From owner-modality@LISTSERV.ARIZONA.EDU Mon Feb 8 21:40:06 1999 Date: Mon, 8 Feb 1999 22:28:53 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Angela J Burnette Subject: Re: Kripke's arguments for a priori contingencies To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Regarding Kripke's arguments for a priori contingencies and a posteriori necessities, I am assuming that the basis for the arguments is the fact that a prioricity is an epistemic notion while necessity is a metaphysical notion. Given this, there is the possibility that necessary truths be known only a posteriori and contigent truths (if true) can be known a priori. Identity statements such as Hesperus is Phosphorus are supposed to be examples of the former because, while the identity (which is between rigid designators?) is necessarily true, we could only find this out through empirical means. This sounds plausible, but I have a problem with Kripke's example of an a priori contingent truth... He argues that the length of S at time 0 is used to fix the referent of the rigid designator 'one metre' by way of definition, which is not to say that the MEANING of 'one metre' is 'the length of S at time 0'; so, even though we can know that 'the length of S at time 0 is one metre' a priori (because that's how we defined it?) this is a contingent truth because the length of S at time 0 could have been different (if it were heated or whatever). Now, 'one metre' is a rigid designator, while 'the length of S at time 0' is not; so the meaning of 'one meter' is not the same as the meaning of 'the length of S at time 0'. But, given this, what exactly is the meaning of 'one metre' if, as Kripke stipulates, the length of S at time 0 could be the only standard being used? In other words, given that there is no standard for 'one metre' besides 'the length of S at time 0' how can 'one metre' be a rigid designator? How could someone reapply the term without reference to the description which is only supposed to fix its reference, not provide its meaning? I don't see why this isn't a problem. If someone can tell me where I've gone wrong in describing Kripke's account, such that I'm asking this question, I'd appreciate it. angela From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 00:47:02 1999 Date: Wed, 10 Feb 1999 00:46:58 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Re: Kripke's arguments for a priori contingencies To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Angela writes: >Regarding Kripke's arguments for a priori contingencies and a posteriori >necessities, I am assuming that the basis for the arguments is the fact >that a prioricity is an epistemic notion while necessity is a metaphysical >notion. Given this, there is the possibility that necessary truths >be known only a posteriori and contigent truths (if true) can be known a >priori. Identity statements such as Hesperus is Phosphorus are supposed >to be examples of the former because, while the identity (which is between >rigid designators?) is necessarily true, we could only find this out >through empirical means. This sounds plausible, but I have a problem with >Kripke's example of an a priori contingent truth... > >He argues that the length of S at time 0 is used to fix the referent of >the rigid designator 'one metre' by way of definition, which is not to say >that the MEANING of 'one metre' is 'the length of S at time 0'; so, even >though we can know that 'the length of S at time 0 is one metre' a priori >(because that's how we defined it?) this is a contingent truth because the >length of S at time 0 could have been different (if it were heated or >whatever). Now, 'one metre' is a rigid designator, while 'the length of S >at time 0' is not; so the meaning of 'one meter' is not the same as the >meaning of 'the length of S at time 0'. But, given this, what exactly is >the meaning of 'one metre' if, as Kripke stipulates, the length of S at >time 0 could be the only standard being used? In other words, given that >there is no standard for 'one metre' besides 'the length of S at time 0' >how can 'one metre' be a rigid designator? How could someone reapply the >term without reference to the description which is only supposed to fix >its reference, not provide its meaning? I don't see why this isn't a >problem. If someone can tell me where I've gone wrong in describing >Kripke's account, such that I'm asking this question, I'd appreciate it. The crucial part of Kripke's account is that when a term is a rigid designator, once one has fixed its reference to a particular object (e.g. "Hesperus" refers to the planet Venus) or to a particular property (e.g. "hot" refers to a certain energy property) or to a certain length (e.g. "one meter" refers to length L, about 39.37 inches), one can then consider modal questions about what the term picks out in alternative possible worlds just by considering what is "that object" or "that property" or "that length" in other possible worlds, and by forgetting about the factors that were involved in fixing reference. So "Hesperus" picks out Venus in all worlds, irrespective of whether Venus is the evening star there. Similarly, one meter picks out length L in all worlds, irrespective whether the stick in Paris has that length there. The reference-fixer serves as a sort of bootstrap, or as a ladder we can kick away once it has done the job of giving the reference. The reference of "one meter" is fixed by looking at what is the length of stick S, but once we have accepted that this is a given length L, then we can consider what is length L in other worlds quite independently of what is the length of S there. And according to Kripke, the modal application conditions of "one meter" go with the former (length L), not with the latter (the length of S). As Kripke notes, it seems that we can quite coherently consider counterfactual possibilities in which the length of stick S was less that one meter (just say it broke, or that it had never grown that long in the first place); so the application of "one meter" to these possibilities can't merely involve the length of stick S. (Incidentally, if you accept that it's necessary that Hesperus is Venus, then presumably you don't think it's necessary that Hesperus is the evening star (as it certainly isn't necessary that Venus is the evenin star). But arguably it's a priori that Hesperus is the evening star (by analogy with the meter case). So you'll then be committed to something like a contingent a priori statement there, i.e. "Hesperus is the evening star".) (One difference between the meter case and the Hesperus case is that one might argue that there are deeper problems with reidentifying lengths across worlds than with reidentifying planets across worlds. What counts as length L in a given world? In reply, I think Kripke would make his suggestion here that we can "stipulate" lengths in our possibilities (e.g. considering the possibility that something is two meters long) rather than have to "find" the length through a distant telescope. But there are certainly some issues here.) What's the "meaning" of one meter? Kripke manages to stay away from that sort of question for the most part. Many people take the moral of Kripke's discussion to be that the "meaning" of the term is distinct from what "fixes reference" to the term. The reference of "one meter" is fixed as "whatever is the length of stick S"; but the "meaning" of the term involves its condition of application to possible worlds, which is different. The term "one meter" picks out length L in all worlds, so the "meaning" of the term involves that every length. Analogously, one might think of the reference-fixation of "Hesperus" as going via whatever is the evening star, but th "meaning" as involving the planet Venus. This is the sort of thing a "direct reference" theorist might say, for example, and there are a couple of passages in Kripke that suggest sympathy (e.g. where he distinguishes "fixing reference" from "giving the meaning"). For my part, I think both reference-fixing conditions and counterfactual application-conditions are in some sense part of the "meaning" of a term. I think of the former as the epistemic application-conditions (primary intension) and the latter as the subjunctive application-conditions (secondary intension). The former can be thought of as the a priori part of meaning; the latter as the (often) a posteriori part. In the 2-D framework, we'll say (roughly) that the primary intension os "one meter" picks out whatever is the length of a certain stick in all centered worlds, and that the secondary intension picks out a specific length L in all worlds. It doesn't matter much which of these we call the "meaning", but I think one can make a case that both are crucial. If one likes, one can think of the a priori conceptual analysis of "one meter" as something like "the actual length of stick S" or "dthat(the length of stick S)". Then we can make sense of most of the phenomena. E.g., we can make sense of Kripke's counterfactual possibility that stick S could be less than one meter by noting that stick S could have been shorter than its actual length (i.e., than its length in the actual world). i.e., there are worlds in which the secondary intension of "one meter" and the secondary intension of "the length of stick S" come apart. But it isn't epistemically possible that stick S is less than one meter, as the primary intensions of "one meter" and "the length of stick S" are the same. --Dave. From agillies@pop.u.arizona.edu Wed Jan 27 09:22:44 1999 Subject: initial discussion--names and intensions Date: Wed, 27 Jan 1999 10:35:36 -0700 x-sender: agillies@pop.u.arizona.edu From: Anthony S Gillies Status: RO Hi all, I was wondering about applying the two-dimensional framework to the arguments that Kripke gives in _N & N_, particularly those against the cluster concept view of names in the first lecture. Does it seem fair to say that the cluster concept view (a la Searle) is a theory about the pimary intensions of names, and Kripke's examples point to the secondary intensions of names? For example: the primary intension of 'Aristotle' can be cashed out as the guy who has some chunk of the properties: pupil of Plato, teacher of Alexander the Great, and so on. This function won't pick out the same guy in all possible worlds, and so "Aristotle was a pupil of Plato" will be contingently true (according to the 1-intensions). Kripke's examples, though, go like this: OK, imagine a counterfactual world where Aristotle was not a pupil of Plato (or any old property we think picks out Aristotle in this world). It is still a counterfactual situation about *Aristotle*. So, the name 'Aristotle' picks out the guy Aristotle in all possible worlds (according to the 2-intensions). So names are rigid designators. What this suggests, I think, is that assuming the two-dimensional picture of reference, Kripke's arguments about how names work is missing a premise: namely, that in natural language we mostly care about the 2-intension of names. So: (1) does this seem like an accurate description of the landscape when we apply two-dimensional semantics to _N & N_?; and (2) do we ever care about the primary intensions of names? --Thony P.S. I've BCC'ed this email to everyone in the class, that way people don't need to scroll through so many address headers at the beginning of every email. "Curious green ideas sleep furiously." From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 01:48:56 1999 Date: Wed, 10 Feb 1999 01:48:53 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Names and intensions To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Here's a reply to Thony's message from two weeks ago re Kripke. >From agillies@pop.u.arizona.edu Wed Jan 27 09:22:44 1999 >Subject: initial discussion--names and intensions >Date: Wed, 27 Jan 1999 10:35:36 -0700 >From: Anthony S Gillies > >Hi all, > >I was wondering about applying the two-dimensional framework to the >arguments that Kripke gives in _N & N_, particularly those against the >cluster concept view of names in the first lecture. Does it seem fair to >say that the cluster concept view (a la Searle) is a theory about the >pimary intensions of names, and Kripke's examples point to the secondary >intensions of names? > >For example: the primary intension of 'Aristotle' can be cashed out as >the guy who has some chunk of the properties: pupil of Plato, teacher of >Alexander the Great, and so on. This function won't pick out the same >guy in all possible worlds, and so "Aristotle was a pupil of Plato" will >be contingently true (according to the 1-intensions). > >Kripke's examples, though, go like this: OK, imagine a counterfactual >world where Aristotle was not a pupil of Plato (or any old property we >think picks out Aristotle in this world). It is still a counterfactual >situation about *Aristotle*. So, the name 'Aristotle' picks out the guy >Aristotle in all possible worlds (according to the 2-intensions). So >names are rigid designators. > >What this suggests, I think, is that assuming the two-dimensional picture >of reference, Kripke's arguments about how names work is missing a >premise: namely, that in natural language we mostly care about the >2-intension of names. So: (1) does this seem like an accurate >description of the landscape when we apply two-dimensional semantics to >_N & N_?; and (2) do we ever care about the primary intensions of names? First a bit of background, which may already be clear to some people, but which is very important and makes some of these issues clearer. There are actually two very distinct strands in Kripke's argument against the traditional view of names. The first, centered on Lecture 1, is what we can call the "modal argument". This makes the point that names are rigid designators, so that there counterfactual conditions of application are different from those of the various descriptions in the vicinity. E.g., for any description D, there's a counterfactual possibility that Aristotle wasn't D. The second, centered on Lecture 2, is what we can call the "epistemic argument". Here, he argues that its not even the case that the reference of names is fixed via description. This is where he brings in the cases of Feynman, Godel, Jonah, etc, and argues that for any description D in the vicinity, it is not a priori that Feynman (etc.) satisfies D. That's because it's always epistemically possible that Feynman didn't satisfy D; we could imagine finding that out. So not only can't we be descriptivists about modal application conditions, we can't even be descriptivists about reference-fixing. To see that the two points are quite distinct, note that the modal argument alone is compatible with the idea that names are "descriptive names": their reference is fixed via some description, but then they pick out the referent in question in all possible worlds. Some names, are arguably like this: e.g. "Jack the Ripper" arguably has its reference fixed to whoever committed the murders, in effect via description, but then picks out that very person (person X) in all possible worlds. Note that for a name such a "Jack the Ripper", one could run the modal argument, but not the epistemic argument: it's not necessary that Jack the Ripper committed the murders (there are worlds where someone stopped him), but it's a priori that Jack the Ripper (if he exists) committed those murders. But Kripke argues that most names are not like this. This is why (in Chapter 2) he argues that the descriptive theory of reference has to be supplanted by a causal theory (or picture) of reference. The modal point, in effect, is a point about secondary intensions. It concerns the application conditions of our concepts to counterfactual scenarios, given that the actual world is fixed. And we can express Kripke's insight by saying that the secondary intensions of names pick out the same object in all worlds. E.g. the secondary intension of "Jack the Ripper" picks out the same person in all worlds, irrespective of whether that person committed any murders there. The epistemic point, in effect, is a point about primary intensions. It concerns the application of our concepts to epistemic possibilities, and concerns matters of reference-fixation. Even though the secondary intension of "Jack the Ripper" goes with a specific person, not the description, the primary intension goes with the description: the primary intension of "Jack the Ripper" picks out (roughly) whoever committed the murders in question in a given centered world. But Kripke's point is that most names are not like this. We can express it by saying that the primary intensions of most names are not captured by any description. Rather, the primary intension involves picking out whatever is at the other end of a long causal chain. Kripke's epistemic argument is somewhat more controversial than his modal argument, and one might argue that there are certain resources available to a description theorist to resist his conclusion. E.g., one might invoke descriptions such as "the entity at the other end of an appropriate causal chain" (causal descriptivism), or "the person called such-and-such around here" (metalinguistic descriptivism). And arguably the 2-D framework can give support to a certain sort of attenuated descriptivism. But that's a complex issue (feel free to raise issues in the area). Kripke's epistemic point re names is widely but not universally accepted in contemporary philosophy; I think it's fair to say that the modal point is fairly close to being universally accepted these days (or as close to universal acceptance as anything gets). Now to Thony's points (some of which are in effect addressed already). Re the cluster theory: pre-Kripke, the modal and epistemic issues weren't carefully distinguished. It may be that in effect, Searle intended this as a theory of both, and in effect was concerned with both primary and secondary intensions. But I think it's arguable that it was most centrally intended as a theory of what fixes actual reference, and so was implicitly concerned with primary intensions. Certainly, the modal argument seems fairly decisive against it as a theory of secondary intensions; one might think it has more promise as a theory of primary intensions. But Kripke in lecture 2 is concerned to argue against it even as a theory of primary intensions (via the epistemic arguments). Someone like Searle might resist by accepting the modal point but not the epistemic point, and arguing that one can still give some sort of modified cluster theory at least as a theory of reference-fixation, if not as a theory of counterfactual application. That sort of debate is to some extent still going today. But anyway, I don't think it's quite right to say that Kripke isn't concerned with the primary intensions of names, given all the discussion in lecture 2. And I think that one could argue that the "missing premise" of his argument against description theories is supplied by the argument in lecture 2. It's true, though, that the modal discussion in lecture 1, which is arguably the central point, is largely concerned with secondary intensions. I think you're right that in some sense, Kripke thinks we are most concerned with secondary intensions of names rather than primary intensions. As noted before, he seems to implicitly see SIs as more closely tied to "meaning", whereas the "reference fixation" involved in PIs isn't really a matter of meaning. Of course matters are complicated by the fact that Kripke doesn't put things in terms of the 2-D framework at all (I'm told he isn't all that keen on it). Still, it's easy to translate much of his discussion into those terms. One further reason for his preference for SIs over PIs may be what we might call the "semantic argument". Arguably, different users of the same name can have quite different primary intensions for it, though the same secondary intensions. E.g., maybe my PI for "Bill Clinton" is different from Hillary's PI for "Bill Clinton", though we have the same SI. And more generally, maybe the PI of a name can vary throughout a community, whereas the SI is constant. If so, and if one adds the premise that the meaning of a term should be something that is shared between users of that term, one can see the case for thinking of the SI, not the PI, as part of a term's "meaning". I think this is one of the reasons why contemporary philosophers of language tend to focus more on names' SIs, not PIs. (Another reason may be that some are skeptical of the very idea of a PI, or haven't thought much about it.) For my part, I think that even if one accepts that the PI of a term may not be the same for all users, there's still a sense in which it is part of the meaning of a term at least for a specific user (and arguably, it corresponds much more closely than the SI to what we can think of as the "cognitive content" of the term for the user). Or one can distinguish between "sentence meaning" and "utterance meaning" (as Marga Reimer does in some of her work) and argue that PIs are at least part of the latter, perhaps. But again, what truly gets to count as "meaning" is to some extent a terminological issue. --Dave. From owner-modality@LISTSERV.ARIZONA.EDU Sun Jan 31 23:18:29 1999 Date: Mon, 1 Feb 1999 00:07:19 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Timothy J Bayne Subject: Comment on Thony; thoughts on primary intensions To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Some thoughts. . .first, a comment on Thony's message, then a thought or two on primary intensions. It's pretty long - sorry. (1) Some comments about Thony's point re the 2 level framework and Kripke's ruminations on proper names. (a) First, an exegetical point. I find it very difficult to nail down the distinction between Searle and Kripke. When you read them very closely (i.e. footnotes and all), you see that Kripke allows for description to play a role in fixing reference, and Searle allows for causal chains to play a role in transmitting reference, and it becomes harder to see where they really disagree. (b) Assuming that a 2 level approach to proper names might work, Thony asks whether we ever care about the primary intension of proper names (i.e. the Searlian descriptive part). Note that Chalmers seems to think that we care about primary intentions *more* than secondary intensions - they are 'cognitive content' of thought (p. 65). I'm tempted to say that our use of a proper name (drawing inferences from it, etc.) involves its primary intension - and thus, at the very least, we do care about the primary intension a lot. But there may be another sense in which we also care about secondary intensions (i.e. the referent itself). (c) A disanalogy between natural kind terms and proper names is that the primary intension of the former is presumably shared by all members of the linguistic community who have the term (even if only by relations of deference, e.g. "elm"), while the descriptive content that I associate with, say, "Bill Clinton" may not be that which others, say ex-interns, associate with that same proper name. I'm not sure whether or not this difference is a difference that makes a difference. (2) (a) Chalmers says that primary intensions may not always be easy to uncover, but they are in principle open to reflection (to the degree that they are determinate). Here's a couple of cases that people have wildly different intuitions about and reflection really doesn't seem to help all that much. (i) I could have been a builder; (ii) I could have been born a century earlier than I was. (iii) I could have been a poached egg. (That last one is meant to be read literally - and it is in fact false in case you are wondering). (iv) I could have been the number 2. Some think that only (i) is true. Some think that (i) and (ii) alone are true. Some think that only (i), (ii) and (iii) are true. Perhaps you think that all four are true. (In which case you are a poached egg.) Is this just a particularly difficult case for the notion of the in principle transparency of primary intension, or does it suggest deeper and more pervasive problems with it? This example also raises another question. Some want to say that it shows that some of my properties are essential to me (perhaps, say, being human), while other properties are not. How do we cash out the difference between these two types of properties in the 2D framework? Perhaps it goes like this. For any possible world P and any property x, if I have x essentially, then it is logically true that I have x in every world in which I exist, and if I have y only contingently, then it is logically true that there are some world in which I exist and have y, and other worlds in which I exist and don't have y, and which worlds are which depends on the natural laws and boundary conditions therein. In this way, perhaps, we can cash out the difference between essential and non-essential properties, without bringing in metaphysical necessity. (b) Chalmers responds to the Quinean point about revisability by saying that his arguments are not affected, because they turn on supervenience conditionals (p. 55). I'm not sure why the fact that the propositions in question are supervenient conditionals should ipso facto bestow on them a certain kind of immunity to revision. First, note that our judgments about laws of nature (i.e. conditionals) are revisable. Second, note that we can distinguish two types of supervenience conditionals. First, there are those conditionals that state that B facts merely are supervenient on some other range of facts, without specifying which range of facts forms the supervenience base. (Actually, this is not really a supervenience conditional, but rather the claim that there is such a conditional.) Second, there are those conditionals which go on to say that B facts are supervenient on A facts. Now, presumably we can be wrong about the latter type of supervenience conditional. It is harder to see how we could be wrong in thinking that a certain type of fact is supervenient. We might go wrong in thinking that, say, aesthetic facts supervene on F facts, but could we err in thinking that aesthetic facts are supervenient in the first place? Chalmers, I take it, claims that we can't go wrong in such have an argument here, but it seems to me that we could mistakenly assume that a certain type of fact is supervenient although it is fact not, and vice-versa (its the vice-versa that DC needs, I think). (c) I think that Chalmers wants to tie in the claim about supervenience conditionals being unrevisable, with the claim that we have transparent access to the primary intensions of our concepts, but I'm not sure that I see the connection. It seems to me that primary intensions are something like application conditions - this is what "watery stuff" seems to come to. But need application conditions involve a conception of whether or not a property is supervenient or not? I don't see it. One's general conception of what is or is not supervenient, and what may be supervenient on what, seems to be a fairly sophisticated thing. But perhaps that's okay. Perhaps all Chalmers needs to say is that if you have the (note: not "my") concept of consciousness, then you have the ability to reflect on it, and to see that it could not be logically supervenient on any other types of properties. So: one can have a perfectly good concept of consciousness without having grasp of the supervenience conditionals that is contained within it and are putatively immune to Quine's worries. Is this the connection? Sorry to be so long-winded. Tim Timothy J. Bayne RM. 213 Social Science Department of Philosophy University of Arizona Tucson, AZ 85721 USA Hm ph. (520) 298 1930 From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 02:37:53 1999 Date: Wed, 10 Feb 1999 02:37:47 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Re: Comment on Thony; thoughts on primary intensions To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Some thoughts on Tim's old message. >Date: Mon, 1 Feb 1999 00:07:19 -0700 >From: Timothy J Bayne >Subject: Comment on Thony; thoughts on primary intensions >To: MODALITY@LISTSERV.ARIZONA.EDU > >Some thoughts. . .first, a comment on Thony's message, then a thought or >two on primary intensions. It's pretty long - sorry. > >(1) Some comments about Thony's point re the 2 level framework and >Kripke's ruminations on proper names. (a) First, an exegetical point. I >find it very difficult to nail down the distinction between Searle and >Kripke. When you read them very closely (i.e. footnotes and all), you see >that Kripke allows for description to play a role in fixing reference, and >Searle allows for causal chains to play a role in transmitting reference, >and it becomes harder to see where they really disagree. (b) Assuming that >a 2 level approach to proper names might work, Thony asks whether we ever >care about the primary intension of proper names (i.e. the Searlian >descriptive part). Note that Chalmers seems to think that we care about >primary intentions *more* than secondary intensions - they are 'cognitive >content' of thought (p. 65). I'm tempted to say that our use of a proper >name (drawing inferences from it, etc.) involves its primary intension - >and thus, at the very least, we do care about the primary intension a lot. >But there may be another sense in which we also care about secondary >intensions (i.e. the referent itself). (c) A disanalogy between natural >kind terms and proper names is that the primary intension of the former is >presumably shared by all members of the linguistic community who have the >term (even if only by relations of deference, e.g. "elm"), while the >descriptive content that I associate with, say, "Bill Clinton" may not be >that which others, say ex-interns, associate with that same proper name. >I'm not sure whether or not this difference is a difference that makes a >difference. Re (a), I think you're right that Searle's and Kripke's views of reference-fixing (i.e. primary intensions) are closer than they may seem. I think the crucial difference at that while Kripke allows descriptions a role, you can't do everything with descriptions, whereas Searle thinks you can. Of course, as you note, some of Searle's descriptions may involve "the thing at the other end of a causal chain", or something like that. This leads to what is sometimes called "causal descriptivism" (a sort of hybrid); but I think it's a view that is more congenial to the descriptivist in some ways (after all, why should they object to having some descriptions involve causation). Kripke's response (e.g. around p. 162, and footnote 38) is to suggest that this is a pretty trivial sort of descriptivism (incorporating the whole causal theory as a part!), but the issue isn't entirely clear. Certainly someone like David Lewis is happy to take this on board as enough descriptivism for the descriptivist's central purposes (though it may be somewhat unlike a traditional descritivism). Re (b), yes, my own view is that primary intensions are in many ways more important than secondary intensions. I think they are more central to thought, to rational inference, and to capturing the way that language describes the world. On my view secondary intensions are mostly crucial to making sense of counterfactual thought and language; that's an important role but somewhat in the background. (Just why contemporary philosophy has then focused so much on SIs becomes something of a question; we'll talk about that a bit in the upcoming discussion of "the tyranny of the subjunctive"). Of course reference is important, but I think it is important qua reference, not qua its role in constituting an SI (to see this, note that reference is just as important for descriptions, like "the inventor of the zip", even though in these cases the SI doesn't involve the actual referent). But all this is just my view and of course is arguable. Re (c), I discussed this in the reply to Thony. I think it is plausible that PIs of names, but not SIs, vary in a community. It's actually not entirely obvious whether or not the same goes for natural kind terms. I think someone might argue that different people could have different PIs for "water" or "gold" or whatever. For example, maybe there could be an isolated chemist who only associated "water" with the "H2O" description, and doesn't know about its appearance. His PI would be different from ours, but arguably he would still be using the term "water". Someone might suggest in reply that he doesn't really have our word, as the "watery" mode of presentation is somehow essential to the word; the resolution of that debate isn't entirely clear. But in any case it seems that at least mild variations in PIs may be possible even in the natural kind case. >(2) (a) Chalmers says that primary intensions may not always be easy to >uncover, but they are in principle open to reflection (to the degree that >they are determinate). Here's a couple of cases that people have wildly >different intuitions about and reflection really doesn't seem to help all >that much. (i) I could have been a builder; (ii) I could have been born a >century earlier than I was. (iii) I could have been a poached egg. (That >last one is meant to be read literally - and it is in fact false in case >you are wondering). (iv) I could have been the number 2. Some think that >only (i) is true. Some think that (i) and (ii) alone are true. Some think >that only (i), (ii) and (iii) are true. Perhaps you think that all four >are true. (In which case you are a poached egg.) Is this just a >particularly difficult case for the notion of the in principle >transparency of primary intension, or does it suggest deeper and more >pervasive problems with it? Hmm, I think you're concerned here not with primary intensions but with secondary intensions. You're considering subjunctive counterfactual possibilities about what could have been; i.e., the sort of modal possibilities Kripke is concerned with in N&N (especially lecture 1). That is, I take it that people who are concerned with these questions are worrying about the same sort of modality as in Kripke's questions about whether I could have descended from a different sperm and egg (given that I descended from these ones). These are matters of secondary intension. Personally, I don't have really strong intuitions about what I (DC) could and couldn't have been in various counterfactual possibilities. Maybe I could have been born earlier; probably not a poached egg or the number 2, but I'm not certain. This goes along with the idea that the secondary intensions of many of our concepts are not all that determinate; on my view, it is often almost a terminological issue which way to go. (There are even times when I'm tempted by the contra-Kripkean thought that water could have been XYZ.) Of course other claim to have much more determinate intuitions about these things, and even when they don't, they think that at least the SI is fairly determinate. (Of course it's not required that we have a priori access to SIs. But one does hope that one has a priori access to SIs conditional on knowing the nonmodal facts (e.g. about actual reference). So insofar as SIs still seem a bit loose even given the nonmodal facts, it ought to be the case that either (a) one hasn't analyzed the case well enough, or (b) the SI is in fact indeterminate, or (c) disagreeing people actually have slightly different concepts.) As to primary intensions, I think these are generally a lot more determinate and accessible, though of course even these can be fuzzy around the edges. When one is considering not subjunctive possibilities but epistemic possibilities about the way the world is, I think our terms often seem to have a pretty determinate application. There may be variation still in borderline cases and in very way-out cases, but there's at least a fairly determinate core; and much of the variation in wild cases can be dismissed as somewhat terminological and so not really important metaphysically. Again, I'll want to make the claim of a priori access, and argue that disagreement involves either (a) inadequate reflection, (b) indeterminacy in PI, or (c) terminological differences. >This example also raises another question. Some want to say that it shows >that some of my properties are essential to me (perhaps, say, being >human), while other properties are not. How do we cash out the difference >between these two types of properties in the 2D framework? Perhaps it goes >like this. For any possible world P and any property x, if I have x >essentially, then it is logically true that I have x in every world in >which I exist, and if I have y only contingently, then it is logically >true that there are some world in which I exist and have y, and other >worlds in which I exist and don't have y, and which worlds are which >depends on the natural laws and boundary conditions therein. In this way, >perhaps, we can cash out the difference between essential and >non-essential properties, without bringing in metaphysical necessity. I think the difference between essential and inessential properties will show up in the SI of a term. We'll say that Fred has property P essentially if "Fred is P" is necessary, i.e. 2-necessary (necessary in the subjunctive sense that Kripke considers). This will hold if the SI of "Fred" picks out a being with P in all worlds. Fred has P inessentially if the SI of "Fred" picks out a being without P in some worlds. In a certain sense, this is cashing out the difference between essential and inessential properties without bringing in any really heavy notion of metaphysical necessity. One still needs the basic space of worlds, but these can just be the conceivable worlds, more or less. Once one has worlds plus SIs, one has enough to understand the distinction. To get SIs, one needs (a) a good enough a priori grasp of the concept in question, and (b) nonmodal facts about the actual world. (E.g., say water is essentially H2O. Then if one has a full a priori competence with the term "water" and knows the nonmodal facts about the composition of actual water being H2O, one can know that the SI picks out H2O in all worlds. Etc.) Indeed, it seems pretty clear that Kripke's methodology for thinking about what's essential is more or less like this. The only empirical knowledge he needs are some nonmodal facts, e.g. that he's descended from sperm S and egg E, or that water is made of H2O, and so on. The rest seems to follow by a priori analysis: arguably, by the method of conceivability plus a priori analysis of the way our concepts apply to worlds. If this works, we can cash out essentialism relying only on (a) an a priori accessible space of worlds, (b) a full enough a priori grasp of the concepts in question, and (c) some nonmodal facts about the actual world. Sounds promising for modal rationalism! More on Tim's other points later. --Dave. From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 08:54:33 1999 x-sender: agillies@pop.u.arizona.edu Date: Wed, 10 Feb 1999 10:07:53 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Anthony S Gillies Subject: twins To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Dave recently wrote: >In the subjunctive case, we're considering the standard sort of >possibility that Kripke and others discuss. Here, I can imagine >someone arguing that it is essential to sentient human beings (like >me) that they be at least potentially sentient, so that the hypothesis >that I could be a zombie is like the hypothesis (discussed by Kripke) >that I could have descended from a different sperm and egg, etc. If >so, the 2-intension of "me" and "DC", etc, can only pick out >potentially conscious beings, and it's (secondarily) impossible that I >be a zombie; counterfactual possibilities concerning a zombie should >not be described as possibilities concerning me. (I say "potentially >conscious" as it's quite possible that I be temporarily unconscious, >and perhaps that I died before becoming conscious.) I'm not sure >whether this essentialist claim is right, but it has a certain >possibility. This is what I had in mind re my zombie twin. The worry was whether a materialist might be able to force the zombie thought experiment to be run along lines which would require using one's own proper name. If so, there are two options. (1) If we evaluate the scenario under PI's, then consciousness is (or arguably is) included in our PI of ourselves. (2) If we evaluate the scenario under SI's, then the name picks out the same thing in all possible worlds--and it's secondarily impossible that I could be a zombie. Either case looks problematic. Now, I don't know if any materialist would *want* (or bother) to try to come up with a way of forcing the thought experiement to make essential use of one's own proper name, but that's the sort of situation I had in mind. Thony "Curious green ideas sleep furiously." From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 17:09:21 1999 Date: Wed, 10 Feb 1999 17:09:18 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Re: twins To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Thony writes: >This is what I had in mind re my zombie twin. The worry was whether a >materialist might be able to force the zombie thought experiment to be >run along lines which would require using one's own proper name. If so, >there are two options. (1) If we evaluate the scenario under PI's, then >consciousness is (or arguably is) included in our PI of ourselves. (2) >If we evaluate the scenario under SI's, then the name picks out the same >thing in all possible worlds--and it's secondarily impossible that I >could be a zombie. Either case looks problematic. Now, I don't know if >any materialist would *want* (or bother) to try to come up with a way of >forcing the thought experiement to make essential use of one's own proper >name, but that's the sort of situation I had in mind. I'm not sure just what you have in mind by "forcing the thought-experiment to use one's proper name", but it seems to me that that couldn't really be done, since proper names are strictly speaking irrelevant to the argument. The anti-materialist needs to argue that "P -> Q" is not a priori and not necessary, where P is the complete microphysical tuth about the world and Q is a phenomenal truth. They can make that case quite independently of considerations about proper names. Maybe a materialist could try to argue that "P' -> Q" is a priori or necessary, where P' is some variant on P that includes proper names, but even if that conclusion were established, it wouldn't have much bearing on the question of wheter materialism is true or false. At best, what would be going on is the same as what might go on in the case where P'' includes facts about love (say), where we take the view that love requires consciousness. Then P'' will entail some facts about consciousness, but that implies nothing at all about materialism. As long as P doesn't entail Q, then materialism is false. The entailment from P'' just tells us that materialism is false about love, too, as one would expect. Similarly, if P doesn't entail Q but P' does entail Q, where P' includes facts about DC, all that will follow is that materialism is false about DC. Which is just what one would expect if being DC essentially involves being conscious. --Dave. From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 10:45:57 1999 Date: Wed, 10 Feb 1999 11:45:14 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Timothy J Bayne To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Folks, What follows are some impressionistic "minutes" regarding today's meeting, compiled by Tim, Thony and Brad, and loosely stiched together. . Our discussion seemed to focus on two issues, broadly conceived. (1) What are primary intensions? We spent a lot of time trying to work out whether proper names, such as London, have primary intensions, and if so, >> what are they. There was some discussion of whether 'London' and 'Londres' have different primary intensions, if they have primary intensions at all. > > > >> Thony argued that primary > >> intensions aren't epistemic in this sense, they are simply functions from > >> possible worlds to individuals or properties. There seems to be some > >> tension between the cognitive/epistemic role primary intensions are meant > >> to play, and their metaphysical role. > > It sounds odd to say that a PI is a function from worlds to > referents, and then say that different people can have different primary > intensions for the same terms. `Water' is supposed to pick out the > watery stuff in any world in which it is uttered, quite irrespective (I > take it) of what folks know about water (though, presumably, most folks > would know which stuff was the watery stuff--but this needn't be the > case). > > > > >Perhaps related to this, there was some argument about whether even natural kind terms (our example was "water") have primary intensions. Brad claimed that it made sense to say "Water might not have been H2O", but Thony argued that this only made sense to him when the modality was construed epistemically rather than metaphysically. I (Brad) was claiming that it was true on a metaphysical reading. If it were true, that would motivate the notion that "water" has a primary intension (since it is certainly *not* true that H20 might not have been H20). > > > > I guess the question that I (Thony) wanted to raise (though not at all clearly) was > whether the semantically salient thing that names (natural kind terms, > etc.) contribute to sentences/discourses in which they occur is anything > like a primary intension. I'm (Thony) not at all sure, but my intuition seems to > favor SI's in this respect. > > > > >> > >> > >> This got us on to a discussion of whether it is words that are the > >> (primary) bearers of primary intensions, or concepts. I (Tim) think we concluded > >> that both words and concepts are meant to have primary intensions. There > >> was some discussion of whether a word has a single > >> primary intension associated with it. I (Tim) suggested that it is not obvious > >> that this is the case: Imagine that I know that ice, liquid water and > >> steam are the same stuff. I call all three kinds of stufff "water" and its > >> part of my primary intension that all three kinds of stuff are the same > >> stuff. Imagine that my kids are only exposed to liquid water, and I don't > >> teach them that water can boil or freeze. Two questions: (A) When I talk > >> with my kids > >> about water (liquid water), are we using the same word (I guess I want to > >> say yes, tentatively); (B) do we attach the same primary intension to it > >> (I want to say no, tentatively). > >> > >> {It also seems hard to draw a hard and fast distinction between primary > >> and > >> secondary intensions, given the way kids learn words/concepts. If content > >> really has this dual structure, wouldn't one expect it to be reflected > >> more clearly?} > >> > >> But most of our discussion on this point concerned whether "London" has a > >> primary intension, and if so, what is it? We also spent some time talking > >> about the primary intentions of proper names, like 'Brad", and indexicals, > >> like "me". > >> > >> (2) Next, we talked about the debate that Thony and I (Tim) have had regarding > >> the issue of trans-world individuation: does Thony have to be able to pick > >> out his zombie > >> twin, rather than just a zombie who is physical/functional type identical > >> to him, to get the zombie argument going. This got us on to the topic of > >> why it is important for the > >> zombie argument that I know that I am conscious. Here is how I see the > >> argument as running: > >> > >> (A) I know I am conscious > >> (B) We assume that consciousness is logically supervenient (for reductio). > >> Thus, > >> (C) Any being that is physically-functionally type identical to me will, > >> in any possible world, also be conscious. > >> (D) I can conceive of an individual that is physically-functionally type > >> identical to me, but is not conscious. > >> [Such an individual is my zombie "twin" in the sense that we are type > >> identical, but note that it need not be > >> *me* in any deep sense.] > >> (E) Conceivability entails logical possibility. > >> (F) from D and E, it follows that there is a world in which there is an > >> individual that is physically-functionally type identical to me, but lacks > >> consciousness. > >> > >> But (F) is inconsistent with (C), so (B) is false. > >> > >> I think you can run the zombie argument without making any claims about > >> the trans-world identity of individuals. **Although** you might have to > >> make commitments to the trans-world identity of properties, such as the > >> property of consciousness, and this might be problematic. > >> > Thony: This formulation is good, and is enough to establish that either (B) or > (A) is false. Now, my (Thony) thinking yesterday (and perhaps today) was that in > order to avoid someone concluding that (A) is false (and (B) true), > imagining one's zombie twin becomes an exercise in considering > counterfactual possibilities--and these require the 2-intensions. > Further, in these subjunctive contexts, it looks like it is necessary > that I am conscious. > > My main concern was not so much how/whether a materialist might be able > to push the zombie argument into involving crucial use of *my zombie > twin*. But, if one could make such a move, it seems that we get some > strange consequences about conceivability and possibility. > > > > "Curious green ideas sleep furiously." > > Timothy J. Bayne RM. 213 Social Science Department of Philosophy University of Arizona Tucson, AZ 85721 USA Hm ph. (520) 298 1930 From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 19:07:58 1999 Date: Wed, 10 Feb 1999 19:07:54 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Do names have primary intensions To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Some thoughts on the "minutes". >Our discussion seemed to focus on two issues, broadly conceived. >(1) What are primary intensions? We spent a lot of time trying to >work out whether proper names, such as London, have primary >intensions, and if so, >> what are they. There was some discussion of >whether 'London' and 'Londres' have different primary intensions, if >they have primary intensions at all. Actually, I discuss the 'London'/'Londres' case a bit in "The Components of Content". Everyone should read that paper, since it gives a lot more detail on the 2-D framework, though it is primarily cast in terms of concepts, not words. Do proper names have primary intensions? As I said yesterday, they arguably don't have PIs that are universal to all users of a name. It may be that two different users have different PIs for "London", just as for "Bill Clinton", etc. So this might lead one to say that names don't have PIs as a matter of "semantics", where the semantics of a term is supposed to be universal to all users of the term. Still, I think that it is plausible at least that any name has a PI for any user on any occasion of use. We might put this by saying that "name tokens" have PIs (or are associated with PIs), even though "name types" may not. If so, we can arguably see PIs at least as part of the semantics of an "utterance" (cf. Reimer's distinction), and certainly as part of the semantics of an underlying thought. Why think that name tokens have primary intensions? Basically, because like any referring word, a name (token) must have some pattern of application across epistemic possibilities. Just say I use the name "Godel". Then I can consider all sorts of epistemic possibilities about the way the actual world might be, and I can see how the name will refer with each of them. Think of Kripke's own methodology in Lecture 2 -- we consider e.g. the epistemic possibility that the incompleteness of arithmetic was proved by a guy called "Schmidt" and stolen by a guy called "Godel" who published it, moved to Princeton, etc. Kripke notes, in effect, that if that epistemic possibility is actual, our term "Godel" refers to the Princeton guy. That's to say that the primary intension of "Godel" picks out the Princeton guy in the relevant centered world. One can do the same thing for a very wide range of worlds considered as actual (i.e. considered as epistemic possibilities): for a large numbers of such worlds, there seem to be clear facts about what our words will pick out if those worlds are actual. We can think of the primary intension as the "reference-fixing" conditions of our terms, if we like, though actually I prefer to think of it as giving the epistemic application-conditions of our terms, or the "epistemic profile" (as opposed to the "modal profile" of the secondary intension). It's central to the way language works that any referring term has an epistemic profile (at least for a given user on a given occasion): there are facts about how it applies to different epistemic possibilities, and about how it will refer if those possibilities turn out to be actual. How does one evaluate the primary intensions of names that one uses? Basically, take a name, e.g. "London", take a given centered world, and ask oneself "to what does the name refer if that world is actual"? Here, one considers the world as actual, i.e. considers it as an epistemic possibility: "what if the world actually turns out to be that way?". And for a very wide range of such worlds, we have clear intuitions about how the name refers. E.g., if the actual world turns out to contain XYZ in the oceans etc, then we'll say that "water" refers to XYZ. If the actual world turns out to have the Princeton guy stealing the proof from Schmidt, we'll say that "Godel" refers to the Princeton guy. If the actual world turns out to have a non-whale-eaten guy at the other end of a causal chain from our use of "Jonah", we'll say that "Jonah" refers to that guy. Etc, etc. Summing up a primary intension in langauge is often difficult, and the same goes for these cases; what really matters is the function from worlds to referents, not any capsule summary. But if I were to try to get at some of the things that are involved e.g. in the PI of my name "Godel", we might try: "the guy called 'Godel' who's at the other end of a causal chain from my use of the name", or something like that. That's imperfect, as Kripke argues -- e.g. one can consider epistemic possibilities in which the guy my term refers to wasn't called "Godel" at all (I've got his name wrong, or some such). But that just means that we have to refine our view of the PI. Importantly, Kripke's own methodology here relies precisely on evaluating how the term will refer if a given epistemic possibility turns out to be actual, i.e., on evaluating the PI of the term at a world. As for "London" and "Londres": think of Pierre's situation. For him, there are lots of epistemic possibilities. The world he thinks he is in is one with a beautiful faraway city by the name of "Londres", and an ugly one close at hand called "London". Pierre quite reasonably says that if that epistemic possibility is actual (as he believes it is), then his terms "London" and "Londres" name two different cities: "London" picks out the faraway city and "Londres" the city close at hand. So in that centered world, the PIs of "London" and "Londres" give different results. On the other hand, another epistemic possibility for Pierre (at least a broad epistemic possibility in the sense articulated earlier) is the actual possibility: i.e., that the people he got the term "London" from were actually referring to the same city he's living in now, and that it has both beautiful and ugly parts, etc. If Pierre were confronted with the hypothesis that this epistemic possibility is actual, he should rationally conclude that under that hypothesis, his terms "London" and "Londres" pick out the same city. So that's to say that in this centered world (Pierre's actual world!), the PIs of "London" and "Londres" give the same results. So, Pierre's "London" PI and his "Londres" PI give the same result on some centered worlds (including the actual world), but different results on other worlds (e.g. the one he thinks is the case). So they are at least slightly different intensions overall. How to summarize these intensions in language? Again, it's difficult and imperfect, but we might make a first attempt by saying that his "London" PI picks out "the beautiful city I've heard of under the name 'London'", and his "Londres" PI picks out "the ugly city I'm living in", or perhaps "the city I've heard of under the name 'Londres'". Getting the details right will depend on careful consideration of cases, and will depend on just how Pierre's conceptual system is set up, etc. But this gives us enough to see how the PIs might pick out the same extension in one world (the actual world) but different extensions in some other worlds. The fact that Pierre's PIs are different precisely reflects the fact that it is epistemically possible for him that London isn't Londres. This suggests a general principle: when "A=B" is a posteriori (for a user), A and B have different PIs. To see this, one can go through the following reasoning. (1) "A=B" is a posteriori; so (2) it is (broadly) epistemically possible that A is not B, so (3) there is some epistemic possibility in which my terms "A" and "B" pick out different things (when that possibility is considered as actual) so (4) there is some world in which the PI of "A" and the PI of "B" yield different extensions so (5) A and B have different PIs. I hope all the steps here make sense. You can illustrate it by thinking of the London/Londres case, or the Hesperus/Phosphorus case. Note that the principle here is equivalent to: when "A=B" is a posteriori, "A=B" has a contingent primary intension. (N.B. the primary intension of a statement is just the obvious generalization of the primary intension of a term -- in the book I call this a "primary proposition", but I now tend to use the unified terminology.) If A and B have different PIs, there's a centered world where they pick out different extensions, and in which the PI of "A=B" is false. And to say that "A=B" has a contingent PI is just to say that there's a world in which the PI is false. This is an instance of a very important general principle: if a statement S is a posteriori, S has a contingent primary intension. It's fair to say that this is at the heart of the two-dimensional account of a posteriori necessity. The central claim of this account is if a statement S is necessary a posteriori, S has a contingent primary intension and a necessary secondary intension. That's not too hard to illustrate by working through the standard Kripke cases. I note that there are just possibly some counterexamples to this principle which someone might put forward, though they probably won't be the standard Kripke cases. We'll be discussing potential counterexamples later in the course, but feel free to suggest any now. But anyway, one can see how all this hooks in with the need for names to have PIs. Presumably a lot of statements involving names will be a posteriori for many or all users, e.g. "London = Londres", "Cicero = Tully", etc. That's just to say that for a given user, there is the epistemic possibility that the statement is false, i.e., there are some scenarios such that if they turn out to be actual, the statement turns out false. So the very a posteriority of the statement indicates that the names in question have distinctive patterns of applications to epistemic possibilities, i.e. that they have (distinct) PIs. That's enough for one message. Hopefully it gives useful background for thinking about issues in the rest of the minutes. More on the rest later. --Dave. From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 20:47:07 1999 Date: Wed, 10 Feb 1999 20:47:03 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO More on the minutes. >> >> Thony argued that primary >> >> intensions aren't epistemic in this sense, they are simply functions from >> >> possible worlds to individuals or properties. There seems to be some >> >> tension between the cognitive/epistemic role primary intensions are meant >> >> to play, and their metaphysical role. Hmm, I'm not sure exactly what the tension is here. Say more? On my own view, primary intensions are in a sense "epistemic", in that they involve the way our terms apply to worlds considered as epistemic possibities, but at the same time they are perfectly respectable functions for worlds to extensions. And the worlds in question may be perfectly respectable metaphysical possibilities at the same time. E.g., the XYZ-world, with XYZ in the oceans etc, is presumably a perfectly respectable world. Considering the world as counterfactual, our term "water" picks out H2O in that world; considering the world as actual (as an epistemic possibility), "water" picks out XYZ. At least, nothing in Kripke's discussion suggests that the worlds we are considering as epistemic possibilities can't all by metaphysically possible worlds in their own right. That's the centerpiece of the 2-D framework -- one space of worlds, two sorts of semantic evaluation. Somebody might want to argue that some of these epistemic possibilities don't correspond to metaphysical possibilities even in the 2-D way, but such a claim would have to rest on considerations quite distinct from Kripke's. That's something we'll be talking about later on. >> It sounds odd to say that a PI is a function from worlds to >> referents, and then say that different people can have different primary >> intensions for the same terms. `Water' is supposed to pick out the >> watery stuff in any world in which it is uttered, quite irrespective (I >> take it) of what folks know about water (though, presumably, most folks >> would know which stuff was the watery stuff--but this needn't be the >> case). What exactly is odd here? I talked yesterday about some ways in which the PI of a term such as "water" might vary between individuals. Take the chemist who associates "water" only with certain chemical properties (and let's say, intends to use the term entirely without semantic deference). For him, if the XYZ-world turns out to be actual, "water" will still refer to H2O, not XYZ. So his PI for "water" is quite different from ours. He'd be a very unusual user, of course, but it's not obvious that the story is incoherent. One might resist by suggesting that his word "water" is not really our word, and part of me would have some sympathy, but that would be a tricky row to hoe. And without going to this sort of extreme, one can arguably find everyday cases in which two different users of a natural kind terms have slightly different PIs (as witnessed by the fact that they would react to certain epistemic possibilities in different ways). E.g. a miner who is an expert on "chromium" and an airplane worker (I don't know anything about chromium, so let's pretend that it is found in mines and used in airplanes). Arguably the miner could pick out chromium as "the stuff with such-and-such superficial properties that I mine", and the airplane worker could pick it out as "the stuff with such-and-such chemical properties that I work on", or some such. Or at least, the miner might give a little more weight than the airplane worker does to superficial properties. All this is complicated by issues about semantic deference (where users intend to refer to what the community around them calls "chromium", as one gets in e.g. Burge's "arthritis" cases) but I'm taking it here that it's at least possible to use such terms without semantic deference. I take it that Kripke's view is that natural kind terms are like names in this regard as well as in the "rigid designator" modal regard -- i.e., he thinks they don't have any canonical "mode of presentation" under which they are picked out. I.e., no universal PIs for natural kind terms. I don't think this matter is cut-and-dried, and the issues of to what extent natural kind terms are like names here is at least somewhat debatable; certainly, there is an intuition that they have some sort of descriptive "feel". It's a tricky issue. It's true that when thinking about natural kind terms, it's easy to slip into a picture on which every user has the same PI, and I don't think there's much harm in doing that (and it doesn't do huge violence to situation that most of us find ourselves in), as long as we recognize that it may turn out to be be an idealization. In any case, even if Kripke is right, it's not obvious that serious problems arise for us if different users have different PIs for the same terms. For most purposes that will matter for us, it might as well just be a single user's PI that we're interested in. The effect will be that what counts as (say) an explanation of "water" for us may differ from what counts for the chemist, but that won't indicate any major metaphysical difference, just a difference in the way we use the term. The general picture of metaphysical and explanation (abstracting away from surface terminology) will stay the same. >Perhaps related to this, there was some argument about whether >even natural kind terms (our example was "water") have primary >intensions. Brad claimed that it made sense to say "Water might not have >been H2O", but Thony argued that this only made sense to him when the >modality was construed epistemically rather than metaphysically. I (Brad) >was claiming that it was true on a metaphysical reading. If it were true, >that would motivate the notion that "water" has a primary intension (since >it is certainly *not* true that H20 might not have been H20). Hmm. I guess I'd argue that natural kind terms (or term tokens) have PIs in just the same way as I argued that names have PIs: they have a distinctive pattern of application to epistemic possibilities. Re "water might not have been H2O": personally, I find this ambiguous between what I'd call an epistemic and a "subjunctive" reading (I prefer not to use "metaphysical" here since I think both are metaphysical in a way). Given the way the actual world is, I think Kripke's makes a pretty strong case (though not an unarguable case) that there is no counterfactual world in which water is not H2O. If so, the subjunctive reading comes out false. But it's still the case that it's broadly epistemically possible that water isn't H2O, i.e. there a possible world such that when that world is considered as actual, "water is H2O" comes out false. So I think one can argue that the epistemic reading comes out true. One might argue over whether the sentence in question actually has the epistemic reading (one could argue it is intrinsically subjunctive). I personally think language is pretty flexible here, but if one has doubts, one can substitute something more clearly compatible, such as "water might not be H2O". I'm not sure wheter I'm agreeing with Brad or Thony here. I guess I agree with Thony that it's true on an epistemic reading, and the question of the "metaphysical" depends on just what one means by that. If one means a Kripkean subjunctive reading, then I suppose I'm inclined to agree with Kripke that the sentence is false, though I don't think the issue is entirely cut-and-dried. But as I said, I'm inclined to think that even the epistemic reading is "metaphysical", in that it involves evaluating perfectly respectable metaphysically possible worlds. Anyway, I think that whatever one says about this, water comes out having a PI either way. All one needs for a PI is a distinctive pattern of application across epistemic possibilities (considered as actual), and that's just what we seem to find. E.g., it's epistemically possible that water isn't H2O, i.e., there a scenario (the XYZ-world) just that when it's considered as an epistemic possibility, "water is H2O" comes out false, i.e., the PI of "water = H2O" is false in that scenario, i.e., "water" and "H2O" have different PIs. >I guess the question that I (Thony) wanted to raise (though not at all >clearly) was >whether the semantically salient thing that names (natural kind terms, >etc.) contribute to sentences/discourses in which they occur is anything >like a primary intension. I'm (Thony) not at all sure, but my intuition >seems to favor SI's in this respect. That's an interesting question that depends a lot on what we mean by "semantically salient". It's true that given that PIs can vary between users and that SIs don't, one might argue that SIs are most central to semantics. On the other hand, when one looks at what goes on in communication and understandiong (e.g., what I learn when someone tells me that there's water in the sink), I think it's arguable that PIs are more central. It's a tricky issue. >This got us on to a discussion of whether it is words that are the >(primary) bearers of primary intensions, or concepts. I (Tim) think >we concluded that both words and concepts are meant to have primary >intensions. Right. On my view, both words (or word tokens) and concepts have PIs. The treatment in the book is largely directed at words; the treatment in "The Components of Content" is largely directed at concepts. I think one can argue that there is a certain sense on which the PIs of concepts are "primary", in that the PI of a user's token of a word is derivative on the PI of the user's corresponding concept. But that matter isn't entirely obvious. >There was some discussion of whether a word has a single >primary intension associated with it. I (Tim) suggested that it is >not obvious >that this is the case: Imagine that I know that ice, liquid water and >steam are the same stuff. I call all three kinds of stufff "water" and its >part of my primary intension that all three kinds of stuff are the same >stuff. Imagine that my kids are only exposed to liquid water, and I don't >teach them that water can boil or freeze. Two questions: (A) When I talk >with my kids >about water (liquid water), are we using the same word (I guess I want to >say yes, tentatively); (B) do we attach the same primary intension to it >(I want to say no, tentatively). Ah, excellent. That's a nice example of the sort of thing I was talking about before. I think it clearly seems intuitive to say that we're using the same word. Do we have the same PI? Actually, that's a bit tricky. We may have different beliefs, but it certainly isn't the case that every difference in beliefs is reflected in a difference in PIs. What matters intuitively for a PI are the beliefs that a user implicitly takes to be "constitutive", or a priori. It's not obvious to me that your belief that water has three forms is one of those beliefs. E.g., just say it turns out that the liquid around you only has one form, and that the stuff in freezers and that comes out of kettles is actually some entirely different chemical compounds. I think then you'd say that "water" refers to the liquid but not to the other things. If so, the "epistemic profile" of your "water" concept doesn't require that water have the three forms. Arguably your PI for "water" is more centered on the liquid stuff, just like the kids'. So it's not completely obvious that this is a case of different PIs. Still, it might not be too hard to turn it into such a case, or to find other such cases in the vicinity. >{It also seems hard to draw a hard and fast distinction between primary >and >secondary intensions, given the way kids learn words/concepts. If content >really has this dual structure, wouldn't one expect it to be reflected >more clearly?} I'm not sure exactly what the concept-learning considerations are here. On my view, the central structure of our concepts involves PIs, as these give the conditions whereby our concepts apply to the actual world, are most associated with "core beliefs", determine the epistemic possibilities that we take to be live options, etc. SIs have a secondary role, giving the application-conditions of our concepts to counterfactual worlds, and mostly are revealed in counterfactual thinking and in subjunctive contexts. So I suppose that SIs may be revealed in kids' thinking about counterfactual scenarios ("just say I had a twin brother"), and how they describe them, etc. But probably PIs are more central, as these are what are centrally involved in the kids beliefs about the way the actual world is, about the epistemic possibilities they're prepated to entertain, etc. So I think one would probably find it easier to "read off" a PI from a kid's surface conceptual structure than an SI; and indeed, firm SIs may take longer to develop than firm PIs (given that thinking about counterfactual scenarios can be very loose). But this all brings up some very complex and arguable issues (I don't think anything I just said is obvious or uncontestable, and don't worry if it's hard to follow). I hope this all helps to clarify some issues about the way the 2-D framework works. Please go ahead and fire away with follow-up questions, other thoughts, etc. --Dave. From owner-modality@LISTSERV.ARIZONA.EDU Wed Feb 10 22:54:19 1999 Date: Wed, 10 Feb 1999 23:53:35 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Erik A Herman To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Hi, My main issue is Kripke's discussion of water, heat, and light (pg. 128). He says, and I agree, that "we identified water originally by its characteristic feel, appearance and perhaps taste..." But he goes on, and this is where I don't agree, to say that "if there were a substance...which had a completely different atomic structure from that of water, but resembled water in these respects, would we say that some water wasn't H20? I think not." He goes on to make an analogy to fools gold (fools water). My difficulty is in the amount of weight he/we are putting on physics. The reason fools gold isn't gold isn't because it is molecularly different but because the feel/appearance/(taste) part doesn't correspond to gold. My contention is that if everything were the same gold/fools gold, water/fools water than we WOULD say when we finally discovered the difference under the microscope that there are two kinds of water and two kinds of gold-- for one thing it would be impossible to determine which was the *right* one. As it turns out we understand these things in terms of the feel/appearance/taste rather than as the molecular descriptions they happen to have. Likewise, I feel the same way about heat and light. I do not think heat IS the motion of molecules or that light IS a stream of photons, rather they are the phenomenal qualities that happen to be stimulated by molecule motion and photon streams. I realize the importance of separating phenomenal effects from physical events but we shouldn't use the same word to describe both. Am I running into a typical problem? Is there a name for the position I take? More importantly, is there something big that I'm overlooking (I always feel like there is.) Do I have it right that, in this case the primary intension would be the feel/appearance/taste? i.e. the essential qualities? This stuff can get a little confusing. :) Erik H. P.S. I don't see how consciousness could be a priori OR a posteriori -- could we make another category for it?! Also, I can't see, especially in light of everything, how Chalmers could possibly think he's a zombie. Or any of us for that matter. From owner-modality@LISTSERV.ARIZONA.EDU Thu Feb 11 06:10:05 1999 Date: Thu, 11 Feb 1999 06:09:55 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Erik H. raises an interesting point: maybe the reason we say that fools' gold isn't gold is not because it has different internal makeup, but because it has slightly different surface properties. That's an interesting and subtle issue. Things are complicated slightly by the fact that fools' gold is found in a lot of the same places we find gold. Because of that, you may well be right that if it were just the same on the surface, we'd say that there were two sorts of gold, just as we say there are two sorts of jade with different structure, jadeite and nephrite. (After all, both would be at the other end of the relevant causal chain!) But how about if there turned out to be superficially identical fool's gold on a distant planet, or if it were synthesized in a lab? If that were so, I think the chances that we'd call it "gold" would diminish, though it isn't completely obvious. One thing we can say is that having the same intrinsic structure as the stuff we've in in contact with certainly helps get the label of "gold". Think about "carbon": certain superficially very different substances all get the label "carbon" because of their common internal structure, whereas stuff that looked like that with different structure certainly wouildn't get the name. So certainly one important factor in determining reference is having the same internal structure as the stuff we've been in contact with. So that suggests that surface qualities can't be the only factor (as Erik suggested at one point). But this doesn't imply that intrinsic structure must be the only factor, either (as Kripke suggests). The possibility remains that e.g. a disjunction of the right surface quality or the right intrinsic structure is enough to deserve the name. How to assess this? One relevant data point involves different rare-ish metals with the same surface structure, or very similar structure (titanium and chromium, say, though I have no idea). Here, it seems that finding a metal with the same surface structure as titanium isn't enough for us to call it titanium. Titanium goes with a particular atomic structure, and that sems to be that. So one might argue that the same goes for gold. does, too. And what would we say if a time of astronauts find a planet with the superficially identical XYZ? Water or not? Putnam and Kripke bet that we'd say it isn't water -- see "The Meaning of Meaning" for a much more extended argument for this sort of claim. I recall that I was sort of skeptical when I first heard the claim myself, but I've gradually come around. I can still see the intuition, though, and part of me can still go both ways. >Likewise, I feel the same way about >heat and light. I do not think heat IS the motion of molecules or that >light IS a stream of photons, rather they are the phenomenal qualities >that happen to be stimulated by molecule motion and photon streams. I >realize the importance of separating phenomenal effects from physical >events but we shouldn't use the same word to describe both. That's another tricky one. It raises the specter of the old riddle, "if a tree falls in the forst with no-one around, does it make a sound?". To which one standard answer is to say, it makes a sound in one sense but not another. We might say it makes an "objective sound" (the air movement) but not a "phenomenal sound" (the subjective experience). And one could argue that the term "sound" is ambiguous between these. One might make the same ambiguity claim for light and heat: objective light vs phenomenal light, objective heat vs. phenomenal heat. You want to go further and say that terms such as "heat" refer to phenomenal heat, not objective heat. One reason for questioning this might lie in the fact that we sometimes say "it's really hot in there" even though there's no-one who's actually feeling the heat (e.g. inside a furnace). Similarly we can talk about the light sent from Alpha Centauri to others star (which maybe no-one sees, etc). So it seems that objective heat and objective light are at least sometimes called "heat" and "light". Do we sometimes use the terms for phenomenal heat and for phenomenal light. I think maybe so, though it isn't so clear. Kripke seems convinced that in a world where someone feels our "heatish" sensations but caused by something very different in the external world, we'd say there's no heat. But it doesn't seem completely unreasonable to say that there is heat in that world, that the external things are hot, etc. It's a little less clear for "light", but maybe. If you really insist that we shouldn't use the same word for both, my own feeling is that there is slightly more of a case for using the term for the objective kind than for the phenomenal kind. But of course it is ultimately a terminological issue. One of my fundamental principles in philosophy is that nothing of vast importance ever really turns on a terminological issue. But of course the way the terminology of natural language works is of some interest in its own right. >Am I running into a typical problem? >Is there a name for the position I take? > >More importantly, is there something big that I'm overlooking (I always >feel like there is.) > >Do I have it right that, in this case the primary intension would be the >feel/appearance/taste? i.e. the essential qualities? I can't think of a name, offhand, but I think your view isn't entirely unusual. Some people have argued that the reference of our terms such as "heat" has gradually shifted from the phenomenal kind to the objective kind as science has progressed: i.e., we have gradually "objectified" the term. So when Goethe argued that light couldn't be explained by science, it wasn't so much that he was proved wrong as that the subject was changed. Either way, the PI of "heat" etc will involve the appearance. If we think "heat" refers to phenomenal heat, the PI will just pick out that particular sort of sensation in all worlds, as will the SI. If (like Kripke) we think it refers to objective heat, the PI in a given world will pick out the external property that causes those sensations around the center of the world. The SI will pick out the *actual* external property (the motion of molecules, given the way the'actual world turned out) in all worlds. (So the PI can be roughly summarized as "the cause of heat sensations", and the SI as "the motion of molecules".) >P.S. I don't see how consciousness could be a priori OR a posteriori -- >could we make another category for it?! I take it you mean knowledge of consciousness. I'm sympathetic with your point here -- there's certainly some sense in which we know about our consciousness non-empirically and certainly, so it can seem a priori. And some philosophers have argued that it is a priori. But i think for a number of reasons it is best to see it as a posteriori. One's knowledge of consciousness is knowledge in virtue of experience, which is just what a posteriori knowledge involves. It isn't a "truth of reason" in any clear sense. One can argue that phenomenology and introspection are just as "empirical" as external observation: one'measures trhe first-person data, the other the third person data. And finally, a lot of philosophical issues make more sense if we see this as a posteriori. So I think there's a good case for putting it in this category, even if it has a somewhat different feel from the paradigm cases of a posteriori knowledge. > Also, I can't see, especially in light of everything, how Chalmers >could possibly think he's a zombie. Or any of us for that matter. I don't think I'm a zombie! All I said is that it's not a priori that I'm not a zombie, and that it's 1-possible that I'm a zombie. That's entirely compatible with the claim that I am certain that I'm not a zombie. It just turns out that I am certain of something that is a posteriori. That seems OK to me, but if there's a problem, I'm interested to hear it. As for you, I'm not certain you're not a zombie, but I think you're not, and arguably I know you're not. But still, it is certainly a broad epistemic possibility that you're a zombie; i.e., it isn't ruled out a priori. --Dave. From owner-modality@LISTSERV.ARIZONA.EDU Fri Feb 12 14:17:22 1999 Date: Fri, 12 Feb 1999 15:16:38 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Timothy J Bayne Subject: Proper Names Comments: To: David Chalmers To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO A thought on proper names. . . There is an interesting tension between two aspects of the proper names that we use. On the one hand, we want a proper name to be unique to the individual whose names it is. (This fact, or rather the absence of this fact, is the basis for Monty Python's Australian Philosophers sketch, in which everyone is called Bruce.) On the other hand, individuals do share proper names, lots of people are called Bruce. It is interesting to note that there is anecdotal evidence that kids resist calling another individual by a proper name that they have already attached to an individual:If my sister is called Jane, then it's wrong to call this other person Jane as well. They seem to over-extend the principle that proper names aren't like nouns and adjectives, they are meant to apply to unique individuals. How does this relate to the "meaning" of proper names? Consider a conversation in which I am talking about my sister Jane, and not my friend Jane. Or, better, a conversation in which I'm talking about my sister Jane, and someone interupts to tell me that "Jane wants to talk to me", and I interpret the second "Jane" to be about my friend. In order to keep the references of the various tokens of Jane straight, I presumably have to bind some kind of descriptive content to them. At some level, don't I have to be thinking "Jane-my-sister" or "Jane-my-friend", or something with *some* descriptive content? The precise nature of the descriptive content needed here is not obvious, but it is clear that a traditional descriptivist suggestions won't work: "The person that everyone around here calls *Jane* won't work", because there are lots of people around here called Jane. In other words: in general, the meaning of a proper name can't just be its referent, 'cos most proper names have a number of referents. One could respond to that by saying that *Jane* in one context means something different from *Jane* in another context, because it is used to refer to a different person. Thus, Jane/Jane would be just like bank(financial)/bank (river). But the analogy with "bank" seems to me to play into the descriptivist's hands. When I say "I saw you at the bank yesterday" I know that I either mean river bank or financial bank 'cos of the descriptive content I associate with the token. Unless I also associate descriptive content with 'Jane', how do I (or my interlocuters) know which Jane I am talking about? Tim Timothy J. Bayne RM. 213 Social Science Department of Philosophy University of Arizona Tucson, AZ 85721 USA Hm ph. (520) 298 1930 From bayne@U.Arizona.EDU Fri Feb 12 14:17:19 1999 Date: Fri, 12 Feb 1999 15:16:38 -0700 (MST) From: Timothy J Bayne To: David Chalmers cc: MODALITY@LISTSERV.ARIZONA.EDU Subject: Proper Names Status: RO A thought on proper names. . . There is an interesting tension between two aspects of the proper names that we use. On the one hand, we want a proper name to be unique to the individual whose names it is. (This fact, or rather the absence of this fact, is the basis for Monty Python's Australian Philosophers sketch, in which everyone is called Bruce.) On the other hand, individuals do share proper names, lots of people are called Bruce. It is interesting to note that there is anecdotal evidence that kids resist calling another individual by a proper name that they have already attached to an individual:If my sister is called Jane, then it's wrong to call this other person Jane as well. They seem to over-extend the principle that proper names aren't like nouns and adjectives, they are meant to apply to unique individuals. How does this relate to the "meaning" of proper names? Consider a conversation in which I am talking about my sister Jane, and not my friend Jane. Or, better, a conversation in which I'm talking about my sister Jane, and someone interupts to tell me that "Jane wants to talk to me", and I interpret the second "Jane" to be about my friend. In order to keep the references of the various tokens of Jane straight, I presumably have to bind some kind of descriptive content to them. At some level, don't I have to be thinking "Jane-my-sister" or "Jane-my-friend", or something with *some* descriptive content? The precise nature of the descriptive content needed here is not obvious, but it is clear that a traditional descriptivist suggestions won't work: "The person that everyone around here calls *Jane* won't work", because there are lots of people around here called Jane. In other words: in general, the meaning of a proper name can't just be its referent, 'cos most proper names have a number of referents. One could respond to that by saying that *Jane* in one context means something different from *Jane* in another context, because it is used to refer to a different person. Thus, Jane/Jane would be just like bank(financial)/bank (river). But the analogy with "bank" seems to me to play into the descriptivist's hands. When I say "I saw you at the bank yesterday" I know that I either mean river bank or financial bank 'cos of the descriptive content I associate with the token. Unless I also associate descriptive content with 'Jane', how do I (or my interlocuters) know which Jane I am talking about? Tim Timothy J. Bayne RM. 213 Social Science Department of Philosophy University of Arizona Tucson, AZ 85721 USA Hm ph. (520) 298 1930 From owner-modality@LISTSERV.ARIZONA.EDU Fri Feb 12 19:52:56 1999 Date: Fri, 12 Feb 1999 20:51:59 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Timothy J Bayne Subject: RE Proper Names To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO With respect to my previous comment about proper names. . . .(Re)reading the preface to *Naming and Necessity* tonight I was reminded that Kripke discusses just this point, and seems to think that it's of no real relevance. I'm not pursuaded that he's justified in dismissing it in the way he does. He seems to think that you just bring in pragmatics or something to work out which Aristotle you're talking about, i.e. to work out which proposition you mean to express, and then plug his rigid designation account into the proposition that one has isolated. Well, maybe, but maybe not. But it's not obvious to me that we can simply dismiss the issue of homonymous names to speaker meaning or pragmatics or something. I don't have an argument here, but Kripke's discussion seems to me to be a little underwhelming. From owner-modality@LISTSERV.ARIZONA.EDU Fri Feb 12 21:51:49 1999 Date: Fri, 12 Feb 1999 21:51:41 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Re: Proper Names To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Lots of interesting issues about multiple uses of the "same" name in Tim's message. Starting from the end of the message: >In other words: in general, the meaning of a proper name can't just be its >referent, 'cos most proper names have a number of referents. One could >respond to that by saying that *Jane* in one context means something >different from *Jane* in another context, because it is used to refer to a >different person. Thus, Jane/Jane would be just like bank(financial)/bank >(river). But the analogy with "bank" seems to me to play into the >descriptivist's hands. When I say "I saw you at the bank yesterday" I know >that I either mean river bank or financial bank 'cos of the descriptive >content I associate with the token. Unless I also associate descriptive >content with 'Jane', how do I (or my interlocutors) know which Jane I am >talking about? Hmm. The view that the meaning of a proper name is just its referent is very popular in contemporary philosophy (held e.g. by direct reference theorists), perhaps even so popular as to be conventional wisdom. I think those who hold this view will respond by saying that the two instances of "Jane" here are simply two different names. One name means X, and one name means Y. It's just that somewhat coincidentally, the two names are spelled and sound the same, or something like that. See for example Kripke's discussion on pp. 7-8, where he suggests that "uses of phonetically the same sounds to name distinct objects count as distinct names". Of course one might object by saying there's an intuitive sense in we say that Jane 1 and Jane 2 "have the same name". Here a direct reference theorist (and others) might respond e.g. by saying that there are two ways of individuating names, either by their phonetics (etc.) or by their meaning; and that while phonetic individuation sometimes plays a role in ordinary practice, it is meaning-individuated name types that correspond best to the deep structure of language, or that matter most in philosophy, or something like that. As usual the issue isn't cut-and-dried, but I think the direct reference theorist has some resources here. Your objection in turn raises a really central issue in the debate over the semantics of names: how do I know the referent of my name? The direct reference theorist might say, well, you know that the referent of "X" is X, but it might be objected that this is trivial and unhelpful. In practice something more seems to be involved in the way in which we identify the referent of our names, i.e. the way in which we come to know who or what the referent is. This seems closely tied to what we might call the "mode of presentation" of the name. The direct reference theorist may well accept that any use of a name has a mode of presentation associated with it by the speaker, or at least some way of identifying the referent, but they will deny that the mode of presentation is part of the "meaning" of the name. It might merely be something in the psychology of the speaker, for example, rather than a truly semantic phenomenon. (Often, a direct reference theorists will sharply separate "semantic" and "cognitive" issues.) Others (especially description theorists) prefer to think of the mode of presentation of a name as something in the realm of meaning and semantics. One consideration in favor of the direct reference theorist is the point (which we've already noted a couple of times) that different users of the same names may well have different associated modes of presentation. E.g., Hillary's and my use of "Bill Clinton". One might then say it follows automatically that the mode of presentation can't be part of semantics, if the semantic features of a term are common to all users of a term, as it is often thought they must be. In response, it might be argued that the mode of presentation is at least part of "utterance meaning" rather than "sentence meaning", or something like that, where the former can be specific to a single use or user of a term, where the latter is required to be universal. Here again Marga Reimer's work on distinctions like these is relevant; I don't have the relevant papers with me (and may be misremembering them a bit), but fortunately Marga has just joined the mailing list -- any contribution is welcome! At some point these debates about what truly counts as "meaning" and "semantics" become somewhat terminological (that's another of Marga's themes, I think), but the substantive issue remains of whether there are modes of presentation associated with names (whether or not they are part of semantics), and if so what they are. Tim in effect argues that there must be something like this, in order that we can know what we're talking about, and I'm sympathetic. One possibility is that these modes are descriptions; that would make things easy, as we could know that the referent is whoever satisfies the descriptions. But maybe there are other possibilities, e.g. that the modes of presentation are causal chains linked to our tokens, and to identify a referent we follow down the causal chain associated with the tokens (given that we somehow know which causal chain to start following). My own view, you won't be surprised to hear, is that these modes of presentation are precisely primary intensions. It's the case that any name (as used by a user) has its own epistemic application conditions, i.e. conditions under which it applies to epistemic possibilities, under the hypothesis that those possibilities turn out to be actual. The fact that users know what they're talking about corresponds to the fact that users have the conditional ability to identify the referent of a name given sufficient empirical information about the actual world -- that is, they have the ability to evaluate a name's primary intension at a world. So they have "cognitive access" to their names' primary intensions, and in a sense the primary intension represents the "cognitive content" of the name as used by them. And at the same times, the primary intension determines reference. The primary intension needn't be a description in any obvious sense, but as we've seen its plausible that it will always have some associated descriptive content (core beliefs, etc). How does all this relate to the dual-name case? I think the issue Tim raises (about knowing what we're talking about) applies even to cases without ambiguity, but it comes up in a sharp form here. When I use the names "Jane 1" and "Jane 2" (so to speak), I know (or believe) that these refer to different people. This reflects itself in my ability to identify the referent of each, given relevant information about the external world. So in many/most/all epistemic possibilities considered as actual, I can determine the separate referent of the names. This is just to say that I have cognitive access to quite different primary intensions for each name. The question then arises about just what these primary intensions involve at the psychological/cognitive level; i.e., what is it that grounds my cognitive ability to identify the referents? The description theorist will say that it is the descriptions I associate with the name: in effect, the primary intension is given by some core or cluster descriptions. Someone else might say it is given by a reactive disposition of some kind, not necessarily mediated by descriptions. My own view is that a primary intensions needn't be represented as a description per se, but that nevertheless it will correspond to what we might think of as "tacit descriptive knowledge" about the referent. E.g., even if we identify the referent (with Kripke) by following a causal chain from a token, our very ability to do this suggests a tacit knowledge that the referent is at the other end of the causal chain, which one might think of as tacit descriptive knowledge about the referent in some sense. So on this view, there will at least be some sort of loose link between primary intensions and tacit descriptive content and the cognitive level. Applying all this back to the Jane cases, the distinct cognitively accessible primary intensions for the two names suggests that there must at least be some sort of different descriptions tacitly associated with each name, that will emerge in the way reference is identified. Just what sort of descriptions they are will depend on the details of the case. I think one is likely to say something quite different depending on whether one is acquainted with the referent in question or has merely heard about them and picked up the name. In the case of acquaintance (e.g. the sister/friend case), I take it one will have tacit descriptive "immediate" knowledge of the people in question and that some of this will be crucial to reference fixation (e.g., knowledge of their roles vis-a-vis you). In the case of non-acquaintance (say I have names for David Hilbert the mathematician and David Hilbert the philosopher, though I haven't met either), we'll have deference to a community of prior users of the name, and things get a bit subtler. One might try using surface descriptive knowledge (e.g. of career) to disambiguate, but I'm not sure this gets to the heart of things, as arguably one could be wrong about those things, and there are cases where one has little such knowledge. I think it's better to note that one knows (at least tacitly) that there are two names here (or one knows that there are at least potentially two people), and one knows that one picked up each name from the community in a slightly different manner. So in each case, one fixes reference by deferring to the community, but in different ways. Maybe for one I defer to the mathematical community (where I heard the name) and the other to the philosophical community. Or maybe I don't remember much at all about which community or users I got the names from, but still, I implicitly defer in each case to "whatever users I picked up *this* name from", as opposed to "whatever users I picked up *this* name from" for the other (in effect demonstrating the different name tokens). That arguably does enough to get started on the different causal chains by which reference through deference is fixed. Anyway, I think you're right that the ability to refer differently with the two names (or name tokens) suggests that we have some sort of tacit descriptive knowledge associated with each that reflects our ability to do so (though in the last case above, the difference in the knowledge is pretty thin). Just whether this difference in associated descriptive content (and difference in primary intension, etc) is enough to refute the direct reference theorist isn't quite clear, as they still have the option of saying that it's a matter of psychology (say), not semantics, and that the true semantics of the name still just involves the referent. But in any case it suggests that there is something important going on in the vicinity. This is already too long, and I think we've covered most of the bases now, but a few quick notes on the other bits. >How does this relate to the "meaning" of proper names? Consider a >conversation in which I am talking about my sister Jane, and not my friend >Jane. Or, better, a conversation in which I'm talking about my sister >Jane, and someone interrupts to tell me that "Jane wants to >talk to me", and I interpret the second "Jane" to be about my friend. In >order to keep the references of the various tokens of Jane straight, I >presumably have to bind some kind of descriptive content to them. At some >level, don't I have to be thinking "Jane-my-sister" or "Jane-my-friend", >or something with *some* descriptive content? > >The precise nature of the descriptive content needed here is not obvious, >but it is clear that a traditional descriptivist suggestions won't work: >"The person that everyone around here calls *Jane* won't work", because >there are lots of people around here called Jane. I hope I've mostly covered this above. One might argue that the mere existence of the two names just suggests that you have two different mental tokens (say) "Jane 1" and "Jane 2", and some sort of tacit rule for figuring out when someone's "Jane" talk corresponds to one or other symbol (e.g., your family speaking -> Jane 1, your friends speaking -> Jane 2). But the fact that you have the ability to know who you're talking about in a deeper sense, i.e. the ability to identify the referent of each in the world, suggests as above (at least to me) that you must at least have some sort of deeper tacit descriptive content associated with each. You're right that the nature of the descriptive content isn't obvious. It's a nice example to refute one description (the one you give) that description theorists sometimes offer. It seems that one would need to appeal to either (a) different bits of "first-order" descriptive knowledge (e.g. in the acquaintance case), or (b) different bits of "metalinguistic" descriptive knowledge (in the deference case). And as you say (in effect), the metalinguistic knowledge can't be specified just by using the word "Jane". One will either have to use knowledge of the different communities or users that one picked up the name from, or at least, the knowledge that each token is part of a separate chain of uses, and that each token is causally connected to a point where you picked it up from somewhere. The last case is an example of the very thin "tacit indexical causal descriptive knowledge" that we talked about above, but I guess it is enough to enable one to pick out the different referents, in the case where one has very thin knowledge associated with the name. The general question of what descriptions a description theorist should appeal to to handle the Kripke cases (e.g. the Feynman, Godel, Jonah cases) and others is a really interesting one that I hope we get a chance to spend some time on at some point. Arguably there are resources that Kripke doesn't consider. His discussion on pp. 160-162 is worth looking at. He concedes (as in footnote 38) the possibility of a relatively trivial "causal descriptivism" that encapsulates the whole causal theory of reference into a description -- not exactly a strong form of descriptivism, but maybe enough for some descriptivist purposes. There are also metalinguistic options to think about, of which he considers some but arguably not the best. Interestingly he alludes (on p.162) to one of the better of these, "Let 'Glumph' denote the man called 'Glumph' by the people from whom I got it (whoever they are)", but only in conjunction with the "trivial" causal description above. One could argue that this is better than any of the metalinguistic options he considers explicitly, although it is still imperfect -- e.g. it might turn out that you misheard the name, so that you're the only one who uses 'Glumph' but you still refer. But then one might try "the person I heard of under the name 'Glumph'", or "the person referred to by the people from whom I got the name 'Glumph', under the name causally responsible for my use of 'Glumph'". And maybe bring in "this very name 'Glumph'" to get around the ambiguity problems Tim raises above. It gets tricky, of course, but then the description theorist never said there has to be a universal description formula for all names. Arguably, going for a "cluster" description with one of the metalinguistic descriptions just mentioned as one of the components of the cluster (probably the central component) can handle almost all of the Kripke-style cases and other cases of deference, and can go along way to grounding something like a causal theory of reference in something like a description theory. But that's a long story. Personally one of the reasons I like the primary-intension framework is that it bypasses this need to come up with specific descriptions, which can obviously get very complex quite quickly. Nevertheless I think there are ways in which one can use the framework to lend support to an attenuated version of the description theory. --Dave. P.S. I just got this. >With respect to my previous comment about proper names. . . .(Re)reading >the preface to *Naming and Necessity* tonight I was reminded that Kripke >discusses just this point, and seems to think that it's of no real >relevance. I'm not persuaded that he's justified in dismissing it in the >way he does. He seems to think that you just bring in pragmatics or >something to work out which Aristotle you're talking about, i.e. to work >out which proposition you mean to express, and then plug his rigid >designation account into the proposition that one has isolated. Well, >maybe, but maybe not. But it's not obvious to me that we can simply >dismiss the issue of homonymous names to speaker meaning or pragmatics or >something. I don't have an argument here, but Kripke's discussion seems to >me to be a little underwhelming. I think you're right that Kripke doesn't get at all the central issues. He mostly says that this issue is irrelevant to issues concerning rigidity, and I think he's more or less right about that. But arguably it is important for issues about the fixation of reference, e.g. for issues about causal theories and descriptive theories of reference-fixing, of the sort he describes in Chapter 2. There has to be some story about how the different tokens get their distinct referents, and as we've seen, this isn't entirely trivial. Incidentally I heard someone a while back give a paper with the thesis, contra received wisdom but in sympathy with your view, that all uses of name tokens such as "Jane" are in fact tokens of the same name. So there is just one name "Jane", one name "Paul", one name "David Hilbert", etc. She addressed the problem of how the name gets to refer differently in different cases by arguing that names are implicit demonstratives -- in effect one is saying "that Jane", "that David Hilbert", etc, with different intentions to demonstrate in each case. A demonstrative like "that" can have different referents in different contexts and with different intentions, but it's still one word. Same on this view for "Jane". I'm not sure the view is completely unproblematic, but it's interesting. From owner-modality@LISTSERV.ARIZONA.EDU Sat Feb 13 10:41:56 1999 Date: Sat, 13 Feb 1999 11:40:43 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Rachael J Parkinson Subject: Re: Proper Names To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO I have some questions about the Kripke reading- specifically on pages 123-133. On page 123 Kripke acknowledges that we could be mistaken about gold having the atomic number 79. But he argues that given that gold does have the atomic number 79, something could not be gold without having the atomic number 79. Kripke makes light of Kant's mistaken assertion that gold is a yellow metal- but it seems that it is *possible* that just as we found out that all gold isn't yellow, we may find out that we are mistaken about gold having the atomic number 79. Kripke argues that we can pick gold out in all possible worlds by its atomic number. But if he was arguing 300 years ago, perhaps he might have made a similar argument- that gold is a yellow metal in all possible worlds. This brings up the question of what properties are essential. I think the answer is more difficult to pinpoint in cases like 'what makes a tiger a tiger' then for 'what makes gold, gold?' On page 121, Kripke seems to suggest that the external appearance of a tiger is not essential - we must look at its internal structure to determine whether it is a true tiger. My question is two-fold. First, if we were to discover a microscopic virus that happened to have the same internal structure as a tiger- would we recognize it as a tiger? Second, Kripke asserts that certain properties like quadrepedal, tawny yellow, and carnivorous are not essential to tigers, we might find that all these properties are optical illusions, but couldn't we discover the same of the internal structure of tigers? I guess I have a problem with how to identify natural kinds, particularly in reconizing which properties are essential and which are not. In light of past mistakes (like gold is a yellow metal) I am not sure that we can ever know that we have actually discovered the essence of a natural kind, or what picks it out in all possible worlds. Perhaps you could say, given *our* concept of gold, what picks it out for us is the atomic number 79. But if it seems logically possible to be wrong about gold always having the atomic number 79 in this world, then why couldn't we be wrong about it in other worlds? As our science develops we discover that our original theories about light and heat were mistaken. I'm not sure how this works- but it seems like if enough of the data (what we recognize as heat and light) don't match up with the theory- then we find a new one. It is still possible to discover that heat is not just the motion of molecules- I suppose that to accomplish such a discovery we would have to be using our primitive notion of heat by which we identify it as a characteristic effect produced on our nerve endings. If heat just *is* motion of molecules then it looks like instead of reformulating the theory- we would have to call that other kind of thing that produces a hot sensation in us- something other than heat. It just doesn't seem that "heat is motions of molecules" or "water is H2O" are as necessary as "a bachelor is an unmarried man." What do you think? -Rachael From owner-modality@LISTSERV.ARIZONA.EDU Sat Feb 13 17:03:29 1999 Date: Sat, 13 Feb 1999 17:52:44 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Laurence A James Subject: Re: Proper Names To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Hi, I would like to respond to some of Rachael's questions. First, you mention that Kripke doesn't deem properties such as quadrepeedal and tawny yellow essential to tigers. I think the reason for this, is not so much that he is worried about optical illusions, but rather more about cases where a tiger is missing a leg (or about one of those white tigers for instance). Both of these types of tigers are still tigers regardless of their minor physical defects. Secondly you mentioned the possibilty of finding a virus that might have the same internal structure as a tiger, and asked whether or not this should be a tiger too? However, I think Kripke was being a little more direct when he mentioned internal structure. I think he was referring to something like the tiger's DNA structure. And, as far as i understand genetics, it will be quite unlikely that a virus would have the same genetic structure (i.e., DNA) as a tiger. As for Gold having atomic number 79. This is an essential property of Gold. Gold and only gold has this atomic number. Now, we might find something that looks similar to gold, infact almost identical, but I don't see the reason why we would have to call it gold. I think that Kripke makes a strong claim here - that for something to be Gold, no matter whether in this world or a possible one, it must have the atomic number 79. Similar looking things with different atomic numbers will not be Gold. Perhaps as an illustrative example, consider Platinum and White Gold. I assume that there is a common distinction between these two. We all agree that they are different, but Platinum and white Gold share many of the same properties. They are both pliable and malleable, they both have a silvery color, etc.. One last thing you mentioned was that it did not seem like "a bachelor is an unmarried man" and "water is H2O" have equal degrees of necessity. You had remarked that it seems in fact that water is H2O does not seem to be as necessary as being a bachelor and unmarried. Perhaps, this comes from the fact that bachelor, simply entails unmarried man, but it doesn't seem to be the same for water (and H2O). However consider the following: Water, white Vinegar, and Parafin. All three look the same. They are all liquid, they all evaporate, they can all make you wet, etc. Now it is clear that these three are different. I think we can see Kripke's point here. When water was found and initially dubbed Water, the liquid took that name necessarily (given a correct dubbing ceremony by competent speakers). Given that the liquid H2O took the name water necessarily, and that that liwquid was H2O, water is now necessarily H2O, just like Gold has atomic number 79. So, as I think Kripke would say, both the bachelor and example have the same degrees of necessity. One further thing to think about, and perhaps an issue that makes this a little unclear, is that the statement "water is H2O" is necessary a posteriori, while that of a "bachelor being an umnmarried man" is necessary a priori. This might be why it does not seem like the two statements have the same degree of necessity, but, I think that Kripke argues for these two modes of 'coming' to the necessary (a posteriori and a priori) quite forcefully. Anyways, I hope these comments have helped a little. -Larry From owner-modality@LISTSERV.ARIZONA.EDU Sat Feb 13 21:09:40 1999 Date: Sat, 13 Feb 1999 21:09:35 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Epistemic and modal issues To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Rachael raises some interesting issues re gold. To address them properly, we have to (as always) sharply distinguish Kripke's modal and epistemic points. His epistemic points concern what we "could find out" about gold in the actual world. His modal point concerns the nature of gold across possible worlds, given that it is a particular way in the actual world. The "gold" discussion in Lecture 3 is somewhat tricky because he uses different aspects of the case to make both modal and epistemic points, but I think one can still separate them. Maybe in this message, I'll take the opportunity here to say some general things about the distinction between the modal and epistemic points and about the structure of Kripke's discussion, in case it's useful for some people. This will also tie in to the 2-D framework. Then in the next message I'll try to address Rachael's discussion. It's sometimes tempting to run Kripke's modal and epistemic points together, because both make arguments against description theories of meaning. But the arguments concern very different issues. The modal point generally argues that for a name or natural kind term N and a description D, it is not *necessary* that N is D (there is a possible world in which N is not D). The epistemic point usually argues that it is not *a priori* that N is D (we could find out that N is not D). As we've seen, the points are quite separable, e.g. by considering cases such as "Jack the Ripper" or "Hesperus" where arguably the modal point holds but the epistemic point doesn't. In Kripke, the modal point is tied to the thesis that names and natural kind terms are rigid designators, picking out the same thing or stuff in all possible worlds. The epistemic point is tied to the causal theory of reference, holding that our names pick out their referent in the actual world not by description but by a causal chain. These two main themes in Kripke -- rigid designation and the causal theory of reference -- are easy to run together but should be kept well apart. How is all this reflected in the structure of Kripke's book? He goes back and forth a bit, but in essence, the central parts of Lecture 1 are devoted to modal issues and rigid designation, the central parts of Lecture 2 are devoted to epistemic issues and the causal theory, and Lecture 3 has some of each. In more detail (though still imperfectly), I'd divide things up as follows: Lecture 1 pp. 22-39: preliminaries (names and descritions, apriority, necessity) pp. 40-63: modal issues (rigidity: Nixon, meter stick, Aristotle, Hesperus) pp. 64-70: general issues re description theory (mostly epistemic) Lecture 2: pp. 71-73: general issues re description theory (mostly epistemic) pp. 74-78: modal issues (recap: Aristotle, Hitler) pp. 79-97: epistemic issues (non-descriptive reference: Feynman, Godel, etc) pp. 97-105: modal issues (necessity of identity, Hesperus=Phosphorus, etc) Lecture 3: pp. 106-116: modal issues (necessity of identity and of origins, essences) pp. 116-123: epistemic issues (natural kinds: gold, tigers) pp. 122-133: modal issues (natural kinds: gold, water, heat, etc) pp. 134-143: general considerations (both epistemic and modal) pp. 144-155: mind/body problem OK, now what about his discussion of gold, in particular. As suggested above, the first part of the discussion (esp. pp. 116-119) focuses on epistemic issues, and the second part (esp. pp. 123-125) focuses on modal issues. In the first part, he argues epistemically (contra Kant) that it is not a priori that gold is a yellow metal. Here he argues that we could find out (much to our surprise) that gold isn't a yellow metal, e.g. because there was a massive optical illusion. In the second part, he argues that given that gold has atomic number 79 in the actual world, it is necessary that gold has atomic number 79. Here he argues that in a counterfactual situation in which the ubiquitous yellow metal didn't have atomic number 79 (e.g. because it was all a yellow pyrite), it wouldn't be gold. Note that both these methods of argument consider hypothetical scenarios, but that we think about the scenarios in very different ways. In the first (epistemic) argument, we take a scenario (the optical illusion scenario) and consider what we would say if we found out that it was actual. In the second (modal) argument, we take a scenario (the ubiquitous pyrite scenario) that is explicitly counterfactual, and we consider how to describe that counterfactual scenario. Here there is no thought of "what to say if this is actual"; rather, we say "given that the actual world is as it is (with gold = element 79), how do we describe the counterfactual scenario". This difference can be naturally translated into the 2-D framework, so I'll spend a few paragraphs doing that, to help get a feel for that framework. To put things in the 2-D terminology, the first argument just mentioned requires us to *consider a scenario as actual*, and the second argument requires us to *consider a scenario as counterfactual*. Considering a scenario as actual involves considering it as a broadly epistemic possibility (a way the actual world may turn out to be); considering a scenario as counterfactual involves considering it as a "subjunctive" possibility (a way a world could be, given that the actual world is fixed). Kripke's epistemic arguments generally involve considering a scenario (e.g. the optical illusion scenario) as actual and thinking about how to describe it. Kripke's modal arguments generally involve considering a scenario (e.g. the ubiquitous pyrite situation) as counterfactual and thinking about how to describe it. It's the former that's relevant to what is a priori, and the second that is relevant to what is necessary, in Kripke's sense. Taking epistemic arguments first: Kripke often wants to argue that some statement ("N is D", "Godel proved the incompleteness of arithmetic", "gold is a yellow metal") is not a priori. To do this, he finds a scenario that he considers as actual (considers as an epistemic possibility), and argues that the statement is false when that scenario is considered as actual (he puts this by saying that if the scenario turned out to be actual, we'd say that the statement is false). So he argues that "gold is a yellow metal" is not a priori by arguing that it comes out false when the optical illusion scenario is considered as actual. This illustrates the general principle: if a statement is a priori, it is true in all scenarios considered as actual. Or to go all the way into 2-D terminology: if a statement is a priori, it has a necessary primary intension. The primary intension of a statement is defined as the function that takes a possible world and returns the truth-value of the statement in that world when the world is considered as actual. When we think about an epistemic possibility and think about what to say if it is actual, we are just evaluating the primary intension of a statement there. So when Kripke is arguing that a statement would come out false if a given scenario (e.g. the proof stealing scenario, or the optical illusion scenario) is actual, he is in effect establishing that the primary intension of the statement is false in that scenario. So the statement has a contingent primary intension, which corresponds to the fact that it is a posteriori. Doing this over again with modal arguments: Kripke often wants to argue that some statement ("N is D", "Hesperus is the evening star", "stick S is a meter long") is not necessary. To do this, he finds a scenario that he considers as counterfactual, and argues that the statement is false when that scenario is considered that way. So he argues that "Hesperus is the evening star" is not necessary by arguing that it comes out false when a given scenario (e.g. Venus was hit by a comet thousands of years ago) is considered as counterfactual. This illustrates the general principle: if a statement is necessary, it is true in all scenarios considered as counterfactual. Or to go all the way into 2-D terminology: if a statement is necessary, it has a necessary secondary intension. The secondary intension of a statement is defined as the function that takes a possible world and returns the truth-value of the statement in that world when the world is considered as counterfactual. When we think about a possibility "subjunctively" and think about how to describe it (given that the actual world is fixed), we are precisely evaluating the secondary intension of a statement there. So when Kripke is arguing that a statement is false of a given counterfactual scenario (e.g. the Venus-comet scenario, or the pyrite scenario), he is in effect establishing that the secondary intension of the statement is false in that scenario. So the statement has a contingent secondary intension, which corresponds to the fact that it is necessary. >From all this, we can see the crucial point that Kripke's epistemic points always (in effect) concern the primary intensions of our statements and terms, and that his modal points always (in effect) concern the secondary intensions of our statements and terms. For example, the modal thesis that names are rigid designators is in effect the thesis that names' secondary intensions don't correspond to descriptions, but pick out the same individual in all worlds. The epistemic argument against the descriptive theory of reference and for the causal theory in effect corresponds to the thesis that names' primary intensions aren't given by descriptions, but by picking out whatever lies at the other end of certain causal chains. OK, now back to the gold discussion. In effect, his epistemic point re gold (that it isn't a priori that gold is a yellow metal, e.g. because of the optical illusion scenario) is intended to count against any descriptive theory of reference-fixing for gold, and for some sort of causal account instead. His modal point re gold (that it necessarily has atomic number 79, e.g. by considering the pyrite scenario) is in effect intended to establish that a term like gold is in some sense a rigid designator (picking out the same stuff across counterfactual worlds). The theses are separable, as usual. In effect, the first concerns the primary intension of "gold" (it picks out what's at the end of a causal chain, not what satisfies a description), and the second concerns its secondary intension (it picks out a constant atomic kind, not what satisfies a description). In this way, he aims to show that natural kind terms are in *both* respects like names. Causal reference-fixing, and rigid designation. Corresponding to these distinct theses, we get somewhat different results for the primary and secondary intensions of "gold", as suggested above. To illustrate the difference explicitly, consider a possible world where there are superficially goldy pyrites everywhere in the mines and such, and nothing with atomic number 79 there. As Kripke notes in effect, when we consider that as a counterfactual scenario (with the actual world fixed), "there's gold in the mines" comes out false there, and "gold has atomic number 79" comes out true. (The pyrites aren't gold, element 79 is.) So, the secondary intension of "there's gold in the mines" is false in this world, and the secondary intension of "gold has atomic number 79" is true there. But when we consider that world as actual, we get different results. If our actual world turned out that way (with all the stuff in the mines around us being pyrite), then we'd say that gold turns out to be a pyrite, and that it turns out not to have atomic number 79. I.e., when we consider that world as actual, "there's gold in the mines" comes out true, and that "gold has atomic number 79" comes out false. I.e., the primary intension of "there's gold in the mines" is true in this world, and the primary intension of "gold has atomic number 79" is false there. So for each of these statements, their primary intensions and secondary intensions give quite different results when evaluated in this world. This reflects the fact that the secondary intension picks out atomic number 79 in all worlds, but the primary intension certainly doesn't. Rather, the primary intension picks out something like the substance at the other end of a causal chain from the center. As far as this example is concerned we might be tempted by something like "the yellowish substance in the mines around us" as approximating the primary intension, but Kripke's illusion scenario suggests that that may need to be modified a bit (whether there is any modified description that can do the job is still controversial). In any case, despite the anti-descriptive intention of both the epistemic and modal arguments, we can see that their upshots for the primary and secondary intensions of terms like "gold" is quite different. --Dave. P.S. I hope this, and my other messages re the 2-D framework, help to flesh out the discussion in the book and the paper. You can consider these messages part of the "readings" for next week's discussion of 2-D issues. Sorry if I've belabored some of these points, but they're really crucial to what comes later, so I'm hoping all this will soon be second nature, if it's not already. From owner-modality@LISTSERV.ARIZONA.EDU Sat Feb 13 23:52:03 1999 Date: Sat, 13 Feb 1999 23:10:50 -0800 Sender: "Philosophy 596B: Mind and Modality" From: David Chalmers Subject: Re: Proper Names To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Rachael writes: >I have some questions about the Kripke reading- specifically on pages >123-133. > On page 123 Kripke acknowledges that we could be mistaken about gold >having the atomic number 79. But he argues that given that gold does have >the atomic number 79, something could not be gold without having the >atomic number 79. Kripke makes light of Kant's mistaken assertion that >gold is a yellow metal- but it seems that it is *possible* that just as we >found out that all gold isn't yellow, we may find out that we are mistaken >about gold having the atomic number 79. Kripke argues that we can pick >gold out in all possible worlds by its atomic number. But if he was >arguing 300 years ago, perhaps he might have made a similar argument- that >gold is a yellow metal in all possible worlds. OK, there are both epistemic and modal points here. I'll take the above sentence by sentence. Kripke's initial acknowledgment is the epistemic point that it isn't a priori that gold has atomic number 79. His following argument is the modal point that it's necessary that gold has atomic number 79 (given that it has that atomic number in the actual world). Kant's "mistake" is the epistemic claim that it's a priori that gold is a yellow metal. Kripke argues that this is not a priori, e.g. by raising the epistemic possibility of an optical illusion. You respond by in effect making the epistemic point that it isn't a priori that gold has atomic number 79, either. That seems right, and Kripke would agree. He thinks (modally) that gold's atomic number is necessary, but he doesn't think (epistemically) that it is a priori. Note that Kripke's central point contra Kant isn't that we've found out that gold isn't a yellow metal (I don't think we've found this out!), but that we *might* find it out (witness the epistemic possibility of an optical illusion). Correspondingly, his modal claim that gold is necessary element 79 and not necessarily a yellow metal isn't based on the fact that gold has turned out to be element 79 and hasn't turned out to be a yellow metal -- as far as I know, it's turned out to be both. The point is that given the way things have turned out (in which gold is a yellow metal with atomic number 79, etc), it is necessary that gold has atomic number 79. Your last sentence raises an interesting issue. 300 years ago, might Kripke have said that it was necessary that gold is a yellow metal? Well, presumably if he'd lived 300 years ago he'd have had very different philosophical views, but let's set that aside by thinking of old-Kripke as having developed a similar abstract philosophy, but as having different empirical knowledge (he thinks that gold is a yellow metal, but doesn't know about atomic number). Certainly the mere fact that he thought gold is a yellow metal wouldn't make him think its yellowness was necessary: as above, the inference isn't from "gold has property X" to "gold necessarily has property X". Rather, it's from "gold has deep property X" to "gold necessarily has property X",m where the "deep" properties are some special class. Kripke-1970 thinks that gold's deep properties involve atomic number, not surface qualities. The residual question, then, is whether Kripke-1670 might have thought that gold's deep properties involved its yellowness. I suppose he might have thought that gold was a homogeneous stuff with not many further properties than its surface properties. If he'd thought that, he might well have thought that it was necessarily yellow. If on the other hand, he thought it had some unified deep microstructure, maybe he'd have thought it had that microstructure essentially, even though he didn't know what that microstructure was. Things are complicated by the fact that the whole idea of microstructure as the unified basis of surface properties hadn't really developed then. So it's hard to know what to say. But in any case, it does seem that if Kripke-1670 had said something different about gold, it would have been grounded in some ignorance about the empirical facts. Bringing the analogy back, Kripke-1970 probably shouid acknowledge that *if* he is wrong about the empirical facts about actual gold, then he could be wrong in his modal claims. For example, if it turns out that gold is a compound, then we might say it's that compound necessarily, not element 79. Or if it turns out that gold has some even deeper structure, "quark structure XY", then maybe we'll end up saying that gold has quark structure XY necessarily, and perhaps not atomic number 79. But all that depends on the empirical facts, which Kripke is happy to take at face value from the scientists. If the empirical facts turn out differently, the specific modal claims may change, but the general philosophical claims will stay pretty similar (the inference from "gold has deep underlying structure X" to "gold necessarily has X"). Could even more radical empirical revision make us change even the abstract philosophical framework? Maybe that's not impossible, but it's not easy to see exactly how it would go. > This brings up the question of what properties are essential. I >think the answer is more difficult to pinpoint in cases like 'what makes a >tiger a tiger' then for 'what makes gold, gold?' On page 121, Kripke >seems to suggest that the external appearance of a tiger is not essential >- we must look at its internal structure to determine whether it is a true >tiger. My question is two-fold. First, if we were to discover a >microscopic virus that happened to have the same internal structure as a >tiger- would we recognize it as a tiger? Second, Kripke asserts that >certain properties like quadrepedal, tawny yellow, and carnivorous are not >essential to tigers, we might find that all these properties are optical >illusions, but couldn't we discover the same of the internal structure of >tigers? Thq question of what properties are essential is a modal point. Kripke thinks that tigers have their DNA structure (or some such) essentially, i.e. that it is necessary that they have this structure. Your second point here brings together both modal and epistemic points. The claim that quadripedal/yellow/carnivorous aren't essential to tigers is a modal point: it's not necessary that tigers have these properties, since there are possible worlds where they don't. The idea that we might find out that these are optical illusions is an epistemic point: it's not a priori that tigers have these properties, since we might find out that they don't. I think you're right that we could find out that tigers don't have the DNA structure, but that's an epistemic point (it's not a priori that tigers have this structure) rather than a modal point, so it doesn't threaten Kripke's claim that tigers have the DNA structure essentially (given that they have it in the actual world). What if a microscopic virus turns out to have the same DNA structure as a tiger? That's an interesting one. Of course, as Larry points out, we're unlikely to find this out, and if we did discover it, it would cause our whole biological theory to be revised. My guess is that if we discovered this, we'd probably give up on the claim that DNA structure is the relevant deep structure, and so we'd give up on the claim that tigers have their DNA structure essentially: we;d have to find some other deep property that's responsible for the macroscopic differences. Whatever happened, it seems unlikely to say that we'd say that the virus is a tiger (though just possibly, there's some outlandish scenario where we would say they were members of the same natural kind, like carbon and diamonds?). There's also the question, what should we say about a possible world where there is something with the DNA structure of a tiger but looks and behaves like a virus. That's a purely modal issue now (we're assuming in the actual world, the DNA and tiger-ish properties go together). Presumably there are weird possible worlds like this. Do we call them worlds where tigers look like viruses, or not? My guess is that we wouldn't, and that we'd say the virus-ish things aren't tigers. If that's so, that means the property of being a tiger isn't identical to the property of having the DNA structure in question. It may still be that the DNA structure is necessary for being a tiger across possible worlds, but it isn't sufficient: perhaps one needs some sort of organic properties as well. E.g., one might argue that tigers are essentially animals, or essentially mammals, or some such, as well as being essentially that DNA structure (something can have more than one essentially property). This raises more general questions about just what the essential properties of natural kinds are, and it's hard to give a complete answer easily. Personally, I find there are many cases about which I don't have firm intuitions, and it may be that beyond a certain point the facts about which properties are essential become indeterminate, or terminological, or some such. > I guess I have a problem with how to identify natural kinds, >particularly in reconizing which properties are essential and which are >not. In light of past mistakes (like gold is a yellow metal) I am not sure >that we can ever know that we have actually discovered the essence of a >natural kind, or what picks it out in all possible worlds. Perhaps you >could say, given *our* concept of gold, what picks it out for us is the >atomic number 79. But if it seems logically possible to be wrong about >gold always having the atomic number 79 in this world, then why couldn't >we be wrong about it in other worlds? That's a good point. Kripke will agree that insofar as we can be wrong about what gold is made of in the actual world, we can also be wrong about what's gold in other possible worlds. All he really wants to commit to is the claim that *if* gold has atomic number 79 as its deep structure in the actual world, then it has atomic number 79 across all worlds (or something like that). Presumably, he does actually believe that gold has that atomic number, so he also believes that gold has that atomic number essentially, but he needn't claim that he is certain of either (and they certainly aren't a priori). What matters here is more the abstract philosophical structure, and the conditional connection between gold's structure in the actual world and its structure across possible worlds. (One might well argue that it's not a priori that gold has atomic number 79 as its deep structure, and its not a priori that gold necessarily has atomic number 79, but it *is* a priori that if gold has atomic number 79 as its deep structure in the actual world, then gold necessarily has atomic number 79. At least, Kripke's philosophical methodology seems to suggest that he regards that conditional connection between nonmodal facts about the actual world and modal facts about possible worlds as a priori. This is an instance of a general pattern that one might discern in Kripke's claims about a posteriori necessity. The necessary modal truths in question may not be a priori, but they are entailed a priori by certain nonmodal facts about the actual world. E.g., it isn't a priori that Hesperus is Phosphorus or that Hesperus is necessarily Phosphorus, but it is a priori that if Hesperus is Phosphorus (in the actual world), then Hesperus is necessarily Phosphorus. Similarly for other cases.) > As our science develops we discover that our original theories >about light and heat were mistaken. I'm not sure how this works- but it >seems like if enough of the data (what we recognize as heat and light) >don't match up with the theory- then we find a new one. It is still > possible to discover that heat is not just the motion of molecules- I >suppose that to accomplish such a discovery we would have to be using our >primitive notion of heat by which we identify it as a characteristic >effect produced on our nerve endings. If heat just *is* motion of >molecules then it looks like instead of reformulating the theory- we would >have to call that other kind of thing that produces a hot sensation in us- >something other than heat. It just doesn't seem that "heat is motions of >molecules" or "water is H2O" are as necessary as "a bachelor is an >unmarried man." What do you think? > -Rachael Here again, I think Kripke would agree that it's not a priori that heat is the motion of molecules, though it is necessary that heat is the motion of molecules (given that the motion of molecules causes heat sensations in the actual world). So we could find out that heat isn't the motion of molecules. If the scenario you describe turned out to be actual, I think Kripke would agree that we'd reformulate our theory, and say that the new thing that really causes the sensations is heat. Kripke would insist, I think that "water is H2O" and "heat is the motions of molecules" are as necessary as "a bachelor is an unmarried man". He'd say that the difference is just that the latter necessity is a priori, while the first two are a posteriori. I think there is a way of accommodating your idea that the first two are "less necessary" then the third, though. The third is both necessary and a priori. So in this case, the sentence has a necessary primary intension and a necessary secondary intension (it's true across all epistemic possibilities and across all counterfactual possibilities). Or in the terms of the book, it is both 1-necessary and 2-necessary. The first two, on the other hand, are necessary but a posteriori. So each of them has a necessary secondary intension but a contingent primary intension (they're not true across all epistemic possibilities). That's to say, they are 2-necessary but 1-contingent. (Or in the terms Evans uses, they are "superficially necessary" but "deeply contingent".) In effect, Kripke always identifies "necessity" with 2-necessity (necessity of secondary intension), so all these come out "necessary" by his lights. Still, the 2-D framework seems to isolate a reasonable sense in which the first two are "less necessary" than the third. --Dave. From owner-modality@LISTSERV.ARIZONA.EDU Sun Feb 14 10:07:15 1999 Date: Sun, 14 Feb 1999 11:06:23 -0700 Sender: "Philosophy 596B: Mind and Modality" From: Timothy J Bayne Subject: Re: Proper Names To: MODALITY@LISTSERV.ARIZONA.EDU Status: RO Regarding the following, see below: > Incidentally I heard someone a while back give a paper with the > thesis, contra received wisdom but in sympathy with your view, that > all uses of name tokens such as "Jane" are in fact tokens of the same > name. So there is just one name "Jane", one name "Paul", one name > "David Hilbert", etc. She addressed the problem of how the name gets > to refer differently in different cases by arguing that names are > implicit demonstratives -- in effect one is saying "that Jane", "that > David Hilbert", etc, with different intentions to demonstrate in each > case. A demonstrative like "that" can have different referents in > different contexts and with different intentions, but it's still one > word. Same on this view for "Jane". I'm not sure the view is > completely unproblematic, but it's interesting. > Burge seems to have advanced something like this view. According to a secondary source, Burge argued in a 1973 JP piece that proper names don't refer to unique individuals, but to a kind of individual, namely the kind with individuals of that name in its extension. The idea, I take it, is that "Alfred' is a kind name that refers to all the individuals called Alfred. Does this mean that *Alfred* changes its meaning when another person gets called Alfred? I.e., Alfred did mean [x, y ] when they only things called Alfred were x and y, but now it means [x, y, and z]? That can't be right. Perhaps the theory is that Alfred refers to the property (universal) of being such that Alfred is one's name - in which case (on most theories of universals), *Alfred's* meaning would not depend on the objects it was true of. *Alfred* would be a bit like *chair*, on this account. Being an Alfred would be a rather extreme nominalistic property. Burge also brings in a demonstrative elemenet to single out which Alfred one is talking about it a particular situation. Anyway, apparently Burge used the fact that such sentences as 'there are lots of Alfreds in this department' to support his claim. Not having read Burge's paper I don't have a settled opinion of his account. Tim Timothy J. Bayne RM. 213 Social Science Department of Philosophy University of Arizona Tucson, AZ 85721 USA Hm ph. (520) 298 1930