Philosophy 125: Metaphysics
[ Home ] [ Syllabus ] [ Sections ] [ Lecture Notes ] [ Study Questions ] [ Paper Topics ] [ Links ] [ Pictures ] |
Given that sets have their members necessarily, the trope theorist is committed to the claim that the set that is courage could not have had a different membership.... On the trope theorist's account, however, concrete objects, persons, are courageous just in case they have a trope that is a member of the set that is identical with courage. But if that set could not have had more or fewer members than it does, we have the result that there could not have been more or fewer courageous individuals than there, in fact, are.
Carefully reconstruct this argument. Then, offer some criticisms of the argument. Do you think this is a good argument? Does it really expose a serious problem for the set-theoretic trope theorist?
Now, assuming that "Courage is a virtue" is true (as it is), we can deduce the following:
3. "Socrates is courageous" is true only if "Socrates is virtuous"
is true.
Explain how this follows from (1) and (2), on the assumption that "Courage
is a virtue" is true. The fact that (3) can be derived
from the conjunction of the trope-theoretic accounts of subject-predicate
discourse and abstract reference is very bad news. To
see this more plainly, note that, since we can run this same derivation
for all courageous
people (explain how this would go!), we can then deduce from the
conjunction of these two trope-theoretic accounts that:
4. All courageous people are virtuous.
But, this claim is false, as we have seen in our discussion
of Austere Nominalism! Remember, it was because (4) is false that
the Austere
Nominalist had to bring in the primitive "ceteris paribus" clauses
in their
account of abstract reference. We have just shown that the conjunction
of the trope theorist's accounts of abstract reference and subject-predicate
discourse has a false consequence. Therefore, it appears that they
can't both be true. Discuss.
- Loux (page 106) says:
Although the central premise of the argument ("Difference of attributes entails difference of bundles") was formulated in bundle-theoretic terms, that premise is merely an instance of a more general principle governing the constituent-whole relation; for if it is true that difference in attributes entails difference in bundles, it is true only because it is true that difference of constituents entails difference in constituted wholes or complexes. But the substratum theorist no less than the bundle theorist construes the attributes associated with an ordinary object as its constituents [viz., (PCI)]. Accordingly, if the bundle theorist is committed to denying that the concrete object emerging from a change is ever numerically identical with that entering the change, so, it would seem, is the substratum theorist.
This suggests that the problem with identification through changes in attribute is the very same problem for both the bundle theorist and the substratum theorist. But, intuitively, is that right? What if a particular changes all of its attributes? Would there, on a substratum view, be anything that (in some sense) could persist through such a global change in attributes? What about on the bundle theory? [Also, see questions #2 and #3, below. As in #3, below, both (PCI) and its converse – the constituent-indiscernability of identicals – will be relevant here.]Loux (page 110) says:
… if subject-predicate discourse presents problems for the bundle theorist, it presents analogous problems for the substratum theorist. The substratum theorist claims that substrata are the items to which we ultimately ascribe attributes; but, then, substrata had better be things we can pick out as identifiable objects of reference. [but] … substrata are bare; they are things that in themselves have no attributes.…there is nothing in a bare substratum, taken by itself, that would enable us to pick it out as something distinct from other things. If a bare substratum is to be identified, it can only be by reference to the attributes with which it is compresent. Those attributes, however, are just the attributes that can be truly ascribed to it. But, then, the substratum theorist would seem to confront the same sorts of difficulties he poses for the bundle theorist.Do you think – as Loux suggests here – that this problem is the same problem faced by the bundle theorist? Or, is there a deeper problem here for the bundle theorist? [Hint: distinguish the metaphysics of identity (as it pertains to subjects) from the epistemology of identification (as it applies to subjects). There does, intuitively, seem to be a "part" of the identity of the substratum theorist's objects that is independent of the attributes it happens to have. See question #3, below.]
Explain what Loux means when he says:
…if the attribute did not enter into the constitution of the object, that object would not exist. On the bundle theory, every true subject-predicate claim is a mere elaboration of the essence of a concrete object. And here we confront what is, perhaps, the central difference between between the bundle theory and the substratum theory; for whereas the bundle theory must construe all true ascriptions of attributes as holding of necessity, the substratum theorist insists that none does.
This is supposed to be the crucial difference between bundle theory and substratum theory. Do you agree with Loux that a bundle theorist must think of all the attributes of a particular as necessarily exemplified by that particular, and that the substratum theorist must disagree (and view attributes as contingently exemplified)? How does this jibe with the fact that Loux also claims that both the bundle theorist and the substratum theorist is committed to principle (PCI), on page 113? Doesn't changing attributes imply changing the constituents? If so, why doesn't this imply a change in identity both for the bundle theorist and the substratum theorist (after all, they both accept (PCI))? [Note: everyone who accepts (PCI) is bound to accept its converse – the constituent-indiscernability of identicals – which says that identity entails indiscernibility with respect to constituents.] If it does, then how can what Loux says in this quote be true? Wouldn't it then also be the case from the point of view of the substratum theorist that "if the attribute did not enter into the constitution of the object, that object would not exist"?- After discussing the bundle theory, which he claims requires particulars to have all of their attributes necessarily, and substratum theory, which he claims requires particulars to have none of their attributes necessarily, Loux (p. 123) describes the choice between these extremes as follows:
If the idea of an entity completely lacking in essential attributes is, as it seems to be, problematic, then the substratum theory is not an attractive option. And if it is, as it seems to be, possible for numerically different concrete objects to be qualitatively indiscernible, then any version of the bundle theory that endorses a realist interpretation of attributes would appear to be unacceptable. It looks as though we have only two options: to join forces with bundle theorists like Hume and Williams who embrace a trope-theoretic interpretation of attributes, or to follow the austere nominalist and deny that concrete particulars have any ontological structure…
That is, Loux characterizes the choice between substratum theory and bundle theory as a choice between a theory which requires that no attributes be essential (or necessary), and a theory which accepts the Identity of Indiscernibles (II). Can you think of a reason why he doesn't express the dilemma here entirely in terms of the disagreement over how many attributes are necessary or essential? In fact, why does he bring in (II) at all? Isn't there a much easier way to make the choice of bundle theory seem unattractive – for both realists and trope theorists (and in a way that is directly related to his complaint about substratum theory), simply by appealing to the fact that it makes every attribute essential? This is especially puzzling, since the (II) argument doesn't apply to trope theory versions of bundle theory, but the "superessentialist" objection applies to both.
- The question at the very end of my handout on rigidity, abstract reference, and predication, which is as follows. On Loux’s reconstruction of the bundle theory, however, all attributes of particulars come out necessary (and for the same reason – Loux assumes that “the set of ___ tropes” is always rigid). Can a trick like the one I applied above to the trope theory of universals be applied to the trope/bundle theory of particulars? If so, what are the consequences of such a move? In particular, is it possible for Socrates to be contingently courageous on such a view? And, if so, how (that is, why is his courage contingent on such an account)?
- Explain why the bundle theorist is prohibited from appealing to spatio-temporal properties when trying to identify a distinguishing property of the (presumed to be distinct) spheres in Black's universe. That is, explain why such properties are "impure". Do you think the bundle theorist should even try to argue that the spheres do have a distinguishing property? Or, do you think the better strategy is for them to simply turn the tables on the substratum theorist, and try to argue that they are really one and the same sphere after all? Along these lines, consider the replies of Hacking and Zimmerman.
- Loux claims that the bundle theorist errs in "…restricting the attributes relevant to the characterization of a concrete
particular to those the realist calls properties." Why can't the realist bundle theorist include kinds (or relations) among the universals it considers to be constitutive of particulars? Does Loux provide an argument for this claim anywhere? If so, examine it, if not, can you think of an argument for this claim that might be suggested by Loux's remarks about bundle theory?
- Loux claims that:
Aristotelians deny that there is a special problem of explaining how concrete particulars can be numerically different from each other. They insist that the multiple instantiation of a kind is, by itself, sufficient to secure the existence of numerically different particulars. Each of its instantiations is a particular that is numerically different from the others.
How is this supposed to work, exactly? Try to give a precise rendition of how the "multiple exemplification of a kind" is supposed to explain how (II) can be false. This is connected with question #7, above. Loux seems to be assuming that kinds do not count among the universals in the indiscernibility clause of (II). But, even if this is assumed, it is not entirely clear how "multiple exemplification of a kind" is going to help establish that a and b are distinct (for some a and b). After all, if a = b, then, presumably, a belongs to kind K if and only if b does. So, how does the fact that a and b "both" belong to K imply that a and b are distinct? What else is being assumed here?