to look at the impetus behind rigid designators (the answer to the question you asked is at the end:) — Mongrel

The quote from Soames is very useful, thank you. My perennial difficulty with all these arguments is that I don't understand how the examples are examples of necessity. To take his four...There are Soames's known by different first names; there are Saul Kripke's that are pet rabbits as well as human beings; 'this table is made of clay' is an entirely contingent statement; the idea that a table is 'made' of molecules is contestable.

I know I have a great reluctance to accept the project of making natural language over into logical form. Perhaps I'll never get over it.

Or perhaps a light bulb is going to go on soon in a great room where all the wise people are gathered and I finally understand what you all mean? I do hope so. In the meantime I hope you'll all carry on part-enlightening me while talking to each other :)

[The proper name] designates the same thing in all possible worlds. — StreetlightX

In an important sense, this is true. Each proper name refers to one specific thing in any possible world. By using a name, we can talk about a thing in any world we might imagine. We can, for example, talk about the Mandarin speaking Obama in our own world. He might even converse with "our" Obama.

The issue is that it is frequently misread as a question of "essence." Instead of recognising each individual thing named in all possible worlds, people treat it as a question of a single thing with the same property in any world. Even you did this above. Supposedly, the property of "Obamaness" makes these those people same, as if anyone called "Obama" was the same person in any possible world, rather than altogether different people who happened to be named "Obama" or have some other similarity.

'this table is made of clay' is an entirely contingent statement — mcdoodle

Why do you say that statement is contingently true?

As long as what is what they name the Earth?

Rigid designation isn't dependent on essential properties.

It's not that hard people, rigid designation means the individual denoted is invariant over worlds of evaluation. That's it.

You're making this into something it's not. Yes, Kripke talks about essential properties. No, that's not going to help you understand what rigid designation is to begin with.

Rigid designation is also not dependent on the law of identity...

Really, you're all making this way more complicated than it has to be.

Being Barack Obama is what Barack Obama does, presumably. We don't stipulate that. He instantiates that property just by being who he is. That should be obvious.

The point is that in imagining alternate scenarios we stipulate that it is he, not some other person or counterpart, that we are making alternate conjectures about. So it is stipulated that we are talking about him, but it isn't stipulated who he is or what it means to be identical to him.

It might be that if the alternate scenarios get too far afield from an individual's actual properties, we have trouble imagining how this could be the 'same' individual. But that just means we have trouble understanding how this individual could have such-and-such properties. This is easy to encode into the model theory with Kripke's machinery by limiting the possible worlds that exist in the model (which would effectively be a way of introducing essential properties). If you're uncomfortable with that, you could instead just say that modals are context-sensitive, and have changing accessibility relations, depending on how far we're willing to depart from actuality. A world that departs too far (say, by making Napoleon a horse) is logically possible, but for most intents and purposes will never be considered for counterfactual evaluation. That said, I personally don't see anything logically incoherent about any such stipulation, even if it's generally a weird thing to stipulate. 'If I were a horse, I'd be a racehorse,' and so on.

Well, think of something like "water is H₂O". According to Kripke, this is necessarily true. It doesn't make sense to consider a counterfactual world where water has some other chemical composition. The identity of a thing as being water is determined by certain facts about it. I don't see why the same can't be said about other things like the Earth, Napoleon, and Obama.

So it is stipulated that we are talking about him, but it isn't stipulated who he is or what it means to be identical to him.

We might stipulate that, but the logic (or metaphysics) of identity might be such that such a stipulation is incoherent. I can stipulate that in some possible world triangles have four sides, or that the square root of 2 is a rational number, but then my stipulation is in error.

And I believe @andrewk's position is that the logic (or metaphysics) of identity is such that a counterfactual person sharing the identity of an actual person doesn't make sense.

Being Barack Obama is what Barack Obama does, presumably. We don't stipulate that. He instantiates that property just by being who he is. That should be obvious.

This is vacuous. I might as well say that being a triangle is what a triangle does or being a bachelor is what a bachelor does or being a squiloople is what a squiloople does. Such claims don't really claim anything. A far more meaningful thing to say is that being a triangle is what three-sided shapes do or being a bachelor is what unmarried men do or being a squiloople isn't a meaningul state of affairs, being that "squiloople" doesn't mean anything.

Well, think of something like "water is H2O". According to Kripke, this is necessarily true. It doesn't make sense to consider a counterfactual world where water has some other chemical composition — Michael

I don't agree with Kripke on this, and his belief seems to stem from scientism that's at odds with ordinary usage. In ordinary usage 'water' likely denotes a qualitative kind not essentially bound to any chemical structure. In a context of chemistry in which we take ourselves to mean by 'water' some specific compound, that's different.

In other words, I don't in ordinary use see anything incoherent about supposing that water had a different chemical structure. But I do see something incoherent about supposing water is, say, qualitatively like zinc. That doesn't seem to make sense, since it would then lack the qualities that water has in virtue of being some qualitative kind.

And I believe andrewk's position is that the logic (or metaphysics) of identity is such that a counterfactual person sharing the identity of an actual person doesn't make sense. — Michael

The error is supposing that there is a such thing as 'an actual person' versus 'a counterfactual person.' This is so in some logics, like Lewis' counterpart logic, but the standard view is that there's simply individuals. There's no actual Obama and other Obama, such that we can ask 'are these two identical?' Rather there's just Obama – and we can ask of him, that one guy, what might have been if such-and-such.

Rid yourself of the notion of a 'counterfactual person.'

We might stipulate that, but the logic (or metaphysics) of identity might be such that such a stipulation is incoherent. I can stipulate that in some possible world triangles have four sides, or that the square root of 2 is a rational number, but then my stipulation is in error. — Michael

It might be, but I don't think it is. In the case of 'triangle,' this is a count noun, not a proper name, so it's different: I agree that properties are (albeit sometimes very loosely) essential for falling beneath the extension of a count noun. For example, I'd be hard pressed to suppose that something were a metal yet had no extension.

As for 2, it's not clear whether this is a sort of proper name or not. Suppose it is a proper name, and suppose it denotes a certain object, say the set {{}, {{}}}. Now, there's something about this set that makes its root irrational, and that'll hold regardless of what properties you assign to individuals in worlds. In other words, what varies from world to world in your model theory may be such that it is not modelable to have this set have a rational root relative to a world: the logic simply won't allow it.

You can do this for a human being too, if you want. But I just think it doesn't reflect ordinary language use – we don't think of people as having many essential properties, even if some properties may be more or less conversationally relevant. It doesn't even sound totally absurd to me to say something like, 'If I were the number 2...' It's a bit mystical, a bit Pythagorean, but OK, so what? That only shows that the only time such a supposition is going to be workable is if we're weird Pythagoreans. But, you say, a person can't be a number! True, but that's not the point, for then I can just say that I am not essentially a person.

The point is that individuals are treated as just being these sorts of pegs in ordinary language, for the non-logical cases. It's an interesting fact that we can do this in spite of metaphysical protestations to the contrary.

Rigid designation isn't dependent on essential properties.

It's not that hard people, rigid designation means the individual denoted is invariant over worlds of evaluation. That's it.

You're making this into something it's not. Yes, Kripke talks about essential properties. No, that's not going to help you understand what rigid designation is to begin with. — The Great Whatever

Yes, I think there's been a bit of a mix up between rigid designators and counterfactuals. We don't need the latter to make sense of the former. Even if we agree that counterfactuals make no sense, and that I cannot be anything other than exactly what I am now, the term "Michael" (when used to refer to me) refers always and only to me in whichever possible world we discuss; it'll just be that I don't exist in these (non-actual) possible worlds.

It's exactly because of rigid designators that @andrewk is able to say that no Mandarin-speaking person in any possible world is Obama.

The point is that individuals are treated as just being these sorts of pegs in ordinary language, for the non-logical cases. It's an interesting fact that we can do this in spite of metaphysical protestations to the contrary. — The Great Whatever

So we talk about people as if they're just "referential pegs". But the question, then, is whether or not people are just such things. If they're not – if them being who they are is by virtue of certain (physical?) facts about them then our ordinary talk is in error, which is andrewk's position. And if they are, then it seems to be that them being who they are is a linguistic imposition (what else does it mean to say that they're "referential pegs"?).

But you seem to deny both of these, which is why I can't make much sense of your position.

An individual isn't essentially a person. That individual may be a person in the actual world. So for instance I'm a person, and an individual, but I might have been a frog (maybe even a sentient one). So it makes no sense to say that 'people are just referential pegs.' The individual in modal logic is held in abstraction from its properties – and being a person is such a property.

An individual isn't essentially a person. That individual may be a person in the actual world. So for instance I'm a person, and an individual, but I might have been a frog (maybe even a sentient one). So it makes no sense to say that 'people are just referential pegs.' The individual in modal logic is held in abstraction from its properties – and being a person is such a property. — The Great Whatever

This seems consistent with a fictionalist/quasi-realist interpretation. You just seem to be saying that we talk about individuals as if they're some abstract thing that can be separated from their actual properties. But the key thing about fictionalism/quasi-realism is that such talk is wrong (even incoherent), and I think that it's this that andrewk is addressing, claiming that individuals aren't some abstract thing that can be separated from their actual properties. So maybe you're just talking past each other?

I'm not interested in mass revision of ordinary language or mass error theory. Such talk isn't 'wrong' to the extent that the sorts of counterfactuals we've been discussing can obviously be true.

And it's clearly not incoherent because everyone but andrewk understands it. His idiosyncratic philosophical commitments simply force him to reject it, but that's the worse for him.

In other words, if the logics designed to account for the most ordinary linguistic phenomena are at odds with what you think, there's a problem, insofar as you're interested in those phenomena and not metaphysical theorizing divorced from it.

So are you saying that individuals are abstract things? This is why I claimed that this entails an anti-realist interpretation of identity (unless you're a Platonist and are going to argue for the independent existence of abstract things?).

There's some sort of misunderstanding here. An individual is whatever it is. So I'm an individual, and I'm concrete, not abstract – a person, of flesh and blood. But I'm not essentially a person, since I could coherently suppose I weren't one. When I talk about what I am, this means what I actually am. To say that I 'am' some abstract thing underlying all my properties is wrong: it's just that I can suppose myself to be any number of other things.

If by 'individual' you mean what's in the model theory of modal logic, then that's a bit of formal machinery, and this formal object, if you like, is just a sort of peg. But in the 'real world' (the only world that there is), there's no such thing as a free-floating individual without properties – it's just that we model individuals separately from worlds in modal logic to reflect the fact that we cans suppose them to have different properties.

It's not that hard people, rigid designation means the individual denoted is invariant over worlds of evaluation. That's it. — The Great Whatever

What makes some x invariant in all possible worlds is that that x is that x, implying the LOI.

We don't talk about names. We talk about objects. Who thought Kripke meant something else?

The concept of some eternal essence that resides in a realm of immaterial forms is not something Kripke talked about. I'm not even sure how that came up. However, if I talk about Obama-the-person, it is not logically possible that the object I'm talking about is not a person. I'm not sure how much more obvious that could be.

There's some sort of misunderstanding here. An individual is whatever it is. So I'm an individual, and I'm concrete, not abstract – a person, of flesh and blood. But I'm not essentially a person, since I could coherently suppose I weren't one. When I talk about what I am, this means what I actually am. To say that I 'am' some abstract thing underlying all my properties is wrong: it's just that I can suppose myself to be any number of other things. — The Great Whatever

This all depends on what sort of thing the term "I" refers to, i.e. the ontological nature of the self. If the self just is some set of concrete properties then nothing that doesn't have these properties can be coherently supposed to be you, even though you might (erroneously) claim otherwise. For counterfactuals to work it is required that the ontological nature of the self (or whatever thing is referred to by the rigid designator) isn't just as some set of concrete properties (that the counterfactual does away with). I don't see how this can be done without claiming that the ontological nature of the self (or whatever thing is referred to by the rigid designator) is as some abstract thing, and so to argue for either Platonism or for identity as a conceptual/linguistic imposition.

If by 'individual' you mean what's in the model theory of modal logic, then that's a bit of formal machinery, and this formal object, if you like, is just a sort of peg. But in the 'real world' (the only world that there is), there's no such thing as a free-floating individual without properties – it's just that we model individuals separately from worlds in modal logic to reflect the fact that we cans suppose them to have different properties.

That we can suppose individuals to have different properties is the very thing that is being questioned. If individuals just are some set of concrete properties then we're not actually supposing them to have different properties; we're just considering something else entirely, and erroneously claiming these two different things to be in some respect the same thing.

Conversely, if we are actually supposing these individuals to have different properties, then being a particular individual is an abstract state-of-affairs, leading to either Platonism or anti-realism.

That we can suppose individuals to have different properties is the very thing that is being questioned. If individuals just are some set of concrete properties then we're not actually supposing them to have different properties; we're just considering something else entirely, and erroneously claiming these two different things to be in some respect the same thing. — Michael

And it's a valid point. It's a decent philosophical topic. But Kripke was addressing the nature of inquiry. We build knowledge one verified proposition at a time. We're always sort of straddling the known and the unknown with the assumption that there's a logical relationship between the two. In this scenario, the known is the aposteriori necessity: necessary because I'm using it to narrow down possible worlds under consideration as I ponder what could be.

It's this pronged image that results in what would be a conundrum if we were involved in a metaphysical thesis. But it's not. The possible worlds we're considering are abstract objects.

"Michael" (when used to refer to me) refers always and only to me in whichever possible world we discuss; — Michael

The problem is that this is stated like a descriptive fact, and the fact in question is an empirical fact--what makes it true or false is what people use "Michael" to refer to.

As long as what is what they name the Earth? — John

As long as people name it "Earth" then that's what we're talking about in terms of names/designators.

This all depends on what sort of thing the term "I" refers to, i.e. the ontological nature of the self. If the self just is some set of concrete properties then nothing that doesn't have these properties can be coherently supposed to be you, even though you might (erroneously) claim otherwise. For counterfactuals to work it is required that the ontological nature of the self (or whatever thing is referred to by the rigid designator) isn't just as some set of concrete properties (that the counterfactual does away with). — Michael

What I am telling you is that counterfactuals do work, and in this way, Q.E.D. Your other option is massive error theory or linguistic revisionism, or an adoption of an alternate framework like counterpart theory.

I don't see how this can be done without claiming that the ontological nature of the self (or whatever thing is referred to by the rigid designator) is as some abstract thing, and so to argue for either Platonism or for identity as a conceptual/linguistic imposition.

I think I've already explained this, and I'm not sure what you mean. Obviously 'I' refers to the speaker, who is generally a person. A person is concrete, not abstract. Identity isn't a linguistic imposition – someone's self-identical just by being whoever they are. Nor is there any Platonism involved. It's just that a person can be supposed to be other than what they actually are, and this is modeled in modal logic by having the domain of individuals and the set of worlds that assign them properties distinct.

Although it should be noted that it's possible for a name to refer to the set of properties that an individual bears relative to a world, rather than that individual, via an operation called Montague lifting. This preserves rigidity in spirit, since the properties are 'centered' on a single individual across worlds, even though technically relative to different worlds the properties change, and the individual can be recovered by an inverse operation known as Montague lowering. This doesn't make an individual 'equivalent to' a set of properties, though, which strictly speaking sounds like nonsense to me, a category error.

Now you're doing it too... letting your ontological prejudices interpret for you. What did you call it? Linguistic revision?

Jeez.

I think to claim like andrewk does that larges swathes of discourse need to be reinterpreted because they are literally senseless (and we only say these things because we say literally senseless things all the time) is either linguistic revisionism or some sort of elaborate error theory. Am I wrong?

Andrewk's primary point appeared to be that an alternate Nixon isn't the actual Nixon and so contingency speech can't be taken literally.

You answered that by saying that contingency speech isn't about an alternate Nixon. It's about Nixon.

First of all, the point Andrewk made (sans the crap about processes) is one that Spinoza and Leibniz would agree with. Schopenhauer would agree with it, so it's not straight from the loony bin. Does pondering that get us any closer to understanding Kripke? No.

It appears to me that you're doing the same thing on the other side. You seem to be insisting that all possibility is logical possibility. You also seem to want to say that we can't pick objects out by specific properties (and by way of this, identify essential properties). We do it all the time, so aren't you revising language use if you insist on this?

Kripke wanted to separate logical possibility from epistemic possibility (something his forebears didn't want to do). That and the rigid designator are the tools he thought he needed to establish that there can be informative necessary truths. Did he succeed?

Andrewk's primary point appeared to be that an alternate Nixon isn't the actual Nixon and so contingency speech can't be taken literally. — Mongrel

This was just my point. Hence the revisionism, insofar as 'contingency speech' is an ordinary feature of language as revealed through counterfactuals. Andrewk is poised to reject an entire grammatical construction as literally nonsensical.

From the Wikipedia entry on a posteriori necessity:

Kripke’s causal theory of reference is a necessary component of a posteriori necessity.

The causal theory of reference is an explanation of how terms acquire specific referents. Referring terms that it may include are proper names, natural kinds, and logical terms. A causal theory of reference claims that the referent of a name is fixed by the original act of naming. Future uses of the name continue to successfully refer to the referent via a causal chain. The chain can then, in principle, be traced back to the original act of naming. Kripke indicates that this explanation is more appealing than the descriptive theory of names developed by Bertrand Russell and Gottlob Frege. The descriptive theory says that names are attached to a uniquely identifying description of that thing. A name’s semantic contents are identical to the descriptions associated with them.

Consider these two theories’ approach to the following question: To whom does the name “Richard Nixon” refer?

The causal/ historical theory could proceed thusly: Suppose the parents of a newborn baby boy gesture to their child and declare “We shall name him Richard Nixon!” From that time onward, that person is referred to as “Richard Nixon”. When someone uses the name “Richard Nixon”, that act can be traced back to the initial baptism of the child by his parents. The name refers to the object originally designated as “Richard Nixon”.

According to the descriptivist theory of names, the referent might be fixed like so: the name “Richard Nixon” refers to the man who won the 1968 election, the 37th President of the United States, etc. Only one person satisfies this description and that is “Richard Nixon”. This could be true for a number of different unique descriptions that identify “Richard Nixon”. So the name “Richard Nixon” would refer to whoever satisfies the particular unique description.

My answer to "To whom does the name 'Richard Nixon' refer" is: To whomever (or whatever) the person using the name "Richard Nixon" on a particular occasion has in mind as the extension of the term.

My answer to "To whom does the name 'Richard Nixon' refer" is: To whoever (or whatever) the person using the name "Richard Nixon" on a particular occasion has in mind as the extension of the term. — Terrapin Station

A simpler, better answer: "Richard Nixon" refers to Richard Nixon.