Professional Documents
Culture Documents
ESSENTIALISM: PART 2
1. Ihe Mind-Body Problem
Perhaps the most controversial segment of Kripkes Naming and Necessip is its
closing treatment of the mind-body problem. O n essentialist grounds, he
argues for the claim that no token mental event is identical with a physical
event, that no mental event-type is identical with a physical event-type and
that the mind is not identical with the body. In each case, Kripkes strategy
is the same: to show that for a given mental item A (the mind or a mental
event-type or a mental event-token) and for a given physical item B,
purportedly identical with A, there is a property that A has essentially and
that B does not or there is a property that B has essentially but A does not.
Given this difference in essential properties, it follows that A is not identical
with B after all. One such property is, according to Kripke, the property of
being mental. A, the mental item has this property essentially or necessarily,
but, Kripke, claims on intuitive grounds, B does not.2
In contrast to the formally similar Goliath case, in the mental-physical
cases many philosophers have explored the matter of whether the intuitions
of contingency that apparently lead to the claims of non-identity can be
reconstrued in Kripkean fashion. Kripke himself takes up this matter and
concludes that a reconstrual is not available and thus he stands by the claims
of non-identity. However, without questioning Kripkes essentialist framework,
many others have argued that, in at least some of the mental-physical cases,
an appropriate reconstrual is available and thus, on this view, the relevant
arguments for non-identity do not go through. In particular, some have
defended a claim of token-token identity along these lines.3
1. Strictly, in the case of event-tokens and event-types, Kripke only draws conclusions for the
type pain and for tokens of that type. But there seems to be no reason why his conclusions
cannot be generdised.
2. Naming And Necessip, pp. 144-155; Identity and Necessity, pp. 98-101.
3. See Richard Boyd, Materialism Without Reductionism:What Physicalism Does Not Entail,
in Ned Block (ed.), Readings in Philosophy ofPTchology, vol. i (Harvard University Press, 1980)
pp. 67--106 (see esp. pp. 83-85); Olaf Gjelsvik, A Kripkean Objection to Kripkes Argument
against Identity Theories, Inquiy, vol. XXY (1987), pp. 435-450; and Colin McGinn,
Anomalous Monism and Kripkes Cartesian Intuitions, in Block (ed.), Readings in Philosophy
of Ppsychoolo~,vol. 1, pp. 156-157. For related approaches, see Steven R. Bayne, Kripkes
Cartesian Argument, Philosophia, vol. xviii (1988), pp. 265-269; Michael Della Rocca,
Kripkes Essentialist Argument Against the Identity Theory, Philosophical Studies, vol. lxii
(1993), pp. 101-1 12; Fred Feldman, Identity, Necessity, and Events, in Block (ed.), Readirgs
in Philosopfy .f PgcholoQ, vol. 1, pp. 148-155; and Thomas Nagcl, The Vm Fmm Nowhm
(Oxford University Press, 198G), pp. 46-49. Yablo defends some of the Kripkean arguments
on this matter in The Real Distinction between Mind and Body, Canadian journal of
Phihsoply, Supplementary Volume xvi (1990), pp. 149-20 1, and Is Conceivability a Guide
to Possibility?,Philoso& and Phenommological Research, vol. lii (1 993), pp. 1-42.
81
0 Blarkwell Publishers Ltd. 19Y6, 108 Cowley Road. Oxford OX4 IJF, UK and 258 Main Street,
82
0 Blackwell Publishen Lrd. 1996
83
0 Blackwell Publishers Ltd. 1996
originates from A (w3 is such a world). But (2) and (3) together entail that
there is no such possible situation. Thus, given (l), (2) and (3), we must
conclude that (4) is false, i.e. that T1 could not have come from B. The form
of the argument is quite simple. (2) and (3) entail that the consequent of (1)
is false. Thus for (l), (2) and (3) all to be true, the antecedent of (1) must be
false. That is, it must not be the case that it is possible for T1 to originate
from B.
Similar considerations would show that T1 could not have originated from
my non-A acorn or indeed from any thing whatsoever besides A. Indeed,
similar considerations could also be used to attempt to establish origin
essentialism for other organisms besides trees, for other natural objects besides
organisms, and for artifacts.
This form of argument is also available to generate kinds of essentialism
besides origin essentialism. The general argument structure is the following.*
One can argue that, on the assumption that x is the F, being the F is essential
to x, $-one is willing to maintain each of the claims:
This kind of argument has been put forth as establishing that each set contains
its members e~sentially.~
Further, the widely-discussed view that a natural
7. See Salmon, Ref.r.lce and Essence, p. 199. Each of these further cases, howevrr, would require
specific qualification of and refinements in the above kind of argument. In particular, one
would need to account for the fact that, in many c a m , we would allow that an object
might have had a somewhat different origin, but not a completely different origin. For
various views on this problem, see William R. Carter, Salmon on Artifact Origins and Lost
Possibilities, Philosophuul Review, vol. xcii (N83), pp. 223-23 I ; Roderick Chisholm, Identity
vol. i (1967), pp. 1-8; Robert Coburn,
Through Possible Worlds: Some Questions, NONOUS,
Individual Essences and Possible Worlds, in French, Uehling and Wettstein (eds.), Midwest
studzes in Philosopb, vol. xi, pp. 165-183; Forbes, 7he Metaphysics of Moduli&, Ch. 7, and
Worlds and States of Affairs: How Similar Can They Be?, in Kevin Mulligan (ed.), Language,
Tmth and onto lo^ (Kluwer, 1992), pp. I18--132; Lewis, On he Pluralirj of Worlds, pp. 243248; Thomas J. McKay, Against Constitutional Sufficiency Principles, in French, Uehling
and Wettstein (eds.),Mdwest Sludies in Philosophy, vol. xi, pp. 295-304; Quine, Worlds Away,
Journal $Philosophy, vol. lxxiii (1976), pp. 859-863; Salmon, Rtjkme and Essence, pp. 229252, Fregean Theory and the Four Worlds Paradox: A Reply to David Over, Philosophical
Books, vol. xxv (1984), pp. 7-11, Impossible Worlds, Anahsir, vol. xliv (1984), pp. 114-117,
Modal Paradox: Parts and Counterparts, Points and Counterpoints in French, Uehling
and Wettstein (eds.), Mdwest Studks in Philosophy, vol. xi, pp. 75-120, and The Logic of
What Might Have Been, Philosophuul &zim, vol. xcviii (1989), pp. 3-34.
8. See Van Cleve, Why a Set Contains Its Members Essentially, pp. 595-596; and Yabloh
review of Forbes, pp. 335-336.
9. Forbes, Ihe Mefuphyics of Moduli&, Ch. 5; van Cleve, Why a Set Contains Its Members
Essentially.
84
d Blackwell Publishers Ltd. 199G
kind (such as water, gold, or cats) possesses its physical structure essentially
can also be reached by this type of argument.
Lets return to the argument concerning T 1. This argument is undeniably
very significant because, as Yablo points out in a related context, it brings
out hidden connections between superficially independent de re modal
claimsnl-viz. between the claim of sufficiency and the claim of necessity.
Even so, we can still wonder whether this argument allays Quinean doubts
about a principled basis for treating some properties as essential to TI and
others as not. We can see right away that it does not. This is because the
argument for essentiality of origin assumes (2)) the claim that being an oak
tree that originates from A is sufficient for being T1. As we saw, this
assumption is tantamount to the assumption that certain properties such as
spatial location, time of existence, etc. are non-essential to T1. Thus a certain
property is classed as essential only by virtue of the fact that we have classed
certain other properties as non-essential. But for a Quinean who is concerned
about the legitimacy of classing properties as essential or as non-essential,
such an assumption of non-essentiality is precisely the kind of claim whose
legitimacy is doubtfii. Thus, by assuming (2), the above essentialist argument
does not really address the Quinean concerns.12Certainly (2) is a very strong
intuitive claim but for one like Quine, who would demand a stronger basis
than mere intuition to rest modal claims on, an assertion of such an intuition
is not enough to jus@ (2) and hence not enough to j u s q the above argument
for origin essentialism. By itself the argument does not go far enough.
Forbes, however, attempts to go further. In an important line of thought,
he attempts to offer a justification for claims such as (2) (and hence for origin
essentialism as well) that stems from the concept of identity. The conceptual
truths about the identity relation that Forbes has in mind are:
85
6 Blackwell Publishers Ltd. 1996
Consider the supposition that things could have been exactly as they are
except that the steel tower in Paris opposite the Palais du Chaillot is
different from the one actually there. To make sense of this supposition, it
is not permitted to imagine that the tower is made of different metal from
the metal which actually constitutes it, or that it has a different design, or
designer, or history. The on& respect in which the imagined situation is to
differ from the actual world is in the identity of the tower. The extent to
which such a difference seems unintelligible is some measure of the
plausibility of the view that transworld differences must be g r 0 ~ n d e d . l ~
It is not immediately clear what Forbes means by intrinsic in (6), but the
way in which he would invoke (6) to justfi (2) is clear. For (2) to be false,
there must be a possible situation in which an oak tree that results from A is
not T1, even though, in the actual situation, T1 is indeed the oak tree that
results from A. Call the counterfactual oak tree that results from A T2 (cf.
w3 described earlier). T 2 # T1, but what can ground this non-identity? We
need to find a difference between T 2 and T1 (as it is in the actual world)
that explains or grounds their non-identity. As we stipulated, they do not
differ in origin. Perhaps they differ in location and, forat least some of their
existence, in material constitution. Such differences are genuine, but, Forbes
would say, extrinsic and so incapable of grounding the diversity in this case.
Forbes would make the same claim for any other differences between the
trees which share the same origin. He would, therefore, conclude that T 2 =
T1 after all and he would thus just+ (2), the claim that Tls actual origin is
sufficient for being T 1.
As I said, it is not immediately clear what Forbes means by intrinsic. In
particular, while the spatial location of a tree is, arguably, extrinsic, it seems
more intuitive to regard material composition as intrinsic.16Yet Forbes must
regard material constitution as extrinsic. He goes some distance toward
clarlfylng and justlfyrng his distinction between intrinsic and extrinsic, but
in order to challenge (6) it is not necessary to explore this distinction further.
13. Forbes, In Defense of Absolute Essentialism, pp. 6-7. I have added numbers and changed
some of the lettering.
14. Forbes, ih Me&p&cs ofMo&l@, p. 130.
15. ih Mc&p/ysks ofModali& p. 128.
16. As Forbes himself notes (In Defense of Absolute Essentialism, p. 10).
17. In Defense of Absolute Essentialism, pp. 9-14.
86
0 Blackwell Publishers Ltd. 1996
87
8 Blackwell Publishers Ltd. 1996
(neMetup/ysia of
course of events, one in which B' (i.e. B) ceases to exist at tl and A' (Le. A)
continues to exist forever. The transworld diversity between A' and B is
grounded simply in the transtemporal diversity within u between A' at t2 (a
time after tl) and B at to (a time before tl).
This is a sophisticated line of thought, but to begin to challenge it, we can
ask what difference grounds the transtemporal non-identity between A' at
t2 and B at to? This non-identity, l i e all non-identities, would for Forbes
require grounding. Forbes's answer, as suggested in the passage quoted above,
is that A' at t2 is spatiotemporally continuous with A at to, whereas B at to
is not spatiotemporally continuous with A at to. Instead, B at to is, of course,
spatiotemporally continuous with B at to.
This account of the transtemporal non-identity of A' at t2 and B at
depends crucially on the claim that A at to # B at to, for i f A at to were
identical with B at then, it seems, by being spatiotemporally continuous
with A at to, A' at t2 would also be spatiotemporally continuous with B at to.
But Forbes explicitly denies that there is such continuity.
To summarise the argument so far: given the branching conception of
possible worlds, Forbes grounds the non-identity of B and A' in the
transtemporal non-identity of A' at t2 and B at to. This non-identity in turn
is, at least in part, grounded in the non-identity of A at to and B at to. (This
is diversity at a time, instead of transtemporal diversity.) But now the inevitable
question arises: what grounds the non-identity of A at to and B at to? To
answer this, one might appeal to the fact that A at to and B at to occupy
different spatial locations. But, as we have seen, Forbes holds klausibly) that
location is an extrinsic property, so this difference could not, for him, ground
this non-identity. Perhaps the non-identity is grounded in the fact that the
globes are made of two different quantities of matter. This might be seen as
an intrinsic difference, but since the two quantities themselves are not
identical, we would still need to find some intrinsic grounds of non-identity.
Let's try a different tack. Perhaps the diversity at a time between A and B
is grounded in some transtemporal difference between A and B. This,
however, will not work because, as we have just seen, the transtemporal
differences in this case seem to be grounded in diversity at a time. Thus, on
pain of circularity, we cannot appeal to transtemporal differences to ground
the diversity at a time.
Similarly, Forbes cannot appeal to transworld or modal differences between
A and B because, given his branching conception of possible worlds, such
differences in this case are grounded in transtemporal differences and,
ultimately, in the non-identity at a time of A and B. Thus again, for Forbes,
it would be circular to ground the non-identity of A at to and B at to in
transworld differences that are themselves grounded in that diversity at a time.
I can see no other way available to Forbes for grounding the diversity here
and so I conclude that if Forbes invokes the branching conception of possible
worlds to block Adams's argument for ungrounded transworld non-identities,
he must accept cases of ungrounded intraworld diversity at a time. Although
in the examples he provides to elucidate (6),Forbes focusses on transtemporal
and transworld non-identity, (6) clearly entails that intraworld diversity at a
88
0 Blackwell Publishen Ltd. 1996
time must also be grounded. Thus it seems that by avoiding one kind of
counterexample to (6),Forbes is forced to embrace another. At the very least,
then, we can say that Forbes has not offered a convincing case for (6).22
Perhaps a fd-back position is available. Perhaps Forbes can grant that
there are cases of ungrounded intraworld diversity at a time, but still maintain
that it is a conceptual truth that transworld diversity must be grounded. This
more limited conceptual truth might be all Forbes needs for his argument
for various forms of essentialism. Such a position may be consistent, but it
seems troubliigly ad hoc. It is not clear why one kind of diversity should
permit ungroundedness, while another does not allow for this. Certainly
Forbes does not provide the materials for a justification for treating intraworld
diversity at a time differently from transworld diversity. Without such a
justification, the fall-back position would be entirely unmotivated.
Since a defence of (6), or at least of the grounding of transworld nonidentity, is crucial to Forbess attempt to provide a principled basis for
distinguishing between the necessary and contingent features of a thing and
since Forbess defence has failed, it appears that Forbes does not provide this
basis. It seems, then, that the kind of argument philosophers use to attempt
to establish so many different essentialist claims has not been made immune
to Quinean doubts about essentialism.
In 1968, before the heyday of essentialism inaugurated by Kripkes Naming
and Necess$v, Richard Cartwright lamented the inability to resolve very
many debates over the essentiality or accidentality of a given property. He
concluded that
The existence of such cases, even in such large number, does not show
that there is simply no distinction between essential and accidental attributes
of an object. But it does show that the distinction is a good deal less clear
than essentialists are wont to suppose.23
At least some of the challenges raised in this paper show, I believe, that,
despite some very significant and groundbreaking work in the intervening
quarter-century, Cartmights assessment is still correct.24
YALE UNIVERSITY
22. For other criticisms of the branching conception of possible worlds, see Eugene Mills,
Forbes Branching Conception of Possible Worlds, Anabsb, vol. l i (1991), pp. 48-50 and
Yablo, review of Forbes. Salmon also rejects (6),though for reasons different from mine
(Modal Paradox: Parts and Counterparts, Points and Counterpoints, in French, Uehling
and Wettstein (eds.), Midwest Studies in Philosofly, vol. xi, pp. 75-120, appendix). For some
responses by Forbes, see Worlds and States of Affairs: How Similar Can They Be?.
23. Richard Cartwright, Some Remarks on Essentialism,Journal ofPhiLosofhy, vol. lxv (1968),
p. 626.
24. I am indebted to Teresa Robertson for very helpful discussions. My research was supported
by a Morse Fellowship from Yale University.
89
Q Blackwell Publishen Ltd. 1996