University of Southern California Dissertations and Theses

The virtue of reasonableness: on the normative epistemic implications of public reason liberalism

(USC Thesis Other)

The virtue of reasonableness: on the normative epistemic implications of public reason liberalism

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE VIRTUE OF REASONABLENESS:
ON THE NORMATIVE EPISTEMIC IMPLICATIONS OF PUBLIC REASON LIBERALISM

by

Michael Douglas Ashfield

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
PHILOSOPHY

December 2021

Copyright 2021 Michael Douglas Ashfield

ii

EPIGRAPH
Is there no virtue among us? If there be not, we are in a wretched situation. No
theoretical checks—no form of government can render us secure. To suppose that
any form of government will secure liberty or happiness without any virtue in the
people, is a chimerical idea.
—James Madison, Speech to the Virginia Ratifying Convention, June 20, 1788
1

The alternative then set, and still before us, is not one between classical virtue and
liberal self-indulgence, but between cruel military and moral repression and
violence, and a self-restraining tolerance that fences in the powerful to protect the
freedom and safety of every citizen, old or young, male or female, black or white.
Far from being an amoral free-for-all, liberalism is, in fact, extremely difficult and
constraining, far too much so for those who cannot endure contradiction,
complexity, diversity, and the risks of freedom. The habits of freedom are
developed, moreover, both in private and in public, and a liberal character can
readily be imagined. It is, however, by definition not to be forced or even promoted
by the use of political authority. That does not render the tasks of liberalism any
easier, but it does not undermine its ethical structure.

—Judith Shklar, Ordinary Vices, 1984, 5

1
See Kloppenberg 1998, 33

iii

DEDICATION
For Amanda, Charis, Emmett, Mom, & Dad

iv
ACKNOWLEDGEMENTS
This dissertation has been a very long time in the making, so I will surely fail to remember every
debt of gratitude I’ve incurred along the way. But I must begin by gratefully acknowledging the
patience and support of my dissertation committee members, David Albertson, Jonathan Quong,
Mark Schroeder, and especially my supervisor, Ralph Wedgwood. Ralph’s many insights have
challenged me, kept my dissertation project alive, and made my work better. Jon and Mark have
likewise been generous, patient, supportive, constructive and challenging in all the ways that a
supervisee hopes for. In addition, Ralph and Jon’s co-supervision of my area examination, Ralph’s
multiple seminars in the normative subdisciplines, Jon’s seminar in political philosophy, and his
supervision of a subsequent independent study of public reason, all provided me with the
foundations I needed to explore the epistemology of public reason liberalism. Finally, David joined
my committee, as the external member, when completion seemed particularly uncertain, and his
kindness was a great encouragement as I redoubled my efforts to finish.
In thinking over my many friends and colleagues at USC, who have played a role in shaping
the present work, I wish to begin by thanking the members of my graduate student cohort, Maegan
Fairchild, Wooram Lee, Jonathan Wright, and especially Renee Jorgensen, who was the first
person to suggest that I focus my dissertation project on the epistemology of public reason.
Together with these and many of our fellow grad students, including Rima Basu, Caleb Perl, Steve
Bero, Aness Webster, Ara Astourian, Tanya Kostochka, Nathan Howard, Alex Dietz, Joe Horton,
Nicola Kemp, Sean Donahue, Paul Garofalo, and many others, I workshopped many of the ideas,
which found their way into the dissertation, through seminars, Spec Society talks, and discussions
in the grad lounge, coffee shops, bars, and the occasional hike. Your contributions are innumerable
and deeply appreciated. Likewise, the tireless logistical support and warm encouragement of

v
USC’s Philosophy Departmental Office staff across the years, especially that of Cynthia Lugo and
Natalie Schaad, has been indispensable to my success.
I also want to thank my many generous commentators and interested colleagues at
conferences and workshops over the years, including especially Aaron Ancell, Nathan Brett,
Michael Da Silva, Andrew Fenton, Kyle Johannsen, Federica Liveriero, Andrew Molas, Emily
Podhorcer, David Rondel, and John Wilcox, as well as helpful audience members at the 2018
Berkeley-Stanford-Davis Graduate Conference, 2016 MANCEPT Workshop of Theories of Public
Reason, 2016 Canadian Philosophical Association, 2016 Rocky Mountain Ethics Congress,
colloquia talks to the philosophy departments of Dalhousie University and UC Santa Barbara, and
multiple presentations across several years to both the Canadian Section of the International
Association for Philosophy of Law & Social Philosophy (CS-IVR) and the Atlantic Region
Philosophers’ Association (ARPA). There are far too many of you to name here, but suffice it to
say that your many helpful comments, criticisms, and words of encouragements sustained my
confidence in the value of my dissertation project through countless moments of doubt.
Finally, I want to thank my loving and supportive extended family, first and foremost, my
wife, Amanda, who has carried me through the lows, while cheering me every step of the way to
completing my studies. I also want to thank my parents, Doug & Doris, who have loved me
unconditionally, provided endless amounts of encouragement to me, and allowed Amanda, our
two children, and I, to live with them in Fredericton, New Brunswick, Canada throughout the
pandemic, which enabled me to focus much more time on my work, and made those months,
isolated in our family bubble, a very special and memorable time for our children. I love you all
more than I can say.

vi
Table of Contents

EPIGRAPH ......................................................................................................................... ii
DEDICATION ................................................................................................................... iii
ACKNOWLEDGEMENTS ............................................................................................... iv
ABSTRACT ....................................................................................................................... xi
I. The Concept of the “Reasonable” ................................................................................... 1
1. The Job Description of the “Reasonable” ................................................................. 10
1.1 The “Reasonable” are Committed to the Publicity Constraint ........................... 13
1.2 The “Reasonable” accept “Reasonable” Pluralism ............................................. 15
1.3 The “Reasonable” Are Epistemically Competent ............................................... 17
1.4 The “Reasonable” are Sincere............................................................................. 19
1.5 A Full Job Description for the Concept of the “Reasonable” ............................. 21
2. Liberalism Anti-Perfectionism: The Overarching Problem and Two Guiding Aims 23
2.1 The Overarching Problem Arising from the Special Circumstances of Politics. 24
2.2 Anti-Perfectionist Liberalism’s Secondary Aim of Neutrality ........................... 25
2.3 A Political Conception of Liberalism ................................................................. 28
2.4 Liberal Anti-Perfectionism: Liberal Theory Tailored to Solving the Overarching
Problem ............................................................................................................... 31
II. Enoch’s Challenge to Public Reason Liberalism ......................................................... 34
1. Getting Epistemological about Reasonable Disagreement ....................................... 36

vii
1.1 Enoch’s Trilemma ............................................................................................... 39
1.2 Outline of Remaining Sections ........................................................................... 45
2. Trying to do Without a Political Conception ............................................................ 46
2.1 Rawlsian Reasonableness of Persons .................................................................. 47
2.2 The Burdens of Judgment ................................................................................... 48
2.3 Problems with the Burdens View ....................................................................... 49
3. A Political Conception of Reasonable Doctrinal Disagreement ............................... 53
3.1 Rawlsian Criteria of Reasonable Doctrinal Disagreement ................................. 53
3.2 The Non-Vacuity of the Publicity Constraint ..................................................... 54
3.3 The Lexical Priority of Moral Decency Over Epistemic Competence ............... 56
3.4 Epistemic Competence: Letting Many Flowers Bloom ...................................... 58
3.5 A Political Conception of Reasonable Doctrinal Disagreement ......................... 59
4. The Private Epistemological Commitments of the Reasonable ................................ 60
4.1 Deflationary Hypotheses about Disagreement .................................................... 61
4.2 Principled and Evidential Epistemological Hypotheses about Disagreement .... 62
4.3 Sophisticated Psychological Hypotheses and the Second Horn ......................... 66
III. Virtue Reliabilism for Liberal Anti-Perfectionists ..................................................... 69
1. Three Challenges for Anti-Perfectionist Virtue Theories ......................................... 69
2. Liberal Virtue Theories ............................................................................................. 71
2.1 Perfectionist and Miscellaneous Liberal Virtue Theories ................................... 72

viii
2.2 Anti-Perfectionist Liberal Virtue Theories ......................................................... 75
3. The Problem of Praiseworthiness and Virtue Reliabilism ........................................ 84
3.1 Rawls on the Moral Worth of Acting from Virtue ............................................. 85
3.2 Audi on the Moral Content of Self-Consciously Virtuous Action...................... 88
3.3 The Robustness of Reliabilist Virtues ................................................................. 89
4. Virtue Theory and the Metaethics of Civic Duty ...................................................... 91
4.1 Boettcher on Less Stringent Accounts of Civic Duties ...................................... 91
4.2 Boettcher on the Metaethics of Virtue-Theoretic Accounts of Civic Duty ........ 93
5. The Situationist Challenge, Social Norms, and Virtue Externalism ......................... 95
5.1 The Updated Situationist Challenge to Liberal Virtue Theory ........................... 96
5.2 Revising Virtue (Internalism) in Vain ................................................................ 97
5.3 Externalism about Liberal Civic Virtue .............................................................. 98
5.4 Externalism and Social Norms ............................................................................ 99
6. Conclusion: Virtue Reliabilism for Liberal Anti-Perfectionists ............................. 102
IV. Epistemological Externalism for Liberal Anti-Perfectionists ................................... 104
1. The Prevailing Epistemological Internalism about Reasonableness ...................... 107
1.1 Internalism about Epistemic Rationality ........................................................... 108
1.2 Internalism about the Epistemic Dimension of Reasonableness and Public Reasons
........................................................................................................................... 109
2. Epistemological Externalism about Reasonableness? ............................................ 113

ix
2.1 Liberal Anti-Perfectionists and Epistemological Externalism .......................... 114
2.2 Wall’s Perfectionist Epistemological Externalism ........................................... 117
2.3 So Why Not Externalism? ................................................................................ 119
3. Public Reason: Somewhere Between Truth and Consent ....................................... 122
3.1 Truth: Unnecessary and Insufficient for Good Faith ........................................ 123
3.2 Persuasion: Insufficient for Good Faith ............................................................ 124
3.3 Ideal Interlocutors: Unnecessary for Good Faith .............................................. 125
3.4 Transparent Persuasion: Unnecessary and Insufficient for Good Faith ............ 125
3.5 Good Faith Debate: Sincere, Cooperative, Adversarial Persuasion .................. 126
4. Epistemological Externalism for Liberal Anti-Perfectionists ................................. 128
4.1 Virtue Reliabilism for Liberal Anti-Perfectionists............................................ 128
4.2 Reliabilism about the Highly General Virtue of Reasonableness ..................... 130
4.3 Virtue Reliabilism about Reasonable, Civil Public Inquiry .............................. 132
4.3 Epistemological Externalism about Reasonableness ........................................ 134
4.4 The Instrumental Value of Reasonableness ...................................................... 135
4.5 Why Putting Cooperation First Provides Us with Enough ............................... 136
5. Putting Externalistic Reasonableness to the Test .................................................... 138
5.1 Accura of Utopia ............................................................................................... 139
5.2 Aimy of Pantopia .............................................................................................. 140
5.3 Public Reasoning in Utopia vs. Pantopia .......................................................... 141

x
5.4 Virtue Externalism about Public Justification .................................................. 144
6. Reasonableness and Luck ....................................................................................... 146
6.1 Reasonable Dispositions in New Brunswick vs. Nigeria .................................. 147
6.1 Kendra and Pristine in New Brunswick ............................................................ 148
6.2 Oba and Wemimo in Nigeria ............................................................................ 150
6.3 Intellectual Dispositions are Reasonable Relative to One’s External Social
Conditions ......................................................................................................... 152
7. Conclusion: Epistemological Externalism for Liberal Anti-Perfectionists ............ 155
V. (Dis)solving the Idealization Dilemma for Public Reason Liberalism ...................... 157
1. The Idealization Dilemma for Public Reason Liberalism ....................................... 159
1.1 The Under-Idealization Horn ............................................................................ 159
1.2 The Over-Idealization Horn .............................................................................. 161
1.3 Why not Lower the Bar? ................................................................................... 162
1.4 Desiderata for Avoiding the Dilemma .............................................................. 166
2. Toward a Dissolution of the Idealization Dilemma ................................................ 166
2.1 The Road Taken: Internalistic Reasonableness + Information ......................... 167
2.2 The Road Not Taken: Externalistic Reasonableness ........................................ 169
2.3 Getting there by the Road Not Yet Taken......................................................... 170
BIBLIOGRAPHY ........................................................................................................... 173

xi
ABSTRACT
This dissertation explores the normative epistemic implications of public reason liberalism’s
notion of being “reasonable.” Chapter I lays out the job description of this complex normative
concept. Chapter II explains how public reason liberal theorists can articulate their own theory, in
a way that sheds light on the epistemological commitments and doxastic attitudes of “reasonable”
parties to doctrinal disagreement, while avoiding epistemological controversy to satisfy a demand
that liberal theory itself remain neutral in such controversies. In this chapter, I also defend some
thin epistemological constraints on the “reasonableness” of disagreement.
Chapter III turns to the question of how best to understand the normative status of
“reasonableness” and its associated publicity and fair-mindedness norms of public reason. I show
that a virtue theoretic account of “reasonableness” is very well-positioned to play the roles that
public reason liberals have assigned these concepts. On this kind of view, public reasoning in
accordance with the norms, is one centrally important manifestation of “reasonableness.”
However, by adopting a virtue-theoretic approach, I open my account to several attacks that have
been levelled against existing accounts of liberal virtues, so in this chapter I also defend a new
reliabilist account of liberal virtue that, I argue, is able to fend off these attacks.
Chapter IV builds on this reliabilist picture, arguing that epistemic rationality (almost
universally understood as an internal notion or matter what’s going on inside people’s heads) can’t
provide all the epistemic idealization that public reason theorists need from “reasonableness.”
After all, you can have a person who’s perfectly rational on the inside, but so isolated in their fake
news bubble, and so systematically misinformed about important social facts, like distributions of
wealth, income, and opportunity, the state of the nation’s finances, and other relevant empirical
questions, that they’re too ready to accept false or misleading “justifications” for exercises of

xii
power. So “reasonableness” must also have some externalist epistemic component.
“Reasonableness” just can’t be purely a matter of what’s going on inside someone’s head, how
sincere they are, or how carefully they’re reasoning, etc.
Accordingly, I advance what I call a “cooperation first” account of “reasonableness,” on
which this virtue consists partly in reliably tracking facts of relevance to the exercises of power
we seek to publicly justify to one another. So reasonableness is first and foremost a matter of
whether and how we succeed in cooperating fairly with fellow members of our society in all
matters of civic life—not merely of how we intend or take ourselves to be interacting with them.
There’s an external socio-political fact of the matter as to whether one’s being reasonable when,
for example, one attempts to publicly justify some exercise of power, or weighs the merits of such
attempts. This ensures that an agent’s contributions to our shared social life and shared task of
democratic governance are informed and constrained by reality.
Finally, in Chapter V, I argue that my account is well-positioned to avoid a well-known
dilemma for idealizations of the “reasonable.” Public reason theorists face a choice. They can
idealize the reasonable just a little, stipulating to a few thin, uncontroversial, normative
requirements (e.g., practical rationality), limiting the number of people excluded from the
constituency of the “reasonable,” and ensuring this constituency is a big tent. But too little
idealization leaves the “reasonable” so ideologically diverse as to make it impossible to find
justifications, acceptable to all, capable of sustaining enough exercises of power for an orderly and
functional society. In that case, we accept either anarchy or illegitimate governance. Alternatively,
public reason theorists can idealize more to limit membership among the reasonable, making it
easier for its smaller membership to reach determinate agreements about justified exercises of
power and authority. But too much idealization makes membership in public reason’s constituency

xiii
too exclusive. The resulting liberal theory becomes unrecognizable as one designed to respect the
pluralism. So public reason theorists need to find some principled account of a moderate
idealization of the reasonable, which avoids both horns of this dilemma.
The key to providing this principled account, I argue, is to recognize that internalism about
“reasonableness” necessitates all the public reason theorist’s idealization to be baked into agents’
heads. But once we force all idealization inside agents’ heads, only complicated and implausibly
demanding constraints are able to rule out their being too credulous of far-fetched, dubious
possibilities. I argue public reason theorists should recognize that sometimes it’s in virtue of social
or political circumstances, rather than internal features, that agents fail or succeed in being
“reasonable.” So public reason theorists should idealize in a way that requires environmental
conditions (minimally) favorable to “reasonableness,” creating space for more moderately
idealized agents to count as “reasonable.”

1
I. The Concept of the “Reasonable”
Imagine you are a pedestrian on Hollywood Boulevard, on your way to an important job interview,
when you encounter a major street closure, extending at least a block North and South of
Hollywood for a few blocks ahead of you. Your appointed destination seems to be right in the
middle of the affected area. You are unsure why roads are closed to traffic: it could be a movie
shoot, preparation for a major entertainment awards show, a police action, or security measures
for a Presidential motorcade. The fact that LAPD is managing pedestrians provides you no clue,
as you know they might be involved in any of the above. Furthermore, there does not seem to be
any rhyme or reason to how the adjacent sidewalks and crosswalks are (or are not) affected by the
closure. You assess the situation and, after some circuitous navigation, you arrive at the corner of
Hollywood and Highland. The walk light signals, it looks safe to cross, and it does not appear you
will get in anyone’s way, so you start to cross. But as you near the other side of the street, a police
officer orders you to turn back without providing a reason. Annoyed, you comply, and retrace your
steps back along the sidewalk for a block, head South, then zig-zag your way back. You finally
arrive through an even more circuitous route, at the same corner you were prevented from reaching
11 minutes earlier, but nothing you have seen or heard in the intervening minutes explains the
officer’s order. You are now certain to be late for the interview, which will significantly hurt your
chances, and it seems you have been wronged. You do not question the LAPD’s authority to
manage pedestrians, whatever might be going on, and neither do you suspect that you were singled
out for any discriminatory reason, like the color of your skin (this time, at least). In this case, the
absence of a compelling justification for the seemingly arbitrary exercise of authority, it seems
like a norm has been violated. But what kind of norm?

2
Intuitively, most people think it is wrong for an agent to boss others around without
providing sufficient justification, even if we think the agent giving the orders has legitimate
authority to do so. Of course, sometimes we make exceptions: in an emergency, we forgive first
responders or responsible bystanders for taking charge and barking orders to “stay back!” or “apply
pressure here!”. But other things being equal, we expect those giving commands to justify them.
As in the above non-emergency example, we expect police to provide justification for cordoning
off a public area or right of way, pulling someone over, or detaining someone, even if we don’t
question their authority as police officers to do so. Indeed, journalists spend a significant amount
of their time seeking justifications for various exercises of authority (or power) by the government,
businesses, religious organizations, and other actors in society, and ensuring proffered
justifications are made available for public scrutiny. Of course, the more arbitrary an exercise of
authority (or power) seems, and/or the greater the moral or practical significance of what’s at stake,
the more likely we are to question the legitimacy of the relevant authority.
These concerns bear considerable resemblance to the kinds of things that an important class
of political theorists worry about. It’s sometimes said that liberalism’s “moral lodestar” is the
conviction that the exercise of political power or authority is legitimate only when it can be
publicly justified.
2
Public reason liberals make the public justification of political authority and of
liberalism itself a central aim of their political philosophy. But public reason liberal theorists
typically don’t think the requirement of public justification extends to personal interactions on the

2
Quong 2011, 2–3; see Macedo 1990, 78; Waldron 1993, 36–7.

3
street with police officers. John Rawls, for example, limits public reason’s site of application to
the so-called public political forum—that is, the discourse of judges, government officials,
candidates for public office and their campaign managers.
3
On the related question of which kinds
of entities stand in need of public justification, sometimes called the question of public reason’s
scope, Rawls limits these to the constitutional essentials and matters of basic justice, “but not in
general [to] all the questions to be settled by the legislature within a constitutional framework.”
4

Therefore, if the norm violated on the street is one of public reason, its site and scope are
much broader than those theorized by Rawls. Given how much is morally and practically at stake
in the political legitimacy of the state, and its (often coercive) exercises of authority or power, it’s
not surprising that public reason liberals have devoted special attention to the problem of publicly
justifying and legitimizing the state’s authority. However, I see no reason to think that something
akin to the norms of public reason do not apply to less consequential and more mundane exercises
of authority or power in everyday life.
Suppose there’s a more general norm requiring that any exercise of authority (or power)
be justified to those subject to it. As public reason liberals point out, there’s a problem. People
disagree—often, very deeply—about what constitutes a satisfactory justification for some type or
token exercise of authority or power. Of course, there are almost always (at least) a few
unreasonable people who are unwilling to accept any plausible justification for certain exercises
of authority or power. But even among reasonable people, there’s often plenty of room for

3
Rawls 1999, 575.
4
Rawls 2001, 91. One reason to limit the site this way is that, “we had a vote and your side lost,” is a satisfactory
procedural public justification for some things but not for constitutional essentials, like civil rights. Were there to be
an amendment to repeal the first amendment, it would be a constitutional crisis. Some things should be outside the
reach of democratic governance.

4
disagreement, especially in large contemporary free societies where a great variety of moral,
political, religious, and philosophical outlooks flourish. Suppose the police officer told you to turn
back at the crosswalk, because the sidewalk ahead was getting more crowded than he likes. Some
would unquestioningly accept the word of a police officer, but unless the sidewalk was clogged
with an immobile mass of people, most of us would find this ridiculous and unsatisfying. This
diversity of opinion, much of it stemming from moral, political, religious, and philosophical
pluralism, makes actual unanimous acceptance of a candidate justification impossible to secure in
practice.
One might contend that so long as an agent A doesn’t dispute some exercise e of authority
(or power), or actively take steps to avoid being subject to e, A tacitly accepts e as being self-
evidently justified or forfeits her claim to e’s justification. But this looks like an unworkable
alternative. Just because I comply with the police officer’s order, rather than disobeying or
disputing it, this doesn’t mean that I accept it as being self-evidently justified. Neither do I
plausibly forfeit my claim to deserving some justification for the seemingly arbitrary order. There
may be considerable costs or obstacles to openly disputing some exercises of power or authority,
like missing my job interview altogether if I anger the police officer and end up being detained,
and even more formidable obstacles to actively resisting or avoiding the reach of some authorities
or powers, especially those of the state, including incarceration. To the extent that avoiding such
obstacles may be costly (or even coercive), this undermines the suggestion that tacit acceptance
can be inferred from a lack of disputation of or resistance to some exercise of authority (or power).
Enter the idea of public reason: public reasons are reasons that appeal only to facts, beliefs,
or political values one could expect any “reasonable” disagreeing party to take seriously, whatever
their personal moral, political, religious, or philosophical commitments might be. So public

5
reasons are those objects of overlapping consensus among the “reasonable,” once we’ve set aside
pervasive moral, political, religious, and other philosophical controversies. Drawing only on these
objects of overlapping consensus ensures there’s a common ground justification for the exercise
of authority or power, at least among the “reasonable.”
But one might wonder why there needs to be a common ground justification at all. Why
isn’t it enough for all “reasonable” disagreeing parties to have or receive some sufficiently
satisfying justification for an exercise of power or authority e, such that they converge on accepting
e, regardless of whether there’s any justification they unanimously accept? Mere convergence, as
opposed to an overlapping consensus, may be fine for some purposes, where little more than e
itself hangs on this justification. But the more tasks of social coordination among the “reasonable”
that depend on their interpretation of e’s justification, the more unworkable mere convergence
becomes.
Reaching mere convergence on the justification for any consequential exercise of authority
or power is bit like agreeing to the text of a constitution without being able to agree on how to
interpret or amend it. In the case of a constitution, we clearly need to reach deeper or wider
agreement on something more than just the wording of the document. Of course, on the question
of what justifies some consequential exercises of authority or power, we may not be capable of
coming to any deeper or wider agreement than mere convergence. But in such cases, we often also
agree to a set of fair procedures for settling further disputes that remain when we’ve reached
bedrock disagreement (e.g., procedures for amending a constitution, electing officials, making
appointments, passing laws in a divided legislature, or resolving disputes through the civil law).
In so doing, we rely on procedural justice to generate the wider and deeper agreement needed to
sustain stable cooperation. For if these procedures generate verdicts in a way that’s seen to be fair

6
and impartial, “reasonable” people will accept their legitimacy, making them objects of
overlapping consensus among the “reasonable,” and hence suitably public reasons on which to
base further social coordination. This procedural approach to supplementing mere convergence
with consensus illustrates the inadequacy of the former, and the need for common ground
justification of the exercise of authority or power that can sustain robust social coordination.
So those considerations constituting the common ground among the “reasonable” are
public.
5
Public reasons are to be contrasted with those moral, political, religious, or philosophical
commitments of individuals that, while personally important to some, are controversial,
idiosyncratic, or sectarian, and hence rejected by other plausibly “reasonable” members of society.
Since these commitments fail to be public in the relevant sense, let’s call them non-public
doctrines.
6
The public and non-public are mutually exclusive and jointly exhaustive. Non-public
doctrines may be suitable grounds for justifying exercises of authority or power within a particular
community organized around them. It may be fine for a religious community to make authoritative
pronouncements about the conduct it expects of its membership on the basis of some theological
or philosophical doctrine: Jains, for example, are expected to refrain from all forms of violence
and injury against any living thing, in accordance with the principle of ahimsa. However, such
non-public doctrines are unsuitable grounds for justifying exercises of authority or power to
society at large. So it would be inappropriate to advocate a legal prohibition or attempt to enforce

5
This is known as weak consensus view of the structure of public reason; see Quong 2013, §5.
6
Although I will call them “doctrines,” no requirement of sophistication or systematization is implied. A child
typically has at least the beginnings of a worldview, with both public and non-public contents, despite a lack of
sophistication and systematization. This much is sufficient for a child to accept moral, political, religious, and/or
philosophical “doctrines” in the attenuated sense I intend.

7
a strict social norm against eating meat, by appeal to ahimsa in a society characterized by a
plurality of “reasonable” doctrinal perspectives.
After all, “reasonable” people can “reasonably” disagree about what counts as a
satisfactory justification. Public reason theory’s insight is that by justifying an exercise of authority
(or power) only by appeal to suitably public reasons, addressers of public reasons ensure their
addressee has the right kind of reason (though perhaps an inconclusive one from the addressee’s
point of view) to accept it (or, on some accounts, no conclusive reason of the right kind to reject
it). Rawls describes the norms of public reason as being twofold:
1. The Publicity Constraint: Individuals (or communities) must be able to explain to one
another how the exercises of authority (or power) they support and for which they advocate,
such as political agendas, public policies, legal reforms, or social norms, can be justified
by suitably public reasons;
7

The publicity constraint tells you that when you make your contribution to justifying exercises of
power or authority, you are to narrow your focus on advancing those reasons you could expect
any “reasonable” disagreeing party to take seriously, whatever their personal moral, political,
religious, or philosophical commitments might be, and to hold others to this same standard. There’s
no prohibition on bearing witness to whatever non-public considerations you may have for
favoring a particular position,
8
but justification must ultimately rest on public reasons. The second
norm of public reason is:
2. The Fair-mindedness Constraint: Individuals (or communities) must be willing to listen to
others and be fair-minded in deciding when to accommodate their views.
9

7
See Rawls 1993, 217.
8
For more on this idea, see e.g., Schwartzman 2011 and 2012.
9
See Rawls 1993, 217.

8
The fair-mindedness constraint tells you that when you listen to others’ contributions to the
justification of exercises of power or authority, you’re to be open to a broader set of considerations
than those you would individually favor, were there no need to coordinate with and accommodate
others.
Public reason liberals theorize about what justifications suitably idealized persons would
accept for exercises of authority (or power), in accordance with the norms of public reason. These
idealized persons are referred to as the “reasonable.” You’ve probably noticed, however, that I
have yet to tell you who the “reasonable” are. What is it that qualifies someone or their view(s) as
being “reasonable” rather than “unreasonable”? Public reason theorists have different ways of
precisifying the notion of the “reasonable.” Rawls assumed the “reasonable” to be practically
rational,
10
and this idealization has gone uncontested. Beyond this, some have advocated primarily
moral conditions on the idealized “reasonableness” of persons,
11
while others have advanced
purely epistemological accounts of the theorized “reasonableness” of disagreements.
12
But I think
the relevant kind of “reasonableness” places moral and epistemic requirements on persons and
their disagreements. I argue moral, practical, and intellectual virtues are all indispensable to the
job description of the “reasonable.” Considerable attention has been given to the moral
idealizations of “reasonable” persons, but with some notable exceptions,
13
less attention has been

10
See e.g., Rawls 2001, 6–8.
11
See Wenar 1995, Klosko 2000, Kelly & McPherson 2001, and (tentatively) Nussbaum 2011.
12
See Thomas Nagel’s (1987) defense of liberalism on the basis of epistemic restraint, Brian Barry’s (1995; cf. Wall
1996, 94–100) appeal to skepticism about conceptions of the good, and more recently, separate attempts by Fabienne
Peter (2013) and Federica Liveriero (2015) to ground a commitment to political liberalism in the epistemology of peer
disagreement.
13
Chief among them, the highly original and trail-blazing work of Gerald Gaus.

9
devoted to making epistemological sense of “reasonable” disagreement, especially among those
closest followers of Rawls.
David Enoch is among the most prominent thinkers to have considered the prospects for
making sense of the normative epistemological implications of public reason theory in general,
and “reasonableness” in particular. Enoch questions whether we can make epistemological sense
of whatever it is that “reasonable” people believe about the “reasonableness” of people who reject
their most deeply held moral, political, religious, or philosophical commitments, and/or the
“reasonableness” they ascribe to the competing commitments of disagreeing parties:
Take the reasonable Catholic. Presumably, she believes in Catholicism. Perhaps
she also believes that other religions have many advantages to them, or that atheists
too can be good people, and that she should interact with all these others as free and
equal, and so on. But she also believes, I take it, that Catholicism is the one true
religion. If no reasonable person can consistently believe that, then no reasonable
person can remain committed to Catholicism, as this is presumably a part of the
very content of Catholicism; if it isn’t, feel free to pick an example of a
comprehensive doctrine of which the analogous claim is a part. What, then, does
this reasonable Catholic believe? What does she believe, in particular, about
(reasonable) non-Catholics?
14

So “reasonable” persons (or communities) have “reasonable” commitments to certain non-public
moral, political, religious, and/or philosophical claims. “Reasonable” persons simultaneously
regard at least some disagreeing parties and/or their incompatible commitments as “reasonable”
too. But when a reasonable agent A thinks of another agent B as being reasonable, despite some
deep disagreement between A and B over religious, moral, political, or philosophical doctrine,
what does this amount to?
15

14
Enoch 2017, 136.
15
Enoch 2017, 137.

10
1. The Job Description of the “Reasonable”
Before I attempt to address this and related questions, however, I’ll begin by providing an overview
of the normative and explanatory work, with which the concept of “reasonableness” is tasked. For
public reason theory’s semi-technical, stipulative concept of “reasonableness” has a complicated
job description. “Reasonableness” is predicated (among other things
16
) of persons, communities,
their moral, political, religious, and/or philosophical doctrines, and their disagreements over these
same matters. This is a semi-technical, stipulative use of the term, but my initial pre-theoretic
references
17
to what reasonable or unreasonable people would or wouldn’t accept show that the
term ‘reasonable’ is in the right conceptual ballpark. As used by public reason theorists, the term
restricts the set of those to whom exercises of authority (or power) ought to be justified. Public
reason theorists maintain that the existence of this set is entailed by a true moral requirement of
public justification for the exercise of authority or power. Once the shape of this set is understood,
‘reasonable’ is used as a name for this set. However, the concept isn’t supposed to be reserved
only for the theorist’s use. The concept defines a set of people, but the people in that set deploy
their own conceptions of the “reasonable” in making judgments about which people to trust, and
which views to accommodate.
Following Ralph Wedgwood’s work on the nature of rationality,
18
I think “reasonableness”
is best understood as a social and political virtue of persons. This virtue is manifested in the
formation of “reasonable” attitudes like epistemically rational beliefs, moral judgments, and in the

16
For an exhaustive list of predications in Rawls’s Political Liberalism (1993), with references, see Wenar 1995, 34,
fn. 3.
17
See its initial uses on page 2 above.
18
Wedgwood 2014, especially section II on the structure of virtue.

11
performance of “reasonable” actions, like practically rational or enkratic choices.
“Reasonableness” is also abstractly manifested by paradigmatically “reasonable” views or actions,
even when the agent in question is an otherwise irrational, incompetent, biased, bigoted, or narrow-
minded person. For just as a vicious person may perform a paradigmatically virtuous act without
virtuous motives, an overall “unreasonable” individual may espouse paradigmatically
“reasonable” views or perform “reasonable” actions. So the “reasonableness” we predicate of
persons picks out some fairly stable disposition of character to act in certain ways, while the
“reasonableness” we predicate of non-public moral, political, religious, and philosophical
doctrines reflects our judgment that they are typically formed and maintained by “reasonable”
persons. The “reasonableness” we predicate of communities organized around a shared moral,
political, religious, or philosophical perspective, may either pick out the abstract “reasonableness”
of their shared doctrinal perspective, despite the fact that individual members of the community
may not individually manifest any “reasonableness” in holding the view, or reflect the judgment
that members of this community tend to be “reasonable” persons.
It is important to observe that the “reasonableness” of a particular view or doctrinal
perspective does not entail its publicity, and the “reasonableness” of a person does not entail the
publicity of all their commitments. A non-public doctrine may be “reasonable,” insofar as a
plausibly “reasonable” person could manifest their “reasonableness” in espousing it. But only
unanimous consensus among all the “reasonable” entails the publicity of a view or doctrinal
perspective. For example, particular commitments like the principle of ahimsa or Catholic
hylomorphism are plausibly “reasonable,” insofar as plausibly “reasonable” persons can espouse
these views. Some version(s) of the broader doctrinal perspectives of Jainism and Catholicism are
likewise plausibly “reasonable.” But neither these particular doctrines, nor the broader doctrinal

12
perspectives of Jainism and Catholicism, respectively, are suitably public. By contrast, the
“unreasonableness” of a particular view, like a racist attitude toward a particular ethnicity, or of a
doctrinal perspective, like Nazism, does entail its non-publicity. For no plausibly “reasonable”
person can manifest their “reasonableness” by espousing such views or doctrinal perspectives, and
it follows that they cannot be objects of overlapping consensus among the “reasonable.”
But given that there are plausibly “reasonable” Catholics and Jains, there’s room for lots
of disagreement among the “reasonable.” For our purposes, the disagreements of special interest
are doctrinal disagreements between “reasonable” parties over the truth of one or more party’s
non-public doctrine(s). The fact of “reasonable” pluralism entails that at least many such doctrinal
disagreements are “reasonable” in large, contemporary, free liberal-democratic societies.
Pre-theoretically, reasonableness is a scalar notion, as evidenced by ordinary usage of
phrases like ‘more reasonable’ and ‘less reasonable.’ It’s possible the stipulative notion of
“reasonableness” employed by public reason theorists likewise admits of degrees. In that case, the
important question is whether someone, some community, some view, or some disagreement is
sufficiently “reasonable” above a threshold. Alternatively, “reasonableness” could be a binary
range property: binary in that it either obtains or it does not, and hence does not vary by degrees,
but its obtaining depends on possessing some other, scalar property (or properties), obtaining
within a specified range.
19
But whether “reasonableness” is itself scalar, or instead a binary range
property depending on other scalar properties, the threshold of sufficient “reasonableness”, or
range in which “reasonableness” obtains, is set relative to several requirements.

19
Carter 2011, 548.

13
1.1 The “Reasonable” are Committed to the Publicity Constraint
Firstly, the “reasonable” constitute public reason’s constituency, in the sense that they’re expected
(in both the normative and predictive senses) to adhere to both of the norms of public reason,
including the publicity constraint. “Reasonable” individuals accept
20
that some moral, political,
religious, or philosophical commitments, while personally important to them, are nonetheless
unsuitable grounds for justifying exercises of authority (or power) to society at large, because
being too controversial, idiosyncratic, or sectarian, they are rejected by other plausibly
“reasonable” members of society. The “reasonable” might give public voice to their own non-
public reasons for advocating some exercise of authority or power e. But whenever e’s justification
is a matter of doctrinal disagreement among parties subject to e, or e is to be carried out on behalf
of such parties, the “reasonable” are expected to provide a suitably public, common ground
justification for e.
To illustrate why adherence to the publicity constraint is plausibly a necessary condition
on the “reasonableness” of persons, let’s contrast it with what would seem prima facie to be a
strong competitor to public reasons: truth. Consider a police officer, Constable Duke, who orders
you to go back the way you came, offering only as justification his Catholic understanding of the

20
Following Leland and van Wietmarschen 2012, 726, n. 7, I use ‘acceptance’ as a catch-all term for various pro-
attitudes, so as to stay neutral on disputes about what types of attitudes are expressed by normative and evaluative
judgments, and the contents thereof.

14
divine right he wields as a duly appointed public servant.
21
Suppose further, that his version of
Catholicism is the true religion and that Duke knows this.
22
Duke is in fact morally and
epistemically justified in acting as he sees fit, in virtue of being a divinely instituted authority. But
non-Catholic that you are, it seems that Duke fails to justify his exercise of authority to you.
Moreover, it seems in some way unfair to expect you to accept as justification the dictates of
Catholicism (even though we’re assuming them to be correct!).
Contrast this with a case where a Constable Doak orders you to stay back to make way for
the imminent arrival of the President’s motorcade. As it happens, Doak is mistaken. The President
has decided to cancel his appearance on Jimmy Kimmel Live! due to his unpopularity and his fear
of being publicly humiliated, but Doak has yet to be informed that the motorcade has changed
course. Even so, concern for the orderly and safe movement of a President—even a highly
unpopular one for whom you did not vote—seems like the right kind of justification for
constraining your behavior as a pedestrian. In fact, even if the person commanding you to get back
was just some busy-body, rather than a police officer specially tasked with helping to protect the
President, and you were engaging in an act of civil disobedience to protest the President’s
decisions, the busy-body’s concern for the orderly and safe movement of the President would still
be the right kind of reason to offer as justification, even if you didn’t in the end find it conclusive.
So intuitively, even though Constable Doak is mistaken, it seems to me that she successfully

21
Perhaps Officer Duke believes that the Apostle Paul’s words apply, granting him the authority to do almost anything
he judges to be good in the performance of his duties: “For the authorities are not a terror to good conduct, but to bad.
Would you like to live without being afraid of the authorities? Then do what is right, and you will receive their
approval. For they are God’s servants, working for your good. But if you do what is wrong, you should be afraid, for
it is not without reason that they bear the sword. Indeed, they are God’s servants to administer punishment to anyone
who does wrong. Therefore, it is necessary for you to be acquiescent to the authorities, not only for the sake of God’s
punishment, but also for the sake of your own conscience” (Romans 13:3–5; International Standard Version).
22
This is a version of David Estlund’s Pope case; see Estlund 2008, 5.

15
justifies her command to you. To the extent that you share my intuitions in this pair of cases, they
seem to demonstrate that truth is neither sufficient nor necessary for justifying the exercise of
authority (or power) to those subject to it, whereas the public reason advanced in the second case,
though mistaken, seems appropriate. If correct, this tells strongly in favor of adherence to the
publicity constraint being a necessary condition on the “reasonableness” of persons.

1.2 The “Reasonable” accept “Reasonable” Pluralism
The “reasonable” also form a diverse constituency for public reason, with (at least) many
competing and incompatible moral, political, religious, and philosophical doctrines being
represented among them. At least many of these diverse doctrines are supposed to be “reasonable”
as well, and so too are the deep doctrinal disagreements between their adherents. Furthermore, this
formal desideratum, that some sufficient degree of pluralism and disagreement be deemed
“reasonable,” also needs to vindicate enough of our pre-theoretic intuitions about which views are
reasonable and which ones are not, for example, that some interpretation of post-Vatican II
Catholicism is a reasonable set of doctrines, that some version of humanistic atheism is reasonable,
and that tin-foil-hat conspiracies are not.
This fact of “reasonable” pluralism is central to public reason liberals’ aim of bringing
people with allegiances to mutually incompatible and competing worldviews to a consensus on a
shared set of reasons or principles, which can justify a workable political agenda and legitimize a
broad range of exercises of authority necessary for sustaining a stable liberal-democratic society.
If “reasonableness” can’t be developed so as to imply that a sufficient diversity of doctrinal
disagreement is “reasonable,” then even if everyone in a liberal society were to agree that they
ought to find common ground with others on shared principles, they could still disagree so much

16
about who counts as “reasonable” that they would recognize no obligation to justify their favored
exercises of authority (or power) to many or most other members of society subject to them.
To illustrate, consider Doc,
23
an atheist who happens to be a uniqueness theorist about
epistemic rationality:
UNIQUENESS: Given a body of total evidence, there is a unique doxastic state that it
is rational to be in.
24

Doc thinks, given the total body of evidence for atheism, epistemically rational and hence
“reasonable” people assign atheism a credence of 1.
25
He agrees that consensus among the
“reasonable” is required for achieving and maintaining the just exercise of political authority.
However, Doc fails to recognize those outside his relatively small like-minded atheist community
(who all and only have a credence of 1 in atheism) as “reasonable.” If Doc could be counted among
the “reasonable,” this would be disastrous for public reason liberalism. For we ordinarily think
such insular sectarian views are ipso facto unreasonable. If the conception of “reasonableness”
succeeds only in adding a new evaluative dimension along which most disagreeing others are
deemed “unreasonable,” public reason theory ends up only exacerbating existing sectarian
divisions in a pluralistic society.
Fortunately, public reason’s conception of “reasonableness” screens out people like Doc,
and the acceptance of reasonable pluralism is the mechanism whereby this is accomplished. One
has to accept the fact of reasonable pluralism to count as “reasonable.” One must accept that at
least many people are “reasonable,” and that at least many doctrinal disagreements are also

23
Dr. Richard Dawkins, perhaps.
24
Hedden 2015, 468.
25
Cf. Leland and van Wietmarschen’s discussion of their “rationalist fundamentalist”: 2012, 730.

17
“reasonable,” and this informs the willingness of the “reasonable” to make fair accommodation of
a diversity of doctrinal perspectives. Public reason theory provides no empirical proof of the fact
of reasonable pluralism, because the fact of reasonable pluralism isn’t an empirical claim like
Hobbes’ fact of scarcity. Instead, the fact of reasonable pluralism is, in part, a normative claim.
Some doctrinal commitments are compatible with it, and some are not. If you accept a moral or
religious doctrine incompatible with liberalism, you’ll think liberalism is the worse for that reason.
Likewise, if you accept any non-public doctrine incompatible with the fact of reasonable pluralism,
you’ll think public reason liberalism is the worse for that reason.
26
If, however, you adhere to the
fair-mindedness constraint, you’re willing to listen and give fair consideration to accommodating
a significant diversity of doctrinal perspectives.

1.3 The “Reasonable” Are Epistemically Competent
Leif Wenar, following Rawls, argues that the “reasonable” must possess the intellectual powers of
judgment, thought, and inference. But the possession of these epistemic capacities alone isn’t
sufficient for “reasonableness.”
27
For the possession of these capacities is a matter of facts entirely

26
Notice, however, that Doc isn’t forced into his position by his epistemological commitment to UNIQUENESS alone.
Instead, it’s the combination of UNIQUENESS and his exceedingly demanding standard of rationality for evaluating the
“reasonableness” of others’ views, requiring perfectly rational credences, which leads him to disqualify so many of
his fellow citizens from the constituency of the “reasonable.” He could instead accept a threshold view of
“reasonableness,” on which those attaining a sufficient, though sub-optimal, degree of global epistemic rationality,
with respect to a number of socially and politically relevant subject matters, count as reasonable. For rationality
plausibly admits of degrees (see Wedgwood 2014, section V). This view makes room for UNIQUENESS theorists to
accept the fact of “reasonable” pluralism, so less conceptual space is ruled out by its acceptance than it might initially
appear.
27
Wenar 1995, 37; see Rawls 1993, 81; 15–35.

18
internal to the epistemic subject, and such facts cannot guarantee epistemic success in the external
world. For even if the “reasonable” aren’t expected to get things entirely right, in order to be trusted
as competent fellow cooperators on fair terms, the “reasonable” must be in sufficient contact with
facts relevant to fair cooperation and social coordination.
28

To see why, consider Dek.
29
Unlike Duke, Dek is committed to observing the publicity
constraint. Unlike Doc, Dek is also committed to the observing the fair-mindedness constraint too.
Dek is discerning in his decisions about which doctrinal perspectives to accommodate and which
ones to ignore or reject, reflecting his minimally morally decent intentions, but where Doc is
rigidly narrow-minded about whom to accommodate, Dek embraces the fact of “reasonable”
pluralism, believing that at least many of those with whom he has some deep doctrinal
disagreements are nonetheless “reasonable.” Dek, however, is epistemically incompetent. Dek
suffers from confirmation bias, fails to update his beliefs in light of new evidence, fails to resolve
apparent inconsistencies in his commitments, and is generally dismissive of any consideration that
would challenge his worldview. Assuming we can coherently conceive of such an epistemically
incompetent person successfully conforming to the norms of public reason, his success could only
be a function of sheer luck. For while Dek might have the best of intentions, only by an incredible
coincidence could he be sufficiently in touch with relevant facts to make good on them. But since
“reasonableness” is a virtue, its manifestation is incompatible with being merely lucky.
Furthermore, as Dek gains notoriety for his epistemic incompetence, this undermines his ability to
assure others of his trustworthiness. And it’s difficult to see how the “reasonable” can collectively

28
See Wedgwood 2014, 327.
29
A popular shorthand for “Don’t Even Know.”

19
succeed in achieving a stably just equilibrium, unless they have a better grip on reality than Dek.
Therefore, “reasonable” persons must satisfy some minimal standards of epistemic rationality and
inference, while actively seeking out evidence and responding to it appropriately, in order to ensure
they’re sufficiently in touch with facts relevant to fair cooperation and effective coordination with
“reasonable” others to be successful in these endeavors.

1.4 The “Reasonable” are Sincere
One final idealization of the “reasonable” remains. “Reasonable” persons may disagree about
whether or not a public reason is conclusive, but when the “reasonable” address public reasons to
one another for the purposes of justifying exercises of authority or power, they must do so
sincerely. That is, the “reasonable” person who addresses a public reason (or set of reasons) R as
sufficient justification for an exercise of power or authority e is expected to (a) believe R is public,
and (b) believe that R provides sufficient reason(s) for justifying e.
30

To illustrate why sincerity is a further necessary condition on the “reasonableness” of
persons, consider Deke. Deke thinks the state should ban transgendered persons from using public
bathrooms not corresponding to their legal sex, insisting they should be legally required to use the
bathroom corresponding to their sex assignment at birth. Deke’s motivating reasons for favoring
this position derive from cis-normative religious assumptions about God’s intentions for Creation.
But Deke realizes his motivating reasons are controversial, sectarian, and hence non-public. So he

30
Schwartzman 2011, 385; cf. Rawls 1993, pp. 226, 242, 253, 390; 1999, pp. 578, 581, 585. Notice that this
requirement permits the “reasonable” to sincerely address insufficient or inconclusive reasons R for an exercise of
power or authority e to their addressees in the course of public reasoning, so long as they don’t present R as providing
sufficient or conclusive justification for e.

20
frames his support for the bathroom law in terms of “public safety concerns,” which he knows to
be baseless. Whenever he publicly advocates for the law, he appeals to a duty to protect children,
especially girls, from predatory adults, especially men, who would exploit a blanket permission on
using the bathroom of one’s choice to prey sexually on the vulnerable. He does this even though,
let us suppose, he knows that transgendered persons are, in fact, far more likely to be the victims
of an assault in a public bathroom than the perpetrators.
In this case, Deke games the publicity constraint by advancing an ostensibly public reason
for his favored draconian legal regime, which he does not in fact believe, since he knows it to be
false, and a fortiori does not believe it to be suitably public. For while real threats to public safety
are conclusive reasons of the right kind for enacting measures against them, insofar as Deke
disingenuously co-opts the language of “public safety concerns” to advocate for his favored law,
he is deceptive.
Suppose Deke is forced to admit that there are no legitimate public safety reasons to legally
require transgendered persons to use public bathrooms corresponding with their sex assignment at
birth. So he retreats to another ostensibly public justification for the law: respect for religious
freedom. Concern for preserving religious freedom is a suitably public consideration to advance
in defense of an exercise of power or authority, and we’ll assume that the private practice of a cis-
normative religion deserves protection from legal interference. Even so, Deke knows that his
favored legal measure isn’t justified by a concern to protect religious freedom, because rather than
protecting the rights of religious associations to make their own policies about bathrooms on their
private property, the proposed law seeks to impose their religious views on others everywhere in
public. Still, even though Deke realizes this, he hopes others will be persuaded by his sloppy
reasoning anyway.

21
In this case, Deke games the publicity constraint by being insincere in a slightly different
way, for in this case he doesn’t advance a putative public reason that he knows to be illegitimate,
but a true public reason which he knows not to justify his proposal. In both cases, Deke attempts
to exploit the fair-mindedness of “reasonable” others and muddy the epistemic waters for those
seeking evidence relevant to the proposed law’s merits. For those competent enough to see through
Deke’s dishonest public reasoning, his conduct undercuts his standing as a trusted fellow
cooperator, and if he manages to deceive enough people, it can significantly erode social
confidence in the “reasonableness” of others. So whether Deke fails to (a) believe his proffered
justification is public, as in the former case, or fails to (b) believe that his legitimately public
provides sufficient justification for his proposal, his insincerity violates the spirit of the publicity
constraint. Therefore, sincerity is a necessary condition on the “reasonableness” of persons.
31

1.5 A Full Job Description for the Concept of the “Reasonable”
Let’s review the full job description for public reason theory’s semi-technical, stipulative concept
of the “reasonable.” Firstly, the “reasonable” accept and adhere to the norms of public reason:
1. The Publicity Constraint: Individuals (or communities) must be able to explain to one
another how the exercises of authority (or power) they support and for which they
advocate, such as political agendas, public policies, institutions, and social norms, can
be justified by suitably public reasons; and,

31
To be clear, I don’t think public reasons are limited to the subset of an agent A’s motivating reasons that are part of
the overlapping consensus among the “reasonable.” So long as a reason (or set of reasons) R is suitably public, A can
sincerely address R to others as a reason justifying an exercise of authority e, whether or not A herself is motivated by
R. For reasons of transparency, A may wish to go beyond the requirements of the publicity constraint, and explain
why she is not personally motivated by R, because this kind of disclosure can promote trust. But so long as she thinks
R is a legitimately public consideration, there’s no reason for her to think others cannot remain “reasonable,” while
being motivated by R, and since R affords her a means to justifying e to such persons, “reasonableness” does not
prevent her from doing so. For a dissenting position, see Schwartzman 2011.

22
2. The Fair-mindedness Constraint: Individuals (or communities) must be willing to
listen to others and be fair-minded in deciding when to accommodate their views.

Secondly, the “reasonable” accept the fact of reasonable pluralism, which ensures that they regard
at least many doctrinal disagreements to be “reasonable,” and stand ready to make fair
accommodations of new and unfamiliar views, so long as they too are “reasonable.” Thirdly, the
“reasonable” are epistemically competent, satisfying some minimal standards of epistemic
rationality and inference, while actively seeking out evidence and responding to it appropriately,
in order to ensure they’re sufficiently in touch with facts relevant to fair cooperation and effective
coordination with “reasonable” to be successful in these endeavors. And finally, the “reasonable”
are sincere, believing their contributions to public reasoning to be genuinely public, and presenting
them as providing sufficient justification for an exercise of power or authority only if they believe
them to do so.
Clearly, public reason theory’s concept of “reasonableness” has a very complicated job
description. But I think non-theorists have some intuitive conception of the social and political
virtue this concept predicates of members of a free pluralistic society. Like any interesting virtue,
it’s difficult to provide a set of necessary and sufficient conditions on its obtaining, but the above
set of necessary conditions are illuminating enough for our purposes, restricting the set of those to
whom exercises of authority (or power) ought to be justified. And now that the shape of this set is
sufficiently established, we’re almost ready to think more carefully, not only about how public
reason theorists employ the concept, but how the people in that set deploy their own conceptions
of “reasonableness” in making their own judgments about which individuals and communities are
reasonable enough to trust as fellow cooperators in society, and which views are sufficiently
reasonable to merit accommodation. For it is as much with these ascriptions of “reasonableness,”

23
as with the theoretical concept, that Enoch concerns himself, challenging public reason theorists
to make epistemological sense of both.
First, however, we must take a brief detour through some particulars of liberal anti-
perfectionism. For the task of answering Enoch’s challenge might appear to be made more difficult
by the liberal’s anti-perfectionist’s additional theoretic constraint of liberal neutrality. Indeed,
Enoch seems to think the political liberal’s commitment to liberal neutrality creates a distinct
problem for answering his challenge to the epistemological coherence of public reason theory. So
in the next section, I briefly lay out what liberal neutrality is and how it fits into liberal anti-
perfectionism.
2. Liberalism Anti-Perfectionism: The Overarching Problem and Two
Guiding Aims
Liberalism’s foundational commitment is to the freedom and equality of persons (or citizens), and
its guiding conviction is that the exercise of political authority or power is legitimate only when it
can be publicly justified. Liberalism anti-perfectionism further claims that at least much of the
moral, religious, political, and philosophical pluralism, which characterizes large contemporary
free societies is, in both moral and epistemological normative senses, “reasonable.” Liberal anti-
perfectionism, therefore, maintains that the legitimate exercise of political authority or power
cannot be grounded in justifications or reasons which privilege any particular moral, religious,
political, or philosophical view of the good life. Instead, the liberal state should restrict itself to

24
establishing fair conditions for citizens and subjects to pursue their own conception of the good
life.
32

2.1 The Overarching Problem Arising from the Special Circumstances of Politics
Liberal anti-perfectionism aims to solve what it sees as a special overarching problem, created by
the conjunction of:
(A) Human beings’ need for a social cooperation to live a valuable life; and
(B) Our deep reasonable disagreements about what it is to live a valuable life and how best to
secure it (whatever it is) through cooperation.

Call (A) and (B) the special circumstances of politics. Given these special circumstances, we
cannot simply privilege one set of deep normative assumptions to settle all our reasonable
disagreements about the good life and social cooperation, because the problem is that we disagree
precisely about those things, and these disagreements present a formidable obstacle to social
cooperation.
In order to solve this overarching problem, liberalism’s primary aim is to justify the
political legitimacy of a system of cooperation enforced by the state’s exercise of power and
authority, to a subset of those subject to this system: the “reasonable.” But this justificatory project
is difficult in large contemporary free societies characterized by (B), a “reasonable” pluralism of
competing and incompatible worldviews. So liberal anti-perfectionism’s secondary aim is to
provide an articulation of liberal theory that, as much as possible, avoids needless moral, political,

32
Quong 2011, 2–3; see Macedo 1990, 78 and Waldron 1993, 36–37.

25
religious, and philosophical controversy, in order to facilitate the justificatory project. It will help
to begin by explaining this secondary aim first.

2.2 Anti-Perfectionist Liberalism’s Secondary Aim of Neutrality
Hobbes presented his political theory as a response to the fact of scarcity, and it’s natural to assume
that liberal anti-perfectionism is framed as a response to the special circumstances of politics in
the same way. But whereas there’s a fact of the matter about whether there are finite resources,
independent of whether there’s a theorized solution to the political challenge it presents, the special
circumstances of politics aren’t presented as an empirical claim independent of liberalism. Unlike
the fact of scarcity, the special circumstances of politics are, in part, evaluative claims, making
moral and epistemological demands on those who accept them. Some anarchists and libertarians
reject the idea that social cooperation is needed in order to live a valuable life, while theocrats
characteristically reject the reasonableness of disagreements with the religious view they wish to
become politically dominant, but (A) and (B) are both part of the content of liberalism, so to the
extent that someone doesn’t accept either claim they’re not committed to the project of
liberalism.
33

33
Insofar as someone doesn’t accept either (A) or (B), they’re not committed to the liberal project of living together
peaceably in a pluralistic society. Theocrats and fascists reject the reasonableness of disagreements with their view of
the good life, and would like nothing more than to impose their view on the rest of society. Such rejections of (B) are
illiberal and unreasonable. Unless the illiberal present a more pressing threat, Rawlsian political liberals typically
think they should be contained, ignored, or converted where possible. But what about those, including some
libertarians and anarchists, who accept (B) but reject (A), and for this reason, would prefer to pursue their conception
of the good life in peace and relative isolation from coerced social cooperation? Such persons may not be liberals, but

26
The normative content of the special circumstances of politics makes them at least
somewhat normatively controversial. To that extent, it may appear to be in tension with liberal
anti-perfectionism's aim of neutrality. But as Joshua Cohen explains:
Liberalism has always been, inter alia, a political outlook, defined by an emphasis
on personal freedom, religious tolerance, open inquiry, the rule of law, social
mobility, and, in its modern formulations, democratic politics. …[Anti-
perfectionist] liberalism aims to free the formulation of liberalism as a political
outlook, so far as possible, from that wider set of philosophical and religious
commitments, and thus to “put no unnecessary obstacles in the way of…affirming
the political conception.”
34

A liberal theory is anti-perfectionist, just insofar as it succeeds in formulating the liberal outlook
on central liberal-democratic values of personal freedom, religious tolerance, equality of the sexes,
the rights of minorities, freedom of the press, the rule of law, social mobility, and democratic
politics, in a way that avoids implicating liberalism in any other moral, political, religious, or other
independent, long-standing philosophical controversies that are not essential to liberalism’s
content.
Suppose, for example, that a particular version of liberalism makes the following aesthetic
assumptions. Firstly, suppose the artisanal self-expression involved in a florist’s arranging of a
bouquet is so integral to her self-understanding that its free expression is beyond the legitimate
reach of anti-discrimination laws mandating the equal treatment of persons in the provision of
publicly available services. Secondly, suppose short-order cooking enjoys no similarly weighty

their preference for isolation is compatible with others living in liberal systems of social cooperation that respect the
central liberal-democratic values, so they plausibly deserve fair accommodation.
34
Cohen 2009, 6–7; the embedded quotation is from Rawls 2001, 37.

27
claim to freedom of creative self-expression. So while a florist can refuse to arrange flowers for
interracial or same-sex couples, on religious grounds, directing them instead to help themselves to
stock arrangements in a cooler, a short-order cook cannot similarly segregate the services at his
lunch-counter for religious reasons.
35
If it really were the case that these kinds of controversial
aesthetic doctrines were part of the content of liberalism, then liberalism would obviously be self-
defeating: there’s no way that a political theory, taking sides in these kinds of aesthetic
controversies, could address the challenge posed by reasonable pluralism. For these strong
aesthetic assumptions are reasonably disputed from the perspective of many worldviews.
People can reasonably think that certain kinds of creative self-expression are integral to
their self-understanding, to living the good life, and may regard them as sacred, while other
creative acts are judged menial. But imposing these kinds of idiosyncratic distinctions on others in
public life is precisely where they become illiberal aesthetocratic (and, in this case, theocratic)
impulses. Moreover, such assumptions cannot be part of the content of liberalism: if they were,
they would deliver the verdict that a morally upstanding person, who reliably cooperates on fair
terms with her fellow citizens, treating them as free and equal, is nonetheless illiberal for being
utterly pragmatic and aesthetically impoverished in the way she approaches public life. Anti-
perfectionists think this an unacceptably doctrinaire verdict for a truly liberal theory to deliver.

35
Kerri Kupec, a lawyer for florist Barronelle Stutzman, owner of Arlene’s Flowers, argued that, because Stutzman
is an artist, she ought not to be compelled to create an arrangement specifically for the nuptials of gay couple, Curt
Freed and Robert Ingersoll, as it goes against Stutzman’s religious beliefs. Kupec alleged that absent any special
protection for artistic expression, “a gay singer could be forced by the government to perform at a religious conference
that is promoting marriage as a man-woman union” (King 2016). Kupec presumably wouldn’t want to extend this
argument to protecting racially segregated lunch-counters.

28
2.3 A Political Conception of Liberalism
Liberals anti-perfectionists distinctively think that any theory deserving the name of ‘liberalism’
shouldn’t be in the business of making such narrow prescriptions about what it is to live the good
life. So anti-perfectionism aims to articulate liberalism in a way that avoids, as much as possible,
building in or relying on any such needlessly controversial assumptions, and one important
theoretical tool at the anti-perfectionist's disposal is a political conception, which must satisfy the
following condition:
LIBERAL NEUTRALITY: a political conception of a target concept cannot advance any
specific controversial aesthetic, epistemological, ethical, metaethical,
metaphysical, or religious doctrines or values beyond what is already implicit
to the public overlapping consensus.
36

LIBERAL NEUTRALITY is a requirement of a kind of theoretical austerity. For by avoiding needless
controversy, a political conception of a target theoretical notion enjoys compatibility with a wide
variety of moral, political, religious, and philosophical perspectives, “because of its restrained
posture toward values outside its provenance.”
37
Liberal anti-perfectionism aims at providing a
political conception of liberalism itself, so LIBERAL NEUTRALITY requires the liberal theorist to
articulate their theory within the confines of the publicity constraint.
Drawing again on Rawls, Wenar tells us that a political conception:
a. is freestanding, presented independently of any controversial, idiosyncratic, or sectarian
commitments;
b. articulates only political values, but does not claim that these values are more important
than other values; and

36
Cf. Rawls 1993, 10.
37
Wenar 1995, 33.

29
c. is independent of long-standing controversies in philosophy; in particular it has no
distinctive metaphysical doctrine of the person.
38

In Rawls’ own words, a political conception is, “a module, an essential constituent part, that fits
into and can be supported by various reasonable [perspectives or worldviews] that endure in the
society regulated by it.”
39
So a political conception of a target notion has some determinate non-
negotiable content, but it is also importantly and explicitly indeterminate, leaving room for those
who wield the conception to fill it out with reference to their own understanding of the target
notion. Therefore, a political conception can be endorsed and fleshed out from many disparate
points of view. Liberal anti-perfectionism succeeds in its aim of avoiding needless philosophical
controversies just in case it is a political conception of liberalism, satisfying the principle of
LIBERAL NEUTRALITY.
For illustration, consider the notion of a national identity, an important organizing principle
for the model nation-state. Appeals to national identity typically privilege one packet of social
values, social experience, and/or biological ethnicity over others, being wielded both in defense of
the privileged identity and against incompatible identities. In reality, national identities are vague,
indeterminate, and always subject to negotiation and revision, but they aren’t entirely without
content. The more determinate a nationality identity, the more exclusive it is: think, for example,
of the highly specified Nazi German aryan ideal of nationality. Approaching the opposite extreme,

38
Wenar 1995, 33. For condition 1, see Rawls 1993, 10–13, 140-141, 144-45; for condition 2, see xx; 157; and for
condition 3, see xix–xx; 10; 29, n. 31; 154.
39
Rawls 1993, 12; as quoted in Wenar 1995, 33.

30
the state of Canadian national identity has recently invited explicit characterizations as an
“incomplete” or “weak” national identity, with suggestions that Canada may be the world’s first
post-national state.
40
But while Canadian national identity may be logically weak or incomplete,
in that it doesn't entail much substantive content beyond the central liberal-democratic values,
perhaps a love some stereotypically Canadian things, like hockey, and some experience of life
within its vast territory, it’s proven surprisingly socially robust. For these observations of its
incompleteness followed nearly a decade of efforts by a previous government to stoke the embers
of Canadian nationalism, weaken the rule of law and the capacity for open scientific inquiry, and
attack the values of tolerance, diversity, and inclusion. Indeed, because of its logical weakness
Canadian national identity has proven resistant to efforts to precisify its content with an official
set of distinctly Canadian values. In this way, the Canadian national identity looks very much like
a political conception of national identity: apart from commitments to core liberal-democratic
values, the notion of Canadian identity is highly indeterminate and compatible with many widely
disparate lived experiences within Canadian society.
41

Given that the special circumstances of politics are normatively controversial, their
inclusion in a theory of liberalism might initially seem to be at odds with the aim of articulating a
political conception of liberalism. But they are essential to the content of liberalism, so their
inclusion doesn’t amount to a needlessly controversial inflation of the same. Some philosophical
views are compatible with reasonable pluralism and a commitment to social cooperation, and some
are not. If your prior philosophical commitments are incompatible with the fact of reasonable

40
See Foran 2017.
41
Although Canada’s first nations, northern peoples, Métis, and Québecois have a more fraught relationship with
Canadian identity.

31
pluralism, you’ll think liberal anti-perfectionism is the worse for that reason. But evaluating a
society’s pluralism as reasonable involves accepting that (at least many of) its citizens have good
reasons for subscribing to their disparate worldviews, and so realistically can only be expected (in
the predictive sense) to adhere to a single shared worldview by the oppressive use of state power.
But to do so would be to abandon the central liberal-democratic values. Insofar as the committed
liberal is convinced about the reasonableness of pluralism, she will be happy to part ways with
anyone who flirts with such illiberality on the basis of abstract philosophical commitments.
So while political liberalism aspires to a high degree of neutrality concerning matters of
moral, political, religious, and philosophical controversy, this neutrality is not absolute. The central
liberal-democratic values remain non-negotiable commitments even on a political conception of
liberalism satisfying LIBERAL NEUTRALITY. Still, political liberalism’s high degree of neutrality
makes it well-suited for addressing its primary aim, to which I now turn.

2.4 Liberal Anti-Perfectionism: Liberal Theory Tailored to Solving the Overarching
Problem
So in order to solve the overarching problem of political legitimacy, created by the special
circumstances of politics, political liberalism’s distinctive contribution is to provide a political
conception of the liberal theory of political legitimacy. In so doing, political liberalism as much as
possible, avoids needless moral, political, religious, and philosophical controversy, in order to
facilitate its primary justificatory aim. For in addressing the liberal theory of political legitimacy
to the “reasonable” in neutral terms, political liberalism tailors itself to justifying the state’s
exercise of power and authority in large contemporary free societies characterized by a
“reasonable” pluralism of competing and incompatible worldviews.

32
As I said at the outset of section 1, liberalism’s guiding conviction is that the exercise of
political power or authority is legitimate only when it can be publicly justified. Public reason
liberalism’s distinctive claim is that it’s illegitimate to exercise political power over (or on behalf
of) citizens on the basis of reasons they could not be expected to “reasonably” accept.
42
Intuitively,
most of us agree that it’s wrong to boss people around without providing some justification, but
what grounds, explains, or justifies the distinctive normative claim of political liberalism? On the
explanation I favor, the normative claim is an implication of the standing or authority one has to
demand a suitable explanation of the state’s exercise of its power over (or on behalf of) oneself, in
virtue of being a citizen of the state.
43
For to be a citizen in a free democratic society, isn’t merely
to be a subject of the state, but a political authority to the state and a source from which it derives
its legitimacy.
Under conditions of reasonable pluralism, consent-based theories of political legitimacy
are of no help to addressing the challenge reasonable pluralism poses for the state’s legitimate
exercise of its authority. Firstly, unanimous consent is impossible to secure under conditions of
reasonable pluralism. Secondly, tacit consent is implausible: few individuals ever get to choose
their citizenship, and the typical infeasibility of emigration undermines the plausibility of the
suggestion that tacit consent to one’s government is secured by voluntarily remaining a citizen. At
the same time, deeper moral stories that could ground the legitimacy of the state’s exercise of its

42
Or more weakly, on the basis of reasons, which they could reasonably reject.
43
Some political liberals disagree with this characterization of the illegitimacy in terms of a recognition of the standing
citizens’ have as such, but I think alternative explanations (e.g., respect for persons, autonomy, civic friendship, etc.)
must rely on deeper moral foundations, and therefore amount to versions of liberal perfectionism. Here I follow
Hartley & Watson 2014 (425), who in turn follow the work of Stephen Darwall, arguing this respect is most plausibly
understood as a kind of recognition respect for fellow citizens qua citizens.

33
authority, by appeal to what some citizens take to be the whole truth, by the lights of their
idiosyncratic, controversial, or sectarian private commitments about morality, politics, or religion,
are likely to be unpersuasive, divisive, or invidious. The problem of reasonable pluralism cannot
be solved just by privileging one comprehensive set of deep normative assumptions to settle all
political questions, because the root of the problem is that we disagree precisely about those things.
What we need is some kind of practice to get around these deep disagreements.

34
II. Enoch’s Challenge to Public Reason Liberalism
Recall David Enoch’s challenge for public reason theorists to make sense of the normative
epistemological implications of their theory in general, and “reasonableness” in particular. Enoch
questions whether we can make epistemological sense of whatever it is that “reasonable” people
believe about the “reasonableness” of people who reject their most deeply held non-public moral,
political, religious, or philosophical doctrines, and the “reasonableness” they ascribe to the
competing commitments of disagreeing parties:
Take the reasonable Catholic. Presumably, she believes in Catholicism. Perhaps she also
believes that other religions have many advantages to them, or that atheists too can be good people,
and that she should interact with all these others as free and equal, and so on. But she also believes,
I take it, that Catholicism is the one true religion. If no reasonable person can consistently believe
that, then no reasonable person can remain committed to Catholicism, as this is presumably a part
of the very content of Catholicism; if it isn’t, feel free to pick an example of a comprehensive
doctrine of which the analogous claim is a part. What, then, does this reasonable Catholic believe?
What does she believe, in particular, about (reasonable) non-Catholics?
44
So, according to public reason theorists, “reasonable” persons and communities have
“reasonable” commitments to non-public moral, political, religious, and/or philosophical claims,
and they simultaneously regard at least some disagreeing parties and their incompatible

44 Enoch 2017, 136.

35
commitments as “reasonable” too. But when a reasonable agent A thinks of another agent B as
being reasonable, despite some deep disagreement between A and B over religious, moral, political,
or philosophical doctrine, what does this amount to?
45

As we saw in the first section of Chapter 1, public reason theory’s stipulative concept of
the “reasonable” has a complicated job description, and its use isn’t reserved only for the theorist.
The concept defines a set of people, and the people in that set deploy their own conceptions of the
“reasonable” in making judgments about which people to trust, and which views to accommodate.
But as we saw in the second section of Chapter 1, some public reason theorists, namely political
liberals, have an additional commitment to building their theory (including any epistemological
elaborations thereof) in a way that respects:
LIBERAL NEUTRALITY: a political conception of a target concept cannot advance any
specific controversial aesthetic, epistemological, ethical, metaethical,
metaphysical, or religious doctrines or values beyond what is already implicit
to the public overlapping consensus.
46

LIBERAL NEUTRALITY is a requirement of a kind of theoretical austerity, and in effect, requires the
liberal theorist to articulate their theory within the confines of the publicity constraint. So not only
must reasonable persons have “reasonable” commitments of their own, and regard at least some
disagreeing parties and/or their incompatible commitments as “reasonable” too, and retain a
commitment to bracketing their non-public commitments whenever they themselves attempt to
justify the exercise authority (or power), political liberalism additionally requires us to make
epistemological sense of these claims within an epistemologically austere theory.

45
Enoch 2017, 137.
46
Cf. Rawls 1993, 10.

36
Enoch, therefore, identifies a cluster of three closely related challenges for public reason
theorists, to overcome: (1) the challenge of epistemologically accommodating the diversity of
plausibly reasonable personal doctrinal commitments (including the variety of ways individuals
epistemologically evaluate their disagreements with others); (2) the challenge of preserving the
intelligibility of individuals’ commitment to adhering to the publicity constraint; and for political
liberals, (3) the additional challenge of addressing (1) and (2) while respecting LIBERAL
NEUTRALITY’s requirement that the theory itself satisfy the publicity constraint.
47
Building on insights from Catriona McKinnon, Joshua Cohen, and Jonathan Quong, I argue
that, like other concepts central to political liberalism, if the concept of reasonable disagreement
is to play its explanatory role, it too must be articulated as a political conception. Despite the fact
that the ideal of public reason has often been articulated within the context of political liberalism,
no author that I know of has been explicit on this point.
48
And while the point will likely not strike
political liberals as controversial, it has nevertheless not been properly appreciated. I will show
that a political conception of reasonable disagreement allows public reason theorists to surmount
Enoch’s challenges.
1. Getting Epistemological about Reasonable Disagreement
At least some public reason theorists have been tempted to think that the reasonableness of
disagreement consists entirely in the persistence of certain moral attributes in disagreeing parties,
especially their readiness to propose and willingly abide by principles and standards that are fair

47
See Enoch 2017.
48
Though some have clearly explored the prospects of giving a suitably political articulation of the reasonableness of
disagreement; most notably, Leland and van Wietmarschen 2012, who are not themselves public reason theorists.

37
terms of cooperation, given assurance that others will likewise do so. But as we learned from the
example of Dek in the first chapter, the “reasonable” must satisfy a minimal standard of epistemic
competence. Yet, Enoch criticizes public reason theory’s deficit of epistemological detail,
observing that public reason theorists “routinely make seemingly epistemological claims,” which
they “do very little to explain.”
49
The idea of public reasons, capable (at least in principle) of
justifying some exercise of authority (or power) to reasonable others with whom one disagrees,
certainly seems to invite epistemological scrutiny. So do the so-called “burdens of judgment,”
which Rawls posits as explanations for the inevitability of persistent reasonable disagreement, and
include the complexities of weighing and assessing sometimes conflicting bits of evidence. It
seems fair to demand some epistemological clarification of these and related notions. Moreover,
it would be surprising and embarrassing for public reason theorists to learn that no epistemological
light could be brought to bear on foundational assumptions, like the reasonableness of
disagreement, without rendering public reason theory inconsistent, self-defeating, or otherwise
unworkable.
50

49
Enoch 2017, 132.
50
Enoch’s negative argument is impressive in its scope. He ambitiously sets out to canvass and eliminate all plausible
candidate hypotheses, which could address public reason theory’s deficit of epistemological detail. In doing so,
however, he considers answers to some very different kinds of questions, so for my purposes it will help to clarify
precisely what kind epistemic clarification I aim to give.
There are at least two different ways that specific, controversial epistemological hypotheses might
supplement public reason theory (following Leece 2003, 534). Firstly, one could give an epistemic argument for public
reason, which attempts to derive a distinctive normative implication of public reason theory from some purely
epistemic claim(s) supported from within a particular epistemological framework. Enoch considers two proposals of
this kind (each offered as a possible interpretation of Nagel 1987), on which public reason is grounded in the epistemic
sensitivity of moral standards (see Enoch 2017, section 3), or, alternatively, the moral sensitivity of epistemic
standards: i.e., moral encroachment (section 4).

38
Since the semi-technical notion of reasonableness employed by public reason theorists has
an epistemological component, there are normative epistemological conditions on the
reasonableness of disagreement. Given the requirement of LIBERAL NEUTRALITY, these normative
epistemological conditions must be articulated within liberal theory in terms of purely public
considerations. Given that the reasonable have their own conceptions of “reasonableness,” which
they deploy in normative evaluations of other members of society, proffered justifications for the
exercise of power or authority, particular views, larger doctrinal perspectives, and communities

As Enoch nicely explains, purely epistemic arguments for political liberalism instantiate the following
schema:
(i) It is morally impermissible to exercise authority or power over or on behalf of someone, based on a principle
that they do not accept, unless the principle has epistemic status P.
(ii) When it comes to non-public doctrines, nothing has epistemic status P.
(iii) Therefore, it is morally impermissible to exercise authority or power over or on behalf of someone, based on
a non-public doctrine that they do not accept (Enoch 2017, 151).
Epistemic status P cannot simply be ‘being reasonably agreed upon,’ or ‘being part of an overlapping consensus
among the reasonable,’ because public reason theorists standardly understand reasonableness as a partly moral, and
hence not a purely epistemic notion. Since an epistemic argument for public reason purports to argue for a purely
epistemic analysis of ‘reasonableness,’ it cannot simply stipulate to its being purely epistemic.
Arguments of this kind have the advantage of relying on thin moral premises (viz., instances of (i)). In
avoiding appeals to controversial moral values, these arguments exhibit one kind of theoretical austerity. However, in
place of appeals to moral concepts, epistemic arguments for public reason have appealed to controversial
epistemological doctrines, contrary to Liberal Neutality. Therefore, like Enoch, I think the prospects for providing this
kind of controversial epistemic grounding for the publicity constraint are dim. Firstly, it’s difficult to see how any
specification of epistemic status P could make both premises (i) and (ii) seem plausible. Second, it’s difficult to see
how any specification of epistemic status P could remain sufficiently uncontroversial to satisfy the principle of LIBERAL
NEUTRALITY (Enoch 2017, 151), So I do not attempt to provide a purely epistemic argument for public reason.
On my view the fact of reasonable disagreement or explanandum, not the epistemological component(s) of
the explanans, is morally and politically significant, owing to the moral and political challenge it poses for justifying
exercises of authority: epistemology explains why morally decent people reasonably disagree, but the appropriate
political response to the fact of reasonable disagreement is specified in relation to moral ideas, like freedom, equality,
and fairness (Leece 2003, 534). Enoch considers and rejects several epistemological hypotheses, which I argue can
profitably be deployed in this second more modest way. It is primarily with the viability of these kinds of
epistemological explanations of reasonable disagreement which I will concern myself. But the success of these
controversial epistemological proposals, depends on their being deployed as personal supplementations to a suitably
freestanding political conception of reasonable disagreement.

39
organized around them, liberal theory’s articulation of the normative epistemological component
of reasonableness, must be (i) compatible with the plurality of controversial epistemological
commitments espoused by reasonable persons, while also (ii) preserving the intelligibility of their
normative evaluations of what’s reasonable and unreasonable.

1.1 Enoch’s Trilemma
Enoch presses public reason theorists for epistemological clarification, both from the point of view
of reasonable persons party to reasonable doctrinal disagreements, and from the impersonal
perspective of public reason theory (or political liberalism) itself: “the public reason theorist must
fill in the details of the reasonable person—in particular, her beliefs about other reasonable people,
who don’t share her [non-public commitments]—in a way that, first, renders her coherent (and
plausibly reasonable), and second, that can justify the requirement [of public reason] not to rely in
the political sphere on considerations that are controversial (in the relevant way) among the
reasonable.”
51

We may distinguish four broad types of hypotheses concerning disagreement: deflationary
hypotheses, principled hypotheses, evidential hypotheses, and psychological hypotheses.
Deflationary hypotheses render disagreements illusory, faultless, or trivial, and hence
epistemologically nugatory. By principled hypotheses I have in mind appeals to a variety of

51
Enoch 2017, 161.

40
respectable epistemological principles or frameworks, which unfortunately do not fit neatly under
any obvious category.
52
Thirdly, disagreements may owe to differences in subjects’ bodies of
evidence. Explanations appealing to principled differences in the processing of evidence may be
classified as hybrid principled-evidential hypotheses.
53
Finally, disagreements may be thought to
owe to cognitive errors in one or more disagreeing party’s reasoning. As we’ll see, psychological
hypotheses of this sort can be antithetical to public reason theory’s purposes. After surveying an
expansive array of hypotheses concerning the reasonableness of disagreement, Enoch concludes
that all plausible candidates are incompatible with public reason theory.

1.1.1 The First Horn of Enoch’s Trilemma
The first problem is that public reason’s endorsement of any specific hypothesis leaves political
liberalism incapable of explaining all plausibly reasonable citizens’ epistemologically-laden
commitments to their non-public doctrines. For whether the hypothesis is deflationary or non-
deflationary, its official endorsement would force public reason theory to take sides in substantive
debates over the nature of commitment and the epistemological evaluation of doctrinal
commitments.
Enoch considers the prospects for understanding doctrinal disagreement in terms of
nihilism, truth-relativism, skepticism, or non-cognitivism.
54
Disagreement is illusory according to
both nihilism and non-cognitivism, since there is no fact of the matter about which to disagree,

52
What I have in mind will become clearer in §4.2.
53
See e.g., Schoenfeld 2014.
54
See Enoch 2017, §§2.1 and 2.2.3.

41
while it’s faultless according to relativism and trivial according to skepticism. So on all these
deflationary explanations, disagreement is of little epistemological significance, and this might
seem to lighten public reason theory’s explanatory load. For illustrative purposes, suppose some
interpretation of post-Vatican II Catholicism (hereafter simply “Catholicism”) is a reasonable set
of doctrines.
Nihilism about moral, political, religious, and/or philosophical doctrines amounts to
believing that no such doctrine is true, or is of genuine value. So nihilism is strictly inconsistent
with at least some plausibly reasonable doctrines whose truth is part of their content. Catholicism,
like most denominations of Christianity, maintains that Jesus Christ is the Truth. Nihilistic
Catholicism is therefore incoherent. Catholicism is similarly strictly inconsistent with truth-
relativism. Skepticism about non-public doctrines is strictly inconsistent with any doctrine which
includes a commitment to knowledge of the truth. More generally, consider a Moorean-
paradoxical commitment of the form “p, and no one can know whether p.” If a commitment to any
reasonable doctrine entails a commitment to its truth, then a skeptic’s commitment to that doctrine
is Moorean-incoherent; Catholicism plausibly involves such a commitment. Although Enoch
discusses non-cognitivism in a different section of his paper, its implications are of a piece with
those just discussed. For while commitment to a reasonable non-public doctrine is plausibly more
than a merely cognitive affair, usually involving at least some non-cognitive states (e.g. a
resolution to live one’s life a certain way), any reasonable non-public doctrine that entails a
commitment to its truth cannot understand commitment as being non-cognitive all the way down.
Moreover, it seems the reasonable Catholic must, “think that atheists, Orthodox Jews, Buddhists,
Shiite Muslims, and many others adhere to false non-public doctrines. All of these, she must
conclude, have false beliefs about such central things as what is of value, what is important in life,

42
what makes life worth living, and so on.”
55
So it isn’t just a matter of accommodating the diversity
of plausibly reasonable understandings of one’s own doctrinal commitments, but also the variety
of epistemologically-laden ways plausibly reasonable individuals evaluate their disagreements
with others.
A thoroughgoing nihilist, truth-relativist, skeptic, or non-cognitivist can coherently endorse
political liberalism together with her non-public doctrine, but if public reason theory were to
officially generalize any such view, it would be impossible for the non-nihilist, non-skeptic, non-
relativist, or cognitivist adherent of any doctrine to coherently endorse political liberalism, and
vice versa.
56
Any such officially generalized view would disqualify too many plausibly reasonable
doctrines and their adherents, greatly narrowing public reason’s constituency so as to undermine
the individual’s rationale for fulfilling the requirements of public reason,
57
while also making
political liberalism unacceptably controversial.
58

1.1.2 The Second Horn of Enoch’s Trilemma
The second problem is that psychological hypotheses, which attribute doctrinal disagreements to
cognitive errors in reasoning, threaten to undermine the reasonable citizen’s motivation for

55
Enoch 2017, 137 (my emphases).
56
That is, a thoroughgoing non-nihilist, non-skeptic, non-relativist, or cognitivist can also coherently endorse political
liberalism together with her non-public doctrine, but if public reason theory were to officially generalize any such
view, it would be similarly impossible for the nihilist, truth-relativist, skeptic, or non-cognitivist adherent of any
doctrine to coherently endorse political liberalism.
57
And thereby fall to the second horn of Enoch’s Trilemma (though for different reasons than those I’ll discuss in
relation to psychological hypotheses below).
58
And thereby fall to the third horn of Enoch’s Trilemma. Evidently, the three horns of the trilemma are closely
related, but each nonetheless represents a distinct theoretical desideratum for a satisfactory epistemological account
of reasonable doctrinal disagreement.

43
restraining herself from appealing to her non-public doctrines, in accordance with public reason’s
distinctive publicity constraint. Hypotheses of this kind are common in political discourse, ranging
from respectful arguments to the effect that a particular commitment is “illogical” or “irrational,”
to wholesale dismissals of religions or political philosophies as “backward,” “biased,” “crazy,”
“ignorant,” “stupid,” and worse. Of course, in these cases the intent is to at least undermine the
reasonableness, or to demonstrate (or assert) the unreasonableness of the target doctrine, so it’s
unsurprising such attitudes toward the cause(s) of doctrinal disagreement undermine any
motivation the individual might have for observing the publicity constraint. However, Enoch
argues that even more sophisticated psychological hypotheses about the causes of doctrinal
disagreement, which explicitly aim at preserving the evaluation of a disagreement as reasonable,
undermine the individual’s motivation for restraining herself from appealing to her non-public
doctrines. For we may expect in the predictive sense that people will make errors, but we expect
in the normative sense that people not make errors. So if doctrinal disagreement just comes down
to others making mistakes (however understandably so), why shouldn’t one expect (in the
normative sense) those in error to respond reasonably to correction?
59
Surely, this isn’t
prohibitively controversial. But if we’re correcting errors, it’s not clear why we should hold back
from telling others the whole truth, by the lights of our non-public doctrines.
Attributing doctrinal disagreements to cognitive error is complicated by the fact that, in
some cases, cognitive errors may have some causal influence over an epistemic subject’s chain of

59
Enoch 2017, 147; see Leland and van Wietmarschen 2012, 726.

44
reasoning without thereby rendering their conclusion (and resulting disagreement) unreasonable
simpliciter. So let’s clarify that if A thinks her doctrinal disagreement with B owes sufficiently to
cognitive error in B’s reasoning, then A takes B’s disagreement to generate no reason
60
for A to
restrain herself from appealing to her non-public doctrines, in accordance with the publicity
constraint.
1.1.3 The Third Horn of Enoch’s Trilemma
So far we’ve seen that Enoch argues that public reason theory’s official endorsement of any
deflationary (or non-deflationary) hypothesis concerning disagreement:
1. leaves public reason theory too doctrinally demanding to accommodate all plausibly
reasonable citizens’ epistemologically-laden commitments to their non-public doctrines
(including commitments about the nature of commitment and epistemological evaluations
of others’ commitments).

It’s also alleged that official endorsement of any psychological hypothesis concerning
disagreement:
2. undermines the reasonable citizen’s motivation for restraining herself from appealing to
her non-public doctrines, in accordance with public reason’s distinctive publicity
constraint.

The horns of this dilemma reflect two difficulties of filling in the details of reasonable doctrinal
disagreement in a way that both renders persons’ commitments coherent (and plausibly
reasonable), and preserves the individual’s motivation for adhering to the publicity constraint.
However, a closely related constraint, constituting the third horn of a trilemma, arises for political

60
I need to think more about whether this is the right way to think about it. Perhaps thinking in terms of whether
particular doctrinal disagreements occasion the publicity constraint is misguided.

45
liberalism’s official endorsement of any specific epistemological hypothesis concerning
disagreement whatsoever, be it principled, evidential, or deflationary.
Enoch points out the implausibility of ascribing sophisticated epistemological
understandings of reasonable doctrinal disagreement to ordinary citizens. After all, we can’t expect
all reasonable citizens be epistemologists, and we can’t expect them to all agree on matters of
epistemological controversy.
61
So political liberalism’s endorsement of any specific
epistemologically controversial hypothesis:
3. renders the content of political liberalism itself too controversial to satisfy LIBERAL
NEUTRALITY, and hence self-defeating.
62

This third horn is really just the impersonal theoretical flip-side of the first horn, but it
represents a distinct desideratum for the political liberal. I call the resulting trilemma Enoch’s
trilemma.

1.2 Outline of Remaining Sections
In this paper I argue that a political conception of reasonable doctrinal disagreement provides the
desired epistemological clarification from the impersonal perspective of public reason theory,
while leaving room for many flowers to bloom from the point of view of reasonable persons, who
are free to supplement the political conception with any combination of a large range of compatible

61
Enoch 2017, 146.
62
This third horn assumes an epistemological explanation of reasonable doctrinal disagreement, which helps to justify
the requirements of public reason, must itself satisfy these requirements, but someone might deny that rationale for
these requirements must be reflexive in this way; e.g., Wall 2002, 388–389.

46
private epistemological commitments. Enoch’s critical discussion does not engage with any
proposal of this kind.
63
In section 2, I begin by considering the bleak prospects for doing without
a political conception, using Rawls’s most influential treatment of the epistemological component
of reasonable doctrinal disagreement for illustration. In section 3, I then turn to Rawls’s under-
discussed criteria of reasonable doctrinal disagreement to develop a suitably political conception
of reasonable doctrinal disagreement, which ensures the non-triviality of the publicity constraint,
prioritizes assurances of cooperation in light of the mutual assurance problem, and leaves room
for a diversity of private epistemological commitments to flourish among reasonable citizens.
Finally, in section 4, I explain in more detail how reasonable citizens may supplement the political
conception of reasonable doctrinal disagreement with various private epistemological
commitments while avoiding the horns of Enoch’s trilemma.
2. Trying to do Without a Political Conception
Suppose that public reason theorists had to rely on a non-political conception of reasonable
doctrinal disagreement. How would that work? Rawls’s seminal discussion of the reasonableness
of persons illustrates the difficulties posed by Enoch’s trilemma for these kinds of views.

63
Enoch does consider what I understand to be a proposal of this kind, in the work of Leland & van Wietmarschen.
However, he interprets the significance of their proposal differently.

47
2.1 Rawlsian Reasonableness of Persons
Wenar influentially characterizes Rawls’s strategy in Political Liberalism as one of explicitly
defining ‘reasonable person’ and ‘reasonable comprehensive doctrine’, while leaving the meaning
of other terms, including ‘reasonable [doctrinal] disagreement,’ to be derived with reference to
these primitive definitions.
64
Rawls’s definition of ‘reasonable comprehensive doctrine’ fails to
rule out comprehensive (or non-public) doctrines that are clearly unreasonable in his sense, such
as islamic fundamentalism, white supremacism, or rational egoism.
65
So Wenar suggests that the
definition of ‘reasonable person’ alone be taken as primitive. Smoothing over the attending textual
difficulties in Rawls’ account as Wenar suggests yields the following five Rawlsian conditions on
the reasonableness of persons; reasonable persons:
1. (a) possess two moral powers–the capacities for a sense of justice and for a conception of
the good; (b) possess the intellectual powers of judgment, thought, and inference; (c) have
a determinate conception of the good interpreted in the light of some [non-public
doctrine(s)]; and (d) are able to be normal, fully cooperating members of society over a
complete life; and,
2. are ready to propose and willingly abide by principles and standards that are fair terms of
cooperation, given assurance that others will likewise do so.
3. recognize the burdens of judgment;
4. have a reasonable
66
moral psychology; and

64
Wenar 1995, 37. Contrary to Wenar, I think Rawls’s descriptions of the fact of reasonable pluralism provide distinct
criteria for reasonable doctrinal disagreement, which can help us to develop a political conception of reasonable
doctrinal disagreement, so I consider them in the third part of this paper. Before I do so, however, it’s helpful to
understand the standard objections to Rawls’s account of the reasonableness of persons.
65
Wenar 1995, 35–36; see Rawls 1993, 59.
66
Obviously, the meaning of ‘reasonable’ here cannot depend on the meaning of ‘reasonable person’ on pain of
circularity.

48
5. recognize the five essential elements of a conception of objectivity.
67, 68

Since my aim is to clarify the epistemology of reasonable doctrinal disagreement, it will most
illuminating to focus on any objections levelled against the epistemological elements of these
conditions. And while the epistemic content of Condition 1b has proven uncontroversial, the
burdens of judgment have drawn considerable criticism.

2.2 The Burdens of Judgment
Rawls’ suggestion with the burdens of judgment is that people accept the permanence of doctrinal
disagreements, because they believe that the exercise of our rational faculties is subject to burdens
which make disagreement inevitable, even in ideal conditions. Moreover, Rawls’s burdens view
is standardly interpreted as implying that citizens, “all accept the same explanation of why they
cannot expect to convince one another of their views through the exercise of reason, even in the
best of conditions. In that case, people who recognise these burdens of judgment accept that
reasonable pluralism is ineradicable.”
69
Rawls describes the following burdens:
A. Relevant evidence is often conflicting and complex, and therefore hard to evaluate;
B. We may disagree about the weight different considerations are to be given;

67
Wenar 1995, 37; for condition 1, see Rawls 1993, 81; 15–35; for condition 2, see 48–54; for condition 3, see 54–
58; for condition 4, see 81–86; and for condition 5, see 110–112.
68
Wenar 1995, 57. Wenar contends that Rawls added conditions 3–5 to his earlier conception (captured by conditions
1 and 2) in an effort to answer questions about why an overlapping consensus is necessary, how it could come about,
and how it could persist. Unfortunately, the resulting elaborations make the Rawlsian conception of the reasonable
too controversial, and hence subject to reasonable rejection. Wenar summarizes his case against the controversial
implications of conditions 3–5 this way: “The burdens of judgment displace firm religious faith. The conception of
the person underlying the reasonable moral psychology conflicts with the conceptions of other philosophical theories.
Political constructivism excludes prevalent accounts of the sources of normative authority” (57).
69
McKinnon 2002, 50 (my emphasis). See also Wenar 1995, 48.

49
C. Our concepts are vague, so we must rely on judgment and interpretation;
D. The way each person assesses evidence and weighs values is shaped by her particular life
experiences (her ethnicity, class, place in the division of labor, etc.);
E. It’s difficult to make an overall assessment of an issue when there are normative
considerations of different force on both sides; and
F. In any society some selection of values must be made from among all that might be realized
(less a source of than occasion for disagreement).
70

Rawls says the main purpose of the burdens is to underwrite toleration and thus public reason.
71
The
idea is that people will only practice toleration and restrain themselves to promulgating public
reasons in support of their political agendas if they attribute the permanence of pluralism to the
fact that everyone stands under inescapable burdens of judgment.
72

2.3 Problems with the Burdens View
The burdens view faces three significant objections. The first is that acceptance of the burdens is
not necessary for endorsing toleration. Sufficient reasons for practicing toleration can be secured
from within a non-public doctrine, independently of a commitment to the burdens, such as Locke’s
liberal perfectionist argument in his Letter Concerning Toleration.
73
Wenar, meanwhile, points to
Catholicism’s unequivocal teaching on liberty of conscience and freedom of thought:
This Vatican Synod declares that the human person has a right to religious freedom.
This freedom means that all men are to be immune from coercion on the part of
individuals or of social groups and of any human power, in such wise that in matters
religious no one is forced to act in a manner contrary to his own beliefs. Nor is
anyone to be restrained from acting in accordance with his own beliefs, whether

70
Wenar 1995, 41; see Rawls 1993, 55–57.
71
Wenar 1995, 42; see Rawls 1993, 58–59, 61.
72
McKinnon 2002, 51.
73
See Klosko 2000, 22.

50
privately or publicly, whether alone or in association with others, within due limits.
The Synod further declares that the right to religious freedom has its foundation in
the very dignity of the human person, as this dignity is known through the revealed
Word of God and by reason itself. This right of the human person to religious
freedom is to be recognized in the constitutional law whereby society is governed.
Thus it is to become a civil right.
74

Furthermore, Rawls does not show that non-acceptance of the burdens makes one impermissibly
intolerant.
75
The second objection to the burdens view, is that acceptance of the burdens is not sufficient
for endorsing toleration. As George Klosko observes, Rawls does not show that acceptance of the
burdens causes one to moderate one’s attitude towards others or their commitments to their non-
public doctrines: “It is an empirical question whether people are able to behave appropriately
without particular kinds of [commitments, including acceptance of the burdens of judgment].
Without strong evidence that this is not possible, we should not exclude any cooperative
individuals from the [constituency of the reasonable].”
76

But the third and most important objection, for our purposes, is that the burdens of
judgment are strictly incompatible with some plausibly reasonable worldviews. As such, they fall
to the first and third horns of Enoch’s trilemma. They are, for example, “repudiated outright” by
Catholic doctrine:
Through divine revelation, God chose to show forth and communicate Himself and
the eternal decisions of His will regarding the salvation of men. That is to say, He
chose “to share those divine treasures which totally transcend the understanding of

74
Declaration on Religious Freedom, article 2; as quoted in Wenar 1995, 42–43; see n. 17.
75
Klosko 2000, 21.
76
Klosko 2000, 21.

51
the human mind.” ... This sacred Synod affirms, “God, the beginning and end of all
things, can be known with certainty from created reality by the light of human
reason” (see Rom. 1:20); but the Synod teaches that it is through His revelation
“that those religious truths which are by their nature accessible to human reason
can be known by all men with ease, with solid certitude, and with no trace of error,
even in the present state of the human race.”
77

In fact, most of the world’s major historical religions characteristically (if not universally) reject
the burdens. As Wenar forcefully summarizes, “to ask Catholics and other believers to accept the
burdens of judgment is to ask them to abandon—unnecessarily—fundamental aspects of their faith
and their attitude toward it.”
78

Rawls introduces the burdens to explain the fact of reasonable pluralism, and I agree with
Rawls that some explanation is needed, but like Wenar and McKinnon, I see no need for one
officially accepted explanation, “instead of a variety of different (and possibly conflicting)
explanations in the minds of those citizens who care to consider the issue.”
79
Indeed, the first horn
of Enoch’s trilemma highlights the problem for any public reason theory that endorses a single
epistemological story: doing so makes public reason theory too doctrinally demanding to
accommodate all plausibly reasonable citizens’ commitments to their epistemologically-laden
non-public doctrines. Thus, requiring reasonable persons to accept the burdens view is needlessly
exclusionary.
80
The standard interpretation of Rawls’s specification of the epistemological

77
Dogmatic Constitution on Divine Revelation, article 6; as quoted in Wenar 1995, 44; see n. 20.
78
Wenar 1995, 45. Rawls does not think the burdens of judgment imply hesitancy, uncertainty, or skepticism about
comprehensive doctrines (1993, 62-63, 150-54; see also McKinnon 2002, 51). However, Wenar and Barry think they
do imply skepticism, and this too would make them incompatible with several plausibly reasonable worldviews.
79
Wenar 1995, 48.
80
McKinnon 2002, 52.

52
component of reasonableness is untenable. Therefore, we need some other way to clarify the
epistemology of reasonable doctrinal disagreement.
81
However political liberals end up defining ‘reasonable doctrinal disagreement’, its
definition must function as a political conception. While this claim should not strike political
liberals as controversial, it has failed to be properly appreciated. For the fact of reasonable
pluralism is the defining problem to which political liberalism addresses itself. As such, an
affirmation of the fact of reasonable pluralism is part of political liberalism’s content. So if public
reason theory endorses a specific controversial account of reasonable doctrinal disagreement in
the course of affirming the fact of reasonable pluralism, ipso facto this constitutes a failure of the
project to articulate a truly political conception of liberalism: the theory becomes self-defeating.
This is the threat highlighted by the third horn of Enoch’s trilemma. Unfortunately, standard
characterizations of reasonableness, including the reasonableness of disagreement, have not been
insulated against this threat.

81
Wenar’s preferred solution is to reject Rawls’s conditions 3–5, retaining conditions 1 and 2 as a minimal conception
of the reasonable. He believes the burdens add nothing essential to the presentation of justice as fairness, and that
specific substantive liberal commitments to things like toleration and public reason are secured by the first and second
attributes of the reasonable (i.e., conditions 1 and 2), obviating the need for elaboration. (Other philosophers have
endorsed similar conceptions of the reasonableness of persons, including Klosko 2000, 21 and Kelly and McPherson
2001; Nussbaum 2011 tentatively endorses something similar: see 30–31 n. 50.)
If this was so, we could simply define ‘reasonable doctrinal disagreement’ with reference to this minimal
conception of the reasonable (i.e., as disagreement between minimally reasonable persons). However, conditions 1
and 2 by themselves are too thin to secure specific substantive normative commitments, like that of religious tolerance.
To illustrate, consider that what it is to cooperate on fair terms will mean different things to the libertarian, the socialist,
the communitarian, the communist, or the liberal, regardless of whether they possess all the attributes of condition 1,
or what mutual assurances they have of others’ cooperation. (Unless, of course, “principles and standards that are fair
terms of cooperation” is read de re.) So Wenar’s minimal conception of the reasonable isn’t a viable alternative.

53
3. A Political Conception of Reasonable Doctrinal Disagreement
3.1 Rawlsian Criteria of Reasonable Doctrinal Disagreement
82

The challenge remains to demonstrate that an epistemologically respectable, public notion of
reasonableness can accomplish all that is required of it by public reason theory. A natural place to
begin is with Rawls’ canonical descriptions of the fact of reasonable pluralism, which provides the
following criteria, according to which reasonable doctrinal disagreement:
1. is not a function of ( 𝛼𝛼) irrationality or logical error, ( 𝛽𝛽 ) prejudice or bias, ( 𝛾𝛾) self- or group-
interest,
83
or ( 𝛿𝛿 ) willfulness,
84
and
2. persists even where parties freely
85
exercise their faculties of reasoning under favorable
conditions.
86

These criteria are of obvious epistemological significance, but they’ve received surprisingly little
attention. For our purpo𝝐𝝐 ses, we may think of these as criteria for believing one’s fellow citizen(s)
to be reasonable, despite some doctrinal disagreement.
Just as we observed in our earlier discussion of cognitive errors in relation to the second
horn of the trilemma, the sources of cognitive error listed in 1𝛼𝛼 –𝛿𝛿 very plausibly admit of degrees,
and any factor among 1𝛼𝛼–𝛿𝛿 might wield some small causal influence over an epistemic subject’s
chain of reasoning without thereby rendering a disagreement unreasonable simpliciter. So the first

82
These are to be distinguished from the five Rawlsian conditions on the reasonableness of persons surveyed earlier
in §4.1.
83
Including class interest; Rawls 1993, 37.
84
Along with 1 𝛼𝛼 – 𝛿𝛿 , Rawls lists “blindness” as another vice that’s incompatible with reasonable doctrinal
disagreement. What Rawls seems to have in mind is “peoples’ understandable tendency to view the political world
from a limited standpoint” (Rawls 1993, 37).
85
See Rawls 1993, xxvi.
86
For condition 1 𝛼𝛼 , see Rawls 1993, 54; for 1 𝛽𝛽 – 𝛿𝛿 , see 58; for condition 2, see 398.

54
criterion is most plausibly read as saying that the more a disagreement owes to irrationality,
prejudice, bias, self-interest, group-interest, or willfulness, the less reasonable it is.
The second criterion, meanwhile, suggests that a disagreement is reasonable simpliciter so
long as it would persist were the disagreeing party’s exercise of his or her reasoning faculties
idealized to the point that it manifests the virtue of “reasonableness,” and is exercised under
conditions of freedom, where the reasoner would be under no duress or threat of coercion.
87

Interpreted in the light of the moral conditions on “reasonableness,” the relevant “faculties of
reasoning” ought not be understood in a narrowly epistemic sense, but as including powers of
moral reasoning as well. Still, as my aim here is to clarify the epistemology of reasonable doctrinal
disagreement, I’ll focus primarily on the epistemic idealization at work in our normative
expectations of the reasonable.

3.2 The Non-Vacuity of the Publicity Constraint
Firstly, as we learn from the example of Doc (in section 1.3) the epistemic idealization at work in
our normative expectations of the reasonable must not render the publicity constraint vacuous. To
put things a different way, the publicity constraint shouldn’t be trivial in the sense that only those
who believe what one takes to be the truth may count as reasonable. The publicity constraint
requires reasonable persons to refrain from appealing to at least some of the non-public doctrines
that are personally important to them, as unsuitable grounds for publicly justifying political
institutions, policies or agendas, because many of their fellow citizens cannot be expected to

87
Rawls 1993, xxvi.

55
reasonably accept appeals on such grounds. As R.J. Leland and Han van Wietmarschen observe,
this would not be the case if the reasonable could have any normative expectations they like. So if
the publicity constraint is not to be vacuous, reasonable citizens cannot be too demanding of their
fellow citizens.
88
If A thinks everyone would converge on her own moral, political, religious, and/or
philosophical views, and all doctrinal disagreements would disappear, if only everyone was
epistemically competent to some high degree D, then A must also accept that some degree D-minus
of epistemic competence below D, where doctrinal disagreements still obtain, is epistemically
competent enough for the purpose of being “reasonable.”
89
Simply put, the bar can’t be so high
that it entails truth, by A’s lights as to what’s true. Furthermore, the more doctrinal disagreement
an agent accepts as “reasonable,” without imperiling her own moral decency, or undermining her
assurance of fair cooperation from those with whom she “reasonably” disagrees, the more
“reasonable” she is.
As we saw earlier, Doc the atheist UNIQUENESS theorist goes wrong by insisting on an
exceedingly demanding standard of epistemic competence for evaluating the “reasonableness” of
others, on which only the epistemically perfect, whom he supposes will all converge with absolute
certainty on atheism, count as “reasonable” in virtue of this feature alone. But what of a more
permissive atheist, who counts those less certain about atheism as “reasonable,” just so long as
they remain atheists and share roughly the same moral, political, and other philosophical
commitments? Such stringent requirements of like-mindedness among the “reasonable” still rule

88
Leland and van Wietmarschen 2012, 727.
89
Leland and van Wietmarschen call this formal constraint “the balancing condition” (2012, 729).

56
out the possibility of a truly “reasonable” pluralism. For the fact of “reasonable” pluralism entails
something stronger than the mere non-vacuity of the publicity constraint; it requires that at least
many doctrinal disagreements, across moral, religious, political, and other philosophical lines,
count as “reasonable.”
One might object that this is too doctrinally demanding, because it requires reasonable
citizens to accept claims about the fundamental limits of human reason, or the view that basic
moral, religious, political, or philosophical facts are unknowable by the human intellect. After all,
any such requirement would violate LIBERAL NEUTRALITY. However, the fact of reasonable
pluralism requires reasonable citizens to believe no such thing: it entails only that, at all levels of
epistemic competence necessary for reasonableness, at least many doctrinal disagreements
obtain.
90

3.3 The Lexical Priority of Moral Decency Over Epistemic Competence
According to Rawls, a political conception must not only respect the principle of LIBERAL
NEUTRALITY, it must also be worked out by appeal to ideas implicit in the public political culture
of a constitutional democracy.
91
Thankfully, it is a commonplace of liberal democratic political
culture to regard a plurality of incompatible moral, political, religious, and philosophical
perspectives as being nonetheless compatible with moral decency. It’s widely accepted that there
are (at least some) morally decent consequentialists, Kantians, contractualists, Rossian pluralists,
virtue ethicists, liberals, conservatives, democratic socialists, libertarians, communitarians,

90
Leland and van Wietmarschen 2012, 744.
91
Rawls 1999, 584.

57
Christians, Jews, Muslims, Hindus, Jains, Sikhs, Buddhists, indigenous spiritists, animists, New-
Ageists, pagans, syncretists, atheists, agnostics, secularists, humanists, as well as morally decent
persons who don’t have considered views or who hold some less-than-perfectly-coherent view
about some or all of the above. Moreover, in a well-ordered liberal society, the acceptance of
reasonable pluralism justifies a presumption
92
that others, regardless of creed, are sufficiently
morally decent and intellectually competent to be reliable cooperators on fair terms. It seems to
me implicit in liberal democratic political culture that citizens care primarily about whether others’
can be relied upon in this way.
Given desires for this kind of assurance, a first constraint on a political conception of
reasonable doctrinal disagreement is that it makes the normative expectation of moral decency
lexically prior to the normative expectation of their epistemic competence. The epistemic
credentials of disagreeing parties to doctrinal disagreements should be of concern only insofar as
their epistemic (in)competence supports or undermines their moral decency.
Contrary to Leland and van Wietmarschen, who favor a demanding standard of epistemic
competence, I contend that, depending on what’s at stake in a particular doctrinal disagreement,
modest normative expectations of epistemic competence are often sufficient for the purposes of

92
Little attention has been paid to the various attitudes public reason theory requires reasonable persons to adopt
toward others whom they deem reasonable, despite doctrinal disagreement. Enoch asks: when a reasonable person A
thinks of another person B as being reasonable, despite B’s rejection of A’s religious, moral, or political outlook, what
does this amount to? What does A believe of B? (2017, 136). As I suggest here, it isn’t simply a matter of what A
believes of B. In a well-ordered society, there is an important practical role for a (defeasible) presumption in favor of
others’ moral decency or trustworthiness, and secondarily, in favor of their sufficient epistemic competence.

58
the reasonable. After all, Leland and van Wietmarschen concede that there’s no guarantee that
even highly epistemically competent people accept the central liberal- democratic political values,
and people who deny these values are unreasonable even if they rank among the most epistemically
competent reasoners on matters of political morality.
93
Moreover, modest epistemic expectations
help to ensure that a robust diversity of doctrinal disagreements are regarded as reasonable.

3.4 Epistemic Competence: Letting Many Flowers Bloom
Recall that the burdens view suffers from a kind of over-demandingness, because the burdens
themselves are needlessly exclusionary of some plausibly reasonable perspectives. It might be
feared that Rawls’s use of the term ‘irrationality’ threatens to make criterion 1 𝛼𝛼 similarly
controversial. After all, it’s disputed whether rationality entails logical or a priori omniscience,
94

and, of course, there are the so-called “rationality wars” over the naturalistic credentials of standard
normative notions of (epistemic) rationality.
95
However, the concept of reasonableness doesn’t
entail any particular account of rationality, and is therefore compatible with any account of
rationality so long as it is, in some sense, normative. Leland and van Wietmarschen nicely illustrate
this point. As they explain, our normative epistemic expectations of reasonable persons have the
following structure:
if you think someone should accept p, then you think she would accept p under
some suitable psychological idealization. The aspects of the person’s psychology

93
Leland and van Wietmarschen 2012, 730.
94
E.g., Smithies 2015.
95
For a brief but helpful overview, see Rysiew 2016, §6.3.

59
that are relevant to this idealization are those that determine how well she is situated
to make a judgment about p. These aspects will differ depending on what type of
proposition p is (e.g., whether it is a meteorological proposition or an aesthetic one).
But, speaking generally, we can say that a person is better situated to judge whether
p insofar as she has access to considerations relevant to p and possesses intellectual
virtue in settling questions in the domain of p.
96

Crucially, there’s no restriction on how reasonable persons understand intellectual competence (or
epistemic rationality). So in addition to the variations in how competence may be assessed across
domains or subject matters, there may also be variation across reasonable perspectives as to how
competence is understood and assessed with respect to a specific subject matter. In other words,
public reason theorists can let many epistemological flowers bloom with respect to how reasonable
citizens understand and evaluate their disagreements.

3.5 A Political Conception of Reasonable Doctrinal Disagreement
So on my view, a reasonable doctrinal disagreement:
1. is between morally decent parties;
2. is between parties sufficiently epistemically competent to be relied upon to cooperate on
fair terms, and is therefore not a function of excessive ( 𝛼𝛼) irrationality or logical error, ( 𝛽𝛽 )
prejudice or bias, ( 𝛾𝛾 ) self- or group-interest, or ( 𝛿𝛿 ) willfulness; and
3. persists (or is expected to persist) even where parties freely exercise their faculties of
reasoning under favorable conditions.

No restriction is placed on how reasonable persons understand epistemic competence or
rationality. Since reasonableness admits of degrees, some suitably small degree of causal influence
from factors 2𝛼𝛼–𝛿𝛿 on a disagreeing party’s chain of reasoning is compatible with their

96
Leland and van Wietmarschen 2012, 726.

60
conclusion(s) and resulting disagreement(s) being nonetheless sufficiently epistemically
competent to count as reasonable.
Recall that a political conception is compatible with a wide variety of non-public doctrines,
because it:
a. is freestanding, presented independently of any non-public commitments;
b. articulates only political values, but does not claim that these values are more important
than other values; and
c. is independent of long-standing controversies in philosophy; in particular it has no
distinctive metaphysical doctrine of the person.

My account of reasonable doctrinal disagreement satisfies each of these conditions. It can therefore
serve as, what Rawls calls, a “module” in the thinking of reasonable citizens, who are free to
supplement it with any of a wide variety of private epistemological commitments. In addition, this
political conception of reasonable doctrinal disagreement can be used to articulate the fact of
reasonable pluralism, public reason theory, and political liberalism in a way that respects the
principle of LIBERAL NEUTRALITY.
4. The Private Epistemological Commitments of the Reasonable
As Wenar and McKinnon first observed, nothing requires reasonable citizens to all embrace the
same general epistemological account of disagreement, nor must they understand the epistemic
status of all reasonable doctrinal disagreements in the same way. A variety of different (and
possibly conflicting) explanations may obtain in the minds of those citizens who care to consider
the issue, without posing a threat to public reason or political liberalism. Explaining the remaining
details of what’s going on in particular cases of disagreement can be delegated to each individual’s
private epistemological commitments, leaving the attendant epistemological controversies outside
the scope of public reason theory, and hence not subject to political liberalism’s principle of

61
LIBERAL NEUTRALITY. Public reason theory doesn’t need to generalize any particular, controversial,
epistemological hypothesis concerning disagreement. So long as an epistemological explanation
is compatible with the political conception of reasonable doctrinal disagreement, reasonable
persons may help themselves to its explanatory virtues as a supplement to the political conception.
So in effect, the epistemological neutrality of my political conception of reasonable doctrinal
disagreement neutralizes the threat of the first and third horns of Enoch’s trilemma, while the
second condition screens out any psychological hypotheses about the causes of doctrinal
disagreement that would fall to the second horn.
The rest of this chapter proceeds as follows: in sections 6.1 and 6.2, I look more carefully
at how deflationary (and non-deflationary) hypotheses, principled hypotheses, and evidential
hypotheses concerning doctrinal disagreement avoid the first and third horns of Enoch’s trilemma,
when operating only as private supplements to the political conception of reasonable doctrinal
disagreement. I then anticipate and respond to a version of Joseph Raz’s objection to epistemic
abstinence, which might be levelled against my view. I conclude this section by taking a closer
look at the sophisticated psychological hypotheses which fall to the second horn of the trilemma
in section 6.3.

4.1 Deflationary Hypotheses about Disagreement
As I noted in discussing the first horn of Enoch’s trilemma (in §1.3.1), the thoroughgoing nihilist,
truth-relativist, skeptic, or non-cognitivist can coherently endorse political liberalism together with
her non-public doctrine (so can the thoroughgoing non-nihilist, non-skeptic, non-relativist, or
cognitivist). So long as none of these epistemologically-laden non-public doctrines concerning the
nature of commitment and doctrinal disagreement are officially endorsed by public reason theory,

62
there’s neither a problem of over-demandingness (pace the first horn of the trilemma), nor a
problem of self-defeat (pace the third horn of the trilemma). So the political conception of
reasonable doctrinal disagreement can be supplemented with any of the above from the point of
view of reasonable citizens. And, of course, a reasonable citizen may, for example, be a non-
cognitivist about aesthetic judgments and disagreements, while remaining a cognitivist about
moral judgments and disputes.

4.2 Principled and Evidential Epistemological Hypotheses about Disagreement
In addition to deflationary hypotheses concerning doctrinal disagreement, a wide variety of
respectable epistemological hypotheses appealing to coherentist, conservatist, subjective
Bayesian, and other versions of rational permissivism,
97
epistemic “near misses”
98
(understood in

97
See e.g., Podgorski 2016 for a nice overview of several representative versions of rational permissivism; cf. Enoch
2017, §2.2.2. As Podgorski explains, UNIQUENESS is standardly understood as the denial of the permissivist thesis, but
Enoch formulates the permissivist thesis this way:
PERMISSIVENESS: Roughly, in a given state of evidence with regard to p, there is more than one credence
that it is maximally rational to have regarding p, or perhaps that regardless of maximal rationality,
there is more than one credence that it is acceptable to have (2017, 140).
PERMISSIVENESS is compatible with the UNIQUENESS thesis, because even if there is only one doxastic state that it is
uniquely rational to be in, given a total body of evidence, there may nonetheless be more than one doxastic state that
it is acceptable to have. By obscuring this standard point of epistemological contention, PERMISSIVENESS threatens to
also obscure a way in which the political conception of reasonable doctrinal disagreement is epistemologically
ecumenical: it is compatible with both permissivism and UNIQUENESS.
Aside from concerns about the first and third horns of the trilemma, which are neutralized by the political
conception of reasonable doctrinal disagreement, Enoch observes that what’s needed to explain some plausibly
reasonable doctrinal disagreements is permissivism about the traditional doxastic attitude of belief (not just degrees
of belief or credence), in order to make it permissible, for example, to believe both Catholicism and its denial, and
Enoch thinks this is implausible (see 2017, 141). But while some versions of permissivism have trouble making good
on this result, others (like Podgorski’s favored Dynamic Permissivism) do not.
98
Enoch 2017, §2.4, considers whether a reasonable Catholic can think non-Catholics have made a mistake, but not a
bad one (i.e., a “near miss”). Since rationality plausibly comes in degrees, the reasonable Catholic may think that the

63
terms of suboptimal but nonetheless acceptable degrees of rationality
99
), as well as the privacy of
evidence,
100
are all available to the reasonable citizen, as private supplementations to the political
conception of reasonable doctrinal disagreement. Even if highly controversial or false, insofar as
the above hypotheses allow reasonable citizens to regard other parties to reasonable doctrinal
disagreements as sufficiently epistemically competent to be relied upon to cooperate on fair terms
(rather than chalking the disagreement up to an excess of ( 𝛼𝛼 ) irrationality or logical error, ( 𝛽𝛽 )
prejudice or bias, ( 𝛾𝛾) self- or group-interest, or ( 𝛿𝛿 ) willfulness), they are acceptable from the
perspective of public reason theory and political liberalism. However, this invites a familiar
criticism, which may be understood as a kind of revenge of the third horn of Enoch’s trilemma.
Joseph Raz objects to what he calls the epistemic abstinence of any version of political
liberalism that refrains from claiming its doctrine of justice is true.
101
As Raz explains, if a doctrine
of justice is true, it’s true in virtue of deeper normative foundations than those afforded by the
relatively shallow foundation of an overlapping consensus or convergence on suitably public
political principles. But such deeper normative foundations are the stuff of non-public doctrine.

beliefs of non-Catholics are almost right or may be thought to have responded non-optimally, but well enough. Enoch
thinks some doctrinal disagreements are hard to accommodate on this kind of view:
If, for instance, we want (some versions of) Catholicism to count as reasonable, and (some versions
of) atheist comprehensive liberalism to count as reasonable, and if no version of Catholicism views
any version of atheism as a near miss, then public reason theorists cannot understand the
reasonable’s beliefs in terms of such near misses” (2017, 147).
I think the “near miss” locution obscures the potential for developing this kind of response in terms of suboptimal
rationality. It’s not obvious that reasonable versions of Catholicism and atheist comprehensive liberalism can’t view
their respective counterparts as acceptably though suboptimally rational. This kind of proposal is also compatible with
the UNIQUENESS thesis.
99
See Staffel 2015’s discussion of measures of overall incoherence of credence functions.
100
If one’s experience is part of one’s evidence, as seems plausible, then it’s obvious that much of our evidence is
radically private, such that we can’t hope to share anything but the most broad outlines of that evidence using
testimony. So we cannot expect that testimony is always capable of bridging the gap between distinct subjects’ total
bodies of evidence.
101
Raz 1990, 9.

64
So to assert the truth of a doctrine of justice, or claim its truth as a reason for its acceptance, is to
violate the principle of LIBERAL NEUTRALITY. Therefore, political liberal theorists have good reason
to practice epistemic abstinence.
For similar reasons, I’ve argued that public reason and political liberal theorists should
practice epistemic abstinence with respect to what hypotheses reasonable citizens should accept
concerning doctrinal disagreement. But Raz argues that this move is unsustainable:
A theory of justice can deserve that name simply because it deals with these matters,
that is, matters that a true theory of justice deals with. In this sense there are many
theories of justice, and they are all acceptable to the same degree as theories of
justice. To recommend one as a theory of justice for our societies is to recommend
it as a just theory of justice, that is, as a true, or reasonable, or valid theory of justice.
If it is argued that what makes it the theory of justice for us is that it is built on an
overlapping consensus and therefore secures stability and unity, then consensus-
based stability and unity are the values that a theory of justice, for our society, is
assumed to depend on. Their achievement – that is, the fact that endorsing the
theory leads to their achievement – makes the theory true, sound, valid, and so forth.
This at least is what such a theory is committed to. There can be no justice without
truth.
102

Raz’s objection seems to be that in order for a purported theory t to count as a real theory of justice,
it must both (a) deal with the appropriate subject matter (in this case, justice), and (b) recommend
itself to us on the basis of its correctness with respect to that subject matter. Raz’s objection seems
to be that a theory displaying epistemic abstinence cannot satisfy requirement (b). Political

102
Raz 1990, 15. It helps to distinguish Raz’s objection from an implausibly strong claim that might be thought to
underwrite it. One might claim that it’s a constitutive feature of what it is to be a theory of justice that it (explicitly)
claim to be the one true theory of justice, and hence part of the true comprehensive normative theory. On this
interpretation, unless a veridic claim of this kind is (explicitly) built into the content of a purported theory t, t simply
cannot be a real theory of justice, as a matter of conceptual necessity. In general, this desideratum would disqualify
too many purported theories across domains from being real theories: after all, how many scientific theories have been
canonically formulated to include a(n explicit) claim to being true? This desideratum appears no less implausible when
understood as a domain-specific requirement of normative theories (though, no doubt, there are some normative
ethicists who stand ready to add a veridic claim to their normative theory if it helps their cause against public reason
or political liberalism).

65
liberalism commends to us its ability to secure stability and unity, and in so doing Raz thinks it
falls short of being a real theory of justice.
Following Raz, it might be objected that my political conception of reasonable doctrinal
disagreement fails to give us a real theory of reasonable doctrinal disagreement for an analogous
reason. For while the political conception deals with the subject matter of reasonable doctrinal
disagreement, it commends itself to you the reader on the basis of its ability to avoid each horn of
Enoch’s trilemma and preserve public reason’s ability to perform its ecumenical function within
political liberalism.
However, as in the case of political liberalism, this is not the only sense in which it
commends itself. Political liberalism and public reason theory alike, recommend themselves to
prospective liberals on the basis of appearing true by the lights of their reasonable worldviews.
Political liberalism and public reason theory abjure to justify themselves in terms of any specific
non-public doctrine, but it nonetheless falls to their adherents to do so on the basis of their own
non-public doctrines. Therefore, from the personal perspective of the reasonable, there’s no
problem of political liberalism or public reason theory failing to recommend themselves on the
basis of their correctness with respect to their subject matters, per (b).
From the impersonal perspective of the theories themselves, there is no stated set of non-
public doctrines in virtue of which either theory claims to be true, but crucially, this doesn’t mean
that neither theory says there is no fact of the matter. Instead, borrowing an analogy from Bayesian
models of vague credences, we may think of the theoretical perspectives of political liberalism and
public reason theory as being modeled by a committee made up of all their reasonable adherents.
Adherents endorse these theories for a variety of their own non-public reasons, so there is no single
fact of the matter about what it is, in virtue of which, political liberalism or public reason theory

66
claim to be correct, but there is nonetheless a fact of the matter. Therefore, there is no problem
arising from the epistemic abstinence of the political conception of reasonable doctrinal
disagreement, and no threat of a revenge of the third horn of the trilemma.

4.3 Sophisticated Psychological Hypotheses and the Second Horn
So, the reasonable citizen has an impressive toolkit of deflationary, evidential, and principled
epistemological hypotheses at her disposal for supplementing her understanding of doctrinal
disagreement in a way that both renders her coherent (and plausibly reasonable), and can justify
the publicity constraint. However, Enoch argues that even sophisticated psychological hypotheses
about the causes of doctrinal disagreement, which explicitly aim at preserving the evaluation of a
disagreement as reasonable, undermine the individual’s motivation for adhering to the publicity
constraint, and thereby fall to the second horn of the trilemma. These sophisticated hypotheses
include attributing to the disagreeing party a performance-error in reasoning on a particular subject
matter despite their overall epistemic competence, or an excusable error.
103
Enoch ascribes the first of these proposals to Leland and van Wietmarschen, who favor a
view on which the reasonable are expected to manifest a very high level of epistemic
competence.
104
Suppose the reasonable Catholic thinks reasonable non-Catholics may be highly
epistemically competent. One can believe a proposition p, while also believing that someone else

103
See Enoch 2017, §§2.3 and 2.4. It may be that one or more of these proposed explanations amounts to a version of
an “inconclusive evidence” story (see §2.2), but I won’t address that question here.
104
Leland and van Wietmarschen 2012, § III. I don’t think Leland and van Wietmarschen have to accept this
characterization of their view. Doctrinal disagreements, even among highly epistemically competent persons, can be
equally well explained by some versions of permissivism, or as epistemic near misses, or in terms of the privacy of
evidence, so they needn’t appeal to performance errors, as Enoch contends.

67
who is highly epistemically competent on matters relating to p, and who is exposed to the same
evidence, nonetheless believes not-p. Just as when two otherwise reliable thermometers provide
different readings we presume one result to be the result of a performance error, and use the rest
of the available evidence to decide which is most likely to be correct, there’s no big mystery here.
So thinking that highly epistemically competent people disagree with you about the truth of
Catholicism does not preclude one from thinking that Catholicism is true.
105

However, given the high level of expected epistemic competence assumed in Leland and
van Wietmarschen’s picture, what can explain the remaining disagreement over Catholicism?
Enoch infers that the best explanation available to the reasonable Catholic is for her to attribute
some kind of glitch in reasoning to all otherwise highly epistemically competent non-Catholics,
such that these glitches account for their non-Catholicism. If this is right, it invites impalement on
the second horn of Enoch’s trilemma: “why exclude from political debate considerations that are
only controversial because of such glitches? This doesn’t seem to rationalize excluding these
considerations.”
106
I agree with Enoch that this kind of glitch-theoretic explanation is untenable.
For it seems that a glitch must be a function of some kind of ( 𝛼𝛼) irrationality or logical error, ( 𝛽𝛽 )
prejudice or bias, ( 𝛾𝛾) self- or group-interest, or ( 𝛿𝛿 ) willfulness, which overrides the otherwise
reliable epistemic competencies of a disagreeing party on a particular occasion. But if A
understands the cause of B’s disagreement in this way it undermines (or will at least tend to
undermine) A’s confidence in B’s epistemic competence, giving A a reason to correct B so as to
preserve or restore B’s ability to cooperate on fair terms. But once A engages in correcting errors,

105
Enoch 2017, 144.
106
Enoch 2017, 146.

68
it’s not clear why she should hold back from telling B the whole truth, by the lights of her non-
public doctrines. In this way, even this sophisticated psychological hypothesis undermines the
reasonable citizen’s motivation for adhering to the publicity constraint, and it thereby falls to the
second horn of Enoch’s trilemma.
For similar reasons, reasonable doctrinal disagreement cannot be understood as the result
of an excusable error. On this proposal, it’s hypothesized that there are considerations that do not
justify a belief, but that nevertheless negate the believer’s responsibility, culpability, or
blameworthiness for believing unjustifiably. Even if there are such considerations, it’s not obvious
why the excusable but nonetheless erroneous rejection of one’s non-public doctrines by reasonable
others should make appeals to those commitments publicly inadmissible in political discourse.
Furthermore, since almost any commitment–however ludicrous or repugnant–can be non-culpably
held under certain circumstances, the publicity constraint approaches vacuity, since it may not
exclude any commitment at all. In this way, explaining reasonable doctrinal disagreement in terms
of an excusable error runs afoul of the second horn of Enoch’s trilemma.

69
III. Virtue Reliabilism for Liberal Anti-Perfectionists
Liberal thinkers take a variety of approaches to theorizing civic virtue. Liberal anti-perfectionists
have tended to think that virtuous agents must be praiseworthy in acting in conformity with virtue,
but this normative requirement of praiseworthiness has not been properly justified. It, therefore,
burdens these theories with needless normative baggage. Virtue reliabilism avoids this problem by
dispensing with the requirement of praiseworthiness and remaining metaphysically,
psychologically, and normatively thinner than its rivals. In so doing, it remains better positioned
to count as anti-perfectionist. I show that this thinner account of liberal civic virtue is also able to
respond to recent challenges from James Boettcher (2012) and Emily McTernan (2014). Contrary
to Boettcher, reliabilist virtue-theoretic accounts of civic duty can be as stringent as the liberal
theorist needs them to be. And, contrary to McTernan, reliabilist virtue theories that embrace a
certain kind of externalism about virtue can benefit from empirical insights into the causal efficacy
of social norms. Liberal anti-perfectionist virtue theorists already implicitly accept externalism,
but no one has previously advanced an explicitly reliabilist account of anti-perfectionist liberal
virtue. Yet, as I show, virtue reliabilism fares better than its rivals in responding to all these
challenges, and it does so without straying into perfectionism.
1. Three Challenges for Anti-Perfectionist Virtue Theories
According to Martha Nussbaum (2011), liberal perfectionists, like Isaiah Berlin, Joseph Raz, and
Steven Wall, espouse “a type of liberal political view that spells out a set of controversial
metaphysical and ethical doctrines concerning the nature of value and the good life, and then goes
on to recommend political principles built upon these values” (3). In contrast, liberal anti-
perfectionists, like Ronald Dworkin, Gerald Gaus, Charles Larmore, Thomas Nagel, John Rawls,

70
and Nussbaum herself “hold that the state should be neutral among rival understandings of the
good” (Wall 2019, §3.1). Liberal perfectionists are free to regard distinctively liberal civic virtues
as of intrinsic value and appeal to this value in their normative liberal theorizing. But anti-
perfectionists cannot appeal to the intrinsic value of civic virtue (without thereby falling into
perfectionism). Thus, as Christie Hartley & Lori Watson (2014) explain, liberal perfectionists and
anti-perfectionists “differ markedly with respect to their accounts of the role and content of civic
virtue. Given that anti-perfectionists do not think the state should aim to promote a particular view
of the good life […] one might question whether anti-perfectionist political theories include or are
compatible with an account of civic virtue at all” (415). However, anti-perfectionists can (and
several do) argue that civic virtue is of indispensable instrumental value to a liberal society.
107

Nevertheless, virtue theories face several challenges.
Firstly, M. Victoria Costa (2004) observes that “admitting excellences of character
involves some risk of perfectionism” (151), and I argue that most extant anti-perfectionist virtue
theories have been saddled with unnecessary normative baggage, causing them to stray into
perfectionist territory. Specifically, I argue that several of these theories needlessly require
manifestations of virtue to be praiseworthy. Secondly, James Boettcher (2012) argues that virtue
theories are insufficiently authoritative to capture the putative stringency of civic duties, like
Rawls’s “duty of civility.” Finally, Emily McTernan (2014) contends that social norms are
empirically better suited to the instrumental roles assigned to virtues by liberal theorists. I address
each of these challenges, respectively, in sections 3, 4, and 5 of this chapter, before summarizing

107
See e.g., Costa 2004, 152; Hartley & Watson 2014, 424.

71
conclusions in section 6. However, in section 2, I begin by providing an overview of liberal
theories of civic virtue.
2. Liberal Virtue Theories
The core idea of a virtue is that it is a trait that is manifested in some kind of performance. On the
full-blown Aristotelian view, ethical virtues are necessarily (a) the product of habituation, with
their possession and manifestation being (b) honourable or praiseworthy, and (c) necessary
components of a flourishing life, but rival accounts of virtue may reject or remain neutral about
any or all these distinctive Aristotelian commitments (a)–(c). One of the important distinctions for
our purposes, demarcating different resources for responding to the challenges listed above, turns
on whether virtue is defined in terms of one or both of the following:
Success-Oriented Disposition: a stable, if imperfect, disposition to perform in
ways that are (sufficiently) good, in terms of some norm or value (often glossed
as a disposition to “act in conformity with virtue”);
Insight-Dependent Attempt: seeking what is (sufficiently) good, in terms of some
norm or value, based on some understanding of that norm or value (often
glossed as “acting virtuously” or “acting from virtue”).

Reliabilists, like Ernest Sosa and Julia Driver,
108
think a Success-Oriented Disposition is necessary
and sufficient for possessing the relevant virtue. Most responsibilists, like Immanuel Kant
109
James
Montmarquet, and Michael Slote
110
maintain that a stable disposition toward Insight-Dependent

108
Battaly and van Zyl 2015, 12–14.
109
Wilson & Denis 2018, §4.
110
Battaly and van Zyl 2015, 17–18.

72
Attempts is necessary and sufficient for possessing the relevant virtue. Some responsibilists, like
Thomas Hurka (2006), do not even require a stable disposition in the agent, being satisfied if the
agent possesses the right kinds of occurrent states or attitudes toward a value. Hybrid theorists,
like Aristotle, Rosalind Hursthouse, and Linda Zagzebski
111
maintain that a Success-Oriented
Disposition manifested through Insight-Dependent Attempts is necessary and sufficient for
possessing the relevant virtue. And pluralists take different positions on different virtues. For
example, Heather Battaly (2015) maintains that neither Success-Oriented Dispositions nor
dispositions toward Insight-Dependent Attempts are necessary, but she thinks each kind of
disposition is sufficient on its own for possessing a kind of virtue. Most of these positions are
represented among liberal virtue theorists.
112

2.1 Perfectionist and Miscellaneous Liberal Virtue Theories
Liberal perfectionists exhibit great variation in their treatments of civic virtue, with a tendency
toward pluralism. William Galston (1991) provides a broadly reliabilist survey of a wide range of
generic social, economic, and political liberal virtues, but when he turns to consider the possibility
of “a conception of the virtuous or excellent individual linked intrinsically to liberal theory and
seen as valuable, not instrumentally, but for its own sake” (229), he emphasizes the responsibilist

111
Battaly and van Zyl 2015, 15–17.
112
In addition, Mark LeBar (2013) argues that a range of contemporary virtue ethical theories grounded in neo-
Aristotelian convictions about human flourishing, including those of Slote, Hursthouse, and Nussbaum, leave open
“the possibility of taking on board a particular virtue which would sustain a second-order, justificatory form of [anti-
perfectionist] liberalism” (285), but his argument for this conclusion remains highly schematic.

73
elements of self-direction advanced by famed liberal perfectionists, like Locke, Kant, and Mill.
113

Stephen Macedo (1991) is a responsibilist about the virtue of principled moderation,
114
but his
account of the virtue of self-governance is more complex: Macedo is a reliabilist about the less
praiseworthy expression of self-governance in those who are merely autarchic and a hybrid theorist
about its more praiseworthy expression in those who are fully autonomous.
115
Richard Rorty’s
account of the civic virtue of irony is similarly complex: the ordinary liberal citizen may exhibit a
reliabilist “Golden Mean between a deficiency of commitment to one’s values and beliefs […] and
an excess of commitment” (Curtis 2016, 407), but “ironist intellectuals” have the added
understanding of irony, which makes their expression of this virtue praiseworthy.
116
Richard
Dagger (2000) provides a responsibilist gloss of the civic virtue of respect for autonomy, but he
gives a hybrid description of fair play.
117
Galston, Macedo, Rorty, and Dagger, therefore, seem to
be pluralists of various kinds. James Gregory (2014), meanwhile, advances a hybrid account of a
liberal virtue he calls “balance.”
118

113
See Galston 1991, 229–231.
114
Costa 2004, 159; see Macedo 1991, 72–73.
115
Self-governance grants equal moral and political standing to persons, but Macedo distinguishes between self-
governance directed merely instrumentally (i.e., “autarchy”) and self-governance directed critically (i.e.,
“autonomy”), the latter of which he regards as the ultimate liberal virtue, because it is critically directed toward ends
that enhance one’s experience of self. See Macedo 1991, 215–227 and Burtt 1991, 682.
116
Being “eligible for the Bloomian accolade of “strong poet,” a title that Rorty stretches beyond writers of verse to
apply to innovators across the cultural spectrum” (406); see Curtis 2016, 405–410.
117
Writing of respect for autonomy, Dagger (2000) says that “The virtuous citizen will want to protect and promote
the ability to lead a self-governed life, in himself or herself and others, not only as a matter of right but also as
something that is intrinsically valuable” (196). But he describes fair play as a disposition to recognize “that other
citizens are not merely potential obstacles to or allies in the accomplishment of one's personal projects but partners in
a common enterprise who deserve civil treatment,” adding that its successful inculcation and exercise depends on “the
citizen truly [seeing] himself or herself as part of a common enterprise ... [and regarding] political participation as a
necessary contribution—and perhaps even an enjoyable one—to the good of the community” (197).
118
See Gregory 2014, 87–89.

74
Among liberal theorists not so easily classified as perfectionist or anti-perfectionist there
is some variation but an overall tendency toward reliabilism. For example, John Horton (1996) is
a kind of deep value pluralist about the virtue of toleration, remaining unpersuaded that there is
“any one uniquely rational perspective […] from which the reasonableness or rightness of a
specific substantive conception of the virtue of tolerance can be established. Tolerance is not a
virtue that stands altogether outside the moral and political conflicts it often seeks to mediate” (40).
However, Horton’s value pluralism about the virtue bears some resemblance to Bernard
Williams’s reliabilism about the attitude. Williams (1996) is a hybrid theorist about the virtue of
toleration, but a reliabilist about the attitude of toleration, which he regards as the surer hope for
liberal societies, thanks to its lack of responsibilist psychological demands and its “wider and more
mixed range of [motivational] resources” (27).
119
Teresa Bejan (2017) extols Roger Williams’
reliabilist theory of “mere civility,” arguing somewhat similarly that it has the stronger claim to
being minimalist, in comparison with those advanced by Hobbes and Locke.
120
And Daniel

119
Williams distinguishes between a reliable practice of toleration, an attitude constituting a disposition toward this
reliable practice, and the full-blown virtue of toleration: “A practice of toleration means only that one group as a matter
of fact puts up with the existence of the other, differing, group. A tolerant attitude (toward this group) is any disposition
or outlook that encourages them to do so: it is more likely to be identified as an attitude of toleration if it applies more
generally, in their relations to other groups, and in their views of other groups’ relations to each other. One possible
basis of such an attitude—but only one—is a virtue of toleration, which emphasises (sic.) the moral good involved in
putting up with beliefs one finds offensive” (19). For Williams, the virtue requires that a Success-Oriented Disposition
to practice toleration be the manifestation of an attitude producing Insight-Dependent Attempts, motivated by
recognition of “toleration as an intrinsic value or [...] the respect for autonomy that underlies toleration as a virtue [...]
its worth lies partly in its difficulty, in [...] its demand that one combine the pure spirit of toleration with one’s
detestation of what has to be tolerated” (26). Williams argues that this psychologically demanding dimension of the
virtue “limits the range of people who can possess it. Because of this, it is a serious mistake to think that this virtue is
the only, or perhaps the most important, attitude on which to ground practices of toleration” (19).
120
See especially Bejan 2017, 160 and 163.

75
Weinstock (2006) explores a reliabilist account of reasonableness and its cognate civic virtues in
his deliberate effort to find a middle ground between perfectionism and anti-perfectionism.
121

However, reliabilism’s minimalism and psychological attainability has yet to be fully appreciated
by anti-perfectionists.

2.2 Anti-Perfectionist Liberal Virtue Theories
Hartley & Watson (2014) are very explicit about the need for anti-perfectionists to avoid appeals
to intrinsically valuable civic virtues,
122
so perhaps unsurprisingly, they come closest to
recognizing the anti-perfectionist appeal of virtue reliabilism. Their introductory discussion of
Rawls’s theory of civility omits its responsibilist elements, leaving the impression that Rawls is a
reliabilist about civility (as I argue below, he is not).
123
Their accounts of tolerance
124
and what

121
Weinstock’s proposed virtue aims at practical compromise, rather than a Rawlsian overlapping consensus of public
justification. He provides a brief sketch of the relative value of compromise as a virtuous end, but his account does
not require a reasonable agent to explicitly understand or act on its value. Instead, he describes the reasonable agent
as being “stably disposed to prefer compromise over conflict” (243), “stably disposed to reach compromises with their
fellow citizens, rather than run the risk of continued normative warfare” (244), and possessing all the “cognate virtues
and dispositions which make compromise possible: an ability to understand the value commitments of those with
whom one is engaged in negotiation, and an ability to distinguish within one’s own commitments that which can be
sacrificed for the sake of the greater good of social peace, and that on which one must stand firm on pain of loss of
integrity, and the like” (244).
122
See Hartley & Watson 2014, 424.
123
See Hartley & Watson 2014, 423.
124
Hartley & Watson understand tolerance as “a disposition to engage in acts of toleration in a tolerant way” (2014,
427). Performing tolerant acts tolerantly means “inter alia, that such acts should not be difficult for the agent to
perform” (426), so as to “show full acceptance of central political values” (427, my emphasis), such as respect for
one’s fellow citizens. Therefore, the way in which these acts are performed should convey that they are manifestations
of toleration, as opposed to merely random or accidental behaviors, but they need not involve any understanding or
toleration’s value or Insight-Dependent Attempts to express a tolerant attitude.

76
Rawls calls “full autonomy” are also reliabilist.
125
However, other prominent liberal anti-
perfectionists have generally adopted hybrid theories of civic virtue.
2.2.1 Rawls’s Hybrid Theory of Civic Virtue
In the original (1971) and revised (1999) editions of A Theory of Justice and in Justice as Fairness:
A Restatement (2001), Rawls defines the virtues as “sentiments, that is, related families of
dispositions and propensities regulated by a higher-order desire, in this case a desire to act from
the corresponding moral principles” (1971, 192),
126
which can be individuated on the basis of these
principles.
127
This consistently hybrid definition of virtue helps to make sense of the picture Rawls
develops in Political Liberalism (1993), affirming the “superiority of certain forms of moral
character” or “moral virtues,” including civility, tolerance, reasonableness, and a sense of fairness,
which are counted among the virtues of social cooperation and characterize the ideal of the good
citizen of a democracy (194).
128
I will focus on just two of these four virtues: reasonableness and
civility.

125
With respect to autonomy, citing Rawls (1993, 77–78), Hartley & Watson argue that:
autonomy does not require individuals to incorporate respect for some kind of substantive, nonpolitical autonomy into
their lives. For example, individuals need not value autonomy as self-government as part of their idea of the good.
They are free to non-reflectively pursue any comprehensive doctrine that doesn’t involve the violation of fellow
citizens’ basic rights. It may even be contrary to their comprehensive doctrine to place value on autonomy understood
as self-government. Rather, political autonomy of the sort described here fundamentally concerns participation in
public life, and fully autonomous action publicly expresses to fellow citizens one’s commitment to political
liberalism’s principle of reciprocity (2014, 427–428).
126
Cf. Rawls 1999, 275 and 383; 2001, 126; see also Berkowitz 1999, 24.
127
See Rawls 1971 437; 1999; 383
128
See also Rawls 2001, 116–119, 142

77
Reasonableness seemingly applies to the entire domain of civic life, having two basic
aspects: “the willingness to propose fair terms of cooperation and to abide by them provided others
do [and] the willingness to recognize the burdens of judgment and to accept their consequences
for the use of public reason in directing the legitimate exercise of political power in a constitutional
regime” (54). These dispositions or propensities seem to be regulated by the higher-order desire
for “a social world in which [persons], as free and equal, can cooperate with others on terms all
can accept. [Reasonable persons] insist that reciprocity should hold within that world so that each
benefits with others” (50). So, reasonableness is largely a matter of being committed to fair
cooperation in all matters of civic life.
129
To help specify this highly general virtue, consider
Rawls’s rather detailed explanation of its application alongside civility.
The “ideal of (democratic) citizenship” is a similarly wide-ranging normative feature of
Rawls’s picture, but it imposes a specific moral duty, which he calls “the duty of civility,” defined
by two constraints on public discourse proposing or defending exercises of political power:
Publicity Constraint: Agents must be able to explain to one another how the
exercises of political power, which they defend or advocate, can be justified
based solely on public reasons;
130

Fair-Mindedness Constraint: Agents must be willing to listen to others and be
fair-minded in deciding when to accommodate their views (see 1993, 217).

129
See Dagger 2013, 302.
130
Rawls recognized roles for advancing non-public justifications outside of public reasoning, through other forms of
discourse, which he called declaration, reasoning by conjecture, and bearing witness (see 2005, 465–466;
Schwartzman 2012).

78
In reference to this narrower domain of public discourse, Rawls describes reasonableness and a
readiness to observe these two constraints as “political values,” specifically “values of public
reason,” as well as “political virtues,” “virtues of citizenship,” and “guidelines of public inquiry,”
which serve as “principles of reasoning and rules of evidence in the light of which citizens are to
decide whether substantive principles [of justice] properly apply [to a political question at issue in
a context] and identify laws and policies that best satisfy [those substantive principles of justice]”
(224).
131
As Rawls explains in “The Idea of Public Reason Revisited” (2005; 1997), the ideal of
public reason is realized in the public political forum by “judges, legislators, chief executives, and
other government officials, as well as candidates for public office, [who] act from and follow the
idea of public reason and explain to other citizens their reasons for supporting fundamental
political positions in terms of the political conception of justice they regard as the most reasonable”
(444). This ideal is realized in ordinary citizens by a firm and widespread disposition to “think of
themselves as if they were legislators and ask themselves what statutes, supported by what reasons
satisfying the criterion of reciprocity [i.e., substantive principles of justice], they would think it
most reasonable to enact [while repudiating] government officials and candidates for public office
who violate public reason” (444–445, original emphasis).
132

We can synthesize Rawls’s picture as follows: the ideal of democratic citizenship consists
of several virtues of social cooperation, including reasonableness and civility. Each virtue is a
family of dispositions and propensities regulated by a higher-order desire to act in accordance with
corresponding moral ideals, values, or principles. Reasonableness is regulated by a higher-order

131
See also Rawls 2001, 92.
132
See also Rawls 2001, 89–92.

79
desire for a social order where all are free, equal, and live together on terms that everyone accepts.
Reasonableness and civility have over-lapping application to public inquiry, and in this domain,
they are (also) regulated by a higher-order desire to act in accordance with the moral duty of
civility, characterized distinctively (though perhaps not exhaustively) by the Publicity and Fair-
Mindedness Constraints. Following Costa’s suggestion, we may further distinguish Rawlsian
civility from reasonableness by looking to Cheshire Calhoun’s proposal that civility is a disposition
“to communicate basic moral attitudes of respect, tolerance and considerateness” (2000, 255; see
Costa 2004, 154). Meanwhile, we may suppose that reasonableness, within this domain, is
primarily concerned with one’s own moral and intellectual conduct in judging, for example, when
reasons are suitably public and when there is sufficient justification to accommodate others’ views,
since fulfilling the Publicity and Fair-Mindedness constraints is indispensable to the over-arching
goal of fair cooperation, at which reasonableness aims.
133

That reasonableness and civility are families of Success-Oriented Dispositions (or
propensities) is strongly implied by the success conditions for realizing the ideal of public reason.
That reasonableness and civility require Insight-Dependent Attempts at public reasoning, based on
some understanding of the ideal of democratic citizenship, the ideal of public reason, its political
value, or the duty of civility, is also clear. So, Rawls is most plausibly read as a hybrid theorist
about these central anti-perfectionist virtues.

133
Cf. Rawls 2001, 170; Costa 2004, 163. Obviously, the other virtues of social cooperation, tolerance and a sense of
fairness, have some role to play in these deliberations too.

80
2.2.2 Audi’s Hybrid Theory of Civic Virtue
Robert Audi, a Christian moral perfectionist,
134
nonetheless argues for a position much like the
political liberal anti-perfectionism of Rawls. Audi shares Rawls’s commitments to political
justification and restraint, and his major principles of secular rationale and secular motivation,
developed as part of a framework for separating the secular and the religious in the conduct of
individuals, are presented as “ideals appropriate to a free and democratic society: standards that
contribute to the health and strength of such a society, even if citizens may be within their moral
(and legal) rights in adopting alternative ideals that permit a lesser separation [of Church and
State]” (1991, 66; original emphasis). Audi (2000) advances a hybrid theory of civic virtue:
A virtue is a feature of character with a significant capacity to influence conduct. A virtue
supplies its possessor both with normative reason indicating what sort of thing should be done in
a wide range of contexts and with motivation to do such things for the right kind of reason. Virtue
is not a mere capacity for good deeds, but a settled tendency to do them for an appropriate reason.
Civic virtue in particular, at least in the form appropriate to citizenship in liberal democracy, is
constituted above all in relation to protection and promotion of the flourishing of civil society.
This implies a disposition on the part of civically virtuous citizens to participate in sociopolitical
decisions and a determination to do so with respect for the freedom and autonomy of others. It
implies (within limits) a commitment to having, and a willingness to offer, publicly

134
Christopher Eberle (2002), another Christian perfectionist, is far more critical of liberalism, but proposes his own
justificatory principle inspired by the anti-perfectionist’s Publicity Constraint. According to Eberle's proposal, the
conscientious citizen “should pursue public justification for his favored coercive laws” (68; original emphasis). As he
makes clear, “an obligation to pursue public justification doesn’t imply an obligation to exercise restraint” (70); it
requires only “a sincere and conscientious aspiration to public justification” (75; emphasis added). This relaxation of
the Publicity Constraint leaves Eberle’s account of civic conscientiousness an (illiberal) example of Hurka’s non-
dispositional occurrent-state responsibilism. See Gaus 2003.

81
comprehensible, evidentially adequate reason for one’s sociopolitical conduct, particularly when
that conduct is in support of a law or public policy that would restrict the liberty of citizens (180).
Audi makes it clear that the possession of virtue is compatible with committing some errors
but requires Insight-Dependent Attempts to manifest a Success-Oriented Disposition, calling it “a
practical success, not a theoretical achievement. Its intellectual requirements are quite
latitudinarian and can be met by a self-conscious rule theorist or by a moral skeptic suspicious of
ethical concepts. [...] Having such a sense of appropriateness is part of what it is to understand the
field of a virtue […] Without that understanding, one would not act from virtue” (148–149). So,
Audi is as clear a hybrid theorist as one can be.
2.2.3 Costa’s Theory of Civic Virtue
Costa highlights the need for anti-perfectionists to avoid appeals to the intrinsic value of civic
virtue,
135
and this may explain the reliabilist bent of her discussion, but her overall view is either
a hybrid theory or a virtue pluralism that includes a hybrid theory of (at least) toleration. Costa
endorses Rawls’s theory of civility but expands its application beyond public dialogue,
understanding it “as a disposition to express attitudes of respect in different sorts of social
interactions” (2004, 155). Perhaps because she takes this wider view, she glosses the Publicity and
Fair-mindedness Constraints as part of the “practice of democratic dialogue” (163), rather than
treating them as the characteristic moral principles of civility or the objects of a higher-order

135
Costa 2004, 152.

82
regulative desire to be civil. She is at least somewhat permissive about the kinds of reasons or
motivations one may have for exercising other liberal virtues like moderation or personal justice,
136

suggesting, per reliabilism, that acts, behaviors, or manifestations of disposition need not be based
on any understanding of the relevant value, to count as manifestations of virtue. However, she also
tries to show how normatively loaded and psychologically rich a conception of civic virtue can be,
while remaining compatible with a Rawlsian political conception of justice.
137
And when it comes
to public dialogue, Costa argues that in addition to Success-Oriented Dispositions to “present one’s
position clearly, to understand and appreciate other people’s points of view, and to exercise
moderation in judgment in order to find consensual agreements. Negatively, [...] one should avoid
vices such as charlatanism, empty rhetoric, cowardice in defending one’s convictions, or making
concessions to any position without scruples” (162–163; cf. Berkowitz 1999, 179–180). Costa
understands toleration similarly, as a Success-Oriented Disposition “of character, controlling the
inclination to react in an intolerant way” (157), and she says that toleration’s exercise is also
incompatible with certain vicious motives. But then she goes further, asserting that “the exercise
of toleration is the recognition that certain expressions of intolerance are not compatible with
mutual respect” (158). Thus, Costa is clearly a hybrid theorist about the virtue of toleration, her

136
See Costa 2004, 161 and 163–164.
137
For example, Costa writes: “political virtue is a complex set of dispositions of perception, emotion, judgment,
choice, and behavior that are essential to maintain fair social cooperation among free and equal citizens. Also, as is
true of any other type of virtue, civic virtue consists of a set of excellences that are exhibited in a nonsporadic and
nonarbitrary way. This understanding of virtue has an Aristotelian flavor because it admits the interaction of emotional
and intellectual components in virtuous action, and the importance of molding character in order to generate stable
dispositions […] defending civic virtue from the political perspective involves respecting certain restrictions in the
strategies of justification, but it does not require denying that civic virtue, as a set of stable dispositions, involves a
deep molding of the character of people” (152–153).

83
discussion of the other civic virtues makes their reliabilist requirements clear, and it leaves the
door open to all of them (especially civility) being hybrid virtues as well.
2.2.4 Liveriero’s Virtue Theory of Reasonableness
Federica Liveriero (2020) focuses on the virtue of reasonableness, calling it “a fundamental
concept in the literature concerning political liberalism and public justification” (902). Unlike
Rawls, she does not employ reasonableness as a criterion for assessing the legitimacy of political
decisions or the non-publicity of “comprehensive” doctrines. Instead, she thinks of reasonable-
ness as a practical virtue, with moral and epistemic dimensions, and she is a responsibilist or hybrid
theorist about both. She introduces the following two constraints:
Reciprocity Constraint: agents are willing to publicly discuss fair terms of
cooperation, because they have a sense of justice (in a very minimal, strictly
political, sense) and are motivated by it;
Public Reason Constraint: accepting the burdens of judgement implies that agents
have good reasons for abiding by a general norm of epistemic modesty and
enter into public deliberation mutually respecting each other as both moral and
epistemic authorities (906).

Agents prove reasonable, in Liveriero’s moral sense, by accepting the Reciprocity Constraint and
being willing to abide by fair terms of social cooperation, provided others do the same, mutually
respecting each other’s free and equal moral status. There may be Success-Oriented Dispositions
involved in her conception of what it takes to accept the Reciprocity Constraint, fulfill its
requirements, or respect others’ free and equal moral status, but the Reciprocity Constraint is a
requirement of justice, which itself requires that one be motivated by one’s sense of justice, so
Insight-Dependent Attempts are built into this constraint.
Agents prove reasonable, in Liveriero’s epistemic sense, by accepting the burdens of
judgement and restraining themselves from trying to impose their private perspectives when
publicly debating political matters, respecting the epistemic authority of other agents. Again,

84
Success-Oriented Dispositions may be involved in her conception of what it takes to accept the
burdens of judgement, fulfill its requirements, or exercise the requisite kind of restraint. She also
implies strongly that it involves Success-Oriented Dispositions to (i) settle disagreements with
fellow citizens by appeal to an external epistemic authority, whose authority all parties to the
disagreement recognize as legitimate to resolve the conflict, or (ii) abstain from a dogmatic
attitude, where no such recourse is available. But, again, this epistemic constraint “requires
reasonable citizens to undertake an attitude of epistemic modesty while deliberating public matters
with agents who hold views different from theirs. […] reasonable citizens are […] aware of the
limitations of humans’ doxastic activities […] are ready to engage in public deliberative processes
and to share epistemic authorities with their fellow citizens” (909–910, original emphasis). Hence,
Insight-Dependent Attempts are built into this constraint as well.
3. The Problem of Praiseworthiness and Virtue Reliabilism
As we saw in the previous section, several prominent liberal anti-perfectionists, including Rawls,
Audi, Costa, and Liveriero are hybrid virtue theorists. However, the praiseworthiness of the agent,
and the (moral) value, worth, or goodness of acting from virtue, through Insight-Dependent
Attempts, goes unremarked and plays no clear role for Costa or Liveriero. This psychological
requirement of agential insight into the value of their behavior, therefore, risks burdening these
anti-perfectionist projects with unnecessary psychological baggage, by placing gratuitous demands
on the virtuous. Such superfluous elements are best avoided by a normative political theory that
aspires to avoid the promotion of a particular view of the good life. Anti-perfectionists, therefore,
are best advised to keep their theories as minimal, ecumenical, and attainable as possible. By
contrast, the praiseworthiness of the agent and the (moral) value of acting from virtue is

85
acknowledged in the theories of Rawls and Audi, but, as I will argue, its inclusion is still not
justified.
To begin to see why, consider a case where one agent or group A is entirely motivated by
earning the approval of another agent or group B. B reliably approves of civically virtuous
behavior. A’s Success-Oriented Disposition to act in conformity with virtue, therefore, is not
manifested through Insight-Dependent Attempts to act virtuously, because its members do not act
based on their understanding of the relevant values, but only on the basis of B’s approval.
Nevertheless, A succeeds in manifesting this virtuous disposition.
138
If this is enough, the added
psychological requirement of hybrid theories, that civic virtues be expressed through praise-
worthy Insight-Dependent Attempts, is unnecessary for liberal anti-perfectionism’s instrumental
purposes, so its inclusion is pointless and may reflect a perfectionist instinct to value
praiseworthiness for its own sake. Call this the problem of praiseworthiness.

3.1 Rawls on the Moral Worth of Acting from Virtue
In A Theory of Justice, Rawls distinguishes between natural assets or powers and moral virtues:
natural powers [are] developed by education and training, and often exercised in
accordance with certain characteristic intellectual or other standards by reference
to which they can be roughly measured. The virtues on the other hand are
sentiments and habitual attitudes leading us to act on certain principles of right […]
A good person, then, or a person of moral worth […] has the features of moral
character that it is rational for members of a well-ordered society to want in their
associates (1971, 437; emphasis added; cf. 398).
139

138
For discussion of similar cases, see McTernan 2014, 97.
139
See Rawls 1999, 383–384, cf. 349.

86
Rawls later summarizes that “the moral virtues are excellences, attributes of the person that it is
rational for persons to want in themselves and in one another as things appreciated for their own
sake” (1971, 528; 1999, 468, emphases added). This implies that the praiseworthiness of Rawls’s
hybrid virtues is of intrinsic final value, risking the introduction of perfectionism.
In Political Liberalism, Rawls seems to distance himself from this picture, arguing instead
the political virtues are of indispensable instrumental value, being “identified and justified by the
need for certain qualities of character in the citizens of a just and stable constitutional regime”
(1993, 195 fn. 29).
140
Here he allows that these very same or similar virtues might “also be
nonpolitical virtues insofar as they are valued for other reasons within various other
comprehensive view” (1993, 195 fn. 29, emphasis added), but their final value within anyone’s
private perfectionist system is purely coincidental.
In Justice as Fairness, Rawls similarly recognizes that, “a conception of moral desert as
moral worth of character and actions cannot be incorporated into a political conception of justice
in view of the fact of reasonable pluralism. Having conflicting conceptions of the good, citizens
cannot agree on a comprehensive doctrine to specify an idea of moral desert for political purposes”
(2001, 73). However, he also says that “just institutions and the political virtues would serve no
purpose—would have no point—unless those institutions and virtues not only permitted but also
sustained conceptions of the good (associated with comprehensive doctrines) that citizens can
affirm as worthy of their full allegiance” (2001, 140–141). So, he argues that, in specifying the
ideal of a good citizen of a democratic regime, the political virtues provide “a (partial) conception
of moral worth, [this ideal] is consistent with the priority of the right in both its meaning and can

140
See Costa 2004, 151–152 and 165; Hartley & Watson 2014, 422.

87
be incorporated into a political conception of justice” (142). Furthermore, he clarifies that any final
ends, which all citizens have in common, are political, not comprehensive, including the “one basic
political end” of “supporting just institutions and giving one another justice accordingly” together
with “the other ends [citizens] must also share and realize through their political cooperation”
(199).
141
But, of course, even these lofty ends may not be regarded as final by all who share them
as ends. For example, they might be understood and accepted as means to one’s chief end of
glorifying God, per the Westminster Shorter Catechism.
Nevertheless, synthesizing Rawls’s picture, we can see that the value of his hybrid civic
virtues is purely instrumental, deriving from their support of just institutions and the giving of
justice to one another, which help to provide for a just and stable constitutional regime. These
higher, perhaps final, ends are ecumenical and uncontroversial enough to be acceptable, even
motivating, to agents with a wide variety of (sometimes conflicting) conceptions of the good. So,
to the extent that virtues like reasonableness and civility serve these ends, they are of indispensable
instrumental value, and to the extent that some citizens act reasonably or with civility, based on
their recognition of this value, their actions are praiseworthy, morally good, and we may even
regard their virtue as an excellence, from a purely political, anti-perfectionist point of view.
But if others act merely in conformity with these same virtues and do so reliably—perhaps
through the development of what Rawls would call a natural asset, rather than a virtue—it remains
unclear why that cannot suffice. Why must Success-Oriented Dispositions to behave in these ways
be manifested by Insight-Dependent Attempts, rather than any other kind of manifestation?
Rawls’s psychological requirement, that virtuous agents be motivated by a higher-order regulative

141
See Dagger 2013, 302.

88
desire to adhere to moral principle, seems unnecessary. Thus, Rawls’s virtue theory has a problem
of praiseworthiness.

3.2 Audi on the Moral Content of Self-Consciously Virtuous Action
Audi provides a convincing argument that the normativity of civic virtue is at least partly moral:
“it requires respect for fellow citizens. [And] its possession is clearly incompatible with gross
immorality in the social arena—pillaging, fraudulent voting, cheating on taxes, and the like. If it
can be grounded in non-moral elements, they must be of a kind that can account for these basic
points” (Audi 2000, 150). Audi also rightly points out that “The more self-consciously virtuous an
agent is, the greater the moral content of the appropriate wants and beliefs […] deeds must be
appropriately aimed, in terms of what the agent wants and believes, or they are not moral—in the
sense of morally performed—but merely consistent with morality” (149). However, he goes on to
explain the function of the responsibilist element in his theory: “If I vote merely to avoid the
criticism of peers, then even if the vote is informed, I am not exhibiting civic virtue in my voting;
if my resulting vote is civically justified, that is quite coincidental so far as my motivation is
concerned” (149, emphasis added). So, the role of Audi’s responsibilist requirement is to ensure
that there is a non-accidental connection between the (indispensable, instrumental, anti-
perfectionist) value of the civic virtues and an agent’s motivation for acting in conformity with
them. Audi then argues that instrumentalists or pragmatists, utilitarians, Kantians, Aristotelians,
communitarians, theological ethicists, and intuitionists can all converge on some version of this
hybrid virtue-theoretic picture.
But a version of the problem of praiseworthiness persists. Audi recognizes that it is not
enough for an agent to behave in conformity with civic virtues and do so reliably. For if the agent’s

89
reliability is merely a cosmic accident, it cannot properly ground trust, and if others have reason
to doubt its trustworthiness, it cannot assure them that the agent’s virtuous behavior will continue.
This undermines the (instrumental) value of its contribution to a just and stable social order. But
there is no reason to insist that the non-accidentality of virtue be ensured by a psychological
motivation to act self-consciously from virtue in a morally praiseworthy manner.

3.3 The Robustness of Reliabilist Virtues
Consider the distinction between getting things right by accident and being so disposed as to ensure
that it is no accident you get things right. It is implicit in the thinking of reliabilists that the virtuous
cannot just happen to get things right. Virtuous agents are not those who just luck out; they have
some kind of competence. This competence displays what Philip Pettit (2014) calls robustness—
there is something in the world that ensures it obtains—it is not fragile. Pettit begins with the idea
that there are rich goods of attachment (e.g., friendship), virtue (e.g., honesty), and respect. Rich
goods have a common structure: a rich good provides others with a corresponding thin good
robustly. For example, the rich good of friendship provides a friend with the thin benefit of favor
robustly, the rich good of honesty provides one’s interlocutors the thin benefit of truth-telling
robustly, and the rich good of respect provides others the thin benefit of restraint or non-
interference robustly. To provide a thin benefit robustly is to be disposed to provide it in a way
that is “resilient enough to survive situational shifts” (24), assuring its provision both actually, “as
things actually are,” and counterfactually, “as they would be under certain variations” (46).
142

142
See Hurley 2016.

90
Liberal civic virtues should be understood as rich goods in Pettit’s sense, but the robustness
of the agent’s manifestations of virtue need not be explained by responsibilist or hybrid theoretic
mechanisms. Insight-Dependent Attempts, seeking what is (sufficiently) good, in terms of some
norm or value, based on some understanding of that norm or value, are neither necessary nor
sufficient for guaranteeing the robustness of a virtue. By contrast, a Success-Oriented
Disposition—a stable, if imperfect, disposition to perform in ways that are (sufficiently) good, in
terms of some norm or value—is exactly the kind of feature in the world that guarantees the robust
provision of a thin good by an agent who possesses it. What Bernard Williams calls an “attitude,”
what Rawls might call a “natural asset,” and what reliabilists call a “virtue” are all equal to this
explanatory task, so any further resort to Insight-Dependent Attempts must be justified on some
other explanatory basis. Absent this additional justification, to avoid the problem of
praiseworthiness, liberal anti-perfectionists should adopt virtue reliabilism. A virtue reliabilist can
say, for example, that to be reasonable is just to be robustly and reliably disposed to cooperate with
others on fair terms in all matters of civic life, regardless of how much one understands the value
of cooperating fairly or why one is motivated to do so.
143
Of course, individuals are free to
supplement this thin conception of reasonableness, or any other liberal civic virtue, with thicker
perfectionist virtues, which they personally find compelling, but they should avoid building thicker
conceptions of virtue into their theory of liberalism.

143
Notice, for example, that virtue reliabilism for liberal anti-perfectionists does not require agents to accept virtue
reliabilism or liberal anti-perfectionism. This seems like an especially promising way of developing George Klosko’s
(2000, 20–22) thin behavior-focused (as opposed to a more cognitive) conception of reasonableness, Erin Kelly &
Lionel McPherson’s (2001) notion of political (as opposed to philosophical) reasonableness, or Nussbaum’s (2011)
notion of ethical (as opposed to theoretical) reasonableness; cf. Liveriero 2020, 907–908.

91
4. Virtue Theory and the Metaethics of Civic Duty
James Boettcher (2012) argues that virtue-theoretic accounts of the norms of public reason are
insufficiently authoritative to capture the putative stringency of these “duties” of civility. His
argument proceeds in two parts. First, he attempts to undermine less stringent accounts of duty,
which seem friendlier to virtue theory, then he attacks virtue theory directly. I address each part of
his argument in turn, arguing that both moves fail.

4.1 Boettcher on Less Stringent Accounts of Civic Duties
Boettcher begins by considering the possibility that civic duties might be imperfect or prima facie
duties, having merely pro tanto stringency. Several commentators have proposed broadly optional,
supererogatory, or defeasible interpretations of Rawls’s norms of public discourse. On such views,
these norms give agents some reason, but not conclusive reason, to adhere to them; perhaps their
observance is praiseworthy, but not obligatory; perhaps their non-observance is regrettable, but
not blameworthy; perhaps they provide merely pro tanto or defeasible reasons for action, which
can be defeated, overridden, or outweighed by countervailing considerations.
Bruce Brower (1994) explores an account of public reasoning as a “goal” rather than a
“deontological constraint” (22). Philip Quinn (1997) suggests that a “supererogatory” ideal of good
citizenship sometimes counsels against introducing religious arguments into political debate, but
“imposes no moral requirements” (160). Audi likewise says that civic duty “calls for a measure of
supererogation, in the sense of good conduct not morally or (properly) legally required” (2000,

92
155).
144
And Paul Weithman (2002) thinks of the Rawlsian conception of citizenship as an
attractive ideal, which “might promote various civic goods and provide a standard for assessing
and improving political discourse and activity,” but it is “not a specification that imposes a duty”
(210).
Boettcher raises two general objections to these sorts of accounts: one weak, one strong.
His weaker objection is that, if adherence to the norms of public reason is supposed to express a
norm of equal respect for persons, they are “best understood as giving rise to duties of mutual
respect,” but failures to show due respect to others are “moral wrongs, the avoidance of which is
obligatory (and not merely good)” (160). Yet, if these civic duties are merely instrumental to
observing some stricter moral end, like showing equal respect for persons, they may be plausibly
outweighed by other considerations of equal respect for persons. Indeed, Rawls seems to think that
the norms of public reason apply with less stringency in the lives of private individuals than in the
public political forum, in part, because respecting persons also requires affording them a private
sphere in which to live out their conceptions of the good with less restraint than is required in
public. Those who voluntarily step into the public political forum, however, submit themselves to
its stricter standards of restraint. Some, including Liveriero and myself, think Rawls’s norms of
public reason remain too strict even there.
145

144
Audi says that it is not that “there is never a moral right not to do one’s civic duty, say to vote on a selfish basis—
though a case can be made that there is at most a right not to be coerced to do the virtuous thing, as opposed to a
positive right to vote selfishly. As I have stressed, virtue requires that we do things not strictly demanded by the narrow
morality of rights, for instance by people’s rights against us” (161, original emphasis).
145
See Liveriero 2020, 902.

93
Boettcher’s stronger objection leverages a worry about liberal legitimacy. He argues that
public reason is supposed to facilitate legitimate law and policy-making, but “Liberal legitimacy
would be frustrated if citizens and officials were insufficiently motivated to satisfy what they see
as merely optional requirements of political justification. And they are likely to be more motivated
to the extent that failure to honor public reason results in moral criticism” (2012, 159).
146
However,
liberal legitimacy need not be frustrated, provided that the aggregate rate of non-observance
remains sufficiently low. And liberal societies have mechanisms for praising observance of
imperfect duties, like publicly recognizing exemplary volunteers who serve their communities, as
well as mechanisms for criticizing non-compliance and encouraging greater collective compliance
with more stringent prima facie civic duties, like getting vaccinated during a public health crisis.
Furthermore, notice that the social utility of praising (or blaming) such conduct is completely
independent of the individual agent’s reasons for acting. So even if praise and blame are to play a
role in one’s theory of liberal civic duty, it need not reintroduce praiseworthiness or responsibilism
into one’s account of civic virtue itself. Thus, I think Boettcher fails to show that civic duties must
be understood as strict, indefeasible, or absolute constraints.

4.2 Boettcher on the Metaethics of Virtue-Theoretic Accounts of Civic Duty
Boettcher presses three additional objections against virtue-theoretic accounts of civic duty,
directly. Firstly, he argues that a “Political justification and restraint are not exactly qualities of
character—that is, habituated capacities or characteristics […even a strong and steady

146
McTernan presses a parallel objection against the notion that liberal virtue admits of degrees: “if liberal virtue
comes in degrees, it may not be capable of fulfilling its instrumental roles” (2014, 88).

94
commitment to these practices is only] a quality of character only in the more limited sense that
the citizens who hold such a commitment are thereby regularly disposed to honor public reason”
(161). Obviously, the practices are not themselves qualities of character, but liberal virtue theorists
understand them to be manifestations of civic virtues. Moreover, virtue reliabilists need not accept
that virtues are habituated character traits in the psychologically rich sense assumed by Boettcher;
they might simply be robust, reliable dispositions to act in certain ways.
Secondly, Boettcher argues that virtue theories “leave open the question of why a person
should aspire to be a good citizen in the first place” (161). This charge seems wholly unwarranted.
Rawls, Costa, Hartley & Watson, and Liveriero all provide direct answers to this question, with
broadly instrumental accounts of how good citizens contribute to the legitimacy and stability of a
just liberal society. Audi appeals to the intrinsic final value of a flourishing liberal society,
147
and
Boettcher rightly points out that “such an account runs the risk of being inconsistent with the fact
of reasonable pluralism” (161). But, of course, anti-perfectionist virtue theorists can abjure from
building a (comprehensive) account of flourishing into their liberal political theory, leaving space
for a thousand different flowers to bloom. There will not and need not be an overlapping consensus
on the perfectionist idea that citizens must flourish or be (made) virtuous—full stop, but why could
there not be an overlapping consensus, for example, on the idea that (enough) citizens need to be
(made) robustly and reliably reasonable in civic life?
Finally, citing the example of Rawls, Boettcher says the most plausible interpretations of
civic virtue subordinate it to deontological notions, seeing “the virtues as supporting a citizen’s

147
Though perhaps by this Audi intends nothing more than what Rawls means by our shared political final ends; see
Rawls 2001, 199.

95
efforts to satisfy valid requirements of public reason. Virtuous qualities of character can be
cultivated in order to promote cooperation and motivate citizens to satisfy their civic duties and
obligations, including the idea of public reason” (161). Rawls derives aretaic notions from deontic
ones, individuating the civic virtues by the deontic principles that govern a virtue’s domain.
148

Eamonn Callan (1999) develops a similarly deontological picture of liberal civic virtues. But Audi
adopts the opposite strategy, deriving deontic notions from aretaic ones; for example, he defines
civic duty in terms of what is essential for civic virtue.
149
Mark LeBar (2009) adopts this kind of
strategy too. A wholly anti-perfectionist virtue theory need not, and probably should not, take a
stand on whether the deontic or the aretaic takes priority, leaving this question to be decided by
each individual’s non-public perfectionist normative commitments. But both options seem viable
prima facie, so the confidence with which Boettcher dismisses the latter seems unjustified.
Regardless of which takes priority, so long as civic virtues play an indispensable explanatory role
in one’s liberal theory, the deeper question of metaethical priority, while interesting, remains
irrelevant to anti-perfectionist virtue theories.
5. The Situationist Challenge, Social Norms, and Virtue Externalism
Emily McTernan (2014) attacks the indispensable explanatory role of virtues in liberal theory,
contending that social norms are “an empirically superior alternative to liberal virtues to fulfil the

148
As Rawls explains, “This priority of the right over the good in justice as fairness turns out to be a central feature
of the conception. It imposes certain criteria on the design of the basic structure as a whole; these arrangements must
not tend to generate propensities and attitudes contrary to the two principles of justice (that is, to certain principles
which are given from the first a definite content) and they must ensure that just institutions are stable. Thus certain
initial bounds are placed upon what is good and what forms of character are morally worthy, and so upon what kinds
of persons men should be” (1971, 32; 1999, 28).
149
Audi 2000, 161.

96
instrumental roles within liberalism of securing stable patterns of behaviour from citizens” (94–
95). However, I argue that this proposed alternative is just more grist for the reliabilist’s mill. For
while, it poses a challenge for internalist accounts of virtue, it poses no threat to a certain kind of
externalism about virtue that is consistent with reliabilism.

5.1 The Updated Situationist Challenge to Liberal Virtue Theory
McTernan begins by providing a non-perfectionist characterization of liberal virtue theory’s ends:
“virtue is given instrumental value by many liberal political philosophers, as a means to secure
some desired pattern of behaviour from citizens where institutions alone would not suffice” (85).
To her credit, she also provides a reliabilist characterization of liberal virtue theory’s means:
“liberal virtues are understood as dispositions towards stable patterns of behaviour that are
consistent across differing situations” (87). But as McTernan explains, interactionism in social
psychology (having replaced the older situationist paradigm) points to a fundamental attribution
error in our reasoning: “This error is the tendency ‘to infer wrongly that actions are due to
distinctive character traits rather than to aspects of the situation’. Thus, the revised lesson of
situationism is that we are likely to be wrong regarding the influences on our behaviour, and so
situationism should teach us to be wary of assuming that character traits are what guide behaviour”
(92; embedded quotation from Harman 2000, 223). Notice, however, that the stable dispositions
cited in McTernan’s definition of liberal virtue need not be realized by robust character traits
inhering in the person, as her framing of the updated situationist challenge suggests. Nevertheless,
in light of virtue’s apparent shortcomings, McTernan argues that social norms provide an
empirically superior alternative, as research in social psychology and economics indicates they
“are powerful determinants of behaviour, and secure stable patterns of behaviour from the majority

97
of those who internalise the norm” (95). McTernan remains neutral on the question of the correct
explanation for the causal efficacy of social norms, relying only on Elizabeth Anderson’s (2000)
uncontroversial definition of a social norm as “a standard of behavior shared by a social group,
commonly understood by its members as authoritative or obligatory for them” (170).

5.2 Revising Virtue (Internalism) in Vain
In the face of the situationist challenge, some virtue ethicists have retreated to the position that
virtue is not attainable for most or even many people. McTernan insists rightly that a retreat into
such virtue elitism is not available to liberal virtue theorists, because the civic virtues are supposed
to be attainable for, at least, a majority of citizens. On the other hand, she is perhaps too pessimistic
about the instrumental value of a more attainable notion of liberal virtue that comes in degrees.
150

Nevertheless, McTernan observes correctly that one cannot retreat behind Rawls’s narrowed site
of public reason, because the kind of situational variation that undermines behavioural consistency
has been observed experimentally, not only in public life, but even within Rawls’s narrowly
circumscribed public political forum: “Experiments have suggested, for instance, that judges give
more lenient sentences immediately after lunch; and that a dirty environment, complete with
rubbish on the street, promotes stereotyping and discrimination” (88–89).
151

McTernan anticipates two further responses on behalf of virtue theory suggested by
contemporary personality psychology. The first looks for help from the subjective consistency of
behavior, but as she argues, “For liberal virtue to perform its instrumental role, it must lead to

150
See footnote 41 above.
151
See Danziger, Levav, & Avnaim-Pesso 2011 and Stapel & Lindenberg 2011.

98
objectively consistent behaviour. Thus, what is required is not that an individual, from a personal
subjective perspective, can make coherent her objectively inconsistent behaviour […] but rather
that individuals are objectively [virtuous across different situations]” (93). The second looks for
aid from the objective cross-situational behavioural consistency of some personality traits, like
extroversion, but here the problem is an excess of stability throughout the subject’s life, which
leaves personality traits “too ingrained and stable for the state to easily cultivate those it needs in
citizens” (94). So, the liberal virtue theorist needs to look elsewhere for help, and I argue that help
comes from externalism about virtue.

5.3 Externalism about Liberal Civic Virtue
Internalists about a virtue think that whether an agent possesses (or displays) the relevant virtue
depends entirely on how things are internally for the agent. Externalists about a virtue think that
whether an agent possesses (or displays) the virtue depends in part on how things are in the agent’s
external environment. Thus, the internalist/externalist distinction, as I have drawn it here, is a
mutually exclusive and jointly exhaustive one. Both positions are available to responsibilists,
reliabilists, hybrid theorists, and pluralists alike, and one can take different positions on discrete
virtues. However, the reader will notice that, in pressing the problem of praiseworthiness, and in
pushing back against Boettcher and McTernan’s psychologized characterizations of dispositions,
I have objected against building unnecessary internalist features into one’s liberal virtue theory:
features that are essential to many responsibilist and hybrid theories, but dispensable for
reliabilism. McTernan’s objection provides the impetus to embrace externalism about liberal civic
virtue, but far from being a “concession,” this view is already implicit throughout the anti-
perfectionist literature.

99
In puzzling over how “objective” virtue must be, Audi says that wildly mistaken beliefs
about issues of the day would leave the agent, at best, unlikely to qualify as civically virtuous,
implying externalism about epistemic or intellectual virtues.
152
Trivially, at least, Liveriero is an
externalist about reasonableness, insofar as the reasonable agent “needs at least a second person
with whom to establish a relation of mutual reciprocity” (904). But Costa, summarizing Rawls’s
thoughts in Justice as Fairness, provides one of the clearest statements of the idea that liberal civic
virtues depend crucially on social or environmental factors external to the agent: “The political
virtues that Rawls mentions appear as the by-product of life under the institutions of a well-
ordered society, institutions whose functioning has a nondeliberate educational role. […] Once we
move from Rawls’s well-ordered society to real societies, we face conditions that generate both
civic virtues and vices, conditions in which virtuous behavior may be very costly for some citizens”
(2004, 165, emphases added).
153
Therefore, a variety of externalist positions about liberal civic
virtue are already embraced by anti-perfectionists.

5.4 Externalism and Social Norms
In the face of the situationist challenge, Mark Alfano (2013) suggests a reconceptualization of
virtue as a triadic relation between an agent, a social milieu, and an environment:

Whether you’re inclined to take the more conservative [internalist] route, according
to which social expectations [including, e.g., social norms] (and the signaling
thereof) are non-moral situational influences that induce conduct in accordance
with virtue, or the more expansive [externalist] route, according to which social
expectations [including social norms] (and the signaling thereof) partially

152
See Audi 2000, 149.
153
See Rawls 2001, 56–57, 122, 125–126, 146–148, 156; cf. Hartley & Watson 2014, 429–431.

100
constitute virtue […] virtue (or at least virtuous conduct) can be fostered by
working on any of the relata in the triadic account. For instance, one way to
cultivate someone’s generosity (or at least reliably generous conduct) is to habituate
her to giving. One way to cultivate someone’s honesty (or at least reliably honest
conduct) is to label her honest. One way to cultivate someone’s creativity (or at
least reliably creative conduct) is to see to it that she’s in a good mood. On the
conservative view, virtue inheres in the person, and is artificially supported by the
labeling or the mood elevator. On the expansive view, virtue inheres in the
interstices between the person, the labeler, and the mood elevator—in the relation,
that is, among the agent, the social milieu, and the environment (184–185).

The externalism of Rawls, Costa, and perhaps Hartley & Watson, is friendliest to recognizing the
place of social milieu, in addition to the agent, within Alfano’s proposed triadic relation of virtue.
Audi’s epistemic externalism, meanwhile, seems friendliest to recognizing the causal influence of
more abstract environmental factors, like epistemically (un)cooperative features of one’s
environment, which suggest analogies with affectively (un)cooperative features of one’s
environment, like trash in the streets or mood elevators. So, Alfano’s expansive externalism seems
like a promising route to developing an account of liberal civic virtue that takes the situationist
challenge seriously. On this kind of view, shaping the virtuous agent and shaping the virtuous
social milieu through social norms go hand in hand. As McTernan explains, “When people fail to
live up to a social norm they have internalised, they may feel shame: this is one reason for the
effectiveness of social norms in securing stable patterns of behaviour” (2014, 103). Indeed, once
internalized, a social norm becomes a reliable disposition that can easily be categorized as a
reliabilist externalist virtue. Whereas the virtue reliabilist can say that to be reasonable is to be
robustly and reliably disposed to cooperating with others on fair terms in all matters of civic life,
the addition of externalism allows them to index this to a particular social milieu or environment,

101
for example, requiring the reasonable agent to be sensitive to certain social, historical, political, or
economic facts about their specific society.
154

In defense of the more conservative internalist picture, McTernan disparages thin
conceptions of virtue, “devoid of any account of human flourishing” and “detached from virtue’s
Aristotelian roots,” as “concessions” (101), but these remarks present no serious worries for anti-
perfectionists, reliabilists, or externalists.
155
More substantively, she argues that even if it is
“possible to characterise liberal virtues in such a way as to make them sufficiently akin to social
norms to share in the empirical support of the causal efficacy of social norms […] there is still
good reason to use the term ‘social norm’ rather than ‘liberal virtue’: doing so avoids redundant
normative debates” (99). Here the idea is that ‘virtue’ talk obscures the fact that different means,
to securing reliable objectively stable patterns of behaviour in citizens, have different normative
costs, because the problems besetting virtue-theoretic approaches are different from those
besetting social-norm-theoretic approaches. But this objection seems to just beg the question
against the possibility that virtue theories can respond to the problems besetting social norm
theories. Consider McTernan’s examples. If worries about illiberal “soul engineering” necessarily
apply to all virtue theories, including those that take a deliberately light touch with the “soul,” then

154
Notice, however, that the addition of externalism does not require agents to accept externalism about virtues, or
epistemological externalism, etc.; see footnote 38 above.
155
Nor would we expect such worries to move Plato, Confucius, the sub-Saharan Afro-communitarian virtue ethicists
(see Metz 2012), or the Nahua (Aztec) virtue ethicists of the Florentine Codex, who developed a non-trivially
externalist theory of virtue in an uncooperative “slippery” world, in which agents are prone to slip ups, and for whom
the end of virtue is “rootedness,” not only in one’s body and psyche, but in one’s community and in the world (see
Purcell 2017, 12–13 and 16–17). Moreover, as Thaddeus Metz argues, Aristotle himself is plausibly read as an
externalist about certain virtues, including courage and justice (see 2012, 108–109).

102
what exactly is the difference between this worry, as addressed to anti-perfectionist reliabilist
externalism, and the worry “that social norms are an example of the pernicious effects of social
pressure on freedom” (102)? McTernan does not say. She also claims that worries about the
illiberality of shaming citizens are faced uniquely by social-norm-theoretic approaches, but shame
and shaming figure in prominent liberal and neo-republican virtue theories as well.
156
Thus, I think
McTernan fails to show that no version of externalism about virtue can benefit from empirical
insights into the causal efficacy of social norms.
6. Conclusion: Virtue Reliabilism for Liberal Anti-Perfectionists
Liberal thinkers have taken a variety of approaches to theorizing civic virtue. Liberal anti-
perfectionists have tended to adopt hybrid virtue theories that require agents to be praiseworthy in
acting in conformity with virtue, but this normative requirement of praiseworthiness has not been
properly justified. It, therefore, burdens these theories with needless normative baggage. Virtue
reliabilism avoids this problem by dispensing with the requirement of praiseworthiness and
remaining metaphysically, psychologically, and normatively thinner than its rivals. In so doing, it
remains better positioned to count as anti-perfectionist. This thinner account of liberal civic virtue
has also proven able to respond to recent challenges from Boettcher and McTernan. Contrary to
Boettcher, reliabilist virtue-theoretic accounts of civic duty can be as stringent as the liberal
theorist needs them to be. Contrary to McTernan, reliabilist virtue theories that embrace a certain

156
See McTernan 2014, 103 and Nussbaum 2006. For liberal discussions of shame, see Rawls 1971, 442–446; 1999,
388–391; Dagger 1997, 29 and 71; and Berkowitz 1999, 101–103. For a neo-republican account that sounds especially
like what McTernan seems to have in mind, see Pettit 1997, 225–227 and 253–257; cf. Costa 2004, 166. For his part,
Audi says that “backing by the appropriateness of such action as strong moral criticism is not a defining property of a
moral virtue” (2000, 162).

103
kind of externalism about virtue can benefit from empirical insights into the causal efficacy of
social norms. Although most liberal anti-perfectionist virtue theorists have already implicitly
accepted externalism, no one has previously advanced an explicitly reliabilist account of anti-
perfectionist liberal virtue. Yet, as we have seen, reliabilism fares better than its rivals in
responding to these challenges, and it does so without straying into perfectionism.

104
IV. Epistemological Externalism for Liberal Anti-Perfectionists
According to Martha Nussbaum (2011), liberal perfectionists, like Isaiah Berlin, Joseph Raz, and
Steven Wall, espouse “a type of liberal political view that spells out a set of controversial
metaphysical and ethical doctrines concerning the nature of value and the good life, and then goes
on to recommend political principles built upon these values” (3). In contrast, liberal anti-
perfectionists, like Gerald Gaus, Charles Larmore, Jonathan Quong, John Rawls, Kevin Vallier,
and Nussbaum herself “hold that the state should be neutral among rival understandings of the
good” (Wall 2019, §3.1). Rawls, Larmore and their followers have developed accounts of
“political liberalism,” while Gaus and his followers have defended versions of “justificatory
liberalism.” Call the union of these two families of theories “public reason liberalism.”
It is sometimes said that liberalism’s “moral lodestar” is the conviction that the exercise of
political power or authority is legitimate only when it can be publicly justified (Quong 2011, 2–
3).
157
The public reason liberal conception of legitimacy holds that (prospective or continued)
exercises of power or authority must be justified to those subject to them, those in whose name
they are exercised, or both—roughly, to citizens. Political liberals argue that these justifications
must be made (available) in terms of reasons that are shared qua reasons by (some normatively
significant, e.g., “reasonable,” subset of) this constituency. Justificatory liberals argue these
justifications must be in terms of reasons that support the same exercises of power or authority,
whether or not they are shared qua reasons by all. Such reasons or justifications are suitably
“public.” This standard of publicity is taken to be importantly different from a standard of truth,

157
See Macedo 1990, 78; Waldron 1993, 36–37.

105
which would require all and only true justifications, or a standard of consent, which would require
all and only justifications to which everyone actually consents.
Many have understood these notions of public reasons and the reasonableness of their
constituency, to be at least partly epistemological, having epistemically normative implications.
However, several public reason theorists have argued that anti-perfectionists should reject
epistemological readings of these notions. George Klosko (2000) argues that they make it harder
for agents to count as reasonable without any empirical evidence that epistemological requirements
actually ensure more reasonable behavior. Erin Kelly & Lionel McPherson (2001) similarly argue
that they pre-emptively restrict the constituency of the reasonable too narrowly and intolerantly.
And Nussbaum argues that they promote specific, controversial understandings of what is
epistemically valuable, without proper justification, thereby falling into a kind of perfectionism.
158

Nevertheless, many public reason theorists have assumed or defended some epistemic dimension
of reasonableness, public reasons, or public justification.
159

As I explain in section 1 of this chapter, anti-perfectionists overwhelmingly assume or
defend epistemological internalism about these notions. However, in section 2, I survey the
externalist implications of some public reason liberal claims. In particular, I look at Wall (2010,
2014), who, unperturbed by the objections of Klosko, et. Al., embraces liberal perfectionism and
defends epistemological externalism about reasonableness and public justifications, prioritizing

158
Nussbaum 2011, 33. David Enoch (2017) argues instead that liberal anti-perfectionists face a trilemma:
epistemological readings of reasonableness either (1) cannot explain every reasonable individual’s epistemically-laden
commitments to their non-public doctrines, (2) undermine reasonable individuals’ motivations for restraining
themselves in conformity with public reason’s publicity constraint, or (3) fall into perfectionism. See Chapter 2 for an
extended response to Enoch.
159
See, e.g., Ancell 2019, Audi 2000, Barry 1995, Estlund 2008, Fuerstein 2013, Gaus 1996 and 2011, Leece 2003,
Leland & van Wietmarschen 2012, Liveriero 2015 and 2020, Nagel 1987, Peter 2013 and 2019, Quong 2011, Rawls
1993, Vallier 2014, van Wietmarschen 2018, Wall 2010 and 2014, and Wenar 1995.

106
the need to justify political decisions on the basis of good epistemic reasons over other concerns.
And I argue that prominent anti-perfectionist objections against epistemological externalism are
not decisive. However, in section 3, I contend that there is good reason to worry that strong
externalism threatens to collapse the standard of public reason into a truth standard, leaving no
room for public reason to do independent work. Therefore, taking the advice of Rawls, I look to
the public political culture to see if there is an implicit uncontroversial standard of public reason
to which we might refer. Following Regina Rini (2020), I argue that the ordinary standard of good
faith debate entails neither truth nor consent, and because it is implicit to the public political
culture, its acceptance presupposes no excessively controversial views.
Therefore, in section 4, taking inspiration from Klosko, Kelly & McPherson, and
Nussbaum, I develop a “Cooperation First” account of reasonableness, on which reasonableness
is first and foremost a matter of being a reliably fair cooperative participant in all matters of civic
life. On this view, successfully cooperative behavior determines who is reasonable. But I argue
that in order to successfully cooperate with other members of one’s society, reasonable agents
must be sensitive to facts about how the burdens and benefits of cooperation are distributed in
one’s society, including social, historical, political, or economic facts about their society. Thus,
reasonableness is an epistemologically externalist virtue, but I argue that this does not collapse the
standard of public reason into a truth standard, because in matters of public inquiry, where
reasonableness is manifested alongside the virtue of civility, the virtuous practice of public
reasoning is governed by the standards of good faith debate, which requires neither truth nor
consent. Thus, I develop a Cooperation First account of reasonableness that is nonetheless
epistemologically rigorous. Before concluding Chapter IV in section 7, I put my view to the test
in a series of thought experiments, demonstrating in sections 5 and 6 that, whereas externalist

107
excellences can ensure successful cooperation in public reasoning, internalist excellences are too
weak.
1. The Prevailing Epistemological Internalism about Reasonableness
A number of liberal anti-perfectionists object to building any epistemic dimension into the
normative requirements of reasonableness. George Klosko (2000) advances a thin conception of
reasonableness, which “centres more on how people behave than on what they believe” (20,
emphasis added), arguing that more “cognitive” conceptions of reasonableness make it harder for
agents to count as reasonable without any empirical justification for the assumption that cognitive
requirements actually ensure more paradigmatically reasonable behavior. Erin Kelly & Lionel
McPherson (2001) similarly distinguish between a more epistemically demanding “philosophical”
sense of reasonableness and a thinner behavior-focused “political” sense of reasonableness,
arguing that the philosophically unreasonable should be tolerated “as long as their philosophical
unreasonableness does not spill over into political unreasonableness” (40). And Martha Nussbaum
(2011) tentatively endorses a similar distinction between “theoretical” and “ethical” senses of
reasonableness, arguing that the more theoretical promote specific, controversial understandings
of what is epistemically valuable, without proper justification, thereby falling into a kind of
perfectionism.
160
Many public reason liberals, however, assume or defend some epistemic
dimension of reasonableness and most assume epistemological internalism about it.

160
See Nussbaum 2011, 29–31 n. 50

108
1.1 Internalism about Epistemic Rationality
To help clarify the thesis of epistemological internalism, consider its least controversial instance.
According to epistemological internalists about epistemic rationality, there is no environmental or
external world condition, like sensitivity, safety, or truth, on the rationality of doxastic attitudes or
processes. So epistemological internalism about the rationality of doxastic attitudes (e.g., belief or
credence) or processes (e.g., reasoning or updating) is the view that a subject S’s doxastic attitude
or process is rational purely in virtue of states internal to S. Therefore, if we take two individuals
A and B who are perfect internal duplicates, whenever A is epistemically rational, B is too, even if
B is a brain-in-a-vat being systematically misled inside Robert Nozick’s experience machine. With
few exceptions,
161
internalism about epistemic rationality is the overwhelming consensus view.
As Stewart Cohen and Juan Comesaña say, “[t]he traditional view in epistemology is that we must
distinguish between being rational and being right.”
162
Epistemological externalism is defined as
the negation of internalism, so externalism about rationality is compatible with rationality being
partly a matter of facts internal to a subject, but it denies that rationality depends solely on these
internal facts. Internalism about epistemic rationality is a philosophically respectable position, and
henceforth I will simply assume it is correct.
163
But what of internalism about reasonableness?

161
Especially, Williamson 2013 and 2017, and Silva 2018.
162
Cohen and Comesaña (forthcoming).
163
Some principal considerations in favor of internalism about rational belief include deontological arguments
concerning the ethics of belief. For it’s thought that whether one has fulfilled one’s intellectual duties is entirely an
internal matter, so if epistemic rationality is purely a matter of fulfilling one’s intellectual duties, epistemological
internalism about rationality follows. Another important source of motivation for internalism are intuitive judgments
that (i) it seems irrational to believe based on a method that reliably tracks your environment in the absence of internal
evidence (e.g., Norman the Clairvoyant; see Bonjour 1980), while (ii) it seems rational to believe based on internal
evidence, even if this evidence is systematically misleading about your environment (e.g., the New Evil Demon; see
Lehrer and Cohen 1983); see Poston, http://www.iep.utm.edu/int-ext/.

109
1.2 Internalism about the Epistemic Dimension of Reasonableness and Public Reasons
One general consideration in favour of epistemological internalism about reasonableness is that
treating reasonableness as a purely internal matter (for example, of having a more or less
epistemically rational doxastic system) makes it easier to account for the fact of reasonable
pluralism. For in that case, no feature of any party’s environment could disqualify them from being
reasonable, making it easier to suppose that reasonable parties to doctrinal disagreements abound.
This consideration does not seem to figure explicitly in liberal thinking, but several of the most
influential discussions of the epistemology of reasonableness identify only internalist requirements
on reasonable subjects. In this section, I survey some representative examples.
1.2.1 Leif Wenar’s Epistemological Internalism about Reasonableness
Following John Rawls, while streamlining and omitting aspects of his proposal, Leif Wenar (1995)
offers two requirements on the reasonable-ness of persons, on which they:
1. possess two moral powers–the capacities for a sense of justice and for a conception of the
good; (b) possess the intellectual powers of judgment, thought, and inference; (c) have a
determinate conception of the good interpreted in the light of some [non-public
doctrine(s)]; and (d) are able to be normal, fully cooperating members of society over a
complete life; and,
2. are ready to propose and willingly abide by principles and standards that are fair terms of
cooperation, given assurance others will likewise do so (37).
164

All of these conditions can be fulfilled by a brain-in-a-vat.
Condition 1b is the only requirement of much epistemological significance, and the possession of
these meager intellectual powers of judgment, thought, and inference are purely matters of what is

164
For condition 1, see Rawls 1993, 81; 15–35; for condition 2, see 48–54.

110
going on inside someone’s head. So Wenar espouses an epistemologically internalist account of
what it takes for the individual to be reasonable.
1.2.2 Gerald Gaus’s Epistemological Internalism about Public Reasons
In Justificatory Liberalism (1996), Gerald Gaus articulates a version of epistemological
accessibilism, about what it is for an agent to have a good (public) justification or reason.
Accessibilism starts with an agent’s mental states at a time and looks at how they could evolve
(setting aside deviant causal chains
165
). For an agent A to have a reason or justification is for A to
have a belief that figures in A’s deliberation, or for there to be some further conclusion that could
be reached or accessed by a sound deliberative route from A’s existing set of commitments.
Because A’s epistemic justification at a time t depends, in part, on further conclusions A
could reach in the right way from A’s existing belief system at t, but which are not part of A’s
belief system at t, Gaus calls it open justification. Although he classifies it as a weakly externalist
view, Gaus uses ‘externalist’ in a different sense from the way epistemological externalism is
typically defined. For Gaus, it is externalist just insofar as it evaluates what reasons or justifications
an agent has at t from an external perspective, taking accessible, but unaccessed, external reasons
at t into account, even though they are not part of the agent’s internal perspective or evidence at t.
However, the core idea of open justification is that it “takes a person’s current system of beliefs
and asks, first, whether given this system that person is committed to accepting some new piece of
information, and, second, whether that person is then committed to revising his or her system of
beliefs in the light of that new information” (31–32, my emphasis). As he clarifies, it is not a

165
See Gaus 1996, 30.

111
question of whether one would actually revise their belief system in light of new information, but
of whether revision would be called for, in some normative sense, by their existing commitments.
So epistemic justification or having good reason for belief is evaluated “relative to a system of
reasons and beliefs ... the point of reference is always the person’s system of beliefs and reasons,
suitably modified to take account of new information and criticism” (63).
Agent-relative accessibility does all the work of individuating an agent’s reasons at t, and
open justification is a function of these reasons, so is it really an epistemologically externalist
view? According to Gaus, ‘fact’-talk is a way of talking about the world, while ‘reason’-talk is a
way of talking about cognitive systems, and “in particular the connections that are appropriately
made within a cognitive system. External reasons are perhaps best understood as reasons that the
person does not now possess, but that her system of beliefs and epistemic norms commits her to
accepting” (34).
166
Better still, we might say that external reasons are cognitive moves to which A
is committed to making at t, but which A has so far failed to make at t. External reasons, in Gaus’s
sense, therefore, only count as reasons for an agent in virtue of that agent’s mental states, and not
truly in virtue of facts about the external world. Gaus’s account, therefore, brackets the external
world, making open justification an internalist account of the epistemic component of
reasonableness.

166
He continues: “This does not involve positing a unique sort of reason; it still understands reasons as cognitive
states.”

112
1.2.3 Kevin Vallier’s Epistemological Internalism about Public Reasons and Reasonableness
Kevin Vallier (2014), following Gaus, expresses the internalist conviction this way: “If public
reason liberals do not embrace justificatory internalism about reasons, they cannot plausibly claim
that [public] reasons are rationally recognizable, as agents may lack psychological access to the
relevant reasons. So a reason is sufficient only if it is epistemically justified in the internalist way,
and override/defeat reasons that contradict it” (104–105). One can be a fully reasonable agent so
long as one has completely discharged one’s epistemic duties to seek reasons, where one’s
epistemic duties are tempered by what is realistic to expect of (slightly below) average agents, and
the issue of whether one has fulfilled one’s intellectual duties is typically understood to be a matter
of how things seem from the inside to the agent.
167
So if reasonableness is purely a matter of
fulfilling one’s intellectual duties, epistemological internalism about reasonableness follows.
Vallier says that all genuine versions of public reason liberalism share something like the
following standard conception of reasonableness, on which a citizen of the well-ordered society is
reasonable because she possesses dispositions to:
(i) Engage in public justification, or to offer justifications for her own preferred principles and
abide by the justified principles proposed by others.
(ii) Recognize the “burdens of judgment” which imply reasonable pluralism.
(iii) Reject the repression of other reasonable points of view.
(iv) Reason rationally, or to largely avoid common errors in reasoning.
(v) Rely on methods of reasoning that others can share or access (146–147).
168

Insofar as these dispositions are all plausibly individuated internalistically, the standard conception
looks internalist prima facie, with (ii), (iv) and (v) all being epistemologically significant. I have

167
See footnote 11 above.
168
Vallier draws this summary of Rawls’s conception of the reasonable from Gaus 1996, 132; See Rawls 1993, 49–
52, 53–58, 60, 76, 119, 162–163, 229; cf. Vallier 2014, 178, endnote 4.

113
already explained the internalist thrust of the burdens of judgment (ii), and reasoning rationally or
avoiding common errors in reasoning (iv), like the avoidance of Rawls’s earlier enumerated vices
1𝛼𝛼–𝛿𝛿 , is also plausibly a matter of states internal to the subject. So the first four of these five
enumerated dispositions are all matters of facts internal to reasonable agents. (v) is ambiguous
between an internalist reading and an externalist one, but according to Vallier, it’s simply “part of
the standard conception of justificatory reasons, and so not essential to conceptions of
idealization”(148). So the standard conception of reasonableness seems quite thoroughly
epistemologically internalist.
In addition to what he calls the standard conception of reasonableness, Vallier distinguishes
another parameter of standard idealization for public reason liberals, which he calls the information
parameter. This information parameter is explicitly externalist, but the very fact of its divorce from
the reasonableness parameter tells just how deeply the internalist assumption about reasonableness
runs.
2. Epistemological Externalism about Reasonableness?
Strongly externalist accounts of reasons or justifications for political decisions, like those advanced
by Joseph Raz (1986) and David Enoch (2014, 2015), leave no role for the publicity of reasons
and are self-consciously incompatible with public reason liberalism. Steven Wall (2010, 2014), on
the other hand, positions himself as a public reason liberal and embraces epistemological
externalism about reasonableness and public justification, but his is a liberal perfectionist account
that is unpalatable to most other public reason liberals. Among anti-perfectionist public reason
liberals, however, there are still some indications of openness to epistemological externalism about
reasonableness and public reasons.

114
2.1 Liberal Anti-Perfectionists and Epistemological Externalism
In Justice as Fairness: A Restatement (2001), Rawls says that reasonableness and practical
rationality, while distinct, are “complementary ideas entering into the fundamental idea of society
as a fair system of social coop-eration.”
169
He is less explicit about the relationship between
reasonableness and epistemic rationality, but in Political Liberalism (1993) Rawls seems to say
that reasonableness requires that one’s beliefs respond to reasons and evidence in (more or less)
epistemically rational ways,
170
though he also suggests this requirement is “rather minimal” (60
n.13).
171
Rawls’s canonical descriptions of the fact of reasonable pluralism stipulate that
reasonable disagreement:
1. Is not a function of ( 𝛼𝛼) irrationality or logical error, ( 𝛽𝛽 ) prejudice or bias, ( 𝛾𝛾 ) self- or group-
interest,
172
( 𝛿𝛿 ) willfulness; or ( 𝜀𝜀) blindness; and,
2. Persists even where parties freely
173
exercise their faculties of reasoning under favorable
conditions.
174

These criteria are also of obvious epistemological significance. Avoidance of the vices enumerated
in 1𝛼𝛼–𝛿𝛿 is plausibly a matter of states internal to the subject. But setting aside the ableist language
of 1𝜀𝜀, it is less clear whether “peoples’ understandable tendency to view the political world from
a limited standpoint” (1993, 37) is purely an internal matter. Moreover, the freedom to exercise
one’s faculties of reasoning unimpeded (e.g., by anything in one’s environment), and the general
favorability of conditions for their exercise, as indicated in the second condition above, are most

169
Rawls 2001, 6.
170
Ancell 2019, 416; see Rawls 1993, 59 and 62; cf. Freeman 2007, 346.
171
See Wall 2010, 479.
172
Including class interest; Rawls 1993, 37.
173
See Rawls 1993, xxvi.
174
For condition 1 𝛼𝛼 , see Rawls 1993, 54; for 1 𝛽𝛽 – 𝜀𝜀, see 58; for condition 2, see 398. I am indebted to Aaron Ancell
for these observations and citations; for discussion see Ancell 2019.

115
plausibly read as externalist requirements on reasonable disagreement. But sadly, the externalist
elements of Rawls’s criteria have gone largely ignored.
Rawls’s much-discussed “burdens of judgment” suggest reasonable disagreement is
inevitable because our rational faculties face obstacles, for even under ideal conditions:
A. Relevant evidence is often conflicting and complex, and therefore hard to evaluate;
B. We may disagree about the weight different considerations are to be given;
C. Our concepts are vague, so we must rely on judgment and interpretation;
D. The way each person assesses evidence and weighs values is shaped by her particular
life experiences (her ethnicity, class, place in the division of labor, etc.);
E. It is difficult to make an overall assessment of an issue when there are normative
considerations of different force on both sides; and
F. In any society some selection of values must be made from among all that might be
realized (Wenar 1995, 41).
175

These influences on our rational faculties explain their fallibility, and the resulting inevitability of
disagreement between agents. Of course, some of these influences are external (especially D and
F, but arguably A as well). However, the point seems to be that they are nonetheless compatible
with being (non-ideally) epistemically rational on the inside. For if we cannot help but be burdened
by these fallibilities, then (in the sense of ‘ought’ that we care about—the ‘ought’ of epistemic
rationality) it is just not the case that we ought to employ our rational faculties free from these
fallibilities. Therefore, as fallible epistemic agents, we are without epistemic blame for any
disagreements inevitably resulting from these limitations. The suggestion, therefore, seems to be
that internalistic epistemic rationality is what really matters, epistemologically speaking, for being
reasonable in a disagreement. So again, the externalist dimensions of these burdens have been
largely ignored.

175
See Rawls 1993, 55–57; F is less a source of disagreement than an occasion for it.

116
Vallier’s standard conception of reasonableness also admits of externalist interpretation.
Interpreted through Gaus and Vallier’s justificatory internalism about what it is for an agent to
have a (public) reason, disposition (v) of the reasonable, to rely on methods of reasoning that others
can share or access, places an externalist constraint on reasonableness. For on their accessibilist
interpretations of justificatory reasons, public justification or public reason-giving is successful
only if the reason(s) advanced can be internally accessed by the addressee(s). So while they
epistemically assess the reasonableness of an addressee of public reasons internalistically, the
reasonableness of agents advancing or addressing public reasons to others seems to depend on
external facts about the minds of those other agents. Thus, if we were to build an epistemic virtue
around this constraint, it would require the reasonable to be reliably disposed to successfully
tracking facts about what is internally accessible to reasonable others. But this externalist
implication also seems to have been overlooked by anti-perfectionists.
Robert Audi (2000) is more unambiguously open to some version of epistemological
externalism. In puzzling over how “objective” liberal civic virtues must be, he says that wildly
mistaken beliefs about issues of the day would leave the agent, at best, unlikely to qualify as
civically virtuous,
176
suggesting an external world success condition on belief for at least some
liberal virtues. Unfortunately, Audi does not develop this aspect of his view in further detail.
Finally, Fabienne Peter (2019) distinguishes between agreement-dependent reasons
grounded in subjective reasons, which “reflect our perspective on the situation—as given by our
beliefs and/or our evidence” (152), and agreement-independent or objective “normative practical
reasons [which] are, or are given by, objective facts” (148). Whereas public reason liberals

176
See Audi 2000, 149.

117
typically argue that justification of political decisions is necessarily in terms of an agreement
grounded in subjective reasons, Peter defends the view that in some epistemic contexts,
justification of political decisions is necessarily in terms of an agreement grounded in subjective
reasons, while in in other contexts, justification of political decisions is necessarily in terms of
objective reasons. For example:
If we have sufficiently robust knowledge of what political decision is objectively correct,
I agree that political legitimacy is undermined if the wrong decision is made. To see the plausibility
of this, suppose a political regime makes decisions that are obviously morally impermissible. For
example, suppose a regime issues laws that violate essential human rights of some minorities
without any redeeming justification. Such decisions do, I want to maintain, lack legitimacy
because they disregard moral knowledge that is sufficiently robust (168–169).
In this way, Peter defends context-sensitivity about political justification, on which public
justifications may be purely pubic or agreement-dependent, purely agreement-independent, or
perhaps mixed, depending on the epistemic context.

2.2 Wall’s Perfectionist Epistemological Externalism
Steven Wall (2010) expresses sympathy for strong externalism about reasons, on which an agent
can have “a reason to do something, even if there is no sound deliberative route that can take him
from his current set of reasons and beliefs to the recognition that he has the reason in question”
(126). On this basis, Wall defends a liberal perfectionist account of ethical and epistemic
reasonableness, which Federica Liveriero (2020) helpfully summarizes this way:
Wall claims that being reasonable, both from an ethical and epistemological perspective,
implies acknowledging the intrinsic normativity attached to the concept of individual rationality

118
and being able to respect the requirement of not rejecting a reason that is actually a reason for me,
notwithstanding the fact I might not recognize it from my internal perspective. […] an agent is
epistemically unreasonable when she rejects a particular conception of justice for bad epistemic
reasons (Wall 2014, 471). Wall characterizes reasonableness as the epistemic virtue of being able
to acknowledge that there are reasons that we—as epistemic agents—cannot reasonably reject
(912, original emphasis).
This is a perfectionist definition of reasonableness because it appeals to the intrinsic value
of a specific, controversial account of normative rationality. Wall argues that anti-perfectionists
like Klosko, Kelly & McPherson, and Nussbaum pay too high a price for prioritizing the ethical
over the epistemic in their expectations of the reasonable, because in lowering the epistemic bar
for who counts as a reasonable, they increase the size of the constituency to whom public
justification is owed, allowing more parties to have a say about which exercises of power are
acceptable, and thereby making it harder for the reasonable to come to agreement on policies that
“would be effective at promoting sound ideals or values associated with leading a good life (such
as, e.g., state measures to promote appreciation of excellence in the arts and sciences or state-
funded educational measures to promote individual autonomy” (473). For a liberal perfectionist
like Wall, who is quite happy for liberalism to spell out a set of controversial metaphysical and
ethical doctrines concerning the nature of value and the good life, and then recommend political
principles built upon these values, “a model of political legitimacy that defends a strong
inclusiveness requirement for citizens’ opinions and doxastic reasons tends to obstruct the
realization of justice and let go of the search for the objectively correct conception of justice”
(Liveriero 2020, 912). Peter’s (2019) defense of context-sensitive externalism about political

119
justification seems to be motivated by similar considerations, though she is careful to avoid
perfectionist reasoning.

2.3 So Why Not Externalism?
2.3.1 Gaus Against the Bare Correspondence View of Externalism
One of the principal arguments against epistemological externalism concerning reasonableness or
public reasons comes from Gaus. In The Order of Public Reason (2011), he argues that we
distinguish external reasons from facts relevant to A’s deliberation, which A fails to believe as fact,
or which remain inaccessible to A by any deliberative route we can realistically expect A to
undertake, because “We employ the concepts of reasons and rationality to make others intelligible
to us [...] or at least understand them, as fellow agents. But when we ascribe to them reasons that
are truly inaccessible to them as agents, to insist that they have a reason φ is simply a way of saying
that φ is right, but the rightness is not something that can enter into their agency” (232–235). Gaus
thinks we must not ascribe reason(s) R to ordinary agents, when the warrant for R is beyond what
can be accessed through the deliberation we can realistically expect from average (or slightly
below average) agents.
177
For implicit to our reason-based moral practice (including public
reasoning), “is a conception of others as free and equal autonomous persons, who act on an internal
ought that they have reason to endorse” (258, emphasis added). So, as he concludes in Justificatory
Liberalism, “we must reject the idea of an inherently impersonal reason. [...] there exist states of
affairs that humans simply can never cognize. [...] it is mysterious what could be meant by saying

177
See Gaus 2011, 257.

120
that they provide reasons to believe B. “Reasons for whom?” is the proper query. Facts become
reasons when they enter into cognitive systems with inferential norms and are able to justify
acceptance of a belief” (1996, 35).
Gaus argues we should be internalists about the epistemological component of reasonable
belief, because it aims at understanding the internal perspective of agents to whom we seek to
publicly justify political power. To reject internalism, Gaus suggests, risks constructing a theory
of the epistemic component of reasonable beliefs, which justifies power to someone or something
other than those to whom justification is owed. This is a powerful consideration, but it only favors
internalism if no externalist account of reasonableness can accommodate it equally well. However,
in trying to show that externalism cannot accommodate this insight, Gaus considers an implausibly
strong version of externalism, which treats external facts of the matter concerning a proposition
believed as all that is relevant to evaluating whether the belief is reasonable. Call this the bare
correspondence view. Gaus argues the bare correspondence view is embarrassed by cases like the
following:
Betty believes that the Apollo moon landings were faked; they were, she believes,
televised from an Air Force base in California. Today we would conclude that the
reasons for her belief could not have been that they were actually faked. But
suppose that we discover in ten years that they really were faked; do we want to say
then that this fact is, after all, the reason for her belief? But why change the account
of her reasons and reasoning just because it turns out that she was correct? Our
original accounting of her evidence, her perceptions, and her reasoning may be as
sound after the discovery as it was before; the mere fact that she turns out to be
correct does not in itself seem to require that we give a new account of her reasons
for her belief that the moon landings were faked.
178

178
See Gaus 1996, 33–34.

121
This is counterintuitive. If Betty’s beliefs are the likely result of conspiracy theorizing, why should
these same beliefs suddenly become reasonable, when it is learned that her conspiratorial thinking
just happened to fix upon some conclusions which were true? Given Betty’s intellectual conduct,
she is lucky to have formed true beliefs, and intuitively, this undercuts her epistemic
reasonableness. The bare correspondence view of reasonableness, being an unusual version of
externalism, cannot accommodate our intuitive judgment about Betty’s unreasonableness. Not
only that, the bare correspondence view suggests that anyone who has ever disagreed with Betty,
believing the moon landings really happened, was unreasonable in this belief, simply because they
were incorrect.
Unsurprisingly, such a view leaves no room for reasonable disagreement whenever there
is a fact of the matter about the question under dispute. Thankfully, externalists need not embrace
the bare correspondence view of reasonable belief. Instead, they must only think that epistemic
reasonableness depends on some facts in the external world, including perhaps, facts about how
agents interact with their environment. So Gaus’s argument against externalism does not tell
against all versions of externalism, and as I will now argue, neither does his argument for
accessibilism tell in favor of internalism over externalism. One way to develop the idea of
accessibility, along externalist lines, is to say that what it is for a fact or reason to be accessible
depends in part on how an agent interacts with their environment, and so believing (and acting)
reasonably depends in part on interacting with one’s environment in the right way.
2.3.2 Liveriero Against Wall’s Externalism
Another argument against externalism comes from Liveriero, who argues that Wall’s externalism
faces a problem of paternalistic disrespect. Wall argues that respect for persons “requires us to
view them as beings who are not stuck with their commitments, but rather as beings that have the

122
capacity to assess, and if called for to revise or abandon, their commitments in response to the
reasons for having them. A view of reasonableness that includes epistemic as well as ethical
elements can take proper account of this aspect of respect for persons” (2014, 478). However,
Liveriero argues that Wall’s view leans too heavily on a moral appraisal respect for persons, which
expresses “positive consideration of the deeds, achievements and character of a person; hence it is
a posteriori, conditional on actual conduct and comes in degrees” (2020, 914). It thereby risks
leading to paternalistic expressions of disrespect when others are scrutinized and deemed
insufficiently well informed. In its place, Liveriero defends an a priori, unconditional, absolute
notion of recognition respect for others, which “treat them as ‘opaque’, without engaging in an
assessment of their personal merits and faults, but simply respecting them as on an equal footing
with us” (915). Specifically, she argues that the reasonable should “ascribe to [others] the default
status of putative epistemic and practical authorities” (916). For my part, I am unconvinced that
appeals to either notion of respect can be decisive in this debate. But Wall’s view raises another
concern.
3. Public Reason: Somewhere Between Truth and Consent
As Quong (2017) explains, the ideal of public reason implies a standard that lies somewhere
between truth and consent:
An account of public reason must find some way of giving the perspective of
individual persons a significant role, without allowing this to collapse into consent:
public reason is not simply a way of identifying those principles to which people
already consent. But equally, public reason must not define those principles that
could be justified to, or be acceptable to, each person as simply those principles that
are true. In either case, the idea of public reason would do no independent work
(§1).

123
Relatedly, warrant or Gettier-proof, knowledge-level justification also seems too strong, since it
arguably entails truth,
179
and Wall’s seems too strong for similar reasons. So how do we go
externalist without going too far? And how do we find an account of a standard that lays
somewhere between truth and consent, that is not itself too controversial, idiosyncratic, doctrinaire,
or perfectionist to count as suitably public? Rawls has a suggestion:
look to the public political culture of a democratic society, and to the traditions of
interpretation of its constitution and basic laws, for certain familiar ideas that can
be worked up into a conception of political justice. It is assumed that citizens in a
democratic society have at least an implicit understanding of these ideas as shown
in everyday political discussion, in debates about the meaning and ground of
constitutional rights and liberties, and the like. [...] Even though such ideas are not
often expressly formulated, nor their meanings clearly marked out, they may play
a fundamental role in society's political thought and in how its institutions are
interpreted” (2001, 5–6).
So, ideally, we want a standard whose “acceptance presupposes no particular comprehensive view,
and its fundamental ideas are familiar and drawn from the public political culture” (Rawls 2001,
33), and popular notions of dealing or arguing “in good faith” seem like a promising place to look.
Good faith is arguably a general organizing principle of the common law, associated with the ideas
of honesty, genuineness, fairness, cooperation, and reasonableness.
180
But in this section, I focus
on Regina Rini’s (2020) work on good faith debate in everyday settings.

3.1 Truth: Unnecessary and Insufficient for Good Faith
Rini considers a number of models of what good faith might require. On truth models of good
faith, only true contributions or truth-seeking debate can be in good faith, but such models face

179
See, e.g., Merricks 1995, 1997 and Moon 2012.
180
See Berger & Arntz 2016.

124
two problems. Firstly, we cannot assume there is always a fact of the matter. Our account of good
faith should not automatically count nihilism, truth-relativism, skepticism, or non-cognitivism
about certain domains, even important ones like the moral domain, as bad faith positions in a
dispute. Disagreement may be illusory according to both nihilism and non-cognitivism when there
is no fact of the matter about which to disagree, while it is faultless according to relativism and
trivial according to skepticism. So, on these views, disagreements (within a given domain) can be
of little epistemological significance, but they are not necessarily without social significance to
the disputants, and so long as interlocutors take the dispute seriously, true contributions and truth-
seeking do not seem necessary for them to dispute in good faith. Secondly, some paradigmatically
bad faith disputants are devoted to the truth, but being manipulative zealots, they are dogmatically
devoted to their version of the truth and willing to use any effective rhetorical trick others into
accepting their version of the truth. So true contributions seem unnecessary and truth-seeking
seems insufficient for making contributions to debate in good faith. So, the ordinary standard of
good faith does not entail truth.

3.2 Persuasion: Insufficient for Good Faith
The case of the manipulative zealot also illustrates that good faith requires more than an attempt
to persuade one’s opponent. Manipulative zealots seem to be able to aim at and succeed in
persuading their interlocutor,
181
through the use of manipulative techniques their interlocutor

181
At least, on the conventional understanding of the verb ‘persuade’ this seems possible. A more controversial
substantive account of persuasion may rule out the manipulative zealot’s behavior as a genuine instance of persuasion;
see, e.g., Rini 2018. However, the liberal anti-perfectionist should avoid relying on such accounts when looking to the
public political culture for guidance.

125
would reject, were they aware of them. But this seems like a paradigm case of debating in bad
faith. So attempts at persuasion are not sufficient for good faith debate.

3.3 Ideal Interlocutors: Unnecessary for Good Faith
A philosopher might be tempted to think that the ideal of good faith requires us to act like ideal
interlocutors, but this kind of model faces a different problem: the obvious ubiquity of fallible
interlocutors. Firstly, the ordinary everyday sense of acting in good faith does not obviously
require agents to be paragons of virtue. Secondly, even ideal interlocutors have to cut corners when
engaging with others who are more fallible. The Fields Medal winning mathematician, who dumbs
things down for me, when some complex mathematical ideas become relevant to our debate, is not
obviously being non-ideal. I am the non-ideal interlocutor, she is virtuously making allowances
for my weaknesses, and in so doing she seems to be acting in good faith, even if she can no longer
fulfill the demands of an ideal interlocutor model. So ideal interlocutors are not necessary for good
faith debate.

3.4 Transparent Persuasion: Unnecessary and Insufficient for Good Faith
Perhaps the real lesson of the manipulative zealot is that a good faith interlocutor persuades others
only by means they can thoughtfully accept. But this requirement still seems too strong. Consider
cases of adversarial demonstration. When Bayard Rustin, a close advisor to Dr. Martin Luther
King Jr., and Malcolm X debated the merits of Black nationalism and the goals of separation vs.
integration in 1961, neither thought there was any chance of persuading the other; each was trying
to persuade the audience. Still, both argued in good faith with their interlocutor, each earned the
respect of the other, and the exercise seemed so valuable to both men that they eventually decided

126
to take their debate on the road in 1962.
182
Though each was moved to some degree by the other’s
points, they never reached anything approaching agreement on these issues. This illustrates that
good faith disputes obtain even where deep disagreements persist, so it does not entail reaching a
resolution to which everyone consents.
But before we give up on the transparent persuasion model, perhaps we can revise it to
allow transparent audience persuasion, such that the good faith interlocutor sincerely aims at
persuading her audience (which may or may not include her interlocutor) using methods her
audience can thoughtfully accept. Unfortunately, even this fairly sophisticated model of good faith
is vulnerable to an even more sophisticated example of bad faith manipulation, which Rini calls
cynical puppeteering. The cynical puppeteer manipulates fair argumentative points to set up a
cheap shot. For example, they might present a fair criticism, in order to make hay out of the fact
that the fair rejoinder is complex or nuanced. The audience might thoughtfully accept the cynical
puppeteer’s criticism and even thoughtfully accept that the puppeteer’s interlocutor owes them an
answer, because, of course, in some sense they do, but the interlocutor is still forced to either walk
into a rhetorical trap or avoid the trap by not engaging with a fair criticism. So even transparent
persuasion does not yet seem like a sufficient requirement to capture the demands of good faith.

3.5 Good Faith Debate: Sincere, Cooperative, Adversarial Persuasion
Therefore, Rini proposes a model, on which a good faith interlocutor aims to participate in a shared
project of sincere, cooperative, adversarial persuasion:
Sincerity: people say what they really think;

182
See Joseph 2020, 117–122.

127
Shared Cooperative Project: each individual’s project of persuading someone is
a subsidiary of a larger shared joint doxastic project of the interlocutors (e.g.,
getting at the facts, identifying the best policy all things considered, etc.);
Adversarial Persuasion: there is genuine disagreement; it is not a mock
disagreement, or just an opportunity to show off one’s argumentative prowess;
interlocutors care about their disagreement and want to persuade others.

This account forecloses on lots of clearly bad faith conduct. Sincerity is obviously incompatible
with lying or bullshitting, but also subtler conversational moves, like attempting to poison the well,
only to retroactively withdraw a contribution from consideration, for example, by saying one was
“just joking.” Aiming to participate in a Shared Cooperative Project is incompatible with
gratuitous violations of Gricean norms, but also for example, with playing the provocateur by
saying gratuitously offensive things, taking demeaning or dehumanizing positions, or flirting with
the incitement of violence, etc., especially against one’s interlocutor. The provocateur typically
crosses a line, then demands not to be held appropriately accountable to ordinary (if implicit)
standards of conversation. But absent mutual agreement to suspend or interrogate them, good faith
interlocutors should expect ordinary everyday standards of conversation to be enforced. Finally,
Adversarial Persuasion is incompatible with doing things like playing the Devil’s Advocate.
Rini’s account also solves each of the problems raised by the counter-examples discussed
above. Canonical shared projects involve getting at the facts, but contrary to the prediction of the
truth model, shared project can also revolve around topics that are (ex hypothesi) non-factual, and
there can be good faith debate about them too. Manipulative zealots may be insincere, and since
they only care about getting at their version of the facts, they are typically unprepared to learn new
ones, so even if they are able to satisfy a simple persuasion model, they may not be able to engage
with their interlocutor(s) in a shared cooperative project. Accommodating fallible interlocutors,
however, is part of engaging in a shared project of cooperative persuasion. Similarly, Malcolm X
and Bayard Rustin engaged in a shared project of sincere, cooperative, adversarial persuasion of

128
their audience, even though neither expected to persuade the other. But the cynical puppeteer,
while aiming at persuasion, is neither sincere nor cooperative.
Since Rini’s model of good faith delivers the right verdicts about these cases it champions
a standard lying somewhere between truth (per the lesson of non-factual disputes) and consent (per
the lesson of Rustin and X’s persistent disagreement). Therefore, to the extent that these
judgements also reflect a standard whose acceptance presupposes no perfectionist views, and its
fundamental ideas seem familiar and drawn from the public political culture, it seems like a
promising candidate for the liberal anti-perfectionist’s standard of public reason. So let us think
about how this idea can inform our theorizing about reasonableness, public reasons, and public
justification.

4. Epistemological Externalism for Liberal Anti-Perfectionists
Given the pervasive assumption of epistemological internalism about reasonableness among anti-
perfectionists, many public reason liberals seem to have overlooked the possibility of building an
externalist epistemic dimension into their idealization or normative expectations of the reasonable.
But insofar as the reasonable are practically rational, their interest in tracking certain truths about
the external world is not merely contingent.

4.1 Virtue Reliabilism for Liberal Anti-Perfectionists
Consider the distinction between getting things right by accident and being so disposed as to ensure
that it is no accident you get things right. It seems implicit in the thinking of many public reason
liberals that the reasonable do not just happen to get things right, and it is similarly implicit in the

129
thinking of virtue reliabilists that the virtuous cannot just happen to get things right. Reliabilists
define virtues as stable, reliable, if imperfect, dispositions to perform in ways that are (sufficiently)
good, in terms of some norm or value, so virtuous agents are not those who just luck out; they have
some kind of competence. This competence displays what Philip Pettit (2014) calls robustness—
there is something in the world that ensures it obtains—it is not fragile. Pettit begins with the idea
that there are rich goods of attachment (e.g., friendship), virtue (e.g., honesty), and respect. Rich
goods have a common structure: a rich good provides others with a corresponding thin good
robustly. For example, the rich good of friendship provides a friend with the thin benefit of favor
robustly, the rich good of honesty provides one’s interlocutors the thin benefit of truth-telling
robustly, and the rich good of respect provides others the thin benefit of restraint or non-
interference robustly. To provide a thin benefit robustly is to be disposed to provide it in a way
that is “resilient enough to survive situational shifts” (24), assuring its provision both actually, “as
things actually are,” and counterfactually, “as they would be under certain variations” (46).
183
Many liberal theorists, including prominent anti-perfectionists, theorize reasonableness as
a civic virtue, and civic virtues should be understood as rich goods in Pettit’s sense. So a virtue
reliabilist can say that to be reasonable is just to be robustly and reliably disposed to cooperating
with others on fair terms in all matters of civic life, regardless of how much one understands the
value of cooperating fairly or why one is motivated to do so.
184
Indeed, this seems like an especially

183
See Hurley 2016.
184
Notice, for example, that virtue reliabilism for liberal anti-perfectionists does not require agents to accept virtue
reliabilism or liberal anti-perfectionism.

130
promising way of developing Klosko’s thin behavior-focused conception of reasonableness, Kelly
& McPherson’s notion of political reasonableness, or Nussbaum’s notion of ethical
reasonableness.
185
Of course, individuals are free to supplement this thin conception of
reasonableness
186

Happily, Rini (2020) also describes good faith participation in debate as a virtuous practice,
and some argument theorists, motivated by similar kinds of considerations, find virtue-theoretic
accounts of argumentation independently plausible and attractive.
187
Moreover, Rini’s account of
good faith is compatible with virtue reliabilism, because it makes no special demands on the good
faith interlocutor, concerning how much they understand or are motivated by the value of
cooperating fairly in a shared project of sincere, adversarial persuasion. So far, so good.

4.2 Reliabilism about the Highly General Virtue of Reasonableness
On the Rawlsian picture, the ideal of democratic citizenship consists of several virtues of social
cooperation, including reasonableness and civility; each virtue is a family of dispositions,
regulated by a higher-order desire to act in accordance with corresponding moral ideals, values, or
principles, and individuated by its governing deontic consideration(s). But let us dispense with
Rawls’s psychological assumptions about regulative higher-order desires, characterizing virtues

185
Cf. Liveriero 2020, 907–908.
186
I defend virtue reliabilism about liberal civic virtues in chapter 3, section 4.
187
See, e.g., Aberdein 2010 and Cohen 2005, 2019. Godden 2016 objects to aretaic accounts of what is intrinsically
valuable in argumentation or reasoning, but since liberal anti-perfectionism is only interested in its instrumental value,
it can remain neutral in that dispute. For my response to similar worries, see my response to Boettcher 2012 in section
5.2 of Chapter 3. I also consider my own cast of vicious public reasoners in Chapter 1, .

131
only as stable, reliable (if imperfect) dispositions to perform in ways that are (sufficiently) good,
in terms of some norm or value.
Reasonableness aims at a social order where all are free, equal, and can live together on
terms everyone accepts, so it is largely a matter of being committed to fair cooperation in all
matters of civic life. Therefore, following Klosko, Kelly & McPherson, and Nussbaum, let all-
things-considered judgments about successfully cooperative civic behavior determine who is
presumed reasonable:
Cooperation First: All-things-considered assessments of others’ reliability in fair
cooperation (or unreliability, non-cooperation, or unfairness) takes priority over
piecemeal evaluations of their cognitive, religious, philosophical, theoretical,
or epistemic (in)competence when assessing reasonableness.

This gives us our first rough approximation of the reasonable constituency. Like Kelly &
McPherson, we will allow that this presumption can be defeated where some form of intellectual
incompetence spills over into behavioral, practical, moral, or political failures to cooperate fairly
and reliably with others.
Reasonableness is manifested in forming and maintaining attitudes like justified beliefs,
rational preferences, moral judgments, and in the performance of actions, including rational
decisions, enkratic choices, and moral acts. These manifestations, among other things, provide
assurance to others that we are generally competent, reliable, fellow participants in society.
Reasonableness may also be understood to be abstractly manifested by paradigmatically
reasonable views or actions, even when the agent in question is an otherwise unreasonable person.
For just as a vicious person may perform a paradigmatically virtuous act without virtuous motives
(e.g., save a drowning child only because they know they are being watched), an overall
unreasonable individual may espouse some paradigmatically reasonable views or perform
reasonable actions. But the reasonableness we predicate of persons picks out a stable disposition

132
to act in certain ways, while the reasonableness we predicate of moral, political, religious, and
philosophical doctrines reflects our judgment that they are often formed and maintained by
reasonable persons. The reasonableness we predicate of communities organized around their
shared moral, philosophical, political, or religious perspective, may either pick out the abstract
reasonableness of their shared doctrines (despite the fact that individual members of the
community may not individually manifest any reasonableness in holding the view), or reflect the
judgment that individual members of the community tend to be reasonable.

4.3 Virtue Reliabilism about Reasonable, Civil Public Inquiry
The disagreements with which public reason theorists typically concern themselves involve parties
with opposing attitudes toward some (prospective) exercise e of authority or power, especially the
institution or revision of coercive laws, public policies, and perhaps social norms. Where (at least)
one member of a society is for e and another is against e, there is a disagreement of (at least some)
social or political consequence. And where we predicate reasonableness of such disagreements,
we either convey the judgment that each party’s attitude toward e is the manifestation of their
reasonable agency, or more weakly, that neither party’s attitude disqualifies them from being
reasonable overall.

In the domain of public inquiry, where reasonable disagreements over the exercise of
power are disputed, reasonableness applies alongside another virtue of social cooperation: civility.
Within this domain, these virtues aim characteristically (though perhaps not exclusively) at
publicity and fair-mindedness, they are manifested through the giving and receiving of public
reasons or justifications for exercises of coercive power, and their aims are realized through

133
successful policymaking (or unmaking) constrained by publicity and fair-mindedness. Rawls
characterizes “duty of civility” as follows:

Publicity Constraint: Agents must be able to explain to one another how the
exercises of political power, which they defend or advocate, can be justified
based solely on public reasons;

Fair-Mindedness Constraint: Agents must be willing to listen to others and be
fair-minded in deciding when to accommodate their views (see 1993, 217).

Rawls also recognizes roles for advancing non-public justifications (outside of public reasoning),
through other forms of discourse, which he called declaration, reasoning by conjecture, and
bearing witness.
188
I tend to agree with Rawls’s many critics, who find this separation too
strict.
189
In light of Rini’s insights into good faith debate, I am inclined to interpret publicity not as
a strict prohibition on appeals to non-public religious, moral, or philosophical doctrines, but as a
strong caution against how easily such appeals can devolve into bad faith forms of zealotry. In
contrast, sincere, cooperative, adversarial use of declaration, witnessing, and reasoning by
conjecture all seem to be legitimate conversational moves to avoid bad faith appeals to such
doctrines.
So, whereas civility is a disposition to communicate respect, tolerance, and consideration
to others in public inquiry,
190
I propose that we understand it to be governed by the discursive
standards of good faith debate, implicit in public political culture. Reasonableness in public
inquiry, on the other hand, is primarily concerned with one’s own moral and intellectual conduct

188
See 2005, 465–466; Schwartzman 2011 and 2012.
189
See, e.g. Liveriero 2020, 902.
190
See Calhoun 2000, 255; Costa 2004, 154.

134
in successfully judging, for example, when reasons are suitably public and when there is sufficient
justification to accommodate others’ views, because these aims serve the over-arching goals of
reasonableness,
191
and as Audi observes, the better informed one is about the issues of the day, the
more likely one is to successfully exercise this virtue. Something like Wall’s all-purpose
externalism about public reasons seems too strong a requirement, but Peter’s context-sensitive
externalism about public justification seems plausible. For some disputes, having the facts is more
important, and while those less informed may still participate in good faith in debates over these
matters, so that public reasoning still does not entail truth (or consent), they are at greater risk of
behaving unreasonably.

4.3 Epistemological Externalism about Reasonableness
Whereas the virtue reliabilist can say that to be reasonable is to be robustly and reliably disposed
to cooperating with others on fair terms in all matters of civic life, the virtue externalist can allow
them to index this to a particular social milieu or environment, for example, requiring the
reasonable agent to be sensitive to certain social, historical, political, or economic facts about their
specific society. In matters of public inquiry, reasonableness is manifested in the agent’s tracking
general facts relevant to fair cooperation, robustly—especially social facts about how the burdens
and benefits of cooperation are distributed in one’s society, but also facts about what is common
knowledge, what is controversial in one’s society, what matters one is well positioned to judge,
which matters others are better positioned to judge, how we might engage in further inquiry in

191
This synthesis of Rawls is defended in Chapter 3, section 2.2.1.

135
order to settle certain disputes, whether it is cost-effective or cost-prohibitive to engage in further
inquiry before acting, etc. We value this highly general virtue of reasonableness for lots of reasons,
but its epistemic dimension provides us with the thin benefit of ensuring that an agent’s
contributions to our shared social life and shared task of democratic governance are informed and
constrained by reality.

4.4 The Instrumental Value of Reasonableness
For as Paul Weithman explains, a liberal conception of justice, including norms of public reason,
is inherently stable, “if a society that is well-ordered by it generally maintains itself in a just general
equilibrium and is capable of righting itself when that equilibrium is disturbed.”
192
An inherently
stable just equilibrium is clearly one of the goals liberalism is designed to secure in an open
pluralistic society, but “stability is threatened if citizens lack sufficient reason to think others will
do their share.”
193
Weithman calls this the mutual assurance problem.
Reasonable citizens are concerned with providing and securing this kind of mutual
assurance, and this assurance is demonstrated to others and secured from others by manifesting
one’s stable disposition to be reasonable in social and political conduct, not only in action but in
attitudes that inform action. In any context where this assurance from others can be threatened by
their conduct, the reasonable give priority to evaluating whether others can be relied upon as fair
cooperators and competent fellow participants in society, and this explains why reasonableness is
about cooperation first.

192
Weithman 2010, 44; see Rawls 1999, 401.
193
Weithman 2010, 46.

136
The Cooperation First constraint means that, even in contexts of public inquiry, the
evaluation of others’ epistemic credentials is not an end in itself for reasonable persons qua
reasonable. So if someone is a reliable cooperator, any remaining epistemic flaws you might detect
in them, which are nonetheless compatible with their reliability as a fair cooperator, are irrelevant
to assessing their reasonableness.

4.5 Why Putting Cooperation First Provides Us with Enough
There is a different kind of evaluative standard that would be driven by assessments of others’
epistemic credentials and would draw the constituency of the reasonable around those meeting
some required minimum level of epistemic competence. For example, R.J. Leland and Han van
Wietmarschen (2012) argue that public reason theorists are committed to a quite demanding
standard of epistemic competence among the reasonable. But I contend that, depending on what is
at stake in a particular disagreement, modest normative expectations of epistemic competence are
often sufficient for the purposes of assuring others of one’s reasonableness. After all, Leland and
van Wietmarschen concede that there is no guarantee that even highly epistemically competent
people accept central liberal-democratic political values, like freedom of conscience, equality
before the law, or freedom of the press, and people who reject these values are unreliable
cooperators for mutual advantage even if they rank among the most epistemically competent
reasoners on social and political matters. And since modest epistemic expectations help to ensure
that a robust diversity of people, perspectives, and disagreements are regarded as reasonable,
194
an

194
See Leland and van Wietmarschen 2012, 730.

137
account that avoids a demanding requirement of epistemic competence is better positioned to
respect reasonable pluralism. So the kind of standard I am advancing draws the constituency
around reliable cooperators, and that standard is only partly epistemic.
To illustrate the plausibility of this constraint, consider a case of small-scale coordination
among (plausibly) reasonable disagreeing parties: the faculty members of a large philosophy
department. In any large philosophy department there’s a wide range of competing
methodological, historical, metaethical, political, normative ethical, applied ethical, agential,
noetic, linguistic, metaphysical, epistemological, aesthetic, and religious views represented in the
department, including departments where everyone’s uncontroversially competent and qualified.
Suppose that most faculty members are quite certain about their favored philosophical views, and
there are some doubts among them about the quality of one another’s epistemic justifications for
holding views incompatible with their own. While there’s certainly a place for healthy debate and
exacting scrutiny of these issues in the seminar room qua philosophers, the department needn’t
settle questions about each other’s epistemic credentials on these issues in order to effectively
perform its functions qua philosophy department, like being a coordinated well-functioning
academic unit in the university, judiciously revising its policies and curriculum as needed,
collegially supporting the teaching and scholarship of its members, responsibly carrying out
procedurally fair job searches, and providing an excellent philosophical education to its students.
This is because their epistemic credentials about deeper philosophical issues are of limited
practical significance to the success of these cooperative projects that constitute the department’s
raison d’être. So if even a philosophy department (composed of philosophers!) can perform its
functions effectively without getting bogged down by concerns over its members’ epistemic
justification concerning all manner of deep disagreements, how much more so can a society?

138
5. Putting Externalistic Reasonableness to the Test
Let us precisify the different respects in which we might predicate reasonableness of various
actions, drawing on a threefold distinction, employed by virtue epistemologists,
195
to assess acts
or performances according to their (1) accuracy, (2) adroitness, and (3) aptness. To illustrate these
distinct act-focused valuations, recall Paris’s act of shooting the arrow that famously struck
Achilles’ vulnerable heel. In addition to wider moral or practical assessments of this action, we
can assess the performance of the shot qua shot. First, the arrow to hit its intended target, so the
shot is accurate. Second, we can assess whether the shot manifests Paris’s skill or competence in
archery, for example, by an expert release of the arrow. Supposing the shot does indeed manifest
Paris’s competence, the shot is also adroit. Third, we might assess whether it is accurate because
it is adroit, that is, whether it is apt. For imagine the arrow was adroitly shot, but another projectile
knocked it off-course, only for Bright Apollo to divert it back on-course, ensuring it found its
target. In that case, the shot would not be apt, for its accuracy in finding its intended target, would
not manifest Paris’s competence, but Apollo’s.
Each of these act-focused valuations has, in turn, an agent-focused analogue applicable to
attempts. For whether or not an archer’s shot successfully finds its target, we can assess the
accuracy, adroitness, and aptness of the agent’s attempt. So consider a case where, in the heat of
combat, Paris fires but Achilles is improbably obscured behind a flying piece of debris at precisely
the right (or wrong) moment. Paris might nonetheless have accurately aimed the shot at Achilles.
Moreover, he might have manifested all his competence as an archer, for example, with his perfect

195
See, especially, Sosa 2007, 22–43.

139
form, even though he failed to predict this highly improbable turn of events. So the unsuccessful
attempt can also be adroit in this agent-focused sense. And supposing his aim is accurate because
of the adroitness with which he attempts the shot, the attempt itself could be considered apt from
an agent-focused evaluative point of view, even though unsuccessful. With these distinctions in
mind, let us consider a case to see how reasonable an agent can be on purely internal or agent-
focused grounds. Within this virtue-theoretic framework, internalist epistemic rationality
corresponds to an agent-focused valuation of doxastic attempts to track the external world. But the
question for us is whether agent-focused valuations, like those canvassed earlier in section 2,
exhaust the epistemic dimension of reasonableness. If not, this suggests that the epistemic
dimension of reasonableness includes some act-focused valuation of success in the external world.

5.1 Accura of Utopia
Accura is a member of the liberal-democratic society of Utopia, and she exhibits the following
Package of Internalist Liberal Excellences (PILE):
PILE: the agent possesses the capacities for a sense of justice and for a conception
of the good; has a determinate conception of the good interpreted in the light of
some non-public doctrine(s); is a normal, fully cooperating member of society
over a complete life; and is ready to propose and willingly abide by principles
and standards that are fair terms of cooperation, given assurance others will
likewise do so. The agent is also committed to central liberal-democratic
political values, like freedom of conscience, religious toleration, freedom of the
press, equality before the law, equality of opportunity, the rule of law, and
democratic governance. The agent is committed to observing the norms of
public reason, including the publicity and fair-mindedness constraints.
Moreover, the agent embraces the fact of reasonable pluralism, believing at
least many of those with whom they have deep doctrinal disagreements are
nonetheless reasonable. Additionally, the agent possesses the intellectual
powers of judgment, thought, and inference, and exhibits (a high degree of) the
internalist’s notion of epistemic rationality, avoiding logical error and fallacious
reasoning, prejudice, bias, excessive self/group-interest, and willfulness, while
updating their doxastic system in light of new evidence she receives, and

140
diligently resolving any inconsistencies arising among her epistemic
commitments.

Indeed, given Accura’s foundational commitments, the doxastic system she’s built for herself is
exceptionally well-ordered. Among Accura’s doxastic attitudes is the belief that poor members of
her society are simply lazy. Relatedly, she believes Utopia is a land of opportunity where people
would not be poor if they worked hard. In fact, Utopia is a land where equality of opportunity is
so perfectly realized that no one fails to avoid poverty, except by complete lack of effort. When
Accura watches Facts News, the reports she receives about equality of opportunity, high social
mobility, the lack of any correlation between one’s parents’ wealth and one’s attainments in life,
etc., are highly reliable, and provide an accurate picture of the contemporary state of affairs in
Utopia. Given her impressive body of evidence, Accura is also internalistically justified in thinking
this particular set of empirical facts is not something she needs to revisit. So, for example, her
credences are highly resilient in light of counter-evidence on this and closely related economic
issues. However, Accura is not generally dogmatic about matters of reasonable disagreement.

5.2 Aimy of Pantopia
Aimy lives in Pantopia. For every attitude Accura has toward Utopia/Utopians, Aimy has the
counterpart attitude toward Pantopia/Pantopians. At one time, Aimy was indistinguishable from
Accura in every respect. Even now, she continues to exhibit the same high degree of internalistic
epistemic rationality. However, since October 7
th
, 1996, Aimy has been watching Faux News, and
apart from superficial differences in naming, she has been receiving the very same reports about
Pantopia that Accura receives from Facts News about Utopia. Like Accura, she believes poor
people in her society are simply lazy. Unfortunately for Aimy, she has gradually been misled by
the subtle machinations of Faux News, whose reports have gradually departed from verisimilitude

141
with the real world. Pantopia is not (now) a land that is anywhere close to perfectly realizing
equality of opportunity. But Faux News is backed by a powerful corporate media empire exerting
unparalleled manipulative control over its audience. So contrary to what Aimy is often told by her
seemingly reliable news source, vanishingly little poverty in Pantopia can actually be explained
by the laziness of poor people.
Like Accura, Aimy embraces the fact of reasonable pluralism, believing in general that at
least many of those with whom she has some deep doctrinal disagreements are nonetheless
reasonable. But she has been deceived about the fact that others reasonably disagree with her about
the causes of poverty in Pantopia, and she is internalistically justified in thinking this particular set
of empirical facts is not something she needs to revisit. For suppose Faux News targets a politically
privileged class of citizens, of which Aimy is a member, for isolation within an echo chamber,
196

to ensure the usefulness of this privileged class to its political agenda. So Aimy is highly resistant
to correction on this issue and related economic issues that figure centrally in the propaganda of
Faux News. But Aimy is not generally resistant to correction on matters of reasonable
disagreement, nor is she more (or less) closed-minded than Accura. Moreover, the isolation of this
politically privileged class of citizens makes them no less eligible to vote, liable to taxation, subject
to the law, etc., so in no ordinary sense can it be said to make them less a part of Pantopian
society.
197

5.3 Public Reasoning in Utopia vs. Pantopia

196
For a useful discussion of the distinction between epistemic bubbles and echo chambers, see Nguyen (forthcoming).
197
My thanks to Emily Podhorcer, for pressing me about the details of this case.

142
It is now present day. Accura and Aimy each make attempts, φAccura and φAimy, to publicly justify
to their fellow Utopians and Pantopians, respectively, a proposal to eliminate the estate tax without
any kind of replacement. In both countries, the estate tax is a tax on property transferred from
deceased persons to their heirs. Only the %0.2 wealthiest estates currently pay the tax in each
country, because it is levied only on the portion of an estate’s value exceeding the specified
exemption level of $5.49 million per person (effectively $10.98 million per married couple). In
Utopia, the estate tax is but a small part of an economic regime that more than adequately ensures
equality of opportunity and high social mobility for all, and this system is very likely to continue
to do so with or without this somewhat gerrymandered tax. In Pantopia, by contrast, the estate tax
serves as one of the few modest correctives to other tax rules that provide massive tax benefits to
income from wealth.
198

Given the state of affairs in Utopia, φAccura is compatible with Accura’s moral decency—
her sense of justice and conception of the good, her being a normal, fully cooperating members of
society over her complete life, her readiness to propose and willingly abide by principles and
standards that are fair terms of cooperation, given assurance others will likewise do so, and her
substantive normative commitments to central liberal-democratic values. On the other hand, given
the actual state of affairs in Pantopia, φAimy is incompatible with Aimy’s sense of justice and
conception of the good. She aims to be a fully cooperating member of society throughout her life;
she is, with the correct information, ready to propose and willingly abide by principles and

198
Suppose, for example, capital gains are taxed at lower rates than wages and salaries. So the top 0.1 percent of
Pantopian taxpayers — those with incomes above $3.1 million — receive 56 percent of the benefit of the preferential
capital gains rates, worth more than $600,000 apiece. Other tax rules allow part of the income of the very wealthiest
to go completely untaxed, even with the estate tax; cf. Huang and Cho 2017.

143
standards that are fair terms of cooperation, given assurance others will do likewise; and she is
committed to (among other things) equality of opportunity and social mobility. But Aimy’s
abnormal epistemic situation—the very unfortunate disconnection between her internal doxastic
system and the external world—leaves her ill-equipped to properly assess what is fair to her fellow
Pantopians. For eliminating the estate tax, without introducing some other kind of taxation that
would do at least as much to address grinding economic inequality in Pantopia, is simply
incompatible with the project of living together on fair terms of cooperation for mutual advantage.
Moreover, Aimy’s abnormal epistemic situation, strongly disposes her to many such behaviors
that are especially relevant to social cooperation. So I submit that Aimy, through no fault of her
own, cannot be counted reasonable by the time φAimy occurs (at least), and probably much earlier.
199

This case demonstrates that without a sufficient grasp on certain facts about the external
world one cannot count as reasonable, and moreover one’s disagreement with others about a
particular exercise e of power or authority cannot be reasonable without a sufficient grasp of the
normatively significant facts relevant to e. For even if Aimy is reasonable in certain respects—for
example, the way in which she rationally forms beliefs or performs certain actions—she’s
paradigmatically unreasonable in other respects of greater practical consequence to the project of
living together on fair terms of cooperation for mutual advantage. For reasonableness is a matter
how we actually interact with fellow members of our society, and not merely of how we take
ourselves to be interacting with them. So reasonableness just cannot be all in the head.

199
I do not insist on the stronger claim that Aimy is all-things-considered unreasonable, as I think the case would have
to be fleshed out in greater detail to settle that further evaluative question.

144
5.4 Virtue Externalism about Public Justification
For all we know so far, it’s possible that φAccura and φAimy might each prove successful. So, for the
sake of argument, let us assume both attempts succeed. Given the centrality of public reasoning
and public justification in the job description of reasonableness, the virtue of reasonableness is
paradigmatically predicated of an agent’s successful justification of exercises of authority or power
on the basis of public reasons. So φAccura and φAimy are paradigmatically accurate in their aim, and
as we have just stipulated, they are accurate in the act-focused sense as well.
Suppose further that each manifests her reasonableness by being civil, and by appealing to
reasons R she believes to be both individually public and jointly conclusive. In that case, each
performance is also adroitly attempted. But while φAimy may be adroitly reasonable in this agent-
focused sense, it cannot be adroitly reasonable in the act-focused sense. Aimy is reasonable in
certain respects, of course. She rationally forms those beliefs on the basis of which she attempts to
publicly justify the repeal of the estate tax, and rational belief formation is one way of manifesting
one’s reasonableness. However, because Aimy gets the relevant normatively significant facts so
badly wrong, φAimy fails to manifest her reasonableness in precisely that respect in which the
attempt is of greatest social and political consequence in the external world.
What about aptness? May we suppose each performance’s adroitness explains its accuracy?
The success of φAccura is not a function of any kind of deviant causal chain. For example, no comedy
of errors leads Accura to sincerely mistake some idiosyncratic set of considerations R1 for
conclusive public reasons, only for her to be bumped on the head, forget R1, and successfully
justify the repeal of the estate tax by appeal to a different set of genuinely conclusive public reasons

145
R2 in her resulting confused state. In that case, φAccura would be successful, despite its inaptness.
200

Happily, in this case, Accura’s performance is apt and, in every respect, a virtuous example of
public reasoning by a reasonable agent.
φAimy accurately and adroitly manifests Aimy’s reasonableness in an agent-focused sense,
and nothing prevents us from charitably supposing the former is explained by the latter, so that the
attempt also aptly manifests an agent-focused reasonableness. But to see why agent-focused apt
manifestation of reasonableness is of little help to the public reason liberal project, notice how
surprising it would be to find that Aimy had succeeded in publicly justifying the estate tax repeal,
given her abnormal epistemic situation. For her actual fellow Pantopians are the target addressees
of φAimy, and Aimy is now badly mistaken about the real state of affairs in Pantopia affecting these
people. The very unfortunate disconnection between Aimy’s internal doxastic system and the
external world leaves her ill-equipped to properly assess what is fair to her fellow Pantopians. If,
somehow, she still managed to successfully publicly justify the repeal, this outcome would not
primarily be a result of Aimy’s agent-focused adroit manifestation of her reasonableness, given its
disconnection with the external world, but of her astonishingly good luck. Therefore, in order for

200
Conversely, an agent may aptly manifest her reasonableness, in the act-focused sense, without her attempted public
justification being entirely successful. To illustrate, suppose Aptish advances R1 and R2 in favor of a proposed
exercise of power e, to his addressee, Addréa. R1 is a conclusive public reason for e, so Aptish succeeds in publicly
justifying e. However, R2 is a falsehood (logically independent from R1). So while Aptish’s advancing of R1 aptly
manifests his reasonableness in the act-focused sense, the advancing of R2 is, at best, apt in the agent-focused sense.
Addréa, being a fair-minded person, recognizes that R1 provides conclusive public justification for e and is thereby
persuaded to accept e, but she also notices that R2 is false. The important question is: how badly does this undermine
Aptish’s reliability as a fellow cooperator? The answer: it depends. The more it seems to Addréa that Aptish
accidentally got R1 right, the more an interaction like this will undermine Addréa’s confidence in Aptish’s
dependability more generally. But the more it appears instead that Aptish got R2 wrong by accident, the less
undermining it will be of Addréa’s confidence in his reliability as a fellow cooperator. Whatever Addréa’s judgment
turns out to be, Aptish’s successful public justification of e remains itself abstractly reasonable (to some degree), and
coordinating on that fact can still advance the cooperative project.

146
an agent to stably assure others of her dependability as a fair cooperator, she must aptly manifest
her reasonableness, in the act-focused sense, and do so reliably (if fallibly) over time.
6. Reasonableness and Luck
Because of how surprising it would be to find that φAimy had succeeded, given Aimy’s abnormal
epistemic situation, the luckiness of her success undermines her ability to aptly manifest her
reasonableness, in the act-focused sense, and thereby assure others of her reliability as a fair
cooperator over time. For an attitude or action relevant to social and political cooperation is a
product of reasonableness-undermining luck if the operative dispositions are lucky to hit their
target, holding external conditions fixed. So if your internal disposition is such that it is just lucky
you hit the relevant target, given your environment, then your disposition is not reliably successful,
so neither you nor others can depend on its successful manifestation.
To be sure, Accura is lucky in a different constitutive respect: holding only her internal
dispositions fixed, she could easily have been in Aimy’s position, had she found herself in an
epistemically hostile environment like Pantopia, rather than Utopia. However, given that Accura
is in Utopia, her internal dispositions ensure that her doxastic processes and attitudes hit their
targets with a high rate of success. Accura’s act-focused aptness in the manifestation of her
reasonableness ensures φAccura is non-accidentally successful, because act-focused aptness
guarantees the right kind of stable, though fallible, causal connection between Accura’s underlying
stable disposition to be reasonable and her success in endeavors like φAccura. This makes Accura’s
reasonableness the kind of practical virtue her fellow Utopians can depend on in social and political
coordination. Therefore, because Accura is well-equipped for epistemic success in her actual
environment, and (in combination with her non-epistemic manifestations of reasonableness) she

147
can be relied upon, as a competent, dependable, fellow participant in cooperative Utopian society,
the fact that she’s (in some sense) lucky to be constituted in a way that makes her successful, given
her epistemic environment, this does not undermine her reasonableness.
By contrast, Aimy’s success is matter of reasonableness-undermining luck. Aimy is in
Pantopia, and given those external conditions, her merely agent-focused reasonableness leaves her
doxastic processes and attitudes poorly suited to hitting their targets (with a high rate of success).
This renders practically irrelevant those purely internal respects in which Aimy is reasonable. For
this reason, despite her occasional good luck, Aimy’s unlikely to assure her fellow Pantopians of
her reliability over time as a competent, dependable, social and political cooperator. Therefore,
purely internal agent-focused manifestations of reasonableness, disconnected as they are from
success in the external world, cannot be the dispositional basis for assuring others of one’s
reliability as a fellow cooperator on fair terms over time. So no matter how epistemically rational
someone might be on the inside, they cannot hope to manifest their reasonableness, in the right
kind of way, without also having a sufficient grasp of relevant facts in the external world. The
epistemic dimension of reasonableness, therefore, consists partly in having stable epistemic
dispositions, which ensure sufficient externalist justification for one’s positions on (prospective)
exercises of power or authority in one’s society. Without sufficient externalist justification, one
may succeed in persuading others and see one’s favored positions prevail, but one cannot aptly
manifest one’s reasonableness in so doing.

6.1 Reasonable Dispositions in New Brunswick vs. Nigeria
To further illustrate why reasonableness-undermining luck is a matter of whether one’s
dispositions relevant to social and political cooperation are lucky to hit their target, holding

148
external conditions fixed (as in Aimy’s case)—not a matter of whether one is lucky to be in a
cooperative environment, holding fixed their internal dispositions (as in Accura’s case)—consider
the following two pairs of public reasoners in two different societies.
Kendra and Pristine live in New Brunswick, Canada. Oba and Wemimo live in Nigeria.
Each one has gradually and unintentionally developed habits of attention, such that Kendra and
Oba are comparatively more attuned to respecting conventionally recognized authorities, while
Pristine and Wemimo are comparatively more attentive to considerations of purity or sanctity,
especially in relation to the human body. Nevertheless, let us stipulate that all four exhibit the
aforementioned Package of Internalist Liberal Excellences (PILE). Indeed, given their
foundational commitments, their doxastic systems are all exceptionally well-ordered. The
following thought experiment shows that even though two individuals, A and B, may be internally
constituted in ways that lead A to be reasonable and B unreasonable in one social milieu, given the
same internal constitutions, we can expect A’s counterpart to be unreasonable and B’s counterpart
reasonable in some other social milieus. This provides a further demonstration that purely
internalist excellences are not sufficient for reasonableness.

6.1 Kendra and Pristine in New Brunswick
Per the above stipulations, Kendra and Pristine are equally reasonable along all agent-focused,
internalistic dimensions of moral, practical, and epistemic evaluation. Pristine’s purity-directed
habits of attention draw her toward information related to things like handcrafted soaps, whole
organic foods, eco-friendly minimalist living, and local indigenous peoples’ notions of the sacred,
leaving her doxastic house very tidy, but rather Spartan with respect to many other kinds of
doxastic furnishings. Yet, inspired by idealistic visions of a wholesome, cooperative, pluralistic

149
free society, she’s committed to participating in public discourse, and always in accordance with
the norms of public reason.
Unfortunately, Pristine has come to have grave concerns about the safety of certain
vaccines. While browsing online, she encountered a cleverly disguised advertisement, passing
itself off as genuine reporting from NB News, that redirected her from NB News’s Facebook page
to The Daily NB Tabloid. For good reasons, she has long been generally distrustful of the big
pharmaceutical companies that produce vaccines and the weak oversight provided by government
regulators. So when she reads, from a seemingly reliable online news source, that children are
developing autism from the measles-mumps-rubella (MMR) vaccine at an alarming rate, she does
a search of news sources that cater to her purity-oriented interests. On balance, these sources seem
to confirm that concerns about the MMR vaccine have merit, and they’ve shown themselves to be
reliable about related matters in the past. Pristine then begins discussing the matter with trusted
friends and acquaintances, so that by the time she begins publicly advocating for changes to the
medical standards of care that mandate MMR vaccination in children, she has built up an
impressive repertoire of anti-vax talking points that (ironically) inoculate her against many of the
best counter-arguments.
While seeking to publicly justify her favored revisions to the medical standards of care,
she appeals primarily to a concern for the safety of innocent children, and secondarily to concerns
for the prevention of fraud by the pharmaceutical industry and resistance to government
corruption. In subsequent media interviews, she also declares her commitment to the sanctity of
the body, which plays an important motivational role for her personally, disposing her to more
readily accept the costs of lacking measles-mumps-rubella immunity, including the loss of herd
immunity that protects those too vulnerable or immunocompromised to be vaccinated. However,

150
she makes no attempt to publicly justify her position on the basis of this idiosyncratic philosophy,
because she’s satisfied that the public considerations to which she’s appealed, all by themselves,
outweigh the alleged benefits of the defective vaccine.
Kendra acknowledges that Pristine is sincere, and moreover, that her appeals to child
protection, fraud prevention, and resisting government corruption are all of the right kind. For
while legal changes to the standards of medical care in the name of bodily purity would be too
invidious in a pluralistic society, protection from potentially harmful, fraudulent vaccines is a
suitably public concern. Nevertheless, Kendra knows that Pristine’s fears about the MMR vaccine
have no basis in scientific evidence or fact.
201
Kendra knows this because her own tendency to
respect authority disposes her to first seek out and then accept the sound reasoning of scientific
authorities who reject the methods and conclusions of anti-vax pseudo-science.

6.2 Oba and Wemimo in Nigeria
Again, per our earlier stipulations, Oba and Wemimo are equally reasonable along all the agent-
focused, internalistic dimensions of moral, practical, and epistemic evaluation. Wemimo’s purity-
directed habits of attention draw him toward information related to things like the agricultural,
religious, and environmental impacts of post-colonialism, the virtues of certain traditional
subsistence farming methods, and local understandings of ritual purity, leaving his doxastic house
very tidy, but rather Spartan with respect to many other kinds of doxastic furnishings. Yet, again,

201
See e.g., Peltola 1998, Meldgaard et. Al. 2002, Mrozek-Budzyn et. Al. 2010, Jain et Al. 2015.

151
inspired by idealistic visions of a wholesome, cooperative, pluralistic free society, he’s committed
to participating in public discourse, and always in accordance with the norms of public reason.
Wemimo has come to have grave concerns about the safety of processed foods, believing
that items being sold by local vendors and grocers are of such dubious provenance that single
servings may prove highly poisonous—even fatal. He’s already generally distrustful of both the
processed food industry and the weak oversight of government regulators. So when he reads from
a seemingly reliable online news source
202
of poisonous counterfeit foods, like plastic rice or
vegetable oil made of recycled oil unfit for human consumption, popping up across the continent
and a local incident in Abuja where two 14-year-olds died and several more were poisoned by
tainted biscuits at a birthday party, he feels compelled to learn more. He does a search of news
sources that cater to his purity-oriented interests. These sources confirm that concerns about
counterfeit processed foods have merit, and they’ve shown themselves to be reliable about related
matters in the past. He decides this is the last straw, and begins publicly spearheading the effort to
bring in much stronger regulations on vendors and suppliers of cheap processed foods.
While seeking to publicly justify his favored regulatory regime, Wemimo appeals primarily
to a concern for the safety of children, and secondarily to concerns for the prevention of fraud by
the unregulated processed foods industry and resistance to government corruption. In subsequent
media interviews, he also declares his commitment to the sanctity of the body, which plays an
important motivational role for him personally, disposing him to more readily accept the costs of
increased regulation. However, he makes no attempt to publicly justify his position on the basis of
this idiosyncratic philosophy, because he’s satisfied that the public considerations to which he’s

202
Nwuneli 2018.

152
appealed, all by themselves, outweigh the alleged benefits of having no red tape to keep processed
foods cheap and readily available.
Oba acknowledges that Wemimo is sincere, and moreover, that his appeals to food safety,
fraud prevention, and resisting government corruption are of the right kind. For while legal
enforcement of healthy eating for health eating’s sake would be too invidious in a pluralistic
society, protection from potentially poisonous, counterfeit food is a suitably public consideration.
Nevertheless, Oba believes these fears about counterfeit processed food stuffs are baseless,
because the economic authorities on which he’s come to rely, insist that the existing legal liability
of businesses and burdensome safety regulations, imperfect as they are, are strong enough to
prevent such obvious frauds. Moreover, seemingly reliable news sources, catering to his deference
to conventional authorities, seem to corroborate these claims. So while Oba concedes that
processed foods are very unhealthy, he thinks there’s no real threat of them being counterfeit. He
therefore rejects Wemino’s proposal as being unworthy of accommodation, and appeals to the
epistemic authority of the economists he trusts in publicly defending the status quo.

6.3 Intellectual Dispositions are Reasonable Relative to One’s External Social Conditions
Given the actual states of affairs in the world, Kendra’s deference to conventional authority in
New Brunswick, and Wemimo’s purity-oriented values in Nigeria are compatible with their moral
decency—their sense of justice and conception of the good, their being normal, fully cooperating
members of society over their complete lives, their readiness to propose and willingly abide by
principles and standards that are fair terms of cooperation, given assurance others will likewise do
so, and their substantive normative commitments to central liberal-democratic values. On the other
hand, given these same states of affairs, Pristine’s attempt to have New Brunswick’s MMR

153
vaccination protocols revised and Oba’s defense of Nigeria’s insufficient food regulatory regime
are each incompatible with their senses of justice and conception of the good. They each aim to be
fully cooperating members of their respective societies throughout their lives; they are, with the
correct information, ready to propose and willingly abide by principles and standards that are fair
terms of cooperation, given assurance others will do likewise; and they’re committed to (among
other things) the central liberal-democratic values. But their abnormal epistemic situations—the
very unfortunate disconnection between their attitudes and the facts—leave each of them ill-
equipped to properly assess what is fair to their fellow New Brunswickers and Nigerians,
respectively. For eliminating effective MMR vaccination protocols (which exist not only to protect
those vaccinated, but to provide herd immunity for the vulnerable and immunocompromised)
especially without introducing some other kind of measure that would do at least as much to
address the public health concerns at comparable costs, is simply incompatible with the project of
living together on fair terms of cooperation for mutual advantage. Preserving an ineffective food
regulatory regime failing to prevent food fraud, is likewise incompatible with these goals. So even
if Pristine and Oba were each to prevail in persuading their peers, neither of them can succeed in
publicly justifying their favored exercises of power and authority. Moreover, their abnormal
epistemic situation, dispose them to other behaviors similarly consequential for social cooperation.
Therefore, like Aimy, Pristine and Oba, through no faults of their own, cannot be counted
reasonable in advocating for these positions.
Again, this pair of cases demonstrates that without a sufficient grasp of certain facts about
the external world one cannot count as reasonable, and that likewise one’s position on a particular
exercise e of power or authority cannot be reasonable without a sufficient grasp of the normatively
significant facts relevant to e. For even if Pristine and Oba are reasonable in certain respects—for

154
example, the way in which they rationally form the relevant beliefs and perform the relevant
actions—they are paradigmatically unreasonable in other respects of greater practical consequence
to the project of living together on fair terms of cooperation for mutual advantage. For
reasonableness is a matter how we actually interact with fellow members of our society, and not
merely of how we take ourselves to be interacting with them.
Furthermore, these cases illustrate why reasonableness-undermining luck is not a matter of
whether one is lucky to be in a cooperative environment, holding fixed their internal dispositions.
For even though Kendra, given her disposition to defer to conventional authority, is lucky to be in
her epistemically cooperative New Brunswick environment, rather than the uncooperative
environment in which Oba finds himself with the same disposition, reasonableness is a matter how
we interact with fellow members of our society. So the sense in which Kendra is lucky does not
undermine her reasonableness. Likewise, even though Wemimo, given his disposition to prioritize
considerations of purity, is lucky to be in his epistemically cooperative Nigerian environment,
rather than the uncooperative environment in which Pristine finds herself with the same
disposition, reasonableness is not a matter of how we would interact with those members of some
possible society of which we might have instead been a member.
To insist otherwise would require us to say that Kendra’s deference to conventional
authority and Wemimo’s purity-oriented values prevent them from being counted reasonable,
because of social realities that obtain in other societies of which they are not a part. But Kendra
and Wemimo are well-equipped for epistemic success in their actual social environments, and (in
combination with their non-epistemic manifestations of reasonableness) each can be relied upon,
as a competent, dependable, fellow participant in their respective societies, which is precisely what
the public reason liberal project requires from its evaluation of individual reasonableness. So

155
reasonableness-undermining luck is not a matter of whether one is lucky to be in a cooperative
environment, holding fixed one’s internal dispositions, but of whether one’s dispositions relevant
to social and political cooperation are lucky to hit their target, holding external conditions fixed.
7. Conclusion: Epistemological Externalism for Liberal Anti-
Perfectionists
As we learned in section 1 of this chapter, over the objections of thinkers like Klosko, Kelly &
McPherson, and Nussbaum, public reason liberals have overwhelmingly assumed or defended
epistemic notions of reasonableness and the publicity of reasons justifying the exercise of power
or authority in a society. Most anti-perfectionists are epistemological internalists about these
notions. However, in section 2, we surveyed the epistemologically externalist implications of some
public reason liberal claims. In particular, we looked at Wall (2010, 2014), who, embraces liberal
perfectionism and defends epistemological externalism about public justifications and
reasonableness, prioritizing the need to justify political decisions on the basis of good epistemic
reasons over other concerns. As I argued, prominent anti-perfectionist objections to externalism
are not decisive. However, in section 3, I argued that there is good reason to worry that strong
externalism threatens to collapse the standard of public reason into a truth standard, leaving no
room for public reason to do independent work. Therefore, taking the advice of Rawls, we looked
to the public political culture for an implicit standard of public reason to which we might refer,
and following Rini (2020), I argued that the ordinary standard of good faith debate entails neither
truth nor consent, and because it is implicit to the public political culture, its acceptance
presupposes no excessively controversial views.

156
Therefore, in section 4, taking inspiration from Klosko et. Al., I developed the
“Cooperation First” account of reasonableness, on which reasonableness is first and foremost a
matter of being a reliably fair cooperative participant in all matters of civic life. On this view,
successfully cooperative behavior determines who is reasonable, but in order to successfully
cooperate with other members of one’s society, reasonable agents must be sensitive to facts about
how the burdens and benefits of cooperation are distributed in one’s society, including social,
historical, political, or economic facts about their society. Thus, reasonableness is an
epistemologically externalist virtue, but I argue that this does not collapse the standard of public
reason into a truth standard, because in matters of public inquiry, where reasonableness is
manifested alongside the virtue of civility, the virtuous practice of public reasoning is governed
by the standards of good faith debate, which requires neither truth nor consent. Thus, I have
developed a Cooperation First account of reasonableness that is nonetheless epistemologically
rigorous, and in putting these ideas to the test, we learned in sections 5 and 6 that, whereas
epistemologically externalist excellences can ensure successful cooperation in public reasoning,
internalist excellences are too weak.

157
V. (Dis)solving the Idealization Dilemma for Public Reason Liberalism
It is sometimes said that liberalism’s “moral lodestar” is the conviction that the exercise of political
power or authority is legitimate only when it can be publicly justified.
203
Since the illegitimate
exercise of political power or authority is unjust, liberals think the justice of a political order also
hangs on its public justification. A system of government in which (at least the majority of the)
exercises of power and authority are legitimate is not supposed to be a big ask. Liberal justice is
not supposed to be something reserved for the perfectly just state in Platonic heaven. On the
contrary, as a modern political tradition, liberalism was first advanced to offer a practicable, stable
alternative to the many tried and failed utopian political ideals on the ash heap of history.
Therefore, liberals believe that, on some practical level, we are required to justify our favored
(prospective) exercises of power and authority to those subject to them, those in whose name they
are exercised, or both. Call this the public justification requirement.
As public reason liberals point out, there is a problem. People disagree—often, very
deeply—about what constitutes a satisfactory justification. There are inevitably some people, with
irrational, misinformed, repugnant, or just very idiosyncratic views, unwilling to accept plausible
justifications for certain exercises of authority or power, or too ready to accept the false, the
incoherent, the nonsensical, or worse. Therefore, recognizing that unanimously acceptable
justifications are often undesirable or practically impossible to secure, public reason theorists
abstract away from the most irrational, repugnant, and idiosyncratic elements of liberal-democratic
society to theorize about what would be unanimously acceptable to some suitably idealized subset

203
See Quong 2011, 2–3; see Macedo 1990, 78; Waldron 1993, 36–37.

158
of society. Public reason liberals distinctively argue that the public justification requirement is
fulfilled so long as relevant exercises of power or authority are justified (or justifiable) by appeal
to reasons or grounds unanimously acceptable to this idealized constituency. Agreement among
members of this constituency, in turn, determines public reason’s content—the set of
considerations that count as suitably public for justifying the exercise of power and authority.
But in free societies, even intelligent, morally decent people still disagree about
inescapably important moral, social, and political questions, and these widespread disagreements
persist, in part, because so many people have good reasons for thinking, valuing, believing, and
acting as they do. Nevertheless, public reason liberals theorize about what considerations these
more level-headed members could unanimously accept for exercises of authority or power. The
idealized members of this subset are referred to as “the reasonable.” As used by public reason
theorists, ‘reasonable’ and its cognates are technical terms predicated of these individuals, their
moral, political, and religious views, any communities they organize around such views, and their
disagreements over the same (among other things
204
).
Anti-perfectionist liberal theorists insist that the liberal state, liberal theory itself, or both
should maintain a strict neutrality between the compelling but incompatible moral, political, and
religious views that abound among the reasonable. So, the idealization of the reasonable is
supposed to leave room for their ideological diversity, rather than simply abstracting away from
all deep disagreements in society. In this way, the mechanism of idealization is supposed to
formally recognize or vindicate a pre-theoretic intuition that even sensible people of goodwill

204
For an exhaustive list of predications in Rawls 1993 with references, see Wenar 1995, 34, fn. 3.

159
inevitably disagree about lots of socially and politically significant issues: what public reason
liberals call the “fact of reasonable pluralism.”
Therefore, in idealizing the reasonable, public reason liberals want to eliminate the
intuitively illiberal from consideration, while preserving as much diversity of opinion as possible.
But there is a tension between realizing these two goals, often pressed as a dilemma, and best
understood as an extensional challenge for public reason’s constituency: how to idealize the
reasonable enough to eliminate the intuitively illiberal without over-idealizing and making this
constituency too exclusive? In this concluding chapter, I show that the Cooperation First account
of reasonableness, developed in chapter 4, is well positioned to dissolve this idealization dilemma
for public reason liberalism.
1. The Idealization Dilemma for Public Reason Liberalism
Let us begin by being more precise about this dilemma.

1.1 The Under-Idealization Horn
On the one hand, public reason theorists can idealize minimally to limit the number of people
excluded from the constituency of the reasonable. In that case, one stipulates a few thin, largely
uncontroversial, normative requirements, like requiring the reasonable to be practically rational.
The main virtue of minimal idealization is that it ensures public reason’s constituency is a big tent,
with room for lots of different worldviews and perspectives, maximizing the amount of ideological
diversity that gets deemed reasonable.
Moreover, by making only minimal normative demands on the reasonable, it is easier for
all reasonable parties to agree on these normative demands, and so agree about who counts as

160
reasonable and perhaps who counts as unreasonable too. This makes the public justification
requirement itself easier to publicly justify and membership in public reason’s constituency more
determinate. This is a very appealing prospect for public reason theorists. Having it be transparent
who the reasonable are (and who they are not) would avoid standoffs where at least one disagreeing
party wrongly takes the other to be unreasonable (or takes their own position to be reasonable
when it is not), thereby preventing unreasonable disagreement from interfering in the process of
public justification.
The extensional problem, however, is that too little idealization leaves public reason’s
constituency so capacious and so ideologically diverse as to make it practically impossible to find
any suitably public justifications for the exercises of power and authority that are essential to
maintaining an orderly and functional society. In that case, we either accept anarchy in practice,
or illegitimately exercise power and authority to maintain governance, while accepting
philosophical anarchism (i.e., skepticism about the legitimacy of state authority). But as I said at
the outset, a system of government in which (at least the majority of the) exercises of power and
authority are legitimate is not supposed to be too big of an ask. Liberal justice is not supposed to
be reserved for the perfectly just state in Platonic heaven, and the idealization of the reasonable is
supposed to obviate the need to be (even philosophical) anarchists. So to the extent that under-
idealization threatens some version of anarchism, it undercuts the purpose of appealing to an

161
idealized reasonable constituency in the first place. Thus, insufficient idealization of the reasonable
makes the public justification requirement too costly for most liberals to accept.
205

1.2 The Over-Idealization Horn
On the other hand, public reason theorists can idealize more to limit membership in the
constituency of the reasonable. After all, the smaller the constituency, the easier it is for its
members to come to determinate agreements about what counts as sufficient justification for
various exercises of power and authority. In this case, one must stipulate thicker, more
controversial, normative requirements, like requiring that the reasonable be perfectly rational,
espouse some specified political values, or possess certain information. Greater idealization can
ensure that intuitively illiberal persons and their views are barred from the ranks of the reasonable,
forestalling the possibility of publicly justifying things like racial segregation. The other main
virtue of increased idealization is that it lowers the cost of accepting the public justification
requirement, by making the public justification of power practically possible, thereby rescuing
public reason liberalism from anarchism.
The extensional pitfall, however, is that too much idealization of the reasonable makes
membership in public reason’s constituency too exclusive. A more restrictive mechanism of

205
A large constituency of the reasonable might get lucky and manage to agree enough to publicly justify exercises of
power sufficient to avoid anarchy, but public reason liberalism is not supposed to be a theory of lucky political
legitimacy. Moreover, too little idealization may also fail to bar intuitively illiberal persons from the ranks of the
reasonable. Racists, sexists, and other kinds of bigots might, for example, count as being practically rational. In that
case, a large constituency might get very unlucky when there are too many persuasive bigots in its midst. If this
element manages to sway enough people to agree with them, they might succeed in publicly justifying intuitively
illiberal exercises of power, like laws mandating racial segregation of public services. So insufficient idealization of
the reasonable threatens public reason liberalism with the prospect of having to deem paradigmatically illiberal
exercises of power politically legitimate. For all these reasons, too little idealization leaves public reason’s
constituency too ideologically diverse, and perhaps too incompetent, morally or otherwise, to do its job.

162
idealization formally recognizes fewer people as reasonable, and therefore ignores more
perspectives on social and political questions. This exclusivity undermines public reason
liberalism’s claim to respecting the fact of reasonable pluralism. If, for example, all and only a few
paragons of “progressive” moral and political sensibilities can count as reasonable, then most
people are ruled unreasonable, and in a way that seems at least suspicious and perhaps viciously
circular or morally arbitrary. Excluding these people from public reason’s constituency does not
mean that they have no rights in a liberal democracy, but it does mean we can legitimately ignore
many of the reasons they advance on behalf of their favored policies and the objections they raise
against those policies they disfavor. In this way, the set of considerations on which legitimate
power and authority can be justified gets narrowly circumscribed to rule out “non-progressive”
sensibilities by fiat. The resulting theory of political legitimacy is no longer recognizable as one
designed to respect the plurality of compelling worldviews one finds represented in a free society.
Furthermore, by making more controversial normative demands on the reasonable, it
becomes more difficult for all reasonable parties to agree on these normative demands and agree
about who counts as reasonable and unreasonable. This makes the public justification requirement
itself more difficult to publicly justify and membership within public reason’s constituency more
indeterminate.

1.3 Why not Lower the Bar?
Several moves can make it easier for the reasonable constituency, at any level of idealization, to
reach agreement about the uses of political power and authority. Each move makes it easier, in
some respect to satisfy the public justification requirement, reducing the threats of anarchism and
thereby blunting the force of the under-idealization horn of the dilemma.

163
Firstly, the theorist might argue to limit the scope of what entities stand in need of public
justification. Jonathan Quong (2004), for example, argues the public justification requirement
applies chiefly to laws. John Rawls (1991) goes much further, narrowing public reason’s scope to
just the constitutional essentials and matters of basic justice, “but not in general [to] all the
questions to be settled by the legislature within a constitutional framework” (91), let alone less
overtly political, but no less consequential exercises of power, like the everyday enforcement of
social norms.
If fewer kinds of things stand in need of public justification, then there are fewer kinds of
debates where irresolvable disagreements can interfere with fulfilling the public justification
requirement. Correspondingly, it is also more likely that we can secure agreement when there are
fewer points of contention. In addition, the more abstract these few remaining points of contention
are, the less likely it is we can tell in advance which one of us will be disadvantaged by any policy
we might propose, and so the more likely that we can agree about the rules of fair play. So whatever
level of idealization the theorist chooses to defend, it is always easier for the reasonable to fulfill
a public justification requirement understood to have narrow scope.
Similarly, the public reason theorist might try to restrict the site of public reason’s
application. For example, Rawls limits its application to what he calls the “public political
forum”—that is, the discourse of judges, government officials, candidates for public office and
their campaign managers, in their capacities as such.
206
As with limiting the scope, restricting
public reason’s application leaves fewer kinds of discourse involved in the justificatory process,

206
See Rawls 1999, 575.

164
and hence fewer occasions for irresolvable disagreements to interfere with fulfilling the public
justification requirement.
The theorist might also opt for a shallow account of public reason’s justificatory depth. All
public reason liberals agree that, in attempting to publicly justify some exercise of power or
authority, reasonable people need not reach an agreement that extends all the way down to first
principles, because first principles are paradigmatically subject to reasonable disagreements. But
there is a question as to how deep agreement among the reasonable has to go in order to be
attainable, while remaining capable of robustly sustaining social cooperation.
Some, like Rawls, argue that only objects of overlapping consensus among the reasonable,
providing common ground justifications for exercises of power or authority, can ensure the
stability needed for social cooperation. But others, like Kevin Vallier, argue that it is enough for
the reasonable to converge on accepting some exercise of power or authority, regardless of whether
there is any common ground justification they all accept. Again, whatever one thinks about the
right level of idealization for public reason’s constituency, the shallower one’s account of this
justificatory structure, the less that reasonable disagreements about deeper questions, not directly
related to a particular exercise of power, are relevant to its public justification. Again, this leaves
fewer disputes where irresolvable disagreements can interfere with fulfilling the public
justification requirement.
Of course, the broader public reason’s scope, the wider its site of application, and the
deeper its required justificatory depth, the more demanding the public justification requirement
becomes, and the more robust the agreement secured among the reasonable when public
justification is achieved. But if one can be convinced that either (i) as matter of empirical fact,
sufficiently stable agreement among the reasonable can be achieved with less, or (ii) a broad-scope,

165
wide-site, and/or deep-agreement version of the public justification requirement is just too
demanding, then the public reason liberal can help themselves to any of the above theoretical
moves, because the narrower public reason’s scope, the more restricted its site of application, and
the shallower its required justificatory depth, the easier it is for the reasonable, at any level of
idealization, to satisfy the public justification requirement.
As a result, responses to the idealization dilemma often trade on one or more of these three
moves. In contrast, the account I offer in this chapter does not rely on any of these moves. On the
contrary, my account assumes a broad-scope and wide-site view of public reason’s application to
social and political life, and my account does not depend on a shallow account of public reason’s
justificatory depth.
207
After all, why should public reason liberalism rest on some gerrymandered
notion of what it means to justify the use of power or authority? If the moral lodestar really is the
conviction that the exercise of political power or authority is legitimate only when it can be publicly
justified, why should we restrict our concerns to matters of constitutional essentials and basic
justice or the dialogue of justices, politicians, political candidates? So, my notion of the public is
much closer to the popular sense of that term, and therefore takes a path of greater resistance,

207
Merely converging on consequential exercises of authority or power will often be like agreeing to the text of a
constitution without being able to agree on how to interpret or amend it. In the case of a constitution, we clearly need
more common ground than just agreeing to the wording of the document. Sometimes we may not be capable of coming
to any deeper agreement. But in such cases we often agree to a set of fair procedures for settling further disputes that
remain when we have reached bedrock disagreement (e.g., procedures for amending a constitution, electing officials,
making appointments, passing laws in a divided legislature, or resolving disputes through the civil law). In so doing,
we rely on procedural justice to generate the wider agreement needed to sustain stable cooperation, when deeper
agreement is not possible to secure. For when these procedures generate verdicts in a way that is seen to be fair and
impartial, people accept their legitimacy, making them suitably public reasons on which to base further social
coordination. So, the procedural approach to supplementing shallow agreement with wider agreement is available
when needed, but where shallow agreement is adequate to sustain stable cooperation, I offer no objection to it here.

166
putting all the theoretical explanatory pressure on our idealization of the reasonable. Nevertheless,
I claim it can still avoid the idealization dilemma.

1.4 Desiderata for Avoiding the Dilemma
However, to avoid the dilemma, it is not enough simply to land somewhere between the extremes
of both horns. A solution to or dissolution of the dilemma needs to incorporate the right kinds of
motivations. Therefore, (1) we need to see that the constituency of the reasonable remains a big
tent, leaving lots of room for meaningful disagreements about religion, morality, and politics; at
the same time, (2) intuitively illiberal persons and their views should be barred from the ranks of
the reasonable; and compared with rival solutions or strategies, (3) our favoured (dis)solution
should make it no harder (and ideally easier) to agree about who counts as reasonable, and perhaps
who counts as unreasonable too, so that (i) the public justification requirement is itself no harder
(and ideally easier) to publicly justify, and (ii) membership in public reason’s constituency is no
less (and ideally more) determinate; (4) our favoured (dis)solution should place liberal justice
closer within reach, so that we can resist (philosophical) anarchism; and (5) we want to avoid
relying on a weak or gerrymandered notion of what it means to justify the use of power or
authority.
2. Toward a Dissolution of the Idealization Dilemma
Why, one might ask, has the idealization dilemma proved so difficult to (dis)solve? Is it because
public reason liberalism in general, or its notion of reasonableness, are inescapably self-defeating?
No, but there is a pervasive epistemological assumption made by public reason liberal theorists,
which makes the dilemma more difficult to solve than it needs to be. What gets mostly taken for

167
granted about the nature of reasonableness is that it is an epistemologically internalistic notion.
208

Epistemological internalism about reasonableness entails that an agent’s attitudes are counted
reasonable purely in virtue of states internal to the agent. So, if we take two individuals A and B
who are internal duplicates, whenever A’s attitudes are reasonable, B’s attitudes are too, even if B
is a brain-in-a-vat that is being systematically misled inside Nozick’s experience machine.
Internalism about reasonableness, therefore, requires all the public reason theorist’s idealization to
be baked into agents’ heads, so to speak. But once we force all idealization inside agents’ heads,
only very demanding and highly ideal constraints are able prevent them from being too credulous
of far-fetched possibilities, and hence too susceptible to being misinformed or misled. Yet, unless
the reasonable can be guarded from excessive credulity and gullibility, they will be unreliable
contributors to and judges of public justification. So there is good reason to want some explicitly
externalist element in our idealization of the reasonable. Unfortunately, the costs of adopting
epistemological internalism about reasonableness have been largely overlooked.

2.1 The Road Taken: Internalistic Reasonableness + Information
In addition to the standard internalistic conception of reasonableness, discussed in section 2.2.3 of
Chapter 4, Vallier (2014) distinguishes another standard parameter of idealization common to
public reason liberalism, which he calls the information parameter. The information parameter is
the only explicitly externalist feature of the public reason liberal’s standard theoretical idealization,

208
See Chapter 4, section 1 for some representative examples of epistemologically internalist accounts of
reasonableness. See section 2 for a more exhaustive survey of the few exemplars of epistemological externalism about
reasonableness among public reason liberals, as well as objections to externalism.

168
and the very fact of its divorce from the reasonableness parameter tells just how deeply the
internalist assumption about reasonableness runs.
The information parameter determines the agent’s information set or total evidential base.
As Vallier explains, public reason liberal theorists add information to the belief-value sets of
idealized citizens, to avoid holding public justification captive to ignorance, while subtracting
partializing information that could bias public justification. Parties to Rawls’s original position
behind the veil of ignorance, for example, “know the general laws of economics and psychology,
but he denies citizens knowledge of their conceptions of the good and contingent characteristics”
(2014, 150). For convenience, Vallier calls any idealization that removes partializing information,
while adding general information, a full information standard. Full information is appealing to
public reason liberal theorists because it (a) avoids attributing (public) justifications based on
falsehoods to idealized agents, and (b) generates some determinacy. Moreover, Vallier says that if
all agents have the same accurate information, they will be more likely to reach agreement. But
while adding information rules out some wild priors, it leaves the unruled-out relative possibilities
exactly the same. And general laws of economics and psychology, whatever those might turn out
to be, will leave (even very wild) priors concerning most possibilities completely untouched.
However, this kind of general information idealization is inefficient from the point of view
of trying to solve the idealization dilemma. It has the effect of ruling less informed people out the
public reason liberalism’s constituency, but not because they are unreasonable. It also still leaves
“fully informed” agents all over the place doxastically, and thereby fails to ensure agreement even
among those reasonable agents who are not ruled out because of their lack of “full” information.
Moreover, although the standard of public reason or public justification is supposed to lie

169
somewhere between truth and consent, the information parameter appears to smuggle much truth
back into the account.

2.2 The Road Not Taken: Externalistic Reasonableness
The standard internalistic idealization of the reasonable cannot tell us what information the
reasonable have about the external world, because it is purely internalist. However, the
epistemologically externalist Cooperation First account, which I develop and defend in Chapter 4,
lets successfully cooperation in all matters of civic life tell us who the reasonable are, following
George Klosko (2000), Erin Kelly & McPherson (2001), and Martha Nussbaum (2011). However,
then it can ask, what information about their social milieu is this constituency tracking? That is the
information to which the virtue of reasonableness makes agents sensitive. Reasonableness makes
agents sensitive to certain information (it need not be all and only the truth), which proves useful
to them in acting reasonably. Then, in abiding by the standards of good faith argumentation when
engaged in public reasoning, the reasonable disseminate the information at their individual
disposals, dispute collectively over its significance and, under the right social and environmental
conditions, evaluate their reasons to arrive at public justifications for the exercise power and
authority, all without enforcing truth or consent as standards, because the standard of good faith
argumentation lies somewhere in between.
This idealization is elegant in that it ties the externalist epistemic parameter of idealization
directly to the central aim of reasonable agents. It, therefore, does not require us to front-load
reasonable agents with lots of general information. Moreover, people can succeed in being
reasonable even in societies where they do not have advanced social sciences that have discovered,
as yet undiscovered, laws of economics, psychology, and so forth. Since the social sciences

170
continue to be contested along political divides, this is good news for theorizing about political
legitimacy and justice in non-ideal societies. Therefore, let us stop focusing solely only on what is
happening inside agents’ heads and consider how well they are managing to track relevant features
of their social and political environments.

2.3 Getting there by the Road Not Yet Taken
To avoid the dilemma, it is not enough just to land somewhere between the extremes of both horns.
A solution to or dissolution of the dilemma needs to incorporate the right kinds of motivations. So,
let us consider whether the Cooperation First view is well positioned to satisfy the five desiderata
we identified earlier, in section 2.4 above.
Firstly, we need to see that the constituency of the reasonable remains a big tent, leaving
lots of room for meaningful disagreements about religion, morality, and politics, and the
Cooperation First view is purpose built to ignore lots of meaningful disagreements about religion,
morality, and politics, that might otherwise tempt us to deem someone unreasonable, focusing on
their ability to behave cooperatively. Secondly, intuitively illiberal persons and their views should
be barred from the ranks of the reasonable, and since we can rely on ordinary pre-theoretic
intuitions about who is successfully cooperating with their fellow citizens and who is not, we have
reason to be very confident that we can bar the illiberal from the ranks of the reasonable. Thirdly,
compared with rival solutions or strategies, our favoured (dis)solution should make it no harder
(and ideally easier) to agree about who counts as reasonable, and perhaps who counts as
unreasonable too, and focusing on our ordinary pre-theoretic intuitions about who is successfully
cooperating with their fellow citizens promises to make this easier. This should make (i) the public
justification requirement itself no harder to publicly justify. Moreover, since the Cooperation First

171
comes packaged with a public political cultural account of the standard of public justification,
drawing on ordinary pre-theoretic intuitions about what it is engage with others and debate in good
faith, this should make it easier to publicly justify this standard. This should also make (ii)
membership among the reasonable more determinate. Fourthly, our favoured (dis)solution should
place liberal justice closer within reach, so that we can resist (philosophical) anarchism, and I think
the Cooperation First account should give us confidence that liberal justice is not reserved for the
perfectly just state in Platonic heaven but can be realized in the world. Finally, we want to avoid
relying on a weak or gerrymandered notion of what it means to justify the use of power or
authority. But in grounding the standard of public justification through public reasoning in the
popular standard for debating others and dealing with them in good faith, implicit to our public
political culture, and in resorting to no special pleading about the scope, site, or structure of public
justification, I have deliberately sought to situate my notion of the public as close to the popular
sense of that term as possible.
If we have a purely internalist account of reasonableness, to get the determinacy we want
we have to bake all the idealization (moral or otherwise) into agents’ heads, but if we change our
idea of what it is to be reasonable, where what really matters is being a reliably fair cooperator in
all matters of civic life, then we do not have to worry so much about the specific mental contents
of peoples’ heads, because we are shifting our concern to what people can be relied upon to do.
This has the effect of reconfiguring the dilemma. The dilemma arises because we are focusing on
the wrong thing. The real issue is not how much mental content we have to idealize the reasonable
as having in common; what matters is whether we can stably cooperate with others on fair terms—
that is the real goldilocks point we are seeking, not some perfectly calibrated set of idealized

172
mental contents. Therefore, my account promises not so much to solve the dilemma, as much as
to dissolve it altogether.

173
BIBLIOGRAPHY
Aberdein, A. 2010. “Virtue in Argument.” Argumentation 24: 165–179. DOI: 10.1007/s10503-
009-9160-0.
Alfano, M. 2013. Character As Moral Fiction. Cambridge University Press.
Ancell, A. 2019. “The Fact of Unreasonable Pluralism.” Journal of the American Philosophical
Association 5 (4): 410–428. DOI: 10.1017/apa.2019.17.
Anderson, E. 2000. “Beyond Homo Economicus: New Developments in Theories of Social
Norms.” Philosophy & Public Affairs 29: 170–200.
Audi, R. 1991. “Religious Commitment and Secular Reason: A Reply to Professor Weithman.”
Philosophy & Public Affairs 20 (1): 66–76.
______. 1993. “The Place of Religious Argument in a Free and Democratic Society.” San Diego
Law Review 30: 677–702.
______. 2000. “Civic Virtue.” In Religious Commitment and Secular Reason (145–180).
Cambridge University Press.
Barry, B. 1995. Justice as Impartiality. Oxford: Clarendon Press.
Bejan, T. M. 2017. Mere Civility: Disagreement and the Limits of Toleration. Harvard University
Press.
Berger, K. P. & T. Arntz. 2016. “Good Faith as a ‘General Organising Principle’ of the Common
Law.” Arbitration International 32: 167–178. DOI: 10.1093/arbint/aiw001.
Boettcher, J. 2012. “The Moral Status of Public Reason.” Journal of Political Philosophy 20 (2):
156–177.
Bonjour, L. 1980. “Externalist Theories of Empirical Knowledge.” Midwest Studies in Philosophy
5: 53–73.

174
Brower, B. 1994. “The Limits of Public Reason.” Journal of Philosophy 91: 5–26.
Burtt, S. 1991. “Books in Review: LIBERAL VIRTUES: CITIZENSHIP, VIRTUE, AND
COMMUNITY IN LIBERAL CONSTITUTIONALISM by Stephen Macedo.” Political
Theory 19(4): 678–683. DOI: 10.1177/0090591791019004013.
Calhoun, C. 2000. “The Virtue of Civility.” Philosophy and Public Affairs 29 (3): 251–275.
Callan, E. 1999. “Liberal Virtue and Moral Enfeeblement.” In David Carr and J. W. Steutel (Eds.).
Virtue Ethics and Moral Education (189–201). Routledge.
Cohen, D. H. 2005. “Arguments that Backfire.” OSSA Conference Archive 8: 58–65.
https://scholar.uwindsor.ca/ossaarchive/OSSA6/papers/8.
______. 2019. “Argumentative Virtues as Conduits for Reason’s Causal Efficacy: Why the
Practice of Giving Reasons Requires that We Practice Hearing Reasons.” Topoi 38: 711–
718. DOI: 10.1007/s11245-015-9364-x.
Cohen, J. 2009. “Truth and Public Reason.” Philosophy & Public Affairs 37(1): 2–42.
Cohen S. & J. Comesaña. “Rationality and Truth.” Forthcoming in Dutant, J. and F. Dorsch, (Eds.).
The New Evil Demon: New Essays on Knowledge, Justification and Rationality. Oxford
University Press.
Costa, M. V. 2004. “Political Liberalism and the Complexity of Civic Virtue.” Southern Journal
of Philosophy 42 (2): 149–170. DOI: 10.1111/j.2041-6962.2004.tb00994.x.
Curtis, W. M. 2016. “Rorty as Virtue Liberal.” Contemporary Pragmatism 13 (4): 400–419. DOI:
10.1163/18758185-01304004.
Dagger, R. 1997. Civic Virtues: Rights, Citizenship, and Republican Liberalism. Oxford
University Press.

175
______. 2013. “Citizenship as Fairness: John Rawls’s Conception of Civic Virtue.” In Jon Mandle
and David A. Reidy (Eds.). A Companion to Rawls (297–311). Wiley Blackwell.
Danziger, S., J. Levav, & L. Avnaim-Pesso. 2011. “Extraneous Factors in Judicial Decisions.”
Proceedings of the National Academy of Sciences 108: 6889–6892.
Eberle, C. 2002. Religious Convictions in Liberal Politics. Cambridge University Press.
Enoch, D. 2014. “Authority and Reason-Giving.” Philosophy and Phenomenological Research 89
(2): 296–332.
_____. 2015. “Against Public Reason.” Oxford Studies in Political Philosophy 1: 112–142.
_____. 2017. “Political Philosophy and Epistemology: The Case of Public Reason.” Oxford
Studies in Political Philosophy 3: 132–165. DOI: 10.1093/oso/9780198801221.003. 0007.
Estlund, D. 2008. “Epistemic Proceduralism and Democratic Authority.” In Raf Geenens &
Ronald Tinnevelt (Eds.). Does Truth Matter? Democracy and Public Space. Springer.
Foran, C. 4 January 2017. “The Canada experiment: is this the world's first 'postnational' country?”
The Guardian. https://www.theguardian.com/world/2017/jan/04/the-canada-experiment-
is-this-the-worlds-first-postnational-country.
Fuerstein, M. 2013. “Epistemic Trust and Liberal Justification.” Journal of Political Philosophy
21 (2): 179–199. DOI: 10.1111/j.1467-9760.2012.00415.x.
Friedman, J. 2020. “The Epistemic and the Zetetic.” Philosophical Review 129 (4): 501–536. DOI:
10.1215/ 00318108-8540918.
Galston, W. 1991. Liberal Purposes: Goods, Virtues, and Diversity in the Liberal State. Cambridge
University Press.

176
Gaus, G. 2003. “Religious Convictions in Liberal Politics, Christopher Eberle (Review).” Notre
Dame Philosophical Reviews. https://ndpr.nd.edu/reviews/religious-convictions-in-liberal
-politics/.
_____. 2011. The Order of Public Reason. Cambridge: Cambridge University Press.
Godden, D. 2016. “On the Priority of Agent-Based Argumentative Norms.” Topoi 35: 345–357.
DOI: 10.1007/s11245-014-9296-x.
Gregory, J. 2004. “The Culture of Liberalism and the Virtue of ‘Balance’.” European Journal of
Political Theory 13 (1): 78–94. DOI: 10.1177/1474885112473718.
Harman, G. 2000. “The Nonexistence of Character Traits.” Proceedings of the Aristotelian Society
100: 223–226.
Hartley, C. & L. Watson. 2014. “Virtue in Political Thought: On Civic Virtue in Political
Liberalism.” In Kevin Timpe and Craig A. Boyd (Eds.) Virtues and Their Vices (415–432).
DOI: 10.1093/acprof:oso/9780199645541.001.0001.
Horton, J. 1996. “Toleration as a Virtue.” In David Heyd (Ed.). Toleration: An Elusive Virtue (28–
43). Princeton University Press.
Huang, C. and C. Cho. 30 October 2017. “Ten Facts You Should Know About the Federal Estate
Tax,” cbpp.org. Washington, DC: Center on Budget and Policy Priorities.
www.cbpp.org/research/federal-tax/ ten-facts-you-should-know-about-the-federal-estate-
tax.
Hurka, T. 2006. “Virtuous Act, Virtuous Dispositions.” Analysis 66 (1): 69–76.
Hurley, P. 2 January 2016. “Philip Pettit’s The Robust Demands of the Good: Ethics with
Attachment, Virtue, and Respect.” Notre Dame Philosophical Reviews. Available at:

177
https://ndpr.nd.edu/news/the-robust-demands-of-the-good-ethics-with-attachment-virtue-
and-respect/.
Jain, A. et Al. 2015. “Autism Occurrence by MMR Vaccine Status Among US Children With
Older Siblings With and Without Autism.” Obstetrical & Gynecological Survey 70(8):
479–480.
Jeffrey, R. 2004. Subjective Probability: The Real Thing. Cambridge: Cambridge University Press.
Joseph, P. E. 2020. The Sword and the Shield: The Revolutionary Lives of Malcolm X and Martin
Luther King Jr. Basic Books.
Kelly, E. & L. McPherson. 2001. “On Tolerating the Unreasonable.” Journal of Political
Philosophy 9 (1): 38–55.
King, A. 15 November 2016. “Washington High Court Hears Case Of Florist Who Refused To
Serve Gay Wedding.” NPR. https://www.npr.org/2016/11/15/502111408/washington-
state-court-case-religious-liberty-versus-anti-discrimination.
Kloppenberg, J. T. 1998. The Virtues of Liberalism. Oxford University Press.
Klosko, G. 2000. Democratic Procedures and Liberal Consensus. Oxford University Press.
LeBar, M. 2009. “Virtue Ethics and Deontic Constraints.” Ethics 119: 642–671.
______. 2013. “Virtue and politics.” In Daniel C. Russell (Ed.). The Cambridge Companion to
Virtue Ethics (265–289). Cambridge University Press.
Leece, S. 2003. “Contractualism and Liberal Neutrality: A Defence.” Political Studies 51: 524–
541.
Lehrer, K. & S. Cohen. 1983. “Justification, Truth, and Coherence”. Synthese 55: 191–207.
Leland, R. J. & H. van Wietmarschen. 2012. “Reasonableness, Intellectual Modesty, and
Reciprocity in Political Justification” Ethics 122: 721–747.

178
_____. 2017. “Political Liberalism and Political Community.” Journal of Moral Philosophy 14
(2): 142–167.
Liveriero, F. 2015. “The Epistemic Dimension of Reasonableness.” Philosophy & Social Criticism
41: 517–35.
_____. 2020. “Reasonableness as a Virtue of Citizenship and the Opacity Respect Requirement.”
Philosophy and Social Criticism 46 (8): 901–921. DOI: 10.1177/ 0191453720903492.
Macedo, S. 1990. Liberal Virtues: Citizenship, Virtue, and Community in Liberal
Constitutionalism. Oxford: Clarendon Press.
McKinnon, C. 2002. Liberalism and the Defence of Political Constructivism. New York: Palgrave
MacMillan.
McTernan, E. 2014. “How to Make Citizens Behave: Social Psychology, Liberal Virtues, and
Social Norms.” Journal of Political Philosophy 22 (1): 84–104.
Meldgaard, M.K., et. Al. 2002. “A population-based study of measles, mumps, and rubella
vaccination and autism.” The New England Journal of Medicine 347(19): 1477–1482.
Merricks, T. 1995. “Warrant Entails Truth.” Philosophy and Phenomenological Research 55 (4):
841–855. DOI: 10.2307/2108335.
⸻. 1997. “More on Warrant’s Entailing Truth.” Philosophy and Phenomenological Research 57
(3): 627–631. DOI: 10.2307/2953754.
Metz, T. 2012. “Ethics in Africa and in Aristotle: Some Points of Contrast.” Phronimon 13 (2):
99–117.
Millar, A. 1991. Reasons and Experience. Oxford: Clarendon Press.
Moon, A. 2012. “Warrant Does Entail Truth.” Synthese 184 (3): 287–297. DOI: 10.1007/s11229-
010-9815-2.

179
Mrozek-Budzyn, D., A. Kiełtyka, & R. Majewska. 2010. “Lack of association between measles-
mumps-rubella vaccination and autism in children: a case-control study.” The Pediatric
Infectious Disease Journal 29(5): 397–400.
Nagel, T. 1987. “Moral Conflict and Political Legitimacy.” Philosophy & Public Affairs 16: 215–
40.
Nguyen, C. T. 2020. “Echo Chambers and Epistemic Bubbles.” Episteme 17 (2): 141–161. DOI:
10.1017/epi.2018.32.
Nwuneli, N. 10 March 2018. “Fake processed food is becoming an epidemic in African urban
life.” Quartz (Quartz Media). Available at: https://qz.com/1226112/fake-food-or-fraud-
food-in- nigeria-kenya-and-other-african-countries/.
Nussbaum, M. C. 2006. Hiding from Humanity. Princeton University Press.
_____. 2011. “Perfectionist Liberalism and Political Liberalism.” Philosophy and Public Affairs
39 (1): 3–45. DOI: 10.1111/j.1088-4963.2011.01200.x.
Peltola, H. 1998. “No evidence for measles, mumps, and rubella vaccine-associated inflammatory
bowel disease or autism in a 14-year prospective study.” Lancet 351(9112): 1327–1328.
Peter, F. 2013. “Epistemic Foundations of Political Liberalism.” Journal of Moral Philosophy 10:
598–620.
_____. 2019. “Political Legitimacy Under Epistemic Constraints: Why Public Reasons Matter.”
In Jack Knight and Melissa Schwartzberg (Eds.). Political Legitimacy: NOMOS LXI. DOI:
10.18574/nyu/9781479888696.001.0001.
Pettit, P. 1997. Republicanism: A Theory of Freedom and Government. Clarendon Press.
_____. 2015. The Robust Demands of the Good: Ethics with Attachment, Virtue, and Respect.
Oxford University Press.

180
Podgorski, A. 2016. “Dynamic Permissivism” Philosophical Studies 173(7): 1923–1939.
Poston, T. “Internalism and Externalism in Epistemology.” In The Internet Encyclopedia of
Philosophy. ISSN 2161-0002. http://www.iep.utm.edu/int-ext/.
Purcell, L. S. 2017. “Eudaimonia and Neltiliztli: Aristotle and the Aztecs on the Good Life.” APA
Newsletter on Hispanic/Latino Issues in Philosophy 16 (2): 10–21.
Quinn, P. 1997. “Political Liberalisms and their Exclusions of the Religious.” In Paul J. Weithman
(Ed.). Religion and Contemporary Liberalism (138–61). Notre Dame University Press.
Quong, J. 2004. “The Scope of Public Reason.” Political Studies 52 (2): 233–250.
_____. 2011. Liberalism Without Perfection. Oxford: Oxford University Press.
_____. 2013. “Public Reason.” In Stanford Encyclopedia of Philosophy (Summer 2013 Edition).
Ed. Edward N. Zalta. Available at: https://plato.stanford.edu/entries/public- reason/.
Rawls, J. 1971. A Theory of Justice: Original Edition. Belknap Press of Harvard University Press.
_____. 1993. Political Liberalism. Columbia University Press.
_____. 1999. A Theory of Justice: Revised Edition. Belknap Press of Harvard University Press.
_____. 2001. Justice as Fairness: A Restatement. Cambridge: Harvard University Press.
_____. 2005; 1997. “The Idea of Public Reason Revisited.” In Political Liberalism, Expanded
Edition. Columbia University Press.
Raz, J. 1986. The Morality of Freedom. Oxford University Press.
_____. 1990. “Facing Diversity: The Case of Epistemic Abstinence” Philosophy & Public Affairs
19(1): 3–46.
Rini, R. 2018. “Abortion, Ultrasound, and Moral Persuasion.” Philosophers’ Imprint 18 (6): 1–20.
http://hdl.handle.net/2027/spo.3521354.0018.006.

181
_____. October 26, 2020. “Trolling, Dunking, and Persuading: What is Good Faith Debate in the
Social Media Era?” Inaugural Lecture for the Conceptual Foundations of Conflict Project.
University of Southern California. https://www.philosophy.academy/conflict-
blog/trolling-dunking-and-persuading; https://youtu.be/S6DR4BusQHc.
Rysiew, P. 2016. “Naturalism in Epistemology.” In Stanford Encyclopedia of Philosophy (Winter
2016 Edition). Ed. Edward N. Zalta, forthcoming. Available at:
plato.stanford.edu/archives/win2016/entries/epistemology-naturalized.
Schoenfield, M. 2014. “Permission to Believe: Why Permissivism Is True and What It Tells Us
About Irrelevant Influences on Belief,” Noûs 48(2): 193–218.
Shklar, J. N. 1984. Ordinary Vices. Belknap Press of Harvard University Press.
Silva, P. 2018. “Explaining Enkratic Asymmetries: Knowledge-First Style.” Philosophical Studies
175(11): 2907–2930.
Smithies, D. 2015. “Ideal Rationality and Logical Omniscience.” Synthese 192: 2769–2793.
Schwartzman, M. 2011. “The Sincerity of Public Reason.” The Journal of Political Philosophy
19(4): 375–398.
_____. 2012. “The Ethics of Reasoning from Conjecture.” Journal of Moral Philosophy 9: 521–
544.
Sosa, E. 2007. A Virtue Epistemology: Apt Belief and Reflective Knowledge (Volume I). Oxford
University Press.
Staffel, J. 2015. “Measuring the Overall Incoherence of Credence Functions.” Synthese 192(5):
1467–1493.
Stapel, D. A. & S. Lindenberg. 2011. “Coping with Chaos: How Disordered Contexts Promote
Stereotyping and Discrimination.” Science 332: 251–253.

182
Vallier, K. 2014. Liberal Politics and Public Faith: Beyond Separation. Routledge.
van Wietmarschen, H. 2018. “Reasonable Citizens and Epistemic Peers: A Skeptical Problem for
Political Liberalism.” Journal of Political Philosophy 26 (4): 486–507. DOI: 10.1111/
jopp.12152.
Waldron, J. 1993. “Theoretical Foundations of Liberalism.” In Liberal Rights: Collected Papers
1981–1991. Cambridge: Cambridge University Press.
Wall, S. 1996. Liberalism, Perfectionism and Restraint. Oxford: Oxford University Press.
_____. 2002. “Is Public Justification Self-Defeating,” American Philosophical Quarterly 39(4):
385–94.
_____. 2010. “On Justificatory Liberalism.” Politics, Philosophy & Economics 9(2): 123–149.
DOI: 10.1177/1470594X09345677
_____. 2014. “Perfectionism, Reasonableness, and Respect.” Political Theory 42(4): 468–489.
DOI: 10.1177/0090591714531493.
_____. 2019. “Perfectionism in Moral and Political Philosophy.” In Edward N. Zalta (Ed.). The
Stanford Encyclopedia of Philosophy (Summer 2019 Edition). https://plato.stanford.edu/
archives/sum2019/entries/perfectionism-moral/.
Wedgwood, R. 2014. “Rationality as a Virtue.” Analytic Philosophy 55(4): 319–338.
Weinstock, D. M. 2006. “A Neutral Conception of Reasonableness?” Episteme 3 (3): 234–247.
Weithman, P. J. 2002. Religion and the Obligations of Citizenship. Cambridge University Press.
_____. 2010. Why Political Liberalism?: On John Rawls’s Political Turn. Oxford: Oxford
University Press.
Wenar, L. 1995. “Political Liberalism: An Internal Critique,” Ethics 106(1): 32–62.

183
Williams, B. 1996. “Toleration: An Impossible Virtue?” In David Heyd (Ed.). Toleration: An
Elusive Virtue (18–27). Princeton University Press.
Williamson, T. 2013. “Gettier Cases in Epistemic Logic.” Inquiry 56(1): 1–14.
_____. 2017. “Ambiguous Rationality.” Episteme 14(3): 263–274.
Wilson, E. E. & L. Denis. 2018. “Kant and Hume on Morality.” In Edward N. Zalta (Ed.). The
Stanford Encyclopedia of Philosophy (Summer 2018 Edition). https://plato.stanford.edu/
archives/sum2018/entries/kant-hume-morality/.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Reasoning with degrees of belief

PDF

Public justification beyond legitimacy

PDF

Contrastive reasons

PDF

Rationality and the primacy of the occurrent

PDF

A deontological explanation of accessibilism

PDF

Responding to harm

PDF

Reasons, obligations, and the structure of good reasoning

PDF

Toward a more perfect liberalism: perfectionism in Kantian political philosophy

PDF

Reasoning with uncertainty and epistemic modals

PDF

Units of agency in ethics

PDF

A perceptual model of evaluative knowledge

PDF

Aggregating complaints

PDF

Getting our act(s) together: the normativity of diachronic choice

PDF

Mistaken defense and normative conventions

PDF

Process-oriented rationality

PDF

On intimacy: a philosophical account

PDF

Feeling good and living well: on the nature of pleasure and its role in well-being

PDF

Capturing moods: a philosophical exploration

PDF

Aggregating happiness: seeking and identifying a single plausible unifying theory

PDF

Beliefs that wrong

Asset Metadata

Creator Ashfield, Michael Douglas (author)

Core Title The virtue of reasonableness: on the normative epistemic implications of public reason liberalism

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Philosophy

Degree Conferral Date 2021-12

Publication Date 09/24/2021

Defense Date 08/19/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag anti-perfectionism,epistemology,good faith,idealization dilemma,liberalism,OAI-PMH Harvest,public justification,public justifications,public reason,public reasoning,public reasons,reasonable disagreement,Reasonableness,virtue,virtue externalism,virtue reliabilism

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wedgwood, Ralph (committee chair), Albertson, David (committee member), Quong, Jonathan (committee member), Schroeder, Mark (committee member)

Creator Email ashfield@usc.edu,mdashfield@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15925685

Unique identifier UC15925685

Legacy Identifier etd-AshfieldMi-10100

Document Type Dissertation

Format application/pdf (imt)

Rights Ashfield, Michael Douglas

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu