Thursday, July 21, 2011

Guest Post: An advertisement for Individual Concepts

[This post is by guest blogger Tristan Haze of Sprachlogik.]

We have concepts of particular objects. The recognition of this piece of common sense enables us to solve a cardinal problem in the philosophy of language. Millianism, the view that the semantic content of a proper name is simply its referent, is apparently refuted by Frege's Puzzle (cf. Frege 1952). The major alternative, descriptivism - which comes in several varieties - is apparently refuted by Kripke's famous arguments (cf. Kripke 1980). The very natural view that names are associated with individual concepts allows us to solve Frege's Puzzle and accept Kripke's arguments against descriptivism, as well as his rigid designation thesis, with remarkable ease.

The solution of Frege's Puzzle is that 'Hesperus' and 'Phosphorus' are associated with different individual concepts, hence 'Hesperus is Phosphorus' differs semantically and cognitively from 'Hesperus is Hesperus'. Individual concepts are not descriptions, nor clusters thereof, hence the view is compatible with anti-descriptivism. Furthermore, it is not only compatible with Kripke's rigid designation thesis, but predicts it: if names are associated with individual concepts - concepts of particular objects - then it is immediate that they will designate the same object in all possible worlds where that object exists; designating another object is out of the question, since we are holding fixed the meaning of the proper name - the associated individual concept.

The basic idea above can be developed in various ways, particularly in connection with modality, and I am working on a development at present. The purpose of this advertisement is to bring the basic idea, simply expressed, to the attention of the philosophical community.

The linguist Barbara Abbott has independently made some strikingly similar suggestions in her detailed discussion (Abbott forthcoming). See also Francois Recanati's work on mental files and identity (Recanati 1993).

References
Abbott, Barbara. 'Support for Individual Concepts', forthcoming in Linguistic and Philosophical Investigations vol. 10.

Frege, Gottlob. 1952. first pub. 1893. ‘On Sense and Reference’, in P. Geach and M. Black (eds.) Translations from the Philosophical Writings of Gottlob Frege, Oxford: Blackwell.

Kripke, Saul A. (1980). Naming and Necessity. Harvard University Press.

Recanati, Francois (1993). Direct Reference: From Language to Thought. Blackwell. (Also, look out for Recanati's more recent, as yet unpublished work on mental files and identity.)

-- Tristan Haze

15 comments:

  1. I'm not sure I understand this. What do you mean that individual names are *associated* with concepts? In what way are they associated? I also don't see how it solves Frege's puzzle. If by 'associated" you mean "refers to" or "designates" or something like that then I don't see how that could work at all.

    First of all, it's pretty clear to me that 'Hesperus' designates an concrete object, not a concept.

    Second, the 'is' in 'Hesperus is phospherous' is the 'is' of identity (is it not?). Thus if 'Hesperus' designates a different concept than 'Phosperus' then the statement is false, not true.

    Also, concepts don't seem to be the semantic content of proper names. For then what are the semantic content of conceptual names such as 'elephant' and 'red'? Would they be something like a superconcept? Are you saying that it's concepts "all the way down"?

    ReplyDelete
  2. Not all of us are convinced by Frege's argument:
    cf. http://tomkow.typepad.com/tomkowcom/2008/10/against-inten-1.html

    ReplyDelete
  3. Thanks for the comments. Nchen: by saying that individual names are tied to, or associated with, individual concepts, I *don't* mean that they refer to or designate them.

    In a word, this is meant to solve Frege's puzzle by predicting that 'Hesperus is Hesperus' and 'Hesperus is Phosphorus' express different thoughts.

    Compare Frege's view that names have senses. For Frege, names are associated with, or tied to, senses. You can simply say that they *have* senses, too. (That construction doesn't work so well with the phrase 'individual concept'.)

    My view is in many ways similar to this view of Frege's, and is motivated by some of the same considerations. The key differences are:

    (i) I do not hold that the sense or meaning of a name can be given in the form of a definite description.

    (ii) I do not hold that sense (e.g. individual concept) determines reference. (Cf. this post on semantic externalism.)

    So I agree with you that 'Hesperus' designates a concrete object, not a concept. (That should also deal with your second point, about identities.)

    Regarding your comment: 'Also, concepts don't seem to be the semantic content of proper names. For then what are the semantic content of conceptual names such as 'elephant' and 'red'?'

    First of all, I think the notion of semantic content is a highly theoretical notion which can play different roles in different sorts of account, and so I'm a bit wary about talking, without qualification, about *the* semantic content of a kind of expression. So it's not immediately clear to me what you're intuiting about here, so to speak.

    One can have one sort of theory on which the 'semantic value' of a proper name is just its referent, but this theory - unembellished - might have various limitations qua certain aims. And so one might also have other kinds of theories on which names are assigned concepts, or pairs of concepts and objects, or yet other things.

    Finally, you ask about 'elephant' and 'red'. Corresponding roughly to the three suggestions above for semantic values of proper names: these terms could be assigned concepts (or intensions), extensions (i.e. sets of things), or a pair containing both things. (Again, there will also be other possibilities.)

    ReplyDelete
  4. Sorry, '(e.g. individual concept)' should read '(or individual concept)'.

    ReplyDelete
  5. Hi Tristan, I'd be curious to hear more about how the individual concepts are supposed to work. In particular, in virtue of what is the cognitive significance of our 'Hesperus' concept distinct from that of our 'Phosphorus' concept? The Fregean picture offers a nice explanation of this, in terms of the associated descriptions. What's the explanation on your account?

    ReplyDelete
  6. If I understand you correctly, "concepts" as you call them function roughly the same way that Fregean senses do. It may be more precise to use the technical Fregean term "expresses" instead of "associate" to describe the relationship between names and your concepts. So 'Hesperus' expresses some concept (or Fregean sense perhaps). However, this seems to be a very unusual way to use that term and it seems to be a usage that is very different than concepts as is understood in Frege's understanding or common understanding of that "concepts" as concepts and concrete individual things are usually understood to be ontologically very different. I'm curious as to what specifically you mean by it.

    ReplyDelete
  7. Hi Tristan.

    After we e-mailed about these things I thought I should comment here as well.

    Firstly, I have the same worry as NChen does. Part of the debate here is whether the semantic value of 'Hesperus' is Venus, or some other worldly object, or something like a sense. Maybe we don't have good intuitions but there are (more or less convincing) theoretical arguments for the various options. So I'm not completely satisfied by the kind of response you gave to NChen.

    Here's another possible worry: One might think that it's a desiderate of views about this sort of thing that they allow for several subjects to share the same thought. So it might be worth asking how idiosyncratic the singular concepts are.

    Finally, I'd like to ask what you think about a sort of question that gets discussed in recent work by David Braun. (I've been talking about this with my colleague Andrea Onofri recently, although any silly mistakes in what follows are my fault not his.)

    Suppose Aisha is as rational, reflective, and competent with English as anybody. When she finds out a bit about astronomy (i.e. that Hesperus is Phosphorus) she comes to hold the view that the following can't differ in truth-value:
    (1) Hesperus is visible in the evening.
    (2) Phosphorus is visible in the evening.

    Then she finds out a bit about Babylonians. She comes to the view that the following do differ in truth-value:
    (3) Hammurabi believed that Hesperus is visible in the evening.
    (4) Hammurabi believed that Phosphorus is visible in the evening.

    This gives a semantic version of the Frege puzzle, and it shows that either she has false beliefs about the semantics of English, or its semantics is not innocent, or its semantics is not directly referential.

    It could be that the singular concepts view is just taking a Fregean way out. Then it's not really new, unless it's an improvement over traditional Fregeans somehow.

    Or perhaps the singular concepts view is trying to explain the psychological difference to which Aisha responds. That's a fine project, and one that neo-Russellians need to engage in. But it's not as if that's new either. Pragmatic neo-Russellians have been talking in terms of e.g. modes of presentation for years.

    Can you convince me that there's a genuine need for a new approach?

    ReplyDelete
  8. Thanks very much for all these questions!

    It seems like there are roughly three main directions in which clarification and further argument is wanted. (1) How does my approach solve various puzzles, principally Frege's? (2) What is the ontological/metaphysical nature of individual concepts? (3) What advantage does this approach give us over existing accounts (i.e. sophisticated Millians and Fregeans)?

    To be quite up front: I have nothing much to say in answer to (2), and I'm not sure how in general to deal with this sort of question. After all, we talk happily of all kinds of things, without being able to provide philosophically satisfying accounts of their nature. In this case, I feel like saying: 'You know! You've got lots. Individual concepts are the things which light up in your mind when you think of, or recognize, particular individuals you know.'

    That is hardly satisfactory: in some ways that says too much ('in your mind'), while in another sense it says too little. Its real function is perhaps something like a signal: don't get theoretical about this notion, you already have it, don't try to definite it in terms of others, just try to work with it. In a word, it's a primitive.

    Again, this isn't likely to be very satisfying, and there may be more specific questions in this vein which people will want to raise. Well, please do, but note that my overall strategy is roughly this: to defer (2) for as long as possible, giving such good answers to (1) and (3) that (2) loses its teeth, loses its stickiness.

    Given this, I don't think I have anything of further use to say in answer to NChen at present. I will respond to Richard and Tom's questions in a separate comment or two.

    ReplyDelete
  9. Hi Richard,

    '...in virtue of what is the cognitive significance of our 'Hesperus' concept distinct from that of our 'Phosphorus' concept? The Fregean picture offers a nice explanation of this, in terms of the associated descriptions. What's the explanation on your account?'

    As I see it, this question needs to be taken back a step. By that, I mean that Frege's story isn't first and foremost a way of answering that question, but rather this one: In virtue of what do 'Hesperus is Hesperus' and 'Hesperus is Phosphorus' differ in cognitive significance?

    To this question, Frege and I both have basic answers, and they are quite parallel. Frege says that these propositions differ in cognitive significance in virtue of the second one containing a different name 'Phosphorus' which has a different sense from 'Hesperus'. I say the same thing, with 'is associated with a different individual concept' in place of 'has a different sense'.

    In both cases, the difference of significance at the propositional level is made by one at the level of names. And here, too, Frege and I are quite parallel. 'Hesperus' and 'Phosphorus' differ in cognitive significance in virtue of having different senses, and in virtue of being associated with different individual concepts, respectively.

    However, because Frege holds that senses are associated with, or given by, definite descriptions, he gets a further level of difference-making, so to speak, and thus gets to say things to which nothing in my account corresponds.

    Frege uses senses as difference-makers, I use individual concepts. But Frege can also give difference-makers for his difference-makers. He can say something like: the sense of 'Hesperus' differs in cognitive significance from that of 'Phosphorus' in virtue of their being associated with different descriptions. But why do the descriptions have differing cognitive significance? There may be an answer in terms of further difference-makers for that question too, but at some point one has to stop cashing out the difference in terms of further entities which themselves have cognitive significance. On my account as it stands, we reach that point with individual concepts - and I say that's OK, because we're going to reach it pretty soon no matter what.

    So, to your question about what makes the cognitive difference between the concept of Hesperus and the concept of Phosphorus, I don't think I can, or need to, say much. These concepts are different concepts, and so they differ in cognitive significance. (I think I'd say that it's of the nature of concepts that when you have two different ones, they differ in cognitive significance.)

    What do you think?

    By the way, I am not saying dogmatically that this is all that could be said in answer to a question like yours. It might give rise to any number of interesting investigations and answers.

    ReplyDelete
  10. Hi Tom,

    It looks like my response has run over-length, so I'll post it in two parts. Part 1:

    'Part of the debate here is whether the semantic value of 'Hesperus' is Venus, or some other worldly object, or something like a sense.'

    First I want to emphasize that it's not a question of just those two options. Two other key options are: (i) something like a pair consisting of an individual concept (or non-reference-determining sense) and a referent, and (ii) a function from possible worlds to individuals.

    (i) corresponds to one natural way, among others, of individuating content - namely, where identity of intension (sense, concept) and extension is necessary and sufficient. (cf. this post.)

    'Maybe we don't have good intuitions' - as I see it, the issue is not so much whether our intuitions are good, as whether there is any single, definite thing we are intuiting about. And I really can't see that there is. The closest example to such an intuition which I can think of here is the feeling, which I am no stranger to, that full-blooded, naive, 'mere tag' Millianism "can't be right", is perhaps incoherent, some kind of strange blindness or confusion, etc. There may be other phenomena which are similarly broad. But I think we also need to be looking at more aim-relative intuitions, arguments, opinions etc.

    'but there are (more or less convincing) theoretical arguments for the various options' - in my view, the primary theoretical motivation for Millianism is essentially that descriptivism doesn't work (Kripke), and vice versa (Frege's puzzle). I accept both premises, which together imply that something other than Millianism or descriptivism is needed. Hence individual concepts: like senses, they solve Frege's puzzle, but unlike senses, they are compatible with Kripkean intuitions, and his thesis that names are rigid designators.

    Of course, descriptivism may also be motivated by the desire for some kind of account of "how reference is determined". Sometimes, this question about names gets tied up with questions about "how language gets to be about the world" - see, for example, the SEP entry on Reference. Obviously this is important, but I'd better not get too caught up in it now.

    The about reference-determination for this discussion, I think, is that my individual concepts are non-reference-determining (in keeping with Putnam's Twin Earth argument, semantic externalism etc.). This is bound up with the fact that individual concepts are Kripke-friendly, while Fregean senses are not.

    'One might think that it's a desiderate of views about this sort of thing that they allow for several subjects to share the same thought. So it might be worth asking how idiosyncratic the singular concepts are.' - In brief, I think there are lots of different ways to individuate thought/content. Given semantic externalism, there's the question of whether intensional (conceptual) identity, extensional, or both, are required. There's the type/token distinction. And, perhaps most relevantly, one can carve up types in different ways, and at different granularities. For obvious reasons, we often carve things up in such a way that two people are said to have the same thought, perhaps despite quite substantial cognitive differences. But in such a case we can also go finer-grained, and say they have distinct thoughts.

    ReplyDelete
  11. Part 2:

    Regarding what you call 'the semantic version of Frege's puzzle': isn't this essentially just the old problem, urged by Quine, of the apparent failure of substitution of coreferring singular terms salva veritate in intentional contexts? I.e. the problem of the existence of intensional-with-an-S contexts? Or am I missing something?

    Anyway, about Aisha from this case, you say 'either she has false beliefs about the semantics of English, or its semantics is not innocent, or its semantics is not directly referential'. I'm not sure exactly what 'innocent' and 'directly referential' mean here, but I think I would go for that last option. I think that substitutivity of co-referring singular terms salva veritate can fail.

    To explain how I make sense of this with respect to belief-reports, let me quote part of a comment I made elsewhere:

    'In short, my view is that the name 'Hesperus' in a belief report like:

    (A) 'Ralph believes that Hesperus is F'

    can be read as doing two things at once. (1) specifying the object of Ralph's belief, and (2) specifying the concept (or mode of presentation) via which he has it. On such a reading, (1) could be expanded to:

    (B) Ralph believes, of Hesperus, via his Hesperus-concept, that it is F.

    (A similar thing could be done for the 'F', but I'll keep it simple.) Some belief reports, on the other hand - purely de re belief-reports - may be read as only specifying the object. (A) read this way could be expanded to:

    (C) Ralph believes, of Hesperus, via *some* concept(s), that it is F.'

    (For the full comment see here or here.)

    You say: 'It could be that the singular concepts view is just taking a Fregean way out. Then it's not really new, unless it's an improvement over traditional Fregeans somehow.'

    I'm happy to call it a Fregean way out, if this expression just points to the fact that my solution of Frege's puzzle, and related puzzles, parallel's Frege's (cf. my reply to Richard above). One difference: I'm no Frege scholar, but as I understand it, about intentional contexts, Frege says something like: the sense of a singular expression becomes its reference. I don't say that. But the stuff about belief-reports above is quite reminiscent of this, only less problematic.

    As for the question of novelty, I don't want to make any claims to it, although of course I hope to be making some sort of progress on these issues.

    I think my view is an improvement over traditional Fregeanism. First and foremost, it embraces semantic externalism, and doesn't come with a questionable theory of reference. Kripke's arguments against Fregeanism don't apply to my view. I think that constitutes a significant advantage.

    Connected with this, I think my view puts us in a far better position to make sense of the modal phenomena drawn attention to by Kripke. If sense determined reference, it is hard to see how modal rationalism could fail to be true.

    I don't want to rule out all kinds of modal rationalism: there probably are useful concepts of necessity and possibility which are such that ideal conceivability does entail possibility, etc. But I do want to clarify a set of notions - metaphysical or subjunctive modal notions - which don't fit with rationalism. (The necessary aposteriori, the contingent apriori, etc.) For more on how individual concepts, as I understand them, help with this, cf. this post on modality.

    What do you think?

    ReplyDelete
  12. I'm sympathetic to a lot of that. If we can make sense of believing of Hesperus via one's Hesperus concept that it is visible in the evening that would be an advance. And it looks like individual concepts are a route worth pursuing.

    Is the following a fair summary? The semantic proposal isn't really novel. The psychological proposal (potentially) is. The novel psychological proposal might be used to give a better version of some more traditional semantic proposals, e.g. Crimmins & Perry's.

    I do have a concern about what you said in these two posts. At the beginning of the first you suggest that the semantic value of a name (in context) will be a pair of its referent and some salient individual concept. I suppose that you will need some story about how it gets to be that simple subject predicate sentences don't end up attributing properties to those pairs. I think there is also going to be an issue with 'Hesperus is Phosphrous', unless we think that those who believe that lack distinct Hesperus/Phosphorus concepts.

    Later in the second part you seem to suggest a different proposal. When you paraphrase (A) as (B) there is no reference to pairs. Maybe the pairs view isn't the way you want to go anyway.

    ReplyDelete
  13. Hi Tom,

    It's good to have some sympathy! You ask if it's a fair summary to say that my semantic proposal isn't really novel, and the psychological proposal potentially is. I'm not a historiographer of philosophy, and am not concerned to make specific claims of novelty, so I guess I don't know. Secondly, I'm not completely sure what you have in mind when you separate semantic and psychological proposals. Still, I have no real objection to the summary you give.

    When I say that names are tied to, or associated with, individual concepts (and add that these aren't given by descriptions, don't determine reference, etc.), this seems to suggest something both psychological (about our cognition) and semantic (in the sense that it suggests that individual concepts should become involved in semantic theory).

    'At the beginning of the first you suggest that the semantic value of a name (in context) will be a pair of its referent and some salient individual concept. I suppose that you will need some story about how it gets to be that simple subject predicate sentences don't end up attributing properties to those pairs.'

    I don't think I need a real story here, anymore than Frege needed a story for why names in ordinary contexts refer to their referents rather than their senses or something else, or anymore than Chalmers needs a story about why names don't refer to functions (or pairs thereof) from possible worlds to individuals. All that's needed, I think, is an insistence on the separation of the notion of semantic value from that of reference.

    'I think there is also going to be an issue with 'Hesperus is Phosphorus', unless we think that those who believe that lack distinct Hesperus/Phosphorus concepts' - OK, I *do* need a special story here. I say that someone who believes that Hesperus is Phosphorus is such that their Hesperus- and Phosphorus-concepts are functioning as aspect concepts, united under a common master. Think of the master concept as a single individual concept which has detachable parts, and which can be broken into multiple full-fledged individual (i.e. not merely aspect) concepts. So if they think 'Maybe I'm wrong, maybe Hesperus isn't Phosphorus after all' they are in effect considering making this separation. (This is all a bit sketchy, and has to do with the special nature of identity statements, or "the identity relation". There's a bit more in the modality post on Sprachlogik, linked to above.)

    'Later in the second part you seem to suggest a different proposal. When you paraphrase (A) as (B) there is no reference to pairs. Maybe the pairs view isn't the way you want to go anyway.' - I think we need to separate between giving a semantic value for an expression and paraphrasing it. I think lots of philosophically useful paraphrases are going to fail to mention the semantic values of the expressions involved in what is being paraphrased. So there's no obvious contradiction here.

    (This is perhaps a bit too hand-wavy, but: We could imagine a semantic theory capable of handling belief-reports and the like, where names are mapped to concept-object (or perhaps more generally, intension-extension) pairs. Since there are often no syntactic indicators for different readings (i.e. only extension specified (de re), both extension and intension, or only intension), intentional verbs would indicate the need for a contextual parameter making this decision. But whatever route is taken, all the relevant material will then be available in the pairs.)

    The full extent of the theoretical utility of intension-extension pairs is not something I've tried to work out, nor am I primarily concerned with the technical development of a semantic theory. But I think the idea has legs, and don't see any serious incompatibility between it and my paraphrases of belief-reports.

    ReplyDelete
  14. See this paper for an updated treatment of modality using individual concepts, as well as an updated exposition (informed by the above discussion) of my notion of individual concepts.

    The paper is called 'Toward an Understanding of De Dicto Subjunctive Modality'. My book's working title is now Necessity and Conceptual Systems.

    ReplyDelete

Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)