A plurality of values
by
Alan Carter
Abstract
Many maximizing normative theories are monistic in resting upon one core value. But
such theories generate highly counter-intuitive implications. This is especially clear
in the case of hedonistic utilitarianism. But an analysis of why we find those
implications counter-intuitive implies that we ought to subscribe to a plurality of
values. For example, the Repugnant Conclusion implies that we should value a high
level of average happiness, while the Problem of the Ecstatic Psychopath implies that
we should value either a large quantity of total happiness or a large number of
worthwhile lives. The problems posed by pleasure-wizards, on the other hand, imply
that we should include a non-utilitarian value: namely, equality. And only when such
values are kept in play simultaneously can the Repugnant Conclusion, the Problem of
the Ecstatic Psychopath and the problems posed by pleasure-wizards all be avoided,
thereby demonstrating the superiority of pluralist over monistic normative theories.
Words in main text: 10,177
Words in notes: 3,196
Words in abstract: 150
Alan Carter
Professor of Moral Philosophy
Department of Philosophy
69 Oakfield Avenue
University of Glasgow
Glasgow
G12 8LT
United Kingdom
Direct line: (+44) 141 330 4273
Fax: (+44) 141 330 4112
E-mail: A.Carter@philosophy.arts.gla.ac.uk
Webpage: http://www.gla.ac.uk/Acad/Philosophy/Personnel/Alan/carter.html
A plurality of values
Many maximizing normative theories are monistic in resting upon one core value. But
such theories generate highly counter-intuitive implications. This is especially clear
in the case of hedonistic utilitarianism. But an analysis of why we find those
implications counter-intuitive implies that we ought to subscribe to a plurality of
values. For example, the Repugnant Conclusion implies that we should value a high
level of average happiness, while the Problem of the Ecstatic Psychopath implies that
we should value either a large quantity of total happiness or a large number of
worthwhile lives. The problems posed by pleasure-wizards, on the other hand, imply
that we should include a non-utilitarian value: namely, equality. And only when such
values are kept in play simultaneously can the Repugnant Conclusion, the Problem of
the Ecstatic Psychopath and the problems posed by pleasure-wizards all be avoided,
thereby demonstrating the superiority of pluralist over monistic normative theories.
1. The presumption of value-monism
Wouldn’t it be nice if we were to discover that the physical universe was reducible to
only one kind of fundamental entity? Wouldn’t it be nice if we discovered that when
one of those entities combined with other entities of the same kind in the various
possible ways in which they can interlock, all other (larger) entities were thereby
produced? And wouldn’t it be nice if all physical, chemical, biological, psychological
and sociological laws were reducible to one simple law regarding the behavior of the
one kind of fundamental entity? There would be an elegance, simplicity and
tremendous explanatory power displayed by the theory that described such an entity
and its behavior. It is not surprising, therefore, that scientists and philosophers alike
should prefer to discover, and have thus sought, a single fundamental entity (whether
it be an elementary particle, a wave-function or a Lewisian spacetime point), along
with a single grounding law.
Wouldn’t it be nice, too, if we were to discover that the moral universe was
reducible to only one kind of valuable entity—or one core value, for short? And
wouldn’t it be nice if we discovered that all moral injunctions could be derived from
one simple principle concerning the one core value, with the simplest and most
natural thought being that we should maximize it? There would be an elegance,
2
simplicity and tremendous justificatory power displayed by the normative theory that
incorporated the one simple principle. The answers to all moral questions would, in
theory at least, be both determinate and determinable. It is hardly surprising,
therefore, that many moral philosophers should prefer to identify, and have thus
sought, the one simple principle that would, hopefully, ground morality.1 And it is
hardly surprising that many moral philosophers, in seeking the one simple principle,
should have presumed, explicitly or tacitly, that morality must ultimately be grounded
upon the maximization of a solitary core value, such as quantity of happiness or
equality, say.
Now, the assumption—what I shall call the presumption of value-monism—that
there is to be identified a single core axiological value that will ultimately ground all
of our correct moral decisions has played a critical role in the development of ethical
theory, for it clearly affects our responses to certain thought-experiments, and, in
particular, our responses concerning how our normative theories should be revised or
concerning which ones ought to be rejected. To demonstrate how this is so, allow me
to begin by considering two thought-experiments that seem to invite intuitive
responses that themselves appear to call for a significant revision to our normative
theories.
2. The Repugnant Conclusion
One normative theory that many have entertained is classical hedonistic utilitarianism,
where we are enjoined, directly or indirectly, to maximize one core axiological value:
the balance of pleasure over pain—or the greatest happiness, for short—for all those
who are deemed morally considerable. But there is a standard thought-experiment that
appears to require the abandonment of classical utilitarianism. Imagine that society A
contains half the number of people as those contained in society B, but the people in B
are, on average, slightly more than half as happy as those in A. If we have to decide
which of these two societies is better, then classical utilitarianism tells us that B is the
better of the pair because it contains the greater total quantity of happiness. And were we
to bring about society B we would thereby maximize the sole core value lying at the
core of classical utilitarianism.
1
See, for example, Henry Sidgwick, The Methods of Ethics, 7th ed. (Indianapolis: Hackett, 1981), Bk.
III, ch. xi & Bk. IV, ch. ii.
3
But if classical utilitarianism tells us that society B, with its larger population, is better
than society A, with its, on average, happier population, and if B contains half the
number of people as those in society C, but the people in C are slightly more than half as
happy, on average, as those in B, then classical utilitarianism also tells us that society C,
with an even larger population and even less happy people, on average, is better than
society B. And this argument can be iterated until we reach society Z, which Derek Parfit
describes as “an enormous population,” all with
lives that are not much above the level where they would cease to be worth living. A life
could be like this either because its ecstasies make its agonies seem just worth enduring, or
because it is painless but drab. Let us imagine Z to be of this second kind. There is nothing
bad in each of these lives; but there is little happiness, and little else that is good. The people
in Z never suffer; but all they have is muzak and potatoes. Though there is little happiness in
each life in Z, because there are so many of these lives Z is the outcome in which there
would be the greatest total sum of happiness.2
Classical utilitarianism thus seems to imply that, as Parfit observes, “[c]ompared with
the existence of very many people—say, ten billion—all of whom have a very high
quality of life, there must be some much larger number of people whose existence, if
other things are equal, would be better, even though these people would have lives that
are barely worth living.”3 As this implication strikes many with repugnance, it has, of
course, become widely known as the Repugnant Conclusion.
However, it should be noted that our intuitions can go somewhat astray, here. For
many might regard this implication as repugnant because of how unbearable they would
eventually find a world in which the only thing to eat was potatoes and the only
entertainment consisted of muzak. Surely, one would become bored to death with the
lack of variety; and after years of muzak and potatoes one would desperately want
nothing more than to end it all. But then, this wouldn’t be an instance of the Repugnant
Conclusion, which describes a world in which everyone’s life is worth living; not one
in which everyone would rather die. Hence, the Repugnant Conclusion requires
magical muzak and magical potatoes: muzak that no one ever tires of hearing and
potatoes that no one ever tires of eating. But still, even when this is taken into
2
Derek Parfit, “Overpopulation and the quality of life” in Peter Singer (ed.), Applied Ethics (Oxford:
Oxford University Press, 1986), p. 148.
3
Ibid., p. 150.
4
account, it remains the case that having to regard the drab society Z as better than
society M, say, appears sufficiently counter-intuitive to require some major revision to
utilitarianism.4
3. The Problem of the Ecstatic Psychopath
We could, of course, avoid the Repugnant Conclusion by revising utilitarianism so as
to regard as the single core value the level of average happiness, rather than the total
quantity of happiness, of all those who are deemed morally considerable. This would
lead to the Average Theory, which, in contradistinction to the classical Total Theory,
enjoins us to bring about, directly or indirectly, the society with the highest level of
average happiness. And that is clearly not the society of the Repugnant Conclusion.
But now consider a second thought-experiment: If the least happy person were to
die painlessly, then the level of average happiness would rise. So, at least one
construal of the Average Theory5 seems to enjoin that person’s killing. And if the next
least happy person were also to die painlessly, the level of average happiness would
again rise. So, the Average Theory, at least on one construal, seems to require the
death of that person, as well. And this argument can be iterated until we are left with the
single, happiest person; for the world containing only that person has a higher level of
average happiness than any other. But a defender of the Average Theory is likely to
object that the one remaining person would be extremely miserable as a result of the
deaths of everyone else. Well, imagine that the remaining person is a psychopath with
no interest whatsoever in other people—a psychopath in a perpetual state of ecstasy,
4
There have, of course, been a number of attempts at avoiding the Repugnant Conclusion. But none to
date appears adequate. One approach is to reject the transitivity of the relation “x is better than y”. See,
for example, Larry S. Temkin, “Intransitivity and the mere addition paradox,” Philosophy and Public
Affairs 16, 2 (1987): 138–87, especially p. 157, n. 24, and Stuart Rachels, “A set of solutions to Parfit’s
problems,” Noûs 35, 2 (2001): 214–38. But this seems far too high a price to pay, especially if a
solution can be found that retains transitivity. Some, on the other hand, have simply chosen to pay a
different price by living with the Repugnant Conclusion. For example, rather than abandon
utilitarianism, John Broome prefers to doubt our intuitions concerning large numbers of people. See
John Broome, Weighing Lives (Oxford: Oxford University Press, 2004), pp. 57–8 & 212–13. Not
surprisingly, many would rather abandon, or at least significantly revise, utilitarianism.
5
For a number of different construals of the Average Theory, see Thomas Hurka, “Average
Utilitarianisms,” Analysis 42 (1982): 65–69, and Thomas Hurka, “More Average Utilitarianisms,”
Analysis 42 (1982): 115–19.
5
and therefore considerably happier than any of her victims had been. But a defender of
the Average Theory might proceed to object that even a psychopath needs other
people to supply her with food, shelter, and so on. Well, imagine that the ecstatic
psychopath has a machine that satisfies her every need. But a defender of the Average
Theory might then object that once the psychopath dies, the level of average
happiness will plummet to zero, which is lower than the level that would have
obtained had the psychopath not eliminated everyone else. Well, imagine that there is
a cloning machine that produces another, equally ecstatic, psychopath the moment she
dies.
As this is the world with the highest level of average happiness, then it appears that
it would be morally right, according to at least one construal of the Average Theory,
for the ecstatic psychopath to have killed, painlessly, every other person. (And note:
regarding sentient nonhuman animals as morally considerable6 exacerbates the
problem, for it seems that the ecstatic psychopath would be morally required to
exterminate them, too, given that they would be less happy than she was.) Now, it
might be our intuitions against killing that renders so counter-intuitive the conclusion
that a psychopath, accompanied by a machine to meet her every need and another to
clone her at death, was morally right to have killed everyone else. So, imagine instead
that, without anyone being killed, we have the choice of bringing about a world that
contains millions of very happy people or a world containing only an ecstatic
psychopath, a machine to meet her every need and a cloning device. The Average
Theory, on all construals of that theory, enjoins us to bring about, directly or
indirectly, the second world because it is the world that contains the highest level of
average happiness. Yet this is so counter-intuitive that it seems that we must reject the
level of average happiness as our ultimate value.
But we have also seen that the Repugnant Conclusion appears to imply that we
should no less reject the total quantity of happiness as our value. We seem to be left
high and dry, and it appears that utilitarianism ought to be discarded completely. 7 In
6
See, for example, Peter Singer, Animal Liberation: A New Ethics for Our Treatment of Animals (New
York: Avon Books, 1977).
7
It might be thought that utilitarianism can evade such counter-intuitive implications by adopting the
Person-Affecting Principle. See Jan Narveson, “Utilitarianism and new generations,” Mind 76 (1967):
62–72. But this principle generates the (even more counter-intuitive) Non-Identity Problem. See
Thomas Schwartz, “Obligations to posterity” in R. I. Sikora and Brian Barry (edd.), Obligations to
Future Generations (Philadelphia: Temple University Press, 1978), and Thomas Schwartz, “Welfare
6
sum, the presumption of value-monism forces us to conclude that neither total nor
average happiness are of ultimate value. For if there is only one core value to be
maximized, and if maximizing a single candidate value generates a seriously counter-
intuitive result, then it cannot be the sought-after value.
4. A case for plural values
But if we hold to be highly counter-intuitive the injunction to bring about, directly or
indirectly, a world whose sole sentient inhabitant is an ecstatic psychopath, it would
appear that we find this to be so because either such a low total quantity of happiness
or such a small number of worthwhile lives as is contained therein is of disvalue. But
the best explanation for this, it would seem, is that its contrary—namely, a large total
quantity of happiness or a large number of worthwhile lives—is valuable. On the
other hand, if we also hold to be highly counter-intuitive the injunction to bring about,
directly or indirectly, a world containing trillions of barely happy people, it would
appear that we find this to be so because such a low level of average happiness as is to
be found within such a world is of disvalue. But the best explanation for this, it would
seem, is that its contrary—a high level of average happiness—is valuable.8 In short,
surely the most plausible reason for the counter-intuitive nature of any mooted moral
requirement to bring about, directly or indirectly, the world of the ecstatic psychopath
is that either a large total quantity of happiness or a large number of worthwhile lives
is of value; and surely the most plausible reason for the counter-intuitive nature of any
mooted injunction to bring about, directly or indirectly, the world of the Repugnant
Conclusion is that a high level of average happiness is also of value.
How is it that we fail to notice something so obvious? I submit: because we are
inclined to dismiss summarily any value that fails to satisfy our desire for the one core
judgments and future generations,” Theory and Decision, 11 (1979): 181–94. However, see Alan
Carter, “Can we harm future people?” Environmental Values 10, 4 (2001): 429–454, for an attempt at
drawing some of the sting out of this particular problem. For different responses to the Non-Identity
Problem, see Derek Parfit, “On doing the best for our children” in M. D. Bayles (ed.), Ethics and
Population (Cambridge, MA: Schenkman, 1976), and Derek Parfit, “Energy policy and the further
future: the identity problem” in Douglas MacLean and Peter G. Brown, Energy and the Future (Totowa,
NJ: Rowman and Littlefield, 1983).
8
Perhaps it is its contradictory—not too low a level of average happiness—that is of value? This might
suggest a satisficing, rather than a maximizing, theory. If this seems more plausible, then see note 16,
below.
7
value—in other words, because of the presumption of value-monism. And whereas
locating the one core value at the heart of the one simple moral principle would be
intellectually satisfying in the extreme, traditional value-monism, which holds that
there is only one core value, is, quite simply, not true of our normative universe.
In a nutshell, then, the pursuit of the most elegant theory leads to the presumption
of value-monism. Because a solitary core value is then sought, each potential
candidate is examined in isolation. And the resulting failure of all those values when
each is considered separately is then taken as even stronger evidence against a
plurality of values, because of how many candidate values are thereby presumed to
have been ruled out. But this is to beg the question against a plurality of values from
the outset. And as a consequence of the widespread, uncritical acceptance of the
presumption of monism, we are sidetracked into the search for the Holy Grail of the
one core value—which can only serve to obstruct the development of an adequate
normative theory. Thus, our penchant for elegance can easily lead us away from the
truth and into error.
But once the presumption of monism is discarded, each of a plurality of values can
be viewed as generating counter-intuitive results only when it is misconstrued as the
sole value in play;9 for these counter-intuitive implications disappear once all
axiological values take to the field simultaneously, as I aim to demonstrate.
Consequently, and as counter-intuitive as it initially appears, the low level of average
happiness manifest in the Repugnant Conclusion does not demand, as is widely
thought, that we reject as a value either the total number of worthwhile lives or the
total quantity of happiness, even if it does provide compelling grounds for rejecting
the Total Theory. And nor does the small number of worthwhile lives or the small
total amount of happiness manifest in the world of the ecstatic psychopath demand
that we reject the level of average happiness as a value, even if one or other
consideration provides compelling grounds for rejecting the Average Theory. In other
words, and contrary to common belief, the Repugnant Conclusion counts in favor of a
value, not against it—as does the Problem of the Ecstatic Psychopath. And, as should
become transparent, it is only because of the presumption of monism that we fail to
see this. But given that our considered responses both to the Repugnant Conclusion
9
C.f. “[F]ollowing one good to the end may be catastrophic, not because it isn’t a good, but because
there are others which can’t be sacrificed without evil.” Charles Taylor, Sources of the Self: The
Making of the Modern Identity (Cambridge: Cambridge University Press, 1989), p. 503.
8
and to the Problem of the Ecstatic Psychopath are only justifiable if either the total
number of worthwhile lives or the total quantity of happiness, on the one hand, and
the level of average happiness, on the other, are both of value, then we have good
reason for regarding traditional value-monism, whether in the form of, say,
utilitarianism or Pure Egalitarianism,10 as fundamentally mistaken.
5. Total happiness or worthwhile lives?
We have seen that there is reason to believe that average happiness is valuable. And
we have also seen that there is reason to believe that either total happiness or the
number of worthwhile lives is valuable, too. But which? If, as Parfit notes, an
additional group of people all live worthwhile lives and produce no adverse effects,
should their level of wellbeing be below that of the rest, and even if this implies an
inequality in the world, it “does not seem to justify the view that the extra group should
never have existed. Why,” he asks, “are they such a blot on the universe?” 11 And, surely,
they aren’t. Perhaps the best explanation is because an extra life that is worth living and
which causes no problems at all would be a good thing.12
Of course, one reason why it might be assumed that the number of worthwhile
lives is not a value is the distaste that might be felt at the prospect of an overcrowded
world. But this would be to assume that the only way in which the number of
worthwhile lives can be increased is synchronically. But we can guarantee a larger
number of lives overall if many of them live at different times than if they all live at
once.13 And if we consider the largest number of worthwhile lives as occurring
diachronically, then we can evade the counter-intuitive implication of having to bring
about an overcrowded world. So, imagine that we have a comfortable number of
10
For the distinction between Pure and Pluralist Egalitarianism, see Derek Parfit, Equality or Priority?:
The Lindley Lecture (Kansas: University of Kansas, 1995), p. 5.
11
Parfit, “Overpopulation and the quality of life,” op. cit., p. 152.
12
As Parfit surmises: “it is good if any extra life is lived, that is worth living.” Ibid., p. 147. However, for a
challenge to the view that additional lives are of value, see Jan Narveson, “Moral problems of
population,” The Monist 57 (1973): 62–86. For a cogent rebuttal, as well as arguments generally in
support of valuing the number of worthwhile lives, see Broome, Weighing Lives, op. cit., chs 10–12 & 14.
13
Robin Attfield deploys a similar consideration in order to defend the maximization of utility as an
ultimate value. See Robin Attfield, The Ethics of Environmental Concern, 2nd ed. (Athens, GA: University
of Georgia Press, 1991), pp. 127–8. However, as we can see, it also facilitates a defense of the number of
worthwhile lives as one of our core values.
9
people living at one time. Most of us think that it would be better if another
generation were to succeed them. And it would be better still if yet another generation
succeeded those successors. Regardless of whether or not there is a duty to preserve
our species, it seems that the greater the number of generations of people living
worthwhile lives, the better the world would be. And this seems to imply that the
greater the number of worthwhile lives, the better.
Hence, we have reason for regarding the number of worthwhile lives as a value.
So, let us provisionally regard it as such. If we also hold that average happiness is a
value, too, then a large number of worthwhile lives enjoying a high level of average
happiness would thereby provide a high total quantity of happiness, which would,
derivatively, be of value. (And this is why, perhaps, we might mistake the total
quantity of happiness for a core value.) So, assume that both average utility and
number worthwhile lives are core axiological values, and also assume that their
product—the total quantity of happiness—is only a derivative value. If we are faced
with a choice between bringing about, directly or indirectly, population A, whose sole
member is an ecstatic psychopath, or population Z (of the Repugnant Conclusion),
then we would be facing a choice between maximizing the level of average happiness
or maximizing the number of worthwhile lives. How might we compare these values
in such a manner as to make rational moral choices? Or put more pointedly, if there is
a plurality of values, then how can there be a determinate answer to moral questions,
even in principle?
6. Axiological two-dimensionality
Brian Barry argues in his early work that we could model trade-offs between
principles such as equity and efficiency in a manner that parallels the way in which
micro-economists employ indifference-curves to model how we might swap grapes
for potatoes.14 What is of note is that it is not the case that in deciding how many
grapes should be exchanged for each potato, 1 potato straightforwardly equals 10
grapes, say. For if I have 1,000 grapes and no potatoes, I may well swap 100 grapes
for 1 potato. But if I have 100 potatoes and no grapes, I may well swap 5 potatoes for
1 grape. So, if, in Figure 1, the number of grapes were to be measured along the
14
See Brian Barry, Political Argument (London: Routledge, 1965), p. 5. Unfortunately, Barry never
pursued this insight as far as one might have wished.
10
horizontal x-axis and the number of potatoes were to be measured along the vertical y-
axis, we would be prepared to trade grapes for potatoes in the manner represented by
the indifference-curves EV, DW, CX and BY.
What if we could adopt an analogous strategy in order to compare not grapes and
potatoes, but core values, such as the level of average happiness and the number of
worthwhile lives? For, surely, we do think that if there is an extremely low level of
average happiness and a colossal number of worthwhile lives, then it would have
been better if there had been fewer worthwhile lives and a higher level of average
happiness. And, surely, we do think that if there is an extremely small number of
worthwhile lives and a colossally high level of average happiness, then it would have
been better if there had been more worthwhile lives and a lower level of average
happiness.15 The number of worthwhile lives and the level of average happiness are,
therefore, values that we do, as a matter of fact, believe can be traded.
But then, what can we possibly mean by saying that it would have been “better” if
there had been more of one value? If we were to call values such as the number of
worthwhile lives and the level of average happiness “contributory values”, and if we
were to call what they contribute towards “overall value”, then we would be able to
claim that when we say that it would have been better if there had been more of one
value, we are saying that there would have been more overall value if there had been
more of the contributory value in question.
So, in order to model how contributory values can be traded so as to maintain
overall value, imagine that the x-axis of Figure 1 measured, instead of the number of
grapes, the number of worthwhile lives,16 and the y-axis measured, instead of the
number of potatoes, the level of average happiness. If we consider a very large
population at a very low level of average happiness, as represented by point Z, then a
society with noticeably fewer people would be no better and no worse so long as there
was the right increase in the level of average happiness. On the other hand, if there
was a very small population at a very high level of average happiness, as represented
by point A, then, so long as there was the right increase in the number of worthwhile
15
See Thomas Hurka, “Value and population size,” Ethics 93, 3 (1983): 496–507, especially pp. 498–9.
16
If it is felt that the earlier argument for maximizing the number of worthwhile lives lacks cogency, but
that within a certain range it is better that more live worthwhile lives than that fewer do, then simply
regard the scale of the x-axis of Figure 1 as confined to that range. The same could be said for any axis
we might subsequently consider. And this is how we might accommodate a satisficing theory; for,
surely, while a value remains below the satisfactory level, the more the world has of it, the better.
11
lives, a society with a considerably lower level of average happiness would be equally
valuable, overall. Were we able to plot trade-offs between different contributory
values accurately so as to maintain overall value, then we could map “isovalue-
contours”—contours resulting from some variably weighted, aggregate core-value
function. So, consider the imagined isovalue-contours mapped in Figure 1: Every
point on the dotted curve BY possesses equal overall value.17 Moreover, every point
on the solid curve CX possesses equal overall value, too. Yet any point on CX is better
than any point on BY. Furthermore, every point on the solid curve DW possesses equal
overall value. Yet any point on DW is better than any point on CX. In addition, every
point on the solid curve EV possesses equal overall value. Yet any point on EV is
better than any point on DW. 18
Why, precisely, would it be the case that any point on the curve DW, for example,
was better than any point on the curve CX?19 Point T on DW would be better than
point S on CX both in terms of the number of worthwhile lives and in terms of the
level of average happiness (as the dashed lines make clear). So, T would have to be
better overall than S. Now, there are points on DW that would be an improvement on
T in terms of the level of average happiness; but that would be exactly compensated
for by the greater number of worthwhile lives that T represents. And there are points
on DW that would be an improvement on T in terms of the number of worthwhile
lives; but that would be exactly compensated for by the higher level of average
happiness represented by T. Because every point on DW would be of equal overall
value, and because one of those points (namely, T) is better than S, then every point
on DW would have to be better than S. Now, there are points on CX that would be an
improvement on S in terms of the level of average happiness; but that would be
exactly compensated for by the greater number of worthwhile lives that S represents.
And there are points on CX that would be an improvement on S in terms of the
number of worthwhile lives; but that would be exactly compensated for by the higher
level of average happiness represented by S. This is why S would be equal in overall
value to every other point on CX. But if every point on DW is better than S, and if S is
17
Actually, it is the worlds represented by those points that are of equal overall value. But from now on,
to avoid an overly cumbersome exposition, when I say that a point is of a certain value I shall mean
that the world it represents is valuable to that degree.
18
It should be noted that, unlike the representations in Figure 1, it is quite possible that the slope and
even the shape of the isovalue-curves could change as one moves further from the origin.
19
I thank Philip Percival for pressing me to provide the following explication.
12
equal in overall value to every other point on CX, then any point on DW would have
to be better than any point on CX.20
For parallel reasons, any point on curve EV would have to be better than any point
on curve DW, and any point on curve CX would have to be better than any point on
the dotted curve BY. In short, the further north-east the isovalue-curve is from the
origin O, the more valuable, overall, would be any point falling upon it.
7. Introducing practicability
But as every point on an isovalue-curve possesses the same overall value as any other
point on that curve, then, were there no further considerations, it is difficult to see
how we could rationally choose to bring about some particular society rather than one
with a greater number of worthwhile lives or one with a higher level of average
happiness. For Figure 1 simply presents an array of possible worlds without any
indication of which outcomes we could actually bring about. The isovalue-contours
tell us nothing whatsoever about the physical limitations governing our actions. So,
imagine that in Figure 2—where the x-axis again measures the number of worthwhile
lives, and where the y-axis again measures the level of average happiness—the dashed
curve AZ (call it “the practicability frontier”) represents the boundary of all
practicable outcomes. In other words, all of the possible populations that we could
practicably bring about are represented by points that fall within the area OAZ. (And it
seems reasonable to assume that the dashed AZ in Figure 2 would take such a shape
because finite resources, for example, could be used to make a small number of
people extremely happy, a moderate number of people happy or a very large number
of people barely happy.)
If we now superimpose Figure 2 onto Figure 1, we obtain Figure 3. Point T in
Figure 3, because it falls on the dashed curve AZ, is a practicable outcome. But it also
falls on curve DW, which is further to the north-east from the origin O than any of the
isovalue-contours (drawn or not drawn) upon which fall the other points either on the
dashed AZ or between the dashed AZ and O. Thus, of every practicable outcome, T
20
There would, of course, be points on the curve CX that were an improvement on point T in terms of
the number of worthwhile lives; but that would be more than compensated for by the higher level of
average happiness that T represents. There would, equally, be points on CX that were an improvement
on T in terms of the level of average happiness; but that would be more than compensated for by the
larger number of worthwhile lives represented by T.
13
would be the optimal. Any outcome that fell upon EV would be even better, but none
is practicable—none is within our power to bring about.
In other words, if we are consequentialists, and even though we are working with
two core values—the number of worthwhile lives and the level of average happiness
—we would have, in this case, a determinate answer to the question of which society
it would be best, directly or indirectly, to bring about: namely, that society which
corresponds in population size and level of average happiness to point T. And
crucially, such a society would be neither the world of the solitary ecstatic psychopath
nor the overpopulated society of the Repugnant Conclusion, which, were they
practicable, would be represented by points A and Z, respectively. Those points fall on
an isovalue-curve (the dotted curve BY) so close to the origin O that we would find
them wholly unacceptable, given that point T, falling on DW (and far further to the
north-east from O than BY), is a practicable outcome. Indeed, the dotted BY is so close
to the origin that we might regard all points on it as simply beyond the pale.21 As such,
we might well refuse even to countenance the possibility of A or Z ever constituting
morally acceptable outcomes. Hence, we have the basis for a form of two-dimensional
value theory—a normative theory that recognizes two different contributory values—
that, at least in principle, provides an answer simultaneously to the Problem of the
Ecstatic Psychopath and to the Repugnant Conclusion.
So, it should now be evident why it is that when two contributory values are on
stage simultaneously, we could avoid the counter-intuitive implications that befall a
reliance upon either value performing on its own. Consequently, neither the
Repugnant Conclusion nor the Problem of the Ecstatic Psychopath require the
abandonment of either the total number of worthwhile lives or the level of average
happiness as contributory values. To the contrary, it is only by retaining both values
that we can solve both problems together. And, clearly, it is the presumption of value-
monism that has played a major role in obstructing the development of this solution.
In short, because it has been widely assumed that there must be only one core value,
each contributory value has been assessed on its own; and each then falls prey either
to the Repugnant Conclusion or to the Problem of the Ecstatic Psychopath. But what
neither value can achieve by itself, both can manage in concert.22
21
This argument thus supports a key premise in Alan Carter, “Moral theory and global population,”
Proceedings of the Aristotelian Society 99, 3 (1999): 289–313.
22
It is now also possible to see how greater sense might be made of “the greatest happiness of the
14
8. Two preliminary objections
Now, it seems possible that the slope of the isovalue-curves relative to that of the
practicability frontier could be very different to what Figure 3 indicates. Figure 3 is
based, in part, upon Figure 1; and Figure 1 presupposes that the slope of the isovalue-
curves will be neither too vertical when approaching the x-axis nor too horizontal
when approaching the y-axis. But it might be objected that, were the isovalue-curves
in Figure 1 re-drawn with significantly different slopes, Figure 3 might fail to rule out
as suboptimal either point A or point Z.23
But if the level of average happiness were not very important relative to the
number of worthwhile lives whenever a gigantic population experienced an extremely
low level of average happiness (as at point Z in Figure 3), then the Repugnant
Conclusion would not be so counter-intuitive. Given just how counter-intuitive the
Repugnant Conclusion appears, the isovalue-curves cannot, therefore, be relatively
vertical when nearing the x-axis, for that would represent a proportionally lower
weighting of average happiness at Z. On the other hand, if the number of worthwhile
lives were not very important relative to the level of average happiness whenever a
tiny population enjoyed an extremely high level of average happiness (as at point A in
Figure 3), then the Problem of the Ecstatic Psychopath would not be so counter-
intuitive. Given just how counter-intuitive the Problem of the Ecstatic Psychopath
appears, the isovalue-curves cannot, therefore, be relatively horizontal when nearing
the y-axis, for that would represent a proportionally lower weighting of the number of
worthwhile lives at A.24 Thus, A and Z will indeed be suboptimal, unless, that is, the
practicability frontier of Figure 3 is implausibly horizontal when nearing the x-axis or
implausibly vertical when nearing the y-axis.
The above objection concerns the slope of the isovalue-contours. A second
greatest number,” which is peculiar in having two maximands: for it could be construed, contra
Narveson, as signaling “the highest level of average happiness and the greatest number of worthwhile
lives;” and we could represent the outcome that it enjoins, directly or indirectly, by means of point T in
Figure 3.
23
I am indebted to Nicholas Baigent for pressing this objection.
24
For a different argument justifying the assumption that the shape of the isovalue-contours would be
convex when viewed from the origin O, see Alan Carter, “Some groundwork for a multidimensional
axiology,” forthcoming.
15
objection that might be raised at this juncture concerns the shape of the practicability
frontier. For Figure 3 is also based, in part, upon Figure 2; and Figure 2 appears to
presuppose that the practicability frontier would be concave when viewed from the
origin O. What, though, if this assumption were to be challenged? If the practicability
frontier were, instead, convex when viewed from the origin, then it would still be the
case that there would be an optimal outcome as long as the practicability frontier were
less convex than the isovalue-contours. And this is because, with the exception of the
point representing the optimal outcome, every point that falls on the practicability
frontier or between the practicability frontier and the origin (which exhausts all
practicable outcomes) would lie on an isovalue-curve closer to the origin than the
isovalue-curve upon which falls the point representing the optimal outcome; and
hence would represent a suboptimal outcome.
So, all that would be required for a determinate optimal outcome, were we able to
plot the various curves, would be that, when viewed from the origin O, the isovalue-
contours were more convex than the practicability frontier. But this could be safely
assumed to be the case as long as a fall in the level of average happiness was tradable
for a significantly smaller increase in the number of worthwhile lives whenever there
was a very large level of average happiness, and as long as a decrease in the number
of worthwhile lives to be brought into existence was tradable for a significantly
smaller rise in the level of average happiness whenever a very large number of
worthwhile lives was already guaranteed. The satisfaction of these two conditions to a
sufficient degree would generate asymptotic isovalue-contours, and should thus
simultaneously answer both objections. And our responses to the Problem of the
Ecstatic Psychopath and to the Repugnant Conclusion suggest that both conditions
are, in fact, adequately satisfied.25
9. Diminishing marginal utility and equality
I have argued that there is reason to regard at least two contributory values as
necessary for any adequate normative theory: (1) the level of average happiness; and
(2) the number of worthwhile lives. But should the contributory values be confined to
25
For responses to a number of other objections that might be leveled against the axiology adumbrated
here, see Alan Carter, “A pre-emptive response to some possible objections to a multidimensional
axiology with variable contributory values,” forthcoming, and Alan Carter, “How to solve two addition
paradoxes and avoid the Repugnant Conclusion,” forthcoming.
16
this pair?
In order to answer this question, let me return to classical utilitarianism.
Utilitarians have gone to some length to show that their normative theory does not
enjoin us to act in a manner that is wildly at odds with commonsense morality.26 One
consideration that utilitarians have sought to deploy for bringing utilitarian injunctions
more in line with commonsense thinking about moral matters is that of diminishing
marginal utility; and this manoeuvre has proven especially valuable because one set of
powerful objections to utilitarian thinking focuses upon its seeming disregard for how
happiness is to be distributed. But as a poor person would derive more happiness from
an additional $1,000 than would a millionaire, then, ceteris paribus, an equal
distribution should ordinarily result in more happiness in the world than an unequal
one.27 Hence, utilitarians have a seemingly impressive argument for why maximizing
happiness will not lead to severe inequalities, which would be a counter-intuitive
implication in the eyes of many: for a greater total quantity of happiness could be
argued to result from a less, rather than from a more, unequal distribution of
resources, given that one’s happiness at the margin tends to decline the more
additional resources one enjoys. And because the classical utilitarian holds that we are
enjoined to bring about the greatest quantity of happiness, then he or she will
conclude that we should distribute resources relatively equally.
But one apparent oddity is that utilitarians usually appear both relieved and
contented by their deployment of this argument. This is odd because if the only value
is the total quantity of happiness or the level of average happiness, then utilitarians
should, unlike those in the grip of commonsense morality, be wholly indifferent to
whether or not utilitarianism implies severe inequalities.28 But perhaps their
26
Jeremy Bentham, for example, points to “secondary mischiefs”, such as the pain of apprehension, to
show why, at least at times, innocent persons should not be sacrificed for the common good. See
Jeremy Bentham, “An introduction to the principles of morals and legislation” in idem., A Fragment on
Government with an Introduction to the Principles of Morals and Legislation (Oxford: Blackwell,
1948), p. 265–6.
27
The ceteris paribus clause is required because it might be objected that the anticipation of an equal
distribution would remove incentives; and therefore it might be argued that a guaranteed equal share of
the cake would result in a smaller cake, and thus in less happiness overall.
28
It should also be noted that the assumption that monetary increases result in diminishing marginal
utility across the board can be challenged. See Harry Frankfurt, “Equality as a moral ideal” in idem.
The Importance of What We Care About (Cambridge: Cambridge University Press, 1988), pp. 138–44.
17
contentment lies purely in showing that monistic utilitarianism is less inconsistent
with commonsense morality than is often thought, thereby making it less implausible
as a normative theory?
10. The price of pleasing pleasure-wizards
Even if the utilitarian’s sole source of contentment in showing that, because of
diminishing marginal utility, his or her normative theory is not as incompatible with
egalitarianism as it would otherwise have been is that it makes the theory less
implausible to the unconverted, there remains a telling objection to its adequacy with
respect to the question of distribution. For unfortunately, as Amartya Sen notes,
different people have different utility functions. So, imagine a utility-monster29 or a
pleasure-wizard: namely, a person more efficient than others at converting resources
into her own personal happiness. Such a being would undermine the egalitarian
pretensions of classical utilitarianism, because, as Sen writes,
if person A as a cripple gets half the utility that the pleasure-wizard person B does from
any given level of income, then in the pure distribution problem between A and B the
utilitarian would end up giving the pleasure-wizard B more income than the cripple A.
The cripple would then be doubly worse off: both since he gets less utility from the same
level of income, and since he will also get less income. Utilitarianism must lead to this
thanks to its single-minded concern with maximizing the utility sum.30
But if we imagine an extreme pleasure-wizard, this problem doesn’t just apply to the
disabled person. For classical utilitarianism would seem to enjoin us to give all the
resources above those required for keeping everyone else at a level where their lives
would be barely worth living to the pleasure-wizard because she would thereby gain
so much pleasure as a result that there would be a greater total amount of happiness in
the world than would result from either a less unequal distribution of resources or
from a less unequal distribution of happiness. We might therefore imagine the
Repugnant Conclusion, except for the addition of a mind-bogglingly ecstatic pleasure-
wizard. But if classical utilitarianism enjoins such a massive inequality—where
everyone but the ecstatic pleasure-wizard lived at the level of the (egalitarian) society
29
See Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), p. 41.
30
Amartya K. Sen, “Equality of what?” in idem., Choice, Welfare and Measurement (Oxford: Blackwell,
1982), p. 357.
18
(Z) of the Repugnant Conclusion—then we would, in the considered judgment of
many, have arrived at an even more repugnant conclusion.
Now, utilitarians are prone to dismiss objections based upon such mythical beings
as extreme pleasure-wizards because utilitarianism was supposedly developed in
order to deal with real-world problems; and such pleasure-wizards, to put it bluntly,
do not exist and, what is more, never will. But their opposites do. And the opposites
of pleasure-wizards—namely, those who are unusually inefficient at converting
resources into happiness—suffice to ruin the utilitarian’s egalitarian pretensions.
Consider, for example, those who suffer from, what are currently, incurable diseases.
If such persons are in great distress at the prospect of their untimely deaths from their
maladies, then an increase in their happiness would require that a huge proportion of
society’s resources be diverted towards finding a cure for their rare condition. Any
attempt at a genuine equality of happiness would drag everyone down to the level of
these unfortunates. Thus, the total amount of happiness is maximized by diverting
resources away from those who are unusually inefficient at converting resources into
happiness. In other words, if the goal is, solely, to maximize the total amount of
happiness, then giving anything at all to such people and spending anything on cures
for their illnesses is a waste of valuable resources. Hence, given the actual existence
of such unfortunates, the maximization of happiness requires a considerable
inequality in its distribution.
But if real-world examples suffice to show that the utilitarian cannot avoid certain
counter-intuitive implications (such as the injunction not to provide any help
whatsoever to the worst-off), matters get considerably worse when fictive thought-
experiments are further developed (and one might hope that the correct normative
theory was capable of dealing with purely imaginary examples, too). For suppose that
the ecstatic psychopath turned out also to be an extreme pleasure-wizard! Monistic
classical act utilitarianism must enjoin her painlessly to kill everyone else so that she
could convert the resources they would otherwise have used into her own individual
happiness. But if it is, again, suspected that it is the killing that strikes us as immoral,
then imagine, instead, choosing between bringing about, without anyone having been
killed, either a world whose sole sentient inhabitant is an ecstatic, psychopathic
pleasure-wizard or a world containing millions of moderately happy people. As both
the highest level of average happiness and the greatest total quantity of happiness
would be found in a world where the psychopathic pleasure-wizard enjoyed all of its
19
resources, then the Total Theory, and not just the Average Theory, enjoins us to bring
about such a world when we are able to do so rather than a more populous one. But
that is counter-intuitive in the extreme. In a word, if our moral intuitions count for
anything, then monistic classical utilitarianism must surely be false.
However, if both the number of worthwhile lives and the level of average
happiness are of value, then a psychopathic pleasure-wizard would not be morally
required to kill everyone else. Nor would we be morally required to bring about,
without killing anyone, a world where the solitary, ecstatic, psychopathic pleasure-
wizard enjoyed all of its resources. We would, instead, be enjoined to choose the
population represented by point T in Figure 3, which is far more numerous.
Unfortunately, we would also be morally required to keep everyone with the
exception of the extreme pleasure-wizard at the level of happiness they would
experience in the Repugnant Conclusion, and to transfer the resources that they would
otherwise have enjoyed to the pleasure-wizard. For this would result in both a large
number of worthwhile lives and in the highest level of average happiness for that
population (and in the greatest total amount of happiness at that size of population,
too). But, as noted earlier, this seems even more repugnant than the Repugnant
Conclusion. The presence of the pleasure-wizard thus turns utilitarianism into the
converse of Robin Hood’s moral philosophy: we should take from the poor to give to
the rich! So, simply relying on the two contributory values that might have been
thought to ground utilitarianism—the number of worthwhile lives and the level of
average happiness (which, when multiplied together, generate, as a derivative value,
the total happiness)—would not enable us to avoid all seriously counter-intuitive
implications.
But there is an even more telling problem that the extreme pleasure-wizard poses
for utilitarianism. As there would be both a greater total quantity of happiness and a
higher level of average happiness in the world were such a pleasure-wizard to exist—
even if that meant that everybody else suffered in being denied the resources needed
to produce in her such unimaginable ecstasy—the consistent utilitarian should
bemoan the fact that pleasure-wizards are to be found only in our fantasies. Perhaps
one might hold the view that suffering overrides another’s pleasure—the core
intuition of negative utilitarianism—and this would rule out this particular
conclusion? But then think again of the world containing one ecstatic pleasure-wizard
and a large population of normal people whose lives are just barely worth living. That
20
world would have a greater total quantity of happiness and a higher level of average
happiness than is to be found in this world of ours. So, the consistent utilitarian should
greatly regret the non-existence of pleasure-wizards; and the utilitarian should do so
even when the existence of extreme pleasure-wizards would morally require everyone
else to be no more than barely happy.
As I have yet to meet a utilitarian, and certainly not a monistic one, who admits to
thinking that the world would be a better place if it contained an extreme pleasure-
wizard living alongside a very large population all at that level of happiness where
their lives were just barely worth living, I must conclude that utilitarians are deceiving
themselves about the values they actually hold dear. Of course, monistic utilitarians of
the hedonistic brand do genuinely think that they believe that the only ultimate value
is the total amount of happiness or the level of average happiness. But if they do not
bemoan the lack of pleasure-wizards, then they must surely value equality directly,
even if they hide that fact from themselves. And this suggests that the smile of
contentment on the faces of utilitarians after they have deployed diminishing marginal
utility in an attempt to show that their normative theory is not incompatible with
egalitarianism has more to do with their valuing of equality than they are prepared to
admit (even to themselves).
11. Axiological three-dimensionality and beyond
Now, if we find the injunction to bring about a world containing billions of barely
happy people living alongside an ecstatic pleasure-wizard (who enjoys a huge
quantity of resources at their expense) to be at least as counter-intuitive as the
injunction to bring about a world containing trillions of barely happy people, as, I
suspect, most of us do, then it appears that this is so because such an inequality is of
disvalue. But the best explanation for this, it would seem, is that its contrary—
equality—is valuable. In short, surely the most plausible reason for the counter-
intuitive nature of the injunction to bring about, directly or indirectly, a world
containing billions of barely happy people living alongside an ecstatic pleasure-
wizard is that equality is also of value. Consequently, due to the problems posed to
utilitarianism by pleasure-wizards, we now have reason to factor in a distributional
component as well as average happiness and the number of worthwhile lives.
Clearly, many think that if there were an extremely low level of average happiness
21
as a result of pursuing equality—the leveling-down objection to egalitarianism—then
it would have been better if there had been less equality and a higher level of average
happiness. And many also think that if there were a colossally high level of average
happiness in a world of tremendous inequality, then it would have been better if there
had been a lower level of average happiness and less inequality. The degree of
equality and the level of average happiness are, therefore, values that we do, as a
matter of fact, believe can be traded. But if we now recognize three contributory
values—number of worthwhile lives, average happiness and equality 31—how, it might
be challenged, could there still be a determinate answer to any moral questions?
Given that we have already seen how an answer could, at least in principle, be
provided when there are two contributory values, the reply is surprisingly easy:
simply imagine that a third axis has been added to our graph, and imagine that three-
dimensional isovalue-planes and a three-dimensional practicability frontier can be
plotted, as in Figure 4 (where the x-axis measures the number of worthwhile lives, the
y-axis measures average happiness, and the z-axis32 measures the degree of equality).
Point T would fall on the three-dimensional practicability frontier AZU, and thus
would represent a practicable outcome. But it would also fall on the isovalue-plane
DWG, which would be further to the north-east and out of the page from the origin O
than any of the other imaginable isovalue-planes upon which would fall the other
points within the volume OAZU (which would represent all practicable outcomes). T
would therefore constitute the optimal outcome. Hence, adding an additional value
would not render the identification of the best practicable world any less
determinate.33 And even with these three values, we would still be on as strong a
31
It might be the case that “equality” is a composite of more than one value. See Alan Carter, “Value-
pluralist egalitarianism,” Journal of Philosophy 99, 11 (2002): 577–99. Indeed, it is possible that
“average happiness” might turn out to be a composite of more than one value, too. Alternatively, it
might transpire that “average perfectionism”, rather than “average happiness”, is an ultimate value. See
Thomas Hurka, Perfectionism (Oxford: Oxford University Press, 1993), pp. 71–5. I thank Jon Tresan
and Kirk Ludwig for pressing me on this point. Furthermore, it might also be the case that “number of
worthwhile lives” actually stands as a proxy for something else that increases pari passu with its
increase. I thank Luc Bovens for pressing me on this point.
32
The z-axis should be viewed as coming out of the page.
33
Further dimensions measuring additional values (such as autonomy) could in principle be handled
mathematically, but they could not be depicted by means of graphs. But that does nothing to diminish
the heuristic value of this general approach.
22
ground as we were previously for avoiding both the Repugnant Conclusion and the
Problem of the Ecstatic Psychopath. But the additional valuing of equality would also
sidestep the distributional problems that pleasure-wizards pose for utilitarianism (and
it would do so in the same way that valuing average happiness enables the Repugnant
Conclusion and valuing the number of worthwhile lives enables the Problem of the
Ecstatic Psychopath both to be sidestepped).
It is worth emphasizing that we would have been unable to solve these problems
simultaneously had we chosen total happiness instead of the number of worthwhile
lives as one of our core values. Imagine that the axes on Figure 4 measured total
happiness, average happiness and equality. As the world inhabited by the solitary,
psychopathic, extreme pleasure-wizard displays a high level of average happiness, a
large total quantity of happiness and complete equality, it would constitute the
optimal outcome. These three values on their own, therefore, cannot explain why any
injunction to bring about such a world is so counter-intuitive. But were total happiness
to be resolved into number of worthwhile lives and average happiness, then as the
world of the solitary, psychopathic, extreme pleasure-wizard contains only a single
worthwhile life, such a world would clearly be suboptimal. This seems to confirm our
original choice as contributory values average happiness and number of worthwhile
lives, to which we have subsequently added equality.
It should also be noted that multidimensional isovalue-planes and a
multidimensional practicability frontier would only need to be deployed when the
core values cannot all be maximized simultaneously. In short, this idealized normative
technology is especially suited to dealing with worst-case scenarios. For in many real-
world cases, a reduction in inequality will lead to an increase in the level of average
happiness. However, it might be thought that an increase in the number of worthwhile
lives would always reduce the level of average happiness. But that would be
obviously so only when people live beyond the Earth’s carrying capacity. And if too
many people live at the same time, they will so damage our planet’s life-support
systems that there will be no future generations at all. But more people can live on
this fragile planet of ours over a considerable period of time than can be crammed
onto it at any one time. Hence, the greatest number of worthwhile lives would, in fact,
need to be measured diachronically rather than synchronically, and it would require
people living sustainable lifestyles.34 This suggests a measure of equality between
34
See Attfield, op. cit., p. 128.
23
generations; and there seems no compelling reason why sustainability necessitates an
unacceptably low level of average happiness.35
12. Indirect consequentialism, virtue ethics and deontology
Were it assumed that we ought to try to bring about the optimal point T in Figure 4
directly, then, clearly, we would have a form of act consequentialism. And thus far, I
might have given the impression that this is what a plurality of values entails. But is
this the only option allowed by such a multidimensional axiology? Ideally, the
axiology would tell us which “bundle” of a set of contributory values contained the
greatest amount of overall value,36 given the various bundles that are, in actual fact,
practicable. But identifying the practicable world that contains more overall value
than any other does not necessarily tell us what to do. For it might be the case that the
direct pursuit of the greatest quantity of overall value would be self-defeating. In a
35
Perhaps it is because several values can often be promoted by one action that traditional value-
monism has appeared less implausible than it would have otherwise. I thank Hugh LaFollette for this
observation.
36
What, precisely, is this “overall value” that could, at least in principle, be measured? Well, it might be
a simple, unanalyzable, nonnatural but real property. It might be a natural property that is not reducible
to other natural properties. It might be what a fully rational and fully informed version of ourselves
would desire that we, as we are in certain respects, desire. It might be construed as the subjective
preferences of an individual evaluator. It might be what the majority of the members of a culture judge
favorably. It might be what merits a certain response in an evaluator sensitized to the saliency of the
morally relevant features. It might be the strength of a pro-attitude. Or it might simply be a fiction
projected onto the world as a result of our nature (perhaps “tinged” with the evaluator’s culture or
modified by intersubjective adjustments). It might be something else altogether. My aim has been for
the normative approach adumbrated here to be as neutral as possible with respect to the ontology of
value. If I have been successful in this endeavor, then regardless of whether one is, metaethically, some
variety of realist, subjectivist, cultural relativist, sensibility theorist, emotivist, norm-expressivist,
quasi-realist projectivist or error-theoretic projectivist, then some of the strengths of value-pluralism as
a normative theory should now be more apparent than before. This notwithstanding, and merely for the
record, I should add that my suspicion is that some form of error-theoretic projectivism is the most
plausible metaethical position. See Alan Carter, “Humean nature,” Environmental Values 9, 1 (2000):
3–37, and Alan Carter, “Projectivism and the Last Person Argument,” American Philosophical
Quarterly 41, 1 (2004): 51–62.
24
word, a value-pluralist axiology37 can ground an indirect form of consequentialism38
no less than can the axiology of classical utilitarianism.39
How, precisely? Considerations of space militate against anything but the briefest
of sketches,40 so it will have to suffice for the present if we simply imagine that we
could most effectively approach the optimal point T in Figure 4 by promulgating
certain obligatory rules. For example, morally right actions could be construed as acts
in conformity with those rules that are such that the greatest overall value (found at T)
would be advanced by their violators being liable to incite blame or to arouse feelings
of guilt. Such rules would likely resemble those that deontologists often advocate. But
such deontic rules—rules that it would be morally wrong to break—would be
grounded, instead, upon a value-pluralist axiology. And they might be expected to
consist of moral requirements that, for example, were conducive to protecting and
promoting the greatest number of worthwhile lives, the highest level of average
happiness and the most equal distribution of resources and welfare. But we would also
expect those rules not to promote, say, the highest level of average happiness at a
great cost to equality. Hence, we can see why such rules, like most, if not all, moral
rules, would incorporate exceptions (and many exceptions seem to be an inexplicable
feature of moral rules grounded on a monistic axiology). In addition, value-pluralism
seems, in general, nearer to commonsense morality than monistic utilitarianism. And
37
Does talk of “overall value” imply that the axiology presented here is, rather than counting as a form
of value-pluralism, actually a form of value-monism in disguise? If it does, then it is emphatically not a
form of “value-monism” as traditionally understood; for just as total happiness could be resolved into
average happiness and number of worthwhile lives, overall value, which increases as we move in
Figure 4 from the origin O to the north-east and out from the page, is resolvable into a plurality of
contributory values. Value-monism, as it is traditionally construed, mistakes one of those component
values for value as such. I thank Timothy Williamson for pressing me to provide this clarification. For
more on the terms “value-monism” and “value-pluralism”, see Alan Carter, “Two senses of ‘value-
pluralism’,” forthcoming.
38
See Peter Railton, “How thinking about character and utilitarianism might lead to rethinking the
character of utilitarianism” in idem., Facts, Values, and Norms: Essays toward a Morality of
Consequence (Cambridge: Cambridge University Press, 2003).
39
For an illuminating account of John Stuart Mill’s indirect utilitarianism, see John Gray, Mill on
Liberty: A Defence, 2nd ed. (London: Routledge, 1996).
40
For a more substantial argument, see Alan Carter, “Indirect multidimensional consequentialism,”
forthcoming.
25
just as rule utilitarianism seems less at odds with it than act utilitarianism,41 a value-
pluralist form of rule consequentialism might appear even closer still to commonsense
morality (which is not to deny that the latter may need revising).
Alternatively, we might imagine that we could most effectively approach the
optimal point T in Figure 4 by encouraging certain motives42 or by inculcating certain
dispositions. Virtues might, for example, be construed as dispositions that are such
that the greatest overall value (found at T) would be advanced by praising those who
possess them,43 and vices might be construed as dispositions that are such that the
greatest value would derive from those possessing them being liable to provoke
denigration or to arouse feelings of shame. A value-pluralist axiology could,
therefore, be employed to ground a form of virtue ethics. 44 And one virtue endorsed
by that axiology could well be respect for those rules that are such that the greatest
value overall would be advanced by their violators being liable to incite blame or to
arouse feelings of guilt.
Up until now, the argument has been purely consequentialist in form. But there is
nothing to prevent deontological constraints from being factored into a normative
model of this general type. For it seems highly implausible that rights, say, are so
important that their protection always trumps the general welfare. In a word, it
appears nothing short of fanatical to insist that one should never violate another’s
rights even to avoid global catastrophe. But in order to avoid such fanaticism, one
would have to be prepared to trade some rights violations for general wellbeing, say.
The normative technology imagined above could easily include a further dimension
measuring deontological considerations,45 if that were deemed necessary.
41
See Alan Carter, “Is utilitarian morality necessarily too demanding?” in Timothy Chappell (ed.),
Demandingness in Ethics (Basingstoke: Palgrave, forthcoming).
42
In other words, a value-pluralist could easily adapt the argument in Robert M. Adams, “Motive
consequentialism,” Journal of Philosophy 73 (1976): 467–81.
43
C.f.: “[I]n distributing our praise of human qualities, on utilitarian principles, we have to consider
primarily not the usefulness of the quality, but the usefulness of the praise…”. Sidgwick, The Methods
of Ethics, op. cit., p. 428.
44
Virtues may turn out to be intrinsically, rather than merely instrumentally, valuable. For an account of
intrinsically valuable virtues that could easily be made compatible with the axiology adumbrated here,
and which could also be construed as supporting it, see Thomas Hurka, Virtue, Vice, and Value
(Oxford: Oxford University Press, 2001), esp. ch. 3.
45
For example, we could simply add a dimension measuring 1/(some function of the number and
seriousness of rights violations). And in order to visualize the inclusion of such a deontological
26
13. Concluding remarks
I have argued that the presumption of value-monism can prevent utilitarians, for
example, from seeing that (1) the total number of worthwhile lives, (2) the level of
average happiness and (3) equality (at the very least) must all be of value. And it
appears that there must be a plurality of such core values, for only then do our
considered responses to the Problem of the Ecstatic Psychopath, to the Repugnant
Conclusion and to the problems posed by pleasure-wizards make any consistent sense.
But if this is so, then we should reject the presumption of monism. And once it has
been discarded, we are no longer driven to conclude that neither average happiness
nor the number of worthwhile lives nor equality are valuable; for they only generate
counter-intuitive implications when each is regarded as the sole value in play. 46 In
short, whereas it has been standard to marshal, for example, the Repugnant
Conclusion and pleasure-wizards for dismissing certain values in order that a targeted
normative theory may be rejected, the novel role they have played here is that of
establishing certain values as contributory values, and, thereby, of justifying a form of
value-pluralism.
Once we have grounds for accepting some form of value-pluralism, the key
axiological question becomes: What is the relationship between the core values—
including those of quantity, quality and distribution? When evaluating different
possible outcomes, we sometimes find ourselves trading the number of worthwhile
lives for average happiness, and trading average happiness for equality. It is an
interesting question how we are able to trade such core values. But the absence of an
answer should not incapacitate our normative project. For it is a plain fact of moral
consideration, we could change one of the axes in Figure 4 so that it measured, instead, the
deontological value in question. If that contributory value were such as to require exceptional
circumstances for it to be outweighed, then that would be reflected in the slope of the isovalue-planes.
46
For a survey of some of the problems facing monistic egalitarianism, see Alan Carter, “The quest for
an egalitarian metric,” Critical Review of International Social and Political Philosophy 7, 1 (2004):
94–113; and for an account of how the counter-intuitive implications of monistic egalitarianism can be
avoided when certain conceptions of what ought to be equalized are combined with average happiness
within a value-pluralist framework, see Carter, “Value-pluralist egalitarianism,” op. cit. For an
argument indicating that egalitarianism might be more pluralist than egalitarians usually presume, see
Alan Carter, “A defense of egalitarianism,” Philosophical Studies 131, 2 (2006): 269–302.
27
life that we just do engage in such trade-offs.47 So, adopting this as a fundamental
starting point, we proceeded by trying to model those trade-offs by imagining the
shape, were we able to plot them accurately, that isovalue-contours, measuring overall
value, would take, where those contours result from the various bundles of any two
contributory values. And it seems clear that they would be convex when viewed from
the origin. We then tried to imagine the shape that the practicability frontier would
take, when it represents bundles of those contributory values, were we able to plot it.
And it seems reasonable to assume that it would be concave when viewed from the
origin. Given that we would then expect there to be a point where the practicability
frontier would be tangential to an isovalue-curve further to the north-east from the
origin than any other that included points representing a practicable bundle of
contributory values, that point (point T in Figure 3) would represent the bundle of
greatest overall value. Thus, abandoning value-monism does not necessarily entail
that there would no longer be a determinate answer to any moral questions—even if
the answer is only one in principle.
A striking feature of this idealized normative technology is that it reveals more
perspicuously than has hitherto been the case how it is that maximizing, monistic
axiologies will, inevitably, turn out to be inadequate.48 For, given all the practicable
outcomes (as indicated by the shape of the practicability frontier AZ in Figure 3), the
point where the practicability frontier comes closest to the axis measuring the sole
value recognized by the monistic theory in question represents the practicable bundle
that contains more of that value than any other. Hence, any maximizing, monistic
axiology will pull the outcome it identifies as the most valuable towards an axis; 49 and
47
For one response to several arguments that seem to imply the incomparability or, alternatively,
incommensurability of values, see Ruth Chang, “Against constitutive incommensurability or buying
and selling friends,” Philosophical Issues 11 (2001): 33–60, especially pp. 51–2. Also see Donald
Regan, “Value, comparability, and choice” in Ruth Chang (ed.), Incommensurability, Incomparability,
and Practical Reason (Cambridge, MA: Harvard University Press, 1997).
48
Furthermore, it manages to do so while preserving the transitivity of the relation “x is better than y”.
For a point (say, T in Figure 3) on any two-dimensional isovalue-plane is better than a point (say, S) on
one closer to the origin, and that point (S) is better than a point (say, Z) on an isovalue-plane even
closer to the origin, while the first point (T), of course, remains clearly better than the third (Z). See
Alan Carter, “A solution to the purported non-transitivity of normative evaluation,” forthcoming.
49
Indeed, this tendency to mislocate the optimal outcome by pulling it towards an axis will also be
displayed by any pluralist theory that accords lexical priority to one of the values it recognizes.
28
hence it will pull it onto an isovalue-curve, such as the dotted curve BY, which is far
too close to the origin O. As there is a practicable bundle of core values (namely,
point T) that falls on an isovalue-curve (DW) far further to the north-east from the
origin, and hence of far greater overall value, any theory that enjoins any outcome
falling on the dotted curve BY is clearly wanting. But that is precisely what any
maximizing, monistic axiology will do. To drive the point home for any maximizing,
monistic theory, one merely needs to measure on one of the axes in Figure 3 whatever
it is that the monistic theory in question insists to be of sole value. 50 Put another way,
by pondering the shapes that the isovalue-curves and practicability frontier would
have to take, we can see that, in pursuing only one core value, any maximizing,
monistic axiology is inevitably doomed to flout our other contributory values—which
is why a number of the implications arising from maximizing, monistic axiologies are
condemned to strike us as counter-intuitive, for they will score badly in terms of some
core value other than the one the axiology in question seeks to maximize.51
An additional contributory value, as we have seen, can easily be accommodated by
moving from the two-dimensional isovalue-curves of Figure 3 to the three-
dimensional isovalue-planes of Figure 4. And the strategy of imagining such idealized
isovalue-planes is highly instructive in showing simultaneously why it is that the
Repugnant Conclusion, the Problem of the Ecstatic Psychopath and the problems
posed by pleasure-wizards, as well as the problems posed by wholesale leveling
down,52 represent situations of severely suboptimal overall value. Thus, a
multidimensional value theory is uniquely suited to dispatching in one fell swoop the
Repugnant Conclusion, the Problem of the Ecstatic Psychopath, the problems posed
50
Is Pure Egalitarianism a maximizing moral philosophy? If its goal is to minimize inequalities, even if
confined within a certain range, then we can re-interpret its goal as, within that range, maximizing 1/
(the degree of inequality). And this would make it count as a monistic, maximizing theory in the sense
intended, and subject to the present criticism.
51
For the observation that this is clearly so with respect to the maximization of total happiness, see
Tyler Cowen, “What do we learn from the Repugnant Conclusion?” Ethics 106, 4 (1996): 754–75,
especially p. 758.
52
Wholesale leveling down will appear desirable only if the level of average happiness (or the total
quantity of happiness) is deemed to be of no value. For it will score highly on an axis measuring
equality; but it will score badly on an axis measuring the level of average happiness. Diverting
resources to pleasure-wizards, on the other hand, will score highly on an axis measuring the level of
average happiness; but it will score badly on an axis measuring equality.
29
by pleasure-wizards and the problems posed by leveling-down, and, along with them,
the Total View, the Average View53 and Pure Egalitarianism.54
But to allay one possible fear, while the accurate plotting of the isovalue-planes
and practicability frontier of Figure 4 would allow us, in practice, to identify the
optimal outcome for three ultimate values, my argument against traditional value-
monism does not rest on the assumption that they can, in fact, be plotted. For the
strategy has been, rather, merely that of imagining the general shapes that such planes
and frontier would take. And if we are justified in thinking that the isovalue-curves
and practicability frontier of Figure 3 are approximately of the correct shape, then that
provides reason to conclude that the isovalue-planes and practicability frontier
depicted in Figure 4 are roughly of the correct shape, too.55 But if this is so, then that
is sufficient to adjudge maximizing, monistic theories inadequate, because it then
becomes abundantly clear that they will fail to direct us towards the optimal outcome.
For, crucially, as Figure 4 visibly reveals, the determinate axiological result (namely,
point T) would be neither the Repugnant Conclusion (as seemingly enjoined by the
53
Parfit assumes that what is required to solve the problem of the optimum population is “Principle (X)”
or “Theory X”. For the latter, see Derek Parfit, Reasons and Persons (Oxford: Oxford University Press,
1984), p. 390. Unfortunately, Parfit is unsure of the details of such a theory. However, elsewhere,
where he seeks “Principle (X)”, he is explicit in stating that “the best-known candidates for the role of
(X)” are “the Average and Total Views.” Parfit, “Energy policy and the further future,” op. cit, p. 296n.
We, on the other hand, have seen that there is reason to conclude that neither view will do.
54
Others have explored the use of indifference curves within moral philosophy. See, for example,
Barry, Political Argument, op. cit., p. 5; John Rawls, A Theory of Justice (Cambridge, MA: Harvard
University Press, 1971), pp. 37–40; Susan L. Hurley, Natural Reasons: Personality and Polity (Oxford:
Oxford University Press, 1989), pp. 69–81; and, most fruitfully, Hurka, Perfectionism, op. cit., pp. 88–
98. But no one to my knowledge has successfully shown how isovalue-planes when combined with a
practicability frontier can simultaneously solve the major problems generated by the Total View, the
Average View and Pure Egalitarianism.
55
Even if, as a matter of fact, we are unable to plot Figure 4, this does not prevent it from providing us,
at the very least, with a convenient rule of thumb: (I) it is better to have a higher level of average
happiness than a lower one, except when that would result in too few worthwhile lives or too great an
inequality; (II) it is better to have more worthwhile lives than fewer, except when that would result in
too low a level of average happiness or too great an inequality; and (III) it is better to have less
inequality than more, except when that would result in too few worthwhile lives or too low a level of
average happiness. Moreover, even if it cannot be plotted, Figure 4 serves to remind us that, when
engaging in our moral deliberations, we should pause to consider whether all relevant core values have
been fully accommodated.
30
Total View), the world of the solitary, Ecstatic Psychopath (as seemingly enjoined by
the Average View) nor, say, an impoverished, leveled-down world (as might,
arguably, be enjoined by Pure Egalitarianism). Thus, in avoiding the Repugnant
Conclusion, the Problem of the Ecstatic Psychopath and the problems posed by
pleasure-wizards, as well as the problems posed by any unmitigated demand to level
down, we appear to have identified an axiology that is far more consistent with our
considered moral judgments than any entailing these counter-intuitive implications.
But if the fate of all maximizing, monistic value theories is to entail counter-
intuitive implications, then it is difficult to see how any non-maximizing, monistic
theory would fare any better,56 especially when any monistic normative theory,
whether maximizing or not, in excluding either equality or average happiness or the
number of worthwhile lives (or even the total quantity of happiness) as core values
will fail to justify our considered responses to the problems posed by pleasure-wizards
or to the Repugnant Conclusion or to the Problem of the Ecstatic Psychopath. All
monistic normative theories must, therefore, be insufficient at best.
Finally, given that a value-pluralist axiology can ground forms of both act and
indirect consequentialism, and that multidimensional isovalue-planes would be
capable of incorporating deontological considerations, then it should now be apparent
that value-pluralism has far more to recommend it (and that traditional value-monism
has considerably less going for it) than has been fully appreciated to date.57
56
Interestingly, Mark Timmons argues that certain seemingly monistic theories are considerably more
plausible when interpreted as pluralistic. See Mark Timmons, Moral Theory (Lanham, MD: Rowman
& Littlefield, 2002), passim. However, he concludes from this that the most plausible normative
theories will therefore necessarily display a measure of indeterminacy. The normative technology
proposed here implies that this need not be the case.
57
I am grateful to audiences at the Universities of Edinburgh, Florida, Glasgow, Hull, Miami,
Nottingham and South Florida at St Petersburgh, Georgia State and North Carolina State Universities,
Trinity College, Dublin and the London School of Economics and Political Science. A shortened
version of this paper was also presented at the Joint Sessions of the Aristotelian Society and the Mind
Association held at the University of Aberdeen.
31
B C D E
A
T V
S
W
X
Y
O Z
Figure 1
32
A
T
O Z
Figure 2
33
B C D E
A
T V
W
X
Y
O Z
Figure 3
34
D
A
T
W
G
O Z
U
Figure 4
35