The Frustrating Problem For Four-Dimensionalism
(forthcoming in Philosophical Studies)
A.P. Taylor
1. Introduction
Metaphysical theories often have important consequences for our ethics.
Thus we have reason to carefully reflect on our ethical commitments when
deciding whether to endorse a given metaphysical view. We should strive for an
ontology that best complements the normative views we hold most deeply. When
we discover that a metaphysical theory has unforeseen and unappealing ethical
consequences, this should give us at least some reason to doubt that theory.1
Perhaps no metaphysical theory has enjoyed a more rapid growth in the
last century than the four-dimensionalist theory of the persistence of physical
objects.2 According to four-dimensionalism objects are “spread out” across
spacetime, and have temporal parts as well as spatial parts.3 They persist by
having different temporal parts in different regions of spacetime. This contrasts
with another popular theory of persistence, the three-dimensionalist account,
according to which objects have only spatial parts and are “wholly present” at
1
This will no doubt strike many readers as an inversion of the usual order of philosophical
operations, according to which we should do the metaphysics first and worry about the ethical
consequences later. However, I think there is something to be said for preferring metaphysical
views that complement our existing ethical beliefs and commitments with a minimum of revision.
Perhaps some revision will be necessary in the end, but revision should be a last resort.
2
The ever-growing list of philosophers who endorse four-dimensionalism, includes some of the
most important figures in twentieth century philosophy: Armstrong (1980), Carnap (1967), Lewis
(1983a; 1983b; 1986), Quine (1960), Russell (1927), and Whitehead (1920) all held versions of
the theory; as well as many excellent and able philosophers of twenty-first century, such as
Balashov (2010), Hawley (2001), Heller (1999), Hudson (2001), and Sider (2001).
3
According to Sider (2001, p. 59), x is an instantaneous temporal part of y at instant t =df (i) x
exists at, but only at t, (ii) x is part of y at t, and (iii) x overlaps at t everything that is part of y at t.
Except where noted, I shall use ‘instantaneous temporal part’ and ‘stage’ synonymously. Thus a
person-stage is just an instantaneous temporal part of a person. The term ‘subperson’ has a wider
extension (see Section 3 below for details).
1
each moment of their existence.4 Given the widespread interest in the view, it is
surprising that very little has been said about its ethical consequences.5 The
defenders and opponents of four-dimensionalism have largely focused on the
metaphysical arguments for and against the view and neglected to think deeply
about its ethical implications. The present work is intended to help fill this lacuna.
I do so by focusing on the concept of well-being (‘welfare’ or ultimate
good). Since well-being is an important concept in ethics, the ethically serious
four-dimensionalist should have good reason to be interested in developing an
account of well-being that best coheres with her chosen ontology.
Perhaps the most widely held theory of well-being amongst philosophers is
the desire satisfaction account according to which an individual’s life goes better
for them if more of their desires are satisfied and fewer are frustrated.6 In this
paper, I develop a new ethical argument against four-dimensionalism, namely,
that four-dimensionalism and the desire satisfaction account of well-being are
incompatible. This is because the four-dimensionalist ontology includes many
more beings with desires capable of frustration than we ordinarily suppose. As a
result, for every person whose desires are satisfied, there will be many shorterlived individuals (‘person-stages’ or ‘subpersons’) who share the person’s desires
but who do not exist long enough to see those desires satisfied; not only this, but
in many cases their desires are frustrated so that the desires of the beings in whom
they are embedded as proper temporal parts may be fulfilled. I call this the
frustrating problem for four-dimensionalism. In the first half of the paper I will
4
Three-dimensionalism has had many prominent defenders including: Baker (2000), Chisholm
(1976), Geach (1972a), Johnston (1992), Lowe (1989), van Inwagen (1990), and Zimmerman
(1999).
5
To my knowledge, the best discussions of the ethical consequences of four-dimensionalism are to
be found in Hudson (1999, 2001, pt. II), and Olson (2010).
6
Advocates of desire-satisfactionism, of one form or another, include: Brandt (1979), Rawls
(1971), Overvold (1980), Haslett (1990), Noggle (1999), Heathwood (2006).
2
lay the groundwork for understanding the frustrating problem, culminating in a
elucidation of the problem itself, and then in the second half, I will examine six
possible responses to the frustrating problem on behalf of the four-dimensionalist,
(i) the Parfit (1984) inspired claim that identity is not what matters, (ii) the
personal pronoun revisionism of Noonan (2010), (iii) the indirect concern account
of Hudson (2001), (iv) the sensible stages account of Lewis (1986), (v) a
multiple-concepts account of desire satisfaction, and (vi) a No Desire View
according to which subpersons have no mental states and thus no desires to
frustrate. I argue that none of these solutions will help the four-dimensionalist; she
does better to reject the desire satisfaction theory, while the defender of the desire
satisfaction theory does better to reject four-dimensionalism.
2. Four Dimensionalism
As I shall use the term ‘four-dimensionalism’ refers to an ontology of
persisting physical objects (including human persons) that combines perdurantism
with a commitment to mereological universalism. Perdurantism identifies
physical objects with ‘fusions’ or ‘sums’ of temporal parts. These fusions are
colloquially referred to as ‘worms’ on account of their elongated four-dimensional
shape. Rather than “sweeping through time” as on the three-dimensionalist’s
picture, objects are “spread out across time.” The basic idea of perdurantism is
often conveyed via an analogy with spatial parts.7 Just as a road might extend
from Boston to New York City, so you extend from your birth to the present. And
just as we can subdivide the road into its Boston part and its New York part, we
can subdivide you into various instantaneous ‘stages’ or ‘time slices.’ Each of
these stages will have a certain height, weight, eye color, and set of beliefs, which
7
For a paradigmatic example of such a use of the spatial analogy, see Sider (2001, chapter 1).
3
will determine your height, weight, eye color, and set of beliefs at the time at
which that slice exists. It is important to point out that for the perdurantist it is the
properties of the stages that determine the properties of the worm. If you, the
person-worm, are presently sitting in front of a computer, then this is because you
have a temporal part located at this particular region of spacetime that is sitting in
front of a computer. Mereological universalism is a theory of the composition of
objects according to which for any group of two or more objects, however
temporally or spatially gappy, the xs, there is another object y such that the xs
together compose y. For example various pieces of metal and wood combine to
form your desk, and various bits of hydrogen make up the Sun. One benefit of
mereological universalism is that it gives us a world complete with all of the
objects of ordinary everyday acquaintance. But it also tells us that there many
more objects than the ones we’re used to thinking about. As Quine puts it, the
theory tells us that physical objects are “simply the content, however
heterogenous of some portion of spacetime, however disconnected or
gerrymandered.”8 Thus, not only is your desk an object composed of many
smaller objects, but likewise the coins in your pocket, my left shoe, and the Eiffel
Tower constitute an object, albeit one for which we have no name (and in which
we have no interest). If mereological universalism is true, then there are many
many more objects in the world than we previously believed. And the objects of
everyday acquaintance are but a small subset of the objects that populate our
environment.
3. Subpersons
According to the four-dimensionalist human persons are physical objects like
8
Quine (1960, p. 171).
4
any others, which is to say that human persons are composed of temporal parts.
We can be arbitrarily sliced, and our parts summed, like any other physical object.
The resulting slices are called ‘person-stages’ or ‘person-slices’ when they are
temporally minimal parts that exist only instantaneously, and they are called
‘person-segments’ when they are temporally extended parts that exist through
intervals of varying sizes. For every discrete instant of time at which you exist
there will be a person-stage, or minimal subperson as I will call them, composed
by your temporal part at that time and only that time.9 And for every discrete
interval of time lasting less than your entire lifespan there will be a personsegment, or extended subperson as I will call them, constituted by your temporal
part(s) at those times and only those times. (For ease of exposition, I will usually
refer to both minimal subpersons and extended subpersons simply as
‘subpersons,’ accept where the argument requires that I make a distinction
between them.) There is one subperson that is composed of your temporal parts
while you are reading this sentence, for example, and another subperson that is
composed of your temporal parts while you are reading the whole paper, and still
another subperson that is composed of your temporal parts from the moment of
your birth until midnight tomorrow. All of these subpersons share your present
temporal part, thus if you are seated then they are also seated. And all of them
likewise share your brain, and your thoughts. If you are thinking about your
favorite café in Paris, or where you left your keys, then they are also thinking
about Paris and your keys. What’s more none of them is identical with the others,
since each is composed of a different set of temporal parts. There are also
temporally gappy subpersons that are composed of, for instance, all and only your
temporal parts while you are lecturing on Kant’s categorical imperative; or while
9
Olson (2010) refers to these beings by the plural ‘subpeople.’
5
you happen to be in the State of Wisconsin; or while you are facing the North
Pole. There is even a subperson composed of a single temporal part at some
vanishingly small instant of your life.
What is life like for one of these subpersons? David Lewis (1983b)
characterizes their existence in a provocative passage:
It does many of same things that a person does: it talks and walks and
thinks, it has beliefs and desires, it has a size and shape and location. It
even has a temporal duration. But only a brief one, for it does not last
long…it begins to exist abruptly, and it abruptly ceases to exist soon after.
Hence a stage cannot do everything that a person can do, for it cannot do
those things that a person does over a longish interval.10
Here Lewis is talking about momentary temporally extended subpersons (note:
Lewis calls them ‘person-stages,’ but his stages seem to be a little more robust
than the point-like temporally minimal subpersons favored by the stage theorist
since he conceives of them as lasting at least long enough to think and it is not
clear that any temporally minimal subperson could think on its own).11 They walk,
talk, think, believe, and have desires. Presumably they can feel pleasure and pain.
They have, in essence, all that typically makes one at least a prima-facie candidate
for moral status. Lewis stresses that we must not think of his person-stages as the
building blocks out of which persons are made. This makes them disanaloguous
with the spatial parts of your desk. Consider the first piece of wood that was hewn
and shaped for your desk. By itself it was not a desk, it had to be added to a larger
fusion first. There is no possible world where an exact physical duplicate
10
11
Lewis (1983), p. 76
For more on the stage theory see Hawley (2001) and Sider (1997, 2001).
6
counterpart of that piece of wood is itself a desk. By way of contrast, your first
person-stage had everything required for personhood.
Some of the subpersons that constitute you are vanishingly short-lived, but
others are nearly as long lived as you. In fact at least one subperson will stretch
from the moment of your birth until the moment just before you die. This
subperson will do virtually everything you do. The only difference between you
and it will be a single instant of further existence. Surely that subperson will have
everything required for moral status. After all, that subperson is an exact physical
duplicate of some counterpart of yours at another possible world (the world where
you die a moment sooner than you do in the actual world) and at that world your
counterpart is a person.
3. Desire Satisfaction Theory
The four-dimensionalist ontology includes many subpersons for each person.
This raises the question of our ethical obligations with respect to these
subpersons. If subpersons have everything that is required for moral consideration
on a par with the moral consideration we commonly afford persons, then this
means that we have an obligation, in so far as it is possible, to extend to them our
characteristic ethical treatment of persons. This in turn gives us a reason to look
for a theory of well-being that will be able to equally treat both persons and
subpersons.
‘Welfare’ or ‘well-being’ refers to that which is ultimately good for some
individual. Theories of well-being are generally taken to answer the question: in
virtue of what do we determine that an individual’s life is going well for them or
poorly for them? They give us a set of necessary and sufficient conditions for
determining whether an individual’s life goes well for them, or poorly for them.
7
Theories of well-being are thus tied into the notion of an individual’s best
interests or ultimate good. Whatever advances someone’s best interests or
promotes their most ultimate good will be that which increases their well-being. It
seems plausible to suppose that what makes my life go well for me is tied in to
what makes my life seem to go well to me, namely the degree and frequency of
the pleasures it presents to me, and the lack of pains. Call the view on which wellbeing consists of a balance of pleasure over pain hedonism.12 Meanwhile, the
difficulty in isolating a single structural feature, value, or good that makes a life
good for a person, has lead some philosophers to argue that a good life is one a
life supplied with a range of the right sorts of goods. Such views are referred to as
objective list theories.13 Finally, many philosophers maintain that what makes
some individual x’s life go well for x is that on balance more of x’s desires are
satisfied, and fewer are frustrated. This view is known the desire satisfaction
theory. As Haybron (2008, p. 23) notes desire satisfaction theory is “currently, the
theory to beat . . . [the] dominant account among economists and philosophers
over the last century or so.”
Desire satisfaction theories of well-being come in many forms. But I will
confine our discussion to the most basic form of the view, which simply holds that
a person’s life goes well for them if it contains on balance more satisfied desires
and fewer frustrated desires, and that a person’s life goes worse for them if it
contains on balance more frustrated desires and fewer satisfied desires. If this
view is correct then what makes my life go well for me is that is that my desires
are related to states of affairs in the world such that those states of affairs fulfill
12 Philosophers who’ve endorsed hedonism include: Feldman (2004), Crisp (2006), Bradley
(2009), and Heathwood (2006).
13
Philosophers who can be characterized as defending an objective list theory include: Rawls
(1971), Ross (1930), Frankena (1973).
8
my desires. Thus, on this view, well-being consists in a relation which takes in the
first place subjective mental states of the individual (i.e. desires), and relates them
to states of affairs of in the world in the second place.14
On the four-dimensionalist picture this amounts to a relation between some
temporal parts of a person, namely the ones possessing the relevant desire, and
some states of affairs in the world. Recall that on the four-dimensionalist view, it
is the subperson that has the mental states and the person only has those mental
states if the subperson has them. So every desire that a person has will be the
desire of one or more of their subpersons.15
4. The Frustrating Problem
The curious thing about the four-dimensionalist’s picture of the world is that it
contains far more beings with desires than we ordinarily suppose. For every single
person we count there are many more subpersons. This in turn leads to there being
more individuals whose desires can be frustrated than we ordinarily suppose. And
in fact, given that we can arbitrarily slice and sum subpersons, it follows that for
every desire a person has there will be many subpersons who share that desire, but
who do not exist at the time when that desire is fulfilled. Thus they are not
constituents of the states of affairs that satisfy their desires. Their well-being is
severely adversely affected. The result is a world filled with individuals whose
desires are routinely frustrated. Not only this, but in many cases their desires are
14
One common refinement on this view has it that what really determines one’s well-being is that
one’s ideal or all things considered desires are fulfilled, and not merely one’s subjective, or
egotistical desires. I want to avoid this refinement and consider, instead, the simple view just
outlined. I do this for two reasons: (i) ideal views have been critiqued for their inherent elitism,
and not all desire satisfaction theorists accept them, and (ii) the Noonan style response that I will
consider in section 5.2. works best if we take the subperson’s interests to align with their
egotistical propositional attitudes.
15
In section 5.6 below I consider the possibility that this claim is wrong and that, while persons
have mental states, subpersons do not.
9
frustrated so that the desires of the beings in whom they are embedded as proper
temporal parts may be fulfilled. For example, consider the case of a person who
decides to undergo a course of chemotherapy (involving severe sickness, pain,
discomfort, and mental anguish) in order to combat their cancer. In making the
decision to do undergo treatment, the person causes many subpersons to suffer in
order that they (the person) might eventually enjoy a period of recovery and good
health. But the subpersons also desire recovery and good health, though they shall
not exist when that state of affairs obtains. Thus the desires of the subpersons are
frustrated and their best interest and ultimate good is slavishly sacrificed for that
of another individual. I call this the frustrating problem for the fourdimensionalist.
At this point the four-dimensional desire satisfaction theorist might reply in a
sort of platitudinous way that “life is not fair,” or that “the world can be a cruel
place.” Perhaps it’s true that things are going badly for more individuals than we
previously recognized, but this in itself is no reason to abandon either fourdimensionalism or the desire theory. Imagine that we found a new island of the
coast of South America where the population was suffering from the most severe
drought and disease, this would be unfortunate, but adding to the population of the
disenfranchised hardly requires that we make wholesale changes in our
metaphysical and ethical views.16
But this response misses the mark. It is not simply that four-dimensionalism
adds to the population of suffering individuals that makes for the frustrating
problem, the problem is that it adds to the population of suffering individuals
whose interests could plausibly be said to differ from those of the persons in
16
I am indebted to Mark Spencer for first bringing the platitudinous reply to my attention.
10
whom they are embedded as temporal parts, and that in doing what is ultimately
good for persons we must do what is ultimately bad for their subpersons.
Setting aside the platitudinous reply, in the rest of the paper I want to turn to
an examination of six further responses on the part of the four-dimensionalist. As
we shall see, each of these responses faces its own difficulties.
5. Six Responses to the Frustrating Problem
In this section I will examine six possible responses to the frustrating
problem: the Parfit-inspired claim that identity is not what matters, the prounoun
revisionism of Harold Noonan, Hudson’s indirect account, Lewis’s sensible stages
view, a multiple concepts account of desire satisfaction, and the No Desire View
according to which subpersons have no mental states. I will argue that none of
these responses are convincing and that, as a result, the four-dimensionalist has
reason to reject the desire satisfaction theory. A corollary of this conclusion is that
the adherents of the desire satisfaction view have at least some reason to reject
four-dimensionalism.
5. 1. Identity is Not What Matters
Faced with a world filled with frustrated subpersons, the four-dimensionalist
might be tempted to adopt the Parfit-inspired move of arguing that identity is not
what matters.17 If this is correct, then it won’t matter that the subperson who
desires a given state of affairs is not identical with the subperson (or person) who
exists when that state of affairs obtains. All that will matter is that someone exists
when that later state of affairs obtains, and that that someone is psychologically
17 A similar strategy is pioneered by Olson (2010). However, Olson’s target is not fourdimensionalism per se, but rather what he calls “the generous ontology” (as opposed to the sparse
ontology he favors). The generous ontology appears to be his name for any ontology that allows
for merelogical universalism, whether four-dimensionalist or three-dimensionalist.
11
continuous and connected to the subperson with the desire in an appropriate way.
A theory such as this preserves the importance of desires, and the states of affairs
that satisfty them, and recognizes the separateness of subpersons (and persons),
but at the same time it minimizes the instances of frustration in the world by
allowing later subpersons (and persons) to obtain the satisfactions of an earlier
subperson’s desires.
In Reasons and Persons (1986) Derek Parfit defends the thesis that “identity
is not what matters in survival.” This contrasts with our commonsense thinking
about practical deliberation. On the commonsense view, we have more reason to
be concerned about what happens to some future individual if it is the case that we
are identical with that individual, and less reason to be concerned if it is the case
that we are not identical with that individual. Parfit argues, contrary to this view,
that there are cases in which we will have reason to be concerned for a future
individual even though it will be indeterminate whether or not they are identical
with us. Why be specially concerned for them? Because for Parfit, it is not
identity, but rather psychological continuity and connectedness that is the relation
that matters (or as he calls it “relation R”).
The centerpiece of this argument is the division case. Here’s how Parfit
lays it out for us:
My body is fatally injured, as are the brains of my two brothers. My brain
is divided, and each half is successfully transplanted into the body of one
of my brothers. Each of the resulting people believes that he is me, seems
to remember my life, has my character, and is in every other way
12
psychologically continuous with me and he has a body that is very like
mine.18
What happens to me in the division case? According to Parfit there are five
possibilities: (i) I cease to exist and two new people come into existence as a
result of the operation; (ii) I survive as one of the twins A, but not the other twin
B; (iii) I survive as B but not as A; (iv) I survive as one person scattered across
both A and B; and finally (v) it is an “empty question” whether I survive or do not
survive—nothing much hinges on the answer. Parfit argues the best thing to say in
this case is (v), whether I survive and how is an empty question. Defenders of the
commonsense view that it really matters whether some future individual is
identical with me are likely to bristle at this suggestion, but that is because they
mistakenly take personal identity (or ‘identity’ for brevity’s sake) to be the
relation that matters. All that truly matters, Parfit tells us, is that I will be Rrelated to both of the products of my division. Identity, for Parfit, is simply
relation R plus a uniqueness, or no branching, clause and it is thus trivial and
extrinsic. Thus it cannot be a component of what matters.19
The four-dimensionalist could try to avoid the frustrating problem by
helping themselves to Parfit’s claim that identity is not what matters and arguing
that as long as some person or subperson y, at time t2, when state of affairs S
obtains, is R-related to the earlier subperson x, who desires, at t1, that S, then x’s
desire is satisfied, and this is all that is required for x’s well-being to be positively
affected. For example, if the subperson(s) currently located in my chair desires
that this paper is well-received by my peers, then so long as some later (person or)
subperson who is R-related to the subperson(s) in my chair is satisfied by the
18
19
Parfit (1986), pp. 254-255.
For Parfit, personal identity consists in relation R + uniqueness.
13
obtaining of the state of affairs consisting in this paper being well-received by my
peers, then the well being of the present subperson(s) will be positively affected.
In short, the four-dimensionalist could argue that, while well-being for persons
(and longer lived subpersons) consists in having the desires of earlier temporal
parts satisfied by states of affairs involving one’s later temporal parts, for shorter
lived subpersons—in virtue of their short lifespans—well-being consists in having
their desires satisfied by states of affairs involving later persons or subpersons
who although not strictly identical with them are nevertheless R-related to them.
Making the Parfit inspired response to the frustrating problem requires the
four-dimensionalist to argue that identity does not matter. But that argument is not
available to the four-dimensionalist given her beliefs about the necessary and
sufficient conditions for identity. David Lewis (1983a) agreed with Parfit that
what matters is psychological continuity and connectedness, but he also believed
that we could save the “commonsense platitude” that identity matters by
recognizing that whenever we had psychological continuity and connectedness,
we had all that was necessary for identity. He responded to the division case by
arguing that what it really shows us is that there are two persons there all along.
The two persons share all of their pre-operation stages, and diverge later on as a
result of the operation. Parfit’s argument for identity not mattering relies on his
claim that identity consists in relation R + uniqueness. Defining identity this way
allows Parfit to reject the concept of identity, while maintaining that something,
namely relation R, is what matters to us. But the four-dimensionalist takes relation
R to be identity. She cannot then deny that identity matters, because were she to
do so, she would have nothing left which did matter. And clearly, since we are
concerned for our future survival, something does matter.
14
Given that something does matter, and that if four-dimensionalism is true,
all that can matter is identity, I contend that the four-dimensionalist has at least a
prima-facie reason to resist making a Parfit inspired move in light of the
frustrating problem.
5. 2. Pronoun Revisionism
Another solution that might appeal to the four-dimensionalist is a variation on
Harold Noonan’s response to the thinking animal problem of Eric Olson (2002,
2007). Olson contends that when a human animal comes to constitute a human
person, then there will be too many thinkers located wherever the animal is
located. After all if the person uses its brain to think, and the animal uses the same
brain, then both should be thinking. But this leads to the absurdity, among others,
that one cannot know when one says “I am hungry” whether one is referring to the
animal or to the person. Noonan (2003, pp. 205-208) responds that since ‘I’ is a
personal pronoun then it must always refer to a person, and not the animal. Olson
(2007) calls this the pronoun revisionism view. The four-dimensionalist might
help themselves to a similar thesis to avoid frustrating problem, that is, they might
argue that whenever the subperson expresses a desire using the personal pronoun,
the desire it expresses is a desire for a state of affairs involving the person rather
than itself. Thus all of a subperson’s desires would in fact be desires on behalf of
the person, and if this is the case, then we needn’t worry about the subperson not
being around when the desire is satisfied, so long as the desires of the person are
satisfied.
Noonan insists that it is analytically true that the referent of the personal
15
pronoun must always be a person. He thinks it is not enough to have thoughts
about thoughts, but rather an individual must have the right sort of psychological
persistence conditions in order to be the referent of ‘I’. Thus the animal cannot be
the referent of ‘I’, since its persistence conditions will not be psychological but
rather biological. The interesting claim for our present purposes is that even
though the animal can utter the pronoun ‘I’ it cannot use that pronoun to refer to
itself since it always refers to the person with psychological persistence
conditions. According to Noonan the animal can think about itself, just not in the
first person mode. Whenever the animal thinks “I am a person”, it is not thinking
something false, since ‘I’ refers not to itself but to the person it constitutes.
Likewise the four-dimensionalist might argue that the subperson will not
have the proper persistence conditions to be the referent of the personal pronoun.
For example, Hudson (2001, p. 121-127), argues that the best candidate for the
title of person, is a being that (i) is not a proper part of any other thinking being,
and (ii) is a what he calls a “maximal C-possessor.” To be a C-possessor,
according to Hudson, is to possess what he calls “characteristics C,” a collection
of cognitive capacities including self-consciousness and first-person intentional
states, that are usually agreed to be relevant to personhood. To be a “maximal” Cpossessor is for it to be the case (i) that all of one’s temporal parts possess
characteristics C, and (ii) that one not be a proper, temporal, part of any Cpossessor. Many subpersons according to a Hudson-like view will be Cpossessors20 (i.e., all of their temporal parts will possess characteristics C) but
they will fail the anti-parthood criterion, thus, will have not have the proper
persistence conditions for personhood. As a result whenever any such subperson
20 It bears reiterating that not all subpersons are instantaneous temporal parts. Many subpersons
are longer-lived fusions of such parts, being sufficiently temporally “thick” to possess thoughts,
beliefs, desires, and so forth.
16
thinks “I desire that my cancer goes into remission” it refers not to itself or its
own desires, but the person of whom it is a part, and oto the person’s desires.
But I think this Noonan inspired response fails. I shall offer three arguments
against it. First, we could always introduce a new subpersonal pronoun, Schmy,
that refers to the subperson rather than the person. In which case desires expressed
using Schmy would be desires of the subperson and not simply desires on behalf
of the person. This would allow the subperson to formulate and express its own
desires, in which case four-dimensionalist would again have to give some account
of the well-being of the subperson as distinct from the well-being of the person.
Second, not all of our desires are expressed using pronouns. For example very
young children have desires (e.g. to be fed or to sleep) even though they do not
express them self-consciously using the personal pronoun. And brainwashed
individuals likewise might have desires that they have been programmed either to
forget or ignore. So it appears that being able to use the personal pronoun is not
necessary for having desires. If this is the case, then it doesn’t matter that
subpersons cannot use the personal pronoun to express desires of their own. They
may be like the brainwashed. A third argument is that we could get at the desires
of the subperson indirectly via the substituted judgment of the person. The person
can simply ask themselves what they’d want in the subperson’s place, I can
imagine that I will go out of existence at midnight and then ask, would I rather
work on the paper, or go out for a few martinis, and this would give the
subperson, who will in fact go out of existence at midnight, a way of thinking
about its own desires without use of a personal pronoun. If these arguments are
correct then pronoun revisionism will not help the four-dimensionalist. What
desires an individual has does not, it seems, depend on their ability to use the
personal pronoun to refer to themselves as Noonan contends.
17
5. 3. Hudson’s Indirect Account
Yet another solution to the frustrating problem is the indirect account
offered by Hudson (2001). He argues that even though the four-dimensionalist
ontology gives us a world populated with many more thinking and desiring
beings, or C-possessors, than we previously imagined, this need not threaten our
commonsense approach to ethical deliberation. This is because, as he says
In absolutely every case, what we do to the part we do to the person.
Accordingly, whereas our moral behavior toward a person is restricted
directly insofar as the person is an object of direct moral concern, our
moral behavior towards person-parts is restricted indirectly insofar as any
consequence of an action on a person-part is always a consequence of an
action for a person.21
In other words, so long as we have accounted for the well-being of persons, we
needn’t go the further step of accounting for the well-being of subpersons, since,
in directly procuring what is good for persons, we indirectly procure what is good
for their parts. In terms of the desire satisfaction theory, what matters for a
Hudsonian view is that we are attentive to the desires of persons. We needn’t
worry ourselves about the desires of subpersons, since whatever desires the
subpersons have, they will also be the desires of persons, and so long as we are
fulfilling the desires of persons then we will be indirectly fulfilling the desires of
the subpersons.
But this seems false. Consider again the case of the cancer patient who
decides to undergo chemotherapy. The person (i.e., the “maximal C-possessor)
will both undergo the chemotherapy and obtain the benefits of recovery and
21
Hudson (2001), p. 165.
18
health. Thus, its desire for health will be satisfied. But the subperson, who exists
only during the most painful period of the chemotherapy, will likewise desire
recovery and health, but it will not exist long enough to enjoy the satisfaction of
its desire. Now, suppose that the cancer diagnosis is made very early on, when the
tumor presents no discernible symptoms. The patient could enjoy weeks of
continued good health and normal activity. But the chemotherapy will involve
gruesome suffering and indignity over the same period of time. By deciding to
undergo the chemotherapy now, the person, and present subperson, decides to
inflict suffering on a later subperson, so that the person, and even later
subpersons, can enjoy the improved health and well-being the person desires. And
to make matters worse, the suffering subperson will inherit its beliefs and desires
from the earlier (sub)person(s), thus it will slavish sacrifice its own well-being for
that of another being. The upshot is that we are certainly doing something to the
part, i.e., the subperson, that we are not doing to the person; we are not making
the person sacrifice their well-being for that of someone else. If this is correct,
then the Hudsonian response is not a good one.
5. 4. Lewis’s Sensible Stages
Some remarks by David Lewis (1986, p. 126) suggests that a solution to
the frustrating problem that appeals to rationality of subpersons and the notion
that what one subperson desires can be fulfilled by later subperson so long as the
pair are appropriately causally related. In a brief footnote where he responds to a
criticism made by Mark Johnston, Lewis writes
I agree that my view makes it fair to think of desires as belonging in the
first instance to my present stage, and derivatively to the persisting sum of
many stages. And I agree that that what my present stage wants is not to
finish the book itself – it’s a sensible stage, so it knows that can’t be
19
expected. But that’s not to say that it only cares what happens, never mind
how. There is a middle ground. My present stage wants the book to be
finished in the fulfillment of its present intentions—there’s the egocentric
part—and that will only happen if the proper sort of causal continuity
binds together my present stage with the one that finishes the book. The
continuity thus desired is part of the continuity that unifies mereological
sums of person-stages into persisting people.
At first blush, Lewis’s account seems similar to the Parfit-inspired account with
identity not mattering. After all, Lewis thinks that what matters is not that the
stage that does the desiring is the same as the stage which is a constituent in the
state of affairs that fulfills the desire, but rather merely that some appropriately
related stage is a constituent in that state of affairs. But it is important to keep in
mind that for Lewis any appropriately related future stage will be identical and
thus identity will matter for Lewis.
One might think this coheres well with the desire satisfaction theory, for on a
such a theory it might seem that our well-being can be affected after we have
ceased to exist—well-being doesn’t stop at the grave, it outlives us, though we are
its subjects. In terms of the present discussion the four-dimensionalist might think
that even though a given subperson will not exist at some future time, its wellbeing at the time at which it does exist might be affected by what happens to an Irelated counterpart at that time.
However, such a solution faces difficulties. For one thing, it is difficult to
imagine how one’s well-being can be affected by states of affairs that occur after
one has ceased to exist. This has lead several desire satisfaction theorists,
including Overvold (1980), Fuchs (1991), and Dorsey (forthcoming), to argue for
a restriction of desire satisfactions to only those states of affairs involving the
20
desiring subject as a constituent. Thus, in terms of the present argument, no states
of affairs involving a later subperson as a constituent will suffice to affect the
well-being of an earlier subperson. (this also seems to be a consideration for those
who hold that satisfactions should be restricted to states of affairs concurrent with
desires on account of the problem changing desires (e.g. some hold that if at time
t1 I desire to climb a mountain but later no longer desire to do so, it will not add to
my well-being if I am later made to climb the mountain)).
Even if we reject this restriction of desire satisfaction another problem
remains. Perhaps Lewis is correct in supposing that his sensible subpersons will
not mind a later subperson getting the satisfaction of their desire. But it is hard to
imagine a subperson accepting with equal sensibility the fact that it must suffer
great pain or mental anguish in order that a later subperson might get some
satisfaction. It seems the opposite of sense to suppose that a subperson would
accept such a state of affairs willingly were they not slavishly subservient to the
larger beings in whom they are embedded.22
5. 5. The Multiple Concepts of Desire Satisfaction View
A fifth response to the frustrating problem is what I call the multiple concepts
of desire satisfaction view.23 In the previous pages I have been assuming that
whatever the correct account of desire satisfaction is, it must be a unified “single
concept” account. In other words I have assumed that whatever makes for desire
22
Perhaps some will think that the stages would endure the suffering out of their concern for the
person. Just as one might willingly sacrifice one’s own life to save or improve the life of one’s
child or spouse. But in cases of such extreme familial sacrifices, the individual is usually aware of
how their sacrifice will benefit their loved one and they are at liberty to decide for themselves
whether to make the sacrifice or not. But the subperson who sacrifices for the person is not at
liberty to decide to do otherwise, nor are they aware of what is they are doing (they think they are
the person).
23 To my knowledge, this view has no precursor in the desire satisfaction literature, but it is
inspired by recent work on the conceptual analysis of causation. Philip Dowe (2000), among
others, has argued that we need multiple concepts of causation to account for both positive or
“biff” causation, and negative causation (or causation by void or absence).
21
satisfaction with respect to persons must also do so with respect to subpersons.
But what if there is no unified account? What if desire satisfaction consists in one
thing for persons and in quite another for subpersons? For instance desire
satisfaction for persons could consist in having the desires of one’s earlier
subpersons fulfilled by states of affairs involving one’s later subpersons, while
desire satisfaction for subpersons could consist in being properly causally related
to, or psychologically continuous with, some future person or subperson (who
enjoys the satisfaction of the desire). Thus making desire satisfaction an (at least
in part) internal relation for persons and an external relation for subpersons. If
this was the case, then the desires of subpersons could be fulfilled by other
subpersons, though not in the same way that a person’s desires are fulfilled. This
would again allow the four-dimensional desire satisfactionist to avoid the
frustrating problem.
However I think this sort of multiple concepts view of desire satisfaction is
troublesome as well. First, as with any multiple concepts view (in other areas),
we’d face the difficulty of stating precisely what it is about the two views that
make them both instances of desire satisfaction. Outwardly they look very
different. One is an intrinsic relation between the temporal parts of a given
aggregate (i.e. a person), and the other is an extrinsic relation involving nonidentical individuals (two or more supersons). In virtue of what feature would we
judge that either sort of relation makes for desire satisfaction? And even if we
could settle on precisely what that feature is, why not make that feature the
grounds for desire satisfaction rather than the relations just described?
Second, keep in mind that many of your subpersons were they to exist as
counterparts in another possible would have everything that is required for
personhood. For instance consider the subperson, in the actual world, that
22
stretches from the moment of your conception until the moment just before you
die. Call this subperson Almost. Surely Almost will have a counterpart in another
possible world where you did not live for that additional moment. Almost will be
an intrinsic physical duplicate of his/her counterpart, but Almost’s well-being will
depend one sort of one sort of desire satisfaction, while his/her counterparts wellbeing will depend on something else in another. It seems strange that two things
that are intrinsic physical duplicates could vary with respect to what makes for
their well-being.
5. 6. The No Desire View
Throughout the foregoing discussion we have been laboring under the
assumption that, as Lewis claims, subpersons are thinking, believing, and desiring
beings, and that persons have the mental states they do because their subpersons
have those mental states intrinsically. But one way for the four-dimensionalist to
avoid the frustrating problem is to deny that subpersons have the mental states
Lewis and others attribute to them. It is at least possible that some fourdimensionalists might deny that subpersons have any intrinsic mental states
whatsoever. Perhaps persons have mental states, but the person’s mental states,
rather than being derived from the intrinsic mental states of her subpersons, arise
from facts about how her subpersons are constituted and arranged. If subpersons
have no mental states, then it follows that they have no desires to frustrate, in
which case the frustrating problem dissipates. Call the view on which subpersons
have no desires the No Desire View.24
While it is true that most four-dimensionalists are unlikely to accept the
No Desire View, there may nevertheless be some plausible argument(s) that can
24 I would like to thank anonymous reviewer for suggesting this avenue of response to the
frustrating problem.
23
be given in its favor. In this final section of the paper I will present and critically
evaluate what I take to be the two best such arguments.
Sidney Shoemaker (1999, p. 304) argues that a being must have the right
kind of psychological persistence conditions in order to be the subject of mental
states. He is concerned with resisting the overcrowding problem that arises if we
believe that both the human animal and the Neo-Lockean psychological person,
which it constitutes, are subjects of mental states (in virtue of the fact that both
use the same brain to think). In order to avoid this problem he argues (i) that
mental states play characteristically functional roles in the subject’s cognitive
dynamics, and (ii) that only beings with the right persistence conditions can have
mental states that play such functional roles. The friend of the No Desire View
could make a similar maneuver arguing that, like human animals, subpersons lack
the appropriate persistence conditions to allow them to play the functional roles
that are characteristic of mental states. If only those beings with the psychological
persistence conditions of persons have mental states, and subpersons lack those
persistence conditions, then it follows that they will not be subjects of mental
states, and thus they will have no desires to frustrate.
But, here we need to recall the distinction we made above between
minimal subpersons and extended subpersons (See section 3). Minimal
subpersons have a point-like duration, while extended subpersons range in
duration from nearly point-like to nearly as long-lived as the person. For the No
Desire View to hold, it must be the case that only persons have desires. And while
many critics and advocates of four-dimensionalism might agree that minimal
subpersons are too short-lived to play the characteristic functional roles of mental
states, it is much less obvious that extended subpersons cannot do so. At best, the
24
functionalist argument will only eliminate minimal subpersons and extended
subpersons at the minimal end of the extended spectrum as subjects of desire.
Though perhaps we needn’t grant the No Desire View even this much. It
might be the case, as Strawson (1994) claims, that part of what is essential to
intentional mental states such as belief and desire is a characteristic ‘feeling’ or
qualitative state. If so then, any being that possesses this qualitative state might be
said to have, or at least to share in, the desire that it is about. Even a minimal
subperson with such a qualitative might be said to have a desire. In which case the
frustrating problem returns. While Strawson’s view is admittedly controversial,
there is another argument lurking. Katherine Hawley (2001, p.65) claims that we
have no reason to think that having mental states must be intrinsic properties of
minimal subpersons (or as she calls them ‘person-slices’). She argues:
My present stage has interests and acts, partly because it is
appropriately related to other stages which have appropriate
properties. In the same way, this piece of wood is a table-leg partly
because it is appropriately related to other pieces of wood. It is no
less a table-leg for all that.
On Hawley’s view the four-dimensionalist is at liberty to maintain that a
minimal subperson does have mental states so long is it appropriately related (e.g.
via immanent casual ties, or the counterpart relation) to some other appropriately
propertied subpersons. If Hawley is right, then even the most minimal subperson
can be said to possess mental states, and thus desires. In which case, again, the
frustrating problem returns.
Perhaps, instead of a commitment to functionalism, the No Desire View
could be generated from an intuitively appealing exclusion principle which says
that no being with mental states could be embedded as a proper, temporal (or
25
spatial) part of another being with mental states. Thus, while the person has
mental states, none of her proper parts could have them. If this were the case, then
only whole persons would be subjects of desire.
But while it may appeal to a three-dimensionalist such as Olson, or van
Inwagen, it is not at all clear why a four-dimensionalist should accept such a
principle. In fact, Hudson (2001, p. 144) explicitly rejects such a principle.
According to Hudson’s formulation,
“human persons are those (spatially and temporally gappy) spacetime
worms (i) that are not proper parts of other human persons, (ii) that are
maximal C-possessors, (iii) whose person-stages are united by a certain
relation of psychological continuity and connectedness, (iv) whose later
person stages bear an appropriate causal-dependence relation to earlier
person stages, and (v) which are most often found somewhen within the
lifespan and somewhere beneath the skin of a living human organism”
Most important for our present purposes is Hudson’s (iii). According to Hudson a
human person’s person-stages are united by their psychological continuity and
connectedness. If this is so, then it follows that person-stages, i.e., subpersons
must possess mental states.
Furthermore In his story involving Hopeful, Vital, Feeler, Thinker, and
Cheerful, (pp. 116-117) Hudson makes it explicit that some thinking beings,
namely Thinker (the Neo-Lockean psychological person) and Cheerful (the
person with a fully-developed psychology), are proper parts of larger thinking
beings, namely, Feeler (the merely sentient being) and Vital (the living human
organism). Both Feeler and Vital think in virtue of having proper parts that think.
Even if one were to argue that contrary to the Neo-Lockeans, the person is
identical to Feeler rather Thinker, the sentient-person would still be embedded in
26
Vital, the larger living organism, which is also a thinking being. So, if fourdimensionalism, can allow for the person to be embedded in another thinking
being (the merely sentient being), why think that it cannot accommodate thinking
beings embedded as proper (temporal) parts of the person?
I want to conclude this section by pointing our that if, on the grounds of
the frustrating argument (or a similar ethical argument), some advocates of fourdimensionalism are moved to adopt a position like the No Desire View, then,
given that this view is unlikely to convince many, for the reasons we have
discussed, I think that this result should count as a modest success for the critics
of four-dimensionalism. The most plausible version of four-dimensionalism
(whether it be the worm or the stage theory) is one on which subpersons have
their mental states, and thus their desires, intrinsically.
6. Conclusion
In the foregoing discussion, I have argued that four-dimensionalism, when
combined with the desire satisfaction theory of well-being, leads to the frustrating
problem. I then discussed six responses to the frustrating problem on behalf of the
four-dimensionalist. First, she could endorse the Parfit inspired strategy—
claiming that identity is not what matters. But this undercuts one the supposed
advantages of four-dimensionalism, namely its conservative approach to identity
mattering. Second, the four-dimensionalist could endorse the personal pronoun
revisionism of Noonan. But the subperson could have interests, well-being, at
stake despite its inability to refer to itself using the personal pronoun. Third, the
four-dimensionalist could opt for a Hudson-style indirect concern account. But
this will fail, since it turns out that if four-dimensionalism is true, then it is false
that whatever we do to the part we do to the person. Fourth, the four27
dimensionalist could argue, as Lewis does, that it is sensible for their subpersons
to do what is in the interests of the person. The problem here is that it does not
seem sensible to subject oneself to a lifetime of suffering purely for the benefit of
another. Fifth, The four-dimensionalist could adopt a multiple concepts account of
well-being, however this will not allow them to explain why one concept of wellbeing will apply to a supberson and another to the subperson’s intrinsic physical
duplicate counterpart. Finally, the four-dimensionalist could argue endorse the No
Desire View, which says, contra Lewis, that subpersons have no intrinsic mental
states and therefore no intrinsic desires to be frustrated. I argue that while it is
plausible that temporally minimal subpersons have no mental states, it is less
plausible that longer-lived subpersons lack them. Thus there will still be many
extended subpersons whose desires will be frustrated. And if Hawley or Strawson
are in the right, then there really is no problem with attributing mental states to
minimal subpersons. Furthermore even if my argument for this last claim fails to
convince my readers, I hope to have at least shown that the most plausible version
of four-dimensionalism faces considerable difficulty.
So what response remains for the friend of four-dimensionalism? I would
counsel her to reject the desire satisfaction account of well-being, and opt instead
for one of its theoretical rivals, i.e. hedonism or the objective list theory, unless
similar problems could be generated for both of those views. The corollary of this
claim is that inasmuch as one is committed to the desire-satisfaction theory, they a
reason to reject four-dimensionalism, especially of the sort endorsed by Lewis.25
25
I would like to thank an anonymous reviewer at Philosophical Studies for many helpful
comments, especially those that lead to the development of the argument in section 5.6. I would
also like to thank David Hershenov, Neil Williams, and Mark Spencer, for their tireless
contributions at every stage of the development of this paper. I would to thank Berit Brogaard and
Shane Babcock for many hours of helpful conversation on these matters. Finally, I would like to
thank Maureen Donnelly, Randy Dipert, Ken Shockley, Matt Lavine, Patrick Ray, and the
28
References
Armstrong, D. 1980. “Identity Through Time.” in, Van Inwagen, P., ed., Time
and Cause: Essays Presented to Richard Taylor. Dordrecht: Reidel.
Baker, Lynne Rudder. 2000. Persons and Bodies. Cambridge University Press.
Balashov, Yuri. 2010. Persistence and Spacetime. Oxford University Press.
_____. 2003. “Restricted Diachronic Composition, Immanent Causality, And
Objecthood: A Reply To Hudson”. Philosophical Papers 32: 23-30.
Brandt, Richard. 1979. A Theory of The Right and The Good. Clarendon Press.
Broad, C.D. 1923. Scientific Thought. New York: Harcourt Brace.
Carnap, Rudolf. 1967. The Logical Structure of The World. trans. Rolf George.
University of California Press.
Chisholm, Roderick. 1976. Person and Object: A Metaphysical Study. La Salle:
Open
Court Publishing.
Crisp, Roger. 2006. Reasons and the Good. Clarendon Press.
Dorsey, Dale. (forthcoming). “Desire-Satisfactionism and Welfare as Temporal.”
Ethical Theory and Moral Practice.
Dowe, Philip. (2002) Physical Causation. Cambridge: Cambridge University
Press.
Feldman, Fred. 2004. Pleasure and The Good Life: Concerning the Nature,
Varieties,
and Plausibility of Hedonism.
Frankena, William. 1973. Ethics. Prentice Hall Publishing.
Fuchs, Alan. 1991. “Posthumous Satisfactions and the Individual Welfare.” The
audience at the University at Buffalo works-in-progress colloquium for their many helpful remarks
on an earlier draft of this paper.
29
Journal of Philosophical Research 16: 345-351.
Geach, Peter. 1972. “Some Problems about Time.” in Logic Matters. Oxford:
Basil Blackwell.
Haslett, D.W. 1990. “What is Utility.” Economics and Philosophy 6(1): 65-94.
Hawley, Katherine. 2001. How Things Persist. Oxford Clarendon Press.
Heathwood, Chris. 2006. “Desire Satisfactionism and Hedonism.” Philosophical
Studies 128(3) 539-563.
Heller, Mark. 1999. The Ontology of Physical Objects: Four-Dimensional Hunks
of
Matter. Cambridge University Press.
Hudson, Hud. 1999. “Temporal Parts and Moral Personhood.” Philosophical
Studies
93: 299-316.
_____. 2001. A Materialist Metaphysics of The Human Person. Cornell
University Press.
Johnston, Mark. 1992. “Constitution is Not Identity.” Mind, 101: 89-95
Jubien, Michael. 1993. Ontology, Modality, and the Fallacy of Reference.
Cambridge
University Press.
Kagan, Shelley. 2009. “Well-being as enjoying the good.” Philosophical
Perspectives
23 (1): 253-272.
Lewis, David. 1983a. “Survival and Identity”. in Philosophical Papers I. Oxford
University Press.
_____. 1983b. Postcript to “Survival and Identity.” in Philosophical Papers I.
Oxford University Press.
_____. 1986. On The Plurality of Worlds. Wiley Blackwell Press.
Lowe, E.J. 1989. Kinds of Being. Oxford Blackwell.
Noonan, Harold. 2010. “The Thinking Animal and Personal Pronoun
Revisionism”.
Analysis 70(1): 93-98.
_____. (2003) Personal Identity, London: Routledge.
Noggle, Robert. 1999. “Integrity, the Self, and Desire Based Accounts of the
Good.”
30
Philosophical Studies 96(3): 301-328.
Nozick, Robert. 1974. Anarchy, State, and Utopia. Basic Books
Olson, Eric. 2002. “Thinking Animals and the Reference of ‘I’. Philosophical
Topics 30 (1): 189-207
_____. 2007. What Are We. Oxford University Press.
_____. 2010. “Ethics and The Generous Ontology.”
Overvold, Mark. 1980. “Self Interest and the Concept of Self-Sacrifice.”
Canadian
Journal of Philosophy 10(1):105-115.
Parfit, Derek. 1986. Reasons and Persons. Oxford Clarendon Press.
Quine, W. V. O. 1960. Word and Object. Cambridge: MIT Press.
Rawls, John. 1971. A Theory of Justice. Harvard University Press.
Raz, Joeseph. 2004. “The role of well‐being.” Philosophical Perspectives 18 (1):
269–
294.
Ross, William. 1930. The Right And The Good. Oxford University Press.
Russell, Betrand. 1927. The Analysis of Matter. New York: Harcourt Brace.
Searle, John. 1992. The Rediscovery of The Mind. Cambridge: MIT Press.
Sider, Theodore. 2001. Four Dimensionalism. Oxford University Press.
_____. 1997. “All The World’s a Stage.” Australasian Journal of Philosophy, 74:
433-453.
Sidgwick, Henry. 1981. The Methods of Ethics. Hackett Publishing.
Strawson, Galen. 1994. Mental Reality. Cambridge: MIT Press.
Tooley, Michael. 1997. Time, Tense and Causation. Oxford Clarendon Press.
van Inwagen, Peter. 1990. Material Beings. Cornell University Press.
Whitehead, Alfred. 1920. The Concept of Nature. Cambridge University Press.
Zimmerman, Dean. 1996. “Persistence and Presentism.” Philosophical Papers,
25: 115-126.
31