PHIL 3710/LING 3332: Philosophy of Language Fall 2008 Instructor: Classroom: Times: Office: Office hours: Phone: Email: Website: Wylie Breckenridge Uris Hall 369 TR 11:40 – 12:55 Goldwin Smith 320 By arrangement 5-6823 breckenridge@cornell.edu www.people.cornell.edu/pages/jwb274 (go to the ‘Teaching’ section) Overview: This course offers a broad overview of contemporary issues in the philosophy of language. Textbooks: Lycan, W. (2000), Philosophy of Language (Oxford: Routledge). Martinich, A. P. (ed.) (2008), The Philosophy of Language, 5th edn. (Oxford: Oxford University Press). Also helpful: Morris, M (2007), An Introduction to the Philosophy of Language (Cambridge: Cambridge University Press). Miller, A. (2000), Philosophy of Language (Oxford: Routledge). Platts, M. (1997), Ways of Meaning (Cambridge, MA: MIT Press). Stanford Online Encyclopedia articles Structure: Each week we will cover one chapter of Lycan’s book, which you should read in advance. In class we will discuss various key issues that Lycan covers, and possibly some that he does not. You will also be assigned a reading from the Martinich collection each week, your understanding of which will be tested by either a quiz or a written summary. Assessment: Participation (attendance, discussion), 10% Reading comprehension (quizzes, summaries), 30% Critical writing: one 7-10 page (double spaced) paper, 25% one 10-12 page (double spaced) paper, 35% Schedule: See over… -1- Schedule Tuesday Thursday Aug 26, 28 Introduction Sep 2, 4 Definite descriptions Russell: On Denoting Sep 9, 11 Names as definite descriptions Kripke: Naming and Necessity Sep 16, 18 Names as directly referring Evans: Causal Theory of Names Sep 23, 25 Traditional theories of meaning Frege: The Thought Sep 30, Oct 2 Grice on meaning Grice: Meaning Oct 7, 9 Verificationism Quine: Translation and Meaning Use theories of meaning First paper due Oct 14, 16 Oct 21, 23 Truth conditional theories of meaning Davidson: Truth and Meaning Oct 28, 30 Intensional semantics Stalnaker: Assertion Nov 4, 6 Semantic pragmatics Kaplan: DThat Nov 11, 13 Speech acts Searle: The Structure of Illocutionary Acts Nov 18, 20 Implicature Grice: Logic and Conversation Nov 25, 27 Dec 2, 4 Metaphor Second paper due -2- PHIL 3710/LING 3332: Philosophy of Language Week 3: Names as directly referring Possible worlds 1. Philosophers often appeal to possible worlds. Here is an argument that there are such things: things are a certain way, but there are various ways that things might have been; call these ways ‘possible worlds.’ 2. Robert Stalnaker takes there to be such a thing as the world, and that these ways are properties of the world. So possible worlds are properties of the world – they are different kinds of thing from the world (so it is misleading to call them ‘possible worlds’). 3. David Lewis agrees that there is such a thing as the world, but takes these ways to be worlds as well, of the same kind as our world. So possible worlds are more of the same kind of thing as our world. Our world is the actual world; the others are merely possible worlds. This position is called modal realism. 4. I will follow tradition and follow Lewis. Rigid designators 1. Designators (e.g. ‘London’, ‘The queen of England’(?)) designate things not just in the actual world but also in other merely possible worlds. 2. For a designator d to be rigid is for d to designate the same thing in every world in which that thing exists – for there to be a thing t such that d designates t in every world in which t exists. 3. For d to be strongly rigid is for d to designate the same thing in every world – for there to be a thing t such that d designates t in every world. 4. For d to be weakly rigid is for d to be rigid but not strongly rigid – for there to be a thing t such that d designates t in every world in which t exists, and such that there is at least one world in which t does not exist. 5. Note: If Williamson is right and existence is necessary, then rigid designators are all strongly rigid and there are no weakly rigid designators. What if Lewis is right about counterparts? Question: if d designates nothing in any world then is it rigid or non-rigid? 6. Important point: we are talking about what d designates in a world w when d is uttered in the actual world, not when it is uttered in w. The issue is not what d might have meant, but what it actually means. -3- 7. How can we tell if a designator d is rigid or not? Here is one way: d is rigid just in case there is a true reading of ‘d might not have been d’. Why is that so? Because the sentence has a true reading just in case there is a world w distinct from the actual world @ such that the thing that d designates in w is distinct from the thing that d designates in @, and that is so just in case d is not rigid. According to this test: ‘Wylie’ is rigid: ‘Wylie might not have been Wylie’ has no true reading. ‘The president of the US’ is not rigid: ‘The president of the US might not have been the president of the US’ has a true reading. 8. Note: a. The test only works for designators that actually designate something. b. The test tells us whether d is rigid or non-rigid, but not, in the former case, whether it is weakly or strongly rigid. Relativize to uses? 1. Perhaps we should be more careful and talk about rigid vs. non-rigid uses of designators. Perhaps on the true reading of ‘The president of the US might not have been the president of the US’, the first occurrence of ‘The president of the US’ is used rigidly, whereas the second occurrence is used non-rigidly. 2. Perhaps a clearer illustration: ‘The president of the US might not have been George Bush’. This has a true reading, on which ‘The president of the US’ is used non-rigidly, and a false reading, on which it is used rigidly. 3. But perhaps these are not rigid uses of ‘The president of the US’. Perhaps on these uses either of the following is going on: a. It is understood as short for ‘The actual president of the US’ b. It is only being evaluated at the actual world 4. What about names – do they have non-rigid uses? Here is some reason to think so: ‘In every class, Wylie is the smartest kid.’ Perhaps even ‘Wylie might not have been Wylie’ has a reading on which it is true. Maybe easier to get: ‘Wylie is not always Wylie’. An argument against descriptive theories of names 1. Kripke appealed to the fact that names are rigid designators to argue that they are not abbreviated definite descriptions: ‘Wylie might not have been Wylie’ has no true reading; -4- ‘The best philosopher at Cornell might not have been the best philosopher at Cornell’ has a true reading; So ‘Wylie’ does not abbreviate ‘The best philosopher at Cornell’ 2. This sort of argument might rule out some definite descriptions as candidates for what ‘Wylie’ abbreviates, but it does not rule out them all: ‘The actual best philosopher at Cornell might not have been the actual best philosopher at Cornell’ has no true reading, so perhaps ‘Wylie’ abbreviates ‘The actual best philosopher at Cornell.’ 3. The point is that some definite descriptions are rigid, and in fact some are strongly rigid: ‘The positive square root of 121’ 4. Perhaps we could also claim that names do abbreviate definite descriptions, even nonrigidified ones, but take mandatory wide scope. 5. Perhaps ‘The actual best philosopher at Cornell might not have been the actual best philosopher at Cornell’ does have a true reading. But if so, this seems to be due to there being two ways of understanding ‘actual’, and perhaps we can just stipulate that ‘Wylie’ abbreviates ‘The actual best philosopher at Cornell’, where ‘actual’ is understood one way rather than the other. Appeal to scope ambiguity? 1. Perhaps we can argue against ‘Wylie’ abbreviating ‘The actual best philosopher at Cornell’ by appealing to scope ambiguities: ‘The actual best philosopher at Cornell might not have been the actual best philosopher at Cornell’ is ambiguous between the following two readings (even if they are truth-conditionally the same): It is possible that: the actual best philosopher at Cornell is not the actual best philosopher at Cornell The actual best philosopher at Cornell is such that: it is possible that he is not the actual best philosopher at Cornell There is no such ambiguity in ‘Wylie might not have been Wylie’ So ‘Wylie’ does not abbreviate ‘The actual best philosopher at Cornell’ 2. But is it right that there no such ambiguity in ‘Wylie might not have been Wylie’? Direct versus indirect reference 1. Perhaps the difference we should be getting at is one of direct versus indirect reference. 2. There seems to be a difference between how we evaluate ‘Wylie is handsome’ and ‘The actual best philosopher at Cornell is handsome’ at other worlds. In the first case we -5- latch onto a guy in the actual world and carry him around from world to world to see if he is handsome at those worlds. In the second case we carry around a property for picking out a guy and see if the guy picked out in each world is handsome at that world. There is a difference in how we get a guy at a world, even if it is the same guy we get at each world. 3. Say that names directly refer, whereas definite descriptions indirectly refer. 4. Think of the difference as being between what they contribute to the propositions expressed by the sentences in which they occur. Names contribute particulars; definite descriptions contribute functions from worlds to particulars (or something like that). 5. So there are two distinctions for designators: rigidly referring vs. non-rigidly referring (with the former splitting into strong and weak); directly referring vs. indirectly referring: Directly referring Indirectly referring Strongly rigid Weakly rigid Non-rigid 6. Perhaps the argument against descriptivism should go as follows: names are directly referring, definite descriptions are not directly referring, so names are not abbreviated definite descriptions. Necessary a posteriori truths? 1. Kripke argued that there are such things: It is necessary that Hesperus is Phosphorus It is a posteriori that Hesperus is Phosphorus. Therefore, there is an x which is necessary but a posteriori: that Hesperus is Phosphorus But is this a valid argument? The causal theory of names 1. We have discussed four facts about names that are problematic for a Millian (i.e. direct reference) theory of names. 2. There is another question about names: When a name refers to an object, why does it refer to that object and not some other? Why does ‘Guy Pierce’ refer to that person and not some other person? 3. Descriptivists have a story to tell about this (but note: Russell would reject the question): it refers by fit. But non-descriptivists need to give some other story. -6- 4. Kripke proposes the following causal-historical picture: a. Person 1 uses the name n to refer to an object a by baptizing a with the name n. This is reference fixing. b. The reference may be fixed by ostension (‘Hesperus’). This requires standing in an appropriate causal relation to a. It may also be fixed by description (‘Neptune’). But the name does not thereby become synonymous with the description. Kripke allows that reference fixing by ostension is a special case of reference fixing by description. c. Person 1 can then go on to use the name n to refer to a. d. Person 2 can use n to refer to a by having an appropriate intention: an intention to use n to refer to whatever it is that Person 1 uses n to refer to. This is reference borrowing. Example: overhearing in the bar. e. Person 3 can borrow from Person 1 or Person 2. And so on. This generates a chain of reference preserving links. 5. Issues: a. b. c. d. e. f. Ambiguous names Naming after The imposter cat Numbers and other abstract objects Wagera Indians (Evans, p. 318) Change in reference: ‘Madagascar’ 6. Evans (‘The Causal Theory of Names’) is most concerned about Kripke’s version of the causal theory being unable to account for change in reference of a name. 7. Evans: Suppose a speaker believes that there is a unique thing which is F, G, H, and so on, and intends to refer to it using the name N. Then what he actually refers to with N is the dominant causal source of this belief. 8. Cases: a. I deduce that there is a tallest man in the world and that he is over 6ft tall. I name him ‘Bruce’. Evans allows that this might be accounted for by a description theory. b. The inventors of a legend about ‘Tony’ should not count as the referent of ‘Tony’. c. If I keep mistaking Bill for this guy then perhaps ‘Bill’ will start to refer to this guy. d. There can be deferential uses of names: Turnip. -7- PHIL 3710 / LING 3332: Philosophy of Language Questions for First Paper Due: Length: Worth: In class, Thursday 16th October 7 – 10 pages (double spaced) 25% of final grade Answer one of the following questions: 1. What theory of definite descriptions does Russell put forward in ‘On Denoting’? What does the theory have going for it, and what problems does it face? 2. What are Kripke’s arguments against descriptive theories of names in Naming and Necessity? How decisive are they? 3. In ‘The Causal Theory of Names’, Evans argues that Kripke’s causal theory of names faces a problem to do with reference change. What is the problem? How does Evans think the problem should be solved? 4. In ‘The Thought’, Frege describes things that he calls thoughts. What, according to Frege, is a thought? Why does he claim that they are not ideas, and why does he claim that they are not things of the outer world? 5. In ‘Meaning’, Grice arrives at an account of what a speaker means by a sentence in terms of what the speaker intends. Describe the considerations that led him to this account. Can you think of any counterexamples to this account? I shall be pleased to see you using clear, concise, and short sentences, and making it excessively clear what you are doing at each stage of the paper and why you are doing it. Pretend that I am lazy, stupid, and uncharitable. -8-