Short answers only for spring 2003 questions. All typos/thinkos are mine, due to lack of sleep. ----- 1. Is conceivability a reliable guide to possibility? I do not think that conceivability is a reliable guide to logical possibility. That is, I do not believe that conceivability entails possibility. I will try to clarify my notion of possibility and present a strategy for producing counterexamples. It seems that Descartes and Hume believed in the entailment; Descartes relies on it in his sixth Meditation; Hume says that nothing we can imagine is absolutely impossible. What Descartes' reliance valid? Yablo, Pryor, Chalmers, and many others seem to call this into question. I fall into their camp, believing that conceivability need not imply possibility. It really does seem that the fact that I can conceive something does not mean it is genuinely possible. So I just need a counterexample. That is, what's something we conceivable but not possible? What is meant by possible? Not physical possibility or conceptual possibility (the ability to be conceived), but metaphysical/logical possibility. What is something logically impossible? Not darned much; basically things which are self-contradictory (3 sided square, A and not A, anything contradicting definition and supervenience). Is a 3 sided square a counterexample? If you think you can conceive on sure. But most people can't conceive such a thing. Here's a much better example: Consider a claim like Goldbach's conjecture (unknown whether true or false). We can conceive that it is true. Similarly, we can conceive that it is false. But surely it is either true or false, not both. In fact, there is a necessary truth/falsity (which means the negation is not possible). So that's something we can conceive of something (namely one of its affirmation or denial) that is impossible. More generally, unknown and a posteriori truth propositions seem to lead to a problem. Consider "Hesperus is Phosphorus". Following Kripke, this is an example of necessary (thought not a priori) truth. Clearly at one point it was thought/conceived that H was not P even though this is impossible. So it seems conceivability doesn't entail possibility. Nor is the converse true, for things could be possible without being conceivable. Conceivability and logical possibility (as opposed to conceptual possibility) are very different notions, and while they do align with some frequency, conceivability is in no way an infallible guide to possibility. ----- 2. Can questions concerning numerical identity plausibly be settled by convention? Are they sometimes merely caused by confusions about the meanings of our words or the context in which we find ourselves? What is lost by understanding numerical identity in any of these ways? ----- 3. It is sometimes said that truth is relative to a conceptual scheme, or a version, or some other such thing. What, if any, good reasons are there for saying such a thing? Are there good reasons to deny this? Relativism about truth, that truth is relative to a conceptual scheme, is a somewhat common view. To some extent, it may be considered to be the view of Putnam and Davidson. Putnam, drawing on Quine, takes the moral of indeterminacy/inscrutability of reference to be that reference is always relative to a background structure or conceptual scheme. Reference doesn't tie words to items of the mind-independent outside world. Same for truth. It makes sense to talk about these within a scheme, but not without. It seems right to say this because this seems to be how we think. Our conception of reference/reality is mediated by our categorizing principles, background theories, language, etc, so there are likely to be multiple possible schemes (for different people/theories). Also seems right to say this because this seems like the only way we can get at truth, through some sort of epistemic method; if it's mind-independent, how do we get at the correspondence to the outside world? It also seems wrong to say truth is relative because precisely what we want out of truth is that it is objective, mind-independent (at least for the Realist). And how can this be if it is relative to a scheme? ----- 4. Are there fictional objects? Sketch and assess the most plausible responses to this question. Are there fictional objects? Why ask such a thing? Because it's clear we want to be able to talk about Hamlet, Holmes, Jupiter, Pegasus, etc, and not only talk about them but assign truth-values to propositions about them. Consider examples: Hamlet is unmarried. Holmes is the greatest detective. Augustus worshiped Jupiter. Pegasus doesn't exist. What is being said here? Are there commitments to the existence of fictional objects? Quine, Russell, Plantinga, among others say we need not make ontological commitments to fictional objects. Parsons, following Meinong, thinks there are fictional objects. Quine considers "Pegasus doesn't exist." He doesn't think this is saying of Pegasus, some object, that Pegasus doesn't exist. There's no such commitment. Instead, Quine says Pegasus doesn't name/refer to an object, but to a description. Following Russell, Pegasus can be paraphrased away (roughly, x such that x Pegasizes). Quine distinguishes between meaning and naming (gulf between them). We can understand the meaning of the sentence without some fictional object Pegasus. There's no need for Pegasus, so there's no Pegasus (desert landscape). Parsons seems more fixed in our intuitions about every day speech. Referring to fictional objects isn't failing to refer. When we talk about Holmes there is something we are talking about --- Holmes. He concedes that maybe the objects can be paraphrased away, but there is no general strategy for paraphrasing that he accepts. Clearly, Quine's theory depends on such a method of paraphrase. What do I think? I think if we can make sense of claims about "fictional objects" we don't actually need them to exist. But we do need something, maybe like Meinong's subsistence I guess, but I'd prefer to say definedness. Holmes (unlike possible men) is defined. Namely, Holmes has a description, so we can say statements involving him are make-believedly true/false following (Evans). ----- 5. Can a useful distinction be drawn between natural and non-natural properties? Can a useful distinction be drawn between natural and non-natural properties? What are natural/non-natural properties? How would the distinction be useful? How could it be made? Following Lewis, natural properties are a special subset of properties. They roughly correspond to Armstrong's sparse Aristotelian universals. Less roughly, they correspond to instantiated, undefined, intrinsic universals. Such properties/distinctions Lewis believes (and I'm inclined to mostly agree) are very useful: useful for duplication, resemblance, supervenience, divergent worlds, materialism, causation, causal laws, language. For instance, two things are duplicates iff they have the same natural properties. Causal laws involve regularities of natural properties. As good as they sound, Taylor provides an objection. The joints are mysterious; where do we carve the natural properties? Our decisions seem to be grounded in human classificatory practices, not any absolute mind-independent natural properties. This seems like a real problem for Lewis' theory, but still, if we could find natural properties, they sound pretty useful. ----- 6. Can problems in metaphysics -- personal identity, the dispute between three and four dimensionalists, differing views about essence and accident, etc -- be resolved? If so, how (and how will we know when they are resolved)? If not, what is the point of analytic metaphysics? ----- 7. How do we rank the properties possessed by an object into those (if any) that are essential and those (if any) that are accidental? Do you believe that this presents a difficulty for the essentialist? The essentialism holds that an object have essential properties. Then it seems the essentialist needs to provide an account of what these properties are. What are essential properties? They are necessary, such as a bachelor being unmarried. Accidental properties, in contrast, are not necessary. (Modality de re, possible worlds.) One and the same (3D) object can persist through change in accidental properties, but it must have the same essential properties. This is why the distinction is so important/useful. It explains persistence through change. But which properties are essential? Some, most famously Leibniz, say all of an objects properties are essential to it (individual essence). Some, take the other extreme and say none of an objects (nontrivial) properties are essential. They could change over time or be different in different possible worlds. Others take a moderate position that some properties are essential and some are accidental. This seems more or less right to me (in fact, I think its likely essential properties have to do with how an object is fixed by reference). But still, which properties are essential? This question is asked by Plantinga (Could Socrates have been an alligator (a rational one)?), Kripke (considering a lectern made of wood or ice), and Chisholm (biblical reference). Of these only Chisholm seriously considers an answer. Develop Adam/Noah example: If Noah and Adam can be the same in possible worlds through minor change, eventually Noah becomes Adam and vice versa, but this seems wrong. So what's the safeguard? Say some of their properties are essential and cannot be changed. If Adam has unique essential properties (and so does Noah), there's no way they can be confused. We can pick out one individual across multiple possible worlds because the essential properties are fixed. But Chisholm points out that even if Adam and Noah do have some essential properties (as it would be useful for them to have), we have no procedure for finding out what they are. (I sure can't think of any; we'd have to be able to inspect them in all possible worlds.) Furthermore, Chisholm says we have no way of knowing whether there even are any essential properties. This is a big problem for the essentialism who is left with the burden of proof in explaining how to identify (and know that one has identified) essential properties. ----- 8. Mereological essentialism, the position defended by Roderick Chisholm according to which an object cannot gain or lose any of its parts, strikes most of us as quite implausible. But given his distinction between identity in the loose and popular sense and identity in the strict and philosophical sense, how do you go about objecting to this doctrine on philosophical grounds? According to Chisholm an object cannot gain or lose any of its parts; it has them all essentially. Paralleling essential properties, Chisholm is saying that objects have essential parts; not only that, all parts are essential. For the 4Dist (object made up of temporal parts) this leads to the conclusion that, for instances Descartes couldn't have lived for a longer/shorter time than he did. Even in 3D examples, mereological essentialism seems problematic. Surely a cat can be a cat without its tail, and even more surely living organisms survive constant change in cells. Why then might Chisholm say such a thing? The answer I think is that he sees change in parts as inconsistent with Leibniz' Law (if A and B are the same thing they have all the same properties/parts). And Chisholm doesn't want to throw LL out. But often this isn't what we mean by identity. We want to say that things persist in change through property (old and young tree). And Chisholm realizes there is a kind of loose identity. Here he's following Butler and Hume/Reid. Butler says a tree that alters its composition retains its identity only in a loose and popular sense, not a strict and philosophical sense. Chisholm considers possible meanings for the loose is of identity (part of, same kind as, relative identity, holds position of, etc); none of these are as strict as the kind of identity he wants. He considers Hume's claim that we feign identity when it is convenient. We create logical/fictional constructions of, for example, a tree, which is the succession of parts connected by such things as resemblance, contiguity, and causality. This (something like a 4D view almost) seems to prevent us with a way to explain persistence through change, which seemed to me like the biggest problem with mereological essentialism. So my motivation for objecting to Chisholm disappears. But it can't be quite this easy. Persistence through change in parts can't just be explained through loose identity. Or can it? Isn't this all we want out of, for example, the ship of Theseus? If loose identity is really all we want, what purpose does the strict identity serve, and why does Chisholm cling to it (and mereological essentialism)? This is where Chisholm needs a further response. ----- 9. Does any plurality of objects, no matter how disparate and dissimilar, compose a single additional object, their mereological sum? Does any plurality of objects compose a single additional object, a mereological (part/whole) sum? There are three possibilities: 1) there is no mereological composition; 2) there is some mereological composition, but it is restricted; 3) there is unrestricted mereological composition. The first seems clearly wrong as we can provide many counterexamples (a house made of a bricks, a deck of cards made up of cards, an organism of cells). Consider the third. For any filled region there is an object. This commits us to infinitely many objects, some of which sound rather odd, like Cartwright's object composed of the Eiffel Tower and the Old North Church or can Inwagen's object composed of two people shaking hands. But is this really a problem? Why can't there be infinitely many such objects? I don't think is really a problem, but if we do accept this, I think we need a new notion of special composition to tell us when there are everyday mereological objects (like the house, but not the tower-church). The second option seems to best fit our intuitions. We can come up with some instances of mereological composition, and we want to deny others. But there is one difficulty with defending this position (and moderate/compromise positions in general). We want some composition and not all, but where do we draw the line? As Lewis says, such borders seem indeterminate/vague. We may want to retreat to the third position here, as it seems the most defensible. Any filled region can be validly referred to as the object of a reference. We just need to add the qualification that not all of these are everyday objects, and as I've said before, we still want a way to pick out such objects. I can only come up with some loose guidelines. We don't want to say that an object must be connected; the deck of cards isn't, nor really is anything on the atomic/subatomic level. It does seem though that we can use proximity as part of a set of rough guidelines. Cells in an organism are near each other. We can also use function here; an organism has a common function, as does a house. A house also has structure. We also want to say something about similarity; cards are similar to each other. That's the best I can offer, and that doesn't settle the vagueness of the middle position. So I maintain any filled region contains an object, no matter how weird it may be. We can use the guidelines I developed to help us in our everyday understanding, but not such much in our philosophical endeavors. ----- 10. Our talk is vague. And so there can be vague identities, in so far as it is a vague matter as to whether two terms name the same thing. But could identity itself be a vague matter? Present and adjudicate among representative answers to this question. Could some identity questions themselves be vague or indeterminate? Following a short paper by Evans, this question has been the focus of much attention. Some, like Parsons borrowing from Chisholm, say yes. Others, like Lewis, Sorensen, and Garret) say no --- vagueness is only reference/semantics. Parsons thinks that there are examples of identity questions that are genuinely indeterminate. One example is Theseus (Chisholm). If a ship can survive change in parts and a new one is built with remains of old, there are two ships; Parsons concludes there is no fact of the matter as to which of the two ships is the original ship. Similarly, Parsons considers person/body and a cat made of many parts (a variation on Tibbles). But do these present a genuine problem with identity? Parson seems to think so, and these are very hard questions. But many others do not, and I'm inclined to agree. The problem seems to be, as Lewis says, semantic indecision (he's talking about objects, but I think the same applies to identity). Also against Parsons' view, Sorensen argues that there is no genuine vagueness presented by those identity puzzles. Vagueness is in thought/language. Following Quine, what appears to be a problem with identity is a problem with how the speaker fixes reference. It seems like something like this can resolve at least the problem of Theseus. One ship materially is the original ship; the other ship continuously is the original ship. The only problem, was that the question as to which was the original ship was not clearly stated. ----- 11. What relation must my beliefs and desires have to my choices, and my choices to my actions, if my action is to be reckoned to be free? While some people believe that determinism is a threat to free will, I believe in a form of strong compatibilism where agent-determinism is required. That is, the agent of an action must be the determinor of the action for the action to be free. Agents/people have liberty to act in a certain way and freedom from constraints in acting in accord with liberty. This is what is needed for "the kind of freedom worth wanting" (moral responsibility, punishment, etc). This follows Hume and Locke. Locke says "to be determined by one's own judgment is no restraint to liberty." So what relation must beliefs and desires have to choice? We must be able to think freely, rationally, not brainwashed, insane, etc. Judgment requires examination, suspension of one's beliefs, and rationality. ----- 12. Can properties be individuated by their causal powers?