Intrinsic Properties and Natural Relations John Hawthorne Philosophy and Phenomenological Research, Vol. 63, No. 2. (Sep., 2001), pp. 399-403. Stable URL: http://links.jstor.org/sici?sici=0031-8205%28200109%2963%3A2%3C399%3AIPANR%3E2.0.CO%3B2-O Philosophy and Phenomenological Research is currently published by International Phenomenological Society.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/ips.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact
[email protected].
http://www.jstor.org Wed Jun 6 07:27:06 2007
Philosophy and Phenomenological Research Vol. LXIII, No. 2, September 2001
Intrinsic Properties and Natural Relations JOHN HAWTHORNE
Syracuse University
Assuming that we find the concept of naturalness coherent, we shall no doubt wish to allow that certain relations count as highly natural. Many of us will think that various spatio-temporal and causal relations-is the cause of, is spatially separated from, is later than-are highly natural. Some of us will think that various basic semantic and mentalistic relations-refers to, attends to, believes ....---are highly natural. Some will think that various logicomathematical relations-being the successor of, being identical to.. .-are highly natural. Consider a highly natural relation R. Given the existence of R, there will also be a monadic property of bearing R to something. Call this the existential derivative of the relation. If R is highly natural, the existential derivative may not be quite as natural. But it will be pretty natural. For example: if loving is a highly natural relation, then being a lover is a pretty natural property. Some relations can only hold between a thing and itself. Other relations can only hold between a thing and something that is neither itself nor part of itself. Other relations are more flexible: Consider the relation attending: One might attend to oneself. One might attend to part of oneself. One might attend to something that is neither oneself nor part of oneself. The existential derivative of the attending relation is thus independent of accompaniment. The existence of a contingent thing that is wholly distinct from me is compatible with my attending to something and also with my failing to. The absence of a wholly distinct contingent thing is similarly compatible with both the presence and absence of the existential derivative. Some existential derivatives may be independent of accompaniment in less obvious ways: Perhaps the causal relation is independent in this way because it is possible to be self-caused, or else because it is possible for a thing to cause part of itself. Perhaps distance relations are independent because it is possible for a thing to be bilocated and hence at a distance from itself. It need not be our concern here to figure out which existential derivatives are
SYMPOSIUM
399
independent of accompaniment. For present purposes, it suffices if we agree that there are some existential derivatives that are (a) pretty natural and (b) independent of accompaniment. I shall persist with the example of attending to something, though the reader should view this as a placeholder for whichever existential derivatives she thinks fit the bill. The property of attending to something is not an intrinsic property. Suppose I am attending to nothing but a chair. A lonely duplicate of me attends to nothing. The existential derivative of the attending relation is thus not an intrinsic property of me.' Any definition of 'intrinsic' thus needs to provide a means of classifying it as extrinsic. As far as I can see, the view presented in David Lewis' 'Redefining ' I n t r i n ~ i cfails ' ~ in this regard.
Here is Lewis' new definition:
P is intrinsic iff
(1) P is independent of accompaniment (2) (P and accompanied) is no more natural than P (3) (P & lonely) is no more natural than P
(4) (-P & accompanied) is no more natural than -P; (5) (-P & lonely) is no more natural than -P. It is beyond dispute that our existential derivative passes the first test. What remains are the relevant judgements of comparative naturalness. One is tempted to protest that the notion of naturalness is too vague to yield results in the cases we are about to consider. But let us put such temptation to one side, and play the game as best we can. (It will of course do Lewis no good to protest in this way, since his method of classifying properties as intrinsic or extrinsic requires that such judgments be in principle available.) It is important to notice that, unlike a gerrymandered disjunction, our existential derivative appears to codify a bona fide similarity. If A is a lonely cube and B an accompanied sphere, then, sure enough, A and B share the property of being either a lonely cube or an accompanied non-cube3: but this marks no very natural similarity between A and B. By contrast, the property
'
400
I assume we do not wish to relativize matters so that we count the existential derivative
an intrinsic property of Jones (who is attending to his leg) but not of me (who is attending
to the table).
This issue of PPR. This paper refines upon the view presented in Lewis and Rae
Langton's 'Defining 'Intrinsic", reprinted in Papers in Metaphysics and Epistemology,
Cambridge University Press, 1999, p. 116-32. The worry that follows tells equally against
the Lewiskangston definition.
Cf. Defining 'Intrinsic', p. 119.
JOHN HAWTHORNE
of attending to something does appear to mark a genuine similarity among its instances. It is not, at first blush, a gerrymandered disjunction in disguise. Of course, that property is equivalent to the disjunction 'attending to something and alone or attending to something and accompanied', but as Lewis is well aware, that fact is, by itself, neither here nor there. After all, predicates that express intrinsic properties are also equivalent to disjunctions of that form. In sum: our existential derivative is independent of accompaniment and does not appear to be disjunctive in any "gruesome" way: we should thus suspect that it will make trouble for Lewis' definition. One conception of naturalness that Lewis himself takes seriously confirms this suspicion immediately. Suppose the qualitative world is, at bottom, constituted by a set of elite universals. We may then think of relative naturalness in terms of ease of definability by a canonical language that includes various standard logical particles and a predicate for each elite universaL4 Now suppose, as seems plausible, that some of the elite universals are relations and that one or more of the elite relations is such that its existential derivative is independent of accompaniment. Let 'attending' name one such relation. The property attending to something will be easier to define in terms of the canonical language than either attending to something and alone or attending to something and accompanied: It will thus be reckoned more natural. (Similar remarks apply, mutatis mutandis, to the predicates relevant to parts (4) and (5) of the Lewis definition.) Even if we retreat to a more inchoate, less regimented, conception of degree of naturalness, relying on intuitive judgements concerning the degree to which some property marks a joint in nature, Lewis will fare no better. Begin with an easy case: The property of being an electron seems more natural than the conjunctive property of being an electron and being within fifteen feet of a bat. Why is this? After all, the conjunctive property secures even more similarity between its instances than is secured by the former p r ~ p e r t y . ~ Well, the line that separates the class of things that are both electrons and fifteen feet from a bat from everything else in the world just does not strike us as anything like a joint in nature. If God were to play a game of pointing to classes of possibilia that corresponded to ways that certain things were genuinely similar to each other and not likewise similar to anything else, we should hardly expect the class of things that were electrons and fifteen feet Thus: "The less elite are so because they are connected to the most elite by chains of definability. Long chains, by the time we reach the moderately elite classes of cats and pencils and puddles; but the chains required to reach the utterly ineligible would be far longer still". 'Putnam's Paradox,' reprinted in Papers in Metaphysics and Epistemology, p. 66. Similarly, for any intrinsic property F, the conjunction of F plus being accompanied guarantees more similarity (though not intrinsic similarity) than being F and lonely, so it will hardly do for Lewis to retreat to a conception of naturalness that makes a property more natural to the extent that it secures similarity between its instances. SYMPOSIUM
401
from a bat to be divinely ostended (assuming, as Lewis must, that not all classes are to be on equal footing for the purposes of this game). Relatedly, we would be extraordinarily surprised if the property of being an electron and fifteen feet from a bat showed up in a relatively fundamental law of nature. Our intuitive reactions do not shift very much when invited to compare 'is attending' with either 'is attending to something and accompanied' or 'is attending to something and lonely'. Suppose the mentalistic relation of attending to something were fundamental. Then we should not be surprised if God pointed to the class of attenders when playing the 'point to a joint in nature' game. And we would not be all that surprised if the property of attending to something showed up in a fairly basic law of nature. By contrast, we would be extremely surprised were God to point to the class of lonely attenAnd we would be very surprised were either the property of being a lonely attender or the property of being an accompanied attender to figure in a law of nature. Consider next the comparisons relevant to conditions (4) and (5) in Lewis' account, involving the negative existential derivative. Admittedly, our intuitions are less firm, since none of the properties in question appear particularly natural. Still, insofar as we do not throw up our hands at the invitation to judge comparative naturalness, one will incline to count the conjunctions as the loser when it comes to degree of naturalness. Is there any conception of naturalness that will work for Lewis here? One can perhaps rig something up. Let us stipulate that the degree of naturalness of a property be determined by how much intrinsic similarity it secures between things that instantiate it. On this unusual conception, the property of being an electron will be no more, no less natural than the property of being an electron that is fifteen feet from a bat, no more nor less natural than the property of being a lonely electron. Meanwhile, the property of being a sphere that is negatively charged will be reckoned more natural than the property of being negatively charged. And finally, the property of being a lonely attender will count as more natural than the property of being an attender, since being a lonely attender is a sufficient condition for the intrinsic property of attending to oneself or part of oneself. We should not, however, be either interested or surprised to learn that this conception of naturalness will do the trick: after all, it was defined in terms of degree of intrinsic similarity. In earlier work Lewis opted for a notion of intrinsic property that is defined in terms of perfect similarity: A property is intrinsic just in case it never divides perfect duplicates. A pair of things are perfect duplicates just in
'
Don't make the mistake of equating the class of lonely attenders with the class of things that attend to themselves or part of themselves. One can be accompanied and still attend to oneself.
402 JOHN HAWTHORNE
case they have exactly the same perfectly natural proper tie^.^ Assuming, as seems plausible, that existential derivatives are never perfectly natural (even if the relations from which they are derived are perfectly natural), then no worry is posed by them. Accept a framework in which one endorses a category of perfectly natural properties, and all may be well. Reject the framework, and one may well be left with a view that treats 'intrinsic' (or else 'duplication') as a primitive notion.8. 5~
'
See 'New Work for a Theory of Universals,' reprinted in Papers in Metaphysics and Epistemology, Cambridge University Press, 1999, p. 27. A pair of things share perfectly natural properties just in case there is a one-one mapping of the parts of one to the parts of the other that preserves the perfectly natural properties of the parts and the perfectly natural relations between the parts. If we allow ourselves only the apparatus of standard quantified modal logic, rejecting quantification over mere possibilia, we shall do well do opt for 'intrinsic' as our primitive, since we will lack the resources to define intrinsicness in terms of duplication. (Note in this connection that the sentence 'x is F and necessarily, anything that duplicates x is also F' only succeeds in bringing intra-world duplicates of x into focus.) I am most grateful to Karen Bennett, Cian Dorr, Tamar Szab6 Gendler, David Lewis, Ted Sider and Brian Weatherson for helpful discussion.
'
SYMPOSIUM
403
