Professional Documents
Culture Documents
37 Virtue’s Reasons
New Essays on Virtue, Character, and Reasons
Edited by Noell Birondo and S. Stewart Braun
42 Moral Skepticism
New Essays
Edited by Diego E. Machuca
Benjamin Sachs
First published 2018
by Routledge
711 Third Avenue, New York, NY 10017
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2018 Taylor & Francis
The right of Benjamin Sachs to be identified as author of this work
has been asserted by him in accordance with sections 77 and 78 of
the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Names: Sachs, Benjamin, author.
Title: Explaining right and wrong : a new moral pluralism and its
implications / by Benjamin Sachs.
Description: 1 [edition]. | New York : Routledge, 2017. | Series:
Routledge studies in ethics and moral theory ; 43 | Includes
bibliographical references and index.
Identifiers: LCCN 2017044434 | ISBN 9781138307353
(hardback : alk. paper)
Subjects: LCSH: Ethics—Methodology. | Normativity (Ethics)
Classification: LCC BJ37 .S23 2017 | DDC 170/.44—dc23
LC record available at https://lccn.loc.gov/2017044434
ISBN: 978-1-138-30735-3 (hbk)
ISBN: 978-1-315-14230-2 (ebk)
Typeset in Sabon
by Apex CoVantage, LLC
To my parents, Marcia and Raymond, for their
unfailing support
Contents
Acknowledgments x
Introduction 1
3 Against Monism 32
5 Non-Rossian Pluralism 83
10 Conclusion 158
This book has been a long time in the making—about 13 years if you start
counting from when I began writing what turned out to be Chapter 9. In
that time I’ve received an immense amount of help in writing this book,
both from colleagues at the various places where I’ve worked and studied
and from friendly acquaintances in the broader philosophical community.
They’ve been incredibly generous to me and most of them have asked for
little to nothing in return.
My most sincere thanks goes to Volko Strahovnik and an anonymous
referee for Routledge, each of whom read and commented on the full draft
manuscript. Thanks as well to Andrew Weckenmann and Alexandra Sim-
mons at Routledge for shepherding me through the publication process.
To the people who initially helped me to work out what this whole
project was about and provided encouragement as I was getting started—
Nick Beckstead, Molly Gardner, Dale Jamieson, S. Matthew Liao, Japa
Pallikkathayil, Doug Portmore, Bill Ruddick, Samuel Scheffler, and Russ
Shafer-Landau—I’m grateful. My gratitude as well to Katherine Hawley,
who helped me with my prospectus.
Marcia Baron, John Broome, Jessica Brown, Jonathan Dancy, Kather-
ine Hawley, Dale Jamieson, Joseph Kisolo-Ssonko, Brian McElwee (twice),
Douglas Portmore, Mike Ridge, Toni Rønnow-Rasmussen, Russ Shafer-
Landau, Justin Snedegar, Christine Straehle, Saranga Sudarshan, Jens Tim-
merman, and Pekka Väyrynen each read and commented on a draft of a full
chapter of this book and for that I thank them sincerely.
Finally, I want to express my appreciation for the long list of people with
whom I’ve had helpful conversations about the ideas presented in this book—
a list that I suspect I’ve left incomplete (apologies to those I’ve forgotten to
mention): Chrisoula Andreou, Derek Ball, Greg Bognar, Anna Burkard, Ben
Colburn, Rowan Cruft, Miranda Fricker, Patrick Greenough, Dan Haus-
man, Katherine Hawley, Ulrike Heuer, Dale Jamieson, Hallie Liberto (many
times), Brian McElwee, Tristram McPherson, Uri Leibowitz, Collin O’Neil,
Theron Pummer, Wlodek Rabinowicz, Martin Rendall, Benedict Rumbold,
Ben Saunders, John Skorupski, Nick Treanor, Pekka Väyrynen, Leigh Vicens,
Daniel Viehoff, Kenny Walden, Jonathan Way, and Brynn Welch.
Acknowledgments xi
The ideas presented here were initially presented in conference/workshop
talks and seminars at the London School of Economics and Political Science,
the University of Glasgow, the Ethics and Explanation Workshop at Not-
tingham University, the University of Stirling, The Conference of the Inter-
national Society for Utilitarian Studies, The Joint Session of the Aristotelian
Society and the Mind Association, the University of Sheffield, the British
Society for Ethical Theory annual conference, Cal Poly San Luis Obispo,
King’s College London, the University of Connecticut, the University of
Massachusetts-Boston, the University of Utah, the Polish-Scottish Philoso-
phy Workshop, and my home institution, the University of St. Andrews.
Finally, I thank: Oxford University Press for permission to reprint, in
Chapter 3, material from my “Non-consequentialist Theories of Animal
Ethics”, Analysis Reviews 75 (2015): 638–54; Springer for permission to
reprint, in Chapters 4 and 5, material from my “Direct Moral Grounding
and the Legal Model of Moral Normativity”, Ethical Theory and Moral
Practice 18 (2015): 703–16; and John Wiley and Sons for permission to
reprint, in Chapter 7, material from my “The Status of Moral Status,” Pacific
Philosophical Quarterly 92 (2011): 87–104.
Introduction
Say what you will about act-consequentialism and Kantianism, at least they
hold out the promise of teaching us something that we didn’t already know.
Unapologetic act-consequentialists like Peter Singer and Shelly Kagan are
famous for telling us that widespread beliefs about infanticide and about the
moral obligatoriness of donating large portions of one’s wealth to charitable
causes are mistaken. And Kant offers us very surprising conclusions about
masturbation, dishonesty, and suicide.
By contrast, in the last 30 years or so we’ve been introduced to a lot of
normative ethical theories that not only don’t yield bold conclusions about
what we morally ought to do but in fact appear to be specifically designed
not to. Even act-consequentialists and Kantians have gotten in on the act.
Act-consequentialists have modified their theory by adding features such as
agent-relative value and by attributing value to integrity and the upholding
of rights, and these alterations serve to moderate what would otherwise be
act-consequentialism’s strange and demanding set of obligations. Kantians,
meanwhile, have been hard at work showing that we can stay within Kant’s
basic moral framework while arriving at more common-sense conclusions
about masturbation, deception, suicide, etc.
Contemporary act-consequentialists and Kantians are joined in this effort
by those philosophers who prefer to build a normative ethical theory around
the ideas of rights and duties. Although it may seem, superficially, that we
leave ourselves very little flexibility once we commit ourselves to the exis-
tence of rights and duties, given how those concepts have a bit of an abso-
lutist air about them, this turns out not to be so. Deontologists are quick to
point out that rights and duties can be pro tanto. This makes the question of
what one morally ought to do a matter of weighing considerations, at least in
those cases that aren’t entirely one-sided. Noticing this, an increasing number
of non-consequentialists these days dispense entirely with talk of rights and
duties and simply talk in terms of moral considerations or reasons.
My view on this trend is that if, at the end of the day, what we get out of
our moral theorizing is a theory that yields all the conclusions that we were
ex ante inclined to accept, what was the point of the whole endeavor? I had
hoped moral theorizing would be a way of learning something about what
I morally ought to do.
2 Introduction
This is not to deny that we should continue to talk of moral reasons. I have
my own theory about where moral reasons come from and what explanatory
work they do; I explain it in Chapters 4–5. What I do deny, however, is that
moral reasons talk is helpful in normative and applied ethics. Moreover, it’s
pernicious, because it’s readily available to be used as a tool for avoiding
ever having to admit the falsity of a substantive claim about right and wrong
that one was initially inclined to accept.
Avoiding this conservative bias in moral theorizing is easy if you’re
attracted to some very concrete all-encompassing generalization about what
one morally ought to do. For instance, act-consequentialists have the idea
that one always ought to maximize the good, and Kantians have the idea that
one always ought to obey the categorical imperative, and these commitments
serve as dialectical anchors that enable act-consequentialists and Kantians
to pull off the Ulysses-like feat of avoiding the temptation to accommo-
date every case-specific intuition that comes along, if they’re inclined to try
(which, as I noted already, most of them these days are not). But what if, like
me, you’re not attracted to any version of moral monism?
For moral pluralists, the answer has traditionally been that one should
use the method of reflective equilibrium, which involves sometimes adjust-
ing one’s theoretical beliefs in light of one’s case-specific (or what I will here-
after call ‘verdictive’) beliefs but also being willing to adjust one’s verdictive
beliefs in light of one’s theoretical beliefs. I will show, however, that we’ve
been too quick to accept this answer. We haven’t gotten clear on why one
would ever be justified in adjusting one’s verdictive beliefs in light of one’s
theoretical beliefs, and therefore as yet we can’t legitimately claim to know
why the method of reflective equilibrium is the correct moral methodology.
And I suspect that it’s our ignorance as to why the method is correct that
explains why we pluralists are always running afoul of it—why, that is, we
customarily adopt either the strategy of being biased toward holding on to
our verdictive intuitions when they conflict with our theoretical convictions
or the strategy of rigging things so that the theoretical claims we accept
deliver no verdictive implications at all because they’re merely pro tanto.
My view is that we can vindicate the classic understanding of moral theo-
rizing, on which it’s a way of correcting one’s mistaken beliefs about what one
morally ought to do and also (naturally) a way of learning which theoretical
claims are correct, if and only if we treat moral theorizing as an explanatory
project. Therefore, I begin the book by arguing, in Chapter 1, that the theo-
retical claims contained in normative ethical theories should be represented,
for the purposes of assessing the relative merits of those theories, as explana-
tory claims—claims about what explains the verdictive facts. I then dis-
cuss how we should go about assessing the plausibility of explanatory claims
and how we should understand the dialectical relationship between claims as
to what explains the verdictive facts (“explanatory claims”) and claims as to
what those verdictive facts are. I endorse the method of reflective equilibrium
and the widely held view of that method on which it requires two sets of
Introduction 3
intuitions as inputs, one of which should be verdictive intuitions. I simply add
that if I am right that moral theorizing should be understood as an explana-
tory project, then our explanatory intuitions should constitute that second set.
Chapter 2 offers a detailed picture of a moral methodology that includes an
appeal to our explanatory intuitions.
Beginning in Chapter 3 I deploy the method of reflective equilibrium,
understood the way I argue in Chapters 1–2 that it ought to be understood.
In that method it is standard practice to begin by identifying ‘fixed points’—
intuitions so firmly held that they constitute non-negotiable demands on
the resulting sets of theoretical and verdictive beliefs. Using a small num-
ber of such fixed points among explanatory intuitions, I argue for discard-
ing the three most popular versions of moral monism: consequentialism,
Kantianism, and contractarianism/contractualism. Then, in Chapter 4, I use
another such fixed point as a basis on which to reject the most popular
version of moral pluralism, namely Rossianism. These arguments are based
not on idiosyncratic intuitions, but rather on widespread intuitions that we
moral philosophers rarely use on account of our failure to recognize that
moral theorizing is an explanatory project.
The arguments of Chapters 3–4 leave us with a family of moral views I
call Non-Rossian Pluralism (NP), and in Chapter 5 I flesh out and defend
that family. Next, in Chapters 6–7, I get more specific about which kind of
NP we should accept. I argue in Chapter 6, again on explanatory grounds,
that NP should be welfaristically distribution-insensitive—i.e., that it should
deny that it matters morally in itself how welfare is distributed. And in
Chapter 7 I argue that NP should deny that facts about the moral status (or
moral standing or moral considerability or membership in the moral com-
munity) of a living thing can be part of the explanation of why treating it a
certain way is impermissible, permissible, or obligatory.
The last two chapters continue the project of giving further shape to NP
while also exploring its implications for real world cases. In Chapter 8 I
argue that there is no such thing as a duty of rescue, and based on that
claim I argue for an absolutist position with respect to the ethics of harmful
research on animal subjects. Finally, in Chapter 9 I focus my attention on
the distribution of health care resources, arguing that there is nothing that
could explain the truth or falsity of the proposed substantive answers to
the questions in this area and that we are therefore forced to conclude that
the questions either have no answer at all or no answer that ordinary moral
theorizing can reveal.
Chapters 8–9 constitute an admission that NP has drastically counter-
intuitive implications. Since I endorse the method of reflective equilibrium
and thus the use of verdictive intuitions in moral theorizing, I cannot take
this outcome lightly, and I make this problem the topic of Chapter 10, my
conclusion. There I propose an admittedly oversimplified overall assessment
of the prospects for NP and its competitors, on which NP has the advan-
tage of according well with our explanatory intuitions and the drawback of
4 Introduction
violating many of our verdictive intuitions, while many of its competitors
have the advantage of consistency with our verdictive intuitions and the
fault of doing violence to many of our explanatory intuitions. Based on an
argument to the effect that our explanatory intuitions are, for now, at least
as worthy of our trust as our verdictive intuitions, I conclude with the claim
that accepting NP is provisionally at least as justified as accepting any of its
competitors.
One caveat, however: If one does wish to simply dismiss NP, and with
it all the highly counterintuitive conclusions of Chapters 8–9, I admit that
there is a perfectly legitimate reason to do so even if one cannot identify a
flaw in my reasoning. This entire book is framed by its first chapter, which is
an effort to understand how we should approach moral theorizing; yet there
is no argument in this book for the claim that we should engage in moral
theorizing as our way of arriving at our substantive moral beliefs, though
I do see myself as undermining, in §3 of Chapter 2, the most influential
argument against the fruitfulness of moral theorizing. Still, this book’s lack
of an argument in favor of moral theorizing means that the controversial
conclusions contained in Chapters 8–9 are really disguised conditionals of
the form “If we should base our substantive moral beliefs on the output of
a sound process of moral theorizing then we should believe [some substan-
tive moral claim].” If, having read those chapters, the reader wants to reject
those controversial conclusions without identifying a flaw in my arguments
for them, I cannot quibble. The message of this book is that we can hang on
to all of our common-sense verdictive beliefs, but if we do then we shouldn’t
pretend that those beliefs are supported by the solid doxastic justification
that moral theorizing can provide. And if our beliefs aren’t justified in that
way, then the question becomes whether they’re justified at all.1
Note
1. We might have to fall back on an appeal to the existence of a reliable faculty of
verdictive judgment, as the particularist Jonathan Dancy does (2004, Chapter 8),
or assert that some verdictive moral claims are self-evident. I won’t pursue the
matter here.
1 Normative Ethical Theorizing
as an Explanatory Project
1. Introduction
My goal in this chapter is to establish that when we assess normative ethi-
cal theories we should represent their generalizations as explanatory claims
and assess those claims for intuitiveness. In §2 I argue that if we do this then
moral theorizing can be a way of gaining knowledge about what one mor-
ally ought to do. In §3 I take up an argument to the effect that doing this
constitutes bad methodology, while in §4–5 I consider and reject two argu-
ments to the effect that doing this is not necessary for moral theorizing to be
a way of gaining knowledge about what one morally ought to do. Finally,
in §6, I note the implications this proposed methodology has for the theory
and practice of the method of reflective equilibrium.
6. An Interesting Corollary
Before drawing this chapter to a close I want to point out that once we
acknowledge the points made above, we open up the possibility of resolving
a tension between the theory and the practice of moral reasoning.
Most moral philosophers believe that the method of reflective equilibrium
is the right method to use when trying to determine which normative ethical
theory to accept. This method involves first arriving at various considered
judgments, some verdictive and some general, and then making adjustments
to those judgments so as to make them consistent with each other. The offi-
cial dogma regarding the method is that whether an initial considered judg-
ment is verdictive or general should make no difference to one’s willingness
to adjust or discard that judgment.14 However, in practice this rule is fre-
quently violated, as anyone familiar with the literature in normative and
applied ethics will have noticed. Philosophers generally demonstrate a much
greater reluctance to give up on a verdictive judgment than to give up on a
general judgment. What’s going on here?
My explanation is twofold. First, most moral philosophers seem to believe
that the generalizations contained in normative ethical theories are to be
represented as categoricals or conditionals. If that were true, then it would
make perfect sense to be more reluctant, when deploying the method of
reflective equilibrium, to be biased in favor of giving up general judgments
over giving up verdictive judgments. As I said above (§4), there’s nothing
in a moral categorical or conditional to remain attached to once one has
searched for counterexamples to it. So if one discovers that a candidate
14 An Explanatory Project
categorical or conditional faces several apparent counterexamples, then of
course one should give it up.15
Second, I suspect that many moral philosophers experience an ever-so-subtle
temptation to make their moral reasoning parallel to good scientific reasoning,
and that this is because they’ve failed to notice the disanalogies between
moral and scientific reasoning that I pointed out in §5.16 The generalizations
contained in scientific theories aren’t apt for being judged on intuitiveness
and often posit entities and processes whose existence cannot be confirmed
via observation at least for the time being, whereas their particular judgments
(judgments about natural events) are based on observation. Because of this, it
makes sense in scientific theorizing to hold one’s initial particular judgments
more fixed than one’s initial general judgments. By contrast, we can, as I
argued in §2, represent the generalizations contained in normative ethical
theories as explanatory claims, thereby making them apt for evaluation
based on their intuitiveness or lack thereof. When we combine this insight
with the uncontroversial claim that verdictive claims in ethics, like morally
explanatory claims, are apt for evaluation based on their intuitiveness or lack
thereof, we get to the conclusion that proper moral inquiry is epistemically
balanced where proper scientific inquiry is epistemically imbalanced: there is
no sense in which moral philosophers have better access to the particular facts
than to the theoretical facts, whereas for scientific theorists this certainly is
the case. Because of this balance in moral inquiry there is no reason, as there
is in the scientific case, to be more hesitant to give up one’s initial particular
judgments. Thus, acknowledging the potential role of explanatory claims in
moral reasoning reinforces the correctness of the official dogma of the method
of reflective equilibrium while also explaining why we are so often tempted
into violating it.17
Notes
1. This point has been made before. See Huemer 2005, §5.1; Hooker 2002; Audi
2015, 76.
2. McDermott (2008, 12), Gaut (1994, 33).
3. One might suggest that one could arrive at one’s initial set of moral beliefs by
accepting someone’s testimony. This is true but of course just pushes back the
question, forcing us to ask how the testimony provider arrived at her initial set
of moral beliefs. If we follow the chain backward far enough, we will arrive at
someone trusting a moral intuition.
4. A helpful analogy, suggested to me by an anonymous referee, would be beliefs
about which things are poisonous. True, we don’t have basic tendencies to
believe that certain things are poisonous and others aren’t, but this doesn’t
mean that natural selection has exerted no influence on such beliefs.
5. Both Broad and Ross used this expression.
I say that this ‘possibly’ was Ross’s view since another reasonable interpretation
of what he’s saying on the cited page is that, as a matter of anthropology,
judgments about conditionals originally were arrived at by intuitive induction
from verdictive judgments. This wouldn’t commit Ross to any particular fact
about how you and I, today, arrive at our judgments about conditionals.
An Explanatory Project 15
6. One might respond that things look differently in cases in which I check whether
all As are Bs only to find that a few of them, intuitively, are not. In such a case
it would be reasonable for me to feel torn, and it might be posited that the only
way to account for this is by supposing that my intuition that all As are Bs
remains a reason to believe ‘all As are Bs’. By way of rejoinder, I would insist
that we have another way of accounting for the reason I have to believe ‘all As
are Bs’, namely all those cases in which my verdictive intuition, with respect to
some A, is that it is indeed a B and all the cases in which my verdictive intuition,
with respect to some non-B, is that it is indeed a non-A plus the recognition that
my intuitions regarding those As that appear to not be Bs and those non-Bs that
appear to be non-As could be mistaken.
7. This analogy between observations in science and verdictive intuitions in moral-
ity is endorsed by List and Valentini (2016) and McDermott (2008, 14).
8. The classic criticisms are to be found in van Fraassen (1980; 1989).
9. The argument that I make in this sub-section is based on an argument Ladyman
(2012, 42–6) makes against appealing to the virtue of power as a way of choos-
ing between competing a priori metaphysical theories.
10. One might think that in order to test a normative ethical theory we have to
engage in a certain amount of artificial dividing. Suppose there’s some case, X,
about which we are confident in our verdictive intuitions. We might build up a
normative ethical theory while setting aside the question of what to say about
X, and then use X as a test case to determine whether we’ve built a good theory.
One might worry that if we don’t set aside some cases in this manner then it’s
difficult to imagine how else we might test our theory.
The mistake in this line of reasoning is its distinguishing between the theory-
building stage and the theory-testing stage. Because of the temporality of sci-
ence (i.e., the fact that at different points in time we have different observations
with which to work) there is no choice but to make this distinction in scientific
theorizing. In moral theorizing, on the other hand, there need not be any distinc-
tion; every intuition in which we are confident can be both a theory-building
intuition and a theory-testing intuition. Surely it makes the most sense to use
them this way. If we are confident in an intuition, then why would we want to
tie one hand behind our backs by refusing to take it into account when building
our theory? And when we reflect on what we have accomplished thus far in our
theorizing project, obviously we should want to check whether our theory does
in fact manage to accommodate that intuition.
One might respond that although moral theorizing isn’t temporal in principle,
it’s temporal in practice since some of our intuitions don’t occur to us until after
we’ve constructed a candidate theory. By way of rejoinder I say fair enough, but
this doesn’t undermine the point I am making that there is never any need to
reason about some moral question in the absence of the relevantly authorita-
tive source of knowledge about it. Although I may never have entertained the
question of what the right thing to do in situation X is, once I do entertain that
question I do so while in possession of the relevantly authoritative source of
knowledge about it if there is one.
11. DeGrazia endorses the use, in ethical theorizing, of appeals to coherence (or
what DeGrazia calls “compatibility”) and simplicity (1996, 16–17). Hooker
(2000, 19–23), likewise, advocates the simplicity criterion. Nagel (1979, x), on
the other hand, rejects appeals to simplicity and elegance.
12. Huemer (2009, 218) makes note of and formalizes this kind of argument for
appealing to simplicity (parsimony) in science, though he goes on to construct
three other arguments for the same conclusion (219–25).
13. There is a second extant argument for appealing to simplicity in scientific
theorizing, namely Roger White’s (2005) argument. However, his argument is
16 An Explanatory Project
clearly inapplicable to moral theorizing as it relies on an asymmetry that holds
in the scientific case but not, as I will show in §6, in the moral case: that the
specific facts are verifiable in a way that the theoretical facts are not.
14. As Daniels says, with respect to the set of more particular judgments and the set
of more general judgments over which one reasons in the method of reflective
equilibrium, “There is no set of judgments that is held more or less fixed” (1979,
266).
15. And in that case, one might as well not bother using one’s intuitions about those
categoricals/conditionals as inputs to reflective equilibrium. This thought may
be what G. A. Cohen (2008, 4) was thinking when he said that there is just one
set of inputs to the process—one’s verdictive intuitions.
16. Ronald Dworkin, for instance, assumes that if the method of reflective equilib-
rium is understood as a way of discovering moral facts (what Dworkin calls the
‘natural model’ of reflective equilibrium), then it will be construed as analogous
to scientific inquiry and will include a bias toward the verdictive (1989, 27–9).
17. Here’s a second interesting corollary to the argument of this chapter:
Joshua Greene (2013) has argued, famously, that empirical studies show that
characteristically act-consequentialist moral judgments—judgments based on a
concern to promote the overall good—are driven by controlled cognitive pro-
cesses whereas characteristically deontological moral judgments are driven by
automatic emotional responses, and has gone on to argue, further, that con-
trolled cognitive processes are more likely to guide us to the truth than automatic
emotional responses in many of the cases over which act-consequentialists and
deontologists disagree and that therefore characteristically act-consequentialist
moral judgment is more likely to yield the moral truth of the matter in those
crucial cases than is characteristically deontological moral judgment. (Peter
Singer (2005) endorses this line of argument.)
Given the argument of this chapter, however, it’s not clear that Greene is right
about what’s characteristic of act-consequentialist and deontological moral judg-
ment. I’ve argued that we should choose between competing normative ethical
theories based on their respective explanatory claims, so there’s a sense in which
what’s characteristic of a normative ethical theory is its explanatory claims. By
contrast, if all one is asked to do is what Greene asked the subjects of those
empirical studies to do, namely to reach a verdictive judgment, it is enough (and
in fact most convenient) to appeal to moral conditionals or categorical claims.
(Appealing to explanatory claims is useful only when theorizing.) So it may be
fair to say that the kind of judgment that the subjects of Greene’s studies engaged
in was not characteristic of any normative ethical theory.
(Still, however, Greene could claim that the moral conditionals/categorical
claims instructing us to promote the overall good typically appear in act-
consequentialist theories. This is true, but we should not overstate its impor-
tance. As I’ve argued elsewhere (Sachs 2010a) moral categoricals/conditionals
that are strongly associated with act-consequentialism can be accommodated
within non-act-consequentialism, and vice-versa.).
2 How Should We Choose Between
Competing Explanatory Stories?
1. Introduction
I argued in the previous chapter that moral theorizing makes sense only if it’s
construed as an explanatory project, which means we should judge candi-
date normative ethical theories in part on how good of an explanatory story
they tell, and make use of our explanatory intuitions in so judging. This
chapter tackles the issue of exactly how to do that. I expect most or all of
what I say here to be obvious, perhaps to the point of being boring, but it
constitutes necessary groundwork for the rest of the book, in which I show
that if we follow the strictures laid out in this chapter—that is, if we actu-
ally approach moral theorizing as an explanatory project—this will have
surprising, radical, and important implications for the conclusions we will
reach in normative and applied ethics.
In §2 I get specific about what kind of explanatory claims normative eth-
ical theories should be construed as containing. In §3 I discuss how we
should deal with the apparent existence of counterexamples to all seemingly
plausible moral generalizations, having agreed that those generalizations
should be construed as explanatory claims. In §4, I offer criteria for decid-
ing which explanatory claims to accept (and thus criteria for deciding which
normative ethical theory to accept). Finally, §5 concludes.
2.5 Conclusion
We began this section already having reached the conclusion that the gen-
eralizations contained in normative ethical theories should be represented
How Should We Choose Between Competing Explanatory Stories? 21
as explanatory claims for the purpose of sifting through candidate theories.
The goal for this section was to identify as which kind of explanation they
should be construed. I began by arguing that they should be construed as
generalizations about atomistic explanations. The goal then became to illu-
minate the nature of atomistic explanation. I advocated a grounding model
of atomistic explanation and stipulated three features that an atomistic
explanation must contain.
5. Conclusion
In this chapter I argued for various claims about how we should determine
which grounding explanations and generalizations about them to accept.
I want to show that if we take these lessons seriously when we engage in
moral theorizing, this will have implications for which conclusions we reach.
In the rest of this book I illustrate this point by doing some moral theorizing,
using the method of reflective equilibrium.
How Should We Choose Between Competing Explanatory Stories? 29
I will come to some very surprising conclusions—conclusions that we
don’t see other moral philosophers arriving at despite their using the same
method. I want to be clear, straight off, as to why this is going to happen. It
will happen because, in accordance with the argument I gave in Chapter 1,
the general judgments that I will use as inputs to the method of reflective
equilibrium will be explanatory judgments (which I argued in §2 of this
chapter are judgments about grounding relationships) as opposed to moral
categoricals or conditionals, which is what most moral philosophers use. On
the other hand, my arrival at very non-standard conclusions is not going to
be the result of my initial set of explanatory judgments being idiosyncratic.
My view is that we theorists should be using the method of reflective
equilibrium collectively as opposed to independently. Here I think the anal-
ogy with science is apt. Scientific theorizing is treated as a collective effort;
this is because we don’t think any scientist has special access to the scien-
tific facts, and so we consider it legitimate for the scientific community at
large to sit in judgment of the claims that each individual scientist makes.
Now although moral epistemology is probably less well understood than
scientific epistemology, I think it’s pretty safe to say that nobody has special
access to the moral facts (Driver 2006; Sliwa 2012; McGrath 2008, 2011);16
though there may be certain individuals, such as children and some mentally
ill people, who we can safely say have inferior access. So, analogously, I con-
sider it legitimate for the moral community at large to sit in judgment of the
claims that each individual moral theorist makes. Given this, it makes more
sense for us all to work together from the start, as scientists generally do,
than for each of us to work in isolation and then engage in critical scrutiny
of each other’s theories only after they’ve emerged fully formed. And part of
working together from the start, in the moral case, is pooling our intuitions:
We should be using the method of reflective equilibrium as an effort to build
a view that does as good a job as possible of accommodating the intuitions
that are widely held. Consequently, the explanatory intuitions from which I
start my arguments in subsequent chapters will be ones that, to my under-
standing, are widely held.17
Notes
1. It is important to see that the project of holistic explanation does not reduce to
the project of atomistic explanation. In this case an analogy with science should
be helpful. An atomistic explanatory project in science would be to explain, for
each case of a window shattering, why it did. But one also might try to explain
the patterns and regularities we observe over the domain of window-shattering
and non-shattering events; this would be a holistic explanatory project. Just as
identifying the cause of each window-shattering does not amount to explaining
why window-shatterings distribute the way they do, identifying (for instance)
the explanation of the wrongness of each case of innocent child-harming does
not amount to explaining why verdictive facts about child-harming actions dis-
tribute the way they do. Likewise, atomistic explanation does not reduce to
30 How Should We Choose Between Competing Explanatory Stories?
holistic explanation. Again, the scientific analogy is apt. I might know all the
true scientific generalizations, but knowing why some particular thing happened
involves (at least) knowing which generalization it falls under. In the moral case
I might have the correct set of explanatory generalizations, but knowing why
some particular action is (say) impermissible involves (at least) knowing which
member of the set is applicable in the given case.
2. Since the only atomistic explanatory claims in which I’m interested are the ones
that offer a non-normative explanans, I am correspondingly interested only in
the generalizations that generalize about that kind of atomistic explanation and,
likewise, interested only in normative ethical theories composed of generaliza-
tions of that sort. Thus, while I opened this chapter saying that I wanted to
establish how we should sort between candidate normative ethical theories, I
should now say more specifically that I want to establish how we should sort
between candidate normative ethical theories that purport to take us from the
non-normative to the verdictive.
Fortunately, there are plenty of normative ethical theories that purport to do
that; in fact the ones that don’t are in the minority. For example, Benthamite
utilitarianism relates non-normative facts to verdictive facts directly through its
generalization that the fact that an action maximizes utility explains its being
morally permissible (and obligatory). Rossian pluralism, meanwhile, relates
non-normative facts to verdictive facts indirectly through its generalizations
regarding certain action-types constituting pro tanto duties and its further gen-
eralization to the effect that some action’s being supported by the balance of
pro tanto duties explains that action’s obligatoriness. (For ease of discussion I
will use theories of the Benthamite sort—theories that make their connection
directly—as my case studies for the rest of this chapter, but everything I say
about them will hold as well for theories of the other sort.)
3. Having said that, I can now elucidate my earlier claim that holistic explanations
are composed of generalizations about atomistic explanation. In such a general-
ization what is asserted is that facts of a certain kind bear atomistic explanatory
relationships to facts of another kind (e.g., in general the fact that an action is
a killing explains the moral impermissibility of that action). I will give a more
precise template for such claims at the beginning of §4.
4. The candidates would be constitutive explanation, causal explanation, and nomic
explanation.
Constitutive explanation is a relationship between properties or between objects,
and so fails to meet the second criterion.
And we know, independently of anything I’ve said here, that in the moral
domain we’re not looking for causal explanations; we know this because if X
causally explains Y then there must be a time lag between X and Y, and in the
moral domain there is no time lag between an action’s being (say) impermissible
and its having whatever feature explains its impermissibility. (Thanks to Kath-
erine Hawley for pointing this out to me.)
As to nomic explanation, here’s the disanalogy: In moral explanation, the
explanans and explanandum are facts about the same thing. So, for instance,
the harmfulness of spreading rumors about someone grounds the wrongness of
spreading rumors about someone. This ‘same-thing’ constraint can be added to
the second feature of atomistic explanations mentioned in the main text. With this
feature added in, nomic explanations no longer have the second feature. The law
that explains why the water froze has as its explanandum a feature of the water
(that it froze) and as part of its explanans a feature of the air (its temperature).
5. A second problem, which I will not discuss here, is whether my claim that accept-
ing a normative ethical theory means accepting the existence of grounding rela-
tionships between non-normative facts and verdictive facts could possibly be
How Should We Choose Between Competing Explanatory Stories? 31
acceptable to a moral anti-realist. Suffice it to say I do not intend to rule out
anti-realism and that if there are quasi-realist positions that can allow for talk
of moral facts and the non-normative facts that ground them, then I’d be happy
to have such quasi-realists on my side.
6. Stratton-Lake (2011, 367) believes this is not problematic at all, but I’ll concede
for the sake of argument that it is.
7. For more on how partial grounding might work, see Leary (2017).
8. Raz (2000), Crisp (2000), Strandberg (2008), and Fogal (2016) each argue for
the validity of this worry. For responses, see Schroeder (2007, Chapter 2) and
Scanlon (2014, Chapter 5), both of whom accept the distinction; see also Bader
(2016).
9. To me this seems obvious, and I’ve never seen it disputed. Maguire (2015, 196,
203) assumes it’s true.
10. To my knowledge, this has never before been pointed out in the literature, obvi-
ous though it is.
11. Kagan (1988, 14–15); Parfit (2011, 415). Something close to this claim is made
by Hooker (2002). He says we have intuitions about relevance; that is, about
what kinds of fact about actions count morally in favor of or against our doing
them.
12. Thanks to an anonymous editor for alerting me to this objection.
13. Take W. D. Ross’s theory as an example. He held (or we can construe him as
having held), that five kinds of non-normative facts about an action can make
that action obligatory: facts about fidelity, righting previous wrongs, gratitude,
beneficence, and non-maleficence. For each such fact we can appeal to our intu-
ition as to whether it really can do that; one can generate an intuition about
whether an action can be obligatory in virtue of being what gratitude requires.
But I don’t think one would be able to generate an intuition about whether all
obligatoriness facts are accounted for by this five-item list.
14. For the sake of simplicity, my discussion of scientific explanation is meant to be
limited to the natural sciences, hence allusion above to the ‘natural domain’.
15. Mark Schroeder says this about reasons (Schroeder 2007, 2). See also Sigd-
wick’s talk of the philosopher seeking “unity of principle” (Sidgwick 1981, 6).
In addition, Rawls (1999a, 144) seems to endorse a simplicity criterion, as do
List and Valentini (2016).
16. Or, more carefully, we have no way of knowing of any particular individual that
she does.
17 Specifically, widely held among the people with which whose intuitions I am
familiar. Unfortunately, this is a tiny sliver of humanity.
3 Against Monism
1. Introduction
In this chapter and the next two I investigate the plausibility of various nor-
mative ethical theories. For organizational purposes I divide the discussion
in two: this chapter deals with monistic theories, while the next two concern
pluralistic theories.
Specifically, in this chapter I discuss consequentialism, Kantianism, and
contractarianism/contractualism, all of which are generally categorized as
instances of moral monism on account of the fact that they seem to be most
naturally representable as containing just one moral generalization. Since I
argued in Chapter 2 that we should represent the generalizations contained
in normative ethical theories as type-grounding claims connecting non-
normative facts to verdictive facts, I will represent each of these theories as
containing nothing more than one such type-grounding claim.
To depict act-consequentialism this way we have to classify evaluative facts as
non-normative facts. Only by doing this can we represent act-consequentialism
as being committed to the existence of just one type-grounding relationship
between verdictive facts and non-normative facts—namely the relationship
such that all the verdictive facts are grounded in facts about maximizing the
good.
I go in for this classification on the principle of charity, because all else
being equal, moral monism is superior to moral pluralism. In §4 of Chapter 2
I claimed that all else being equal a normative ethical theory is better to the
extent that it is simpler, and to that end the fewer non-normative facts one’s
theory countenances as grounding the various verdictive facts, the simpler it
is. Consequently when it comes to the simplicity criterion, consequentialism,
since it is a version of moral monism, can’t be beat.
But I will argue in this chapter that all else is not equal. If in order to be
monists we have to accept a type-grounding claim that is itself counterin-
tuitive or yields counterintuitive token-grounding claims or fails to deliver a
great many of the verdicts we find intuitively appealing, then we should con-
sider not being monists. This chapter will focus exclusively on the criterion
of token-grounding claims. The question will be whether the three candidate
versions of moral monism can deliver a plausible set of token-grounding
Against Monism 33
claims. I will begin by arguing that not one of them can do this. From these
three examples I will generalize to an overall conclusion to the effect that the
prospects are dim for discovering a version of moral monism that offers good
token-grounding explanations. This is an important critique, because each
normative ethical theory is nothing more than a set of explanatory claims.
3. Consequentialism
There is an objection to consequentialism that’s been floating about for a
while: that sometimes consequentialism gets the right answer for the wrong
reason. The idea is that in some of the cases in which consequentialism
delivers the correct verdict about a case, its moral explanation—that is, its
explanation of why that verdict holds—is defective. The objection was origi-
nally directed specifically against utilitarianism; Bernard Williams’s (1973,
117) critique of how utilitarianism handles the Jim and the Indians case
might be the first instance of it, depending on how that critique is to be
understood. And it is often said (e.g., Arneson 2000a, 239; Hinman 2008,
227) that Rawls criticized utilitarianism this way. In any event, there are
several casual mentions of this objection, some of them directed narrowly
at utilitarianism and others more broadly at consequentialism, to be found
in the literature.3 On the other hand, there are just two philosophers, as
far as I can tell, who make the objection central to their case for rejecting
consequentialism.4
I aim to rectify this situation by laying out and defending a detailed ver-
sion of the explanatory objection to consequentialism. §3.1 describes what I
take to be the moral explanation template that consequentialism must offer.
In §3.2 I lay out the first objection to the token-grounding explanations
that follow from this template, which is that they are incompatible with the
directedness of wrongdoing—the idea that sometimes we wrong each other.
Then §3.3 contains my second objection to these explanations, which is
simply that they are counterintuitive.
a) the goodness of all the consequences of that action is less that the good-
ness of all the consequences of some alternative action, or
b) that action would be prohibited by the set of rules such that the good-
ness of all the consequences of universal acceptance or adherence to
that set of rules is better than the goodness of all the consequences of
universal acceptance or adherence to any other set of rules.
Wallace and Darwall each argue that consequentialism cannot account for
the just-mentioned kinds of normativity and reactive attitudes. I believe that
this is true, and that correspondingly it is true that consequentialism cannot
account for wronging.6
Identifying the close links between the concept of wronging, the concept
of obligation, and the appropriateness of various reactive attitudes helps to
reinforce how important it is that a normative ethical theory allow that some
Against Monism 37
instances of wrongdoing are instances of wronging. A failure to do this is
also a failure to provide a place in one’s normative ethical theory for these
closely connected notions. Surely we won’t be satisfied with a theory that, for
instance, cannot make sense of one individual being accountable to another
or of one individual being appropriately resentful at the conduct of another.7
My criticism, that consequentialism cannot account for wronging, is based
on the observation that the token-grounding claims implied by CTGC do
not identify a wrong-sufferer. A sample token-grounding claim derived from
CTGC would be (for act-consequentialism) ‘Jones’s φ-ing was impermissible
because the goodness of all its consequences was less than the goodness of
all the consequences of some other action available to her’ or (for rule-conse-
quentialism) ‘Jones’s φ-ing was impermissible because it would be prohibited
by the set of rules such that the goodness of all the consequences of universal
acceptance or adherence to that set of rules is better than the goodness of all
the consequences of universal acceptance or adherence to any other set of
rules.’ Since wronging is a relationship between a wrongdoer and a wrong-
sufferer, those token-grounding claims do not account for wronging.
Not only does this fact about CTGC constitute a strong reason to reject
consequentialism, it also undermines what has typically been perceived as a
strong reason in favor of consequentialism—namely, that it delivers the right
result in the Peter Singer’s Pond Case and others like it in which one can,
at little cost to oneself, prevent a great tragedy befalling another individual.
The usual intuitive response to the Pond Case isn’t simply that one is mor-
ally obligated to save the child, but also that one would wrong the child
by failing to do so. But consequentialism cannot allow for this, because its
token-grounding explanation of the wrongness of failing to save the child
in the Pond Case will simply point out an impersonal fact about the case.
3.2.6 Conclusion
In this section I have objected to consequentialism on the grounds that
CTGC cannot account for the directedness of some instances of wrongdoing.
I began by situating the idea of directed wrongdoing among other familiar
moral notions, each of which requires that in cases of directed wrongdoing
there be a wrong-sufferer. I argued, next, that the token-grounding claims
implied by CTGC never identify a wrong-sufferer, and that for that reason
cannot explain instances of directed wrongdoing.
3.4 Conclusion
I have argued in this section that consequentialism is vulnerable to two
objections: First, its type-grounding claim, CTGC, yields token-grounding
claims that fail to allow that some wrongdoing is directed. Second, many of
the token-grounding explanations that CTGC yields are deeply counterin-
tuitive. Along the way I discussed various minor modifications that could
be made to CTGC that might be thought to help its case. None of them,
however, did the trick.
I do not conclude from all this that CTGC is defective. All I have shown
is that CTGC is not up to certain explanatory tasks. If CTGC were merely
one type-grounding claim in a pluralistic moral theory, then maybe the other
type-grounding claims in that theory could be invoked to do the explana-
tory work CTGC can’t do. I’ll defer until Chapter 8 a discussion of whether
we should accept as part of a pluralistic theory the two ideas into which
CTGC can be decomposed: the idea that morality requires us not to harm
others and the idea that morality requires us to confer benefit on others.
While I don’t (yet) conclude that we should reject CTGC, I do conclude
that we should reject consequentialism, since consequentialism is composed
of nothing but CTGC and CTGC’s token-grounding claims don’t do a good
job of explaining many of the verdictive facts.
But perhaps I have adopted too narrow a conception of consequential-
ism. It is sometimes said that, for consequentialists, facts about goodness
do not ground the verdictive facts; rather facts about reasons ground the
verdictive facts. If a theory of this sort included a distinctly consequentialist
Against Monism 45
way of understanding of reasons, then it might deserve the ‘consequential-
ist’ label. I’m thinking, for instance, of a theory on which all moral rea-
sons are grounded in the good and the strength of our overall moral reason
to do something is proportional to the strength of the good in which it is
grounded, or, alternatively, of a theory on which for something to be good
just is for it to constitute or ground a reason of a certain sort.17
A theory of this sort would be a version of ‘Rossian Pluralism’, a family
of ethical theories that I discuss in the next chapter. So although I have no
interest in insisting that such a theory would not also qualify as a version of
consequentialism,18 I will not discuss it here.
Act only according to that maxim through which you can at the same
time will that it become a universal law.
(Groundwork 4:421)19
So act that you use humanity, in your own person as well as in the per-
son of any another, always at the same time as an end and never merely
as a means.
(Groundwork 4:429)
If we turn these imperative sentences into declarative ones and bring for-
ward their implicit claims about moral grounding, we arrive at:
The fact that one acted on a maxim that one could not have willed
to become a universal law grounds the moral impermissibility of one’s
resulting action.
The fact that one’s action failed to treat one’s humanity or that of
another as an end in itself or treated it as a mere means grounds the
moral impermissibility of that action.
There are many ways of objecting to this way of representing Kantian eth-
ics. The most important objection, for our purposes, would be the idea
that it is a mistake to portray the Categorical Imperative as the entirety of
Kant’s normative ethical theory. Kant does, after all, put forward a Doc-
trine of Right and a Doctrine of Virtue in the Metaphysics of Morals, from
46 Against Monism
which he derives various rights and duties, and according to some interpre-
tations of Kant, these doctrines are supposed to have independent moral
force (specifically, their moral force is not supposed to stem from their being
derivable from the Categorical Imperative).20
I have no argument against interpreting Kantian ethics this way. However,
in this chapter I am interested in examining the prospects for moral monism,
and since it is possible to represent Kantian ethics as a version of moral
monism and many moral philosophers do think of it that way, it is worth
our trouble to investigate the viability of Kantian ethics so conceived. More-
over, even if Kantian ethics is not monistic in letter, it is monistic in spirit.
All of the elements of Kantian ethics, including the Categorical Imperative,
the Doctrine of Right and the Doctrine of Virtue, are derived from the same
basic idea: that we should always act in a way that is consistent with respect
for humanity.
With regard to the animate but nonrational part of creation, violent and
cruel treatment of animals is far more intimately opposed to a human
being’s duty to himself, and he has a duty to refrain from this; for it
dulls his shared feeling of their suffering and so weakens and gradually
uproots a natural predisposition that is very serviceable to morality in
one’s relation with other men . . . Even gratitude for the long service
of an old horse or dog . . . belongs indirectly to a human being’s duty,
namely, his duty with regard to those animals; considered as a direct
duty, however, it is always only a duty of the human being to himself.
(6:443, emphasis original)21
Against Monism 47
Because animals are not autonomous—that is, they lack humanity—they
do not fall under the Categorical Imperative and thus killing them cannot
constitute a violation of any duty to them.22 Kant would have to say this not
just about shooting a dog, but also about the most unimaginable forms of
cruelty, as long as that cruelty is inflicted on an animal.
The problem here is not that the Categorical Imperative, as a type-grounding
claim, is inconsistent with certain verdicts we are inclined to accept. Kant
shows us how to get the right verdicts, at least some of the time. The prob-
lem, rather, is in the token-grounding explanations of those verdicts the
Kantian type-grounding claim implies. What grounds the wrongness of a
given act of cruelty to an animal, apparently, is how it is damaging to one’s
own humanity (O’Neill 1998). So such an act wouldn’t wrong the animal.
This is terribly counterintuitive.23,24
4.5 Conclusion
The argument I have presented in this section is not intended as a dismissal
of Kantian ethics. I have said nothing against the idea that we can derive
moral constraints on our behavior from the idea of respect for humanity.
Rather, I object to using this idea as a starting point when trying to explain
why it’s impermissible to treat animals cruelly. All I object to, then, is monis-
tic Kantian ethics. For all I’ve said the Categorical Imperative might be sal-
vaged as one part of a pluralistic normative ethical theory that has some
other element that offers a different explanation of the wrongness of cruelty
to animals.28
Against Monism 53
5. Contractarianism/Contractualism and Cruelty to Animals
Notes
1. In conversation with act-consequentialists I’ve found that they’re sometimes
inclined to reinterpret their theory along these lines when confronted with the
explanatory objection to their theory that I raise later in this chapter.
2. This interpretation of Kantianism and the subsequent interpretation of
contractarianism/contractualism are suggested by Väyrynen (2009, 101–2).
3. Shaw 2006, 215; Stewart 2009, 63; Schroeder 2007, 38–9; Wringe 2003, 85;
Lichtenberg 1983, 546; Ross 1988, 37–9. Lichtenberg and Ross actually con-
flate this objection with another objection, which is that consequentialism gets
the wrong verdict in some cases. The objections certainly are closely related, I
admit. But as I’ve argued elsewhere (Sachs 2010a), there could be a brand of
consequentialism that is vulnerable to the explanatory objection but gets all the
verdicts right.
58 Against Monism
Stratton-Lake (2011) raises something like the explanatory objection to con-
sequentialism, though he seems to mostly have rule-consequentialism in mind.
Interestingly, he claims that Rossian pluralism escapes this objection—a claim
I’ll deny in Chapter 4.
4. Kymlicka (2002, Chapter 2, esp. pp. 24–5 and 28–32) (see also Kymlicka 1989,
28–9); Stratton-Lake (2011).
Nick Zangwill (2011) offers something like this objection, though it is directed
exclusively at indirect consequentialism and it is not general in the sense of sug-
gesting that indirect consequentialism generally gets the right answer for the
wrong reason; rather his worry is that it does so in a particular kind of case.
Also, this criticism itself is indirect in that for Zangwill it does not on its own
count against indirect consequentialism, but rather does so in conjunction with
a theory of permissible partiality and morally acceptable deliberation.
5. For the sake of simplicity I set aside other versions of indirect consequentialism,
such as motive consequentialism. But the arguments I make here against act-
and rule-consequentialism could be used, mutatis mutandis, against the other
versions of indirect consequentialism as well.
6. Mark Nelson (2015) makes this argument specifically against utilitarianism.
7. I was prompted to point this out after attending a conference in which Samuel
Scheffler suggested that a consequentialist might take it as a badge of honor
that her theory cannot allow for wronging. (Note: Scheffler was not endors-
ing this line of thought.) A consequentialist might say, he pointed out, that the
idea of wronging is an outdated vestige of morality’s early roots as a system for
maintaining peace and order in small communities. Now that the most press-
ing moral problems seem to be more global and impersonal, a consequentialist
might say we have no need for such an interpersonal notion as wronging.
8. Chappell (2015).
9. Whether this is true is controversial. I will grant for the sake of argument that
it is.
10. The position I take in this paragraph follows Strawson (1962), Darwall (2006,
66–7) and Wallace (2007, 30–1).
11. Darwall (2006, 310–13). This is an aspect of a more general point Darwall makes,
which is that there is no in-principle reason why agent-neutral principles of conduct
couldn’t be derived from an agent-relative theory of moral obligation (such as Kan-
tianism or contractualism). See Darwall (2006, Chapter 1 n.16, Chapter 4 n.32).
12. Parfit (2011, 417–18) is explicit about this.
13. Darwall (2006, 15–17) makes the same point a broader context: He says that
to propose that the social utility of someone’s having a reactive attitude justi-
fies their having that attitudes is to make a “wrong kind of reason” error with
respect to reasons for attitudes.
14. Thanks to Claire Benn and Fiona Woollard for mentioning this possibility and
aiding me in thinking through it. I later discovered that a quite sophisticated
version of this view has been defended in print. See Roberts (2002).
15. Thanks to Pekka Väyrynen for pushing me to address this version of act-
consequentialism. That an act-consequentialist might make this move is noted
in Kymlicka 2002, Chapter 2. The best contemporary example of an act-
consequentialist who at least might believe this is the utilitarian Peter Singer, who
at times describes the principle of utility as falling out of the more basic moral
fact that everyone’s interests count equally. See, for instance, Singer (1993, Chap-
ter 1). Alternatively, it is open to us to say that the injunction to weigh everyone’s
(equally strong) interests equally just is the principle of utility (Hare 1984, 107)
or just is the act-consequentialist moral principle (Scheffler 1982).
16. The consequentialist could respond at this point by saying that correctly
accounting for someone’s personal good is a comparative matter—a matter of
Against Monism 59
assigning the correct relative weights to the personal good of the various indi-
viduals. But this can’t be right. When more than two people are affected by
some action, mere comparative facts to the effect that A’s personal good should
be weighted this much relative to B’s and B’s personal good should be weighted
that much relative to C’s, and so on, may leave it indeterminate what the right
thing to do is. There needs to be some way of calibrating the various relative
weightings so as to yield an overall fact as to which course of action will pro-
duce the most (properly weighted) personal good. But if there is an overall cali-
bration that covers everyone, then for any wrongful action involving more than
two individuals it’s going to be indeterminate which particular miscalibration
was involved, and thus indeterminate who was wronged.
17. See, for instance, Portmore (2011).
18. Urmson (1974, 114–15) does so insist.
19. The translation I’m relying on for this quotation and the next is Kant (2011).
20. Pogge (1997), for instance, argues that we can arrive at Kant’s universal prin-
ciple of justice, part of his Doctrine of Right, without starting at the Categorical
Imperative. But O’Neill (2012) contends that the principle is indeed supposed
to follow from the Categorical Imperative.
One might suggest, as a less extreme departure from classic monistic Kantian-
ism, that the various duties that Kant derives from the Categorical Imperative
are supposed to be grounds of verdictive facts and that the Categorical Impera-
tive is supposed to explain why those duties have their grounding power when
they do. But this is the alternative understanding of Kantianism I discuss and
discard in Appendix A.
21. The translation that I’m relying on here is Kant (1996).
22. Jens Timmermann helpfully distinguishes the question of whether Kantian eth-
ics allows for a duty to not do X to an animal from the question of whether
Kant’s ethics allows for a duty to an animal to not do X to it. Timmermann
(2005) argues for an affirmative answer to the former question. If his argument
works, it means Kantian ethics is consistent with the judgment that certain ways
of treating animals are in themselves to be avoided. In other words, Kantian
ethics wouldn’t have to fall back on the claim Kant himself makes that avoid-
ing treating animals in certain ways is a means to the end of avoiding treat-
ing human beings immorally. However, as Timmermann admits, his argument
doesn’t get us to the conclusion that Kantian ethics is consistent with the judg-
ment that certain ways of treating animals wrong them (or, in Kantian terms,
violate a duty to them). What I want to explore in this section is whether there
is some route to that conclusion.
23. Kant admits this and attempts to explain away the intuition. See his comments
on amphiboly in The Metaphysics of Morals 6:442.
24. One worry I have about Kantian ethics is that this problem may be more wide-
spread than generally recognized. Traditionally, the problem is supposed to be
that what appear to be violations of duties to animals are construed as violations
of a duty to oneself. So Kantian ethics appears a bit too self-regarding. My con-
cern is that Kantian ethics may be entirely self-regarding. In the Groundwork the
Categorical Imperative is derived from the idea of rationality. Kant says that act-
ing rationally means giving laws to oneself, and from this initial premise he goes
on to argue that rationality requires willing in accordance with the Categorical
Imperative. Thus the Categorical Imperative emerges as a self-imposed constraint
(see Korsgaard 2012, 18). Another aspect of Kantian ethics that should make us
suspicious is its insistence that it is only maxims, and not actions, that are to be
subject to the Categorical Imperative. One’s maxim is the principle on which one
decides what to do. So it appears that Kantian ethics, at least in the first instance,
judges our practical reasoning and not our actions (Herman 2016, 34).
60 Against Monism
Looked at this way, Kantian ethics seems to be a lot like virtue ethics, where
the basic idea is to offer a standard for being a good person and only deriva-
tively offer a standard for treating others well (or not offer one at all). (This
affinity between Kantian ethics and virtue ethics is noted by Baron (1997). The
version of Kantianism that seems most aptly described this way is the most
recent version of Korsgaard’s Kantianism (Korsgaard 2009), on which moral
constraints are a byproduct of one’s effort to constitute oneself as an agent.)
Whatever the merits of virtue ethics, a central weakness of that approach to
ethics is that, unless it is merely one element of a pluralistic normative ethical
theory, it isn’t consistent with the idea that some of the things we do wrong
others. According to Rosalind Hursthouse, virtue ethics is the view that wrong
actions are wrong because they instantiate a vice: “[W]hat is wrong with lying,
when it is wrong, is . . . that it is dishonest, and dishonesty is a vice. What is
wrong with killing, when it is wrong . . . [is] that it is callous and contrary to the
virtue of charity” (1999, 6, original emphasis removed). That an action instanti-
ates a vice is a fact about the agent, if it is a fact about anyone. It is certainly
not a fact about the individual to which the action was done. So if we cite the
viciousness of an action as explaining its wrongness, nothing we say picks out
anyone (except perhaps the agent) as the wrong-sufferer.
25. In an interesting footnote, Korsgaard offers the following analogy:
imagine a white male who claims that in valuing his own freedom he is
only valuing the freedom of white males: if, unknown to himself, he turned
out to be a black woman . . . then he would agree that his freedom doesn’t
matter. Our response would be that either he’s insincere or deceiving him-
self, that he’s suffering from a failure of reflective imagination.
(2005, 104)
Moral monism is the view that there is just one kind of non-normative fact
that grounds all the verdictive moral facts. Moral pluralism, by contrast,
holds that there are multiple kinds. In the previous chapter I criticized
monism on the grounds that it seems unlikely to be able to offer a compel-
ling picture of how the verdictive facts are grounded. In this chapter my
main goal is to establish that the most popular version of pluralism is vul-
nerable to the same objection.
The most popular version of pluralism is the kind on which the verdic-
tive facts are directly grounded by other normative facts. Of course, like
nearly everyone, the people who accept this view also believe some version
of the claim that normative facts supervene on non-normative facts. There-
fore, assuming that the verdict-grounding normative facts don’t reduce to
non-normative facts, they must claim in addition that the verdict-grounding
normative facts are grounded in non-normative facts. What makes this
view pluralist is its further contention that there are different kinds of non-
normative fact that can ground the verdict-grounding normative facts.
Different versions of this view can be distinguished according to their
divergent views about which normative facts ground the verdictive facts.
One version is the view that the verdictive fact about an action is grounded
the moral reasons for and against performing it. This is the version on which
I will focus here. It is by far the most popular version of the view and mostly
likely other versions of the view can be made compatible with it.
I am going to reject this view on substantive grounds. This may seem odd,
as it seems to be a purely meta-ethical view. But this appearance is based on
a misunderstanding of normative ethics, which Chapters 1–2 were intended
to dissolve. I argued there that when theorizing we should understand nor-
mative ethical theories as offering competing views regarding which facts
ground the verdictive facts. Now the view I’ve been picking out doesn’t say
which facts ground the verdictive facts, and so I admit of course that it isn’t
a normative ethical theory. The view I’ve been picking out, however, does
purport to identify the type of facts that ground the verdictive facts, and
therefore marks out a family of normative ethical theories. In this chapter
I reject this family on the grounds that each member of it will exhibit a
62 Against Rossian Pluralism
serious shortcoming. I’ll label this family of views Rossian Pluralism (RP)
since Ross’s (1988) normative ethical theory, on which non-normative facts
ground prima facie duties and prima facie duties ground the verdictive facts,
is the most well-known member of the family.
In this chapter I offer a two-part argument for rejecting RP: RP is flawed
as a family of normative ethical theories (§1) and there are no meta-ethical
drawbacks to rejecting it (§2). The chapter concludes with §3.
1.1 Introduction
H. A. Prichard and W. D. Ross put Rossian Pluralism on the map. They argued
that certain kinds of action (Ross said there were seven kinds1) are obliga-
tory. If a single action is of more than one kind, such that it constitutes both
the discharging of at least one obligation and the violation of at least one
obligation, the rightness or wrongness of the action is determined by the
combined degree of obligatoriness of the individual obligations (Prichard
1912, 29 n. 1; Ross 1988, 46, 123).
For a long time, Prichard and Ross were in a lonely minority. RP was sim-
ply not very popular. This is because pluralism per se was not very popular.
All the most prominent moral philosophers before Prichard and Ross were
moral monists. This includes Kant as well as the consequentialists Bentham,
Mill, Sidgwick, and Moore. Prichard and Ross’s pluralism—or what went
by the moniker ‘intuitionism’—seemed to be what you were left with if you
wanted to be a non-consequentialist but not a Kantian. Even after Ross’s
The Right and the Good was published, in 1930, monism remained the
dominant view throughout the mid-twentieth century. This was the era in
which consequentialism, and particularly utilitarianism, ruled the landscape.
So for a while, whether you were a Rossian pluralist depended mostly
on whether you were a pluralist, which in turn depended a lot on whether
you found consequentialism or Kantianism attractive. So pluralism, and
by extension RP, was motivated by certain substantive views about right
and wrong. And we know those substantive views were rather unpopular,
because there were very few Rossian pluralists.
Then something changed. A new version of RP—one in which the verdict-
grounding normative facts are reasons—was introduced to the literature
by Kurt Baier (1958, Chapter 3) and J. O. Urmson (1974). By 1989 Shelly
Kagan, in his book The Limits of Morality, was pronouncing RP (though
he didn’t use that label, of course) to be a “familiar picture of how moral
requirements are generated” (1989, 65). And even Kagan himself, a strict
consequentialist, endorsed it (1989, 65; 1988, 17–18)! Nowadays RP is
pretty much taken for granted, as I’m sure the reader has noticed. Act-
consequentialists accept it (e.g., Portmore (2011)); rule-consequentialists
accept it (e.g., Hooker (2002)); non-consequentialists accept it (e.g., Nagel
Against Rossian Pluralism 63
(1986, 203), Rachels (1999, 196)); even people who reject both consequen-
tialism and non-consequentialism—particularists—accept it (e.g., Dancy
(2004)). What happened?
The short answer, I think, is that philosophers started to get very inter-
ested in the nature of reasons. Indeed, the most challenging task for anyone
who wants to make a case against RP is showing that we can, while reject-
ing it, hold on to some very intuitive claims about reasons. I am going to
postpone that meta-ethical discussion until §2. For the rest of this section
I take on the comparatively easy task of showing that RP is defective as a
normative ethical theory.
In other words, if an action has a characteristic that would make that action
morally required (or ‘duty proper’, as Ross would say) in the absence of any
other relevant considerations, we call that characteristic a pro tanto duty. But
this sounds exactly like what a present-day adherent of RP would say a moral
reason is!11 So the strategy of explaining moral residue by reference to pro
tanto duties, which Ross himself adopted (1988, 28), is at risk of being indis-
tinguishable from the proposal to explain it by reference to moral reasons.
To avoid this fate we need to part ways with Prichard and Ross and
say that a pro tanto obligation/duty really is a thing. This on its own is a
troublesome move to make, since the way we speak of obligations/duties
strongly suggests, as Liberman and Schroeder (2016, 107–8) have pointed
out, that there are no such things and that uses of the word “obligation”
are “just nominalizations of the is obligated to relation between agents and
actions”.12 But even allowing that it’s possible that a pro tanto obligation/
Against Rossian Pluralism 69
duty is a thing, the RP theorist still must tell us what thing it is and why we
should believe such things exist. She doesn’t need pro tanto obligations or
duties to fill the role of grounding the verdictive facts; she already has moral
reasons at hand to fill that role. To posit that, in addition to moral reasons,
there are pro tanto obligations or duties is to postulate a new entity simply
for the sake of explaining moral residue.13 So if we asked the RP supporter
what a pro tanto duty or obligation is, the only way she could say some-
thing helpful would be by saying that pro tanto duties/obligations are what
explain moral residue. Consequently, when the adherent of RP tells us that
moral residue arises in a certain case due to the presence of an unfulfilled
pro tanto obligation or duty, she simply tells us that moral residue arises in
that case due to the presence of something whose lack of fulfillment can give
rise to moral residue. And this is no explanation at all.14
This same objection applies to the other proposal we are considering—the
proposal that there is a sub-class of moral reasons such that leaving them
unfulfilled gives rise to moral reasons. The problem, again, is that we have
no independent way of getting a grip on which moral reasons those are. The
only possible identifying description of them would be: the ones that, if not
fulfilled, give rise to moral residue.
I conclude from all this that the RP supporter cannot show how to explain
moral residue, including the appropriateness of resentment and the demand
for an apology, without appeal to wrongdoing. But this was the core of the
RP supporter’s strategy for escaping the directedness objection to RP. The
idea was to show that even though RP explains wrongdoing in a way that
doesn’t pick out anyone as having been wronged, RP can still pick out a
wronged party (that is, the party who may appropriately feel resentful and
demand an apology) since the facts about who has been wronged are not
explained even in part by the facts about what wrongdoing has been done.
1.4 Summary
In this section I have identified two ways in which RP does a bad job of
explaining the verdictive facts. First, it fails miserably on the simplicity
criterion, because it must contain an unimaginably large number of type-
grounding claims covering the relation between non-normative facts and
facts about moral reasons. Second, given its picture of the grounding rela-
tion between moral reasons and verdicts, many cases of wronging will be
mischaracterized by RP as non-directed wrongdoing.15
There’s a very important reason why I start this section with an endorsement
of Reasons Ground Oughts. There are some, such as Joshua Gert (2005) and
John Broome (2013), who want to define reasons by reference to oughts. The
rejection of RP removes the most obvious way of defining ‘moral reasons’
by reference to moral oughts, namely by defining ‘moral reasons’ as the enti-
ties that ground moral oughts. So for those, such as Gert and Broome, who
deny that the concept of a reason is primitive and can be understood on its
own terms, there may arise a worry that rejecting RP causes us to become
unmoored—that it causes us to lose our grip on what moral reasons are. To
those theorists I offer Reasons Ground Oughts, the truth of which raises the
possibility of (but doesn’t require) defining moral reasons as the kind of things
that (can) contribute to overall oughts. Of course, this could not be a complete
definition of moral reasons, since there might be other kinds of reason that
contribute to overall oughts. Something, then, will have to be added to allow
us to distinguish moral reasons from other reasons. Here we can just add in our
favored theory of the distinctiveness of the moral point of view, for instance
The difference between P1 and P1* is that P1 does, while P1* does not,
insist that if X is morally required to φ then if she fails to φ she must be
blameworthy for failing to φ. P1* merely insists that X must be blamewor-
thy. This allows her to be blameworthy for getting herself into a situation in
which φ-ing was morally required of her, as opposed to being blameworthy
for what she did once she found herself in it. This, I think, allows us to say
what we want to say about cases of the second sort, just discussed (assum-
ing there are such cases)—cases in which one has two moral requirement-
generated reasons of the same strength such that one cannot act on them
both: we want to say you can’t be blamed for abiding by one requirement
and not the other. But we want to say that you can be blamed for getting
yourself into that situation. (Then again, maybe it’s possible to blamelessly
end up in such a situation. If so, we would want to reject P1*. But we would
also be robbed of our grounds for supporting P1. Obviously, I would be
content with this turn of events.)
78 Against Rossian Pluralism
If we replace P1 with P1* then the above argument becomes invalid.
Yet, P1* still gets us a strong connection between moral requirements and
blameworthiness, and if we accept Moral Rationalism then we still get a
strong connection between moral requirements and moral reasons. I don’t
see what the drawback would be.
2.3 Summary
In this section I focused my attention on two claims about reasons that strike
most moral philosophers as true and seem to constitute much of the appeal of
RP, Reasons Ground Oughts and Moral Rationalism. I argued that we can hang
on to both even while rejecting RP. Importantly, however, in order to explain
Moral Rationalism while rejecting RP, I had to endorse The Legal Model of
Moral Normativity. So in order to have a fair discussion of the options, I am
obligated to fold The Legal Model of Moral Normativity into the view I favor
and defend the package as a whole. This is what I will do in Chapter 5.
3. Conclusion
In this chapter I argued, first, that there are two valid explanatory objec-
tions to RP. The first is that RP runs afoul of the simplicity desideratum by
containing an enormously large number of type-grounding claims and the
second is that RP provides bad atomistic explanation by being inconsistent
with the existence of directed wrongdoing. I then argued that we can hang
on to both of the claims that seem to underlie the widespread support that
RP enjoys—namely Reasons Ground Oughts and Moral Rationalism—even
while rejecting RP.
Notes
1. Though he ended up whittling the list down to five.
2. For a defense of the ‘reasons come cheap’ thesis, see Schroeder (2007, Chapter 5).
3. For a defense of reasons holism, see Dancy (2004, Chapter 6). I call the rea-
sons holism feature popular because although Dancy uses its truth as a central
premise in his argument for particularism, many non-particularists (known as
‘generalists’) accept it too (e.g., McKeever and Ridge 2006).
4. Indeed, isn’t that their whole point? What Rossian pluralists take themselves to
be getting right, I gather, is that just about any non-normative fact can make a
moral difference.
5. Russ Shafer-Landau has reminded me that even a theory that contains an unimag-
inably large set of non-normative facts moral reasons grounding claims could
do well on the simplicity criterion, provided that it has a unifying explanation of
why the various non-normative facts ground moral reasons.
I concede that this is true. Whether this is damaging to my objection to RP
depends on whether there is a convincing unifying explanation to be found.
Shafer-Landau has suggested the one provided by Mark Schroeder (2007,
Chapter 1–2). Schroeder, who agrees that reasons come cheap (2007, Chapter 5),
holds that for each fact that grounds a moral reason (or, for Schroeder, is a
Against Rossian Pluralism 79
moral reason), r, there is the same explanation of why it does: that fact helps to
explain why the action that r supports is a way of promoting the object of one
of the agent’s desires (29). The problem for Schroeder is the following corollary
of this meta-explanatory claim: if φ-ing would promote the object of one of my
desires, then I have a reason to φ. This causes trouble because Schroeder has a
very expansive theory about what we desire. The problem isn’t that Schroeder’s
theory yields too many reasons, but rather that its yielding a lot of reasons
makes it inconsistent with many of our verdictive intuitions—specifically the
cases where we intuit that one person has a stronger overall reason to φ than
another person does. (This is exactly the sort of verdictive intuition that Schro-
eder designed his theory to accommodate; see the first two pages of Schroeder
2007). If reasons come as cheaply as Schroeder believes, then the reasons each
of us have are likely to coincide to a great degree. (I borrow this objection from
Brian McElwee.)
6. Except (1) when some of the moral reasons for or against the action are exclu-
sionary reasons—reasons whose presence prevents certain other reasons from
counting in favor of or against the action (thanks to Russ Shafer-Landau for
reminding me of this)—and (2) if Gert (2005) is right and some against an
action reasons don’t to contribute to making that action impermissible while
some reasons for an action don’t contribute to making that action permissible.
7. I won’t worry about constructing a thought experiment on the assumption that
there is neither a moral reason favoring nor a moral reason opposing the con-
ferring of sadistic benefit, since I doubt any RP supporter would make this
assumption.
8. For what it’s worth, Owens (2012, 44) says that all wrongings are wrongdoings,
though he doesn’t argue for this.
9. Liberman and Schroeder (2016, 107) and Shpall (2014, 158–61) agree.
10. Hare (1981, Chapter 2), Conee (1987, 241), McConnell (1996, 39–44), Brink
(1994, 220–3), and Dahl (1996, 94).
One strategy that used to be popular is to suggest that moral residue can be
explained by the fact that sometimes when we do what we are morally required
to do we nevertheless harm someone (McConnell (1996, 161), Conee (1987, 242),
Hill (1996, 193–4)). But this, to me, seems like a sub-species of the broader pro-
posal that sometimes when we do what we are morally required to do we never-
theless act contrary to moral reasons, since RP supporters will say that we have
a moral reason to avoid harming people but also moral reasons stemming from
other sources. So this proposal doesn’t merit a separate discussion, in my judgment.
11. David Brink (1994) quite explicitly translates Rossian claims about pro tanto
duty into the language of reasons.
12. We should say, similarly, that uses of the word ‘duty’ are nominalizations of the
is duty-bound to relation between agents and actions.
13. Russ Shafer-Landau suggested to me that a pro tanto duty is an invariably rel-
evant moral reason, while Joseph Raz is known for cashing out the supposed
duty to obey the law in terms of an exclusionary reason. If pro tanto duties are
invariably relevant moral reasons or exclusionary reasons, then postulating pro
tanto duties doesn’t amount to postulating a new entity. But by the same token
it fails to provide a good explanation of moral residue, because, as I already
argued, reasons can’t explain moral residue.
14. A different strategy, proposed by Shpall (2014), is to explain moral residue by
positing that sometimes when we do as we are morally required to do we never-
theless violate a commitment. This is in one way better than the pro tanto duty/
obligation strategy, since we have an independent grip on what a commitment
is (everyday normative talk includes talk of ‘commitments’). But it is in another
way worse, because it doesn’t hold out the promise of explaining the whole
80 Against Rossian Pluralism
range of cases in which moral residue arises. As Shpall more or less admits,
becoming morally committed to something requires an act of will; I must make
a promise, sign my name, take a vow, etc. Yet, there are cases in which moral
residue arises for some agent in spite of his not having engaged in any relevant
act of will. Suppose, for instance, that I face a situation in which I must either
harm A or harm B. If I harm A, I might owe an apology to her, despite the fact
that I never engaged in any act of will that committed me to refraining from
harming her.
15. I’d be remiss if I didn’t mention what I take to be a third compelling objection
to RP. (I omit a full discussion of it here since my goal is to demonstrate the
implications for normative ethics of the methodology laid out in Chapter 2. The
objection I lay out in this footnote has nothing to do with the points I made in
Chapter 2, since it has nothing to do with moral explanation.)
RP, like consequentialism, will have difficulty accounting for the (apparent)
fact that in some situations there is nothing that the agent is morally required to
do. Presumably versions of RP will say that an action is morally required just in
case it is better supported by the moral reasons than all the other actions. Any
version of RP that seeks to avoid implying that there are very few situations in
which there is nothing that one is morally required to do will have to be quite
strict about which non-moral facts ground moral reasons, thus yielding the con-
clusion that there are lots of situations in which the agent has no moral reason
to do anything. I have already expressed skepticism about the plausibility of
this move (which also came up in connection with my first objection to RP).
The other way is to tell a more complex story about how moral reasons come
together to generate moral requirements. This is the route Douglas Portmore
takes (2011, Chapter 5). Similarly, Joshua Gert (2005) tells a complex story
about how reasons per se come together to generate requirements per se.
16. Pekka Väyrynen, in conversation, raised the worry that once I commit myself
to the claim that reasons aren’t deontic, as I did when discussing moral residue
(§1.3.2), it’s no longer available to me to say that reasons can ground oughts,
since oughts are deontic.
In response, I want to admit that, yes, I am committing myself to the idea that
something deontic (an ought) can be grounded by something non-deontic (a
reason). But this should be entirely uncontroversial; everyone who thinks that
moral facts have non-moral explanations accepts this. And my admission
that moral reasons can ground deontic phenomena is no threat to my earlier
assertion that moral reasons cannot ground moral residue. That latter assertion
wasn’t based on any claim to the effect that moral residue is a deontic phenom-
enon. Rather, it was based on a claim that moral residue clearly results from
some deontic phenomenon; that is, moral residue is clearly (what we might call)
a secondary deontic phenomenon.
17. Darwall points out that if the concept of moral obligation were the same as
the concept of being best supported by the moral reasons, then supererogation
would be a conceptual impossibility. Clearly, however, it is not. Darwall (2010,
147–8).
18. Although I do not intend to defend this theory of legal normativity here, I want
to take this opportunity to defuse one likely objection to it. This theory of legal
normativity, like RP, is a claim about grounding. It is not a conditional. If it were
a conditional, there would be serious trouble for it. It would be the claim: if one
is legally required to φ then one has a legal reason to φ. This claim is subject to
all sorts of good counterexamples having to do with unjust laws. This theory of
legal normativity, since it is a type-grounding claim, merely says that under certain
conditions something’s being legally required can ground a reason to do it.
Against Rossian Pluralism 81
19. Of course, it would be discarded by anyone who rejects the existence of a sepa-
rate domain of legal normativity.
20. One might wonder why, then, it is appropriate for some violations of the law to
be punished more severely than others. (Thanks to John Broome for alerting me
to this question.) It wouldn’t be appropriate, I admit, if the sole legitimate pur-
pose of criminal punishment were a legal purpose—namely upholding respect
for the law as such. And I concede that punishment requires a purely legal justi-
fication (i.e., punishment is justified only in cases of law-breaking). But justifica-
tion is merely about when it’s OK to punish; it leaves aside the separate question
of how to punish when it’s OK to punish. To answer that latter question we
need to inquire as to the legitimate purpose(s) of punishment, and we may find
that punishment has a legitimate non-legal purpose—for instance, promoting
the social good, in which case it we would have an argument for the legitimacy
of using punishment for deterrence and prevention. And the deterrence and
prevention motives could explain why it’s appropriate for some violations of the
law to be punished more severely than others—viz, some violations of the law
are more important to deter and prevent than others.
21. One might worry, I suppose, that if legal and moral normativity work in the
same way, then law and morality won’t be distinguishable at all. But there are
several ways to distinguish moral and legal norms; see Hart (2012, 167–80).
22. Scanlon (1998, Chapter 4). Shafer-Landau (2003, Chapter 8) also endorses this
claim. But he doesn’t argue for it; he simply argues for the weaker claim that
if you are morally required to do something then you have a reason to do it.
Broome (2004, 53), meanwhile, says that moral requirements ground reasons,
but he doesn’t say specifically that they ground moral reasons, and I think he
probably would resist that claim. Finally, Temkin argues (2012, §7.4), in an
interesting discussion of intransitivity, that rightness is a good-making feature of
actions and I think endorses, toward the end of that discussion (211), the appar-
ent corollary that rightness is a reason-grounding feature of actions.
23. The ideal way to test its truth would be to deploy the method of difference:
construct two possible worlds, both containing some individual, X, such that
the two worlds are alike in all respects except that in only one of them X is
morally required to φ. Then one would examine those two worlds in order to
see whether X contains a moral reason to φ in the one world that she lacks in
the other. Unfortunately, we can’t do this. Given the supervenience of the moral
on the non-moral, two worlds that differ in terms of what moral requirements
they contain must also differ in some non-moral respect. But we might hope
to deploy a less pure but still effective version of the method of difference, as
Darwall does (2010, 155–6).
This is Darwall’s argument: In one case, I “intentionally tromp on your foot
to gain an advantage in getting to the table with designer underwear on sale
when I could easily have gone around and risked doing without a pair of Cal-
vin Kleins.” In the other case, I intentionally tromp on your foot, but this time
because it “is the only way I can avoid significant injury to myself.” Let’s grant
that I violate a moral requirement in the first case but not the second. So the
desired difference is there. And presumably we all have the intuition that I had a
stronger reason to do otherwise in the first case than I did in the second. So far
so good. But this is not a pure case since there is a second difference: the bur-
densomeness of doing otherwise. This allows us, if we are committed to denying
that moral requirements ground reasons, to insist that the burdensomeness in
the second case attenuated the strength of the reason I had in that case not to
step on your foot, and that it did so directly—that is, not by undermining what
otherwise would have been a moral requirement.
82 Against Rossian Pluralism
24. Many theorists believe that a thing’s being valuable just is its having some fea-
ture (other than its value) that provides reasons. The locus classicus for this view
is Scanlon (1998, 97).
25. Both readings are left open by what Smith actually says.
26. Remember, The Legal Model of Moral Normativity allows that we can have
non-moral reasons to do what we are morally required to do. So it can allow
that being motivated by the fact that φ-ing would harm someone counts as
being motivated by a reason; we just have to concede that facts about harm
ground (or are) non-moral reasons.
27. See Dancy (2004, 16–17) and Stratton-Lake (2011, 370–1).
28. Another, probably less convincing to most people, consideration in favor of
accepting the second element is this: We tend to think that we all have a moral
reason to prevent someone’s φ-ing if her φ-ing would be wrong. Suppose that
what grounds that reason isn’t the fact that the agent’s φ-ing would be wrong
but rather that it would constitute [inflicting harm, or breaking a promise, or
telling a lie, or . . .]. The problem with adopting such a view is that it has the
implication that we have moral reasons to tame nature, since animals in the wild
are always harming each other. For instance, predators kill and parents abandon
their young. If, on the other hand, what grounds the prevention reason is the
fact that φ-ing would be wrong, then there are no implications for intervening in
nature, since when animals do these things they do no wrong. So we should say
that the fact that an action constitutes [inflicting harm, or breaking a promise,
or telling a lie, or . . .] grounds no moral reason for a bystander to prevent it.
Given this, it seems a bit implausible to assert that such a fact does ground a
moral reason for the agent herself not to do it. It would be odd if moral reasons
for agents and moral reasons for bystanders failed to run in tandem.
29. I argue for this claim in Sachs (2006).
30. See, for instance, Darwall (2006, 28) and Portmore (2011, 43–4) for this kind of
blameworthiness-based argument and for citations of other similar arguments.
5 Non-Rossian Pluralism
3. Deliberative Rationalism
The third objection to NP concerns moral deliberation. NP holds that facts
about moral reasons don’t ground verdictive facts. If this were true, then it
would be hard to imagine that thinking about one’s moral reasons could
be a way of deliberating about what one morally ought to do. I expect this
apparent implication of NP will strike many people as highly counterintui-
tive. Many will want to say, I suspect, that if you want to figure out what
you morally ought to do then thinking about your moral reasons is precisely
what you should do. I’ll call this view Deliberative Rationalism. In this sec-
tion I aim to offer an independent argument (an argument that doesn’t use
NP as a premise) that Deliberative Rationalism is false.
a) identify the moral reasons relevant to A: N1, N2, N3, . . ., NX, then
b) identify other situations—B, C, D, . . ., Z—in which the same reasons
apply, then
Non-Rossian Pluralism 89
c) appeal to verdictive intuition to determine what one ought to do in B,
C, D, . . ., Z, then
d) use that information to determine the relative strengths of N1, N2,
N3, . . ., NX, then
e) use that information to determine what one morally ought to do in A.
Sometimes we know a verdictive fact and can use that knowledge to infer
facts about the relative strengths of certain reasons. (This is precisely how
Joshua Gert (2005) suggests we can discover the relative strengths of rea-
sons.) Then we can take that knowledge and work in the other direction,
inferring further verdictive facts from the just-discovered facts about the
strengths of various reasons.
This is sound methodology, I admit. (We might call it ‘lateral’ moral rea-
soning.) But it doesn’t get us what we want. Moral reasoning isn’t supposed
to merely fill in the blank spots in our verdictive belief-set; it’s supposed to
be a way of correcting one’s mistaken verdictive beliefs if one has any. And
this is exactly what we deliberators lack if Deliberative Rationalism is true
and we don’t have intuitions about reason-strengths in hard cases. Without
those intuitions, we don’t have a second set of moral inputs in hard cases
(as opposed to easy cases, where in addition to verdictive intuitions we have
reason-strength intuitions). Moral reasoning becomes a matter of figuring
out how much information one can extract from one’s verdictive intuitions
plus the non-normative facts. But if one reasons in this way, then one will
never be pressured to alter or change any of one’s pre-existing verdictive
beliefs. And that, officially, is my objection to Deliberative Rationalism. In
other words, I reject Deliberative Rationalism because it is incompatible
with the idea that forms the basis of this book—the idea on which the Intro-
duction and Chapter 1 were premised: that moral theorizing can be a way
of discovering that one was mistaken about what one morally ought to do.
5. Conclusion
In this chapter I defended NP against what I take to be the most obvious
objections that would be raised against its structure. I cannot in this section
summarize all the arguments just mentioned. But in the interest of having a
fair comparison of our options in normative ethics I want to make a final
note of an important debt we incur by adopting NP, namely putting our-
selves in a tenuous spot with respect to genuine moral dilemmas. NP does
not, strictly speaking, imply their existence, but it does imply that every
moral conflict is a genuine moral dilemma.
How big this problem is for NP we cannot know until we begin to hone
in on the best version of it, and thus gain some idea of which non-normative
Non-Rossian Pluralism 93
facts it holds out as morally significant and (as a consequence) how frequent
moral conflicts will be on the theory. Having gone some way toward doing
this, I will return to the question of moral conflict in Chapter 10.
***
Before turning to Chapter 6, I want to take the opportunity to explain how
the contents of the second half of this book will be related to what has trans-
pired in the first half. In Chapter 6 I argue that no generalization having to
do with the distribution of welfare or of any object whose moral impor-
tance is dependent on its relation to welfare can ground any verdictive facts.
Since, as I argued in Chapter 2, we should treat normative ethical theories
of being composed of nothing more than generalizations that do explana-
tory work, I therefore conclude that we should exclude such generalizations
from our normative ethical theory. In Chapter 7 I argue, once again drawing
on the methodology of Chapter 2, that we can give very sensible answers
to moral questions about our treatment of animals (and other individuals
less psychologically sophisticated than you and I) without adverting to talk
of moral status, and that furthermore there is no independent explanatory
work that facts about moral status could possibly do.
It will be only in Chapters 8–9 that I begin to bring out the verdictive
implications of accepting NP, especially as by contrast to RP. The big pic-
ture, which I have already hinted at, is that if we put ourselves in the mode
of thinking, as NP says we should that non-normative facts, when they’re
explanatorily relevant, are relevant by virtue of grounding verdictive facts,
we’re likely to wind up with a much shorter list of non-normative facts that
are explanatorily relevant than we would if we were in RP mode—i.e., if we
were thinking that explanatorily relevant non-normative facts are relevant
in virtue of grounding moral reasons. So we’re likely to end up with a more
spare normative ethical theory. Specifically, in Chapter 8 I argue against
duties of beneficence, and using this conclusion combined with the conclu-
sion of Chapter 7 I argue that we should stop conducting harmful research
on animals. And finally, in Chapter 9, I draw on Chapters 6 and 8 to argue
that our best normative ethical theory will have no answers to the questions
that philosophers seem to be trying to answer when engaging in thought
experiments regarding the distribution of scarce health care resources.
Notes
1. If the grounding relation is transitive, which I have thus far allowed for the sake
of argument, then NP encompasses RP; RP would be the view that the verdic-
tive facts are grounded by certain normative facts and (indirectly) grounded by
certain non-normative facts. Likewise, if moral reasons reduce to non-normative
facts, then NP is RP plus the reduction claim. However, there would still be
room for pluralists to disagree about how the non-normative facts/moral reasons
ground the verdictive facts. In particular, most (I take it) would believe that the
non-normative facts/moral reasons jointly ground the verdictive facts. In other
words, the non-normative facts/moral reasons relative to an action are in some
94 Non-Rossian Pluralism
sense combined (perhaps by having their reason-giving strength aggregated) and
it’s this combination that grounds the verdictive fact about that action. This pic-
ture of how the grounding works is the sort of picture that’s quite natural only if
one thinks that the non-normative facts/moral reasons behave in more the way
that non-reductivists would say moral reasons behave and less in the way non-
reductivists would say that non-normative facts behave. It’s moral reasons, not
non-normative facts, that can easily be thought of as having a weight and as being
combinable into a sort of supra-entity (i.e., a supra-reason). So if the ground-
ing relation were transitive or reductivism were true, then although in principle
all pluralists would accept NP most of them would believe that the things that
ground the verdictive facts are things that behave in a reason-ish way.
This version of NP—the picture that most people would accept were they con-
vinced that grounding is transitive or reductivism is true—I consider just as objec-
tionable as RP, and for just the same reasons. Both models of moral grounding
embrace the general idea that the verdictive fact about an action is grounded in all
the considerations relevant to it, an idea to which I objected in §1.3 of Chapter 4.
So neither the truth of the transitivity of grounding nor of reductivism would be
of any help to me. The same is true, obviously, of the falsity of these theses. This
is why I don’t discuss them.
2. But I would like to point out that I’ve already put a dent in the ought-implies-can
argument against the possibility of moral dilemmas. Insofar as there is an argu-
ment that ought implies can, it’s likely to proceed by way of blameworthiness.
The first premise would be that ought-but-failed-to implies blamable-for, and the
second would be that blamable-for implies could have. But I’ve already argued,
in §2.2.5 of Chapter 4, against the first premise.
3. Sinnott-Armstrong (1988) may have been the first philosopher to use moral resi-
due as an argument for genuine moral dilemmas.
4. It also wouldn’t count for much if moral residue could be accounted for given
moral monism. But moral monism can’t allow for moral conflict, and therefore
obviously cannot account for moral residue.
5. Scanlon (2014, Chapter 5) agrees.
6. I first raised this worry specifically against Douglas Portmore’s consequentialism,
which is a version of RP. See Sachs (2013). A similar point has been made by
Richardson (1990, 287–9). McMahan (2000, 105) confronts the same problem
and simply bites the bullet: he says that the direction of discovery is the reverse
of the direction of grounding. What this means, I think, is that he accepts RP but
rejects Deliberative Rationalism.
7. So far I have mentioned just two ways of arriving at a reason-strength belief:
(1) having a reason-strength intuition, and (2) inferring such a belief from one’s
verdictive beliefs. I don’t want to deny that there might be other ways of arriving
at a reason-strength belief; I simply maintain that the other methods don’t solve
the problem for Deliberative Rationalism. Here are three case studies:
Kearns and Star believe that a reason to φ is evidence that one ought to φ,
and the strength of any reason is the strength of the evidence it constitutes,
where the strength of a piece of evidence is the degree to which it raises
the probability of whatever proposition it is evidence for (Kearns and Star
2009, 230–2; see Thomson 2008 for a similar view). For our purposes
the question is: If Kearns and Star were right about what a reason is and
what determines its strength, would this make available a new defense of
Deliberative Rationalism? The answer, unfortunately, is no. One will have
no greater ease in directly discerning the degree to which some fact makes
it more likely to be the case that one ought to φ than one would have in
directly discerning the strength of some reason to φ.
Non-Rossian Pluralism 95
A similar result emerges if we adopt thoroughgoing Humeanism, where
‘Humeanism’ is the view that each reason is grounded in a desire, ‘thorough-
going Humeanism’ is a view on which the strength of a reason is determined
partly or wholly by the strength of the desire that grounds it. Here the ques-
tion is: If thoroughgoing Humeanism were right about what a reason is and
what determines its strength, would this make available a new answer as
to how one can determine how strong one’s reasons are? Again, the answer
is no. For some Humeans (e.g., Smith 1996, 167) the relevant desires are
hypothetical, meaning we would have a good method for discerning the
strengths of reasons if and only if we had a good method for discerning
the strengths of hypothetical desires—which, of course, we don’t. For other
Humeans (e.g., Manne 2016) the relevant desires are actual but the weighti-
ness of the various reasons is based on agents’ hypothetical prioritizations
of the satisfactions of their various desires, meaning that we would have
a good method for discerning the strengths of reasons if and only if we had a
good method for discerning agents’ hypothetical prioritizations—which, of
course, we don’t (Manne (2016, 134) admits as much).
The final case study is Mark Schroeder’s (2007) theory of reason-strengths.
First, for clarification, his view is that reasons don’t have weight, but rather
sets of reasons do (126). And one of the ways for one set of reasons, A, to
be weightier than another, B, “is for the set of all the (right kind of) reasons
to place more weight on A to be weightier than the set of all (the right kind
of) reasons to place more weight on B (138).” So if we’re to have a reli-
able way of figuring out where the balance of reasons to act lay, we need a
reliable way of figuring out where the balance of reasons to place weight
on reasons lay. Do we? Schroeder says some helpful things (141–44), but
they do not, nor do I think they’re intended to, actually constitute a use-
able method for figuring out where the balance of reasons to place weight
on reasons lay. Providing such a method isn’t really necessary given the
agenda of Schroeder’s book. He’s interested in showing how we can explain
why there are weightier reasons for certain people in certain situations to
do certain things—and, in particular, that a thesis that connects reasons
to desires provides a good explanation—as opposed to showing how we
can figure out which reasons for certain people in certain situations to do
certain things are weightier than which.
The agenda for the rest of the book is to get a better idea of how the most plau-
sible version of non-Rossian pluralism (NP) will look. This means identifying
some of the generalizations it will include and others it will exclude. Now we
already know, based on the argument of Chapter 2, that the generalizations
to be included in a normative ethical theory are type-grounding claims. Hav-
ing said this much, I am in a position to state the thesis of this chapter: the
best version of NP will exclude grounding claims concerning the distribution
of welfare; i.e., it will deny that any non-normative fact about the distribution
of welfare can ground the obligatoriness or impermissibility of any action. In
fact I will argue, more broadly, that this is true of all generalizations regarding
the distribution of welfare and all generalizations regarding the distribution
of anything whose moral importance is dependent on its relation to welfare.
Hereafter I’ll refer to this combined class of claims as ‘welfaristic distributive
generalizations’ (WDGs). I am not, however, going to argue that there are no
true moral generalizations regarding the distribution of welfare; I’m actually
going to offer, at the end of this chapter, an optimistic roadmap for identifying
some. What I am going to deny, however, is that there are any generalizations
about the distribution of welfare that do grounding work.
Before beginning the argument, I want to say something about why my
thesis is important given the concession I just made. If no WDG is itself a
true grounding claim, then any true WDG would have to be true in virtue
of being derivable from some true grounding claim. Consequently, the only
thing we should be debating, with respect to any WDG, is whether it is
derivable from something else. This methodological implication is this chap-
ter’s key take-away.
Importantly, this is not the sort of debate that contributors to the literature
on distributive justice/ethics are having with each other. Contributors to that
debate put forward candidate generalizations, which are judged for whether
they are intuitively plausible on their own, deliver plausible case-specific
verdicts, and are simple enough so as to seem non-gerrymandered (which is
probably a way of examining whether they can plausibly be thought to do
explanatory work). I trust this description rings true to anyone who is familiar
with that literature.
The Question of Scope, Part 1 97
This not a sensible debate to have over generalizations that one knows are
not grounding claims. Imagine I asserted the generalization, ‘on every Wednes-
day afternoon, Bobby should give Sue an apple’. You would recognize instantly
that I wasn’t attempting to identify a ground of obligatoriness, and therefore
you wouldn’t worry about its intuitive plausibility, or its implications for cases,
or whether it seems gerrymandered. You would simply ask, “Why?” (I might
respond, for instance, that Bobby promised to give Sue an apple on every
Wednesday afternoon, thus implicitly adverting to some generalization about
promise keeping grounding moral obligatoriness). Odd as it may seem, I’m sug-
gesting that we should treat WDGs in precisely that way.
As to why no WDG could possibly be a true grounding claim, the answer
has to do with the issue of scope. No distribution of a good can be described
without setting its scope. Simply identifying the good to be distributed and
the designated rule or pattern according to which one ought to distribute
it isn’t good enough. One must also specify among which individuals one
ought to distribute the designate good in the designated way. My claim is
that once we specify the scope of any WDG we will find, no matter what
scope we chose, that the resulting WDG fails to constitute a true grounding
claim. We will know this because we will see that it fails to identify some
non-normative fact that is morally important in itself, which is something
a generalization must do in order to qualify as true type-grounding claim.
So, for instance, if the generalization ‘the fact that an action constitutes
the keeping of a promise can ground the obligatoriness of that action’ is a
true type-grounding claim, this means that it is morally important in itself
whether promises are kept.
Now most theorists working in this area defend their favored candidate
WDGs without specifying their scope.1 I do not intend here to pick apart any
of these defenses. They are not a target for me, since they fail to even attempt
the project that I am saying must be done—namely, defending a WDG while
specifying its scope. Instead of attempting to establish a scope, they implicitly
assume one. So my intention here is simply to demonstrate that if they were
to attempt the project, they would fail. Indirectly this will constitute a refuta-
tion of many of the arguments that have been given for certain WDGs.
Before doing all this, however, I want to be more specific about my topic.
I said I’m targeting all WDGs, but theorists differ as to why we’re obli-
gated to distribute welfare (or another good whose moral importance is
dependent on its relation to welfare) in a certain way. Parfit (2002), in a
famous paper, notes that among those who believe that we’re obligated to
bring about a certain distribution of welfare, some of them believe that this
is because that distribution is good, whereas others accept the obligation
on non-evaluative grounds. Furthermore, theorists differ not only in their
explanation as to why we’re obligated to bring about a certain distribution,
but also, of course, in which distribution they would have us bring about.
These distinctions need not delay us, however, since I take as my target all
WDGs. When I need examples of such claims, I will simply abstract away
98 The Question of Scope, Part 1
from the question of why we might be supposed to be obligated to bring
about that distribution. And as to the question of what that distribution is
supposed to be, I will use egalitarian, prioritarian, and sufficientarian distri-
butions as my case studies. Again, however, I intend my thesis to apply no
matter what the distribution.
I begin, in §1, by arguing that there is a problem for WDGs regarding the
distribution of welfare; this argument forms the basis for the overall argument
of the chapter. The problem is this: As I said, every WDG must be assigned a
scope. The widest-possible scope, for generalizations about the distribution
of welfare, would include all individuals with a welfare. However, such a
WDG would have extremely counterintuitive implications regarding what
is morally important in itself due to the fact that among individuals with a
welfare—a class that includes both human beings and snakes—there are rad-
ical natural differences in welfare-potential. To solve this problem we could
go to the opposite extreme of setting the scope of our favored generalization
such that it applies only within groups of individuals who are precisely equal
in terms of welfare-potential. This, however, would render the generalization
practically irrelevant, since it is probably very rare for two individuals to
be precisely equal in this way. The alternative, of course, is to find a middle
ground between the extremely wide scope and the extremely narrow scope,
but the problem is that all the moderate stopping points are arbitrary, and
arbitrary facts can do no explanatory work. The conclusion of §1 will be
that no WDG regarding the distribution of welfare identifies something that
is morally important in itself.
Having made this argument in §1, I explore in §2 the idea, proposed by
Jeff McMahan and Peter Vallentyne, that the focus on welfare is what’s
causing trouble for our distributive generalizations, and that the problem
might be solved by focusing instead on fortune. This seems an appealing
move because unlike welfare there are not drastic natural differences in
potential to be fortunate. However, I argue that the success of some gener-
alization concerning the distribution of fortune rests on the success of some
generalization regarding the distribution of welfare, and consequently the
turn toward fortune is a red herring. In §3 I generalize the lesson of §1–2
and show that if there is a problem for WDGs concerning the distribution
of welfare, then there is a problem for all WDGs (that is, even the WDGs
concerning goods besides welfare whose moral importance is explained by
their connection to welfare). Section §4 is a response to an objection, while
§5 is where I show how we might construct a plausible derivation of a WDG
from a true grounding claim, which amounts to conceding that for all I’ve
said some WDG might be true. Finally, §6 concludes.
4. An Objection
I have argued that all versions of distributive welfarism have no moral force
of their own. One might wonder, however, how much this shows. Admitting
that no one has yet found a way of fully spelling out a version of distributive
108 The Question of Scope, Part 1
welfarism—where ‘fully spelling out’ means specifying a scope—while
avoiding serious difficulties does not commit us to concluding that no ver-
sion of distributive welfarism has moral force of its own. To make this infer-
ence, one might worry, is to impose too high of a standard on those who
are convinced that the distribution of welfare matters morally. We don’t,
for instance, require ourselves to have a fully satisfactory theory of what
constitutes lying before we can consider ourselves justified in believing that
the generalization prohibiting lying has moral force of its own. In the case
of lying, we consider it enough if we are justified in believing that the uncon-
troversial cases of lying are wrong, and wrong because they’re lying. Why
not say that it’s enough if we are justified in believing that the central cases
of failure to distribute welfare appropriately are wrong, and wrong because
they constitute such failures?13
My response is that it would be enough, but we are not in fact justified
even with regard to the central cases. To illustrate the point, I will borrow an
example from Thomas Nagel’s article “Equality.” This is a classic example
from a classic article that sought to motivate a concern for the distribution
of welfare. If any case should do the trick, it’s this one.
Suppose I have two children, one of which is normal and quite happy,
and the other of which suffers from a painful handicap. Call them
respectively the first child and the second child. I am about to change
jobs. Suppose I must decide between moving to an expensive city where
the second child can receive special medical treatment and schooling,
but where the family’s standard of living will be lower and the neigh-
borhood will be unpleasant and dangerous for the first child—or else
moving to a pleasant semi-rural suburb where the first child, who has a
special interest in sports and nature, can have a free and agreeable life.
[. . .] the gain to the first child of moving to the suburb is substantially
greater than the gain to the second child of moving to the city.
(Nagel 2002, 75)
Nagel expects the reader to have the intuition that there is a special moral
pull in favor of moving to the expensive city, and takes this to count in favor
of a moral theory that incorporates distributive concerns as opposed to a
theory that is composed of nothing more than the principle of utility.
But this inference is too quick. There are other ways to explain why the
parent should move to the expensive city (assuming that he should). For
instance, we might think that one of the duties of a parent is a duty to
ensure that one’s child does not suffer unnecessarily. Given the facts of the
case as Nagel presents them, it’s reasonable to read into the case the fur-
ther fact that the only way neither child ends up suffering is if the family is
moved to the city. The broader point, however, is simply that a verdictive
intuition is not, in itself, an intuition in favor any particular generaliza-
tion or family of generalizations. What generalization to endorse in light
The Question of Scope, Part 1 109
of a verdictive intuition is something we have to reason out. And part of
that process involves coming up with a list of candidate generalizations. We
can’t take just any generalizations and examine the extent to which they are
supported by our case-specific intuitions; that would take forever. Rather,
the best strategy is to take a list of generalizations such that each of them
seems sensible on its own—that is, in advance of considering the extent
to which it accords with our verdictive intuitions—and then examine each
member of that list to determine how well it accords with our verdictive
intuitions. (Remember, I argued in Chapter 1 that the generalizations out
of which normative ethical theories should be composed are themselves apt
to be evaluated for intuitive plausibility.) An analogy: We all have the case-
specific intuition that it was wrong of Germany to invade Poland on Friday,
September 1, 1939. We could use this intuition in support of the following
generalization: That a foreign invasion takes place on a Friday can ground
the wrongness of that action. But we wouldn’t actually do this, because that
generalization doesn’t strike us as plausible on its own. Similarly, we have
yet to manage to formulate a complete (that is, with scope included) version
of distributive welfarism that seems to us to have moral force of its own.
So we shouldn’t count our intuition in Nagel’s case in favor of some as-yet-
unspecified version of distributive welfarism.
6. Conclusions
We began this chapter by exploring the idea that some generalization regard-
ing the distribution of welfare has moral force of its own. I argued that none
of them do. Thus, we have made some initial progress in identifying the best
version of NP: We know now that it will not contain any welfaristic distrib-
utive generalizations (WDGs). That is the main conclusion of this chapter.
This conclusion makes for a sobering prospect for those who start from
the idea that distribution matters morally, such as Amartya Sen (1992, 3–4)
and G. A. Cohen (1989, 906), who simply assume that we are in the busi-
ness of determining what it is that shall be distributed equally. If they want
to defend a WDG, they will no longer be able to do so by identifying one
that has intuitive force, gets the cases right, seems non-gerrymandered, etc.
It may, however, be the case that they will manage to identify a grounding
claim the truth of which, taken together with supplemental non-philosophical
premises, implies some WDG. Consequently, philosophers who work on dis-
tribution must not shy away from foundational moral theory. This is the
second part of my conclusion.
112 The Question of Scope, Part 1
Notes
1. Here are some examples of this methodology in practice.
• In defense of welfaristic egalitarianism: Sen 1992; Temkin 2011; Persson
2007; McKerlie 1989; Dworkin 2000.
• In defense of welfaristic prioritarianism: Arneson (2000b). Likewise, Schef-
fler (1982, esp. pp. 31–2) and Feldman (1997, Chapter 8, esp. pp. 161–2)
each deploy this methodology in defending the prioritarian element in their
respective versions of consequentialism.
2. So, for instance, certain emotional capacities are necessary for achieving various
relationship goods, and certain cognitive capacities are necessary for achieving
various intellectual goods.
3. For a defense of such a view, see Parfit (1987, Chapters 10–11).
4. This is Savulescu’s (2011) view.
5. We should beware of allowing ourselves to get fooled by the ease with which we
sort human beings into the categories of psychologically normal and psycho-
logically abnormal. This is easy for us to do because we have a set of familiar
categories of mental disability and, thanks to psychologists and psychiatrists,
relatively well-defined criteria for sorting individuals into those categories.
Let’s not forget, however, that these categories and criteria exist to serve a
set of practical agendas. For instance, we have to decide which individuals
can legitimately plead insanity as a criminal defense, which individuals should
be given welfare benefits on account of their inability to do socially useful
work, and which individuals should be allowed to take controlled substances.
(These, anyway, are some of our practical agendas. We also might have more
sinister agendas when it comes to diagnosing mental disability, but I won’t go
into that here.) These neat distinctions do not map onto any underlying neat
distinctions in levels of mental functioning. Rather, for any domain of mental
functioning there is a continuum from high levels of that functioning to the
complete absence of it. We have to draw a line somewhere, and we can do so
non-arbitrarily if we’re careful, but this doesn’t mean that we’re responding to
some underlying normative fact of the matter as to which levels of functioning
are defective and which are not.
6. Here’s why: Whatever the true theory of welfare is, it’s necessarily true. (All
high-level evaluative truths are necessary truths.) This theory includes some
facts about how the exercise of psychological capacities constitutes welfare and
about how each domain of welfare is irreplaceable. So if we hold fixed the
actual facts about an individual’s psychological capacities (which is what we
should do, given that capacity for welfare is defined in terms of those facts),
then we can derive a necessary fact as to the most welfare that that individual
can have. That latter fact will be, by the definition of ‘capacity for welfare,’ that
individual’s capacity for welfare.
Admittedly, I’ve encountered in discussion a fair bit of resistance to the idea that
there are any differences (never mind necessary differences) in capacity for wel-
fare. My reason for thinking that there are is, as I just explained, simply that many
elements of welfare are realized through the exercise of psychological capacities
that some individuals lack. But this argument rests on an unstated assumption:
that there is a universally applicable true theory of welfare. Some people believe
instead that there is a distinct true theory of welfare for each living thing.
I admit that there is some intuitive plausibility to this. It does seem a bit chau-
vinistic to infer, from the fact that having loving relationships is an element of
welfare for me, that it’s also an element of welfare for the snake, even though
The Question of Scope, Part 1 113
the snake has no capacity for love. It may seem more natural to construct a
theory of welfare specific to that snake, based on our observations and our
theorizing about its nature.
If that’s the agenda, however, then we confront a challenge: How do we make
welfare comparisons between individuals? Suppose that Smith’s welfare scale
runs from a (the least welfare she could have) to z (the most welfare she could
have), while Jones’s welfare scale runs from α to ϖ, and that we know what
point in Smith’s scale Smith occupies and what point in Jones’s scale Jones occu-
pies. From this information, how can we reach a conclusion regarding which
of them has more welfare? It depends on the details of the case, but at the very
least we would have to determine that a particular point in Jones’s scale cor-
responds to a particular point in Smith’s. But this is exactly what we’re now
assuming can’t be done. We’re assuming that Smith has her way of flourishing
based on her nature, and Jones has her way of flourishing based on her nature,
and that the one has nothing to do with the other.
If we cannot make welfare comparisons between individuals, then the egal-
itarian and prioritarian versions of distributive welfarism are at best practi-
cally irrelevant and at worst incoherent, because the inequalities they tell us to
address turn out to necessarily not exist. And as to the sufficientarian version,
it too would be incoherent, as it would be unable to pick out, as its sufficiency
threshold, a level of welfare that for each being with a welfare would map onto
some particular location on that being’s scale of potential welfare. So the claim
that there are no differences in capacity for welfare is no threat to my agenda
here.
7. This is actually a much-simplified version of the theory of fortune that McMa-
han (2002, Chapter 2) proposes.
The major worry about the ratio-based conception of fortune is that it might
imply that having congenital psychological limitations is not a misfortune. Let’s
suppose that some version of the continuity theory of personal identity is true,
and correspondingly that one’s genetically determined psychological make-up is
identity fixing. This being the case, congenital psychological limitations, while no
doubt reducing one’s welfare, will also reduce one’s welfare-potential to a cor-
responding degree, thus leaving the afflicted individual no worse-off ratio-wise.
Since the turn from welfare to fortune is being put forward on my opponent’s
behalf, I’m going to simply assume that this objection to the given account of
fortune can be answered. (For what it’s worth, I think that McMahan (2009)
has done a nice job of undermining the objection.)
8. This claim, as applied to Vallentyne, is a bit of a fudge. Distributive fortun-
ism implies that it matters morally in itself how different individuals’ welfare/
welfare-potential ratios compare, and this isn’t quite what Vallentyne believes.
He believes something closer to the view that it matters morally in itself how
different individuals’ welfare/capacity for welfare ratios compare. This fudge
will make no difference moving forward. What I say against distributive fortun-
ism applies just as well to Vallentyne’s view.
9. In fairness to McMahan, his concern isn’t the prospects for distributive general-
izations but rather the question of what we owe to the severely psychologically
disabled.
10. For an argument to the effect that distributive fortunism is simply false, see
Arneson (1999) and Holtug (2007).
11. See, e.g., Temkin (2011).
12. Arneson (1989) used to endorse equality of opportunity, and for the reason just
mentioned.
13. Thanks to Wlodek Rabinowicz for pushing me to address this objection.
114 The Question of Scope, Part 1
14. This kind of inference—from the moral importance of a particular psychologi-
cal ability to the claim that for an individual to fall within the scope of the the-
ory she must meet some threshold of psychological sophistication—is explicitly
made by Rawls (1999a, 441–3).
15. It might seem disingenuous of me to take seriously, as I have been here, the possi-
bility of grounding welfaristic distributive generalizations in contractarianism or
contractualism, given my earlier hostility to those theories (see Chapter 3). Two
things, however: First, I never said we shouldn’t be contractarians/contractualists;
rather, I said we shouldn’t be monistic contractarians/contractualists, which is
to say that we shouldn’t accept only contractarianism/contractualism. Second,
this book is about normative ethics, but since normative ethics doesn’t exhaust
the normative domain there are other things to be contractarians/contractualists
about. Normative ethics is, I take it, the study of the moral constraints incumbent
on individual moral agents as such. This would mean we need a separate theory
to discuss the special moral constraints, of which there surely are some, that
apply only to collective moral agents, such as states, and only because they’re
such agents. Call this subject matter ‘political morality’. I believe, as I will explain
in Chapter 9, that contractarianism/contractualism is much more plausible as a
political morality than as a normative ethical theory. On the possibility of differ-
ent theories being true in different moral domains, see Bräanmark (2016).
16. Why derivative moral importance but not non-derivative moral importance?
Because of what hypothetical contractors, or individuals trying to agree to a set
of rules that would allow them to justify their behavior to each other, can and
cannot accomplish. What they can accomplish, if contractarianism or contrac-
tualism is true, is to make it the case that we are morally obligated to abide by
some generalization regarding the distribution of welfare. What they cannot do
is make it the case that some such generalization has moral force of its own. No
one can make something have moral importance of its own. The fundamental
moral truths are necessarily true.
17. This conclusion is in the same spirit as some of Parfit’s (2002) comments in his
“Equality or Priority?” In the first four sections of that paper Parfit discusses
a scope objection to welfaristic egalitarianism. Whereas I have been concerned
with the question of scope along the dimension of psychological sophistication,
Parfit is concerned with the question of scope along the dimensions of space
and time. His objection to welfaristic egalitarianism is that it makes no sense to
worry about inequalities in welfare between people who live in different periods
of time or different parts of the world or universe. Parfit then points out that
there is a version of welfaristic egalitarianism that avoids this problem: deontic
egalitarianism. Of course, the deontic egalitarian must explain why some wel-
fare inequalities are unjust and others aren’t. This is broadly in line with what I
argue here; in fact Parfit even suggests, like I do, that an appeal to contractarian-
ism might do the trick.
7 The Question of Scope, Part 2
Non-Distributive Moral Concerns
Since my concern is with whether we need to talk about moral status when
inquiring into how we ought to treat fetuses, disabled humans, animals,
etc., I am not going to discuss whether moral status facts might plausibly be
thought to play the second or third role listed above. If moral status facts
do play one or the other of these roles, it remains true that there is no need
to talk about moral status, since we could instead just talk about the facts
120 The Question of Scope, Part 2
to which moral status facts reduce. In what follows, then, I will discuss the
other two possibilities.
The objection could be sustained if there were such things as (what I will
call) deliberative obligations—obligations to take certain individuals into
account in deciding what one ought to do.13 One might think, on the con-
trary, that we are really only answerable for our intentions and actions; that
we are morally innocent so long as we do the right thing, regardless of how
we arrived at the conclusion that that was the right thing to do.
Nevertheless, there is some intuitive pull to the idea of deliberative obliga-
tions. Unfortunately, no one has ever offered a theory of the scope and basis
of such obligations. This is troublesome, as there are numerous questions that
arise immediately when considering the idea of deliberative obligations. If the
suggestion is that one is always obligated to take into account, when deter-
mining what one ought to do, every individual that can be or at least possibly
can be wronged, the requirement is much too demanding. My high school
English teacher can be wronged, obviously, but I need not take her into con-
sideration when deciding whether I ought to have a sandwich for lunch.14 If,
on the other hand, the suggestion is that one is obligated to take into account
just those individuals whose interests might be affected by what one does,
The Question of Scope, Part 2 123
other problems arise. First, an epistemic one: How can I identify all the people
whose interests could be affected by my contemplated courses of action, given
how easy it is for someone to have an interest in something? (One can have an
interest in something merely in virtue of caring deeply about it and thus pos-
sibly being happy or sad about how it turns out.) Second, even if those cases
in which I have all the relevant information, some problems remain:
The bottom line is that, on the assumption that X’s having moral status is
the fact that X can, or possibly can, be wronged, the knowledge that X has
moral status does not on its own tell us whether one should take X into
account when deciding what one ought to do.
So it is simply not true that when we deliberate about what we ought
to do we are required to take into account all beings that have moral sta-
tus. The actual deliberative obligation, if there is one, must be much more
lenient. And so deliberating in a morally permissible way does not require
knowing which individuals have moral status.
In this section we have found that if ascriptions of moral status merely
tell us that the being in question can be wronged in some way or other, or
possibly can be wronged in some way or other, then such ascriptions tell us
nothing useful. This should not come as a surprise. What is important to
know is in what ways different individuals can actually be wronged.
5. Conclusion
My main goal in this chapter has been to put forward a method of answer-
ing scope questions for non-comparative moral concerns—a method that
is suitable for moral pluralists. I endorsed James Rachels’s method. For
each non-comparative moral concern our task is to think about what non-
normative characteristics an individual would need to have in order to be
wronged that way. Sometimes the answer will be immediately obvious,
while in other cases the answer will become clear only after engaging in a
more substantive investigation of the ultimate source of the moral concern.
In any event, having identified the crucial non-normative characteristic, the
rest of the job is empirical: identify which individuals have that character-
istic. Whichever ones do, they are within the scope of that moral concern.
124 The Question of Scope, Part 2
Secondarily, I set out to resist the idea, prevalent in most non-
consequentialists’ writings on scope questions, that there is a broader or
more unitary question to answer in this area—i.e., some question about
which individuals are in The Scope of Morality per se, or about which indi-
viduals have (how much) Moral Importance. I stressed that these questions
make sense only from a morally monistic point of view, and that insofar as
we moral pluralists attend to them we risk discarding what was attractive
in the first place about our moral pluralism. In particular, I tried to demon-
strate the uselessness of engaging in talk of moral status. Moral status facts
play no independent role in moral explanation and there is nothing we can
say about scope questions using moral status talk that we can’t say, and usu-
ally more perspicuously, without it.
Notes
1. One might think that the issue is whether morality includes distributive con-
cerns at all. But this is not the case. If there is a derivatively true distributive
generalization, then the type-grounding claim that yields it should also yield a
clear delineation of its scope, thus obviating the need for any discussion specifi-
cally of the scope of that distributive generalization. See §5 of Chapter 6.
2. E.g., Harmful to whom?, Breaks a promise to whom?, Deceives whom?, Consti-
tutes ingratitude to whom?, etc.
3. Harman (2011, 728; 2004, 104; 1999, 312, 316; 2003, 176, 179, 186, 193–4);
DeGrazia (2012, 28); McMahan (2013a, 276).
4. E.g., Buchanan (2009, 346, 360); Douglas (2013, 473); Agar (2013a, 72, 73;
2013b, 82, 83); Cohen (2007, 188).
5. This, I think, is how Cohen (2016) understands what he calls “moral standing”.
6. See, e.g., Sumner (1981), Beauchamp and Childress (2009, 64), Cohen (1986),
Fox (1978), DeGrazia (2012), Buchanan (2009), Vallentyne (2005, 423–32).
7. Part of the answer may lie in the fact that it is a version of Rossian Pluralism
(RP)—a view that, as we explored in Chapter 4, has a lot to be said for it in vir-
tue of its compatibility with Reasons Ground Oughts and Moral Rationalism.
In particular, just as moral reasons can be weighed against each other, thereby
providing a way of accounting for how there can be moral conflicts that are
not genuine moral dilemmas, degrees of moral status can perhaps be weighed
against each other. In any event, insofar as this explanatory picture is supposed
to be appealing for any of the reasons that RP was supposed to be appealing, I
refer the reader to my arguments against RP in Chapter 4.
8. That is, on the assumption that there really is a distinction between non-
normative and moral facts and that no amount of the former can ever ground
the latter. These assumptions gave been questioned, of course, but here I grant
them to my opponent for the sake of argument.
9. Goodpaster (1978), Steinbock (1992, 5), Rollin (2006, 34, 44); Bernstein (1998,
114).
10. I thank Collin O’Neill for alerting me to this objection.
11. For a similar assertion, see Harman (2003, 174)
12. I thank Lori Gruen for pushing me to consider this objection.
13. This objection would be made more defensible by substituting ‘when deciding
what to do’ for ‘when deciding what one ought to do’. Other individuals have
reasons to care about my practical reasoning that they lack when it comes to
my moral reasoning, since only the former kind of reasoning necessarily results
The Question of Scope, Part 2 125
in my acting, intending, willing, etc. To avoid confusion, however, I will keep
Regan’s wording. My later objections apply whatever the wording.
14. We would need a very expansive conception of ‘taking X into consideration’ in
order to avoid this problem. In particular, the needed conception of ‘taking X
into consideration’ will not require that I actually entertain the thought of X
when I deliberate. We would have to say that there are some acts of deliberation
we actually engage in that get us credit for acts of deliberation we don’t actually
engage in. For instance, perhaps at one point I thought in general terms about
what I may permissibly do with respect to my teacher, and perhaps this means
I get credit for thinking about her with respect to my sandwich ordering before
ordering my sandwich despite the fact that I don’t actually think of her before
ordering my sandwich. And what would ‘getting credit’ amount to in this case?
It could mean nothing other than that I am off the hook, morally speaking.
But why would I be off the hook? Obviously, I am off the hook because there
is no moral requirement to think of my teacher with respect to my sandwich
ordering before ordering my sandwich. And why would that be? Again, there
is only one available answer. There is no requirement because it is not possible
that this act of sandwich ordering will wrong my teacher. Thus, if we have a
duty to take everyone with moral status into consideration before making all of
our decisions, there is no way to determine exactly what goes into discharging
this obligation without first engaging in some first-order moral reasoning (e.g.,
reasoning about who might be wronged by my act of sandwich ordering). In
other words, we’d need to discharge our deliberative obligations in the course of
figuring out what they are! That being the case, we might as well stop worrying
about deliberative obligations.
15. See Raz (1990).
16. See Williams (1981).
8 Doing Harm and Failing
to Rescue
1. Introduction
My objective in this chapter is to determine whether two purported ground-
ing facts really are grounding facts. First, can the fact that an action harms
some individual make that action morally impermissible? I will argue in §2
that it can. Second, can the fact that an action constitutes a rescue make that
action morally obligatory? In §3 I conclude that it cannot. Finally, in §4, I
build off the conclusions of §2–3 to argue for an abolitionist position on
performing harmful research on animals.
2. Doing Harm
2.1 Introduction
Most cases of harm doing are intuitively wrong, but there is also a non-
negligibly large set of cases of harm doing that intuitively are not. This forces
us to consider the possibility that harm doing is not wrongness grounding.
The best way to motivate that idea would be to find something else, besides
their harmfulness, that most of the intuitively wrongful cases of harm doing
have in common—something not found in the intuitively permissible cases
of harm doing—and identify that something else as the wrongness ground-
ing feature of the actions. The alternative is to insist that the harmfulness of
an action can indeed make it wrong, and that the cases of harm doing that
are not wrong are cases in which enablers are absent or disablers are pres-
ent. I argue in this section for taking this second, more conservative, tack.
Intuitively, there are many differing factors to be cited, for the different
categories, as to why the harmful actions of that type aren’t wrong. In the
first case we’ll cite the presence of consent, in the second case it’s the fact
that the harm was visited upon oneself, while in the third case it’s the pres-
ence of compensation. In the fourth case we’ll avert to the fact that one
did not initiate the harmful sequence. The fifth case is more complicated;
some may think that the fact that it’s self-defense is enough explanation
on its own, while others would hold that the reason goes deeper—perhaps
something about how the harmed party has ceded her Hohfeldian immunity
to harm by making herself a threat to others. The sixth case is also (notori-
ously) controversial, as there is widespread disagreement over what justifies
inflicting harmful punishment, with deterrence, prevention, and retribution
constituting the most popular options.
The seventh case is complicated as well. The question here is why it is
morally permissible to harm someone by taking the job that he otherwise
would have gotten, buying the house for which he put in the next-highest
bid, getting married to the woman who would have taken him as a second
choice, etc. Obviously sometimes this is not permissible, for instance when
one succeeds in the competition by cheating or by doing something that’s
independently morally wrong, or when the competition itself is set up in
an unfair way. (These exceptions are noted by Feinberg (1984, 220) and
Haslett (2003, 640–1).) But when this is permissible, the best explanation as
to why is probably that in zero-sum games we have no choice but to inflict
harm (Haslett 2003, 640–1). There is no competing without at least one of
the competitors being harmed, and if we refrain from competing then we
collectively end up even worse-off.
The apparent diversity of explanations is quite important for our pur-
poses. This is because we have to determine the proper way to conceptualize
what we’ve learned when we discover that there are harmful actions that are
morally permissible because the harm was consented to, and also harmful
actions that are morally permissible because the harm was visited upon the
agent herself, and so on for each of these categories of action. Because of
the diversity of explanations it appears that we’re not going to be able to
pull off the trick I mentioned earlier: finding some feature, x, that’s lack-
ing in all or most of the purported cases of morally permissible harm, which
128 Doing Harm and Failing to Rescue
would then make it plausible to propose that it’s not actually the harmful-
ness of harmful actions that can ground their wrongness but rather their
possessing feature x.
Admittedly, pulling off this trick would be a legitimate prospect if con-
sequentialism were a possibility. If invoking rule-consequentialism, for
instance, were an option here, then one might argue that all seven categories
of harmful action are such that, plausibly, the value-maximizing set of rules
would permit them.
But in Chapter 3 we rejected consequentialism.2 Of course, the rejec-
tion was not based on a complete examination of all the implications of
accepting consequentialism as compared to all the implications of reject-
ing consequentialism, which means that the rejection should be consid-
ered provisional. So if we were simply at a loss here—if we didn’t know
what to say about these varying categories of apparently permissible harm-
infliction—then it would be entirely appropriate to put consequentialism
back on the table.
Positing the existence of these disablers jibes well with our intuitions. It
does strike us as intuitively plausible to suppose, for instance, that the fact
that a harmful action is executed in self-defense can disable what otherwise
would have been the wrongness of that action, and similarly for the other
posited disablers. So that is another mark in favor of taking the conserva-
tive route.
The only apparent drawback that I can identify is that one might want an
explanation of why each of the disablers can do its disabling work. But as
I argue in Appendix A, we should be willing to do without explanations of
this sort despite the reasonableness of the desire for them. Explanation has
to run out somewhere, and it’s no less plausible to chalk up as unexplain-
able the ability of some factor to do disabling work than it is to chalk up as
unexplainable any other work that gets done in the moral sphere.
With that objection disabled (no pun intended), my argument for identify-
ing harm doing as a wrongness grounding non-normative feature of actions
is complete. I now want to investigate the moral relevance of rescuing.
3. Rescuing
3.1 Introduction
Consider Peter Singer’s (1972) famous Pond Case: Jones is walking by a pond
when he hears a distressed shout. He turns his attention to the pond and
discovers that a small child has fallen in and is drowning. Saving the child
would be easy, and also cost-free except for the fact that Jones’s nice shoes
would get muddy and ruined by wading in to the pond.
Now suppose Jones declines to save the child, and the child dies. Our
intuitive belief about this case will be that Jones has violated an obligation.
So ideally the normative ethical theory on which we eventually settle will
deliver that verdict. The question I want to begin with is what grounding
explanation of that verdict we should hope our theory can offer.
One candidate explanation is that saving the child was obligatory on grounds
of its constituting the maximization of the good. This, however, is an act-
consequentialist explanation, and we already rejected act-consequentialism.
Of course, rejecting act-consequentialism doesn’t involve rejecting the idea
that for some wrong actions the explanation of their wrongness is their
failure to maximize the good. But using act-consequentialism’s explanatory
story even just for the purpose of the Pond Case and others like it is a very
unattractive move, as I pointed out in Chapter 3. Surely it’s counterintuitive
to suggest that the wrongness of Jones’s action is grounded in how it was
a failure to maximize the good. Similarly, the rule-consequentialist is put
in the uncomfortable position of suggesting that the wrongness of Jones’s
130 Doing Harm and Failing to Rescue
action is grounded in how it was a failure to abide by the optimal set of
rules.
We should look, then, to non-consequentialist explanations. And when
we look to the non-consequentialists for their view on this matter, what
they tend to do is sort this kind of case under the purported duty of rescue.
Assuming that they mean to invoke the duty of rescue not just as a valid
conditional or categorical connecting moral obligatoriness and rescue but
rather (or in addition) as a grounding claim, their proposal is that in the
Pond Case Jones has an obligation to save the child on grounds of that
action constituting a rescue. This proposal deserves close scrutiny.
5. Conclusion
In §3–4 we arrived at highly counterintuitive conclusions regarding the
Pond Case and harmful animal research, and by extension committed our-
selves to highly counterintuitive conclusions regarding innumerable other
cases that are similar in the relevant respects. At the end of the last chapter I
want to say something about how to proceed in light of this. For now, how-
ever, I simply want to argue that things might not be as bad as they seem. I
will argue with regard both to cases like the Pond Case (§5.1) and cases like
harmful animal research (§5.2) there are sometimes other ways to get to the
verdictive conclusions we want.
Notes
1. This last item requires a bit of explanation. The idea here is that there are cer-
tain goods the supply of which is limited, and so if one has an interest in obtain-
ing such a good that interest qualifies as a “competitive interest”—an interest
142 Doing Harm and Failing to Rescue
that can be advanced only at the expense of setting back another individual’s
similar interest (Feinberg 1984, 218–21).
2. Of course, what we rejected there was consequentialism as a monistic theory,
whereas all that’s being proposed here is to deploy a consequentialist explana-
tion of the wrongness of some wrongful actions, thus leaving open the possibil-
ity of deploying non-consequentialist explanations of the wrongness of other
wrongful actions. Note, however, that in Chapter 3 it was these very kinds of
case—harmful actions—for which we determined that consequentialism cannot
give a plausible explanation of their wrongness.
3. If the reader prefers a basis of comparison other than the two I entertain here,
that’s fine. The objection I’m going to raise against the supposed moral impor-
tance of harm prevention given a comparative account of harm holds regardless
of what the basis of comparison is.
4. This approach strikes me as so obviously right that it cannot be argued for. For
what it’s worth, Feinberg (1986, 149) and Parfit (1987, 374) agree with me.
5. One option would be to construct a theory on which the comparative and non-
comparative accounts of harm are each applicable to, and only to, distinct domains
of action. It’s worth recalling that the non-comparative account of harm was
developed originally by Shiffrin (1999) and Harman (2004; 2009) as a way of
accounting for the apparent possibility of harming an individual in causing her
to exist; it seems clear that the comparative account of harm can’t allow for this
possibility.
6. See, for instance, the absolutist positions on animal research adopted by Regan
(1983), Cohen (1986) and Frey (1980).
7. See, for instance, the non-absolutist positions on animal research adopted by
the consequentialists Singer (1995) and Frey (2011), and by McMahan (2005,
534–5), who is a consequentialist when it comes to our treatment of any indi-
vidual who falls below the ‘threshold of respect’, which includes all or almost
all animals.
8. This seems to be Nussbaum’s (2006, 403–5) and Slicer’s (1991, 121) view, and
they both do indeed use the word ‘tragedy’ to describe the situation.
9. See, for instance, Pogge (2005).
10. Of course, as I emphasized in Chapters 3–5, we should be concerned not just
with getting the intuitive verdict but with getting an intuitive explanation of
that verdict. Admittedly, Pogge’s argument doesn’t yield the conclusion that
acting so as to mitigate the poverty of those around is obligatory-on-grounds-
of-constituting-the-alleviation-of-suffering. So it doesn’t deliver a claim that we
would want to make. However, its explanation of the obligatoriness of acting
in this case—that it constitutes rectifying an institutional harm—is, at least, not
counterintuitive and is also not incompatible with the explanation we would
most want to offer (since it’s possible for an action to be obligatory on more
than one grounds).
9 The Distribution of Health
Care Resources
1. Introduction
In the previous chapter I illustrated how approaching moral theorizing as
an explanatory project would lead us to surprising substantive answers to
general and case-specific ethical questions. In this chapter my aim is to show
how approaching ethical theorizing as an explanatory project will lead us to
surprising conclusions regarding which case-specific questions have answers
that moral theorizing can reveal.
My case study will be the ethics of health care resource distribution, and
my conclusion will be that the ethical questions philosophers have been
asking and arguing about for decades now have either have no substantive
answer at all or no substantive answer of the sort that the tools of moral
theorizing are useful for discovering.1 An implication of this is that a good
chunk of what’s been published in applied ethics in the last 30 or so years
is a red herring.
To give the reader a feel for what my target is here, let me get specific. I
am talking about the literature that tries to work out the importance, for the
question of how health care resources should be distributed, of (among other
things) the number of people a particular resource is used to benefit, the age
of the individuals who need the resource, the extent to which the individuals
who need the resource can benefit from it, the current level of destitution of
the needy people, and the moral importance of giving the needy individu-
als at least a chance to receiving a needed resource. Among the prominent
contributors to this literature are Frances Kamm, Peter Singer, John Broome,
Dan Brock, Alex Voorhoeve, and Michael Otsuka.
Again, as was the case with my discussion of animal research in the pre-
vious chapter, I don’t want to seem as if I’m treating this matter casually. I
don’t take lightly the prospect of discarding as a red herring a large swath of
philosophical writings on any subject, never mind the distribution of health
care resources. For the record, I consider the public provision of health care
to be morally required (Sachs 2008b) and I’ve been interested in health
care resource distribution for longer than I’ve been interested in any other
philosophical issue. Furthermore, the philosophers working on this issue
144 The Distribution of Health Care Resources
are, needless to say, some of the smartest around and the answers they’ve
developed are some of the most interesting and nuanced to be found in any
field of philosophy. As interesting as the questions and proposed answers
are, however, I’ve arrived at the conclusion that there is nothing that could
ground the correctness or incorrectness of these answers. So once again the
problem is explanatory.
That there is nothing that could explain the truth of the proposed answers
follows in a somewhat straightforward fashion from the conclusions I
reached in Chapters 6 and 8. My aim in §2 is to show that this is the case.
Next, in §3, I concede for the sake of argument that I might have been
wrong in my conclusions from Chapter 6 or those from Chapter 8, but I
argue that, even so, there is another argument available to show that the
questions of health care resource distribution have either no answer at all or
no answer of the sort that can be derived through moral theorizing. Finally,
I review, in §4, the various twists and turns (of which there will be many) of
the arguments in §2–3.
Suppose two people each have a fatal disease. Each can be treated and
each would gain twenty years of life if she were treated. But one of
them is already disabled from another cause, so that treating her would
restore her to a life with disability. The other is healthy apart from
the particular disease, and treating her would restore her to a healthy
life. . . . It would be unfair to deny the first person her life on the
grounds that she is already suffering from disability. . . . If resources are
The Distribution of Health Care Resources 145
limited, the state would do more good treating the able-bodied person
than the disabled one. . . . How should we resolve this conflict between
fairness and doing the most good?
(Broome 1994, 37)
Suppose, for example, that we have a choice between saving A’s life and
saving B’s, and alongside B is C who has a sore throat. Our drug that
can save B’s life can also in addition cure C’s sore throat. [N.B.: The
question posed here is whether we should give the drug straightaway to
B & C or instead toss a coin.]
(Kamm 1996, 101)
The first horn is the claim that the class of goods to which the true distribu-
tive principle applies is some very large class of goods encompassing health
care resources and other goods besides. Suppose, for instance, that societal
resources per se should be distributed in an egalitarian manner. What can
we learn from this that would be relevant to the situation of an individual in
possession of a quantity of a certain kind of health care resource? Certainly
not that that quantity should be distributed in an egalitarian manner. To
make such an inference is to commit the fallacy of division—the fallacy of
making an inference from the fact that a class of things has a certain property
to the claim that each sub-class within that class has that property as well.8
Now one might respond that although the pursuit of a certain kind of
distribution of some large class of goods does not entail the pursuit of that
The Distribution of Health Care Resources 151
kind of distribution of any sub-class of goods within that class, the pursuit
of a certain kind of distribution of some large class of goods will necessarily
have some implications or other for how an individual should distribute the
limited sub-class of those goods at her disposal. So the idea would be that
the individual should do her bit. This of course is true, but totally unhelpful,
for we are in no position to judge what “her bit” is, when it comes to such
a monumental task as ensuring that the distribution of some large class of
goods, of which health care resources are just one sub-class, conforms to a
certain pattern.
Notes
1. Recall that in Chapter 1 I defined moral theorizing quite broadly as the project
of reasoning about theoretical and applied ethical questions all at once. So the
thesis, just mentioned, that I seek to advance in this chapter doesn’t follow trivi-
ally from some narrow understanding of what moral theorizing is.
2. There are many more thought experiments like this in Kamm (1996). There are
also thought experiments like this in Broome (1995; 1988), Parfit (2011, 353),
Rakowski (1994), and innumerable other articles and books.
3. There is another possibility, which is that these thought experiments are being
used not to answer some question in applied ethics but rather to advance our
thinking about distributive ethics—for instance the question of the relative
plausibility of egalitarian, prioritarian, and sufficientarian principles. This being
the case, there is a flawed presupposition behind these thought experiments,
namely that distributive generalizations can have moral force of their own. This
cannot be the case, as I argued in Chapter 6.
4. Of course, if we broaden our understanding of what a health care resource is,
then the current problem vanishes. In particular, if money counts as a health
care resource, on account of the fact that it can be used to purchase things that
qualify as health care resources on a narrower construal of that term, then indi-
vidual moral agents do in fact frequently find themselves in possession of surplus
health care resources. But the real-world circumstances that fit this description
are drastically different from the stipulated circumstances of the typical thought
experiments. For one thing, there is often great distance between the agent who
possesses the resource and the individuals who could benefit from it. This is
the case, for instance, for denizens of developed western nations and needy
people in sub-Saharan Africa. For denizens of developed western nations, the
choice they face is which charitable organization to give their money to, not to
156 The Distribution of Health Care Resources
which individual to give their money. Already this is a big difference between
the real world and the stipulations of the thought experiment. In any event, if
these thought experiments really are supposed to advance our thinking about
the ethics of poverty alleviation, then they are terribly misleading because it is
never said that that’s what the thought experiments are for, nor do the authors
of the writings in which these thought experiments appear generally make any
attempt to situate their thoughts within, or respond to, what has already been
said in the philosophical literature on the ethics of poverty alleviation.
5. In fairness to one of the major users of this kind of thought experiment, Frances
Kamm, it should be mentioned that she often notes that it makes a moral dif-
ference who is doing the distributing of the resource (1996, 144, 185; 1994).
But she nonetheless draws conclusions from such thought experiments without
specifying who is doing the distributing.
6. Non-moral features of actions can also, as initially mentioned in Chapter 2
and reinforced several times since then, have moral significance by constituting
enabling or disabling conditions. But we can set this possibility aside for now.
7. The reader might have noticed that I’m now speaking of ‘distributive principles’
as opposed to continuing to speak, as I did in Chapter 6, of ‘distributive gener-
alizations’. I think of principles as claims that state verdictive facts about broad
act-types and are derivable from generalizations. So, for instance, from the gen-
eralization, ‘under certain circumstances an act’s being a failure to distribute
welfare equally can ground its impermissibility’, one can derive the principle,
‘under certain circumstances failing to distribute welfare equally is impermis-
sible’. I’ll continue to focus my attention on distributive principles, as opposed
to the generalizations from which they’re derived, for ease of speaking.
8. The mistaken inference is made by Culyer (2001, 276) and Culyer and Wagstaff
(1993, 452). They begin by accepting the principle of equality of opportunity
to flourish and pointing out that health is a necessary condition for flourishing.
From this they infer that health should be distributed equally. But what actually
follows from their assumptions is that the necessary conditions for flourishing—
the set of them as a whole—should be distributed equally. Anand (2004) makes
the same mistake, except that Anand accepts a different version of equality of
opportunity.
I also suspect, but am not certain, that Ronald Dworkin makes this infer-
ence. Dworkin accepts the principle of equality of resources, and conceives of
health care as a resource. He offers an account of what it is for resources to be
divided equally (the hypothetical auction and the envy test), and then applies
this account directly to the distribution of health care (Dworkin 2000, esp.
Chapter 9) Thus, from his claim that resources should be distributed equally,
Dworkin appears to be inferring that health care should be distributed equally. But
of course there are other resources besides health care. (For further discussion
of this issue, see Jacobs (2004, 178–92).)
Finally, Shlomi Segall appears to make this inference. Segall believes that it
is unfair for one person to be made worse-off (opportunity-wise) than another
due to reasons beyond her control (Segall 2010, Chapter 1), and he infers that
this principle should regulate the distribution of some health care (ibid., 93).
But of course there are other opportunity-affecting goods aside from health
care. Segall seems to rescind this inference immediately after making it, so it is
difficult to determine exactly what his position is.
Larry Temkin (2011, 73–4) pointedly refuses to make such an inference. John
McKie et al. (1996) are also careful not to make this mistake.
9. This basic thought seems to be endorsed in the Universal Declaration of Human
Rights, Article 25 of which says that “everyone has the right to a standard
of living adequate for the health and well-being of himself and of his family,
The Distribution of Health Care Resources 157
including food, clothing, housing and medical care and necessary social ser-
vices”. It’s also endorsed by Powers and Faden (2006, esp. pp. 23, 61), Ram-
Tiktin (2012; 2017), Alvarez (2007), Ruger (2010), Daniels (2008), and Axelsen
and Nielsen (2017, 116–17). There’s now an entire edited volume on sufficien-
tarianism in health care, edited by Carina Fourie and Annette Rid, who them-
selves are sympathetic to the approach (Fourie and Rid 2017, 8).
10. Alvarez (2007), a sufficientarian when it comes to health care, concedes this
point.
11. Hirose (2017) constitutes another example of a version of sufficientarianism
that just is a priority rule. However, it’s a priority rule for ranking states of
affairs, not for ranking actions. Consequently, it is not a target for me since it
doesn’t attempt to pull off the feat that I am arguing a normative ethical theory
cannot accomplish—the feat of saying something helpful about what we ought
to do in cases of health resource scarcity.
12. Powers and Faden (2006) describe the problem of distributive justice in health
care as one of non-ideal justice and are concerned with problems of resource
scarcity, but they do not examine this question.
13. I use the ‘morally favored’ language as a way of keeping the discussion non-
committal enough to cover both kinds of distributive principles—those that tell
us which distribution(s) are (how) morally good and those that tell us which
distribution(s) we morally ought to bring about.
14. Sen (2006, 219–20) makes basically this same point, though he makes the mis-
take of not explicitly restricting his claim to non-comparative principles.
15. Fourie (2017, 196). This is just an instance of the more general truth that part
of what it means to achieve justice in non-ideal circumstances is to work toward
eliminating the non-ideal circumstances. I am not aware of anyone who denies
this claim, and several theorists explicitly affirm it. See, for instance Rawls
(1999b, 89), Simmons (2010), Sreenivasan (2007), Robeyns (2008), Dunn (1990),
and Shields (2017, 92).
16. Efrat Ram-Tiktin, a sufficientarian in health care who appeals to an auxiliary
principle, fails to acknowledge this. See Ram-Tiktin (2012).
17. For instance, suppose that some of the scarce resource one does possess could
be invested so as to yield a greater quantity of that kind of resource at a later
time, thereby guaranteeing that there will be no scarcity at that later time. In
this case the imperative of eliminating the scarcity would require investing some
of the resource while the auxiliary principle, whatever it is, may well require
immediately using all of it.
18. Granted, a great deal of theorizing about non-ideal justice has been done already,
but the two kinds of non-ideality with which that literature is concerned—
non-compliance and the absence of political institutions—are different from the
problem of scarcity, which is our concern here. (These two kinds of non-ideality
became the focus of theorizing on non-ideal justice because Rawls (1999a, 216)
specifically defined non-ideal justice as justice in the circumstance in which
either there is a lack of well-ordered institutions or there is widespread non-
compliance with the demands of those institutions.)
10 Conclusion
Having laid out a moral methodology in Chapters 1–2 and argued against
competing normative ethical theories in Chapters 3–4, I’ve spent the rest of
this book exploring a new family of theories called non-Rossian pluralism
(NP). I’ve been developing it piece-by-piece, and consequently at no point
have we had the opportunity to assess its plausibility as a whole. Of course,
even now we’re not really in a good position to do so, since there are many
topics on which a normative ethical theory has to take a stand that we as
yet haven’t inquired as to what NP will say about. What we can do, and
what I will do in this brief concluding chapter, is to examine how it’s doing
so far, in light of the topics on which we know NP’s stand. Admittedly, when
I say we “know” NP’s stand on a certain issue I mean only that I’ve made
an argument in this book that it should take some particular stand, and of
course any of those arguments could be flawed. For now, though, assessing
my admittedly incomplete version of NP is all that can be done by way of
assessing the prospects for the theory. So that’s what I’ll do.
I’ll start with the bad news for NP, which is that the verdictive implica-
tions I extracted from NP in Chapters 8 (except §2) and 9 were extremely
counterintuitive.1 To begin with Chapter 9: Some of the thought experi-
ments commonly used in the literature on health care resource distribution
have apparently obvious answers. For instance, the lifeboat case in which
one can bring one’s lifeboat to an island that contains a 20-year-old in good
health who will live another 60 good years if saved or an island that con-
tains an 80-year-old in permanently poor health who will live just one more
year if saved, but one cannot do both, yields the apparently clear conclu-
sion that one is morally required to save the 20-year-old. Other thought
experiments are more difficult: For instance, if you have two doses of a pill,
should you give them to the person who needs both doses in order to not
die, or should you give one dose to each of two people who will lose the use
of their legs if they don’t receive a dose? Although people would probably
have contrasting intuitions about this case, it’s a safe bet that nearly every-
one would have what we might call the ‘meta-intuition’ that we can expect
the correct normative ethical theory to instruct us to do one or the other.
Yet, I showed in Chapter 9 that NP implies that all of these intuitions are
Conclusion 159
mistaken. Regarding the scenarios like the ones just given, NP will either tell
us that there’s nothing at all that the agent should do or that the situation
is a genuine moral dilemma, or it will tell us to look somewhere else for an
answer.
This simply compounded the problem we confronted in Chapter 8, where,
again, I argued from NP to some highly counterintuitive conclusions. In par-
ticular, I argued that if we could develop a medical intervention that would
save many, many people from a serious disease, but only at the cost of causing
harm to a few animals who would have to be used as research subjects in
the course of developing the intervention, it would be wrong to do so. This is
just representative of a potentially infinite set of actions, of the type prevent-
something-very-bad-from-happening-to-some-individuals-at-the-cost-of-
causing-mild-harm-to-some-other-individual(s), that will intuitively strike us
as permissible (maybe even obligatory, especially if the other individual is one-
self) but that NP will condemn as impermissible.
This double blow from Chapters 8–9 can be softened a bit, as I tried to
show earlier. In Chapter 9 I argued that insofar as health care distribution
thought experiments are constructed in a realistic way, it is likely that the true
political morality will yield answers to them. And in §5.1–2 of Chapter 8, I
tried to show that accepting the main conclusion of that Chapter—i.e., that
there is no duty of rescue—is compatible with it being morally required to
make significant sacrifices for the sake of impoverished people elsewhere in
the world. Still, in both cases highly counterintuitive results remain.
Actually, “highly counterintuitive” is an understatement. I assume we would
want to take it as a fixed point, for purposes of reflective equilibrium, that
saving the child in the Pond Case is obligatory, that if we can save a million
humans by doing mildly painful research on a few mice that we ought to do
it, and that morality requires us to save the 20-year-old in the lifeboat case
mentioned earlier. If we had started from these fixed points in our effort to
identify the most plausible family of normative ethical theories, we wouldn’t
have given NP a second thought.
So it would be helpful to recall how we got to this point. Perhaps the most
obvious way for a non-consequentialist to get what she wants in the Pond
Case, the case of harmful animal research, and lifeboat cases is to avert to
the balance of moral reasons. If we had this tool at our disposal, then it
would be easy to vindicate our verdictive intuitions about those cases. But
in Chapter 4 we decided to deprive ourselves of this tool. Of course, we had
a good justification for doing so: Grounding explanations that invoke moral
reasons are poor explanations. So what’s become apparent now is that the
price we pay for insisting on good explanations is being able to explain less
of what we want to explain.
Still, however, it’s not clear that we should pay that price. What I call
‘insisting on good explanations’ is just another case of relying on intuitions—
explanatory intuitions, specifically. Furthermore, as I noted at the end of
Chapter 5, NP puts us in a tight spot with regard to moral dilemmas: Since
160 Conclusion
it implies that every moral conflict is a genuine moral dilemma, it implies,
on the assumption that there are a lot of moral conflicts, that there are lots
of moral dilemmas.2
On the other side of the ledger, it bears mentioning that the explanatory
intuitions that guided us toward NP were themselves fixed points; the judg-
ments on which I drew in discarding balancing versions of moral pluralism,
as well as the intuitions I used in Chapter 3 in arguing against the most
popular versions of moral monism, were very strong ones. Moreover, NP
sets us a deliberative task that we have the tools to carry out, while Rossian
Pluralism fails to do so. Furthermore, NP can allow that acts of cruelty to
animals wrong those animals; no other normative ethical theory surveyed
here gets that result.
There is no way to make a serious attempt at totting up the total counter-
intuitiveness of NP and comparing it to the total counterintuitiveness of its
competitors; that would require figuring out how to individuate the (pos-
sibly infinite) cases on which we already know what the various theories
will say and coming up with an interval measure of counterintuitiveness. So
instead I will oversimplify drastically and assess the relative acceptability of
NP and it competitors in their oversimplified situations.
First, some of the competitors should simply be discarded. For instance,
I argued in Chapter 3 that act-consequentialism, Kantianism, and contrac-
tarianism each deliver terribly counterintuitive explanatory implications as
stand-alone normative ethical theories. Furthermore, in their original forms
they also each deliver terribly counterintuitive verdictive implications. Act-
consequentialism, for instance, delivers terribly counterintuitive implications
in the famous Transplant Case (and cases like it) and in cases concerning
giving to charity. Kantianism, meanwhile, delivers terribly counterintuitive
implications about masturbation, dishonesty, and suicide. And contracta-
rianism delivers drastically counterintuitive implications about our moral
obligations to disabled humans. These theories have little to recommend
them. However, there are reconstructed versions of each these theories that
can deliver intuitively plausible verdictive implications; these should stay in
the picture. Rossian Pluralism must be given consideration as well, because
it too can deliver intuitively plausible verdictive implications.
My oversimplification of the respective situations of the various theories
is this: NP has intuitively plausible explanatory implications, while some
of its competitors, including the ones mentioned just now, have equally
intuitively plausible verdictive implications. And NP has terribly counter-
intuitive verdictive implications while some of its competitors, including
the ones mentioned just now, have equally counterintuitive explanatory
implications.
Broadly speaking, then, my impressionistic representation of the conflict
between NP and its competitors is that they each have one very important
virtue and one very important defect.3 This, one might think, leaves us at a
stalemate.
Conclusion 161
One thing I expect that my readers will want to say about this is that this
is no stalemate because it is much more appealing to give up the explana-
tory intuitions with which the competitors to NP were shown, in Chapters
3–4, to be inconsistent than to give up the verdictive intuitions with which
NP was shown, in Chapters 8–9, to be inconsistent. This seems to tilt the
balance against NP. In response, I suggest we take care not to confuse the
phenomenology of finding a claim intuitively plausible with the phenom-
enology of wanting to believe a claim. For instance, I find myself strongly
drawn toward believing that 9/11 was not an inside job. On reflection, I
can distinguish at least two sources of this feeling. One is just the intuitive
plausibility of the claim. But another is that I just want to believe it; I would
feel like a crazy person if I didn’t. But the latter is the wrong kind of reason
for a belief. Similarly, with regard to the Pond Case and others like it, for
many of us there are probably two ways in which we are drawn to believe
that failing to save the child is wrong: The claim is intuitively plausible and
we would feel like bad people if we denied it.4 Again, however, the latter is
the wrong kind of reason to believe it. Likewise, we should be careful not to
confuse the phenomenology of finding a claim intuitively plausible with the
phenomenology of wanting a claim to be true. If we think that some people
are sometimes motivated to do the right thing because it’s right, and if we
think that true moral claims are (or will in the future be) more likely to
be believed than their denials, then we must conclude that a moral claim’s
being true makes it (or will in the future) make it more likely that the action
it recommends will be done or that the action it forbids will not be done.
This makes it rational to care which moral claims are true. It makes it ratio-
nal, for instance, to want it to be true that it would be permissible to save
a million innocent people from death by means of breaking the leg of some
other innocent person. (One is much less likely to be the one than to be one
of the million.) But, once again, wanting something to be true is the wrong
kind of reason to believe it. So while we might feel more reluctant to give
up our verdictive intuitions in the Pond Case, the save-a-million lives case,
and similar scenarios than we are to give up our strongly held explanatory
intuitions, reluctance brings in too much that’s irrelevant; it brings in the
wrong kinds of reason for belief.5 If we concentrate our attention exclu-
sively on how plausible the relevant judgments seem, I contend that we
won’t find a difference.
Thus, we return to the idea that we are at a stalemate. I contend, however,
that we are not. Although we cannot, based on my earlier oversimplified
summary of their respective situations, choose between NP and its com-
petitors based on which does the least violence to our intuitions, we can
choose between them based on the extent to which it is rational for us to
trust the intuitions to which they do violence. There are, as I discussed in
§3 of Chapter 1, two arguments in the philosophical literature purporting
to show that our moral intuitions are not particularly reliable guides to the
moral facts, assuming there are such facts: Greene’s empirical argument and
162 Conclusion
Street’s evolutionary skepticism argument. And I argued that both of those
arguments are at their strongest when targeted specifically against our ver-
dictive intuitions. This makes it rational for us to trust those intuitions less
than our explanatory intuitions.
That conclusion, however, comes along with some enormously important
caveats. First, it’s far from a settled matter whether either of those arguments,
never mind both, have any force at all. Second, the fact that Greene’s empiri-
cal argument has less force against explanatory intuitions may be entirely
down to the fact that no one has as yet attempted to conduct the relevant
kinds of studies on such intuitions; it may be that such studies, were they
to be conducted, would demonstrate that explanatory intuitions are suspect
in just the same way as Greene argues our verdictive intuitions are. Third,
it’s certainly possible that there are other arguments out there to be made—
arguments targeted specifically against the claim that our explanatory intu-
itions are reliable. It might be that the only reason that philosophers have
developed two arguments targeted against verdictive intuitions and no argu-
ments targeted against explanatory intuitions is that explanatory intuitions
have been flying under the radar in virtue of the fact that it hasn’t been
widely acknowledged that they constitute an indispensible input to moral
theorizing.
My conclusion, therefore, must be the following extremely weak one.
Provisionally—for want of a more rigorous appraisal of the strengths and
weaknesses of NP and its competitors as opposed to my incomplete and
impressionistic assessment, and failing a resolution of the debate over the
validity of Greene’s and Street’s arguments, and in the absence of any serious
investigation of the reliability of our explanatory intuitions—accepting NP
is at least as rational as, and maybe more rational than, accepting any one
of its competitors.
Notes
1. I would add, though, that I chose the topics of Chapters 8 and 9 because NP deliv-
ers counterintuitive implications in those domains. (It would have been a much
less interesting task exploring what NP has to say on questions about which it
delivers the verdicts we were already inclined to accept!) So NP might do better
in many domains than it does in the ones I explored.
2. This worry, however, turns out to be less serious than it seemed at the end of
Chapter 5. In Chapters 6–9 I argued for removing certain non-normative facts,
including some very widespread ones—especially the fact that some action would
confer a benefit on someone—from our list of requirement-grounding facts. Hav-
ing done this, we can conclude that moral conflict is a much less frequent phe-
nomenon than we would have initially been inclined to believe.
3. Jeff McMahan (2013b, 34) has argued, on entirely separate grounds, for this same
kind of conclusion—i.e., that it’s unlikely we’ll be able to come up with a norma-
tive ethical theory that doesn’t have drastically counterintuitive implications.
4. This worry is not reasonable when the concern is that one will adopt a mistaken
belief in the opposite direction. No one could plausibly be labeled a bad person
for mistakenly believing that she is morally required to φ. (Cases in which φ-ing
Conclusion 163
is evil do not constitute counterexamples to this generalization; if φ-ing is evil
then what might make you a bad person isn’t believing that φ-ing is morally
required but rather believing that φ-ing is morally permitted.) So insofar as we
allow these considerations to affect our belief-forming practices, we have a bias
toward adopting an overly strict morality.
5. This, I contend, is another good explanation of the phenomenon for which I
already offered one explanation (Chapter 1, §6), namely the phenomenon of
moral philosophers accepting the validity of the method of reflective equilibrium
but frequently violating it anyway (specifically, by being more attached to their
verdictive intuitions than their explanatory intuitions).
Appendix A
Moral Meta-Explanation
Table A.1
Notes
1. Actually, Väyrynen speaks of reason-explaining facts, but he certainly would be
willing to extend what he says about reason-explaining facts, mutatis mutandis,
to verdict-explaining facts.
2. Kagan (1988, 20–1) believes this too. He calls the explainers of the explainers the
“foundations” of normative ethics.
3. This doesn’t amount to an endorsement of normative foundationalism. Norma-
tive foundationalism is a claim in epistemology, whereas I’m talking about nor-
mative metaphysics.
4. In another work (2006, 719–20), Väyrynen offers a third argument—an argument
by analogy. He notes that in the case of epistemic verdicts—like the verdict that
one ought to believe X—we think that explainers aren’t bedrock. We tend to
think that there’s always an explanation as to why some fact explains such a
verdictive fact. So, for instance, if the fact that the dog is a whippet explains the
fact that I ought to believe that the dog is fast, this is presumably in virtue of how
the fact that the dog is a whippet raises the probability of the dog’s being fast, or
something like that.
As Väyrynen admits, the analogy is worrisome for someone like me only if
epistemic reasons and moral reasons share a nature. The idea that they do is, as
Väyrynen notes, widely accepted. But I’m skeptical. I don’t even think moral rea-
sons share a nature with other reasons for action. And my position on this isn’t
motivated by a desire to resist Väyrynen’s argument here. It’s motivated, instead,
by a desire to be able to give good explanations of both the moral verdicts about
actions and the other practical verdicts about actions; I give this argument in
Chapter 5.
5. Suppose, for instance, that I lie to you about where I’ve been for the past three
hours. If I lied to you because of the embarrassment that telling the truth would
have caused me—say, for instance, I was supposed to be taking our kids to the
park but instead was playing video games—I suspect we’ll all have the intuition
that there is nothing here to undermine the impermissibility-grounding power of
the fact that my utterance was a lie. If, on the other hand, I lied to you so as to
not spoil the surprise birthday party that I’m throwing for you—say, for instance,
I spent the last three hours checking out potential venues—I suspect we’ll all have
the intuition that the impermissibility-grounding power of the fact that my utter-
ance was a lie is indeed undermined in this case.
Works Cited
Agar, Nicholas. ‘Why Is It Possible to Enhance Moral Status and Why Doing So Is
Wrong?’. Journal of Medical Ethics 39 (2013a): 67–74.
———. ‘Still Afraid of Needy Post-Persons’. Journal of Medical Ethics 39 (2013b):
81–3.
Albertsen, Andreas and Knight, Carl. ‘A Framework for Luck Egalitarianism in
Health and Health Care’. Journal of Medical Ethics 14 (2015): 165–9.
Alvarez, Allen Andrew A. ‘Threshold Considerations in Fair Allocation of Health
Resources: Justice Beyond Scarcity’. Bioethics 21 (2007): 426–38.
Anand, Sudhir. ‘The Concern for Equity in Health’. In Public Health, Ethics, and
Equity, edited by Sudhir Anand, Fabienne Peter and Amartya Sen, 15–20. New
York: Oxford University Press, 2004.
Arneson, Richard. ‘Equality and Equal Opportunity for Welfare’. Philosophical Stud-
ies 56 (1989): 77–93.
———. ‘What, If Anything, Renders All Humans Morally Equal?’. In Singer and His
Critics, edited by Dale Jamieson, 103–27. Oxford: Blackwell, 1999.
———. ‘Rawls Versus Utilitarianism in the Light of Political Liberalism’. In The
Idea of a Political Liberalism, edited by Victoria Davion and Clark Wolf, 231–52.
Lanham, MD: Rowman and Littlefield, 2000a.
———. ‘Luck Egalitarianism and Prioritarianism’. Ethics 110 (2000b): 339–49.
Audi, Robert. ‘Intuition and Its Place in Ethics’. Journal of the American Philosophi-
cal Association 1 (2015): 57–77.
Axelsen, David V. and Nielsen, Lasse. ‘Essentially Enough: Elements of a Plausi-
ble Account of Sufficientarianism’. In What Is Enough? Sufficiency, Justice, and
Health, edited by Carina Fourie and Annette Rid, 101–18. New York: Oxford
University Press, 2017.
Bader, Ralf. ‘Conditions, Modifiers, and Holism’. In Weighing Reasons, edited by
Errol Lord and Barry Maguire, 27–55. New York: Oxford University Press, 2016.
Baier, Kurt. The Moral Point of View. Ithaca, NY: Cornell University Press, 1958.
Baron, Marcia. ‘Kantian Ethics’. In Three Methods of Ethics, edited by Marcia W.
Baron, Philip Pettit and Michael Slote, 3–91. Malden, MA: Blackwell, 1997.
———. ‘The Ticking Bomb Hypothetical’. In Disputed Moral Issues, 3rd ed., edited
by Mark Timmons, 628–38. New York: Oxford University Press, 2013.
Barry, Brian. Theories of Justice. Hemel Hempstead: Harvester-Wheatsheaf, 1989.
Beauchamp, Tom L. and Childress, James F. Principles of Biomedical Ethics, 6th ed.
New York: Oxford University Press, 2009.
Works Cited 169
Bernstein, Mark H. On Moral Considerability. New York: Oxford University Press,
1998.
Bräanmark, Johan. ‘Moral Disunitarianism’. Philosophical Quarterly 66 (2016):
481–99.
Brink, David. ‘Moral Conflict and Its Structure’. Philosophical Review 103 (1994):
215–24.
Broad, C.D. Five Types of Ethical Theory. London: Kegan Paul, Trench, Trubner &
Co, 1930.
Brock, Dan. ‘Separate Spheres and Indirect Benefits’. Cost-Effectiveness and Resource
Allocation 1 (2003): 4.
Broome, John. ‘Good, Fairness and QALYs’. In Philosophy and Medical Welfare,
edited by JM. Bell and Susan Mendus, 57–73. Cambridge, MA: Cambridge Uni-
versity Press, 1988.
———. ‘Fairness Versus Doing the Most Good’. Hastings Center Report 2 (1994):
36–9.
———. ‘Fairness in the Rationing of Health Care’. In Market Capitalism and Moral
Values, edited by Samuel Brittan and Alan Hamlin, 79–87. n.p.: Edward Elgar, 1995.
———. ‘Normative Requirements’. Ratio 12 (1999): 398–419.
———. ‘Reasons’. In Reason and Value: Themes From the Moral Philosophy of
Joseph Raz, edited by R. Jay Wallace, Philip Pettit, Samuel Scheffler and Michael
Smith, 28–55. Oxford: Clarendon Press, 2004.
———. Rationality Through Reasoning. Malden, MA: Wiley Blackwell, 2013.
Buchanan, Allen. ‘Moral Status and Enhancement’. Philosophy & Public Affairs 37
(2009): 346–81.
Chappell, Richard. ‘Value Receptacles’. Noûs 49 (2015): 322–32.
Clune, Alan C. ‘Biomedical Testing on Animals: An Attempt at a “Rapprochement”
Between Utilitarianism and Theories of Inherent Value’. The Monist 79 (1996):
230–46.
Cohen, Andrew. ‘Contractarianism, Other-regarding Attitudes, and the Moral Stand-
ing of Nonhuman Animals’. Journal of Applied Philosophy 24 (2007): 188–201.
———. ‘Contractarianism and Moral Standing Inegalitarianism’. Dialogue 55
(2016): 639–58.
Cohen, Carl. ‘The Case for the Use of Animals in Biomedical Research’. New Eng-
land Journal of Medicine 315 (1986): 865–70.
Cohen, G.A. ‘On the Currency of Egalitarian Justice’. Ethics 99 (1989): 906–44.
———. Rescuing Justice and Equality. Cambridge, MA: Harvard University Press,
2008.
Conee, Earl. ‘Against Moral Dilemmas’. In Moral Dilemmas, edited by Christopher
W. Gowans, 239–49. New York: Oxford University Press, 1987.
Cornell, Nicolas. ‘Wrongs, Rights, and Third Parties’. Philosophy & Public Affairs
43 (2015): 109–43.
Crisp, Roger. ‘Particularizing Particularism’. In Moral Particularism, edited by Brad
Hooker and Margaret Little, 23–47. New York: Oxford University Press, 2000.
———. ‘Equality, Priority, and Compassion’. Ethics 113 (2003): 745–63.
Culyer, Anthony J. ‘Equity—Some Theory and Its Policy Implications’. Journal of
Medical Ethics 27 (2001): 275–83.
Culyer, Anthony J. and Wagstaff, Adam. ‘Equity and Equality in Health and Health
Care’. Journal of Health Economics 12 (1993): 431–57.
170 Works Cited
Dahl, Norman O. ‘Morality, Moral Dilemmas, and Moral Requirements’. In Moral
Dilemmas and Moral Theory, edited by H.E. Mason, 86–101. New York: Oxford
University Press, 1996.
Dancy, Jonathan. Ethics Without Principles. Oxford: Clarendon Press, 2004.
Daniels, Norman. ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics’.
The Journal of Philosophy 76 (1979): 256–82.
———. Just Health: Meeting Health Needs Fairly. Cambridge: Cambridge Univer-
sity Press, 2008.
Darwall, Stephen. The Second-Person Standpoint. Cambridge, MA: The Harvard
University Press, 2006.
———. ‘But It Would Be Wrong’. Social Philosophy and Policy 27 (2010): 135–57.
DeGrazia, David. Taking Animals Seriously. Cambridge: Cambridge University Press,
1996.
———. Creation Ethics. New York: Oxford University Press, 2012.
Douglas, Tom. ‘Human Enhancement and Supra-Personal Moral Status’. Philosophi-
cal Studies 162 (2013): 473–97.
Driver, Julia. ‘Autonomy and the Asymmetry Problem for Moral Expertise’. Philo-
sophical Studies 128 (2006): 619–44.
Dunn, J. Interpreting Political Responsibility. Cambridge: Cambridge University
Press, 1990.
Dworkin, Ronald. ‘The Original Position’. In Reading Rawls, edited by Norman
Daniels, 16–53. Stanford, CA: Stanford University Press, 1989.
———. Sovereign Virtue. Cambridge, MA: Harvard University Press, 2000.
Feinberg, Joel. Harm to Others. New York: Oxford University Press, 1984.
———. ‘Wrongful Life and the Counterfactual Element in Harming’. Social Philoso-
phy & Policy 4 (1986): 145–78.
Feldman, Fred. Utilitarianism, Hedonism, and Desert. Cambridge: Cambridge Uni-
versity Press, 1997.
Fogal, Daniel. ‘Reasons, Reason, and Context’. In Weighing Reasons, edited by Errol
Lord and Barry Maguire, 74–103. New York: Oxford University Press, 2016.
Fourie, Carina. ‘Sufficiency of Capabilites, Social Equality, and Two-Tiered Health
Care Systems’. In What Is Enough? Sufficiency, Justice, and Health, edited by
Carina Fourie and Annette Rid, 185–204. New York: Oxford University Press,
2017.
Fourie, Carina and Rid, Annette. ‘Introduction’. In What Is Enough? Sufficiency,
Justice, and Health, edited by Carina Fourie and Annette Rid, 1–8. New York:
Oxford University Press, 2017.
Fox, Michael. ‘ “Animal Liberation”: A Critique’. Ethics 88 (1978): 106–18.
Frey, R.G. Interests and Rights: The Case Against Animals. New York: Oxford Uni-
versity Press, 1980.
———. ‘Animals’. In The Oxford Handbook of Practical Ethics, edited by Hugh
LaFollette, 161–87. Oxford: Oxford University Press, 2003.
———. ‘Utilitarianism and Animals’. In The Oxford Handbook of Animal Ethics,
edited by Tom L. Beauchamp and R.G. Frey, 172–97. New York: Oxford Univer-
sity Press, 2011.
Gaut, Berys. ‘Moral Pluralism’. Philosophical Papers XXIII (1994): 17–40.
Gert, Joshua. ‘A Functional Role Analysis of Reasons’. Philosophical Studies 124
(2005): 353–78.
Works Cited 171
Goodpaster, Kenneth. ‘On Being Morally Considerable’. Journal of Philosophy 75
(1978): 308–35.
Gosseries, Axel. ‘Intergenerational Justice, Sufficiency, and Health’. In What Is
Enough? Sufficiency, Justice, and Health, edited by Carina Fourie and Annette
Rid, 121–43. New York: Oxford University Press, 2017.
Greene, Joshua. ‘Beyond Point-and-Shoot Morality: Why Cognitive (Neuro)Science
Matters for Ethics’. Ethics 124 (2013): 695–726.
Hare, R.M. Moral Thinking. Oxford: Clarendon Press, 1981.
———. ‘Rights, Utility, and Universalization: Reply to J.L. Mackie’. In Utility and
Rights, edited by R.G. Frey, 106–20. Minneapolis: University of Minnesota Press,
1984.
Harman, Elizabeth. ‘Creation Ethics: The Moral Status of Early Fetuses and the Eth-
ics of Abortion’. Philosophy and Public Affairs 28 (1999): 310–24.
———. ‘The Potentiality Problem’. Philosophical Studies 114 (2003): 173–98.
———. ‘Can We Harm and Benefit in Creating?’. Philosophical Perspectives 18 (2004):
89–113.
———. ‘Harming as Causing Harm’. In Harming Future Persons: Ethics, Genetics,
and the Nonidentity Problem, edited by Melinda A. Roberts and David T. Wasser-
man, 137–54. Dordrecht: Springer, 2009.
———. ‘The Moral Significance of Animal Pain and Animal Death’. In The Oxford
Handbook of Animal Ethics, edited by Tom L. Beauchamp and R.G. Frey, 726–37.
New York: Oxford University Press, 2011.
Hart, H.L.A. The Concept of Law, 3rd ed. Oxford: Oxford University Press, 2012.
Haslett, D.W. ‘Murder and the Exception for Fair Competition’. Social Theory and
Practice 29 (2003): 631–54.
Hempel, Carl G. and Oppenheim, Paul. ‘Studies in the Logic of Explanation’. Phi-
losophy of Science 15 (1948): 135–75.
Herman, Barbara. ‘Kantian Commitments’. Studi Kantiani XXIX (2016): 131–43.
Hershovitz, Scott. ‘The End of Jurisprudence’. The Yale Law Journal 124 (2015):
1160–204.
Hill, Thomas E. ‘Moral Dilemmas, Gaps, and Residues’. In Moral Dilemmas and
Moral Theory, edited by H.E. Mason, 167–98. New York: Oxford University
Press, 1996.
Hinman, Lawrence M. Ethics: A Pluralistic Approach to Moral Theory. Belmont,
CA: Thomson Wadsworth, 2008.
Hirose, Iwao. ‘Axiological Sufficientarianism’. In What Is Enough? Sufficiency, Jus-
tice, and Health, edited by Carina Fourie and Annette Rid, 51–68. New York:
Oxford University Press, 2017.
Holtug, Nils. ‘Equality for Animals’. In New Waves in Applied Ethics, edited by Jes-
per Ryberg, Thomas S. Petersen and Claerk Wolf, 1–24. Basingstoke, Hampshire:
Palgrave Macmillan, 2007.
Hooker, Brad. Ideal Code, Real World. New York: Oxford University Press, 2000.
———. ‘Intuitions and Moral Theorizing’. In Ethical Intuitionism: Re-evaluations,
edited by Philip Stratton-Lake, 161–83. Oxford: Clarendon Press, 2002.
Huemer, Michael. Ethical Intuitionism. Basingstoke, Hampshire: Palgrave Macmil-
lan, 2005.
———. ‘When Is Parsimony a Virtue?’. The Philosophical Quarterly 59 (2009): 216–36.
Hursthouse, Rosalind. On Virtue Ethics. New York: Oxford University Press, 1999.
172 Works Cited
Huseby, Robert. ‘Sufficiency, Priority, and Aggregation’. In What Is Enough? Suffi-
ciency, Justice, and Health, edited by Carina Fourie and Annette Rid, 69–84. New
York: Oxford University Press, 2017.
Jacobs, Lesley A. Pursuing Equal Opportunities. Cambridge: Cambridge University
Press, 2004.
Kagan, Shelly. Normative Ethics. Boulder, CO: Westview Press, 1988.
———. The Limits of Morality. New York: Oxford University Press, 1989.
Kamm, Frances. ‘To Whom?’. Hasings Center Report 24 (1994): 29–32.
———. Morality, Mortality. Vol. 1. Oxford: Oxford University Press, 1996.
———. Intricate Ethics. New York: Oxford University Press, 2007.
Kant, Immanuel. The Metaphysics of Morals, translated by Mary Gregor. Cam-
bridge: Cambridge University Press, 1996.
———. Groundwork of the Metaphysics of Morals, translated by Mary Gregor and
edited by Jens Timmermann. Cambridge: Cambridge University Press, 2011.
Kearns, Stephen and Star, Daniel. ‘Reasons as Evidence’. In Oxford Studies in Meta-
ethics. Vol. 4, edited by Russ Shafer-Landau, 215–42. New York: Oxford Univer-
sity Press, 2009.
Korsgaard, Christine. ‘Kant’s Formula of Humanity’. In Creating the Kingdom of
Ends, edited by Christine Korsgaard, 106–32. Cambridge: Cambridge University
Press, 1996.
———. ‘Fellow Creatures: Kantian Ethics and Our Duties to Animals’. In The Tan-
ner Lectures on Human Values. Vol. 25, edited by Grethe B. Peterson, 77–110. Salt
Lake City: The University of Utah Press, 2005.
———. Self-Constitution: Agency, Identity, and Integrity. New York: Oxford Uni-
versity Press, 2009.
———. ‘Interacting With Animals: A Kantian Account’. In The Oxford Handbook
of Animal Ethics, edited by Tom L. Beauchamp and R.G. Frey, 91–118. New York:
Oxford University Press, 2011.
———. ‘A Kantian Case for Animal Rights’. In Animal Law: Tier und Recht, edited
by Julia Hänni, Margot Michel and Daniela Kühne, 3–27. Zurich: Dike Verlag,
2012.
———. ‘Kantian Ethics, Animals, and the Law’. Oxford Journal of Legal Studies 33
(2013): 629–48.
Kymlicka, Will. Liberalism, Community and Culture. New York: Oxford University
Press, 1989.
———. Contemporary Political Philosophy: An Introduction. Oxford: Oxford Uni-
versity Press, 2002.
Ladyman, James. ‘Science, Metaphysics, and Method’. Philosophical Studies 160
(2012): 31–51.
LaFollette, Hugh. ‘Animal Experimentation in Biomedical Research’. In The Oxford
Handbook of Animal Ethics, edited by Tom L. Beauchamp and R.G. Frey, 796–
825. New York: Oxford University Press, 2011.
Leary, Stephanie. ‘Non-naturalism and Normative Necessities’. In Oxford Studies in
Metaethics. Vol. 12, edited by Russ Shafer-Landau, 76–105. New York: Oxford
University Press, 2017.
Liberman, Alida and Schroeder, Mark. ‘Commitment: Worth the Weight’. In Weigh-
ing Reasons, edited by Errol Lord and Barry Maguire, 104–20. New York: Oxford
University Press, 2016.
Works Cited 173
Lichtenberg, Judith. ‘The Right, the All Right, and the Good’. Yale Law Journal 92
(1983): 544–63.
List, Christian and Valentini, Laura. ‘The Methodology of Political Theory’. In The
Oxford Handbook of Philosophical Methodology, edited by Herman Cappelen,
Tamar Szabó Gendler and John Hawthorne, 525–53. Oxford: Oxford University
Press, 2016.
Little, Margaret Olivia. ‘On Knowing the “Why”: Particularism and Moral Theory’.
Hastings Center Report 31 (2001): 32–40.
Maguire, Barry. ‘Grounding the Autonomy of Ethics’. In Oxford Studies in Metaeth-
ics. Vol. 10, edited by Russ Shafer-Landau, 188–215. New York: Oxford Univer-
sity Press, 2015.
Manne, Kate. ‘Democratizing Humeanism’. In Weighing Reasons, edited by Errol
Lord and Barry Maguire, 123–40. New York: Oxford University Press, 2016.
Marmot, Michael. ‘Social Causes of Social Inequalities in Health’. In Public Health,
Ethics, and Equity, edited by Sudhir Anand, Fabienne Peter and Amartya Sen,
37–61. New York: Oxford University Press, 2004.
McConnell, Terrance. ‘Moral Residue and Dilemmas’. In Moral Dilemmas and Moral
Theory, edited by H.E. Mason, 36–47. New York: Oxford University Press, 1996.
McDermott, Daniel. ‘Analytic Political Philosophy’. In Political Theory: Methods
and Approaches, edited by David Leopold and Marc Stears, 11–28. New York:
Oxford University Press, 2008.
McGrath, Sarah. ‘Moral Disagreement and Moral Expertise’. In Oxford Studies in
Metaethics. Vol. 3, edited by Russ Shafer-Landau, 87–107. New York: Oxford
University Press, 2008.
———. ‘Skepticism About Moral Expertise as a Puzzle for Moral Realism’. Journal
of Philosophy 108 (2011): 111–37.
McKeever, Sean and Ridge, Michael. Principled Ethics. New York: Oxford Univer-
sity Press, 2006.
———. ‘Elusive Reasons’. In Oxford Studies in Metaethics. Vol. 7, edited by Russ
Shafer-Landau, 110–37. New York: Oxford University Press, 2012.
McKerlie, Dennis. ‘Equality and Time’. Ethics 99 (1989): 475–91.
McKie, John, Kuhse, Helga, Richardson, Jeff and Singer, Peter. ‘Double Jeopardy, the
Equal Value of Lives and the Veil of Ignorance: A Rejoinder to Harris’. Journal of
Medical Ethics 22 (1996): 204–8.
McMahan, Jeff. ‘Cognitive Disability, Misfortune, and Justice’. Philosophy and Pub-
lic Affairs 25 (1996): 3–34.
———. ‘Moral Intuitions’. In The Blackwell Guide to Ethical Theory, edited by
Hugh Lafollette, 92–110. Malden, MA: Blackwell Publishers, 2000.
———. The Ethics of Killing: Problems at the Margins of Life. New York: Oxford
University Press, 2002.
———. ‘Animals’. In A Companion to Applied Ethics, edited by R.G. Frey and
Christopher Heath Wellman, 525–36. Malden, MA: Blackwell Publishing, 2005.
———. ‘Radical Cognitive Limitation’. In Disability and Disadvantage, edited by Kim-
berley Brownlee and Adam Cureton, 240–59. Oxford: Oxford University Press, 2009.
———. ‘Infanticide and Moral Consistency’. Journal of Medical Ethics 39 (2013a):
273–80.
———. ‘Causing People to Exist and Saving People’s Lives’. The Journal of Ethics
17 (2013b): 5–35.
174 Works Cited
McNaughton, David. ‘An Unconnected Heap of Duties?’. In Ethical Intuitionism,
edited by Philip Stratton-Lake, 76–91. Oxford: Oxford University Press, 2002.
Nagel, Thomas. ‘Preface’. In Mortal Questions, edited by Thomas Nagel, ix–xiii.
Cambridge: Cambridge University Press, 1979.
———. The View From Nowhere. New York: Oxford University Press, 1986.
———. ‘Equality’. In The Ideal of Equality, edited by Matthew Clayton and Andrew
Williams, 60–80. Basingstoke, Hampshire: Palgrave Macmillan, 2002.
Nelson, Mark. ‘What the Utilitarian Cannot Think’. Ethical Theory and Moral Prac-
tice 18 (2015): 717–29.
Nussbaum, Martha. Frontiers of Justice. Cambridge, MA: The Harvard University
Press, 2006.
———. ‘The Capabilities Approach and Animal Entitlements’. In The Oxford Hand-
book of Animal Ethics, edited by Tom L. Beauchamp and R.G. Frey, 228–51. New
York: Oxford University Press, 2011.
Oberdiek, John. ‘Lost in Moral Space: On the Infringing/Violating Distinction and
Its Place in the Theory of Rights’. Law and Philosophy 23 (2004): 325–436.
O’Neill, Onora. ‘Necessary Anthropocentrism and Contingent Speciesism’. Proceed-
ings of the Aristotelian Society Supplementary Edition 72 (1998): 211–28.
———. ‘Kant and the Social Contract Tradition’. In Kant’s Political Theory: Inter-
pretations and Applications, edited by Elisabeth Ellis, 25–41. University Park, PA:
Pennsylvania State University Press, 2012.
Owens, David. Shaping the Normative Landscape. New York: Oxford University
Press, 2012.
Parfit, Derek. Reasons and Persons. New York: Oxford University Press, 1987.
———. ‘Equality or Priority?’. In The Ideal of Equality, edited by Matthew Clay-
ton and Andrew Williams, 81–125. Basingstoke, Hampshire: Palgrave Macmillan,
2002.
———. On What Matters. Vol. 1. New York: Oxford University Press, 2011.
Persson, Ingmar. ‘A Defence of Extreme Egalitarianism’. In Egalitarianism: New
Essays on the Nature and Value of Equality, edited by Nils Holtug and Kasper
Lippert-Rasmussen, 83–97. Oxford: Clarendon Press, 2007.
Peter, Fabienne. ‘Health Equity and Social Justice’. In Public Health, Ethics, and
Equity, edited by Sudhir Anand, Fabienne Peter and Amartya Sen, 93–106. New
York: Oxford University Press, 2004.
Pogge, Thomas. ‘Is Kant’s Rechtslehre Comprehensive?’. The Southern Journal of
Philosophy XXXVI, Supplement (1997): 161–87.
———. ‘Severe Poverty as a Violation of Negative Duties’. Ethics and International
Affairs 19 (2005): 55–83.
Portmore, Douglas. Commonsense Consequentialism. New York: Oxford University
Press, 2011.
Pound, Pandora and Bracken, Michael. ‘Is Animal Research Sufficiently Evidence
Based to Be a Cornerstone of Biomedical Research?’. British Medical Journal 348
(2014): g3387.
Powers, Madison and Faden, Ruth. Social Justice: The Moral Foundations of Public
Health and Health Policy. New York: Oxford University Press, 2006.
Prichard, H.A. ‘Does Moral Philosophy Rest on a Mistake?’. Mind 21 (1912): 21–37.
Psillos, Stathos. Scientific Realism: How Science Tracks Truth. Oxford: Oxford Uni-
versity Press, 1999.
Rachels, James. The Elements of Moral Philosophy. Boston: McGraw-Hill, 1999.
Works Cited 175
———. ‘Drawing Lines’. In Animal Rights: Current Debates and New Directions,
edited by Cass R. Sunstein and Martha Nussbaum, 162–74. Oxford: Oxford Uni-
versity Press, 2004.
Rakowski, Erik. ‘The Aggregation Problem’. Hastings Center Report 24 (1994): 33–6.
Ram-Tiktin, Efrat. ‘The Right to Health Care as a Right to Basic Human Functional
Capabilities’. Ethical Theory and Moral Practice 15 (2012): 337–51.
———. ‘Basic Human Functional Capabilities as the Currency of Sufficientarian
Distribution in Health Care’. In What Is Enough? Sufficiency, Justice, and Health,
edited by Carina Fourie and Annette Rid, 144–63. New York: Oxford University
Press, 2017.
Rawls, John. A Theory of Justice, rev. ed. Cambridge, MA: The Belknap Press, 1999a.
———. The Law of Peoples. Cambridge, MA: The Harvard University Press, 1999b.
Raz, Joseph. Practical Reason and Norms. Princeton, NJ: Princeton University Press,
1990.
———. ‘The Truth in Particularism’. In Moral Particularism, edited by Brad Hooker
and Margaret Little, 48–78. New York: Oxford University Press, 2000.
Regan, Tom. ‘The Nature and Possibility of an Environmental Ethic’. Environmental
Ethics 3 (1981): 173–98.
———. The Case for Animal Rights. Berkeley, CA: The University of California
Press, 1983.
Richardson, Henry S. ‘Specifying Norms as a Way to Resolve Concrete Ethical Prob-
lems’. Philosophy & Public Affairs 19 (1990): 279–310.
Roberts, M.A. ‘A New Way of Doing the Best That We Can: Person-Based Conse-
quentialism and the Equality Problem’. Ethics 112 (2002): 315–50.
Robeyns, Ingrid. ‘Ideal Theory in Theory and Practice’. Social Theory and Practice
34 (2008): 341–62.
Rollin, Bernard E. Animal Rights & Human Morality. Amherst, NY: Prometheus
Books, 2006.
Ross William David. Foundations of Ethics. Oxford: Clarendon Press, 1939.
———. The Right and the Good. Indianapolis: Hackett, 1988.
Rowlands, Mark. Animal Rights: Moral Theory and Practice. Basingstoke, Hamp-
shire: Palgrave Macmillan, 2009.
Ruger, Jennifer Prah. Health and Social Justice. New York: Oxford University Press,
2010.
Sachs, Benjamin. ‘Does Rationality Ever Require Immorality?’ PhD diss., University
of Wisconsin-Madison, 2006.
———. ‘Reasons and Requirements’. Ethical Theory and Moral Practice 11 (2008a):
73–83.
———. ‘The Liberty Principle and Universal Health Care’. The Kennedy Institute of
Ethics Journal 18 (2008b): 149–72.
———. ‘Consequentialism’s Double-Edged Sword’. Utilitas 22 (2010a): 258–71.
———. ‘The Exceptional Ethics of the Investigator-Subject Relationship’. The Jour-
nal of Medicine & Philosophy 35 (2010b): 64–80.
———. ‘Going From Principles to Rules in Research Ethics’. Bioethics 25 (2011): 9–20.
———. ‘Reasons Consequentialism’. Journal of Moral Philosophy 10 (2013): 671–82.
———. ‘Contractarianism as a Political Morality’. Proceedings of the Aristotelian
Society CXVI (2016): 49–67.
Sapontzis, Steve. ‘Animal Rights and Biomedical Research’. The Journal of Value
Inquiry 26 (1992): 73–86.
176 Works Cited
Savulescu, Julian. ‘Genetically Modified Animals: Should There Be Limits to Engi-
neering the Animal Kingdom?’. In The Oxford Handbook of Animal Ethics,
edited by Tom L. Beauchamp and R.G. Frey, 641–70. New York: Oxford Univer-
sity Press, 2011.
Scanlon, T.M. What We Owe to Each Other. Cambridge, MA: The Belknap Press, 1998.
———. Being Realistic About Reasons. New York: Oxford University Press, 2014.
Scheffler, Samuel. The Rejection of Consequentialism. New York: Oxford University
Press, 1982.
Schroeder, Mark. Slaves of the Passions. New York: Oxford University Press, 2007.
Segall, Shlomi. Health, Luck, and Justice. Princeton, NJ: Princeton University Press,
2010.
Sen, Amartya. Inequality Reexamined. Cambridge, MA: Harvard University Press,
1992.
———. ‘Why Health Equity?’. In Public Health, Ethics, and Equity, edited by Sudhir
Anand, Fabienne Peter and Amartya Sen, 21–33. New York: Oxford University
Press, 2004.
———. ‘What Do We Want From a Theory of Justice?’. The Journal of Philosophy
CIII (2006): 215–38.
Shafer-Landau, Russ. Moral Realism. New York: Oxford University Press, 2003.
Shaw, William H. ‘Contemporary Criticisms of Utilitarianism: A Response’. In The
Blackwell Guide to Mill’s Utilitarianism, edited by Henry R. West, 201–16. Mal-
den, MA: Blackwell, 2006.
Shields, Liam. ‘Some Questions (and Answers) for Sufficientarians’. In What Is
Enough? Sufficiency, Justice, and Health, edited by Carina Fourie and Annette
Rid, 85–100. New York: Oxford University Press, 2017.
Shiffrin, Seana Valentine. ‘Wrongful Life, Procreative Responsibility, and the Signifi-
cance of Harm’. Legal Theory 5 (1999): 117–48.
Shpall, Sam. ‘Moral and Rational Commitment’. Philosophy and Phenomenological
Research 88 (2014): 146–72.
Sidgwick, Henry. The Methods of Ethics. Indianapolis: Hackett, 1981.
Simmons, A. John. ‘Ideal and Nonideal Theory’. Philosophy & Public Affairs 38
(2010): 5–36.
Singer, Peter. ‘Famine, Affluence, and Morality’. Philosophy and Public Affairs 1 (1972):
229–43.
———. Practical Ethics. Cambridge: Cambridge University Press, 1993.
———. Animal Liberation. London: Random House, 1995.
———. ‘Ethics and Intuitions’. Journal of Ethics 9 (2005): 331–52.
Sinnott-Armstrong, Walter. Moral Dilemmas. Oxford: Basil Blackwell, 1988.
Slicer, Deborah. ‘My Daughter or My Dog? A Feminist Assessment of the Animal
Research Issue’. Hypatia 6 (1991): 108–24.
Sliwa, Paulina. ‘In Defense of Moral Testimony’. Philosophical Studies 158 (2012):
175–95.
———. ‘On Knowing What’s Right and Being Responsible for It’. In Responsibility—
The Epistemic Condition, edited by Philip Robichaud and Jan Willem Weiland,
127–45. New York: Oxford University Press, 2017.
Smith, Michael. The Moral Problem. Malden, MA: Blackwell Press, 1994.
———. ‘Normative Reasons and Full Rationality: Reply to Swanton’. Analysis 156
(1996): 160–8.
Works Cited 177
Southwood, Nicholas. ‘Deliberative Models of Contractualism’. Politics, Philoso-
phy & Economics 7 (2008): 183–208.
Sreenivasan, Gopal. ‘Health and Justice in Our Non-Ideal World’. Politics, Philoso-
phy & Economics 6 (2007): 218–36.
Steinbock, Bonnie. Life Before Birth: The Moral and Legal Status of Embryos and
Fetuses. New York: Oxford University Press, 1992.
Stewart, Noel. Ethics: An Introduction to Moral Philosophy. Cambridge: Polity
Press, 2009.
Strandberg, Caj. ‘Particularism and Supervenience’. In Oxford Studies in Metaeth-
ics. Vol. 3, edited by Russ Shafer-Landau, 129–58. New York: Oxford University
Press, 2008.
Stratton-Lake, Philip. ‘Scanlon’s Contractualism and the Redundancy Objection’.
Analysis 63 (2003): 70–6.
———. ‘Recalcitrant Pluralism’. Ratio (New Series) XXIV (2011): 364–83.
Strawson, P.F. ‘Freedom and Resentment’. Proceedings of the British Academy 48
(1962): 1–25.
Street, Sharon. ‘A Darwinian Dilemma for Realist Theories of Value’. Philosophical
Studies 127 (2006): 109–66.
Sumner, L.W. Abortion and Moral Theory. Princeton, NJ: Princeton University Press,
1981.
Tännsjö, Torbjörn. From Reasons to Norms: On the Basic Questions in Ethics. Dor-
drecht: Springer, 2010.
Temkin, Larry. ‘Justice, Equality, Fairness, Desert, Rights, Free Will, Responsibility,
and Luck’. In Responsibility and Distributive Justice, edited by Carl Knight and
Zofia Stemplowska, 51–76. New York: Oxford University Press, 2011.
———. Rethinking the Good. New York: Oxford University Press, 2012.
Thompson, Michael. ‘What Is It to Wrong Someone: A Puzzle About Justice’. In Rea-
son and Value: Themes From the Moral Philosophy of Joseph Raz, edited by R.
Jay Wallace, Philip Pettit, Samuel Scheffler and Michael Smith, 333–84. Oxford:
Clarendon Press, 2004.
Thomson, Judith Jarvis. Normativity. Chicago: Open Court Publishing, 2008.
Timmermann, Jens. ‘When the Tail Wags the Dog: Animal Welfare and Indirect Duty
in Kantian Ethics’. Kantian Review 10 (2005): 128–49.
Tulodziecki, Dana. ‘Underdetermination, Methodological Practices, and Realism’.
Synthese 190 (2013): 3731–50.
Urmson, J.O. ‘A Defence of Intuitionism’. Proceedings of the Aristotelian Society 75
(1974): 111–19.
Vallentyne, Peter. ‘Of Mice and Men: Equality and Animals’. Journal of Ethics 9
(2005): 403–33.
VanDeVeer, Donald. ‘Of Beasts, Persons, and the Original Position’. The Monist 62
(1979): 368–77.
van Fraassen, Bas. The Scientific Image. Oxford: Oxford University Press, 1980.
———. Laws and Symmetry. Oxford: Oxford University Press, 1989.
Väyrynen, Pekka. ‘Moral Generalism: Enjoy in Moderation’. Ethics 116 (2006):
707–41.
———. ‘A Theory of Hedged Moral Principles’. In Oxford Studies in Metaethics.
Vol. 1, edited by Russ Shafer-Landau, 91–132. New York: Oxford University
Press, 2009.
178 Works Cited
———. ‘Grounding and Normative Explanation’. Proceedings of the Aristotelian
Society Supplementary Edition 87 (2013): 155–78.
Wallace, R. Jay. ‘Reasons, Relations, and Commands: Reflections on Darwall’. Eth-
ics 118 (2007): 24–36.
———. ‘The Deontic Structure of Morality’. In Thinking About Reasons: Themes
From the Philosophy of Jonathan Dancy, edited by David Bakhurst, Brad Hooker
and Margaret Olivia Little, 137–67. Oxford: Oxford University Press, 2013.
Warren, Mary Ann. Moral Status. New York: Oxford University Press, 2000.
White, Roger. ‘Why Favour Simplicity?’. Analysis 65 (2005): 205–10.
Williams, Bernard. ‘A Critique of Utilitarianism’. In Utilitarianism: For & Against,
edited by J.J.C. Smart and Bernard Williams, 77–150. Cambridge: Cambridge
University Press, 1973.
———. ‘Persons, Character, and Morality’. In Moral Luck, edited by Bernard Wil-
liams, 1–19. Cambridge: Cambridge University Press, 1981.
Wringe, Bill. ‘International Justice, Intervention, and the Prevention of Evil’. In Evil:
An Interdisciplinary Approach, edited by Margaret Sönser Breen, 75–94. Amster-
dam: Rodopi, 2003.
Zangwill, Nick. ‘Cordelia’s Bond and Indirect Consequentialism’. In Oxford Stud-
ies in Normative Ethics. Vol. 1, edited by Mark Timmons, 143–65. New York:
Oxford University Press, 2011.
Index