So you're agreeing that we don't look at all possible worlds. We look at all relevant possible worlds... specifically where our rigid designators pick out objects that exist in those worlds.

To interject...there is a paper by Michael Weisberg called Water is not H20. I liked it. Distilled water, maybe, although I gather chemists would prefer greater precision even then. But this topic gets me into hot, erm water. Someone turned me down for a Master's course over a paper i wrote about it. Water just is H20, she exclaimed. How then can heavy water be a form of water? How can the polluted water in my local canal be water?

So you're agreeing that we don't look at all possible worlds. We look at all relevant possible worlds... specifically where our rigid designators pick out an object that exists in that world. — Mongrel

Well, if we are evaluating the de re modal status of a true statement (i.e. inquiring about its being necessary of merely contingently true predication of some object) then we are looking for a possible world at which this statement might be evaluated false. If the object mentioned in the statement does't exist in some possible worlds, then the statement can't be evaluated at all, so those worlds aren't relevant. But when the statements at issue are purported a posteriori necessary statements, then it is not up to us (or up to "speaker's intentions") whether or not such a possible world W exists, or doesn't exists, such that the statement would be evaluated false.

Hence, I may know who Mark Twain is, and also have heard about some writer called Samuel Clemens, and not know that they are the same person. When I make this discovery (in the actual world, through historical investigation) then I can rule out that there is any possible world at which Mark Twain isn't Samuel Clemens. Possible worlds where Mark Twain never was born (whoever else might have been given the same name) aren't relevant to my discovery.

By the way, it seems to me that the reason why we speak of evaluating statements at possible worlds, rather than in possibles worlds, in the technical literature, might be in order preempt the confusion that the statements are to be evaluated in accordance with the alternative linguistic conventions that hold among our Doppelgangers "in" those worlds. This is a mistake that you (and John, if I remember) have a tendency to make. Evaluating a statements at a world means assessing its truth value in the hypothesized circumstances of this world while interpreting the statement in accordance with our own language. Hence, who Mark Twain refers to in some possible world W, at which we are evaluating the statement "Mark Twain is a featherless biped" is insensitive to whatever name Mark Twain goes by in that world. When those conventions are lost track of, endless confusion ensues.

To interject...there is a paper by Michael Weisberg called Water is not H20. I liked it. Distilled water, maybe, although I gather chemists would prefer greater precision even then. But this topic gets me into hot, erm water. Someone turned me down for a Master's course over a paper i wrote about it. Water just is H20, she exclaimed. How then can heavy water be a form of water? How can the polluted water in my local canal be water? — mcdoodle

Yes indeed. The crude identity statement is quite a simplification. Hilary Putnam himself was sensitive to some of those pragmatic consideration, even as early as his original The Meaning of "Meaning" paper, if I remember.

Which is why I rather like the loose phrasing "Water is essentially materially constituted by H2O", since "essentially" has the ordinary language connotation "almost but not quite exactly". It's got to have some fair amount of H2O in it, and heavy water is almost but not quite H2O, isn't it? (For deuterium and tritium also are hydrogen isotopes). Putnam was also, of course, sensitive to the fact that some stuff-names, such as jade, aren't "natural kind" terms at all. And some acceptions of "water" may also be like that.

It is not a "purported alternative France" that I was talking about; it is France. France would still be France if, counterfactually, at some point in time, its history had diverged from its actual history in some inessential respects. — Pierre-Normand

To me it seems that there's your problem, right there ; you haven't been able to specify what are the essential and what are the inessential "respects" that we could use to say that it is a counterfactual France that we are thinking about; or how we could come up with criteria for deciding which respects are essential and which inessential, and also how those criteria could be anything but determined by qualitative.

Maybe you are an actualist. Many of your comments point in that direction. Actualists believe that whatever P is actually true is necessarily true, and whatever Q is actually false is necessarily false. Hence, anything that is possible is necessary, on that view; there is no non-actualized possibilities. But that is a rather contentious metaphysical doctrine.

I don't know; this kind of philosophy is not really my area of interest, so I am not familiar with "actualism". I do believe that whatever has happened has necessarily happened, but I don't believe that whatever happens in the future is necessary now. In other words I am not a determinist; I think the future is radically open, and that there is now a real.diversity of possibility.

In any case, I didn't want to get into modal philosophy at all. I was originally making a point about the differences between the kinds of knowledge exemplified by "Paris is the capital of France" and "the Sun is shining at such and such a location at such and such a time", and TGW and Jamalrob denied that there is any valid distinction in kind between those two propositions. That there is such a valid distinction is all I have been arguing for.

If the object mentioned in the statement does't exist in some possible worlds, then the statement can't be evaluated at all, so those worlds aren't relevant — Pierre-Normand

If the statement can't be evaluated at all at those worlds, then it certainly isn't true at them.

The statement is only true at worlds in which Samuel Clemens exists. And you would disagree with this.... why?

But before we evaluate the statement, it would be important to know what it means. Does linguistic convention preclude multiple meanings of the sentence? Of course not. So by some means, a particular meaning would have to be chosen. For instance, by some means, it would have to be determined that Mark Twain means a person, and the sentence is not answering a question about what Samuel Clemens' pen-name was.

How do you suppose that sort of determination is made (choosing from the various possible meanings of an utterance?)

In any case, I didn't want to get into modal philosophy at all. I was originally making a point about the differences between the kinds of knowledge exemplified by "Paris is the capital of France" and "the Sun is shining at such and such a location at such and such a time", and TGW and Jamalrob denied that there is any valid distinction in kind between those two propositions. That there is such a valid distinction is all I have been arguing for. — John

So far as I can see, that the Sun is shining at some place X at some time T is an empirical and contingent fact of geography/meteorology/astronomy, whereas that France has Paris as its capital city at some time T is an empirical and contingent fact of human history. The Sun would still be the Sun while hidden behind a thick cloud cover and France would still be France with its capital city established away from Paris. I don't see any categorical difference between those two cases in point of topic-neutral metaphysics.

If the statement can't be evaluated at all at those worlds, then it certainly isn't true at them. — Mongrel

Not necessarily. If you analyse all singular terms as definite descriptions (as Russell does in the case of all singular referring expression except for private demonstrative reference to one's own sense data and one's use of the first personal pronoun) then all statements of the form "The X is F" would be evaluated false at possible worlds where nothing answers to the description "X". But proper names (or ordinary demonstratives) aren't like that. If a proper name NN is empty, then the statement "NN is red", say, is not false. It is meaningless. It purports but fails to express a singular thought. Nobody would say that NN isn't red (which would logically follow from "it is not true that NN is red") who knows NN not to exist.

The difference you are still failing to see is that the fact that the sun is shining is directly observable; whereas the fact that Paris is the capital of France is not, The latter is a kind of secondary fact that can be known only by accepting what others have said; it is derivative on the fact that people say that they designate and consider Paris to be the capital. If people ceased to designate and consider Paris to be the capital tomorrow it would cease to be the capital. By contrast, whether the Sun is shining at some place and time has nothing to do with whether people consider it to be shining. Put very simply, the fact that Paris is capital is dependent on thought in a way that the sun shining is not. That is why I say it is a quasi-semantic proposition.To me this is a very simple and self-evident distinction, as far as I can I see no one has yet come up with a good argument against it and I am therefore quite astounded at the degree of resistance to it I have encountered here.

I think our discussion fell off the rails.

I hold that Samuel Clemens is Mark Twain can only be true at a possible world that contains an object picked out by Samuel Clemens.

If you disagree, we have an impasse, but all I can say is I think you're wrong.

I hold that in the case of any utterance, it will have to be sorted out somehow what it means. You can't just point to what you understand to be linguistic convention.

You apparently disagree with that as well. Again.. I think you're wrong.

Since you disagree with both the premises to the argument I think follows from here... you obviously aren't going to agree with the conclusion.

Good discussion. Thanks!

The difference you are still failing to see is that the fact that the sun is shining is directly observable; whereas the fact that Paris is the capital of France is not, The latter is a kind of secondary fact that can be known only by accepting what others have said; it is derivative on the fact that people say that they designate and consider Paris to be the capital. If people ceased to designate and consider Paris to be the capital tomorrow it would cease to be the capital. — John

Yes, of course, social-institutional facts aren't the sorts of facts that physicists and other natural scientists are interested in. Facts of the former kind are, in a sense, human dependent in a way that facts of the latter kind aren't; although I would resist attempts to inflate this common sense distinction into a sharp dichotomy. If human beings were wiped off the surface of the Earth, most objects left behind would lose most of their socially instituted determinations (for instance, monetary tokens would cease to have determinate monetary values, dogs would cease to be pets, etc.), but they would retain their natural-scientific determinations. I am not denying or ignoring this distinction but it seems to have little relevance to the a priori/ a posteriori or the necessary/contingent distinction that are being discussed in this thread. Both of those distinctions cut across the institutional/natural divide that you are appealing to. Many natural facts can be known a priori and/or are necessary, and many social/institutional facts can be known a posteriori and/or are contingent. So I don't see how you can expect to ground the distinctions that are the topic of this thread on the social-institutional/natural divide.

I hold that Samuel Clemens is Mark Twain can only be true at a possible world that contains an object picked out by Samuel Clemens.

If you disagree, we have an impasse, but all I can say is I think you're wrong. — Mongrel

I don't disagree. I quite agree. You misread me. I denied that the evaluation of this sentence at possible worlds where Samuel Clemens doesn't exist is relevant to the determination of the modal status of this sentence. Kripke agrees, since he restricts his attention to possible worlds where the relevant objects exist in his demonstration of the necessity of identity. See the first four pages of his paper Identity and Necessity, or the first few pages of John P. Burgess's paper On a Derivation of the Necessity of Identity.

Unless you are a modal realist, then you can simply hold, as does Kripke, that X is necessarily identical to Y if and only if there is no possibility that X not to be identical to Y, where X and Y are the objects designated by "X" and by "Y", respectively, in the actual world (*). It's not relevant to the evaluation of this claim that X's or Y's existence be contingent, and hence that there might be possible worlds where they don't exist. Likewise, to repeat myself, the evaluation of the a posteriori necessity of water being composed of H2O is insensitive to the mere possibility that the world might have been devoid of water (and thereby, of course, also devoid of H2O).

(*) See Kripke's Naming and Necessity p.14: "I will say something briefly about 'possible worlds'. (I hope to elaborate elsewhere.) In the present monograph I argued against those misuses of the concept that regard possible worlds as something like distant planets, like our own surroundings but somehow existing in a different dimension, or that lead to spurious problems of 'transworld identification'. Further, if one wishes to avoid the Weltangst and philosophical confusions that many philosophers have associated with the 'worlds' terminology, I recommended that 'possible state (or history) of the world', or 'counterfactual situation' might be better. One should even remind oneself that the 'worlds' terminology can often be replaced by modal talk—'It is possible that . . .'"

I hold that in the case of any utterance, it will have to be sorted out somehow what it means. You can't just point to what you understand to be linguistic convention.

You apparently disagree with that as well. Again.. I think you're wrong. — Mongrel

I certainly don't disagree with this either. I am questioning the inferences that you are drawing from this. It is one thing to evaluate what is said by a speaker who makes use of a sentence, accounting for pragmatic considerations and contextual features of the utterance, and another to evaluate the modal status of the claim being made. You wish to make the latter rest entirely on the former, but questions of a posteriori necessity obviously outrun mere considerations of the utterer's intentions. If they would rest entirely on intention and/or convention then those modal claims would be a priori, stipulated by mere fiat.

I wasn't trying anything as ambitious as you are suggesting here. I was just pointing out that the knowledge that Paris is capital of France is known quasi-analytically insofar as it is so as a matter of definition or designation if you don't like the word 'definition' in this context.
And this just to show that the analytic/ synthetic divide is not as clear cut as it is sometimes made out to be.

And this just to show that the analytic/ synthetic divide is not as clear cut as it is sometimes made out to be. — John

True, but then, if the division isn't clear cut, this means that many socially instituted facts (e.g. the values of currencies relative to the gold standard, of the legal/political/administrative functions of cities and institutions) also have an empirical character. Also, if some entity has an institutional status (e.g. some city being the capital of a sovereign state) as a result of the collective will of the people (or of the King, etc.), then if the determination of this human will is itself contingent, then so is the inherited status and official function of the entity.

I don't disagree. I quite agree. You misread me. I denied that the evaluation of this sentence at possible worlds where Samuel Clemens doesn't exist is relevant to the determination of the modal status of this sentence. — Pierre-Normand

It's relevant, and I would say critical to grasping the concept of aposteriori necessary truths that we're talking about statements that are true over a limited number of possible worlds as opposed to true over all possible worlds.

I certainly don't disagree with this either. I am questioning the inferences that you are drawing from this. It is one thing to evaluate what is said by a speaker who makes use of a sentence, accounting for pragmatic considerations and contextual features of the utterance, and another to evaluate the modal status of the claim being made. You wish to make the latter rest entirely on the former, but questions of a posteriori necessity obviously outrun mere considerations of the utterer's intentions. If they would rest entirely on intention and/or convention then those modal claims would be a priori, stipulated by mere fiat. — Pierre-Normand

I understand the concern you're raising, but I think if you followed the line of thought, you'd see that it's not the problem you're envisioning it to be.

As it is, your position seems to leave you endorsing a contradiction. Intention matters when discerning meaning, but not when evaluating modal claims. That just seems crazy to me.

It's relevant, and I would say critical to grasping the concept of aposteriori necessary truths that we're talking about statements that are true over a limited number of possible worlds as opposed to true over all possible worlds. — Mongrel

This is a puzzling remark that you would have to explain. That something is necessary rather than contingent just means that it could not have been otherwise in any circumstance, which is exactly how Kripke urges that the claim that there is no "possible world" where it is false be understood. If there were (per impossibile) a possible world where Samuel Clemens isn't Mark Twain, then this would mean that they aren't necessarily identical, and hence, a fortiori, that the claim of identity is not a posteriori necessary. Likewise, if there were a possible world where water isn't composed of H2O, that would mean that water isn't necessarily composed of H2O, and hence, a fortiori, that this claim of essence isn't a posteriori necessary either.

As it is, your position seems to leave you endorsing a contradiction. Intention matters when discerning meaning, but not when evaluating modal claims. That just seems crazy to me. — Mongrel

That seems rather trivial to me. If someone purports to make some statement of a posteriori necessity, then, in a first step, you may indeed have to pay attention of the circumstances of her utterance, and her communicative intentions, etc., in order to understand what it is that she is claiming to be a posteriori necessary. Then, in a second step, by dint of the fact that her claim can only be known to be true a posteriori, you have to investigate what it is, in the world, that makes it true (e.g. investigate the nature of water, or seek out, by means of investigative journalism, if Clark Kent and Superman really are the same person, as she claims them to be.

In short, it's one thing to assess what it is someone means to say, and another to evaluate if it is true. In the special case where the claim can be known to be true a priori, then, maybe, those two steps collapse into the first one, but it is precisely a distinctive feature of the a posteriori necessary that they do not so collapse.

That something is necessary rather than contingent just means that it could not have been otherwise in any circumstance — Pierre-Normand

This is a confusing statement. Necessary modifies true. I guess it could modify false... that could be managed. You seem to be thinking of some.... thing? as being contingent or necessary. Some thing that could have been otherwise if it's contingent.

How about a statement or proposition can be necessarily or contingently true?

And if a statement is necessarily true, it's true in all possible worlds.

Aside from mountains of unresolved issues involving theory of meaning and reference, theory of truth, and theory of possibility... it's not brain surgery. :)

It's easy to imagine possible worlds where Clemens didn't exist. Clemens is Twain is not true at those possible worlds because it isn't truth apt.

To make a necessarily true statement, we need to say:

If Clemens exists, he is Twain.

That's actually necessarily true. (Assuming you approve of the concept of rigid designators and aposteriori necessity.. not all professional philosophers do.... obviously.)

That seems rather trivial to me. If someone purports to make some statement of a posteriori necessity, then, in a first step, you may indeed have to pay attention of the circumstances of her utterance, and her communicative intentions, etc., in order to understand what it is that she is claiming to be a posteriori necessary. Then, in a second step, by dint of the fact that her claim can only be known to be true a posteriori, you have to investigate what it is, in the world, that makes it true (e.g. investigate the nature of water, or seek out, by means of investigative journalism, if Clark Kent and Superman really are the same person, as she claims them to be. — Pierre-Normand

Having determined the meaning of a statement, one need not at any point abandon that meaning for some convention for the sake of predicating truth.

No philosopher I know of would disagree with that. Do you know of one?

Having determined the meaning of a statement, one need not at any point abandon that meaning for some convention for the sake of predicating truth.

No philosopher I know of would disagree with that. Do you know of one? — Mongrel

You thought I was advancing something crazy and I merely clarified what I meant. Now you seem to agree. But in that case, once it is established what de re necessity it is someone is purporting to express, you haven't shown how this claim being a posteriori necessary could depend in any way on the intentions of the speaker (beyond specifying what she means to say), or -- what has been centrally at issue between us -- how it could depend on some arbitrarily restricted range of possible worlds being single out for special consideration by the speaker (e.g. worlds in which France has Paris as its capital city). You've merely muddled the issue with the insistence the the items under consideration must exist in those possible worlds, which is trivially accepted by everyone (see my earlier Kripke and Burgess references) and has nothing to do with the speaker's particular intentions. Worlds in which water or Samuel Clemens don't exist just don't have any relevance to the evaluation of the modal status of the claims of a posteriori necessity that we've been considering.

This is a confusing statement. Necessary modifies true. I guess it could modify false... that could be managed. You seem to be thinking of some.... thing? as being contingent or necessary. Some thing that could have been otherwise if it's contingent. — Mongrel

Sorry, this was a shorthand. I means something being said; some proposition, or claim of de re necessity.

And if a statement is necessarily true, it's true in all possible worlds. — Mongrel

Yes. That's what I am claiming. And that's what you seem to have been denying consistently:

It's relevant, and I would say critical to grasping the concept of aposteriori necessary truths that we're talking about statements that are true over a limited number of possible worlds as opposed to true over all possible worlds.

You appear, strangely, not to have realized that a posteriori necessary statements are... necessary.

But in that case, once it is established what de re necessity it is someone is purporting to express, you haven't shown how this claim being a posteriori necessary could depend in any way on the intentions of the speaker (beyond specifying what she means to say), or -- what has been centrally at issue between us -- how it could depend on some arbitrarily restricted range of possible worlds being single out for special consideration by the speaker (e.g. worlds in which France has Paris as its capital city) — Pierre-Normand

Is there some reason a speaker could not say "France" and mean a country that has Paris as its capital?

This is the state of this conversation... I really have no earthly idea how you're likely to answer that. You keep trying to think three steps ahead of me instead of just logically evaluating the questions I asked. You say one thing and in the next post say "Of coarse I don't think that..."

I'm honestly finding it tiresome. Aren't you?

Yes. That's what I am claiming. And that's what you seem to have been denying consistently: — Pierre-Normand

Point to where I denied that.

Think about why necessity was once (and by some still is) thought to entirely overlap apriori and analytic. If you do that, you should get a hint as to why it matters if a statements shows up non-truth-apt in a possible world.

Is there some reason a speaker could not say "France" and mean a country that has Paris as its capital? — Mongrel

Whenever anyone speaks about France, one speaks about a country that indeed has Paris as its capital. It is generally understood that it doesn't have Paris as its capital essentially. That is, there are intelligible counterfactual circumstances in which France has (or would lave had) another city as its capital. It would still have been France. But you are rather meaning to say, it seems, that one could conjure up a new meaning for the word "France" such that the 'country' that answers to it has Paris as its capital essentially (which also requires an arbitrary modification of the ordinary concept of a country). I am unsure what the point of doing so would be. For whatever object X that has property P contingently, I can conjure up some concept X* such that X falls under X* and X* has property P essentially. This is a queer and artificial concept reminiscent of Quine's gavagai or Goodman's grue. It doesn't seem to have any bearing en the topic of a posteriori necessary truths, which are truths that we can investigate about, and are empirically falsifiable, rather than being modal truths that we stipulate by dint of arbitrary concept modification as you seemingly envision them to be. Stipulated truths aren't a posteriori.

Point to where I denied that. — Mongrel

I just did, immediately after the ":"

It's relevant, and I would say critical to grasping the concept of aposteriori necessary truths that we're talking about statements that are true over a limited number of possible worlds as opposed to true over all possible worlds.

(My bold)

"... over a limited number ... as opposed to all possible worlds" would appear to contradict "...in all possible worlds."

It is generally understood that it doesn't have Paris as its capital essentially. — Pierre-Normand

True. I don't believe there is anything that precludes a person talking about an object, France, for which it is essential (to that object) that its capital is Paris.

I'll tell you what. I'm going to write out the question and send it to a bunch of living professional philosophers... obviously with Scott Soames being one.

I'll get back to you with the answers (if any). If they all disagree with my assessment, then you'll have that gratification. If even one of them says nothing precludes the scenario, you will be obligated to at least try to understand what I'm saying instead of shoving every word I say back down my throat.

Deal?

"... over a limited number ... as opposed to all possible worlds" would appear to contradict "...in all possible worlds." — Pierre-Normand

It's only in the light of aposteriori necessity that we limit our assessment.

Traditionally, there was no limitation. There is no possible world where a bachelor is not an unmarried man.... you don't need the existence of bachelors in a world for that statement to be true (at that world. :)).