(Routledge Studies in Contemporary Philosophy) Casey Rebecca Johnson (Ed.) - Voicing Dissent - The Ethics and Epistemology of Making Disagreement Public-Routledge (2018)

Voicing Dissent
Disagreement is, for better or worse, pervasive in our society. Not only
do we form beliefs that differ from those around us, but increasingly we
have platforms and opportunities to voice those disagreements and make
them public. In light of the public nature of many of our most important
disagreements, a key question emerges: how does public disagreement affect
what we know?
This volume collects original essays from a number of prominent scholars—
including Catherine Z. Elgin, Sanford C. Goldberg, Jennifer Lackey, Michael
Patrick Lynch, and Duncan Pritchard, among others—to address this question
in its diverse forms. The book is organized by thematic sections in which
individual chapters address the epistemic, ethical, and political dimensions
of dissent. The individual contributions address important issues such as
the value of disagreement, the nature of conversational disagreement, when
dissent is epistemically rational, when one is obligated to voice disagreement
or to object, the relation of silence and resistance to dissent, and when
political dissent is justified. Voicing Dissent offers a new approach to the
study of disagreement that will appeal to social epistemologists and ethicists
interested in this growing area of epistemology.
Casey Rebecca Johnson is an assistant professor of philosophy in the Politics

and Philosophy Department at the University of Idaho. Prior to joining
that department, Dr. Johnson was a post-doctoral fellow on the project on
Humility and Conviction in Public Life at the University of Connecticut’s
Humanities Institute.
Routledge Studies in Contemporary Philosophy
The Legacy of Kant in Sellars and Meillassoux

Analytic and Continental Kantianism
Edited by Fabio Gironi
Subjectivity and the Political

Contemporary Perspectives
Edited by Gavin Rae and Emma Ingala
Aspect Perception after Wittgenstein

Seeing-As and Novelty
Edited by Michael Beaney, Brendan Harrington
and Dominic Shaw
Nature and Normativity

Biology, Teleology, and Meaning
Mark Okrent
Formal Epistemology and Cartesian Skepticism

In Defense of Belief in the Natural World
Tomoji Shogenji
Epistemic Rationality and Epistemic Normativity

Patrick Bondy
From Rules to Meanings

New Essays on Inferentialism
Edited by Ondřej Beran, Vojtěch Kolman, and Ladislav Koreň
Toleration and Freedom from Harm

Liberalism Reconceived
Andrew Jason Cohen
Voicing Dissent
The Ethics and Epistemology of Making Disagreement Public
Edited by Casey Rebecca Johnson
For a full list of titles in this series, please visit www.routledge.com

Voicing Dissent
The Ethics and Epistemology
of Making Disagreement Public
Edited by Casey Rebecca Johnson

First published 2018
by Routledge
711 Third Avenue, New York, NY 10017
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2018 Taylor & Francis
The right of the editor to be identified as the author of the editorial
material, and of the authors for their individual chapters, has been
asserted in accordance with sections 77 and 78 of the Copyright,
Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Names: Johnson, Casey Rebecca, editor.
Title: Voicing dissent : the ethics and epistemology of making
disagreements public / edited by Casey Rebecca Johnson.
Description: 1 [edition]. | New York : Routledge, 2018. | Series:
Routledge studies in contemporary philosophy ; 105 | Includes
bibliographical references and index.
Identifiers: LCCN 2017061301 | ISBN 9781138744288 (hardback :
alk. paper)
Subjects: LCSH: Knowledge, Theory of. | Public opinion. |
Conformity.
Classification: LCC BD181 .V58 2018 | DDC 302/.1—dc23
LC record available at https://lccn.loc.gov/2017061301
ISBN: 978-1-138-74428-8 (hbk)
ISBN: 978-1-315-18118-9 (ebk)
Typeset in Sabon
by Apex CoVantage, LLC
Contents
Acknowledgements vii
Introduction 1
CASEY REBECCA JOHNSON
1 Reasonable Disagreement 10
CATHERINE Z. ELGIN
2 Disagreements, of Beliefs and Otherwise 22

DUNCAN PRITCHARD
3 Dissent: Ethics and Epistemology 40

SANFORD C. GOLDBERG
4 Dissent: Good, Bad, and Reasonable 61

KLEMENS KAPPEL
5 Silence and Objecting 82

JENNIFER LACKEY
6 For the Sake of Argument: The Nature and Extent of Our

Obligation to Voice Disagreement 97
CASEY REBECCA JOHNSON
7 Eloquent Silences: Silence and Dissent 109

ALESSANDRA TANESINI
8 Epistemic Arrogance and the Value of Political Dissent 129

MICHAEL PATRICK LYNCH
vi Contents
9 Emancipatory Political Dissent in Practice: Insights
From Social Theory 140
RACHEL ANN MCKINNEY
10 Speaking and Listening to Acts of Political Dissent 164

MATTHEW CHRISMAN AND GRAHAM HUBBS
11 Responding to Harmful Speech: The More Speech Response,

Counter Speech, and the Complexity of Language Use 182
MARY KATE MCGOWAN
Contributors 201
Index 202
Acknowledgements
I am grateful to many people for their help with this volume. First and
foremost, I’m grateful to the contributors for their thoughtful and thought-
provoking chapters. My thanks also to Michael P. Lynch, Catherine Elgin,
and Mary Kate McGowan for encouraging me to develop this project.
The workshop that helped produce this volume was supported by Humil-
ity & Conviction in Public Life, a project of the University of Connecti-
cut’s Humanities Institute, funded by the John Templeton Foundation. That
workshop (and much of my work from 2015–2017) would not have been
possible without help from Jo-Ann Waide-Wunschel and Nasya Al-Saidy. At
the University of Idaho, my thanks to McKenzie Moore for her work on for-
matting and indexing the volume and to Omni Francetich and the Politics
and Philosophy Department for their support.
Introduction
Casey Rebecca Johnson
In the spring of 2016, I found that I wanted to have a particular conversa-

tion about disagreement. I’d been reading about peer disagreement, social
epistemology, trust and testimony, but also about echo chambers, social
psychology, gossip, and hate speech, and it occurred to me that there could
be something philosophically interesting where these issues meet. Mostly,
I wanted to get together a bunch of philosophers working on epistemol-
ogy, ethics, political philosophy, and their intersections and ask them what
they think about disagreement. We would take lessons from our various
subfields to make sense of realistic, real life, dissent expressed by realistic,
socially embedded epistemic agents. The conversation seemed interesting
and timely, given the state of the philosophical literature on disagreement.
And I was sure that from that conversation we could build a volume on
publically disagreeing.
Then, in the fall of 2016, with the passing of Brexit and the election of
Donald Trump, our philosophically timely conversation took on new import
and gravity. Disagreement, though always present and part of our social and
political lives, became ubiquitous, inescapable, and pressing. And voicing
that disagreement came to feel, to many, like a moral and political impera-
tive. Faced with protests, marches, and free-speech debates, with Twitter
and fake news, our conversation took on new meaning and new scope.
Disagreement has been on the radar of analytic philosophers for some
time, especially as social epistemologists moved us from viewing the knower
as alone in a sea of evidence to a more realistic view of the knower embedded
in a community. We are not, the observation goes, lone Cartesian subjects,
forming our beliefs from the armchair.1 Instead, the social epistemologist
argues, we form our beliefs and come to have what justification we achieve,
in part, by interacting with others.2 And sometimes, those others disagree
with us. So far, so good—the social epistemologist has made progress.
However, the mainstream analytic social epistemologist’s inquiry into dis-
agreement is then almost entirely absorbed by what agents should do about
their beliefs when they disagree with others.3 Should we remain steadfast?
Should we conciliate? Should we capitulate? The work is still, almost exclu-
sively, concerned with the beliefs the agent should or should not form. And
2 Casey Rebecca Johnson
clearly, once an agent comes to know that her peer disagrees, there may be
rational requirements on that agent. However, for disagreement to make a
difference to agents at all, it has to be communicated. That my interlocutor
disagrees with me is only available and relevant to me if it is communicated
or made public. And, when disagreement is made public it has both epis-
temic and ethical consequences.
Other areas of social epistemology and political philosophy have focused
less on peer disagreement and more on the ways social factors interact with
testimony and knowledge. In work on epistemic injustice, silencing, epis-
temic violence, and epistemologies of resistance, social difference makes a
difference.4 Here there is a clear understanding that knowers are socially
embedded, and a commitment to capturing important features of social
context. Social context affects knowers and their access to and facility in
exchanging ideas. But in this literature, disagreement as such has enjoyed
less particular attention. And there are important observations from these
subfields for our account of public disagreement. For one, different people
face different consequences for voicing disagreement. And different social
positions make voicing disagreement easier or feel more natural to some
of us rather than others. Finally, there are different tactics one might take,
depending on one’s social context, to make one’s disagreement public.
Social context matters for the chapters in this volume, though it is more
central for some than others. And the contributions are informed by and
responding to a wide variety of work. The goal is to bring a varied set of
related literatures to bear on what we take to be an important question:
what happens when disagreement goes public?
As in the previous literature, some of the chapters will be asking what should
agents do when they find themselves disagreeing. However, we are not only
concerned with cases that involve epistemic peers, but also those wherein
the epistemic credentials of their interlocutor is unknown or unimportant.
We also question how social, political, and epistemic obligations regard-
ing disagreement are generated. What is it about conversation, inquiry, or
reasonable expectations that obligate agents to make their disagreements
known? To keep them quiet? Under what circumstances do these obliga-
tions arise? What constraints are there on the ways in which we voice dis-
agreements? What is the appropriate response to disagreeable content? Can
silence be a way to dissent? How should we respond to hate speech? These
questions, and related concerns, are the focus of the volume.
The discussion in the volume starts by addressing issues most closely
related to peer disagreement, and gradually broadens in scope to include
increasing amounts of political detail. Throughout, we are concerned with
voiced dissent—disagreement that has come to light or been communicated.
As we move through the chapters, the kinds of disagreements and the topics
of dissent move from the academic and abstract to the realistic and political.
The conversation about these questions is one I’ve wanted to have for a
while. And I was lucky enough to have a version of it in May 2017. With
Introduction 3
the support of the Public Discourse Project and the project on Humility
and Conviction in Public Life at the University of Connecticut (funded by
the John Templeton Foundation), many of the contributors for this vol-
ume gathered in Storrs, CT for a workshop. As you’ll see, in some cases
this resulted in direct reference to one another’s contributions. During the
course of the workshop several things became clear: first, we all agreed that
disagreement was an important and central phenomenon in our epistemic
lives. However, we disagreed about disagreement in a number of impor-
tant ways. For instance, contributors to this volume hold a wide variety
of commitments regarding the nature of disagreement, its epistemic value,
and the right responses to it. Fortunately, those disagreements (and areas of
commonality) about disagreement have been illuminating. I detail them fur-
ther below.
Dissent and Disagreement

Intuitively, a disagreement occurs when two or more thinkers’ beliefs differ
over the truth or the justification of some proposition. Cases in which you
think that p and I think that not-p are clearly instances of disagreement. We
also seem to disagree if you think that p and I’m agnostic as to whether or
not that p (though this is more contentious). More contentious still is a case
in which you fully believe that p and I believe that p but have some reserva-
tions. Does this count as a disagreement?
It will be no surprise that the contributors to this volume do not all agree
about disagreement. We don’t all agree about what genuine disagreement
is—does it require belief? Is it a mere difference in credence for some propo-
sition? We don’t all agree about what disagreement demands of us—are
we rationally required to change our beliefs? Are we epistemically obli-
gated to voice our disagreement? Does disagreement merely indicate a dif-
ference in antecedent commitments? Is it epistemically problematic, or an
expected part of inquiry? And we don’t all agree about whether disagree-
ment demands anything at all.
In addition to disagreements about what disagreement is, there are a
variety of views on when disagreement is reasonable, and when it is good.
These views suggest positions about when disagreement must be attended
to—when do we have good epistemic reasons to attend to those who voice
their dissent to us, and when can they be ignored? When is dissent political?
When can such dissent be justified?
We also differ about what it takes to make our disagreements public. In
the foregoing discussion, I took it that making disagreement public was as
simple as saying it out loud, shaking one’s head, or otherwise communicat-
ing it. That, however, clearly ignores two key complications: first it ignores
the complicated question of the nature of a public.5 Must a disagreement
be expressed in a town square or in a newspaper to be properly public? Is
expressing a disagreement with a going theory to like-minded individuals
making it public? Or must we take our disagreements to the source (so to
speak)?
Second, my simple discussion of voicing disagreement ignores the dif-
ficulty some speakers have in making themselves heard, or in finding an
audience at all.6 Does publicizing our disagreements require that we secure
uptake? Must our disagreement be made explicit? Does silence sometimes
work as dissent? Is public dissent always aimed at policy change? What
factors bear on what we ought to do, conversationally and epistemically, in
the face of public disagreement or dissent? Must a disagreement be sincere
to warrant response, or do conversations with devil’s advocates make epis-
temic demands of us?
Despite these differences, there are some common commitments running
through the chapters in this volume. For one, it is clear that disagreement and
dissent are closely related. Dissent is one way in which disagreement is made
public. Dissent is one way in which we talk about disagreement with an
established or institutional view. There are, however, some intuitive differ-
ences. Disagreement must be sincere, while dissent can be voiced insincerely.
Disagreement need not be voiced at all, while dissent requires, at least, a
communicative act. These two are, therefore, distinct. It is an open question,
and likely a matter of theory choice, if every voiced disagreement counts as
dissent. Here, too, is a point where different contributors differ.
Useful distinctions can be drawn between disagreements that are voiced
in intimate or small conversations, and disagreement that is voiced in a pub-
lic speech, at a protest, or as an op-ed. The publicity of these communicative
acts may be a matter of degree, however, the intentions and the epistemic
effects of each (not to mention the risks and possible repercussions of each)
are often quite different. Some of the following chapters make no use of this
distinction, while others are concerned only with one or the other kind of
voiced disagreement.
With regard to voicing disagreement, one of the goals of this volume is
to attend to real social factors affecting agent’s ability to participate in con-
versation. Informed by work in psychology, legal theory, and political phi-
losophy, we attempt to give due attention to the social facts on the ground
while also offering clear and precise analysis of disagreement and dissent in
the real world.
The Value of Dissent

During the workshop we returned at several points to the value of dissent
and voiced disagreement. One common intuition apparent in the literature
is to treat disagreement as an indication that something has gone wrong.
Disagreement, then, has value, but only as a kind of alarm system. When
we disagree, on this view, something must be done to resolve and get rid of
that disagreement. We might think, on the other hand, that disagreement
has a kind of positive value. Especially voiced disagreement is perceived as
Introduction 5
a kind of democratic necessity. Without it, we run the risk of polarization
and dogmatism—we run the risk, in other words, of only exchanging ideas
with those we find agreeable, driving disagreeing positions further and fur-
ther apart. Further, we might think of public disagreement as an epistemic
necessity. It helps us, as Mill would point out, to be sure of our reasons and
justification.7 Voiced disagreement can also help us signal various things to
our communities—an openness for discussion, for example, or that we’ve
opted out of some practice.
This is not to suggest that we take public disagreement to be of value all
things considered (in all contexts), nor is it to suggest that disagreement’s
costs are negligible. If, for example, I’m at dinner with my partner’s boss,
for example, voicing all of my disagreements might be rude, and practi-
cally inappropriate. If, in some other case, my interlocutor is intransigent
and racist, then voicing my disagreement with his claims might well further
entrench his racism. Voicing disagreement doesn’t always have epistemic
benefits. Further, the costs of disagreeing are not always fairly distributed.
This comes into particularly sharp relief when we consider the ways in
which members of socially subordinated groups face disproportionate cen-
sure for disagreeing. It might be harder for a woman to speak up against
misogyny, for example. These are the kinds of details we are committed to
accommodating in our discussion of voiced disagreement.
List of Articles in This Volume

Catherine Z. Elgin, “Reasonable Disagreement”: Elgin offers a discussion
of some nuances so far neglected in the peer disagreement literature. Elgin
points out that the so-called solutions to instances of peer disagreement
assume that disagreement is a problem to be solved when peers either con-
ciliate or they remain steadfast. The assumption is that when peers disagree
it is a problem because one of them must have made a mistake. But this,
Elgin argues, is itself a mistake. There are many axes along which disagree-
ment occurs, and progress in inquiry requires that we tolerate and explore
together with people who disagree with us along some of these axes. Rather
than being a problem to be solved, Elgin argues that these disagreements can
be of epistemic benefit.
Duncan Pritchard, “Disagreements, of Beliefs and Otherwise”: In his
article, Pritchard explores what is necessary for a genuine disagreement—
crucially, whether such a disagreement requires a difference at the level of
beliefs. Pritchard argues that it usually does. There is a large class of what
might appear to be disagreements are not genuinely so, but instead are the
result of what he calls “dialectical posturing.” The dialectical poseur does
not put any serious epistemic effort into addressing objections to her view,
or to integrate it into the rest of her commitments. Dialectical posturing
might be seen as a kind of manipulation or mere dialectical-role play. While
one might be epistemically obligated to do something in light of genuine
disagreement, one is perfectly in one’s epistemic rights to ignore dialectical
posturing. There are, however, some genuine disagreements that are not
at the level of belief. These are over what Pritchard calls “hinge commit-
ments.” Differences in hinge commitments are not differences at the level
of belief; however, they are nonetheless disagreements and are epistemically
relevant.
Sanford C. Goldberg, “Dissent: Ethics and Epistemology”: Goldberg
explores the pressure that social regularities can generate for members of
certain communities. In particular, he argues that when communities have
signaling practices, these practices give community members some practice-
specific reasons. These include reasons for a member to indicate publically
when she is opting out of the practice. An agent ought not, given the right
background practice, just quietly opt out of signaling. She ought to make
her defection explicit—it ought to be voiced. This pressure is generated by
fairly weak and plausible epistemic and ethical principles and bears directly
on the ethics and epistemology of dissent.
Klemens Kappel, “Dissent: Good, Bad and Reasonable”: Kappel draws a
useful distinction between different ways we diagnose dissent. Taking pub-
lic dissent as his focus, Kappel examines what makes dissent epistemically
good or bad, and what makes dissent reasonable or unreasonable. Kappel
develops an account of reasonable and unreasonable dissent that is simi-
lar to Rawls’s account of reasonable and unreasonable political views, and
offers some suggestions for responding to unreasonable dissent. One key
conclusion that emerges is that when we diagnose dissent as reasonable or
unreasonable we are rarely offering a mere description. More often this
diagnosis is meta-level commentary on the manner in which we should treat
the instance of dissent in question. Thus, the labeling of dissent as reason-
able or unreasonable is itself often a political act.
Jennifer Lackey, “Silence and Objecting”: In her article, Lackey observes
that there is a connection between silence and the duty we have to object
to what we take to be false or unwarranted. She argues, however, that the
central approach to explaining this connection is flawed. The flaw in what
she calls “the cooperative conversation view” is that it excludes some non-
ideal facts about the world, which leads it to make faulty assumptions about
conversers and conversations. The solution is to begin with a non-ideal
theory—one that focuses on features of the actual world like power and
oppression. This kind of theory, as Lackey shows, demonstrates that, for
many, objecting is a luxury.
Casey Rebecca Johnson, “For the Sake of Argument: The Nature and
Extent of Our Obligation to Voice Disagreement”: Johnson builds on the
idea that agents, at least sometimes, have an epistemic obligation to voice
disagreements. Any of four background theories, inspired by influential
work in social epistemology, can generate these obligations. However, each
background theory generates obligations with different characteristics.
Johnson explores these differences by looking at the extent and limits of
Introduction 7
each. Key questions include the conditions under which agents should voice
their disagreement, and to what extent that disagreement must be sincere.
One way of asking this second question is to ask, to what extent are we
epistemically obligated to play devil’s advocate?
Alessandra Tanesini, “Eloquent Silences: Silence and Dissent”: As Tane-
sini observes, silence is increasingly used as a means to protest and resist.
Silence can be, and is used as, a communicative act. This, however, has
gone unnoticed by philosophical accounts of dissent. Tanesini’s chapter fills
this lacuna in the philosophical literature by showing how silence can be
an illocution. In so doing she offers an analysis of what she calls “eloquent
silence.” These silences are, and are intended to be communicative and can
indicate resistance and dissent. This means that silence, contrary to some
accounts, does not always indicate acceptance. Tanesini also identifies some
features of eloquent silences that explain why they are effective means of
expressing dissent.
Michael Lynch, “Epistemic Arrogance and the Value of Political Dissent”:
Lynch writes about what he calls “discursive political dissent,” that is, dis-
sent that is voiced in the form of an argument and that aims at influencing
policy or people’s beliefs about policies. This kind of dissent, Lynch argues,
can be epistemically valuable when it brings about positive epistemic con-
sequences in the right kind of circumstances. Not every case of discursive
political dissent will bring about these positive consequences, especially
given the prevalence of a particular kind of arrogance—epistemic arro-
gance. Epistemic arrogance in a dissenter’s communicative community can
prevent the members of that community from being receptive to the dissent.
And this, in turn, can prevent the epistemically positive consequences from
obtaining.
Rachel Ann McKinney, “Emancipatory Political Dissent in Practice:
Insights from Social Theory”: In her chapter, McKinney tracks dissent before
it becomes public, and recognized by those in power. McKinney argues that
public dissent does not initiate contention, but instead comes after a period
of less visible thought and discourse. Covert dissent, like gossip, rumor, and
graffiti is often instrumental to allow for public dissent. Participating in this
kind of activity can help agents to develop ways of voicing their political dis-
sent publically and also helps crystalize the terms and limits of the content
that is eventually voiced.
Matthew Chrisman and Graham Hubbs, “Speaking and Listening to Acts
of Political Dissent”: In their article, Chrisman and Hubbs are concerned
with the norms on political dissent, which is a subset of political speech.
Chrisman and Hubbs discuss two ways to understand the role of political
speech in a just democracy: the liberal and Rousseauean. These two ways
of understanding the role of political speech in turn offer two distinct ways
of justifying political dissent. Both views, however, are insufficiently mor-
ally neutral, so Chrisman and Hubbs offer a third and preferable account.
Theirs is a speech-act analysis of dissent.
Mary Kate McGowan, “Responding to Harmful Speech: The More
Speech Response, Counter Speech, and the Complexity of Language Use”:
Dissent is voiced disagreement with some objectionable content. Sometimes
that objectionable content is harmful speech. McGowan’s article investi-
gates the widespread view that the remedy for such speech is more speech.
She argues that the so-called “more speech response” fails to capture the
complexities of conversation. Using work in philosophy of language and
the kinematics of conversation, McGowan demonstrates the insufficiency of
the more speech response, and also shows that speech can, nonetheless, have
profound positive effects.
Notes
(Goldman, 1991; Mills, 1988; Sosa, 2010)
1
2 (Fricker, 2007; Goldberg, 2011; Hinchman, 2005; Lackey, 2008)
3 (White, 2005; Elga, 2007; Christensen, 2009; Kelly, 2010; Enoch, 2010)
4 (Coady, 2010; Fricker, 2007, 2013; Hawley, 2011; Kukla, 2012; Maitra, 2010;
Medina, 2012; Pohlhaus, 2013)
5 (Fraser, 1990; Warner, 2002)
6 (Maitra, 2010; Maitra & McGowan, 2010; McGowan, 2009, 2013; Mcgowan,
Adelman, Helmers, & Stolzenberg, 2011)
7 (Mill, 1869)
References
Christensen, D. (2009). Disagreement as evidence: The epistemology of controversy.
Philosophy Compass, 4(5), 756–767.
Coady, D. (2010). Two concepts of epistemic injustice. Episteme, 7(2), 101–113.
Elga, A. (2007). Reflection and disagreement. Noûs, 41(3), 478–502.
Enoch, D. (2010). Not just a truthometer: Taking oneself seriously (but not too seri-
ously) in cases of peer disagreement. Mind, 119(476), 953–997.
Fraser, N. (1990). Rethinking the public sphere: A contribution to the critique of
actually existing democracy. Social Text, 25/26.
Fricker, M. (2007). Epistemic injustice: Power and the ethics of knowing. Oxford:
Oxford University Press.
Fricker, M. (2013). Epistemic justice as a condition of political freedom? Synthese,
190(7), 1317–1332. doi:10.1007/s11229-012-0227-3
Goldberg, S. (2011). Putting the norm of assertion to work: The case of testimony*
assertion. Oxford: Oxford University Press.
Goldman, A. I. (1999). Knowledge in a social world (Vol. 281). Oxford: Clarendon
Press.
Hawley, K. (2011). Knowing how and epistemic injustice. Knowing How: Essays on
Knowledge, Mind, and Action, 283–299.
Hinchman, E. (2005). Telling as inviting to trust. Philosophy and Phenomenological
Research, 70(3), 562–587.
Kelly, T. (2010). Peer disagreement and higher order evidence. In Goldman, A. &
Whitcomb, D (Eds.), Social epistemology: Essential readings (pp. 183–217).
Oxford: Oxford University Press.
Kukla, R. (2012). Performative force, convention, and discursive injustice. Hypatia,
28(3).
Introduction 9
Lackey, J. (2008). Learning from words: Testimony as a source of knowledge.
Maitra, I. (2010). The nature of epistemic injustice. Philosophical Books, 51(4),
195–211.
Maitra, I., & McGowan, M. K. (2010). On racist hate speech and the scope of a
free speech principle. The Canadian Journal of Law and Jurisprudence, 23(2),
343–372.
McGowan, M. K. (2009). Oppressive speech. Australasian Journal of Philosophy,
87(3), 389–407.
Mcgowan, M. K. (2013). Sincerity silencing. Hypatia, 28(3), n/a–n/a.
Mcgowan, M. K., Adelman, A., Helmers, S., & Stolzenberg, J. (2011). A partial
defense of illocutionary silencing. Hypatia, 26(1), 132–149.
Medina, J. (2012). The epistemology of resistance: Gender and racial oppression,
epistemic injustice, and the social imagination. Oxford: Oxford University Press.
Mill, J. S. (1869). On liberty. London: Longmans, Green, Reader, and Dyer.
Mills, C. W. (1988). Alternative epistemologies. Social Theory and Practice, 14(3),
237–263.
Pohlhaus, G. (2013). Discerning the primary epistemic harm in cases of testimonial
injustice. Social Epistemology, 1–16.
Sosa, E. (2010). The epistemology of disagreement. In A. Haddock, A. Millar, & D.
Pritchard (Eds.), Social epistemology (pp. 278–297). Oxford: Oxford University
Press.
Warner, M. (2002). Publics and counterpublics (abbreviated version). Quarterly
Journal of Speech, 88(4), 413–425.
White, R. (2005). Epistemic permissiveness. Philosophical Perspectives, 19(1), 445–459.
1 Reasonable Disagreement1
Catherine Z. Elgin
Normally, we take it that we know rather a lot. We take ourselves to know

the date of the Battle of Trafalgar (1805), the atomic number of gold (79),
the score of last week’s football game (Arsenal beat Manchester United,
2–0), the name of the Queen of Spain (Sofia). We have reasons—often quite
good reasons—for believing the things we take ourselves to know. Evidence
mounts up, eventually reaching or surpassing the threshold required for
knowledge. But typically our reasons are less than conclusive. They just
render our conclusions highly probable. Moreover, some of the reasons
may themselves be epistemically precarious. Perhaps some of the evidence
is misleading or unreliable. Perhaps a source we trusted was misinformed.
Perhaps an inference was unsound. Given our epistemic vulnerability, it is
no surprise that others disagree with us about many of the things we take
ourselves to know. What should we make of that? Typically, philosophers
assume that where parties disagree about a matter of fact at least one of
them has made a mistake. I will argue that that is false. Because judgment
involves choices from within ranges of epistemically acceptable alternatives,
not all disagreements, and not all peer disagreements, about matters of fact
rest on mistakes. This is not to say that the conclusions of both parties are
true. It is to say neither party made an error in coming to her conclusion. If
this is right, it reveals something important about the precariousness of our
epistemic condition.
The standard question with respect to disagreement is how epistemic
agents should respond. Should we hold fast to our opinions, revise or mod-
erate them in the face of disagreement, jump to the other side? When the
question is framed so schematically, the alternatives seem stark:
A. Be steadfast: hold on to your convictions, disregarding disagreement.

B. Conciliate: moderate your conviction in the face of disagreement. Either
suspend judgment about the issue under dispute or lower your credence
in your conclusion.
C. Concede: endorse your opponent’s position, abandoning your own.
We need not think that the same stance is appropriate in every case. Some
disagreements may call for steadfastness, others for conciliation. In yet
Reasonable Disagreement 11
others, the best thing to do is concede that your opponent is right. Let’s
look at some cases.
Hal is and knows himself to be pretty clueless about ornithology and
rightly considers Val quite knowledgeable. He thinks that all goldfinches
look alike. Val disagrees. She thinks goldfinches display sexual dimorphism.
In the face of their disagreement, Hal probably ought to concede that she
is right, or at least suspend judgment. He recognizes that she has expertise
that he lacks. Their situations are not symmetrical. The disagreement gives
Val no grounds for revising her opinion. Since Hal can hardly tell a sparrow
from a hawk, his opinion need not give her pause. It is reasonable for her
to remain steadfast. Ceteris paribus, when an expert and a novice disagree,
the novice should back off. He should then either conciliate or concede. The
expert, however, can reasonably hold fast.
What if parties to a disagreement are epistemic peers? Neither is more
knowledgeable than the other. Here their level of competence matters. Bill
and Jill are on the verge of failing introductory chemistry. Jill thinks that
methane is an organic chemical; Bill thinks that it is inorganic. Given the
precariousness of their grasp of the topic, they should probably suspend
judgment and look it up. If an agent’s grasp of an issue is sufficiently weak,
it should be unsettled by the news that anyone, even an equally ignorant
peer, disagrees.
Cases like these are easily handled. Disagreements among competent
epistemic peers are more troublesome. What should a responsible, com-
petent epistemic agent do on learning that a peer with the same evidence,
background assumptions, reasoning abilities, and epistemic incentives dis-
agrees?2 In what follows I will assume that the peers under discussion are
competent with regard to the issue in dispute. Concessiveness is reasonable
only when an epistemic agent believes that her interlocutor has considerably
greater relevant expertise than she does. Where peers disagree, there is no
reason to be concessive. In peer disagreements, an epistemic agent should
either stick to her guns, or she should conciliate. Kelly (2005) advocates
steadfastness. If Nell considers herself to have been responsible in forming
her opinion, she should dismiss her peer Mel’s opinion, concluding that
he must be in error. Christensen (2007) advocates adopting a conciliatory
stance. Nell should lower her credence or suspend judgment, concluding
that one of the two is in error. Either way, it is widely agreed that at least
one has made a mistake—miscalculated, overlooked evidence, ignored base
rates, reasoned fallaciously, or committed some other cognitive gaffe. What-
ever the precise failing, where peers disagree, at least one of them did some-
thing cognitively culpable. As Sidgwick says, “If I find any of my judgments,
intuitive or inferential, in direct conflict with some other mind, there must
be error somewhere” (1981, p. 342).
Some disagreements, such as Christensen’s restaurant case, exhibit this
profile. When competent agents who are equally good at mental math dis-
agree about how to split the bill, at least one of them miscalculated. The
obvious remedy is to suspend judgment temporarily, do the calculation
12 Catherine Z. Elgin
publicly, and figure out the right answer (2007). Similarly, if they disagree
about the date of the Battle of Trafalgar. That too can be easily checked.
Pretty plainly, parties to the dispute should suspend judgment, pull out their
smart phones, and google the answer. Both questions are easily resolved. To
suspend judgment and check the answer takes little time and effort. Concili-
ation in such cases is relatively cost free.
But many disagreements lack an easy resolution. There is no feasible way
to backtrack and find an error. Sometimes the information that would settle
the dispute is simply not to be had. Ann and Dan, forty-year-olds with good
memories, disagree about which team won the third-grade sack race when
they were in school. No records were kept about this momentous event. Nor
are they in touch with any of their former classmates. That being the case,
their disagreement seems fated to remain unresolved. In cases like this, the
protagonists agree about everything relevant except the fact at the heart of
the dispute. Someone made an error, but there is no way to determine who.
In other cases, however, there is no error. That is not to say that both
claims are true. It is to say that both opinions are products of epistemically
acceptable, perhaps even epistemically impeccable, reasoning. Although one
of the protagonists arrived at a false conclusion, neither made a mistake.
I want to focus on such a case.
Suppose an animal bone with holes pierced in it is found among the arti-
facts in a Neanderthal settlement.3 It looks enough like a primitive flute that
had it been found in a prehistoric Homo sapiens settlement, paleoanthro-
pologists would have no reservations about calling it a flute. But according
to current theory, Neanderthal brains were not complex enough to create
music, nor were their hands dexterous enough to finger a flute. Jen and Ken,
both eminent paleoanthropologists, disagree about whether the artifact is
a Neanderthal flute. Jen thinks it is a Neanderthal flute; Ken thinks it is
not. Their disagreement may be irresolvable. Given the paucity of evidence
about Neanderthals, there is no expectation that anyone will ever come up
with sufficient evidence to settle the matter conclusively. Moreover, some-
thing significant is at stake here. The verdict as to whether the artifact is a
Neanderthal flute will contribute to the ongoing debate about how and in
what respects Neanderthals differed from early Homo sapiens. If Neander-
thals were capable of making music, a host of related hypotheses are seri-
ously off the mark.
Jen and Ken are equally knowledgeable and competent, and reach or sur-
pass the threshold for being qualified to judge the matter under dispute.
They have the same evidence and the same ability to assess the evidence.
Both consider the issue significant. Neither is cavalier. What should they do?
Conceding has no appeal. Neither party to this dispute has reason to think
that the other is in a better position to assess the matter. But both concili-
ation and steadfastness look like live options. What is to be said for them?
The case for steadfastness: Jen knows that she has considered the matter
carefully and judiciously and has taken account of all the available evidence.
She knows that she would have no reservations about her conclusion if she
relied only on the first-order evidence, her background assumptions, and
the inferences she draws from them. She rightly considers Ken her intellec-
tual peer in paleoanthropology. So she is surprised that Ken does not agree
with her. If she is steadfast, she does not, however, consider his disagree-
ment a reason to rethink her position or moderate her confidence level. She
concludes that in this case Ken must have made a mistake. Maybe he was
careless in his reasoning; maybe he overlooked some relevant factor. If he
had thought the matter through as carefully as she did, she thinks, he would
have come to the same conclusion. Ken’s disagreement gives her no reason
to change her mind (see Kelly, 2005). Ken’s situation is symmetrical. He
too recognizes that he thought the matter through carefully and judiciously.
Being equally steadfast, he can only conclude that Jen, although normally
quite solid on such issues, has made a regrettable mistake. Her disagree-
ment gives him no reason to change his view either. If they remain steadfast,
Jen and Ken may permanently be at loggerheads. Since each is inclined to
dismiss the other’s opinion, and to have no incentive to examine the other’s
reasons, neither is in a position to learn from the other. The disagreement
gives them no reason to rethink the issue. The steadfast are intellectually
arrogant. Faced with a disagreement, a steadfast epistemic agent immedi-
ately concludes that the other party must be wrong. They are dogmatic.
They do not think that a peer’s disagreeing gives them any reason to rethink
their position or change their mind. They are, moreover, disrespectful of
those they consider intellectual equals. Indeed, we might wonder whether
they really do consider their opponents intellectual equals, if they are so
quick to claim superiority the moment a disagreement arises.
The case for conciliation: we’ve seen that Jen has thought through the
matter carefully and judiciously. She would have no reservations about her
conclusion if she relied exclusively on the first-order evidence, her back-
ground assumptions, and the inferences she draws from them. But Ken,
whom she considers her epistemic peer, disagrees with her. She has no rea-
son to think that Ken was any less careful and judicious in his approach to
the question, or that he was cavalier in his reasoning or interpretation of the
evidence. One of them must be wrong. But given that they are peers, there is
currently no way to tell which one. She therefore suspends judgment, hop-
ing that further evidence will be found that will decide the issue between
them. Conciliators are intellectually humble. Their epistemic stance is fal-
libilist. Jen thinks she is right, but acknowledges that she could be wrong.
Conciliators are respectful of their epistemic peers. As far as Jen can tell,
Ken’s view has as much going for it as hers. There is no basis for choosing
between them.
It might seem that this shows that when epistemic peers disagree, they
should conciliate. Certainly conciliators seem more congenial than the stead-
fast. But conciliation can be costly. If Jen and Ken take their disagreement as
a reason to suspend judgment about whether the artifact is a Neanderthal
flute, neither of them is in a position to know. They have backed off from
belief about the matter. If Jen’s first-order evidence was actually sufficient
for her to know, then by suspending judgment she sacrifices knowledge. For
knowledge requires belief. One of them is surely right. The artifact either is
a Neanderthal flute or it is not. If both hold fast to their opinions, then one
of them has a true belief. And if the one with the true belief has sufficient
evidence that bears non-accidentally on its truth, one of them knows. If Jen
is right and sticks to her guns, she knows or at least justifiably believes that
the object is a Neanderthal flute. So, it seems, the steadfast are capable of
preserving knowledge while the conciliatory willingly give it up.
Steadfastness, however, instantiates Kripke’s so-called dogmatism para-
dox (2011).
If S knows that p, then p is true.

If p is true, then any evidence against p is misleading.
If S wants to retain her knowledge, she should disregard misleading
information.
So S should be dogmatic and disregard any evidence that seems to tell
against p.
This argument is perfectly general. It applies to first-order evidence as well

as evidence that S might glean from the discovery that T disagrees with her.
It is not strictly a paradox, but it yields an epistemically disheartening rec-
ommendation. Rationality requires that epistemic agents be open-minded;
they should not intentionally blind themselves to evidence, even if that evi-
dence might turn out to be misleading. So if Ken’s disagreement is evidence
that Jen is wrong, Jen should not disregard it. Although the steadfast can
retain knowledge or highly justified belief, they do so by shirking their epis-
temic responsibility. It is just by luck that the evidence that might be gleaned
from taking the disagreement seriously turned out to be misleading. Given
that Jen and Ken are epistemic peers, it is just by luck that Jen (rather than
Ken), by being steadfast, retains her knowledge.
Conciliators, however, might seem spineless. The moment they learn that
a peer disagrees, they back off from full belief. It seems epistemically ill-
advised, and perhaps a bit cowardly, to move too quickly from “I might
be wrong” to “I should immediately abandon my position.” If this is what
conciliation requires, it is not an attractive option. There is, however, some
slack in the conciliationist position. Conciliators hold that on learning that a
peer disagrees, an epistemic agent should either suspend judgment or lower
her credence. Suspending judgment is a lot to ask, at least if that means
having no opinion or thinking one is unjustified in one’s opinion about a
matter. Given the range of disagreement, most of us would have to suspend
judgment about a lot. But it may be reasonable to say that we should lower
our credence in our conclusion. Conciliators could still be fairly confident
that their conclusion holds, but they would have to admit that they are not
fully confident and are not entitled to be fully confident about it. This might
result in a loss of knowledge. Whether it does depends on where the thresh-
old for knowledge is, and how far above the threshold an agent’s belief orig-
inally was. Arguably, it often should result in a loss of knowledge. The agent
would not think that she is, or should consider herself to be, utterly clueless
about the topic or spineless in backing off. But she would and should have
her confidence in her conviction shaken. She might then still be pretty sure
of the conclusion.
Conciliators seem to gain more than they lose. We saw that the steadfast
are dogmatic. They dismiss their opponent’s opinion as mistaken without
further investigation. If steadfast Ken is wrong, he is likely to remain wrong,
since he takes himself to have no reason to rethink the issue. The existence
of the disagreement gives him no incentive to investigate the matter further.
Conciliators have such an incentive. Jen and Ken have the same evidence
and the same background information. On learning that Ken disagrees with
her, Jen has reason to ask, “What does he see that I do not?” She has an
incentive to go back and examine the evidence, the background assump-
tions, her reasoning, and, if she has access to it, his. Has she overlooked
something? Has she reasoned fallaciously? Is there an illuminating perspec-
tive on the issue that she missed? Conciliators are in a better position than
the steadfast to find the mistake that led to the disagreement, assuming of
course, that there was a mistake.
But the assumption that disagreement always rests on a mistake itself is
a mistake. The issue is why the peers disagree. Even in the idealized sce-
nario, where peers are identical in multiple relevant respects, there can be
a variety of reasons. Although it is stipulated that the peers have the same
evidence, background assumptions, reasoning abilities, and epistemic moti-
vation, there is no reason to assume that they use them in the same way. If
they do not, they may arrive at different verdicts. And there may be nothing
clearly mistaken about the way either of them proceeds. In such cases, what
is at issue is not just whether p, but how one ought to determine whether
p. I will urge that our epistemic situation is more complex than is usually
recognized. Peer disagreements bring this complexity to the fore.
Disagreements occur in a multi-dimensional space. Along a variety of axes,
epistemic requirements fix ranges within which acceptable verdicts must lie.
But the requirements are not sufficiently fine-grained to assure uniqueness
or to provide a decision procedure for differentiating among judgments that
fall within the range of acceptability. If parties disagree because they make
different choices within a range of epistemically acceptable options, neither
has made a mistake. Where an agent comes down depends to a considerable
extent on which alternatives she had chosen. Had she made other choices
within the acceptable range, she would have reached a different verdict.
Epistemic peers are not epistemic clones.
By stipulation, epistemic peers have the same evidence. But it does not
follow that they assign the same weight to different bits of evidence. The
disputed artifact was found in a cave in Slovenia, in an area known to have
been populated by Neanderthals. Jen considers this highly significant, think-
ing it constitutes fairly strong evidence that the artifact was crafted by a
Neanderthal. Ken attaches less weight to this datum, noting that the perfo-
rated bone could have been left in the cave by a wandering Homo sapiens,
or been brought there by a Neanderthal, having been crafted elsewhere by
a Homo sapiens. Ken does not deny that the location where the object was
found is evidence, he just considers it relatively weak evidence.
Peers are said to have the same reasoning abilities. But they need not have
the same reasoning styles. Ken tends to reason analogically and to credit
analogical arguments. Jen, although equally adept at analogical reasoning,
prefers inference to the best explanation, considering analogical arguments
rather loose. Ken saw his dog pick up a ball in the yard this morning and
carry it over to her bed. He buttresses his belief that the artifact might have
been transported to the Neanderthal site by relying on the analogy with
his dog’s behavior. Jen thinks that the best—simplest, most plausible—
explanation of an artifact’s being found in a Neanderthal site is that it was
crafted by a denizen of that site—viz., a Neanderthal. Ken is skeptical of
inference to the best explanation, thinking that such liner reasoning tends to
blind one to available alternatives.
In characterizing a newly found object as a Neanderthal artifact, paleo-
anthropologists compare it with objects that are already characterized as
such. If it is close enough to acknowledged Neanderthal artifacts in relevant
respects, it is likely to be incorporated into the precedent class; otherwise
it is apt to be rejected. But everyone recognizes that the current precedent
class contains some items that were wrongly classified. That is, it contains
misleaders. Ken and Jen differ over which items are misleaders. This leads
them to different verdicts about this artifact. Jen credits scored bits of flint
that are thought to have been used as tools. She thinks that if the Nean-
derthals had the cognitive and physical dexterity to carve such tools, they
had the cognitive and physical dexterity to pierce holes in a bone to make
a flute. Ken doubts that the gashes in the flint were intentionally produced.
He suspects that what Jen considers intentional scores are in fact natural
striations, amplified over eons by wind and water. So the similarity to the
flint tools leads Jen to think the perforated bone is a Neanderthal artifact,
while Ken harbors doubts. It may be obvious that an evidence class contains
misleaders. Perhaps various elements are in tension with one another in the
sense that if one is a Neanderthal artifact another probably is not. But it
may not be obvious which bits of evidence are the misleaders. Peers then
may disagree because they differ over which members of the (admittedly
somewhat messy) evidence class they consider worthy of trust.
A subject’s background information about a topic consists of all the infor-
mation she has that directly or indirectly bears on that topic. Although epis-
temic peers by stipulation have the same background information, they need
not draw on the same bits of information or assign the bits they draw on the
same weight. If they do not, background information plays different roles
in their reasoning. Contemporary anatomical theory suggests, albeit weakly,
that organisms with thick phalanges are relatively lacking in fine motor con-
trol. Because Neanderthals had thick phalanges, Ken thinks that they would
have been unable to finger a primitive flute. Jen is not convinced. She consid-
ers the anatomical theory sketchy and its bearing on the case slight.
Many epistemologists follow William James in holding that our over-
arching epistemic objective is to believe as many truths as possible and to
disbelieve as many falsehoods as possible (1983). They typically then focus
exclusively on believing truths. But James’s formula involves a proportion
(see Riggs, 2003). If an agent is epistemically risk averse, she will set a high
threshold for acceptance. Perhaps she will accept only hypotheses that are
.95 probable on the evidence. If she is more daring, she sets a lower thresh-
old, perhaps .90. Both thresholds comply with the requirement than an
acceptable conclusion must be highly probable. The risk averse agent will
accept fewer truths and fewer falsehoods than the more risk tolerant agent.
But neither is epistemically irresponsible. Jen is willing to accept a larger
measure of epistemic risk than her mentor, Ben, is. She concludes that the
artifact is a Neanderthal flute because, having taken all of the previous fac-
tors into account, she deems it 92% probable that it is. Ben, being more
conservative, suspends judgment because he demands a probability of .95.
He agrees with all of Jen’s assessments. He simply does not think that they
yield a high enough probability.
Evidently peers can differ along a variety of epistemically relevant axes:
the weight to assign to evidence, the standards of acceptability, the identity
of misleaders, the relevance and importance of various bits of background
information, their favored styles of reasoning. To limit the cast of char-
acters, I described Jen and Ken as differing along all these axes. Ben was
introduced to show that even with widespread agreement in numerous other
respects, epistemic agents can differ over thresholds of acceptability. But the
factors I’ve mentioned vary independently and can point in different direc-
tions. Someone who sets a high threshold acceptability could easily favor
or disfavor analogical reasoning. Along each axis, both Jen and Ken hold
reasonable views. Neither is making an obvious mistake. Nor is it plausible
that either is making a subtle mistake. It is unreasonable to expect there
to be a precise numerical value for how risk averse one should be; it is
implausible to think that there is an algorithm for identifying misleaders, for
assigning weights to different bits of evidence, or for assessing the credibility
of an analogical argument.
We might think that this just shows that the epistemologists who set the
criteria for peerhood should include more constraints. They should insist
that epistemic peers also agree about the weight of evidence they assign,
the criteria of acceptability, the reasoning strategies they actually use, and
so forth. Then we would get back to the original question: should they be
steadfast or conciliatory? The same arguments could be given on each side
with a bit more precision. Once we control for all the differences I indicated
(and probably others that I overlooked), the familiar alternatives reappear.
The rationale for the original constraints on peerhood was to focus atten-
tion on disagreements that are epistemologically telling. As we have seen, not
all are. Drawing on legal terminology, let us say that a factor is dispositive
when it settles how a disagreement should be resolved. In some cases, imbal-
ances in levels of expertise are presumptively dispositive. Once we recognize
that one party is a novice and the other is an expert, or one party has a rel-
evant cognitive asset and the other a corresponding deficit, it is usually clear
how the disagreement ought to be resolved. But in disagreements between
epistemically competent peers, there are no (even presumptively) dispositive
disparities. To be sure, if we think the current criteria have omitted some
dispositive differences, we can augment the criteria. Elsewhere I suggested,
for example, that besides being equal in the respects standardly recognized,
epistemic peers should be equally well educated in the relevant disciplines
(2010). But the grounds for disagreement that we have canvassed here seem
different. In each of the respects we looked at, it is not at all obvious that
either of the protagonists is in an epistemically superior position. We can
determine how many additional false positives and false negatives will result
from taking .90 rather than .95 as the threshold for acceptability, but that
by itself does not tell us where the line should be drawn. Setting a precise
threshold for how probable a conclusion should be, or how much weight
should be attached to a given bit of evidence seems arbitrary. Moreover, it is
apt to be counterproductive. Philip Kitcher (1990) argues is that the scien-
tific community may best serve its collective epistemic ends by, at any given
time, countenancing a range of conflicting views. When there is a significant
chance that a currently disfavored view is true, if the members of the com-
munity want to believe what is true, they ought not foreclose inquiry prema-
turely. Multiple ways of balancing the value of believing truths against the
disvalue of believing falsehoods are often permissible. There is no reason to
think there is a single optimal balance.
I suggest that we look at the availability of these sorts of disagreements
as epistemic assets. The diversity of ways in which peers (as originally char-
acterized) can still reasonably disagree should sensitize us to the complexity
of a situation, and the epistemic opportunities that are available through
taking different perspectives on it.
We have seen that judgments can be sensitive to choices among epistemi-
cally acceptable alternatives. Had an agent made different choices within
one or more acceptable ranges, she would have justifiably come to a differ-
ent conclusion. This highlights the precariousness of our epistemic condi-
tion. That precariousness is independent of disagreement. The dependence
of Jen’s opinion on choices would be there even if she knew nothing about
Ken’s opinion—indeed, even if Ken had no opinion on the matter. This
might suggest that my discussion just underscores our fallibility. Despite the
agent’s best efforts, she still could be wrong.
Perhaps so. But disagreements among competent peers can be epistemi-
cally fruitful. If properly investigated, they provide resources for a focused
fallibilism. If parties dissect their disagreement, they can discover how the
various epistemically responsible choices affected their verdicts. Jen and
Ben, for example, might come to appreciate that the basis for their dis-
agreement has nothing to do with Neanderthals per se; it is about how
great a risk of error paleoanthropology ought to tolerate. Given their
mutual respect, disagreement might also provide parties with an incentive
to rethink their choices. Ken might, for example, be prompted to reconsider
whether he would view the matter differently if he revised his views about
which elements of the evidence class are misleaders. Jen might rethink her
doubts about the strength of analogical arguments. Even if all parties end
up endorsing their previous positions, their stance vis à vis them is different.
They appreciate that it is not just (or perhaps even mainly) a disagreement
about the status of an artifact; it is also a disagreement about epistemic
methods and standards that affect how one ought to determine the status
of the artifact.
Open-mindedness is a propensity to entertain alternative points of view.
Not all points of view merit attention. None of our protagonists has any
reason to think that the artifact in question is a space alien’s cleverly dis-
guised slide trombone. A competent peer’s disagreement effectively certifies
that a perspective is worth taking seriously. It raises the question: What is
to be said for his way of looking at things? Given that he is a peer, the pre-
sumption is that there is something to be said for it.
The picture I have sketched about the value of disagreement has an addi-
tional payoff. It vindicates the way we do philosophy. Neither the steadfast
nor the conciliatory really engage. The steadfast agent ignores the disagree-
ment and holds fast to whatever she already believed. The conciliatory agent
immediately backs off. (She may then recalculate, or recalibrate, but it is
open to her to simply suspend judgment or lower her credence without
doing anything else.) Few if any philosophers exhibit either sort of behavior.
If anyone were so bold as to simply dismiss a worthy opponent’s disagree-
ment out of hand, concluding that he must be wrong somewhere but there
is no premium in attempting to figure out where, he would not be doing phi-
losophy. Philosophy is not dogmatic. Someone who immediately suspended
judgment on learning that a peer disagreed would not be doing philosophy
either. Philosophy is not spineless. In fact, a lot of us are at home with
disagreement, but rather uncomfortable with agreement. If someone says,
“I find your position entirely convincing,” we are apt to think (if not say),
“Then you weren’t really paying attention.” Faced with disagreement, phi-
losophers engage. We attempt to uncover the basis of the disagreement, and
assess its merits. We look for shared or unshared presuppositions, common
or diverging conceptions, flaws in our or our opponent’s arguments. Some
of us may be sufficiently convinced of our views that we are quietly sure that
our interlocutor has made a mistake somewhere, but we take it that we need
to figure out where. Others may be more open-minded, both about whose
mistake it is and even whether there is a mistake. In any case, we take it that
it is our responsibility to take the disagreement seriously.
Moreover, we take it that there is considerable epistemic benefit in doing
so. We learn more about our own commitments and their vulnerabilities. We
learn more about alternatives to our position and the trade-offs that favor
one side or another. Figuring out exactly where and why you disagree with
David Lewis or Judith Jarvis Thomson is an intellectually rewarding experi-
ence. Taking well-supported philosophical disagreements seriously is intrigu-
ing, informative, and fun. If we took our own practice as a model for how
disagreement in general ought to be approached, we would be better off.
If we recognize that disagreement does not always indicate that some-
one has made a factual error, we can introduce a further stance—one that
I think philosophers take almost by default. Parties might remain committed
to their own position, considering it a good basis for inference and action
when their ends are cognitive (see Cohen, 1992; Elgin, 2017). Nevertheless,
they recognize its vulnerability. This recognition is not idle. It requires them
to be open to, and perhaps even to seek out, emerging counter-evidence and
counter-arguments. This blocks Kripke’s paradox. It gives them a strong
incentive to strengthen and stabilize their position and to protect themselves
should it turn out that they are wrong. It undermines complacency and fos-
ters intellectual respect. Like the steadfast, the committed retain their con-
victions; unlike the steadfast they do not dismiss or disregard disagreement.
Rather they use it, not only as a reminder that they might be wrong, but also
as a probe, to tease out exactly where and why they might be wrong.
If the parties to the disagreement entertain one another’s position seri-
ously and respectfully—if they keep a genuinely open mind to the possibil-
ity that there is something significant to be said for it—they may come to
appreciate weaknesses in their own position as well as strengths in their
adversary’s. This may lead one to concede that the other party is correct.
Alternatively, it may lead her to shore up her own position so that it can
deflect his objections. A third possibility is that together they craft a posi-
tion different from the one with which either of them started, or revise their
methods, standards, or criteria of acceptance. There is no assurance that
they will ever agree. Nor, as Kitcher (1990) argues, should we think that we
would be better off if we could devise criteria that insured agreement in
either the short or the long run. But by entertaining respectable alterna-
tive positions and appreciating their merits, we understand our epistemic
situation better. Moreover, if we appreciate the basis of our disagreement,
we are in a better position to introduce appropriate safeguards. Given our
fallibility, it is in general wise to take into account the possibility that a
well-supported opinion might still be false. We hedge our bets, take out
insurance, make backups, diversify our portfolios, carry umbrellas in the
recognition that our confident expectations do not always pan out. Attention
to disagreement can pinpoint the places where safeguards are most called
for. In cases like the ones we’ve discussed here, the safeguards in question
might involve introducing caveats, or highlighting the ways a conclusion is
based on assumptions that could legitimately be challenged. It might also
involve efforts to stabilize one’s position by showing that the conclusion is
not too dangerously dependent on a controversial choice. If Jen could arrive
at her conclusion without strongly depending on the disputed bits of flint, or
without dismissing analogical arguments, her case would be stronger. If she
sees how Ken’s argument works against her, she can marshal resources for
the defense. When we take disagreements among competent peers seriously,
we gain a richer, more focused appreciation of our epistemic predicament.
Whether we decide to remain committed or to conciliate, we are in a better
position to appreciate the nature, scope, and insights to be gleaned from our
epistemic vulnerability.
Notes
1 Earlier versions of this chapter were presented at the 2017 Voicing Dissent Work-
shop at the University of Connecticut, and at the 2017 meeting of the Canadian
Philosophical Association. I am grateful to the participants in both for their help.
2 These equivalences are obvious idealizations. No two people are exactly alike in
these respects. I follow the literature in making the idealizations, since the point
is that differences of opinion are possible even when there is wide agreement in
abilities and backgrounds.
3 I model this discussion on the controversy over the Divje Babe flute. I focus on
whether it is Neanderthal, although there is also considerable controversy about
whether it is a flute.
References
Christensen, D. (2007). Epistemology of disagreement: The good news. Philosophi-
cal Review, 116, 187–217.
Cohen, L. J. (1992). An essay on belief and acceptance. Oxford: Clarendon Press.
Elgin, C. (2003). Persistent disagreement. In R. Feldman and T. Warfield (Eds.),
Disagreement (pp. 53–67). Oxford: Oxford University Press.
Elgin, C. Z. (2010). Keeping things in perspective. Philosophical Studies, 150(3),
439–447.
Elgin, C. (2017). True enough. Cambridge, MA: MIT Press.
James, W. (1983). Principles of psychology. Cambridge, MA: Harvard University
Press.
Kelly, T. (2005). Epistemological puzzles about disagreement. In T. S. Gendler &
J. Hawthorne (Eds.), Oxford studies in epistemology (Vol. 1, pp. 167–196).
Kitcher, P. (1990). The division of cognitive labor. Journal of Philosophy, 87, 5–22.
Kripke, S. (2011). On two paradoxes of knowledge. In Philosophical troubles
(pp. 27–49). Oxford: Oxford University Press.
Riggs, W. (2003). Balancing our epistemic goals. Noûs, 37, 343–352.
Sidgwick, H. (1981). The methods of ethics. Indianapolis, IN: Hackett.
2 Disagreements, of Beliefs
and Otherwise
Duncan Pritchard
Much of the work on the epistemology of disagreement tends, not unrea-

sonably, to take such disagreements to be at the level of belief. I elaborate
some of the reasons why we might have this particular focus, and part of
that involves arguing that as epistemologists by “belief” we have in mind
a specific propositional attitude. As we will see, with belief so understood
it is hard to see how an exchange could constitute a genuine disagreement
without conflict of belief. Relatedly, I explore a certain core class of “dis-
agreements” that I contend are in fact nothing of the kind when properly
understood, since there is no conflict of belief, and hence lack the epistemic
import that we might have expected them to have. But while I agree that
the main cases of epistemic relevance do involve disagreement about belief,
I nonetheless argue that we miss out something important if we confine our
attention to just these cases. In particular, I argue that belief isn’t neces-
sary for disagreement, and offer an alternative proposal about the nature of
disagreements. While some cases of genuine disagreement without conflict
of belief are not epistemically interesting, I argue that there is at least one
exception in this regard. To this end I will be exploring disagreements at the
level of our hinge commitments, which I maintain are not beliefs. As we will
see, understanding the structure of these kinds of disagreements helps us
to understand that certain strategies for resolving them would be hopeless.
Introductory Remarks
Much of the work on the epistemology of disagreement tends, not unreason-
ably, to take such disagreements to be at the level of belief. You believe that
p, and I believe that not-p, and hence we have a disagreement over the truth
of p.1 I say that this focus on disagreements at the level of belief is not unrea-
sonable since the kinds of toy examples, usually concerned with disagree-
ments between (at least putative) epistemic peers, that are employed in this
literature (folks arguing about the split of the restaurant bill and so forth),
certainly are conflicts of belief. Moreover, as I explain below, there are other
good reasons to think that disagreement, certainly of a kind that has any
epistemic import anyway, will tend to be disagreement at the level of belief.
Disagreements, of Beliefs and Otherwise 23
One thing that we will be doing is exploring what it means for something
to be a genuine disagreement. Part of this process will involve saying more
about the notion of belief in play, since I contend that as epistemologists we
have a quite specific propositional attitude in mind. As we will see, under-
standing this fact helps us to appreciate why we would focus on disagree-
ments at the level of belief in the first place, since it is hard to make sense of
the idea of a genuine disagreement that doesn’t involve a conflict of belief in
just this way. I go on to extract one practical consequence of this, which is
that a certain kind of apparent disagreement of which we are all too famil-
iar is not in fact a genuine disagreement at all. This is the phenomenon that
I term dialectical posturing, and which I claim does not involve a conflict
of belief. Accordingly, it does not have the epistemic implications that we
might have expected it to have. We are within our epistemic rights to sum-
marily dismiss the dialectical poseur.
Generally speaking, I will be defending this contemporary epistemologi-
cal focus on disagreements at the level of belief (albeit claiming as a result
that there are fewer genuine disagreements that we might have hitherto
thought). Nonetheless, while I agree that the main cases of epistemic rel-
evance do involve disagreement about belief, I argue that we miss out some-
thing important if we confine our attention to just these cases. In particular,
I will be claiming that belief isn’t necessary for disagreement, in that there is
a special (albeit unusual) class of genuine disagreement that doesn’t involve
a clash of belief. Accordingly, I will offer an alternative proposal about the
nature of disagreements, which slightly tweaks the standard view. While
some of the cases of genuine disagreement without conflict of belief are not
epistemically interesting (since by their nature we tend to have an indepen-
dent epistemic basis to doubt the epistemic credentials of what they claim),
I contend that there is at least one exception in this regard. To this end I will
be exploring disagreements involving a clash between what Wittgenstein
(1969) termed hinge commitments, which I maintain are not beliefs in our
sense. As we will see, understanding the structure of these kinds of disagree-
ments helps us to realize that certain strategies for resolving them would be
hopeless.
Before closing these introductory remarks, let me lay my cards on the
table as regards the wider epistemology of disagreement debate. My favored
view as regards the epistemology of peer disagreement is one that allows a
subject, at least once she has suitably reflected on the disagreement in hand,
to continue to maintain her belief with the same level of conviction as before.
This is a so-called non-concessionary stance, and although it has some
adherents it tends to be unpopular in the literature, though for reasons that
I think aren’t very compelling (for example, that to be non-concessionary
is to fail to display the virtue of intellectual humility).2 This particular issue
about the epistemology of peer disagreement isn’t quite our concern here,
as much of what we will be discussing will be orthogonal to this topic. But,
as we will see, this issue does get an indirect purchase, in that the cases
24 Duncan Pritchard
that we will be looking at are such that, properly understood, there is less
reason to downgrade our judgments in the light of disagreement than one
might have previously thought. Our reconstruction of such cases thus lends
itself to a more general non-concessionary stance to the epistemology of
disagreement. Basically, we are epistemically more entitled to “stick to our
guns” that we might have hitherto thought.
Disagreement and Belief

Belief gets used in lots of different ways, not just by philosophers but also by
practitioners of other disciplines, such as cognitive scientists. There is at least
an expansive understanding of the term such that it simply picks out very
broad propositional attitude of endorsement aimed at a certain proposi-
tion.3 But this expansive rendition of belief strips it of philosophical interest
precisely because it allows too much. Trusting that p, being pathologically
convinced that p, accepting that p, and many other distinct propositional
attitudes besides, will tend to fall into its remit. Our interest, particularly
as epistemologists, is thus in a distinctive propositional attitude that belief
that picks out, and hence is concerned with a more narrow rendition of the
notion. So let us be pluralists about belief, and allow that there are different
notions here that might legitimately be our concern when using this term,
but nonetheless stipulate that we are here focusing on a particular rendering
of the notion. Articulating a complete account of this narrow conception of
belief would take us too far afield for our purposes here, but I think we can
at least pick out three key distinctive features of this conception. (Note that
henceforth when we talk about belief without qualification, it will be this
narrow conception that we have in mind).
The first feature is core to this conception of belief, since arguably the
other two features I will list simply flow from it. It is that belief excludes
agnosticism, in the following sense: to believe that p is to believe that p
is true, and hence is incompatible with being agnostic about the truth of
p. Note that this doesn’t apply to several propositional attitudes in the
broad vicinity of belief. Consider, for example, the notion of acceptance as
it is commonly found in the philosophy of science.4 A scientist might well
accept the truth of a particular theoretical claim because she holds that this
claim is the one that is best supported by the empirical evidence even while
being agnostic about whether that claim is true. The evidence, after all,
might only be marginally in favor of this claim in light of its competitors,
and this might be a field where the evidence is constantly evolving. Accord-
ingly, it might be rational for the scientist to be somewhat skeptical about
the truth of the target proposition, even while actively accepting it (and, for
example, acting on this basis—e.g., by conducting certain experiments as a
result, putting in a particular grant application, etc.). Acceptance is not the
same as belief in our sense.
The second feature is that belief bears some basic conceptual connections
to (epistemic) reasons, again in virtue of the fact that to believe that p is
to believe that p is true. Of course, the conceptual connections are not so
tight as to exclude the possibility of irrational beliefs, since manifestly such
believing is common. But they do exclude the possibility of a propositional
attitude to p counting as a belief if it is in principle unresponsive to rational
considerations. If, for example, one would continue to endorse p even if one
recognized that one has no reason at all for thinking p to be true, then it is
hard to see why we would count such a propositional attitude as a believing
that p, as opposed to a wishful thinking that p, or a hoping that p, etc. The
point is that in such a case one is clearly manifesting a propositional attitude
that has no essential concern for the truth, but since belief is by its nature
directed at the truth, that excludes it from being a believing.
The third feature of belief that I want to focus upon follows naturally
from the second. It is that it is not always a transparent matter what one
believes. In particular, it does not follow from the fact that one sincerely
thinks that one believes that p that one does believe that p. Instead, one
can be mistaken about whether one’s propositional attitude to p is really
one of belief. The point is that belief is at root tied to action; ultimately
one manifests one’s belief via one’s actions. But this has a consequence that
we can discover that what we thought we believed was in fact not a belief
at all. For example, imagine a parent who takes herself to believe that her
child is innocent of the serious charges that have been brought against him.
Now suppose that she discovers, as the evidence of his guilt becomes over-
whelming (even by her own lights), that her conviction in his innocence is
unaffected. Wouldn’t this show that it was not really a believing at all, but
rather some other kind of propositional attitude (e.g., blind faith in one’s
son’s innocence)?
Putting these three features of belief together, I think we have a handle,
albeit an incomplete one, on a distinctive propositional attitude. In particu-
lar, I think it is this propositional attitude that epistemologists are working
with when they talk about belief. For example, it is this particular proposi-
tional attitude that is widely thought by epistemologists to be a constituent
part of knowing that p. (For example, we would not want a propositional
attitude as a constituent part of knowing that p that was compatible with
agnosticism about the truth of p). Furthermore, notice that it is this propo-
sitional attitude that is plausibly in play in the standard examples used to
motivate and illustrate the epistemology of (peer) disagreement literature.
When two epistemic peers disagree over the split for the restaurant bill, for
example, then they are clearly both believing contradictory propositions in
just this sense (e.g., and not merely accepting different propositions). In
particular, what makes this a genuine disagreement is that they each have an
opposing propositional attitude that involves a commitment to an opposing
truth, and as such excludes agnosticism about that truth.
26 Duncan Pritchard
Notice that the foregoing explains why, for example, certain cases of
apparent disagreement are not genuine because the propositional contents
in play aren’t in conflict. That one person claims that so-and-so is “tall”
in one context, while another person claims that the same person is not
“tall” in another doesn’t suffice for them being in disagreement. Perhaps
one of them is employing a more demanding notion of tallness than the
other (e.g., she is a basketball coach). If that’s right, then there need be no
conflict in beliefs here—in that both might agree that this person isn’t tall-
for-a-basketball-player—and hence no genuine disagreement.5
That disagreement requires conflict of belief would also account for why
some other cases involving an oppositional exchange that has the surface
structure of a disagreement don’t count as genuine disagreements. Suppose
that I endorse p, and someone comes along and begins defending not-p.
Imagine, however, that it becomes clear that my disputant is blatantly lying
about not-p, or is clearly just playing around in their defense of not-p, or is
being otherwise insincere in their defense of not-p. I think one would then
naturally say that such exchanges are simply not disagreements at all, even
though they might have the superficial appearance of one.
I think the same applies to some other kinds of exchanges where there is
no deceit of insincerity involved, but also no conflict of belief. For example,
cases when the other party is playing devil’s advocate, or where one’s oppo-
nent is adopting the stance of the Pyrrhonian skeptic (who, recall, is suppos-
edly not committed to the truth of anything, or hardly anything anyway). In
such cases one is rather playing the game of disagreeing, rehearsing through
the moves of a disagreement, rather than actually disagreeing. And the natu-
ral explanation of why is that there is not a conflict of belief here.6 Although
there is no deception or insincerity in such cases, given that such cases don’t
involve a belief in the opposing proposition, they don’t seem best thought
of being disagreements.
Note too that it is only usually genuine disagreement that imposes even a
prima facie epistemic burden on one. That it is clear that your beliefs con-
flict with mine usually offers at least some reason to reflect on the epistemic
standing of my own belief, unless I regard you as completely uninformed or
otherwise deceived about the subject matter in question. (We don’t need to
get into epistemic peerhood here, as the claim in play is so modest. Even if
I don’t think you are my epistemic peer, I think most epistemologists would
accept that so long as I don’t regard you as completely uninformed or other-
wise deceived about the subject matter in question, then I have at least some
reason to reflect on the epistemic standing of my beliefs as a result of this
disagreement. Note too, in this regard, that such reflection needn’t entail
any revision of one’s previous views, though of course it might. Even the
proponent of a non-concessionary view of peer disagreement, like myself,
can consistently grant that such a disagreement demands further reflection).
In contrast, if I think you are merely playing around or lying when offer-
ing your opposing arguments, then I don’t seem to be under any epistemic
burden at all to reflect on the epistemic standing of my beliefs (though of
course I may do so). The devil’s advocate case is slightly different on this
score, in that it is an exchange that prompts me to reflect on the epistemic
standing of my belief, but notice it does so in no way in virtue of there being
a disagreement.
The foregoing makes clear that there is at least a good prima facie ratio-
nale for epistemologists to focus on disagreements of belief. First, because it
seems that it is only when both parties genuinely do believe what they claim
that we have a genuine disagreement at all. Second, because it seems that it
is only genuine disagreements of belief that impose an interesting epistemic
burden on you. As we’ll see, this picture, while undoubtedly broadly on the
right lines, isn’t quite right, in that it leaves out some important cases (so
we will need to move to a slightly different view). Before we get to that,
however, I want to consider one implication of this account of disagreement,
which is that a certain kind of apparent disagreement of which we are all
familiar is not in fact a genuine disagreement at all, at least when properly
understood. (As we will see, this will be a conclusion that will survive our
move to the tweaked alternative view of the nature of disagreement).
Dialectical Posturing
The kind of case that I want to focus on concerns what I will call dialectical
posturing. I contend that a lot of real-world disagreement about conten-
tious matters, such as the existence of human-made climate change or the
desirability of political correctness, is infected with a certain kind of inau-
thenticity. By this I mean that there are parties to the dispute who, far from
expressing their genuine convictions about the subject matter in hand, are
instead merely playing a certain role, wearing a particular dialectical hat, if
you will. They are dialectical poseurs.
Take the debate about human-made climate change as a case in point.
Here we have an overwhelming scientific consensus on the one hand, and
on the other hand various kinds of conspiracy theories. Note that the oppo-
sition has to take the form of a conspiracy theory given that it is lined up
against a scientific consensus, since how else is one to account for why the
scientific community is so unified on this topic?7 But conspiracy theorists
come in many different kinds, and we should be wary of lumping them
together within a single catch-all category.
The debate about climate change is particularly useful when it comes to
telling different kinds of conspiracy theorist apart. The reason for this is
that science is generally regarded as an epistemically paradigm way of set-
tling empirical truths (at any rate, we can reasonably take it as such for
our purposes). Accordingly, rejecting a scientific consensus places epistemic
burdens on one that simply do not arise to the same extent with regard to
other conspiracy theories. Contending that JFK was assassinated by the FBI,
for example, requires one to suppose that there is far less transparency in
28 Duncan Pritchard
U.S. government than one might wish, but such a contention hardly pits one
against a putatively paradigmatic epistemic source. In contrast, arguing that
the scientific consensus is wrong does bring with it just such a commitment.
We can distinguish between two distinct kinds of conspiracy theorist in this
regard in terms of how they respond to this commitment.
On the one hand, there will be those who take up the challenge of making
rational sense of how such a paradigmatic epistemic source could have so
systematically gone awry in this case. Given the nature of the challenge, this
will require a quite considerable investment of one’s epistemic resources. The
opposing position needs to be researched and understood, and a nuanced
line taken as a result. For example, the proponent of this view, aware that
attributing deception on such a massive scale is not psychologically that
plausible, might instead opt for a view on which the raw data, accessible to
a small subset of the scientific world, is engineered to give public data which
subsequently misleads climate scientists at large. This account also has the
added bonus of being able to allow most climate scientists to be acting in
perfectly good faith in advocating human-made climate change. Moreover,
the sophisticated proponent of the opposing position needs a compelling
motivational story as to why anyone would be inclined to manipulate the
public in the fashion (social and political influence, perhaps?). The details
need not concern us. The point is just that there can be sophisticated conspir-
acy theorists, who manage—and remember that this by its nature requires a
great deal of cognitive resources—to actively integrate the conspiracy theory
into the noetic structure of their other commitments, even when such con-
spiracies are opposed to epistemically paradigmatic sources of knowledge.8
My concern here is not with the sophisticated conspiracy theorist, how-
ever, who raise epistemological problems of a very different nature to those
that presently detain us. I’m rather more interested in what I take to be the
more commonly found unsophisticated conspiracy theorist. This is someone
who simply adopts the stance of the conspiracy, like a coat that one puts on
to face a rainy day (but then takes off again when the weather brightens).
The point is that no serious effort is made by this agent to reconcile the
epistemic challenge posed to their other beliefs by their endorsement of the
conspiracy theory (particularly in the case of a conspiracy theory involving
human-made climate change, which goes against the accepted science of
the day). No meaningful attempt at noetic integration with regard to this
conspiracy theory is attempted. It is, instead, more like a dialectical stance
that they adopt in certain conditions, but which is largely disconnected from
their other commitments. That this is so is easily revealed by how little
thought has been given to why such a conspiracy would be undertaken, why
it hasn’t been exposed already, why science can be so generally effective if it
allows high-level conspiracies of this kind, and so on.
I want to suggest that when people propose conspiracy theories in this
fashion, they are not presenting their beliefs at all, and hence any dispute
with them would not qualify as a genuine disagreement. In particular, the
dialectical stance they take is not rooted in their wider actions, where this
includes the intellectual actions of noetically integrating this stance, but
rather stands apart from their other commitments. In this way it has a very
different profile to a genuine belief, even though it may seem superficially
similar. Note that I’m not here disputing that these subjects may well think
that they believe the conspiracy theory that they put forward, but as we
noted earlier, I don’t think that this is enough to settle the question of
whether belief is present anyway. Instead, I maintain that what is taking
place here is a kind of dialectical role-play—not all that different in kind
from the kind of dialectical role-play we looked at earlier (e.g., playing
devil’s advocate), albeit in a less explicit form—in that this kind of propo-
nent of a conspiracy theory simply enjoys taking this sort of radical dia-
lectical line (perhaps, for example, they imagine themselves as iconoclasts,
or at least as non-conformists, unlike the wider, and more gullible, public
around them).9
This is important, because it has epistemological ramifications. As we
saw in the cases of merely apparent disagreement above—where there was
insincerity involved, for example—we don’t seem to be under any epistemic
obligation to reconsider the epistemic credential of our beliefs as a result of
an inauthentic exchange of opinions of this kind. Note that I am not say-
ing here that the dialectical imposter is lying, since from their perspective,
of course, they are expressing opposing beliefs (remember that our notion
of belief allows that thinking you believe that p does not entail a belief
that p). But there is a level of intellectual disengagement with regard to the
opinion expressed—something which is a kind of insincerity, very broadly
speaking—to the point that it is not an expression of opposing belief. As
such, the mere fact that someone is proposing an opposing proposition in
this case does not have any immediate epistemic ramifications. Epistemo-
logically, it is on a par with someone jokily “asserting” not-p in response to
your assertion of p as a joke.10
Conviction and Disagreement

We have just seen that one consequence of our account of belief is that
certain kinds of dispute are automatically disqualified as being bona fide
disagreements, since they do not involve a conflict of beliefs. I noted ear-
lier, however, that the idea of genuine disagreement being disagreement at
the level of belief wasn’t quite right, and I now want to make good on this
claim. For it turns out that there is a class of cases where there are genu-
ine disagreements without conflict of belief, and so we will need to slightly
qualify our understanding of what constitutes a disagreement. Importantly,
however, this qualification does not undermine our point from the previous
section that disputes with a dialectical poseur do not count as a genuine
disagreement, since the cases where disagreement come apart from belief do
not concern the kind of inauthenticity present in this case.
30 Duncan Pritchard
Think about our example from earlier involving the parent who is not
convinced in her son’s guilt. Recall that we described the case such that the
conviction in play is such that it is entirely unresponsive to reasons, such that
it continues even in the face of, as the parent acknowledges, what is over-
whelming counter-evidence. We noted there that this propositional attitude
doesn’t count as a believing, due its complete unresponsiveness to rational
considerations. The parent after all, even by her own lights, has not only
no rational basis for the truth of the target proposition, but overwhelming
evidence against its truth. But her conviction in her son’s innocence remains
unabated. The point is that this is not a mere case of irrational belief, but
is rather an instance where the link between belief and reasons is so cut off
that it no longer counts as a belief at all.
Here is the question: given the genuine conviction in play, if one gets into
a dispute with the parent about the guilt of her son (during the trial, say),
then would this be a genuine disagreement? I think we would naturally
describe it as such. In particular, this case lacks the kind of play-acting and
insincerity that one finds in the other examples where lack of belief leads to
a lack of a genuine disagreement. In contrast, we here have complete sincer-
ity. Indeed, in keeping with our notion of belief, what we have manifested
here is a propositional attitude that excludes agnosticism about the truth of
p, albeit in ways that have nothing to do with reason.
Moreover, although we might think of this conviction on their part as
manifesting some sort of serious mental illness, I think we would more natu-
rally consider it to be simply a manifestation of deep parental devotion. This
is significant, since we might not class disputes with the seriously mental ill
as genuine disagreements, no matter how convinced the patient might be.
For example, imagine you are psychologist treating someone with Cotard’s
delusion, who is ardently maintaining that they are dead. In putting forward
the opposing view, is one disagreeing with them?11 Personally, my intuitions
on this are unclear. I think the differences between this case and our parent
case are significant enough, however, that our hesitancy regarding the for-
mer needn’t carry over to the latter.
What I am suggesting is thus a slight tweak to the account of disagree-
ment in terms of belief that was offered earlier. Rather than belief being
what is necessary for a bona fide disagreement, what’s required is rather
a genuine conviction on each side (i.e., a conviction that p that excludes
agnosticism about the truth of p). This is something that is lacking in the
cases we looked at earlier involving deceit, insincerity, posturing, and so on.
But it can—unusually—be present in cases where one’s sincere conviction
that p manifests in such a way that it doesn’t count as a believing that p.
Now one might object that this is a pointless qualification for the episte-
mologist to make (it wouldn’t be the first!). After all, it looks to be in the
very nature of cases of sincere conviction without the corresponding belief
that we have an independent basis to ignore the opposing testimony. After
all, the two propositional attitudes are precisely coming apart because of
the failure of the subject’s propositional attitude to be remotely responsive
to counter-evidence. Such cases are thus not epistemically interesting, since
while they may be genuine disagreements, they do not generate any of the
epistemic burdens of normal disagreements, and so can be safely set to one
side. Indeed, notice, a disagreement of this sort would be of a very special
kind. In particular, if one realized that this was the nature of propositional
attitude that our parent had, then one would thereby realize that there sim-
ply is no epistemic purpose served in engaging in this dispute (though there
might be other points to it, of course).
I agree that this kind of case of genuine disagreement without conflict of
belief is not epistemologically interesting (qua disagreements anyway). But
as I will explain in the next section, I think there are other plausible cases
of differences of sincere conviction that don’t (just) involve belief that are
epistemologically interesting.
Hinge Commitments
In his final notebooks, published as On Certainty, Wittgenstein (1969) put
forward a radical picture of the structure of rational evaluation. He argued
that it is in the very nature of any rational system that it presupposes essen-
tially arational commitments, what I am going to call hinge commitments.
In particular, Wittgenstein claimed that these hinge commitments must be
in place in order for one to be a genuine believer (or a doubter, etc.) in the
first place. We are optimally certain of our hinge commitments, and that
enables them to function as the fixed points relative to which we rationally
evaluate other claims (identify what is a good reason for what, what is a
good reason to doubt, and so on). What is core to our hinge commitments
is that they manifest our underlying conviction that we are not radically
and fundamentally in error in our beliefs (what I have elsewhere called the
über hinge).12 Interestingly, Wittgenstein showed that this underlying core
certainty manifests itself in our conviction (in normal circumstances) in
apparently mundane claims, such as that one has hands, that the earth is
round, that one’s name is such-and-such, and so on. These look like normal
empirical claims of a kind for which we would have excellent rational sup-
port. In contrast, Wittgenstein argues that their hinge status means that we
could not make sense of the idea of there being reasons in favor or against
such commitments.13
One consequence of this picture is that it entails that the very idea of a
universal rational evaluations (i.e., evaluating the rational status of our com-
mitments as a whole) is impossible, since such a rational evaluation would
include our hinge commitments, and they are the arational fixed points that
enable rational evaluations to occur. This point obviously counts against
certain skeptical lines of thought, as they typically do proceed by attempting
to doubt all of our commitments en masse (think about Descartes’s famous
metaphor of not inspecting the apples one-by-one, but tipping out the whole
32 Duncan Pritchard
barrel). But it is also counts against traditional anti-skeptical proposals too,
as they attempt to rationally evaluate all of our commitments and show
them to be rationally in order.
This is obviously not the place to explore and defend Wittgenstein’s radi-
cal, and controversial, proposal in detail (something that I have done at
length elsewhere).14 For the purposes of this piece I want to rather run with
this idea and show how it could be usefully applied to certain kinds of fun-
damental disagreement.
There are few important things to notice about our hinge commitments.
Although they are essentially arational and involve optimal levels of cer-
tainty, they can change over time, and indeed will usually do so in (indi-
rectly) rational ways. One of the everyday certainties that Wittgenstein
mentions—which like many of them he got from G. E. Moore—is that he
had never been to the moon.15 This looks like an eminently plausible hinge
commitment in Wittgenstein’s day, but there may come a point in the future
where it will lose its hinge status, at least for some. Moreover, notice why
it would change. Remember that one’s hinge commitments express one’s
underlying über hinge commitment that one is not radically and fundamen-
tally mistaken in one’s beliefs. It follows that as one’s beliefs change—which
can happen in entirely rational ways (attending to the relevant shifts in evi-
dence, and so on)—so one’s set of hinge commitments can alter accordingly.
Another point to notice about our hinge commitments is that by their
nature they cannot by their nature be in conflict with our wider set of beliefs,
since they are manifesting our underlying commitment that these beliefs as
a whole are not radically in error. This is important since not everything
that we are optimally certain about is thereby a hinge commitment. In par-
ticular, the kind of pathological certainty that we noted above with regard
to Cotard’s delusion is not like that. The same goes for our loving parent
who cannot accept her child’s guilt, in that she perfectly well recognizes that
there is overwhelming evidence for this guilt, but simply refuses to counte-
nance this possibility nonetheless.16
One thing that our hinge commitments share with these overwhelming
convictions, however, is that they are not beliefs, at least in the specific sense
of belief that we outlined above (and which is of particular interest to epis-
temology). While, like belief, they involve genuine conviction (one cannot
be agnostic about the truth of a hinge commitment), they don’t stand in the
appropriate basic relations to reasons. In particular, it is in the nature of
one’s hinge commitments that they remain even once one recognizes that
one has no rational basis for their truth. This by itself sets them apart from
beliefs in the sense that we are interested.
Can we be radically divergent in our hinge commitments? Wittgenstein
seems to hold that while we can have different sets of hinge commitments,
they are bound to be largely overlapping. The reasons are complex, but
arguably there is something like a Davidsonian principle of charity in
operation here.17 As Wittgenstein puts it at one point: “In order to make a
mistake, a man must already judge in conformity with mankind” (Wittgen-
stein, 1969, §156).
In any case, there can certainly be some divergence in our hinges, and this
is what I want to focus upon.
The most plausible candidate in this regard is, of course, religious convic-
tion. Those who have such conviction have a hinge commitment that those
who don’t have such conviction lack.18 This is a matter of faith, after all,
which in this context contrasts with belief. I think this is the best way of
understanding religious disagreements, where by this I mean full-on debates
between those with religious conviction and those who lack it. Think, for
example, of some the heated debates that took place in response to the so-
called New Atheism movement, with atheists like Richard Dawkins, Chris-
topher Hitchens, and Daniel Dennett lined up against defenders of the (in
this case Christian) faith, such as William Lane Craig and Alvin Plantinga.
At least on the religious side of the debate, this is not a debate about beliefs,
but rather about their hinge commitments.19
Modulo our previous remarks about the nature of disagreement, this
is a bona fide disagreement, even if it involves hinge commitments rather
than (just) beliefs. There is genuine conviction here, on both sides (unlike,
say, the case of the dialectical poseur). Moreover, unlike the parent cases of
genuine disagreement not involving belief from the previous section, this is
not a case where the propositional attitude in play ensures that we thereby
have independent reason to discount the counter-testimony. If Wittgenstein
is right, after all, we all have hinge commitments, and such commitments
are essential if we are to be rational. It is just that the religious believer has
a different set of such commitments to the atheist.
What is most interesting about disagreements of this kind that involve
hinge commitments, however, is that they are not going to be resolved in the
way normal disagreements about belief are resolved. Throwing more and
more arguments and evidence at the religious believer is not going to change
her religious conviction. One’s hinge commitments, recall, are immune to
rational considerations, at least directly. Understanding this point highlights
how pointless a lot of the New Atheism movement was, at least insofar as
its goal was to change the minds of those with faith.20
Does this mean that there is no way to rationally convince those with
religious conviction to change their mind? Not at all. I noted earlier that our
hinge commitments are not directly responsive to rational considerations,
but that they can change over time in indirect response to rational consid-
erations, as one’s set of beliefs change. There is thus a sense in which one’s
hinge commitments are indirectly responsive to rational considerations.
This points us towards the right way to conduct disagreements of this kind.
Rather than a full-on charge that might well be appropriate in normal cases
of disagreement (at least where there is something very important at stake),
one should instead approach these debates in a “side-on” manner. What
I mean by this is that one should look for common ground, and use that
34 Duncan Pritchard
common ground to change the person’s beliefs in relevant ways. If enough
of those beliefs are changed, this could over time impact upon their hinge
commitments.
Note that this means that there is still room for rational persuasion even
in the case of the sort of fundamental disagreement involved in a clash of
worldviews involving different sets of hinge commitments. In particular, it
is not as if by appealing to hinge commitments one is thereby resigning one-
self to epistemically incommensurate epistemic systems which will brook no
rational debate.21
Concluding Remarks
My aim in this chapter has been to show that once we look at some real-world
cases of disagreement, we discover that there are some wrinkles to be added
to the standard ways of thinking about the epistemology of disagreement.
On the one hand, some apparent disagreements are not genuine at all, where
this has important epistemic consequences for how we should respond to
them. On the other hand, there are some disagreements that are bona fide,
but which aren’t best thought of in terms of a disagreement at the level
of belief. The epistemic upshot of this second kind of disagreement is that
we need to approach it in a very particular kind of way, albeit one that—
crucially—doesn’t necessitate abandoning rational processes of persuasion
altogether.
I noted at the outset that my general approach to the epistemology of
disagreement—which isn’t argued for here, it should be admitted—is quite
non-concessionary. I think the cases we have looked at further reinforce, in
a rather piecemeal and partial fashion at least, this general line of thinking.
After all, if I am right, then there is a class of apparent disagreements that
places no serious epistemic burden on one to downgrade one’s previous
opinions. And there is also a second class of genuine disagreements where
there are important limitations on the relevance of rational persuasion (even
though, as we have seen, there is still a role for rational persuasion nonethe-
less). There are thus more potential reasons to stick to one’s guns in the face
of an apparent disagreement than one might have hitherto supposed.22
Notes
1 Here, for example, is how the relevant Stanford Encyclopedia of Philosophy
entry describes such disagreements:
“Suppose that two people form conflicting beliefs about a given question:
one believes p while the other believes not-p” (Goldman & Blanchard,
2015, §3.4).
2 I defend my view in Pritchard (2012a), where I tackle one standard motiva-
tion for rejecting non-concessionary views, which is the “track-record” argu-
ment. See also Kallestrup and Pritchard (forthcoming) where, following Roberts
and Wood (2007), I defend an account of intellectual humility that consists in
manifesting distinctive other-regarding attitudes, a view that is entirely compat-
ible with maintaining one’s position in the light of peer disagreement (unlike
other, more standard, accounts of intellectual humility in the contemporary lit-
erature). Similar views have been defended by Tanesini (2016), and by Priest
(2016) in unpublished work that I have had the pleasure of reading (and seeing
presented). For an influential, but very different, approach to intellectual humil-
ity, see Whitcomb, Battaly, Baehr, & Howard-Synder (forthcoming).
3 And of course there are understandings of the term where it isn’t even a propo-
sitional attitude at all. See Stevenson (2002) for a helpful survey of a range of
different ways in which the notion of belief is employed. Note that I follow
Stevenson in thinking of “belief” as being, following the mathematician and
computer scientist Minsky’s (2007) memorable phrase (applied to “conscious-
ness” rather than “belief”), a “suitcase” term. The idea is that there is not a
single thing in play in peoples’ usage of this term (not just the folk, but also, for
example, cognitive scientists and so forth), but rather a range of interconnected
concepts. Inevitably, there will be those who will argue that there is a core notion
of belief that trumps the others—an interesting recent account of belief, along
pragmatist lines and which does allow beliefs to not be propositional attitudes
(and also, e.g., to be the result of decision), of just this sort is offered in Zimmer-
man (forthcoming)—but I find this general approach both implausible and also
liable to gloss over important distinctions. In any case, I won’t be arguing here
for the suitcase view of belief, but taking it as a given.
4 I am here in particular thinking of the influential way that van Fraassen (1980)
distinguishes between belief and acceptance, as part of a general motivation for
scientific anti-realism.
5 See Frances (2014, ch. 1) for a useful discussion of such apparent disagreement
cases and why they aren’t genuine.
6 Note that it is not being denied here that these mere apparent disagreements
could develop into genuine ones, as when one disagrees with one’s devil’s advo-
cate about whether such-and-such is a good objection to p (with you believing
that it isn’t, and your opponent believing that it is, say). Now you are disagreeing,
but just not about the truth of p itself. (Not yet anyway—but who knows where
this exchange might go next?). Moreover, notice that in such a case your oppo-
nent is no longer playing devil’s advocate on this specific point anyway, as in this
scenario she really does believe that such-and-such is a good objection to p.
7 There is an obvious exception to this, of course, which is a scientist who discov-
ers something hitherto overlooked that calls the current climate data into ques-
tion en masse. Such an agent could reject human-made climate change and yet
not be committed to supposing that there is any conspiracy in play.
8 In terms of the UK media, for example, the journalist (for The Telegraph) Chris-
topher Booker is one example of a prominent skeptic of human-made climate
change who might fit this description. Note, by the way, that this point about
sophisticated conspiracy theorists doesn’t mean that there aren’t dialectical
poseurs as regards, say, the JFK conspiracy—clearly there are—but only that
they will be harder to spot, as the threshold for being a sophisticated conspiracy
theorist in this case will be lower.
9 Dialectical posturing is closely related to Frankfurt’s (2005) notion of bullshit,
which is characterized by a lack of concern for the truth. The bullshitter likely
isn’t presenting their beliefs either, and relatedly any dispute with them is not
a genuine disagreement, nor does it impose any epistemic burden upon us to
reflect on the truth of our beliefs. There might be some differences though, as
Frankfurt’s view, as I understand it anyway, involves the bullshitter actively try-
ing to persuade you of their position. In contrast, our dialectical poseur prob-
ably doesn’t care whether they convince you of anything; indeed, it will likely be
36 Duncan Pritchard
important to their self-conception that they express views that most others do
not share.
10 Note that I am not denying here that one might be (reasonably) intrigued by the
counterargument and on that basis revise one’s conviction in the target proposi-
tion. But the point is that this epistemic import to your belief that p is not aris-
ing out of the disagreement per se, but out of your willingness to run with the
counterargument (which you are in your epistemic rights to dismiss tout court).
Some epistemologists have suggested that the mere raising of an error-possibility
makes it relevant, in the sense that one needs to rationally discount it thereafter
(this is a guiding thesis of attributer contextualism—see, e.g., Lewis (1996)). But
I don’t think this kind of claim is at all plausible on closer inspection; we need a
reason to take error-possibilities seriously (such as that so-and-so believes that it
is so). (See Pritchard (2010b) for my take on some of these issues, in the specific
context of perceptual knowledge).
11 Note that the way that I describe the phenomenon is very superficial. In actual-
ity, how best to describe the propositional attitude of those subjects who suffer
from the Cotard delusion is very difficult, so this presentation is very much a
caricatured one, simply for the purposes of this chapter. For a useful overview of
the literature on Cotard’s delusion from a psychological perspective, see Berrios
and Luque (1995). (Note especially their use, in common with the literature on
this topic, of the notion of belief to describe the deluded subject’s propositional
attitude. They clearly do not have our more specific notion of belief in mind).
12 See especially Pritchard (2016b).
13 Does that mean that one can’t have rationally grounded knowledge that, say,
one has hands? Although one might put the point that way, it would be a very
misleading gloss, since on this view it is not as if one is ignorant of anything in
failing to have rationally grounded knowledge of this proposition. The crux is
rather that our hinge commitments are not in the market for rationally grounded
knowledge in the first place.
14 See, especially, Pritchard (2016b). See also Pritchard (2012a, 2016b). For some
key alternative readings of Wittgenstein in this regard, see Williams (1991),
Moyal-Sharrock (2004), Coliva (2015), and Schonbaumsfeld (2017). For a
recent survey of work on this topic, see Pritchard (2017).
15 See especially Moore (1925), but also Moore (1939). Arguably, some of the
examples that Wittgenstein employs are from Newman (1979 [1870]). See Kien-
zler (2006) and Pritchard (2015) for some discussions of the (largely unacknowl-
edged) influence that Newman’s work had on the later Wittgenstein, particularly
the notebooks that make up On Certainty. See also endnote 18.
16 This is also the reason why our hinge commitments are not aliefs either (Gendler,
2008a, 2008b). In particular, aliefs can be in tension with what you believe, as
when one believes that flying is safe but alieves that one is in danger in flying.
This cannot be true of one’s hinge commitments.
17 See, especially, Davidson (1983). For discussion of Davidson’s view in this con-
text, see Pritchard (2013). For more discussion of how Wittgenstein deals with
epistemic relativism, see Pritchard (2010a).
18 Indeed, I think one of the chief concerns that Wittgenstein had in developing his
hinge epistemology—and this relates to the influence of Newman (1979 [1870)
noted in endnote 15—is in applying it to religious conviction. For more on the
application of hinge epistemology to the epistemology of religious belief—which
I argue leads to a distinctive position that I call quasi-fideism—see Pritchard
(2011, 2015, forthcoming a; forthcoming b).
19 Why only one this side of the debate? The reason is that I don’t want to take
a stance here on whether atheism is a belief or a hinge commitment (it doesn’t
matter anyway, since as we saw earlier, there are cases of non-genuine disagree-
ment where only one side lacks belief in the target proposition, such as our
dialectical poseur).
20 Construed more charitably, it was rather aimed at those undecided about the
matter. For a philosophical survey of the New Atheism movement, see Taylor
(2017).
21 For further discussion of the relationship between hinge epistemology and epis-
temic incommensurability/epistemic relativism, see Williams (2007), Pritchard
(2009, 2010a), and Kusch (2016).
22 Thanks to Sven Bernecker and Aaron Zimmerman for useful discussions on
topics related to the chapter. Special thanks to Casey Johnson for detailed com-
ments on an earlier version of this chapter. This chapter has benefitted from three
grants awarded by the Templeton Foundation, all of them for projects hosted
at the University of Edinburgh’s Eidyn research center. These are: (i) the ‘Virtue
Epistemology, Epistemic Dependence and Intellectual Humility’ project, which
was itself part of the wider ‘Philosophy and Theology of Intellectual Humility
Project’ hosted by Saint Louis University; (ii) the ‘Intellectual Humility MOOC’
project; and (iii) the ‘Philosophy, Science and Religion Online’ project. In addi-
tion, thanks also to the University of Connecticut’s Humanities Institute, where
I was a visiting professorial fellow while working on this chapter, and where I also
participated in the ‘Humility and Conviction in Public Life’ project, which is also
funded by the John Templeton Foundation.
References
Berrios, G. E., & Luque, R. (1995). Cotard’s delusion or syndrome? A conceptual
history. Comprehensive Psychiatry, 36, 218–223.
Coliva, A. (2015). Extended rationality: A hinge epistemology. London: Palgrave
Macmillan.
Davidson, D. (1983). A coherence theory of truth and knowledge. In D. Henrich
(Ed.), Kant oder Hegel? Stuttgart: Klett-Cotta.
Frances, B. (2014). Disagreement. London: Polity.
Frankfurt, H. (2005). On bullshit. Princeton, NJ: Princeton University Press.
Gendler, T. S. (2008a). Alief and belief. Journal of Philosophy, 105, 634–663.
Gendler, T. S. (2008b). Alief in action (and reaction). Mind and Language, 23,
552–585.
Goldman, A., & Blanchard, T. (2015). Social epistemology. In E. Zalta (Ed.), Stan-
ford encyclopaedia of philosophy. Retrieved from https://plato.stanford.edu/
entries/epistemology-social/.
Kallestrup, J., & Pritchard, D. H. (Forthcoming). Intellectual pride and intellectual
humility. In J. A. Carter & E. Gordon (Eds.), Pride. Totowa, NJ: Rowman &
Littlefield.
Kienzler, W. (2006). Wittgenstein and John Henry Newman on certainty. Grazer
Philosophische Studien, 71, 117–138.
Kusch, M. (2016). De-idealizing disagreement, rethinking relativism. manuscript.
Lewis, D. (1996). Elusive knowledge. Australasian Journal of Philosophy, 74, 549–567.
Minsky, M. (2007). The emotion machine: Commonsense thinking, artificial intel-
ligence, and the future of the human mind. New York: Simon & Schuster.
Moore, G. E. (1925). A defence of common sense. In J. Muirhead (Ed.), Contempo-
rary British philosophy (2nd series). London: Allen & Unwin.
38 Duncan Pritchard
Moore, G. E. (1939). Proof of an external world. Proceedings of the British Acad-
emy, 25, 273–300.
Moyal-Sharrock, D. (2004). Understanding Wittgenstein’s on certainty. London:
Palgrave Macmillan.
Newman, J. H. (1979 [1870]). An essay in aid of a grammar of assent. Notre Dame,
IN: University of Notre Dame Press.
Priest, M. (2016). Intellectual virtue: An interpersonal theory. manuscript.
Pritchard, D. H. (2009). Defusing epistemic relativism. Synthese, 166, 397–412.
Pritchard, D. H. (2010a). Epistemic relativism, epistemic incommensurability and
Wittgensteinian epistemology. In S. Hales (Ed.), Blackwell companion to relativ-
ism (pp. 266–285). Oxford: Blackwell.
Pritchard, D. H. (2010b). Relevant alternatives, perceptual knowledge, and discrim-
ination. Noûs, 44, 245–268.
Pritchard, D. H. (2011). Wittgensteinian Quasi-Fideism. Oxford Studies in the Phi-
losophy of Religion, 4, 145–159.
Pritchard, D. H. (2012a). Disagreement, scepticism, and track-record arguments. In
D. Machuca (Ed.), Disagreement and Scepticism. London: Routledge.
Pritchard, D. H. (2012b). Wittgenstein and the groundlessness of our believing. Syn-
these, 189, 255–272.
Pritchard, D. H. (2013). Davidson on radical skepticism. In E. LePore & K. Ludwig
(Eds.), Blackwell companion to Donald Davidson (pp. 521–533). Oxford: Blackwell.
Pritchard, D. H. (2015). Wittgenstein on faith and reason: The influence of New-
man. In M. Szatkowski (Ed.), God, truth and other enigmas (pp. 141–164). Ber-
lin: Walter de Gruyter.
Pritchard, D. H. (2016a). Epistemic angst. Philosophy and Phenomenological
Research. Online First, doi:10.1111/phpr.12280.
Pritchard, D. H. (2016b). Epistemic angst: Radical skepticism and the groundless-
ness of our believing. Princeton, NJ: Princeton University Press.
Pritchard, D. H. (2017). Wittgenstein on hinge commitments and radical scepticism
in on certainty. In H.-J. Glock & J. Hyman (Eds.), Blackwell companion to Witt-
genstein (pp. 563–575). Oxford: Blackwell.
Pritchard, D. H. (Forthcoming a). Quasi-fideism and religious conviction. In M.
Szatkowski (Ed.), Epistemology of religious belief. Berlin: DeGruyter.
Pritchard, D. H. (Forthcoming b). Faith and reason. Philosophy.
Roberts, R. C., & Wood, W. J. (2007). Intellectual virtues: An essay in regulative
epistemology. Oxford: Oxford University Press.
Schonbaumsfeld, G. (2017). The illusion of doubt. Oxford: Oxford University Press.
Stevenson, L. (2002). Six levels of mentality. Philosophical Explorations, 5, 105–124.
Tanesini, A. (2016). Intellectual humility as attitude. Philosophy and Phenomeno-
logical Research. Online First, doi:10.1111/phpr.12326.
Taylor, J. E. (2017). The new atheists. In J. Fieser & B. Dowden (Eds.), Internet
encyclopaedia of philosophy. Retrieved from www.iep.utm.edu/n-atheis/.
van Fraassen, B. (1980). The scientific image. Oxford: Clarendon Press.
Whitcomb, D., Battaly, H., Baehr, J., & Howard-Snyder, D. (2017). Intellectual humil-
ity: Owning our limitations. Philosophy and Phenomenological Research, 94(3),
509–539.
Williams, M. (1991). Unnatural doubts: Epistemological realism and the basis of
scepticism. Oxford: Blackwell.
Williams, M. (2007). Why (Wittgensteinian) contextualism is not relativism. Epis-
teme, 4, 93–114.
Wittgenstein, L. (1969). On certainty. (G. E. M. Anscombe & G. H. von Wright,
Eds. & D. Paul & G. E. M. Anscombe, Trans.). Oxford: Blackwell.
Zimmerman, A. (Forthcoming). Belief: A pragmatic picture. Oxford: Oxford Uni-
versity Press.
3 Dissent
Ethics and Epistemology1
Sanford C. Goldberg
Overview: The Pressure to Conform in Signaling Practices

In this chapter I want to explore the pressure that certain social regularities
generate for members of the relevant community to conform to the regular-
ity. I will be particularly interested in those regularities that (i) are familiar
to members of the community and which (ii) involve signals, which is to say,
where the presence or absence of a certain feature—a behavior, an exhibited
characteristic, etc.—reliably correlates with some further condition, and so
carries information regarding the likely presence of that further condition
to observers who are knowledgeable of the regularity. The ultimate thesis
of this chapter is that in scenarios with familiar social regularities involving
signals, certain individuals who do not conform to the regularity have some
reason to indicate this publicly. The individuals in question are those who
might reasonably be regarded as participating in the practice. The argument
for this conclusion will appeal to certain relatively weak ethical and epis-
temic principles. If this is correct, then such principles might be seen to lie at
the basis of a plausible ethics of dissent.
Two Social Practices and a Question

During the baseball season in Chicago, at least on the north side, it is com-
mon practice for families to prominently display a “W” flag outside their
home after every Cubs’ victory (to be taken down the next day and in any
case no later than just prior to the start of the next game).
FLYING THE “W”

People display the “W” flag immediately following every Cubs’ victory;
they tend to take it down the next day, or prior to the next Cubs’
game, whichever comes first.
I would speculate that this practice dates back to an era when the chal-
lenge of getting up-to-date news was greater than it is now: originally this
was a way to spread the word quickly that the Cubs had won. Interestingly,
Dissent: Ethics and Epistemology 41
this practice still persists today in the age of immediate information at our
fingertips. What this means is that on game days when the Cubs win, even
those who are oblivious to the various sources of news can quickly come
to learn that the Cubs won their last game, by observing the various “W”
flags. To be sure, not everyone in town participates in this practice. (I am a
Yankees’ fan; my family doesn’t fly the “W.”) Still, enough do, so that (if one
picks a time after a Cubs victory) on any ten-minute walk around my neigh-
borhood in Evanston (just north of Chicago) one can expect to encounter
more than a dozen families who do.
We might then think of this as a rudimentary way to determine whether
the Cubs won their last game: take a walk around the neighborhood, and
look for the displayed “W” flags. If you see one, you can be reasonably
confident that they won their previous game. If you see none, then you can
conclude that either the Cubs haven’t had a game in the last 24 hours, or
else they lost—with the result that if you know that they did play last night,
you can conclude that they lost. In this way, the displaying of the “W”
flag sends the signal that the Cubs won their previous game within the last
24 hours. (Small caveat: this practice was thrown out of whack in the fall
of 2016 when—incredible as it still seems—the Cubs won the World Series,
prompting people to display the “W” flag throughout the off-season.)
To be sure, there is no pressure on anyone to conform to this practice.
No one is expected to participate; no one is expected to make it clear when
they don’t participate. If someone were to get upset with my family for fail-
ing to fly the “W” after a Cubs win, that person’s complaint would fall on
deaf ears. We were under no obligation, we didn’t “owe it” to anyone to
participate, and not being Cubs’ fans ourselves we had no impulse to do so.
I suspect that if we were to buy a “W” flag and fly it on occasion, then we
might be in a different position. Suppose that we buy one and mischievously
fly it whenever we felt like it, whether the Cubs won or lost. In that case,
I can imagine being mildly chastised by our neighbors for failing to conform
to the practice. (I can even imagine being rather forcefully called out by one
of our neighbors who is a life-long, die-hard Cubs’ fan, and who takes his
fandom quite seriously.) To be sure, it’s a free country, and we are legally
permitted to fly that flag as we wish. Even so, here I am somewhat sympa-
thetic with our would-be critics: if there is a familiar up-and-running social
regularity, then there is something rather perverse about exhibiting the trap-
pings of participation while failing to conform to the regularity itself. Still,
whether or not you agree with me on this, I suspect you will agree that my
family is under no pressure to conform to the practice in the first place.
It is interesting to compare the practice of flying the “W” with other
social practices where, it seems to me, there is some pressure on individuals
to conform to the practice, or at least to make it public when they do not so
conform. Imagine a community of middle school students that have broken
up in to different cliques or gangs, each one with its own color. Suppose
that everyone around follows this practice: members are hyper-reliable in
42 Sanford C. Goldberg
conforming to the practice of wearing their group’s colors. Imagine that
there is a group whose practice is that each Monday every member shows
up to school wearing magenta:
MAGENTA MONDAYS
Members of the Magenta gang (as they are called) regularly wear
magenta shirts on Mondays. Magenta being a rare color for a shirt,
few others in this community ever wear Magenta shirts—on Monday
or indeed on any day of the week.
In this scenario, magenta-on-Monday is a hyper-reliable signal of member-

ship in the Magenta gang.
Indeed, it is easy to imagine that once this practice is in place, and had
become widely familiar, there is some pressure on everyone to conform.
There is pressure on gang members to wear magenta on Mondays, and there
is pressure on non-members to avoid doing so. In this respect, the practice
in MAGENTA MONDAYS differs from that in FLYING THE “W.” If this
is correct, we might wonder why that is. What’s the difference?
The Appearance of Pressure to Conform to Ongoing

Social Practices
The main difference, I think, should be clear. In particular, the cases differ
in an important way in connection with the information that is “signaled”
in them. In FLYING THE “W,” the information concerns the Cubs’ base-
ball team, and anyone who flies the flag properly can convey all the infor-
mation that can be conveyed. Assuming that there is a single household
that reliably and competently participates in the practice, one can learn
(in the sense of come to know) all that there is to be known by observing
that family’s behaviors under conditions in which one was aware of their
reliability. That other people fly the flag only makes it easier to acquire the
information quickly in various locations; but what is learned is the same
in each of these locations. By contrast, in MAGENTA MONDAYS, the
information concerns the wearer of the shirt her- or himself. Consequently,
it is not true that there is any one person whose shirt color on Mondays
sends a signal whose content exhausts the information one can learn on
this score. Even if Melissa wears magenta on Monday, this signal does not
bear on Danit’s status in the gang—only the color of the shirt Danit wears
on Monday does that.
I suspect that this difference explains our intuitions about the (lack of)
pressure there is to participate in the practices.
In a case where anyone can send a signal that exhausts the information
at issue, then once a signal is sent by someone, it would seem that the infor-
mation has been given. To be sure, if the signal was sent through a channel
to which only certain observers had access—say, the Salazar family flew
the “W,” but they live in a gated community, and those who don’t live
in that community are not permitted to enter—then the efficient spread of
this information via signaling will require others to do so. But this is not
because more information can be signaled, but because the signal that was
sent (which exhausted the information to be signaled) didn’t reach every-
one. Insofar as restrictions in the signal’s channel are not at issue, it would
seem that once a signal has been sent there is no rationale for others to do
so as well—at least no rationale pertaining to our interest in the information
at issue (whether the Cubs won their most recent game).2
Next, consider the case where each individual participant can signal dif-
ferent information—that is, information with a different (presumably logi-
cally independent) content. And suppose too that what is wanted is the
information at issue (here, who is in the gang and who is not), and that the
information at issue includes information that only one oneself can s ignal—
so that the information conveyed by others’ signals is irrelevant to the
information that would be sent by one’s own signal. (I will use ‘proprietary
information’ to designate the sort of information that pertains to oneself.) In
that case, insofar as we want the full information, it would appear that there
is pressure on everyone to play her/his crucial role—i.e., to provide what
only she can provide—and so (given the desire for complete knowledge) it
would appear that there is pressure on everyone to conform to the practice.
To be sure, this appearance of pressure to conform may be misleading; and
even if it isn’t, the pressure need not be decisive (perhaps others have even
stronger reasons for resisting conformity). Still, there is the appearance of
pressure to conform, deriving from the desire people have to know.
This leads us to the following hypothesis, which I will designate “APTC”
for the “Appearance of Pressure to Conform”:
APTC Given a familiar social practice involving signaling, where each

participant is such that her signal carries proprietary information
that is logically independent of the information carried by the sig-
nals sent by other participants, then there appears to be pressure on
individuals to conform to the practice, deriving from the common
interest in acquiring the knowledge in question.
I do not intend for APTC to settle the question whether this appearance of
pressure corresponds to reality, i.e., whether there really is pressure; for all
APTC says, the appearance may be illusory with respect to whether there
really is any such pressure. APTC merely records something like a datum for
which we would like to be able to account, leaving open the possibility that
the best account will explain this datum away.
In what follows I am going to be arguing that appearances are not illu-
sory here—there really is pressure to conform. My claim will be that this
derives from certain weak (and hence very plausible) ethical and epistemo-
logical principles. After arguing for this, I will go on to ask about how many
practices fall within APTC. It turns out quite a lot; for in a good many
practices part of the content of the signal concerns the signaler herself—in
particular, what she herself believes on related matters.
The Ethical and Epistemic Basis of Pressure to Conform

In the previous sections, I have been speaking of social practices. This talk
has been proper, given that the regularities in question involve norms, and
that the participants themselves hold one another accountable to the norms
(or at least they would be entitled to do so). However, I believe that if we are
to explain APTC, we will need to take a step back. We will need to consider
things that are not already practices with up-and-going norms, but where
there are what we might think of as mere behavioral regularities in play.
Consequently, I will talk about familiar social-behavioral regularities; and
part of my reason for doing so is an interest in seeing how a mere behavioral
regularity might itself become a social practice (governed by norms).3
I want to focus on behavioral regularities that involve signaling. In the
cases that follow, the information being sent by the signal can but need not
be intended by the sender to be communicated. I start with a schematic
description of the sorts of case I have in mind.
Imagine that a number of people start to behave in a certain recognizable
way W, and that there is a very high correlation between behaving in way
W and exhibiting feature F. That is to say, a high percentage of those who
exhibit behavior W also exhibit feature F (and we can suppose that this is
a much higher percentage of people-who-are-F than what is found in the
population at large). To be sure, this is a mere regularity: there are no norms
here, just an empirical regularity in behavior. In this case, behaving in way
W signals (an increased likelihood of) being F, in the sense that behaving in
that way carries information about this increased likelihood. What is more,
people might come to recognize the reliable connection between behaving
in way W and being F. A concrete case of this sort might be this: it is
familiar to many that those who bungee-jump for fun are likely to have
risk-taking personalities. This is not normative, in the sense that there is
no norm whose standard requires people who bungee-jump for fun to be
risk-takers; a person who bungee-jumps for fun but does not have a risk-
taking personality is not violating any rule or standard (e.g., for which she
might deserve censure). But it is a behavioral regularity, and this fact itself
is familiar to most of us.
Alternatively, imagine that there are behaviors W* such that if there is
anyone in a population that exhibits that behavior, this is correlated with
the obtaining of some fact F*. That is to say, a high percentage of those
communities in which one or more person(s) behave(s) in way W* are
communities in which F* holds (and again we can suppose that this is a
higher percentage than what is found in the population of communities at
large). In this case, behaving in way W* signals (an increased likelihood of)
the obtaining of F*, in the sense that behaving in that way carries informa-
tion about this increased likelihood. What is more, people might come to
recognize the reliable connection between behaving in way W* and the
obtaining of fact F*. A concrete case of this sort might be this: those com-
munities in which some person(s) buy(s) kosher food are likely to be com-
munities with a sizable Jewish population. This is not normative, in the
sense that there is no norm whose standard requires communities to be
like this; a community in which there is a lone person buying kosher food,
but no Jews present, is not violating any rule or standard (e.g., for which
censure might be appropriate). But it is a behavioral regularity, and this fact
itself is familiar to most of us.
Now it is clear in both cases that there is no pressure on anyone to behave
in way W or W*. No one is under any pressure to bungee-jump, or to buy
kosher food. (At least no one is under such pressure merely in virtue of being
an individual in an arbitrary community in which there are people who
bungee-jump, or who buy kosher food. In the latter respect, I am imagining
that an arbitrary community is a diverse one, not an Orthodox Jewish com-
munity.) How might it come to pass that there can be pressure on people to
behave in the ways in question? The short answer is that this can come to
pass if (i) non-conformity itself also sends a signal, (ii) the signals sent by
individuals send proprietary information, and (iii) it is reasonable to sup-
pose that there is a general interest in having the information in question.
After presenting some cases in which this is the case, I will highlight what
I regard as the rather weak epistemic and ethical principles that underwrite
this conclusion.
Consider the following case, in which (I contend) there will be pressure
on the individuals to conform to the behavioral regularity. I call this case
EMERGENCY ASSISTANCE
In this community, people regularly confront various types of situations
in which they find themselves in need of emergency aid. The aid itself
is of a sort that requires certain specialized forms of help (involving a
combination of skills, strength, endurance, and expertise) which only
a few select people in the community—the “helpers”—are in a posi-
tion to offer. It is not obvious just by looking who the helpers are. Still,
a good many of them have adopted the habit of wearing a shirt of a
particular color. This is helpful as it enables people to quickly identify
these individuals as helpers, and it enables people to avoid the mistake
of misidentifying someone who isn’t a helper (as a helper). However, the
community itself is one in which there is a deep and abiding resistance
to dress codes of any kind. Consequently, there is no rule on the books
mandating attire for helpers (or anyone else). The wearing of shirts of
this color, then, is in this community a mere behavioral regularity rather
than an up-and-going social practice with norms.
Here, the wearing of a shirt of the relevant color signals (a higher likelihood)
that the individual who is wearing it is a helper. Notice that this information
is useful to have: in the community there is a need for a quick and reliable
way to identify the helpers, and this does it. It would seem here that once
this regularity is sufficiently robust4and sufficiently salient to people in the
community,5 there will be some pressure on everyone to conform to this
regularity. That is, helpers will be under some pressure to show up in a shirt
of the relevant color, and non-helpers will be under some pressure not to
show up in a shirt of that color.
This verdict seems intuitive and plausible. What might explain it? I believe
that it can be explained by six rather weak (and so plausible) principles, four
from ethics, and the other two from epistemology.
I will begin with two rather weak principles in ethics. I designate the first
as the “Reasons to Avoid Anticipated Harm” Principle. It reads as follows:
RAAH If S reasonably anticipates that φing will have harmful effects on

others, then S has some reason not to φ.
I regard RAAH as highly plausible. It merely speaks of S’s having “some

reason” to avoid φing; it does not say how strong this reason is, nor does it
say what it would take for there to be overriding reasons in support of φing
despite the anticipated harmful effects. If we understand “some reason” in
this weak way, as I intend, then RAAH itself seems almost undeniable. (It
appears to be implied by the stronger claim that, all else equal, we shouldn’t
harm others.) Moving on to the second principle in ethics, I designate it as
the “Reasons to Help” Principle, as follows:
RTH If S recognizes that A needs help, then S has some reason to

help A.
I regard RTH as highly plausible. It merely speaks of S’s having “some rea-
son” to help A; it does not say how strong this reason is, nor does it say
what it would take for there to be overriding reasons in support of not
helping. In particular, we might think that the reason RTH generates is suf-
ficiently strong to require S to help A if S can provide this help at little or no
cost to herself, but perhaps not if the cost of doing so outweighs the benefits
to A. In any case RTH is silent on this matter. So it is compatible with any
plausible view one might have regarding the duty to help others. It is also
compatible with the view that one’s duty to help others is an imperfect duty:
perhaps the reasons to help are to be evaluated together, as a set, so that the
duties or obligations one has in this regard are restricted by the help one has
provided to others. (In any case RTH is implied by the stronger claim that
we should help others when we can do so.)
As for the third and fourth ethical principle, both involve connecting doing
things with good/bad effects with doing things that help/harm. Starting
with the bad/harm connection first, we might formulate the “Badness-
Harm Link” (or “HARM” for short) as follows:
HARM To knowingly do something that one can anticipate will have

bad effects on someone is, to that degree, to knowingly do something
that one can anticipate will harm that person.
Moving on to the good/help connection, we might formulate the “Goodness-

Help Link” (or “HELP” for short) as follows:
HELP To knowingly do something that one can anticipate will have

good effects that the subject could reasonably be expected to want
is, to that degree, to knowingly do something that one can anticipate
will help that person.
These, too, seem most plausible, as they merely connect doing things with
good and bad effects with doing things that help and harm.
I move now to the two epistemic claims, which advance two rather weak
epistemic principles which I designate as the “Epistemic Badness of False
Belief” and the “Epistemic Goodness of True Belief” principles. They are
as follows:
EBFB False beliefs are bad from an epistemic point of view.

EGTB True beliefs are good from an epistemic point of view.
EBFB does not imply that false beliefs are bad all things considered, nor
does EGTB imply that true beliefs are good all things considered. They
merely say that, from the epistemological point of view, false beliefs are
bad, and true beliefs are good. Both of these seem uncontroversial. Con-
sider a familiar desideratum on attempts to characterize epistemic justifica-
tion: most epistemologists will agree that it is a desideratum to preserve
the idea that there is an intimate connection between epistemically justified
belief and belief that is likely to be true, or that one has reason to believe is
true. The combination of EGTB and EBFB would appear to underwrite this
desideratum. Now if EBFB is correct, then it stands to reason that whatever
generates false beliefs is, to that extent at least, something that has a bad
epistemic effect; and if EGTB is correct, then it also stands to reason that
whatever generates true beliefs is, to that extent at least, something that
has a good epistemic effect. Correspondingly, to do something which one
can anticipate will most likely lead another to acquire a false belief, is to
do something which one can anticipate will most likely involve having an
effect that is, at least to that extent, epistemically bad;6 and so too, mutatis
mutandis, in the case of something that has a good effect.
But now if we put these six claims together, we reach the following two
conclusions. First, to knowingly bring it about that another’s epistemic
perspective is worse than it had been is to knowingly do something that one
can reasonably anticipate harms that person, epistemically speaking. This
result, of course, puts us in a position to use RAAH to conclude that one
has some reason not to do such a thing. Second, to knowingly bring it about
that another’s epistemic perspective is better than it had been is to know-
ingly do something that one can reasonably anticipate helps that person,
epistemically speaking. This result, of course, puts us in a position to use
RTH to conclude that one has some reason to do such a thing.
I submit that these four ethical principles (RTH and RAAH, as well as
HELP and HARM) together with the two epistemic principles (EBFB and
EGTB), can be used to defend the claim that in EMERGENCY ASSIS-
TANCE people are under normative (moral) pressure to conform to the
behavioral regularity. In that scenario, the mere behavioral regularity is both
robust (no one but helpers ever wear a shirt of the relevant color) and salient
to the members of the community (people are aware that several helpers
wear shirts of this color, and that no one else does). As a result of this, if an
arbitrary community member observes a person not wearing a shirt of this
color, the community member will likely downgrade the probability that
the observed person is a helper. (The observed person will be regarded as a
helper only by those in the community who already know on independent
grounds that the observed person is a helper.) So, too, if an arbitrary com-
munity member observes a person who is wearing a shirt of this color, the
community member will most likely regard the observed person as a helper.
(Again, this judgment will likely be made at least by those in the community
who do not already know that she is a helper).
To get a sense of the plausibility of the principles in application, it is
worthwhile reflecting on cases in which the signal sent by the shirt color
was misleading- the person isn’t as the signal would lead an observer to
believe- with the result that an observer is mistaken. There are two cases to
consider: false positives (in which a shirt of the relevant color is worn by a
non-helper), and false negatives (in which a helper is not wearing a shirt of
the relevant color). I will take these up in order.
In the former, (false positive) case, an observer who relies on the signal
from the shirt color both failed to regard a helper as a helper. Here we
need to consider two variants: one in which the observer has a false belief
to the effect that the person in question isn’t a helper: and in the other the
observer simply lacks the true belief to the effect that the person in ques-
tion is a helper. (We can treat cases in which the community member has a
diminished credence in the hypothesis that the person is a helper as falling
into the latter variant).
Consider the false belief variant first. In this case, the helper in question
has done something regarding which she might reasonably have anticipated
would have had such a result. After all, it is common knowledge that the
behavioral regularity is robust, so it would have been reasonable to suppose
that not wearing a shirt of the relevant color risks leading others to believe
that one isn’t a helper. Given the Epistemic Badness of False Belief principle,
EBFB, together with the Badness-Harm Link, HARM, we can strengthen
this result: the helper has done something regarding which she might rea-
sonably have anticipated would have a harmful effect. And given the Rea-
sons to Avoid Anticipated Harm Principle, RAAH, she has a reason to avoid
doing so. QED.
Consider next the lack of true belief variant. Here, we can reach the same
conclusion, not through RAAH, but through the Reasons to Help principle,
RTH. Here, we might say that given that it is common knowledge that
members of the community often find themselves in need of aid which only
a select few can provide, it is reasonable to think that people would like to
know who can provide this aid. That is, it would be helpful to provide this
information. And since it is also reasonable to think that not wearing the
shirt would leave those not already in the know ignorant about one’s status
as a helper, it is reasonable to think that those not already in the know need
such help. So given RTH, our helper has a reason to provide that help, and
so has a reason to conform to the shirt-wearing regularity. QED.
Reflecting on both variants of this case, we can see that there is some
normative pressure to conform to the behavioral regularity, and we can
appreciate the source of this pressure. The pressure itself is an instance of
the sort of pressure on us to help and avoid harming others, as these ethi-
cal principles are applied to cases in which the harms are at least partly
epistemic in nature. To be sure, the felt pressure might increase or decrease
according to what is at stake: in a case where the absence of true belief or
presence of false belief can be life-threatening, the pressure to conform will
be great; whereas in a case where the absence of a true belief or presence of
a false belief is highly unlikely to lead to any further harms, the pressure will
be minimal. Still, I submit that the pressure is there even in the latter case;
even here individuals have some reason to conform to the behavioral regu-
larity, on pain of bringing about anticipatable harms of an epistemic kind.
We might also note that in such cases the behavioral regularity is one
regarding which the three conditions I mentioned above, (i)-(iii), are met.
For one thing, (i) non-conformity itself sends a signal. Here, non-conformity
takes the form of dressing as one likes (so that helpers might well fail to
wear a shirt of the color in question, or non-helpers might well wear such a
shirt). Clearly, both deviations from the practice send a signal: helpers who
fail to wear a shirt of this color are “signaling” that (there is an increased
likelihood that) they are not a helper, whereas non-helpers who wear a shirt
of this color are “signaling” that (there is an increased likelihood that) they
are a helper. Second, (ii) the signals sent by individuals send proprietary
information. This much should be obvious. And third, (iii) it is reasonable
to suppose that there is a general interest in having the information in ques-
tion. This much was built into the story.
The Role of Practical and Epistemic Reasons
in the Pressure to Conform
When I claim that we have “some reason” to conform to behavioral regu-
larities under the conditions just described, what is the nature of the reason
itself? My claim is that it is a practical reason to avoid doing something one
has epistemic reason to think will have harmful consequences (where these
consequences include, but may not be limited to, epistemic harms). This is
worth spelling out in some detail.
It is helpful in this connection to begin with an epistemic point. In this
light consider the following claim:
EVID If S has evidence E that n% of the members of a population

C conform to practice P, then S has some grounds for the correspond-
ing credence in the hypothesis that some arbitrary member M of
C conforms to P.
(Here the “corresponding credence” is a credence of n/100 in the hypoth-

esis that some arbitrary member M of C conforms to P.) EVID asserts that
S has “some grounds” for this credence given her evidence. Such a claim is
consistent with the idea that there are all sorts of considerations that might
tell against S’s adopting such a credence towards the hypothesis in question.
And it is clear that there are such considerations. For one thing, there can
be other relevant evidence: evidence regarding the individual M in question
(she just isn’t like that), the context (people don’t behave like that around
here), and so forth. Nor is evidence the only thing that might lead one to
refrain from fixing one’s credence in the way EVID states. As we will see
below, when one has doubts as to the morality or justice of the practice or
of the basis for the behavioral regularity itself, one will have moral or justice
grounds for resisting the idea that there is any normative pressure on people
to conform to the regularity or practice; and if so one will (or at any rate
ought to) refrain from drawing any inference regarding an individual made
on the basis of statistics about the prevalence of the regularity in the popu-
lation.7 Even so, it seems that we can say this much: if there is no question
as to the morality or justice of the regularity or of the basis of practice, and
there is no other relevant evidence (beyond the statistics of the population
C) bearing on whether M conforms to P, then the corresponding credence
is warranted. And this is to say that (absent worries of morality or justice)
evidence regarding the conformity rates of a regularity provide epistemic
reasons supporting the hypothesis that some arbitrary member M of this
community is more or less likely to conform to the practice. This much
should be uncontroversial.
Now recall the Reasons to Avoid Anticipated Harm Principle, RAAH,
according to which if S reasonably anticipates that φing will have harmful
effects on others, then S has some reason not to φ. This principle would
appear to cover the case in which one anticipates that there is an increased
risk in bringing about some harm.8 Accordingly, we might modify the prin-
ciple to make this explicit, so that the result is what we might call the “Rea-
sons to Avoid Anticipated Risk of Harm” Principle:
RAAHR If S reasonably anticipates that φing risks having harmful

effects on others, then S has some reason not to φ, where the strength
of the reason reflects the risk itself.
RAAHR postulates practical reasons9 to avoid doing what one anticipates

carries the risk of harming others (where the strength of the reasons cor-
responds to the likelihood of the risk and the badness of the harm).10 Now
let us add the Epistemic Badness of False Belief principle, EBFB, and the
Badness-Harm Link, HARM. Then we reach the conclusion that one has
some reason to avoid behaving in such a way as to risk leading others to
form false beliefs regarding whether one conforms to the regularity in ques-
tion.11 The reason in question is of the sort that RAAHR postulates: this
is a practical reason. The greater the degree of compliance with a behav-
ioral regularity within a population, the more likely it is that others will
assume that an arbitrary member of that practice (likely) conforms to the
regularity—with the result that if one is to avoid (having the harmful effect
of) generating false belief in others, one has some reason to avoid doing
what one has reason to believe is likely to bring about this harmful effect.
Keeping in mind that this holds only if there are no moral or justice con-
cerns about the (basis of the) behavioral regularity itself, we have reached a
conditional conclusion: when there are no such concerns regarding a famil-
iar behavioral regularity, there will be some normative pressure on people to
conform to this practice. This conclusion is supported by the combination
of EVID and RAAHR, together with the ethical and epistemic principles
mentioned above.
Is There Pressure to Conform When the Social

Pressure Is Not Robust?
So far I have been exploring how normative pressure to conform to a behav-
ioral regularity might be generated in cases in which the practice is both
salient to the population at large and sufficiently robust. But can we extend
our result to cases where the regularity, though salient and statistically sig-
nificant, is not as robust as the cases described above? With some slight
modifications, I think we can.
Several things can be said in support of the existence of (perhaps weak)
practical reasons for conformity even in a community in which there is a
familiar behavioral regularity with low conformity levels.
First, such practical reasons might be generated by the desire to avoid
doing something that likely prompts others to take certain (ultimately
unnecessary) precautionary measures. We might reasonably anticipate that
a subject who has even a low credence in the proposition that a given indi-
vidual will conform to a behavioral regularity may go the effort of taking
relevant precautions, so that the subject is prepared in either case. For this
reason, we might think that if it is little or no cost to the individual in ques-
tion to indicate publicly when she does not conform, she has some practical
reason to do so. The risk she runs in not doing so is not so much that she
will mislead others into believing that she conforms when she does not (we
are assuming that others’ credence in that hypothesis is low), as that she
allows others to expend energy (and to take precautions) they need not
have expended (taken). It would seem that if I can prevent you from making
unnecessary efforts, I have some reason to do so.
Second, such practical reasons (to avoid the risk of generating false belief)
might be generated by reflecting on considerations of epistemic diversity.
Different people in a community will have different evidence as to the con-
formity rates for a given practice in their community. Suppose then that we
have a large group of individuals making up a community. We might rea-
sonably anticipate that, while most people in the community have evidence
warranting a low degree of credence in the hypothesis that an arbitrary
individual conforms to the regularity, there will be some who have evidence
warranting a higher credence in the hypothesis. Even if such people are
relatively rare, even so, we might think that individuals who do not con-
form to the regularity risk doing these (evidentially idiosyncratic) people an
epistemic harm. And we might reasonably suppose that the larger the group,
the more likely it is that someone is evidentially idiosyncratic in this way.
Such an evidentially idiosyncratic person is harmed epistemically by those
whose non-conformity is not recognized—leading either to the absence of
(what would have been a helpful) true belief or the presence of a false belief.
Putting all of this together, we might say that the larger (and more epistemi-
cally diverse) the group, the more normative pressure there will be to con-
form to the regularity—lest one risk harming some evidentially idiosyncratic
individual(s) in the group (if there are any).
Third, such practical reasons (to avoid the risk of generating false belief)
might be generated by reflecting on considerations of epistemic uncertainty,
together with a mentality of “playing it safe” when there is such uncer-
tainty. Take a case in which the regularity itself is familiar, but it is unclear
to subjects what percentage of people in the community conform to the
regularity. (This might be so even if, as a matter of fact, the evidence they
have determines a precise answer; my point is that the answer itself need
not be known, as it is not luminous.) In that case, one might appeal to a
principle that (in the absence of stronger reasons) we have some practical
reason to “play it safe” by avoiding behavior that risks leading to the harm-
ful outcome. Interestingly, if enough people think that enough other people
endorse this “play it safe” reasoning, then this will render the practical rea-
sons to avoid the behavior even stronger.12 For in that case, people will think
that since others endorse this reasoning they are likely not risking the harm-
ful effects of non-conformity—in which case if in point of fact others are
non-conforming, they are behaving in ways that they might have anticipated
would lead to the harm of generating false belief regarding various individu-
als’ conformity to the practice.
In short, given a legitimate familiar behavioral regularity regarding which
people have evidence (and so epistemic reason to believe) that some non-
negligible percentage of the population conforms to the regularity, mem-
bers of that community will have some practical reasons to conform, or in
any case to avoid doing something that risks the generation of an epistemic
harm in other observers. This is so even if the evidence suggests that the per-
centage is low (so long as it is not negligible). And similar things can be said,
mutatis mutandis, in the case involving HELP.
Toward an Understanding of the Pressure

for Public Dissent
So far I have been considering cases in which there is a legitimate familiar
behavioral regularity, arguing that in such cases members of that commu-
nity will have some practical reasons to conform, lest they risk generating
harmful epistemic effects. But what can be said about cases in which the
behavioral regularity, though familiar, is (or is regarded as) illegitimate? In
that case, the argument above cannot be used to show that we have some
reason to conform, since that argument is addressed to regularities which
themselves are both familiar and legitimate. But the argument above can
be used to establish that, if there is a familiar behavioral regularity in one’s
community to which one does not conform, then either one regards the
practice as illegitimate or else one has some reason to conform. This result
is of additional interest, as it enables us to see where the basis for public
dissent might come from.
To see this, we might think about the disjunction just described as fol-
lows. Given a familiar regularity that is legitimate, those who do not or
would not conform to the regularity are under pressure to recognize that
they have some reason to make their non-conformity public (lest they do
something that they should anticipate risks having the harmful effects
described above). Consequently, one who has no questions as to the legiti-
macy of a familiar behavioral regularity is under pressure either to conform
or to make her non-conformity public. We might put the same point in
an equivalent fashion as follows: one who does not conform to a familiar
behavioral regularity must either find that regularity illegitimate, or is under
some pressure to make that non-conformity public. This way of putting
things is particularly interesting if we assume further that when a regularity
is immoral or otherwise illegitimate we all are under some pressure to speak
out on the matter. To be sure, this assumption is controversial. Even so, it
has been defended in a variety of different ways by others.13 And if it holds,
then our result is strengthened, since in that case, either way, one has some
reason to make public one’s non-conformity—in the latter case, by speaking
out against the legitimacy of the practice.
I anticipate many people will resist the idea that the existence of a familiar
behavioral regularity exerts this sort of pressure. How can the mere fact that
a practice is familiar force one’s hand in this way? In a sense, this should
come as no surprise. We are social creatures. We draw inferences from oth-
ers’ public behaviors. Insofar as there are familiar practices which carry
various signals, these will shape how we think of those public behaviors.
Insofar as one opts out under conditions in which it is not otherwise clear
that one is doing so, one has to expect the possibility that others will mis-
interpret one’s behavior. Assuming that the resulting false beliefs are a bad
thing, one has some reason to prevent this. Unless, of course, one thinks
that the practices themselves are immoral or unjust. In that case, it is argu-
able that one has a reason to indicate as much publicly.
Normative Pressure to Speak Out in Doubt or Disagreement

So far I have been speaking of behavioral regularities in the abstract. I now
want to consider one very specific regularity—a regularity that is so familiar
that the expectation of conformity here appears to constitute something of a
normative expectation (and the regularity itself a social practice). I refer here
to the behavioral regularity of indicating one’s doubts or disagreements when
one observes a public assertion with which one disagrees or towards which one
has doubts. The result of this regularity, I submit, is a social practice wherein
it is standard to interpret silent non-response as acceptance. After bringing
all of this out, I will conclude by suggesting how this appears to underwrite
something like normative pressure to speak out in doubt or disagreement.
I begin, first, with a description of an idealized (schematic) case, meant to
bring out the points above. To this end, imagine that there is a community—
I will call it Knowville—in which the following behavioral regularity holds
(I call this “DD” for “Disposition to Dissent”):
DD Whenever anyone harbors doubts regarding an assertion made in

one’s presence, one speaks out publicly to indicate these doubts.
Imagine that this regularity constitutes a strict empirical generalization:

there are no exceptions. Now imagine that S is a member of this community
who knows that DD characterizes this community. S is then in a position
to learn something quite specific whenever a public assertion goes unchal-
lenged in her community: S is then in a position to know that others have
no doubts or objections to the assertion. In particular, S is in a position to
exploit what I will call the “Silence-to-No-Dissent” conditional:
SD If audience A observes an assertion and does not publicly dissent,

then A does not harbor doubts regarding the assertion made.
If we assume further that an audience who does not harbor doubts regard-
ing an assertion accepts the assertion,14 then we can transform the above
conditional into the “Silence-to-Acceptance” conditional:
SA If audience A observes an assertion and does not publicly dissent,

then A accepts the assertion.
In short, anyone who is aware that DD holds in Knowville and knows that
A is a community member there is in a position to knowingly regard A’s
silence in the face of an assertion as an indication that A (harbors no doubts
regarding, and so) accepts the assertion. Insofar as members of Knowville
themselves are aware that DD holds there, they can interpret each others’
silence as indicating (lack of doubt, and so) acceptance.
Given what I argued in previous sections, we would predict that there is
some normative pressure on audiences in Knowville to speak out when they
disagree. By “normative pressure” I mean that audiences have (practical)
reasons, reflecting ethical considerations, for doing so. These reasons derive
from the two ethical principles above—avoid harm, help when possible—
and the two epistemic principles—false belief is an epistemic bad, true belief
an epistemic good. With these assumptions in place, people have practical
reasons to avoid doing what they have reason to believe would likely have
a harmful effect, as well as practical reasons to do what they have reason to
believe would likely have a helpful effect.
So far, we have a case that is very much like the sorts of case I was describ-
ing, schematically, above. However, there are reasons to think that this case
is particularly interesting from the perspective of an interest in dissent. For
there are reasons to think that the false beliefs that one risks generating if
one does not conform with the regularity are of particular significance. In
particular, given that non-conformity is a matter of not speaking out when
one disagrees or harbors doubts, one who does not conform is likely to
mislead people into thinking that one has accepted a public assertion under
conditions in which one has not done so. And for creatures like us, whose
pursuit of knowledge is often a highly social affair, the generation of false
beliefs of this sort is likely to be quite significant. Let me explain.
I should begin by conceding an obvious point: knowing when your peers
accept another’s assertion is not always an interesting thing to know. After
all, if your peers are credulous, uncritical, or otherwise have very little back-
ground knowledge with which to assess assertions made in their presence,
then the fact that they accept those assertions, while perhaps an interest-
ing fact about their psychology, is not interesting in any deeper sense. In
that case, knowing whether your peers have accepted another’s assertion
might be of interest to you as you try to predict their behavior (in contexts
in which such beliefs might be relevant), but it would appear to have no
greater value.
But it is easy to see that under certain further conditions, beliefs regarding
whether your peers have accepted an assertion are extremely significant—at
least from an epistemic point of view. For consider that there are conditions
under which peer acceptance is evidence of the acceptability of the assertion.
Admittedly, the conditions themselves are idealized. But it is worth bringing
the point out, as some non-idealized variants obtain in real-life cases. Still
let’s consider the idealized version first. To this end, suppose that the follow-
ing (“General Knowledgeableness”) condition also holds in the community
I am calling Knowville:
GK In general, people—speakers and audience members alike—are

highly knowledgeable regarding the subject matter of conversations
they have with one another.
The fact that GK holds in Knowville is a great boon for those members who
are aware of this. Suppose S (who is also aware that DD holds) is such a
member. Then S is in a position to draw even further inferences from the fact
that her peers have accepted some publicly observed assertion. In particular,
S is in a position to regard the fact that others accept the assertion, as some
evidence of the acceptability of the assertion. We might put this in terms of
another conditional, the “Acceptance-to-Acceptability” conditional:
AA If an audience A accepts an assertion, then this is some evidence that

the assertion is acceptable.
My claim is that those who know that DD and GK hold of Knowville, and
who know as well that A is a member of Knowville, are in a position to
reason on the basis of SA and AA.
The idea underwriting AA itself is not hard to appreciate. If you know
that Sivan is generally knowledgeable on the topic of an assertion that some
third party made in her presence, and you know that Sivan has accepted
the assertion, then you have evidence that someone knowledgeable on the
topic has accepted the assertion—which is a kind of higher-order evidence
of the acceptability of the assertion itself. Just how strong this higher-order
evidence is, as evidence of the acceptability of the assertion, will depend
on various things. Perhaps we can summarize these things by saying this:
Sivan’s acceptance of the assertion indicates the acceptability of the asser-
tion to just the extent that had there been any evidence to the contrary (or
reasons to reject the assertion) Sivan would have possessed that evidence
(or those reasons). Given that Sivan is rational she accepts an assertion only
when she doesn’t have reasons for doubt. So insofar as this counterfactual
holds of her—she would have had such reasons had there been any—her
acceptance indicates that there are no such reasons. Of course, you might
not know anything as strong as this counterfactual. You might only know
that had there been such reasons Sivan would likely have had them. In that
case, her acceptance of the assertion gives you evidence that it is likely that
there are no such reasons.
We can now see why facts about what others believe can be very sig-
nificant indeed- with the result that false beliefs about these facts is a very
bad thing, epistemically speaking. Suppose that one is in a community like
Knowville, where both DD and GK hold. Then AA holds too—in which
case one has excellent grounds for taking the silence of an audience A as
evidence not merely that A has accepted the assertion, but also (and much
more importantly) that the assertion itself is acceptable. If this is so, then
an audience A who leads others to form false beliefs to the effect that A has
accepted an assertion (when she has not) is likely to lead others to false
beliefs not only about what she believes, but also (given her knowledgeable-
ness) about what is the case in the non-mental world. This is bad indeed.
Extending Our Result

What is the bearing of these reflections on communities other than
Knowville? Take an ordinary community like ours. It is not true that there
is a behavioral regularity according to which
DD Whenever anyone harbors doubts regarding an assertion made in

one’s presence, one speaks out publicly to indicate these doubts.
On the contrary, people respond to assertions made in their presence in all

sorts of ways; people often have various motives for keeping their disagree-
ment private. Still, I think we can say this much:
DDWeak It sometime happens that when one harbors doubts regarding an

assertion made in one’s presence, one speaks out publicly to indicate
these doubts.
In fact, I think we can say something stronger than this. Speaking out when
one disagrees is a familiar phenomenon: it is not merely that it sometimes
happens, it is also that we recognize this as a familiar thing when it does
happen. We might put this as follows:
DDFmliar It is a familiar (though far from invariable!) behavioral regular-

ity that when one harbors doubts regarding an assertion made in one’s
presence, one speaks out publicly to indicate these doubts.
I regard DDFmliar as an important claim in at least two senses. First, if true,

DDFmliar suggests that the hypothesis, that audience A’s silence indicates her
acceptance, will be salient to those who observe A. This is to say that it will
be among the “alternatives” that have to be taken into account when observ-
ers try to figure out what is going on in the exchange. But second, DDFmliar
suggests that observers will not only have this as a salient hypothesis, but
that they will assign it some non-negligible credence. If they are rational,
that credence will be determined by their relevant evidence—including the
proportion of silent responses that are cases of acceptance.
Now this is enough to establish that in any community in which DDFmliar
holds, there will be some normative pressure to indicate rejection by speak-
ing out. To be sure, this pressure is weaker than in Knowville. But it still is
there. To see this, we need only invoke the Reasons to Avoid Anticipated
Risk of Harm Principle, RAAHR: if S reasonably anticipates that φing risks
having harmful effects on others, then S has some reason not to φ, where the
strength of the reason reflects the risk itself. If this is correct, then wherever
the practice of publicly manifesting one’s rejection of an assertion is familiar,
to just the extent that it is familiar and widespread, to that extent it would
seem that there is normative pressure to conform to the practice. This is
a point that is true in our world, not only in the imagined community of
Knowville.
Conclusion
In this chapter I have tried to show how something like a practical reason
to speak out in dissent to a statement can be generated by appeal to four
plausible principles in ethics and two plausible principles in epistemology.
The argument itself frames the issue in terms of behavioral regularities that
involve signaling. In a certain subclass of behavioral regularities—those
in which the regularity itself is neither unjust nor immoral, and where (i)
non-conformity itself also sends a signal, (ii) the signals sent by individu-
als send proprietary information, and (iii) it is reasonable to suppose that
there is a general interest in having the information in question—the con-
clusion asserting normative pressure to conform to those regularities can be
obtained by appeal to our six principles. The claim that there is normative
pressure to indicate one’s doubts or disagreement in the face of a publicly
observed assertion is a special case of this more general phenomenon.
Notes
1 With thanks to the audience at the Dissent Conference at the University of Con-
necticut, where I gave this chapter as a talk. Special thanks to Casey Johnson for
written comments on an earlier version of this chapter.
2 Whence the number of people who fly the “W” then? I think it has to do with
signaling not only a Cubs’ victory, but also one’s allegiance to the team and to
the community of like-minded fans. At this point it appears to have become
something ritualistic. There is some proprietary information, then, but it is not
the knowledge that the signal evolved to communicate. (I thank Casey Johnson
for pointing this out, in conversation.)
3 This is a topic that is addressed in great length in Bicchieri 2006.
4 I will call this regularity robust when two conditions hold. First, the likelihood
that someone is a helper, given that they are wearing a shirt of this color, is far
greater than the likelihood that someone is a helper, given only that they are a
member of the community; and second, the likelihood that someone is not a
helper, given that they are not wearing a shirt of this color, is far greater than the
likelihood that they are not a helper, given only that they are a member of the
community.
5 People in the community are aware of the regularity.
6 Namely, making the audience’s epistemic perspective worse (in at least this one
way) than it had been previously.
7 See, e.g., Gendler 2011 for the psychological and epistemological costs on peo-
ple who confront a related situation.
8 A similar point will go with the Reasons to Help Principle, RTH, but I will not
belabor that here.
9 The reasons are practical insofar as they are reasons to avoid the undue risk of harm-
ing others. (With thanks to Casey Johnson for indicating the need to clarify this.)
10 We can argue, by reductio, that RAAHR is true if RAAH is. Suppose (for reduc-
tio) that RAAH holds but RAAHR does not. In that case, S has reason not to
φ when she reasonably anticipates that it will lead to harmful effects, but does
not have reason not to φ when she reasonably anticipates that it risks leading to
harmful effects. Such a result is deeply implausible and arguably incoherent. It is
deeply implausible: if RAAH were true and RAAHR were false, then, contrary to
fact, it should make sense to say, “I have no reason not to φ even though I antici-
pate that the risk of harmful effects of φing is high, but I would have reason not
to φ if I were to anticipate that φing will have harmful effects.” And arguably this
combination is incoherent since it seems that reasonably anticipating that some-
thing will lead to harmful effects just is a special case of reasonably anticipating
that it risks leading to harmful effects (i.e., the case in which the risk is maximal,
or above some suitably high threshold). So I conclude that RAAHR is as plausible
as RAAH.
11 The strength of these practical reasons reflects in part the strength of the epis-
temic reasons just described.
12 When this belief about others is false, we have an instance of the phenomenon
social psychologists call “pluralistic ignorance.” For philosophical discussions,
see e.g., Bicchieri (2006), Hendricks (2010), and Berring, Hansen, and Pedersen
(2014).
13 In his “Beyond Vietnam” speech delivered at New York’s Riverside Church on
April 4, 1967, Dr. Martin Luther King, Jr., famously remarked that “there comes
a time when silence is betrayal.” For a recent extended defense of this idea, see
Hay (2013).
14 This is not entirely uncontroversial, though many epistemologists and philoso-
phers of language accept something like this assumption. Epistemologists accept
something like this assumption when they construe testimony as a default source
of justification, such that audiences are entitled to accept another’s say-so unless
they have reasons for doubt. (This is the so-called “anti-reductionist” view in
the epistemology of testimony.) The view then amounts to a normative version
of the assumption above: audiences are entitled to accept another’s say-so in
the absence of reasons for doubt. And philosophers of language accept some-
thing like this assumption when they construe assertion as a speech act that aims
to update the common ground, and when they assume that in the face of an
observed assertion an audience has only two options (acceptance or rejection).
See, e.g., Stalnaker (1978).
References
Bicchieri, C. (2006). The grammar of society: The nature and dynamics of social
norms. Cambridge: Cambridge University Press.
Bjerring, J., Hansen, J., & Pedersen, N. (2014). On the rationality of pluralistic
ignorance. Synthese, 191(11), 2445–2470.
Hay, C. (2013). Kantianism, liberalism, and feminism: Resisting oppression. Springer.
Hendricks, V. (2010). Knowledge transmissibility and pluralistic ignorance: A first
stab. Metaphilosophy, 41(3), 279–291.
Gendler, T. (2011). On the epistemic costs of implicit bias. Philosophical Studies:
An International Journal for Philosophy in the Analytic Tradition, 156(1), 33–63.
Retrieved from http://www.jstor.org/stable/41487720
Stalnaker, R. C. (1978). Assertion (pp. 147–161). Blackwell Publishers Ltd.
4 Dissent
Good, Bad, and Reasonable
Klemens Kappel
Public Dissent
We all agree that expressing dissent in public is a good thing. But everyone
also admits that this is not always so. Sometimes dissent is bad, or seems
bad, and there are many cases that come to mind: dissent denying climate
change, dissent holding that the evolution theory is false or poorly justified,
dissent denying the finding that smoking causes cancer, dissent insisting on
a link between MMR vaccines and autism, dissent asserting that HPV vac-
cines regularly lead to severe side effects.
It is one thing to say when dissent is good or bad. It is another thing to say
that dissent is reasonable in that it deserves to be heard and accommodated,
or unreasonable meaning that it does not deserve a hearing, or response, or
can even legitimately be targeted and silenced. In this chapter I will propose a
way of thinking about good and bad dissent, and reasonable and unreason-
able dissent. My main concern will be the distinction between reasonable
and unreasonable dissent.
Dissent is basically a case of disagreement. An individual A dissents with
B about some proposition P by expressing her disagreement about P. How-
ever, talk about dissent normally presupposes a particular context featur-
ing a mainstream view which is being contested in assertions by one or
more dissenting non-mainstream voices. So, dissent involves an assertion
(made in writing, speech, or otherwise) of a proposition, whereby a dis-
senter expresses her disagreement with a mainstream view. Following this,
I will for convenience speak about dissenters and mainstreamers, where
mainstreamers assert P, while dissenters deny P, or deny that P is known or
as solidly justified as the mainstreamers think. As we shall see, in addition to
mainstreamers and dissenters, there will be audiences to consider.
I will focus my discussion on what we might call public dissent, whereby
I mean dissent expressed in public. The cases of dissent that I find most
urgent to discuss involve individual or political decision-making contexts,
that is, contexts where the dissent concerns views that may affect individual
choices (such as in the cases of vaccination above), or choices to be made
by policy makers or in our common democratic deliberation. I will consider
62 Klemens Kappel
only dissent regarding factual (non-normative) questions, as dissent and dis-
agreement concerning normative questions in my view raise very different
issues. Finally, as the examples above suggest, I will focus on cases where
dissent involves a mainstream view being asserted or supported by estab-
lished science. Thus, the paradigmatic case of dissent that I will be con-
cerned with features established science proclaiming a mainstream view P,
and a dissenting party publicly rejecting or doubting P (or rejecting that P
is as well established as asserted by the mainstreamers) and where the issue
is important for decisions that we may make as individuals or collectively.
I will start by focusing on the question of when dissent is epistemically
good and bad (section 2). As we shall see, an importantly different question
concerns when dissent is reasonable or unreasonable in a sense that I will try
to specify. I will start developing this account in section 3. The distinction
between reasonable and unreasonable dissent is, I suggest, in certain ways
similar to Rawls’s distinction between reasonable and unreasonable politi-
cal views, and I find it instructive to briefly consider some of the details in
Rawls’s distinction (section 4). In section 5, I outline what I call the relevant
alternative account of reasonable dissent, and propose a more specific inter-
pretation of this account which links reasonable dissent to publicly available
scientific evidence. In section 6, I assess the account, and in section 7, I sum
up some of the features of it. Finally, section 8 briefly considers the question
of how we should respond to unreasonable dissent. My overall aim is to
provide a conceptual framework for thinking about good and bad dissent,
and reasonable and unreasonable dissent, and to consider what may justify
a plausible way of distinguishing reasonable from unreasonable dissent.
Good and Bad Dissent

Dissent can be evaluated from other normative perspectives than the epis-
temic, and these are obviously important in their own right. Generally, the
communicative act of expressing dissent may be a means of having one’s
voice be heard or of expressing one’s identity, and plausibly there is, within
the framework of liberal democracy, a moral right to a say and to be heard,
or express one’s identity. This may imply a right to dissent, and a corre-
sponding duty to make dissent possible, or that the expression of dissent is
itself a good outcome. Generally, I will set aside these normative features of
dissent, and focus on epistemological questions.
The epistemic benefits of dissent have been discussed in general terms
in philosophy of science, in political philosophy, and in social epistemol-
ogy, and I will briefly rehearse the main themes here.1 In general, dissent
can be epistemically productive for a variety of reasons, the discussion of
which dates back at least to Mill: (i) dissent may help pool existing knowl-
edge, for example when dissent brings attention to reasons and evidence
that has gone unnoticed. (ii) Relatedly, but slightly different, dissent may
serve to facilitate a distribution of cognitive tasks. This may happen when
Dissent: Good, Bad, and Reasonable 63
mainstreamers defend P while dissenters explore the alternative hypothesis
that not-P. As Kitcher argues, it may be better that a broader variety of
hypotheses are explored, rather than only the most promising hypothesis.2
(iii) Dissent may also improve the quality and outcome of argumentative
reasoning. Assume that the mainstreamer asserts P and adduces reasons R
for P. Dissenters question the truth of P, and they may do so by pointing
to additional evidence that suggests not-P, by questioning that R is true, or
by asserting that R does not support P after all. In response to this, main-
streamers will (ideally) either rebut the arguments presented by dissenters,
or revise their confidence in P. As a result, mainstreamers (and dissenters)
might move to a more accurate view, or gain a better justification of their
view. (iv) Finally, dissent may help avoid discursive failures due to cascades
or group polarization and similar phenomena. There are important features
of our cognitive psychology such as confirmation bias, identity protective
reasoning, and similar phenomena that underlies (iii) and (iv). We all tend
to look selectively for confirmation of our views, and to ignore reasoning-
errors behind conclusions that we concur with. By contrast, we are much
more inclined to search for and actually detect errors in a pattern of reason-
ing when we disagree with the conclusion to begin with. For these reasons,
argumentative reasoning is often held to be significantly improved by diver-
sity of opinion.3
If we define good dissent as dissent that promotes valuable epistemic ends
among mainstreamers, dissenters, and audiences, we can define bad dis-
sent as dissent that detracts from such ends.4 So, bad dissent is dissent that
detracts from good epistemic ends or is conducive to bad epistemic ends.
When reflecting about what makes good dissent good, we can identify, at
least in very rough outline, the most obvious ways in which dissent may lead
to bad outcomes. Dissent based on false first-order beliefs will sometimes
not add to the existing pool of knowledge or deepen our understanding
of the arguments for the mainstream view, and may therefore be a waste
of time. Similarly, dissent may pursue a hopeless project, again leading
nowhere. Dissent is sometimes so broad in scope that it calls into question
basic norms of evidence and reasoning. If this happens, the dissent, and
responding to it, is likely not to lead to the normal revisions of beliefs in
response to arguments and counter-arguments and therefore not be epis-
temically beneficial. Dissent may be insincere in the sense that dissenters
press what they know to be false beliefs not supported by the evidence, or in
the sense that they pursue a strategy of not revising their beliefs in response
to compelling evidence and arguments, or never to publicly admit defeat
and so on. Insincere dissent of this sort is what one sees in the documented
cases of manufactured dissent involving the claims that tobacco smoking is
a leading cause of cancer, and the findings of anthropogenic climate change.5
Again, insincere dissent may lead to poor epistemic outcomes.
For mainstreamers, evidence of dissent may represent higher-order evi-
dence to which they should rationally respond. If the dissent is based on
64 Klemens Kappel
false beliefs, questions valid epistemic norms, or is insincere, but this is not
yet clear to the mainstreamers, the dissent may constitute misleading higher-
order evidence indicating that there is something wrong about the main-
stream view. Mainstreamers may then be in a position in which they may
believe that they should rationally reduce their confidence in their views in
response to the misleading higher-order information.
In most of the practically important disputes between mainstreamers and
dissenters, most of us are neither dissenters nor defending a mainstream
view, but members of the audience. For members of the audience, dissent,
and how it plays out, is an important piece of social evidence that decisively
influences the epistemic status of audience’s beliefs about the disputed mat-
ters. This is because, typically, members of the audience will not be able to
understand or follow details of an exchange between mainstreamers and
dissenters, because they lack the training, time, or energy. So, to form beliefs
regarding the disputed matters, the audience will have to use other cues that
are more readily accessible and easier to process. I consider it plausible that
members of an audience can increase their degree of epistemic justification
significantly by being spectators to debates that others conduct. Suppose,
for example, that the dissenters bring to the fore a number of reasons that
initially suggest the mainstream view to be wrong. The mainstreamers then
carefully rebut these arguments, and as a result they become reassured in
their conviction that the mainstream view is correct. Suppose, though this is
unlikely to happen in practice, that the dissenters then agree to the rebuttal
of their dissent. Even if the audience doesn’t understand or follow the details
of these rebuttals and revisions, merely observing the exchange can provide
important social evidence in virtue of which members of the audience can
acquire a very high degree of epistemic justification in certain views.
As this suggests, one psychologically important cue is the perception of
how dissenters and mainstreamers adjust their views in response to argu-
ments and reasons proposed in the debate. This explains why public dissent
questioning basic epistemic norms, or dissent based on insincere argumenta-
tion is particularly pernicious, as this results in debates that the mainstream-
ers may never be perceived to win, even if they are in fact right and have
the evidence on their side. For the audience it may not be transparent at all
that one side of the debate questions basic epistemic norms or is insincere,
so the audience may mistake the dissenters’ steadfastness for a high degree
of justification of their views, or the weakness of the mainstream view, and
the audience may then adjust their own view accordingly.
There are other important cues, of course. A crucial one concerns the rela-
tive numbers of mainstreamers and dissenters. A lot of the higher-order evi-
dence reaching the audience is simply due to the impressions that dissenters
are a sizeable group, not by anything they actually do or say during a debate.
A debate can give the impression that dissenters and mainstreamers are more
evenly divided that they actually are, or that dissenters (or mainstreamers)
have formed their beliefs more independently than they actually have.6
Consequentialism About Good and Bad Dissent
I have assumed that good and bad dissent are generally speaking conse-
quentialist notions: good and bad dissent are defined in terms of epistemic
consequences, rather than, say, in terms of intentions of speakers, exercise
of epistemic virtues, or compliance with epistemic norms.
When the definitions are consequentialist, they come with features famil-
iar from consequentialist thinking. First, truth or other desirable epistemic
ends may be promoted through unusual or abnormal causal chains. So,
what normally would be a case of bad dissent may, due to unusual circum-
stances, be good, and vice versa. Second, we are often not able to say with
any degree of certainty whether a particular instance of dissent is good or
bad. Even if it looks good, there might be unforeseen circumstances that
turn out to make it bad, or vice versa. Third, there may be a plurality of
epistemic goods that matters, and sometimes dissent may promote some at
the expense of others: dissent may, for example, promote true belief, but not
knowledge, or dissent may increase the preponderance of justified belief,
but not truth. Fourth, there are important distributional concerns to bear in
mind. As argued, dissent is in great part good or bad in virtue of the ways
that it affects audiences, rather than the dissenting parties and the main-
streamers. Indeed, often the main reason we care about dissent is because of
the effects it may have on audiences. But it is an abstraction to speak about
the audience in the singular. There are many audiences, and they may be
affected in very different ways by a given display of dissent. For example, it
is well documented that nutrition advice provided by health authorities is
absorbed in quite differential ways by various segments of the population.
Similarly, one might expect that dissent about such information may have
very diverse effects on different segments of a population.
There are other implications of defining the notions of good and bad
dissent in consequentialist terms. A rule such as “Offer good dissent, and
no bad dissent” is perfectly well justified and reasonable. However, it will
be generally ineffectual in the following way. Often, when agents sincerely
try to follow that rule the best they can, this will make little difference to
whether the dissent they offer is in fact good or bad. This is because when
dissenters object to P, it is generally because they reject that P is known or
sufficiently justified, but also because they consider it important that this
is communicated to the public, that is, the audiences in question. So, when
they dissent regarding P, they not only regard P as false or insufficiently
justified, but also regard dissent against P as all things considered good, or
likely to be good. So, dissenters can assiduously and sincerely follow the rule
and still offer bad dissent.
Since good and bad dissent about P is defined in ways that implicitly
refers to the truth or falsity of the disputed view, a controversy regarding
P is likely to be mirrored in controversy regarding whether dissent about P
is good or bad. Mainstreamers will think that dissent about P is likely to
66 Klemens Kappel
be bad, in part because they regard P as true, whereas dissenters will think
that dissent is good, in part because they regard P as false. A recent case from
Denmark illustrates this. HPV vaccine administered to teenage girls is consid-
ered very safe and likely to have a significant effect preventing certain kinds
of cervical cancer. However, like in other parts of the world, dissenting voices
have recently publicly asserted that the vaccine actually leads to severe side
effects in a considerable number of cases. After conducting new studies, and
evaluating the available scientific evidence, the health authorities disagree. Yet,
the dissenters, or at least enough of them, are unabated by the expert consen-
sus. The public exposure of this dissent is probably a main reason that there
has now been a very sharp increase in false beliefs about the safety of the vac-
cine, and consequently a steep decline in the number of girls being vaccinated.
The experts, believing that the HPV vaccine is safe and effective, obviously
regard this as a tragic outcome, epistemically and otherwise, and wished that
the dissent had never gotten the traction it had through media exposure. How-
ever, dissenters obviously regard the decline in the acceptance of the vaccine
as a good outcome, both epistemically and in other terms. So, unsurprisingly,
disagreement over the likelihood of serious side effects of the vaccine is mir-
rored in disagreement about whether the dissent expressed was good or bad.
None of the above suggests that the consequentialist distinction between
good and bad dissent is somehow misconceived, or implausible. But there
are two features that nonetheless speak to the feasibility of the good/bad
distinction. First, it is epistemically challenging. Applying the distinction
requires information about likely outcomes that we often don’t have, and
are not in a position to acquire. Second, the good/bad distinction is divisive
in a certain sense illustrated by HPV case above: when we disagree about
some practically important proposition P we also tend to disagree about
whether dissent about P is good or bad, and maybe more frequently in the
hard and bitter disputes that should concern us most. So, evaluating dissent
in terms of good and bad is not going to help reconcile us, but rather adds
an extra layer of contention.
For these reasons, we might be interested in a way of evaluating dissent
that is less epistemically challenging and divisive. This is what we might
to get from a distinction between reasonable and unreasonable dissent to
which I will now turn. Evaluating dissent in terms of reasonability may, one
might hope, be epistemically more manageable and less divisive than evalu-
ating dissent in terms of good and bad dissent.
Reasonable and Unreasonable Dissent

To give an intuitive idea of where we are going, consider first some cases of
dissent that we might initially agree to regard as unreasonable: Holocaust
denial, denial that smoking causes lung cancer, at least some cases of climate
change denial, evolution skepticism, insisting that the Earth is flat, and a
number of conspiracy theories. Compare to instances of dissent that we might
say are reasonable. This could concern questions concerning the effects of
various economic policies, the efficacy of various medical treatments, advise
about nutrition or exercise, or the exact consequences of various global cli-
mate change scenarios, and of course many controversies in philosophy.7
To add content to our notion of reasonable dissent, I suggest that we
specify the functional roles of reasonable dissent. The simple idea is that the
distinction between reasonable and non-reasonable dissent latches on to a
set of norms for public deliberation and public decision making, and that
this in part determines the content of the distinction between reasonable and
non-reasonable dissent. I propose the following:
Reasonable Dissent in Public Deliberation

Reasonable dissent whether P should be taken seriously. Roughly, delib-
erators are obligated to acknowledge reasonable dissent by responding
to it, either by adjusting their confidence in P, or by explaining why
they do not find dissent about P persuasive and thus see no ground for
reducing confidence in P.
To elaborate, one can think that reasonable dissent rests on an error of

some sort, but it must be considered an honest mistake, it is not that dissent-
ers need be ill willed, stupid, or hopelessly ignorant, and it merits a response.
Unreasonable dissent, on the other hand, can be rightfully ignored, and in
some cases it should be ignored. We don’t have to invite proponents of
unreasonable views to round tables; we don’t have to give those views air-
time on radio or television. And it is not only that we do not have to share a
platform with unreasonable views, in some cases we should not.
Reasonable Dissent in Public Decision-Making

Reasonable dissent regarding some question P should have some role
in the justification of public policy. If a policy choice turns on the truth
of some proposition P, then reasonable dissent about P should influence
the policy choice.
Reasonable dissent can be accommodated in policy making in a variety of

ways. Policy makers can look for policy options that are attractive for both
those who believe P and those who deny P, or they can try to find compro-
mise policy options. If policy makers take the truth of P for granted in their
justification of their policy, then at least they should acknowledge that hon-
est dissent about P exists, and they need to justify themselves to dissenters
why they have chosen to rely on P rather than not-P in their policy making.
Using a term from Rawls’s political philosophy, we can say that when
there is reasonable dissent about P, then both the view that P and the dis-
senting view that not-P will be part of public reason.8 Unreasonable dissent,
68 Klemens Kappel
on the other hand, can be ignored in the justification of public policy. Unrea-
sonable views should not be part of the factual basis of what justifies pub-
lic policy, even indirectly in the sense that one ought to acknowledge its
existence and justify not accommodating it. It can and should be ignored
entirely. Unreasonable dissenting views are, in this sense, not part of public
reason.
This gives a rough functional account of the distinction between reason-
able and unreasonable dissent. The question I want to pursue in the follow-
ing sections is this: is there some underlying property (or set of properties)
that we can theoretically identify, and which underlies and justifies this dis-
tinction and the distinctive roles for reasonable and unreasonable views that
I have depicted? To get an idea of what I have in mind, consider first Rawls’s
notion of reasonable political views.
Rawls on Reasonable Political Views

Rawls famously argued that the existence of pluralism in liberal societies
gives rise to a deep problem. In his “The Idea of Public Reason Revisited”
(Rawls, 1997) he writes
Citizens realize that they cannot reach agreement or even approach

mutual understanding on the basis of their irreconcilable comprehen-
sive doctrines. In view of this, they need to consider what kinds of rea-
sons they may reasonably give one another when fundamental political
questions are at stake. I propose that in public reason comprehensive
doctrines of truth or right be replaced by an idea of the politically rea-
sonable addressed to citizens as citizens.
(766)
Let us briefly unpack the idea of the politically reasonable, as Rawls details
it in the pages that follows. A comprehensive doctrine is a view about the
metaphysical, religious, and moral questions that we know of or might have
an interest in. A comprehensive doctrine is liberal if it affirms the basic lib-
eral values of freedom and equality of citizens. A comprehensive doctrine is
non-liberal if it fails to affirm these values. A political conception of justice
is a view of the basic structure of society, i.e., the constitutional essentials
and the fundamental questions of justice. Again, a political conception of
justice can be liberal by affirming the basic freedom and equality of citizens.
A political conception can be non-liberal in that it denies this. A political
conception of justice can be part of a comprehensive doctrine, but it can
also be freestanding in that it is not part of any particular comprehensive
doctrine, but perhaps is supported by overlapping consensus of compre-
hensive doctrines. A comprehensive doctrine that supports a liberal concep-
tion of justice is reasonable, whereas comprehensive doctrines that do not
support liberal political conceptions are not reasonable. Even non-liberal
comprehensive doctrines can support liberal conceptions of justice, so even
they can be reasonable.9
The fundamental problem that Rawls is concerned with is how we justify
coercive policies to one another given that we disagree about very fundamen-
tal normative questions.10 Rawls’s suggestion is that we should acknowledge
a restriction in the reasons we offer to one another when justifying coercive
policies. Basically, we should restrict ourselves to proposing reasons that
are included in our reasonable political conceptions. This is what Rawls
means by public reason. So, his view is that coercive policies should be jus-
tifiable by public reason, where public reason is the set of reasons included
in the set of reasonable political conceptions that are either contained in
reasonable comprehensive views or supported by reasonable comprehensive
views. We can also state this in terms of reasonable individuals, rather than
views. Roughly, a coercive policy is legitimate only if it is justifiable to all
reasonable individuals, where a reasonable individual is someone holding a
reasonable view.
There are many details of interpretation that I will set aside. The impor-
tant thing to note is that democratic legitimacy of coercive legislation is a
matter of acceptability to all reasonable views or reasonable citizens, but we
do not have to answer to non-reasonable views. Nor do we have to rebut
objections to coercive policies in so far as they are based on unreasonable
views.
Clearly, there are important analogies between reasonability in Rawls’s
sense and the sort of epistemic reasonability that is our concern here. Both
rely on a distinction between views that are to be taken seriously in delibera-
tion and political decision making, and views that are not. Both specify the
distinction in terms of functional roles. Rawls’s distinction between reason-
able and unreasonable views is both less epistemically demanding and less
divisive than determining the correct comprehensive view, or determining
what a just society is. These questions are highly complex, and in compari-
son it is relatively easy to determine whether some political conception is
reasonable or not. And in many cases reasonability is less divisive: when you
and I disagree in comprehensive views, we might easily agree that both our
views are reasonable.11
Rawls’s distinction between reasonable and non-reasonable views, and
his specification of the content of public reason, is primarily morally moti-
vated. There are several overlapping restrictions that determine the content
of public reason. One is that only reasonable comprehensive doctrines are
permitted as a part of public reason. Another is the principle of restraint,
which holds that one should appeal only to reasons and values that others
may reasonably be expected to reasonably endorse. This means that one
should not appeal to the parts of one’s comprehensive view that is non-
shared, even if one’s comprehensive view is reasonable. Both restrictions are
closely linked to various ideas summarized in what Rawls calls the principle
of reciprocity.
70 Klemens Kappel
The principle of reciprocity requires that one regard others as free and
equal citizens that one wants to collaborate with on fair terms, conditional
on their doing so as well. This implies that one cannot offer reasons that one
knows that other reasonable individuals cannot accept. On the other hand,
holders of unreasonable views are, by definition, not willing to support rea-
sonable political conceptions of justice, that is, conceptions of justice that
commit to treat others as free and equal citizens. So, they fail to abide by the
principle of reciprocity, and in part this is why we are entitled to set aside
their views. It is not that unreasonable views are inconsistent or incoherent
(though they might be of course), or that holders of unreasonable views are
stupid or are acting in bad faith (though they might be). Clearly, holders of
unreasonable views may be smart and informed people, who act in good
faith and sincerely believe the views they assert. Rather, the reason their
views are excluded from public reason is that they fail to conform to basic
ideals of freedom and equality. They fail to respect the principle of reciproc-
ity, and in great part this is why we are entitled to exclude them from con-
sideration, according to Rawls.
We can thus make a distinction between two parts of Rawls’s account.
One the one hand there is the reasonable/unreasonable distinction which is
specified in terms of content and functional role of public reason. On the
other hand there is the justification for making this distinction in just that
way. As I said, Rawls’s justification is primarily moral in nature in that it
identifies a set of moral reasons that explains and justifies the distinction
between reasonable and non-reasonable views and the content and func-
tional role of public reason.
The Relevant Alternative Account of Reasonable Dissent

My main question now is whether something similar can be said about rea-
sonable dissent. That is, can we identify independent underlying properties
that justify a distinction between reasonable and unreasonable dissent and
support the associated functional roles? In part to motivate the proposal
that I find most plausible, I want to start considering some other options.
I assume that dissent paradigmatically involves making public assertions.
This might naturally lead to the suggestion that reasonable dissent can be
defined in terms of norms of assertions. The idea would be that dissent is
reasonable only if the assertion expressing the dissent complies with the
proper norm of assertion. Brief reflection suggests that this is not plausible.
Consider the knowledge norm of assertion according to which asserting P is
epistemically appropriate only if one knows that P. If the mainstream view
P is true it will be impossible to know not-P, and therefore not possible to
be reasonable to dissent by asserting not-P.12 Clearly, this is too sweeping.
Dissent that not-P is sometimes reasonable, even if the dissenters are in fact
wrong in their claims.
Consider then the main alternative to the knowledge norm of assertion,
which is the justified belief norm of assertion. According to this, asserting
P is epistemically appropriate if one is sufficiently well justified in believing
that P, but one does not need to know P. So, this view would imply that S’s
dissent that not-P is reasonable only if S justifiably believes not-P. I think this
is wrong as well. Dissent asserting not-P can be reasonable even if dissent-
ers are not highly justified in believing not-P, or even don’t believe P at all.
Suppose that the mainstream view holds that a particular medication has no
serious side effects. Albert has recently undergone treatment with that medi-
cation, and has suffered symptoms that he takes to be serious side effects
of the medication. As it happens, these symptoms have never been reported
following the medication in question before. So, assume that Albert dis-
sents regarding the mainstream view that the medication has no serious side
effects. Now, whether Albert’s dissent is reasonable does not seem to depend
on the epistemic status of Albert’s belief that the medication can have serious
side effects. Suppose that Albert’s belief that the medication can have serious
side effects is not epistemically justified to any significant degree, e.g., for
the following reason: the symptoms that Albert has experienced may be a
random occurrence that are not causally related to the medication, but this
obvious possibility is one about which Albert never thought. Despite this
I would say that Albert’s dissent may well be reasonable in the sense I have
outlined. So, a dissenting belief need not be doxastically justified for it to
be reasonable. Indeed, Albert need not even believe that the medication can
lead to these side effects for his dissent to be reasonable. Suppose that Albert
did in fact consider the possibility that his perceived side-effects were merely
a random occurrence, and therefore he does not believe with any confidence
that the medication caused it—he just believes that this is a possibility, and
his dissent expresses this. Still, it seems that Albert’s dissent could be entirely
reasonable.
Similar remarks apply, I think, to the suggestion that reasonable dissent
should be defined in terms of epistemic blamelessness: dissent that not-P is
reasonable if (and only if) one is blameless in believing not-P. Assume that
an agent is epistemically blameless in believing not-P if she is immersed in
an epistemic tradition and simply follows that tradition in the proper way.
Accordingly, given embedment in the appropriate epistemic traditions, one can
be blameless in believing the weirdest things. Any view could qualify as reason-
able dissent, if only the dissenter is immersed in a sufficiently odd epistemic
tradition. Again, this is far from our normal notion of reasonable dissent.13
I suggest that the reason that the above proposals are not right is that
they link epistemic reasonability to epistemic features of the dissenter, that
is, to whether the dissenter knows, justifiably believes, or is epistemically
blameless.14 I think that our pre-theoretical notion of reasonable dissent is
better captured in a different way. Reasonable dissent is not a function of
the individual dissenter’s beliefs or epistemic properties, but a function of
72 Klemens Kappel
the state of the publicly available evidence supporting the mainstream view.
I will now describe an account of reasonable dissent along these lines in
more details, albeit a bit abstractly.
The account assumes that there is a common pool of evidence E supporting
the mainstream view P. E supports P in so far as P is a plausible explanation
of E. So, P is a hypothesis that explains how E could come to be the evidence
that we have. Relative to E there is a set of not-P alternatives, that is, views
which explain the evidence in other ways, that is views that also explain E
while implying that P is false. Assume that some of the not-P alternatives are
relevant, and some are not relevant. Relevant not-P alternatives are those
which E needs to rule out in order for E to provide support for P. So, when
E supports P, this is because E rules out all the relevant not-P alternatives.
Obviously, no matter how strong E’s support for P is, one can always invent
not-P alternatives that E cannot rule out—just think of familiar skeptical
scenarios. But these are not relevant alternatives, so they can be rightfully
ignored. For E to support P we only need that the relevant not-P alternatives
are ruled out, while we can disregard the non-relevant alternatives.15
I suggest that we use this familiar model of evidential support to define
reasonable dissent:
Relevant Alternative Account of Reasonable Dissent

Dissent is reasonable if it asserts a relevant not-P alternative which is
not ruled out by the common pool of evidence E. Dissent is unreason-
able if it asserts a not-P alternative which is not a relevant alternative, or
asserts a not-P alternative which is already ruled out by E.
I suggest that our ordinary pre-theoretical way of talking about reason-

able and unreasonable dissent is approximately captured by this account.
When we say that dissent is reasonable, we assume the existence of a com-
mon pool of evidence supporting a mainstream view, a partition of alter-
natives into relevant and non-relevant alternatives, and we assert that the
dissent concerns a relevant alternative that is not rendered false (or suffi-
ciently improbable) by the common pool of evidence. When we assert that
dissent is unreasonable, we assume that it concerns an alternative that is not
relevant, or is already ruled out by the public evidence.
For the account to mirror how we normally talk about reasonable dissent,
I suggest we need to adopt what we might call an objectivist rather than a
subjectivist interpretation. Roughly, according to the subjectivist interpreta-
tion, reasonable dissent is a matter of S’s dissent being reasonable accord-
ing to S’s own standards, that is, S’s take on E, P, and the space of relevant
not-P options. On an objectivist interpretation, by contrast, for any set of
evidence E and a proposition P, there is a sort of external standard that
settles the partition into relevant and irrelevant alternatives, and reasonable
is defined relative to those standards. It is pretty clear, I think, that we need
the objectivist interpretation of the model to remain in the vicinity of our
pre-theoretical notion of reasonable and unreasonable dissent. Going for
the subjectivist account would potentially include any conspiracy theory
however mad into the realm of reasonable dissent. Building conspiracy the-
ories is not cognitively challenging. For any set of evidence E and proposi-
tion P, an agent S can fairly easily add beliefs to her system of beliefs such
that additional not-P alternatives are deemed relevant on S’s system, while
S still maintains a coherent belief set. Yet, on our ordinary way of thinking,
dissent is not rendered reasonable merely by being underpinned by the dis-
senter’s conspiratorial worldview, and if it did, the notion loses the claim to
the function role that I have specified.
We evidently need to say more about the objective standards. It is not
wrong to define reasonable dissent in relation to objectively correct parti-
tion of relevant and irrelevant alternatives whatever they are. But if we don’t
have a way of indicating what they are, the account will be of limited use.
For the purpose of further discussion, I want to propose a more specific
interpretation of the relevant alternative account of reasonable dissent by
tying the account to the use of established scientific findings and norms of
reasoning within science:
Science-Based Account of Reasonable Dissent

Dissent against P is reasonable if it asserts a not-P alternative which is
relevant and not already ruled out by the publicly available pool of evi-
dence E, relative to generally accepted scientific findings and scientific
standards of evidence. Dissent against P is unreasonable if it asserts a
not-P alternative which is not a relevant alternative, or asserts a not-P
alternative which is already ruled out by E, relative to generally accepted
scientific findings and scientific standards of evidence.
The science-based account of reasonable dissent mirrors Rawls’s definition of

political reasonability, which as we saw, is not tied to objectively correct values
whatever they may be, but to a distinct set of foundational values of liberal
democracy, those of freedom and equality. Political reasonability is not neutral
among worldviews, but requires support to certain specified substantive values.
In the same way, the science-based model of reasonable dissent is not neutral
among different views of what the proper epistemic standards are. Dissent can-
not be reasonable if it disregards scientific standards and findings.16
Of course, the model implies gross simplifications, both in its general form
and when it takes the form of the science-based model of reasonable dissent.
Yet, it might still be useful for structuring our thoughts about reasonable
dissent, even if some of the variables are underspecified or the model presup-
poses things about the world that are just not true.
The model assumes a dichotomous distinction between reasonable
and unreasonable, whereas it might clearly be more natural to think of
74 Klemens Kappel
reasonable and unreasonable dissent as gradual affairs. The science-based
account assumes that there is a uniform set of scientific standards which
applies across all disciplines, where they are actually quite divergent, even
to the extent that practices in some part of science and academia grounds
what other parts consider unreasonable dissent.17
Similarly, the model assumes the existence of a pool of publicly available
evidence relative to which reasonable and unreasonable dissent is defined
which is partly an idealization. Moreover, the interpretation of the account
is extremely sensitive to what we count as public evidence. When does some
item of evidence belong to the common pool? Suppose that there is some
item of esoteric scientific evidence e in favor of the mainstream view P,
where e is both crucial for P, but also such that only specialists understand
why e strongly supports P and even know about the existence of e. Is e then
part of the public evidence E? I suggest that for the model to mirror how we
normally think about reasonable dissent, we would need to assume some
rather liberal standards for something to be part of public evidence. Suppose
that it is publicly known or fairly easily accessible that the relevant part of
the scientific community considers P strongly supported by the available evi-
dence. Suppose that when the scientific community makes this judgment, it
also includes the existence of e, though this detail is known only to special-
ists. So, non-specialists do not know that e exists, or how it relates to P, but
only that some evidence exists which according to the specialists supports P,
and e happens to part of that. I suggest that in this case, e should be consid-
ered part of the publicly available evidence in virtue of social transmission
of knowledge, epistemic dependence, or trust in the scientific institutions.18
This is one reason why scientific consensus and public evidence about scien-
tific consensus is so important for publics.
The account elucidates features of reasonable dissent which we might find
independently plausible. The account explains why dissent may be reason-
able at some stage of inquiry and cease to be at others. In general, reason-
able dissent has a dynamic similar to a conversation, or a project of joint
inquiry. Given a certain state of common evidence and a particular aim of
inquiry, certain objections are appropriate, but others are not because they
are too farfetched, or deemed irrelevant by the evidence already gathered.
As objections are addressed adequately, sometimes by adding new evidence,
objections may cease to be appropriate.
The model explains why a dissenter does not need to know, be justified,
or even believe her dissenting not-P assertion. It seems that one can sincerely
dissent by asserting not-P without even believing not-P. As I assumed previ-
ously, we might find this result independently plausible; compare, for exam-
ple, to the common practice of making insincere assertions in philosophy
discussions as objections to views being discussed. What matters is whether
not-P dissent involves a relevant not-P alternative which is not already ruled
out by the publicly available evidence.
Assume that sincerely asserting not-P requires believing not-P. As we have
seen, one can dissent by asserting not-P, even if one does not believe not-P.
So, we might suggest that sincere dissent does not require sincere assertion.
Rather, we might say that sincere dissent is asserting not-P when one regards
the particular not-P possibility asserted as relevant and not ruled out by
the publicly available evidence. Insincere dissent, by contrast, is asserting
not-P while believing that not-P is not a relevant alternative, or is already
adequately ruled out by publicly available evidence.
It also seems plausible that dissent can take other forms than by flat out
asserting not-P. Sometimes dissenting opinions merely question that the
mainstream view is as well supported as is assumed, or expresses doubt that
P or suspension of judgment whether P, without asserting any particular
not-P alternative. Sometimes dissent overtly asserting some not-P alternative
is really a proxy for these other attitudes, and sometimes one can express
those attitudes without asserting any particular not-P alternative. A more
fully developed account should be amended to explain when such attitudes
are reasonable and when they are not. The obvious suggestion would be to
define reasonability relative to a common pool of evidence: a common pool
of evidence can be such that doubt or suspension of judgment about P is no
longer warranted.
As I said, the model asserts that for a given set of evidence E and a propo-
sition P, there is a partition of relevant and non-relevant not-P alternatives.
If we wish, we can add that this partition is sensitive to contextual or prag-
matic features: when more is at stake, more not-P options become rele-
vant.19 When more is at stake, the space for reasonable dissent expands. Or,
as Lewis suggests, alternatives may become relevant merely by being salient,
and they can become so just in virtue of being the focus of our attention.20
If this is so, then simply paying attention to a dissenting view can serve to
render it reasonable, which would create problems for the functional role
that I assumed reasonable dissent to have.
If the space of relevant alternatives depends on pragmatic features, it will
likely be asymmetrical in some cases, and by implication the space for rea-
sonable dissent will also be asymmetrical.21 Consider a case in which the
social consequences of accepting P while P is false are negligible. On the
other hand, the social consequences of rejecting P while P is true are disas-
trous. In such cases we might propose that there is an asymmetrical set of
relevant alternatives. Relative to E and the assumption that P is correct,
there is a space of relevant not-P alternatives. Relative to E and the assump-
tion that P is false, there is a space of relevant P-alternatives. However, the
second space is larger, because more needs to be done to ensure that we
don’t wrongly reject P. Accordingly, the space for reasonable dissent will be
asymmetrical as well. If the mainstream view holds that P is false, then there
is a large space of reasonable dissent asserting P, as the consequences of a
mistake in rejecting P is a social disaster. If the mainstream view holds that
76 Klemens Kappel
P is true, then the space of not-P dissent is comparatively smaller as nothing
much is at stake.
The account allows us to make sense of more complex disputes. One
important type of dispute concerns what the common pool of evidence is,
or what this evidence rules in or out. The model predicts that a dispute over
this will tend to transfer into a dispute over reasonable and unreasonable
dissent. There can also be disputes over the space of reasonable dissent that
turn on disagreements over the space of relevant alternatives. If the space
of relevant alternatives depends on pragmatic features, we can predict that
there can be disputes over what dissent is reasonable that will do so as well.
Some will hold that dissent is reasonable in view of the grave consequences
of being wrong, whereas others hold that consequences of being wrong
are overstated, and thus the dissent not reasonable. Another feature of the
model is that there might be different publics, recognizing different pools of
evidence E and E’, and dissent may be reasonable relative to E, but not rela-
tive to E’. This may make sense of some disputes over reasonable dissent.
Assessing the Science-Based Account of Reasonable Dissent

In describing the relevant alternative account of reasonable dissent I have
employed the framework of evidence and relevant alternatives, but I don’t
intend important conclusions to depend on the specifics of this particular
framework. If we wish, we can describe the model in terms of doxastic prac-
tices, belief forming processes, methods of inquiry, epistemic virtues, epis-
temic norms, and so on, and explain the same things in these terms. In those
terms, the account roughly says that dissent is reasonable only if it is based
on epistemic practices, belief forming processes, methods of inquiry, or vir-
tues that are actually reliable in the appropriate senses, and that reasonable
dissent also requires that dissenting beliefs can be maintained while based
on such practices, processes, methods of inquiry, or virtues. Reasonable dis-
sent, therefore, requires some measure of appropriate or successful relation
to the nature of the world, or the facts, and therefore also some modicum of
true beliefs about the features of the world that determine which methods of
inquiry are likely to lead us to true beliefs, and which are not.
Consider now whether the science-based account of reasonable dissent
is epistemically demanding and divisive. On the assumption that one can
simply defer to the scientific consensus, including a consensus about what
alternatives are relevant and which are not, and which are ruled out and
which are not, the account is not epistemically demanding. Is it divisive?
Well, among those who accept the account, it is not. They can disagree
about P, and yet fully agree that various ways of dissenting about P are
entirely reasonable, and others are not.
In other important respects the science-based account will be both
demanding and divisive. Suppose, for example, that we disagree whether
there is a scientific consensus, or whether a consensus has been generated
for truth-indicative reasons or not. Or suppose we disagree whether some
option is appropriately ruled out by the current evidence. Or suppose that
we disagree whether some option is relevant or not, maybe because we
ultimately disagree in our assessment of the gravity of the consequences of
being wrong. These disagreements can be internal to those of how to accept
the science-based account. But these disputes are still hard to settle, and
in this sense the account will sometimes be quite epistemically demanding.
Moreover, such disagreements can divisive in the sense that they involve dis-
agreements about whether dissent is reasonable that are trailing the under-
lying disagreements about the facts. If we disagree about the truth of the
mainstream view P, we are likely also going to disagree whether dissent that
not-P is reasonable or not.
Obviously, the science-based account will be divisive in another way, too.
It is not an objection to the account that it refers to an objectively correct
set of epistemic norms. Any plausible account of reasonable dissent will
need to do so, or so I argued above. But some will surely object that the
science-based account illegitimately singles out scientific epistemic norms
or practices. The science-based account of reasonable dissent in effect rules
out as unreasonable any dissent that questions the epistemic authority of
science. Clearly, many will reject this as an unwarranted form of scientism.
Thus, the science-based account of reasonable dissent may, in such cases,
itself be divisive.
For the reasons just given, the account predicts that cases of dissent are
divisive in a quite unsettling way under certain circumstances which are
unfortunately not uncommon. When it is generally assumed that there is
expert consensus that the public evidence decisively supports P, and yet some
dissent, this is because they (i) reject the genuine existence of the expert con-
sensus, or (ii) because they reject its epistemic significance (they grant that
the consensus exists, but believe that it has been garnered for non-truth
indicative reasons), or (iii) because they don’t accept the privileged status
of scientific reasoning in the domain in question. When this is so, they not
only disagree about P, they also insist that certain not-P claims qualify as
reasonable dissent, whereas the science-based account of reasonable dissent
implies that they are not. In those cases, when we disagree about P, we tend
to also disagree about whether dissent about P is reasonable or not.
What sort of defense can one give of the science-based account of rea-
sonable dissent? It is hard to see how one could defend the science-based
model without assuming the superior reliability of scientific standards of
reasoning and evidence, that is, the privileged epistemic status of scientific
norms of reasoning and evidence. But these premises can hardly be given
a non-circular defense. In Alston’s sense, even the best justification we can
offer for the privileged status of scientific standards of reasoning and evi-
dence will be epistemically circular.22 So, there is a (sort of) more funda-
mental justification of the science-based account, but it is one that will not
rationally convince those who oppose the account because they reject the
78 Klemens Kappel
superior status scientific norms of reasoning and evidence. In this respect,
the account shares the predicament of Rawls’s account of reasonable politi-
cal disagreement. Those rejecting the basic values of liberal democracy are
not going to be impressed by Rawls’s justification of his distinction between
reasonable and unreasonable political views, as this merely assumes the
superiority of basic liberal values.
On the other hand, it is worth noting that if one accepts that employ-
ing scientific standards of reasoning and evidence represent our best hope
to get to the truth about non-normative factual matters, then the science-
based account of reasonable dissent makes perfect sense. We should want an
account of reasonable dissent that is no stricter but also no more permissive
than the science-based account. Moreover, the functional roles of reason-
able and unreasonable dissent that I sketched above appear perfectly sen-
sible. We should take note of reasonable dissent in our public deliberation
and in our public decision making, because reasonable dissent by definition
points to lacunas in our best evidence, but by the same token we should
ignore unreasonable dissent.
It might perhaps be suggested that the Millian reasons why dissent may be
epistemically good briefly mentioned in section 2 actually support expand-
ing the space of reasonable dissent wider than the science-based account
would have it. After all, one might suggest, why not reap the epistemic
benefits of debating unreasonable dissent? On reflection, we can see that
this is hardly so. Debating options that are unreasonable according to the
science-based account means debating options that are already discredited
by the scientific evidence, or too farfetched anyway. This is unlikely to serve
epistemically good ends, except when it turns out that the dissent was rea-
sonable after all, or if it helps persuade dissenters to adopt the science-based
model of reasonable dissent. Also, if we accept the prerogative of scientific
norms of reasoning and evidence, it is hard to see why we should expect our
decisions to generally become better by including non-reasonable alterna-
tives in our deliberation.
Reasonable Dissent: Substantive, Divisive, Political

I will end by summarizing and commenting upon some features of the
account as developed so far. I have proposed that reasonable dissent should
be considered a substantive notion in that reasonable dissent is defined with
reference to objectively correct substantive epistemic norms. In broader
terms, being epistemically reasonable requires abiding to some degree by the
correct epistemic norms, trusting the proper experts or sources of evidence,
relying on proper methods, possessing the right epistemic virtues, and so on.
The notion of reasonable dissent being substantive does not yet say anything
about what the correct epistemic norms are. However, the science-based
account of reasonable dissent is what we should insist upon if we gener-
ally believe that epistemic norms and practices embodied in science are our
best bet at arriving at true or justified beliefs concerning non-normative fac-
tual questions. The science-based account of reasonable dissent is, in many
cases, both epistemically manageable and not divisive.
Yet, in other significant cases, the science-based account of reasonable
dissent is divisive in several ways. In familiar cases of recalcitrant politicized
disagreements over scientific facts, verdicts about reasonable and unreason-
able dissent will be contestable. Dissenters will typically assume that the
non-P alternative they assert is relevant and not ruled out by public evi-
dence, whereas mainstreamers will think that it is not relevant, or no longer
a live option. This is because dissenters dispute the mainstreamers’ claims
about what the common pool of evidence is, what it shows, or what alterna-
tives are relevant. Or it is because dissenters do not accept the prerogative
of scientific reasoning and evidence in the determination of what counts as
reasonable dissent. The science-based account will be predictably divisive in
those cases. Mainstreamers will be inclined to dismiss or ignore what they
think of as unreasonable dissent, and dissenters will take this as an affront.
Mainstreamers may even think that repeated unreasonable dissent cannot
be made in good faith, and thus suspect dissenters for foul play or for acting
in bad faith. In some cases they may be right, of course, but they need not
be. As we have seen, the account explains how honest people can disagree
about what counts as reasonable dissent.
I suggest that the account supports saying that expressing judgments
about reasonable and unreasonable dissent is in a certain sense political.
I don’t mean this in a derisive way. By saying that it is political, I merely
mean that the point of asserting that an instance of dissent is unreasonable
or reasonable is rarely just to make a dispassionate descriptive statement
about certain epistemological features of an instance of dissent. It is much
more to endorse that the dissent be treated in a certain way, viz. the func-
tional roles that I suggested above. Moreover, when saying to dissenters
that their dissent is not reasonable, one will often be making a statement
that one knows cannot be supported by reasons and arguments that can
rationally convince that particular audience (for the reasons given above).
This supports suggesting that when saying that dissent is unreasonable—to
those particular audiences—one is more likely endorsing a policy regarding
how their dissent should be treated, a policy with which one knows that the
audience will disagree. The point of saying that dissent is unreasonable can
hardly be to remind the dissenter about the prerogative of scientific stan-
dards of reasoning, and thus it is more natural to see it as endorsing that
prerogative and the criteria of reasonable dissent based on it. This is what
I mean by saying that expressions of judgments about reasonable and non-
reasonable dissent are political. My intention is not to make a sophisticated
claim about the semantics of linguistic expressions of evaluations of dissent.
The point is simpler: why would we bother to point out whether dissent is
reasonable or not reasonable if it were not because we care about the associ-
ated functional roles?
80 Klemens Kappel
Notes
1 The following paragraphs on the epistemic benefits of dissent are based on Kap-
pel, Hallsson, & Møller (2017).
2 See (Kitcher, (1990)
3 For references, see Kappel et al. (2017)
4 This is essentially the underlying assumption in recent contributions such as Bid-
dle and Leuschner (2015; Leuschner (2016)
5 See the famous account in Oreskes and Conway (2010)
6 For a lucid discussion of such issues, see Goldman (2001).
7 Note the similar distinctions in the literature: legitimate/illegitimate dissenting
voices (De Melo-Martín & Intemann, 2014) epistemically detrimental/non-
detrimental dissent, appropriate/inappropriate dissent (Biddle & Leuschner, 2015;
Leuschner, 2016).
8 For more discussion of the role of science in public reason, see Jønch-Clausen
and Kappel (2016).
9 Note that Rawls endorses the constraint on reasonable views that they result
from the exercise of reason, implying that they cannot be grossly inconsistent or
incoherent.
10 Rawls’s concern is actually limited to the design of constitutional essentials or
basic structures of society. We can set this complication aside here.
11 Of course, those who hold unreasonable views will reject the prerogative that
Rawls gives to fundamental liberal values, and with this the privileged functional
role that he assigns to reasonable views. In this sense, Rawls’s thinking on rea-
sonability will be divisive.
12 On this account, if P is true, it might still be reasonable to dissent by asserting that
we don’t have good enough evidence for P, but only if one knows that this is so.
Again, this is a rather demanding requirement on reasonable dissent. Moreover,
if the mainstream view is not only true, but also known, then dissenters cannot
be in position to know that there is insufficient evidence for the mainstream view.
13 Everything here of course depends on how epistemic blamelessness is defined.
One can have less permissive views on epistemic blamelessness.
14 It may well be that the account of reasonable dissent that I develop herein implies
that reasonable dissent can be stated in terms of propositional justification of
epistemic agents.
15 This follows Lewis in broad outlines. See Lewis (1996)
16 Fernando Broncano-Berrocal suggested that we also need a notion of reason-
able dissent within an epistemically unreasonable tradition. This could be pro-
vided with an account having the same structure, but where reasonable dissent is
defined relative to the epistemic norms of the unreasonable tradition in question.
17 Thanks to Fernando Broncano-Berrocal for this observation.
18 For views about this, see Goldberg (2010); Kappel (2014).
19 See Derose (1995; Lewis (1996); Fantl and McGrath (2002)
20 See Lewis (1996).
21 Cf. Douglas’s discussions of values in science, (Douglas, 2009).
22 See Alston (1986) on epistemic circularity, and later work on related issues.
References
Alston, W. P. (1986). Epistemic circularity. Philosophy and Phenomenological
Research, 47(1), 1–30.
Biddle, J. B., & Leuschner, A. (2015). Climate skepticism and the manufacture of
doubt: Can dissent in science be epistemically detrimental? European Journal
for Philosophy of Science. Springer Netherlands, 5(3), 261–278. doi: 10.1007/
s13194-014-0101-x.
De Melo-Martín, I., & Intemann, K. (2014). Who’s afraid of dissent? Addressing
concerns about undermining scientific consensus in public policy developments.
Perspectives on Science, 22(4), 593–615. Retrieved April 18, 2017, from https://
muse.jhu.edu/article/561197
Derose, K. (1995). Solving the skeptical problem. Philosophical Review, 104(1), 1–52.
Douglas, H. E. (2009). Science, policy, and the value-free ideal. Pittsburgh, PA: Uni-
versity of Pittsburgh Press.
Fantl, J., & McGrath, M. (2002). Evidence, pragmatics, and justification. The
Philosophical Review. Duke University PressPhilosophical Review, 111(1), 67.
doi:10.2307/3182570.
Goldberg, S. (2010). Relying on others: An essay in epistemology. Oxford: Oxford
University Press.
Goldman, A. (2001). Experts: Which ones should you trust? Philosophy and Phe-
nomenological Research, 63.
Jønch-Clausen, K., & Kappel, K. (2016). Scientific facts and methods in pub-
lic reason. Res Publica. Springer Netherlands, 22(2), 117–133. doi:10.1007/
s11158-015-9290-1.
Kappel, K. (2014). Believing on trust. Synthese. Springer Netherlands, 191(9),
2009–2028. doi:10.1007/s11229-013-0376-z.
Kappel, K., Hallsson, B., & Møller, E. F. L. (2017). Freedom of expression, diversity,
and truth. In K. Lippert-Rasmussen, K. Brownlee, & D. Coady (Eds.), A com-
panion to applied philosophy (pp. 147–162). Hoboken: John Wiley & Sons, Ltd.
Kitcher, P. (1990). The division of cognitive labor. Journal of Philosophy, 87(1), 5–22.
Leuschner, A. (2016). Is it appropriate to “target” inappropriate dissent? on the
normative consequences of climate skepticism. Synthese. doi:10.1007/s11229-
016-1267-x.
Lewis, D. (1996). Elusive knowledge. Australasian Journal of Philosophy, 74(4),
549–567.
Oreskes, N., & Conway, E. M. (2010). Merchants of doubt:How a handful of scien-
tists obscured the truth on issues from tobacco smoke to global warming. Blooms-
bury Press.
Rawls, J. (1997). The idea of public reason revisited. The University of Chicago Law
Review, 64(3), 765–807.
5 Silence and Objecting
Jennifer Lackey
What are we entitled to infer from the silence of others? This is the ques-
tion that will be at the center of this chapter. More precisely, I will explore
the connection between silence and the duty we have to object to what we
take to be false or unwarranted. I will argue that the central approach to
understanding this connection in the literature—what I call the coopera-
tive conversation view—is an instance of an ideal theory and, as such, it
excludes the way that things are in the actual world, especially for those
who are systematically marginalized. I then show how this exclusion results
in a number of significant problems facing the cooperative conversation
view, ultimately leading to its rejection. Finally, I argue that when our theo-
retical starting point is non-ideal theory, and we focus on conversational
exchanges in which features of the actual world take center stage—such as
power, oppression, and cultural differences—we find ourselves recognizing
that objecting is often a luxury, one that not everyone can afford to make.
This leads to the conclusion that a constraint on any plausible view of the
duty to object is that one’s duty can be directly influenced by one’s social
status.
Ideal Theory
Ideal theory in ethics and political theory, often paradigmatically exempli-
fied by the work of John Rawls in A Theory of Justice, is frequently cri-
tiqued for all that it leaves out of the theoretical picture. Indeed, it is not
the appeal to ideals themselves that is regarded as distinctively problematic,
since non-ideal theorists will also invoke moral ideals, but the absence of
attention paid to the way the world actually is, especially for those who are
oppressed and marginalized in various ways.
This is a point that is developed extensively by Charles Mills, who char-
acterizes ideal theory in a recent book primarily in terms of what is absent
or ignored rather than by what is present. He writes:
What distinguishes ideal theory is the reliance on idealization to the

exclusion, or at least marginalization, of the actual. . . . [I]deal theory
Silence and Objecting 83
either tacitly represents the actual as a simple deviation from the ideal,
not worth theorizing in its own right, or claims that starting from the
ideal is at least the best way of realizing it.
(Mills, 2017, p. 75)
According to Mills, then, the core feature of ideal theory is not the idealiza-
tion itself but, rather, the disregard of varying degrees of the actual world
and the people and institutions in it. He goes on to argue further that ideal
theory will use some or all of a list of concepts and assumptions, including
(i) idealized capacities, (ii) silence on oppression, and (iii) ideal social institu-
tions. Let’s focus briefly on each of these.
With respect to (i), ideal theory often presupposes capacities that are
entirely unrealistic for human agents. This is true of those who are privi-
leged, but especially of “those subordinated in different ways, who would
not have had an equal opportunity for their natural capacities to develop,
and who will in fact be disabled in crucial respects” (Mills, 2017, p. 76).
Moreover, when capacities are idealized in this way, norms and expecta-
tions, along with the corresponding disapprobation, are likely to become
skewed. For instance, if a parent idealizes the cognitive capacities of her
5-year-old child, then she might think it is appropriate to expect him to
be able to sit through math tutoring every day for an hour without inter-
ruption. When he begins to fidget after 30 minutes and cries on the second
day, she might criticize his behavior and regard him as disappointing or
deficient when in fact the problem is her idealization of his capacities. This
example focuses on cognitive capacities, but similar remarks apply in the
moral, political, and epistemic domains.
Regarding (ii), Mills says,
Almost by definition, it follows from the focus of ideal theory that little
or nothing will be said about actual historic oppression and its legacy
in the present or current ongoing oppression, though these may be ges-
tured at in a vague or promissory way.
(Mills, 2017, p. 76)
This is especially problematic when the issues being explored are norma-
tive ones, such as those involving justice, obligations, blameworthiness, and
so on. If, for instance, we are assessing when to hold agents blameworthy
for being bystanders, completely disregarding the vulnerable positions of
members of different oppressed groups would result in holding all of those
present equally responsible for their inaction. But this might be misguided
insofar as those in positions of power have far less to lose when intervening
in morally complex situations than those who are systematically oppressed.
Finally, (iii) focuses on the idealization of social institutions, such as
economic structures and legal systems, which are conceptualized as mod-
els functioning “with little or no sense of how their actual workings may
84 Jennifer Lackey
systematically disadvantage women, the poor, and racial minorities” (Mills,
2017, p. 76). As we saw with the idealization of capacities, this can result in
significant distortions. If, for instance, we’re theorizing about punishment in
the context of an idealized view of the legal system, we might end up with
a radically different conclusion about the moral permissibility of the death
penalty than if we factor in the racism pervading criminal justice.
Thus, in general, Mills objects to ideal theory because of its disconnection
from the actual world—especially the experiences of the marginalized—, the
way this distorts our understanding of phenomena of critical importance,
and the overall impact this has on our ability to achieve the desired results
of the very theories in question:
In modeling humans, human capacities, human interactions, human

institutions, and human society on ideal-as-idealized-models, in never
exploring how profoundly different these are from ideal-as-descriptive-
models, we are abstracting away from realities that are crucial to our
comprehension of the actual workings of injustice in human interac-
tions and social institutions, and we are thereby guaranteeing that the
ideal-as-idealized-model will never be achieved.
(Mills, 2017, p. 77)
According to Mills, non-ideal theory not only avoids these problems, but is
also far better suited to accounting for the perspectives of members of sub-
ordinated groups, which is essential to any normative theory.
The Cooperative Conversation View

I now want to turn to what I will argue is a non-ideal theory of the duty1 to
object2 and show that some of its most serious shortcomings are the result
of its disconnection from the actual world.
In a recent paper,3 Sanford C. Goldberg argues on behalf of the following
claim, which he calls the Default Entitlement to Assume that Silence Indi-
cates Acceptance:
DEASIA Competent language users enjoy a default (albeit defeasible) entitle-

ment to assume that in speech exchanges which are conversations, a
hearer’s silence in the face of an observed assertion indicates acceptance
of the assertion. (Goldberg unpublished, p. 3)
The kind of entitlement Goldberg has in mind is a practice-generated

entitlement,
wherein a social practice is sufficiently widespread and recognized that

those who participate in the practice are entitled to expect that the stan-
dards of the practice are being followed by the other participants, and
where the expectation itself is normative (the sort through which we
hold one another accountable).
(Goldberg unpublished, p. 2)4
Goldberg’s defense of this claim is twofold. The first is empirical: he argues

that, as a matter of fact, people often will assume that a hearer’s silence reveals
acceptance. To support this, he cites quotes from Plato to Mark Twain, Mar-
tin Luther King Jr. to Rabbi Bradley Artson, all purportedly showing that we
standardly take silence as assent. King, for instance, says that “There comes
a time when silence is betrayal,”5 and Artson, commenting on the Jewish
response to injustice, remarks that “our silence and inaction in the face of
contemporary injustice and oppression is akin to assenting to it.”6
The second source of support that Goldberg invokes on behalf of the
DEASIA is normative: he maintains that observers enjoy a presumptive
(albeit defeasible) entitlement to regard silence as indicative of assent.
Goldberg’s arguments here are rich and detailed, but they rest primarily on
the view, stemming from the work of Grice, that conversations are coop-
erative, rational activities and, thus, that silent rejection is uncooperative.
Given this,
insofar as linguistically competent subjects are entitled to suppose that

they are participants in a conversation, they have a reason, deriving
from the practice of assertion itself, to suppose that other participants’
silence in the face of a mutually observed assertion indicates acceptance.
(Goldberg unpublished, pp. 7–8)
That Goldberg’s thesis is relevant to the duty to object should be clear. If

we are entitled to regard silence as assent or agreement, then there has to
be some reason to believe that conversational participants will object when
they disagree with what is said. Goldberg himself characterizes this in terms
of there being normative pressure to signal rejection of a proffered assertion,
but this can easily be understood in the sense of having a duty to object to
what is said. According to Goldberg, this pressure or duty is
a special case of a more general, and more familiar, phenomenon: we are

under normative pressure not to engage in behaviors that recklessly or
negligently risk harming others. In the case of silence, the paradigmatic
sort of harm is an epistemic harm: the hearer who fails to indicate pub-
licly his rejection of the mutually observed assertion recklessly or negli-
gently risks misleading other participants into forming false beliefs—in
particular, as to his own reaction to the assertion, but also, as a possible
consequence, as to the truth-value of the asserted proposition itself.
(Goldberg unpublished, p. 2)
Let’s call this the cooperative conversation view of the duty to object.
86 Jennifer Lackey
What I would like to emphasize now is the extent to which this view
is an example of an ideal theory of the sort discussed earlier. Recall that
Mills argues that the distinguishing feature of ideal theory is its reliance on
idealization to the exclusion or marginalization of the actual world. This is
precisely what Goldberg does in his defense of the cooperative conversa-
tion view. His theoretical starting point is what can be inferred from con-
versations in which there is full cooperation of participants. Moreover, just
as Mills argues that ideal theory represents the actual as a simple deviation
from the ideal, Goldberg relegates to “defeating conditions” the multitude
of ways in which conversations can be radically uncooperative because of
the different situations of those involved.
To make this point clearer, let’s look at how we end up in a different place
than Goldberg does if we start by paying adequate attention to the way
the world actually is. Consider, for instance, some ways in which silence
reasonably and regularly fails to indicate acceptance. First, objections with
traction are a limited epistemic good, and we need to make wise choices
about when to use them. Because of this, silence may be the result of a
simple cost-benefit analysis due to limited resources. While a full discussion
of how to understand objections with traction lies beyond the scope of this
chapter, here is a start: objections simpliciter are assertions that are added to
the conversational context with the aim of correcting the record, but ones
with traction typically involve more, though what this “more” involves can
take on different forms. Sometimes, objections with traction are ones that
are accepted by at least some members of the conversational context. Other
times, they will have weaker functions, such as sowing seeds of doubt about
the targeted proposition, or being factored into the overall evidential basis
of the beliefs of the audience members. Still others will be such that they are
not immediately rejected or defeated. At a minimum, however, we might say
that objections with traction cannot be systematically ignored or silenced by
the members of the conversational context.7 I emphasize “systematically”
because we would be reluctant to say that an epistemic agent’s objections
are effective if they are merely not tuned out in every conversational context
she finds herself in, but never even rise to the level of being factored into
the evidential basis of the corresponding beliefs. Given that objections with
traction are not a limitless good, silence is often the result of using caution
or care with our voices so that our objections matter and are heard when we
raise them. We all do this frequently—parents are silent with their teenagers
for fear of being “tuned out,” colleagues are silent at department meetings
so that they “pick their battles,” and friends are silent on social media so
that they do not become “that person” who is always quibbling. In none of
these cases does silence mean agreement, and yet the contexts may be utterly
indistinguishable from those in which it does.
This brings us to the second way: differences in status or power might
lead conversational participants to not object even when they reject what is
being said. Just as those without power often have fewer social and material
resources than those with it do, so, too, they often have fewer epistemic
resources. For instance, prisoners lack the social status and power that cor-
rectional officers possess, students lack the authority and the epistemic status
of their teachers, and victims often lack the credibility that their assailants
enjoy. Silence here might not be merely the result of choosing to be careful
with one’s epistemic goods, as we all do at times, but instead due to social
structures that make objecting impossible, difficult, futile, costly, and so on.
In a recent article by law professor Patricia A. Broussard, this issue is taken
up directly when she writes:
Black women continue to suffer from trauma they endured as a result

of the dynamics of the societal structure of their world during and after
slavery. Moreover, that social structure, by its very nature, imposed a
code of silence upon Black women, which continues to exist to this
day. There are some aspects of life one does not share and there are
aspects that silence protects. As a result of this culture of secrecy, Black
women, through their silence, have unwittingly enabled and protected
those who have abused them for decades.
There are many societies that embrace a culture of silence, but they
do not have a history of slavery and Jim Crow. Therefore, there are
clearly other factors and dynamics responsible for shaping silent behav-
ior in those societies . . . the veil of silence still worn by Black women is
a remnant of survival tactics adopted to survive slavery and Jim Crow.
(Broussard, 2013, p. 375)
According to Broussard, then, silence from a black woman has a unique

history, one that is grounded in oppression and in no way indicates genuine
acceptance. Audre Lorde makes a similar point when she writes:
I wrote for those women who do not speak, for those who do not have a
voice because they/we were so terrified, because we are taught to respect
fear more than ourselves. We’ve been taught that silence would save us,
but it won’t.
(Lorde, 1983, p. 105)
While the first and second ways are related in that they both can be under-
stood in terms of the distribution and use of epistemic goods, there is a level
of choice involved in the former that is absent in the latter. We all make deci-
sions about when and how to use our voices, but for those without status
and power, the ability to object might simply be denied to them or severely
restricted in various ways.8
The third way silence regularly fails to indicate acceptance is the result
of psychological or cultural differences: some people are shy, reticent, and
conflict-averse, while others are at home with loud protests. Some cultures
have norms and expectations that are fundamentally at odds with raising
88 Jennifer Lackey
objections, especially for certain members, such as women, while others
cultivate frank and explicit debates. With respect to Japanese culture, for
instance, Takie Sugiyama Lebra writes:
It has been shown that silence [in Japanese culture] is not only polyse-
mic but symbolic of logically opposite meanings or emotions. This cer-
tainly generates confusion and misunderstanding for a cultural outsider,
but for the native as well. The silent speaker, too, is likely to have mixed
feelings or rationales. When a woman says she was silent throughout
the period of her husband’s extra-marital indulgence, she can mean her
feminine modesty, compliance, patience, resentment, unforgiveness, or
defiance, and may mean all. A man’s refusal to express tender feelings
toward his wife may be explained not only as embarrassment, but as an
expression of male dignity, or as his true, sincere love, which is beyond
words. In the scene of collective decision-making, silence can be taken
as polite acquiescence or disagreement.
(Lebra, 1987, p. 350)
According to Lebra, then, silence within Japanese culture has wildly differ-
ent meanings, some of which are even at odds with one another. Particularly
noteworthy for our purposes here is that silence can signal either assent or
dissent and that even native speakers cannot easily discriminate between
them. Thus, silence in the face of disagreement might be the default for some
people and cultures, and yet it might be highly unusual for others.
The fourth way silence regularly fails to indicate acceptance is semantic or
functional: even within cultures, silence has many different roles, many of
which are competing. Vernon J. Jensen (1973) is one of the first linguists to
discuss the communicative functions of silence, and he highlights five such
functions, each of which has a positive and a negative value:
1. A linkage function: Silence may bond two (or more) people or it may
separate them.
2. An affecting function: Silence may heal (over time) or wound.
3. A revelation function: Silence may make something known to a person
(self-exploration) or it may hide information from others.
4. A judgmental function: Silence may signal assent and favor or it may
signal dissent and disfavor.
5. An activating function: Silence may signal deep thoughtfulness (work)
or it may signal mental inactivity. (Jensen, 1973, pp. 249–255)9
This list identifies not only five different functions of silence, but also two
different values within each function. Moreover, which function and value
is at work in a particular conversational exchange is far from transparent to
the conversational participants. Silence may, for instance, convey deception
or assent, disagreement, or distance, without any clear signal which of these
is in fact operative.
Finally, silence might be the result of purely pragmatic factors: we have
many obligations and aims, and only so many hours in a day. Thus, silence
might be due to distractions, exhaustion, greater priorities, indifference, and
so on. I suspect most of us would be unable to count how many times we’ve
been silent in a conversation simply because our minds are elsewhere—
perhaps we have an ill relative, or a pressing deadline, or errands to run.
Indeed, it is arguable that this explanation is the most problematic for Gold-
berg’s view, as such purely contingent, highly contextualized features are
not only widespread, but might be entirely opaque to conversational par-
ticipants. I hardly know when my own family members are preoccupied, let
alone all of the participants in my daily conversational exchanges. Given
this, were the cooperative conversation view correct, I would be enti-
tled to infer acceptance from silence all over the place when in fact people
are often just tired and busy.
What these considerations highlight are two features of silence, the com-
bination of which render the cooperative conversation view particularly
problematic: a heterogeneity with respect to the reasons for silence and a
lack of transparency in what these reasons are. As we have seen, the causes
of silence are wildly and importantly diverse. This, by itself, poses a signifi-
cant challenge to the cooperative conversation view, for Goldberg claims
that silence defeasibly indicates assent and yet there is a significant range
of explanations for silence, especially ones that are dependent on diverse
features, such as social status, race, culture, personality, mood, daily events,
and so on. But when this is coupled with the fact that such reasons are often
entirely opaque to conversational participants, it becomes highly question-
able whether we can—in general—reasonably assume anything at all from
silence.
Of course, as should be expected, Goldberg is aware that not every
instance of silence indicates acceptance, and so this is why he says that the
entitlement we enjoy to draw such an inference is defeasible. According to
Goldberg, the entitlement can be defeated by either: (i) non-conversation:
The particular speech exchange is not a conversation—it is not a coopera-
tive exchange—in the first place; or (ii) outweighing explanation: The best
explanation of the listener’s silence appeals to considerations that outweigh
the hearer’s conversation-generated reason to be helpful. I take it, then, that
the five reasons highlighted above for silence would be understood as vari-
ous kinds of defeating conditions. People without power might not be part
of cooperative conversations in the first place, and so a prisoner’s silence
need not indicate acceptance of what a correctional officer says—it may
instead be the result of being or feeling “silenced.”10 Or the best explanation
of a person’s silence might be shyness or preoccupation, and so the default
entitlement to infer assent would be defeated.
90 Jennifer Lackey
To my mind, however, this approach is misguided. Defeating condi-
tions ought to be such that they pick out the non-normal or unusual
against a background of what is normal, the latter being the default.
Thus, on the cooperative conversation view, silence indicating accep-
tance should be the norm, and this is precisely what purportedly gives
it such powerful explanatory value. But this is simply not the case. The
reasons for silence are as deep as they are varied, and they are woven
into the very fabric of our interaction with others. No conversation is
entirely free of differences in the distribution of epistemic goods, status,
power, psychology, cultural expectations, practical constraints, or some
combination thereof. Speaking up against others almost always involves
a calculation—whether conscious or not—that is based on one’s position
and the costs and benefits of dissent on this topic at this time with this
conversational participant.
There is also a point to be made against the support that Goldberg offers
on behalf of the cooperative conversation view. Recall that he provides a
number of quotations that purport to show that, as a matter of fact, we take
silence to indicate acceptance. However, these quotations don’t strike me as
compelling evidence on behalf of his claim, as they seem more normative
rather than descriptive. That is, rather than describing how silence in fact
functions, they seem to aim at motivating people to speak out when they
disagree, especially when the stakes are high. Consider, for instance, Martin
Luther King Jr. saying that “There comes a time when silence is betrayal.”
Notice that King is not saying here that silence always means betrayal but,
rather, that there comes a times when this is the case. In the context of the
speech in which this assertion is offered, he is referring to the Vietnam War,
and he goes on to say that he is making “a passionate plea to [his] beloved
nation” to speak out against it. Nothing about this suggests that King takes
himself to be describing our ordinary conversational practices. In fact, if he
were, it is entirely unclear why he would need to make a passionate plea to
the nation to do what we all are doing anyway. Instead, King is naturally
read as providing a motivational, rather than a merely descriptive, claim
here, and he is plausibly interpreted as saying that the reason the time has
come for silence to be understood as betrayal is because the stakes have
gone up with respect to the Vietnam War. This is at odds with the coopera-
tive conversation view that silence has this same significance in any ordi-
nary conversation.
Similar considerations apply regarding the quotation from Rabbi Bradley
Artson when he remarks that “our silence and inaction in the face of con-
temporary injustice and oppression is akin to assenting to it.” Once again,
notice that he is singling out high-stakes cases—ones where injustice and
oppression are at issue—rather than run-of-the-mill instances of silence.
Moreover, like King, Artson is making this claim in the context of trying
to encourage people to speak out against wrongs that they might other-
wise ignore.11 Thus, if people need to be motivated to object in the face
of disagreement, then Goldberg’s claim that silence, by default, indicates
acceptance is implausible.
There is a still further point against the cooperative conversation view
that should be emphasized: in order for silence to indicate acceptance, not
only does it have to be the case that there is the norm to object that Gold-
berg describes, but it also needs to be followed with at least some general-
ity and regularity. In particular, most conversations need to be cooperative
ones, for it is only the uncooperative nature of silent rejection that enables
us to infer assent from silence on this view. But why should we think that
this is the case? As indicated above, silence can be the result of wildly het-
erogenous factors, many of which are the result of asymmetries in power.
Are poor black women generally in cooperative conversations with wealthy
white men; are untenured faculty members generally in cooperative conver-
sations at department meetings with tenured colleagues; are those who are
incarcerated generally in cooperative conversations with those who are not?
While these are empirical questions, there is evidence, some of which was
offered above, that supports negative answers here. But even if the answers
are instead positive, there is still the issue of the lack of transparency—how
can I tell which ones are cooperative and which ones aren’t? Otherwise put,
how could I possibly infer that your silence means that you’re assenting
when I have no idea whether we are part of a cooperative conversation in
the first place? I might, for instance, do everything possible to cultivate an
environment in which my students feel comfortable to express themselves in
conversation with me, but it obviously doesn’t follow from this that they do.
At this point, I have argued in a fair bit of detail that the cooperative
conversation view faces significant challenges. My purpose in doing so is
twofold: first, I want to show that there are independent reasons to reject
Goldberg’s account, both of what can be inferred from silence and what this
reveals about our duty to object in conversational contexts. Second, and
perhaps even more importantly, I want to reveal the extent to which the
cooperative conversation view gives rise to an ideal theory of the duty to
object, and how such an approach leaves out the way that silence functions
in the actual world, especially for those who are systematically marginal-
ized. If our theoretical starting point is not a conversation in which everyone
has the privilege of being cooperative, but, rather, one in which features of
the actual world take center stage—such as power, oppression, job insecu-
rity, limited resources, cultural differences, and so on—we do not end up at
a place where inferring assent from silence seems plausible. Instead, we find
ourselves recognizing that objecting is often a luxury, one that not everyone
can afford to make.
Non-Ideal Theory
In this section, I will defend one feature of the duty to object that emerges
when we begin, not with the ideal, but with the actual.12 Otherwise put,
92 Jennifer Lackey
I will argue that a non-ideal theory of the duty to object leads to the accep-
tance of the following condition:
social status: One’s duty to object can be directly influenced by one’s

social status.
By “social status,” I include not only those properties that contribute to

differences in power, paradigmatic examples of which are race, sex, gen-
der, and class, but also properties that are more epistemic in nature, such
as authority and expertise, which often accompany professional roles. For
instance, the social status of the President of Northwestern includes not just
that he is a white male, but also that he is an economist and an administra-
tor with years of experience.
If social status is correct, then people have different obligations or are
under various kinds of normative pressure to express their dissent, depend-
ing on who they are and what social position they occupy. Unlike Goldberg’s
account—according to which, absent defeating conditions, an undergradu-
ate student, an incarcerated man, and a black woman all have the same
normative pressure to object as a white, male CEO does—my view holds
that such people quite literally do not have the same duties. Given this,
the default cannot be that silence conveys acceptance because what silence
indicates is normatively linked with the particulars of the individual who
is silent. So, for instance, if a tenured, white, male professor hears a fellow
colleague make a clearly sexist remark, his duty to object might be greater
than that of his black, female, junior colleague.
To see this, I will offer considerations on behalf of this conclusion from
both an epistemic and a moral point of view. Let’s consider the former first:
with great power in a domain often comes greater authority, and thus an
increased likelihood that one’s testimony will have an effect. So, if we assume
that the sexist remark in question is false and that one of our aims as epis-
temic agents is to promote the truth, then the white professor objecting to it
might have more epistemic impact in producing true beliefs, both at the indi-
vidual and the collective level. That this is the case is supported by a recent
study by Kevin Munger (2017), in which he looks at the impact of calling out
racist harassment online via Twitter. To this end, he used “bots” to object to
harassers, varying their identities not only by in-group (white man) and out-
group (black man) membership, but also by the number of Twitter followers
each bot has. Munger found that subjects who were sanctioned by high-
follower white males significantly reduced the use of racist slurs, leading to
the following title of a recent article in The Atlantic: “Why Online Allies
Matter in Fighting Harassment: A clever experiment with Twitter bots shows
that telling people not to be racist can work—but only if it comes from some-
one influential and white.”13 In this same article, the author writes:
“There’s a reason why higher-status members of these communities bear

a larger share of the responsibility for speaking out against racist or
bigoted speech,” says Betsy Levy Paluck, a psychologist at Princeton
University. “This isn’t just a moral judgment but an empirical regularity
that’s been coming out of many research programs: People with higher
status are influencing norms, and with that influence comes responsibil-
ity. If anyone says, I’m not a role model, that’s a wish, not a fact.’’
Here are other examples of this kind: a pediatrician speaking out about
the safety of vaccines, a prosecutor objecting to police misconduct, and a
university official condemning ineffective sexual assault policies on campus.
In each case, the duty to object might be greater for the person in question
than it is for the average citizen in large part because that person’s objecting
is likely to lead to more true, and fewer false, beliefs.
Notice that one assumption operative in my argument here is that our
epistemic duties extend beyond our own beliefs as individuals to include
those of others in our broader communities. While traditional epistemology
focuses almost entirely on obligations with respect to our own beliefs, I see
no reason why we shouldn’t also be concerned with the beliefs of others.
Indeed, it seems clear that we are subject to criticism for knowingly promot-
ing, or even permitting, at least some false beliefs in those around us. If, for
instance, I know that you falsely believe that our colleague cheated on his
partner, or that classes are canceled tomorrow, or even that a local restau-
rant serves vegetarian food, it is objectionable for me to do nothing at all to
correct your belief. Moreover, the false beliefs of those around me will very
likely beget false beliefs in others, eroding the competence and trust rela-
tions of our epistemic communities. Given this, combined with the fact that
one of the central epistemic goals is to maximize true beliefs and minimize
false ones, it follows straightforwardly that those whose voices will have a
greater epistemic impact have a greater duty to use them.
Of course, this argument should not be understood as in any way devalu-
ing the voices of those who already marginalized. Instead, the point is that
in many cases it will be supererogatory for members of the community who
have lower social status to object to what they take to be false or unwar-
ranted, and thus they will be deserving of praise and admiration for doing
so. In contrast, for those who have higher social status, they will often be
doing the minimum that is expected of them when they object, and failure
to do so will leave them open to disapprobation.
Similar considerations apply at the moral level. Surely there is more
moral pressure for the tenured, white, male professor to object to the sexist
remark than there is for his junior colleague, both because he has the social
standing to bring about greater positive change and because there is less
risk of harm for him. For instance, drawing on the Munger research again,
if those with higher social status are more likely to reduce racist slurs, then
their objecting will have a greater chance of eliminating both false beliefs
and wrong or harmful actions by successfully cultivating moral communi-
ties in which there is less racism overall. Moreover, the stakes are typically
far lower for those with higher social status, rendering them less vulnerable
94 Jennifer Lackey
for speaking out, especially about contentious matters. The tenured white
professor, for example, has political, professional, and economic advan-
tages that make him far less exposed to retaliation or other adverse effects
for raising objections. He doesn’t have to worry about being regarded as
stereotypically angry or whiny, and he doesn’t risk losing his job and finan-
cial stability.
Still further, a common concern expressed by some members of marginal-
ized groups, particularly by Black Americans, is the overwhelming burden
that comes with having to constantly explain their experience to others and
challenge the racism pervasive around them, both at the systemic and indi-
vidual levels.14 Their daily lives are described as “exhausting,” often depriv-
ing them of time and energy crucial for making progress in other areas of
value. Given that those with lower social status are typically already shoul-
dering a greater epistemic and moral burden in simply trying to navigate a
system designed to exclude or oppress them, those who do not face these
barriers should step in and take on more when possible. Objecting to what
one takes to be false or unwarranted is precisely an area where this can and
should be done.
A related, though more general, argument on behalf of social status can
be given by focusing on some of the objections raised against the coopera-
tive conversation view. In particular, recall my claim that objections with
traction are a limited epistemic good and, as with all other goods, some
have far more of them than others. Now, notice that we often have different
normative expectations of people, given the amount of goods they possess.
For instance, we clearly have different views about the level of charitable
giving Bill Gates ought to engage in compared with someone who is living
on minimum wage. Indeed, if Bill Gates did not make any monetary contri-
butions to charities, we would be critical of him, perhaps even strongly, but
we have no similar expectations of those with far fewer resources. Similar
considerations apply when the goods are epistemic. Suppose, for instance,
that I have platforms with audiences; I’m listened to and respected; I have
the competence to develop compelling objections; and I have power and
authority. In such a case, I would have a lot more objections with traction
than someone without these privileges, and so just as we expect Bill Gates
to engage in significant charitable giving because of his tremendous wealth,
we should expect me to use my voice more than others who have far fewer
social and epistemic goods.
Conclusion
In this chapter, I’ve argued against an ideal theory approach to understanding
the duty to object, which is exemplified with the cooperative conversation
view, and I’ve taken a first step toward showing how to theorize about this
duty within a non-ideal framework. One conclusion that quickly emerges
is that one’s social status plays a significant role in whether one has such a
duty in the first place, and relegating the differences between us in social sta-
tus to “defeating conditions” masks the critical role they play in our norma-
tive lives. Once we see this, it becomes clear that we cannot determine when
we ought to voice dissent without first looking carefully at our positions of
power and privilege, or lack thereof.
Notes
1 While I will talk about there being the “duty” to object, for those who don’t like
talk of duties, this can be substituted with “obligations,” “demands,” “norma-
tive force,” or “normative pressure,” which I will often use interchangeably with
duties.
2 See Lackey (forthcoming).
3 Goldberg (unpublished).
4 For a detailed discussion of this sort of entitlement, see Goldberg (forthcoming).
5 Source: www.informationclearinghouse.info/article2564.htm, accessed 4 July
2016.
6 Source: www.myjewishlearning.com/article/no-neutrality-silence-is-assent/,
accessed 4 July 2016.
7 For a helpful discussion of the distinction between rejecting and ignoring testi-
mony, see Wanderer (2012). For work on “silencing,” see Hornsby (1995) and
Langton (1993
8 See also Tuerkheimer (unpublished) for a detailed discussion of some of these
issues in relation to reports of sexual assault.
9 See also Jaworski (1993).
10 For work on “silencing” in this sense, see Hornsby (1995) and Langton (1993).
11 Unlike King, however, Artson grounds his specific view in the Talmud.
12 I develop my positive view of the duty to object in Lackey (forthcoming).
13 Source: www.theatlantic.com/technology/archive/2016/11/why-online-allies-
matter-in-fighting-harassment/507722/, accessed 7 December 2016.
14 See, for instance, www.seattletimes.com/seattle-news/education/what-its-like-to-
be-black-on-campus-isolating-exhausting-calling-for-change/ and www.usato
day.com/story/life/people/2014/12/04/the-exhausting-task-of-being-black-in-
america/19894223/.
References
Broussard, P. A. (2013). Black women’s post-slavery silence syndrome: A twenty-
first century remnant of slavery, Jim Crow, and systemic racism—who will tell her
stories? The Journal of Gender, Race & Justice, 16, 373–421.
Goldberg, S. (Forthcoming). Should have known. Synthese.
Goldberg, S. (Unpublished). Assertion, silence, and the Norms of public reaction.
Hornsby, J. (1995). Disempowered speech. Philosophical Topics, 23, 127–147.
Jaworski, A. (1993). The power of silence: Social and pragmatic perspectives. New-
bury Park, CA: SAGE Publications.
Jensen, V. J. (1973). Communicative functions of silence. ETC: A Review of General
Semantics, 30, 249–257.
Lackey, J. (Forthcoming). The duty to object. Philosophy and Phenomenological
Research.
Langton, R. (1993). Speech acts and unspeakable acts. Philosophy and Public
Affairs, 22, 293–330.
96 Jennifer Lackey
Lebra, T. S. (1987). The cultural significance of silence in Japanese communica-
tion. Multilingua: Journal of Cross-Cultural and Interlanguage Communication,
6, 343–357.
Lorde, A. (1983). Black women writers at work (C. Tate, Ed., pp. 100–116). New
York: Continuum.
Mills, C. W. (2017). Black rights/White wrongs: The critique of racial liberalism.
Munger, K. (2017). Tweetment effects on the tweeted: Experimentally reducing rac-
ist harassment. Political Behavior, 39(3), 629–649.
Tuerkheimer, D. (Unpublished). Incredible women: Sexual violence and the cred-
ibility discount.
Wanderer, J. (2012). Addressing testimonial injustice: being ignored and being
rejected. The Philosophical Quarterly, 62(246), 148–169.
Yong, E. (2016, November). Why online allies matter in fighting harassment:
A clever experiment with Twitter bots shows that telling people not to be racist
can work—but only if it comes from someone influential and white. The Atlantic.
Web. Retrieved December, 2016.
6 For the Sake of Argument
The Nature and Extent of Our
Obligation to Voice Disagreement1
Disagreeing
When someone says something I believe to be false I often have the urge to
“correct” them. I feel compelled to make my disagreement known, either by
telling the speaker that I disagree or by explaining my own view. The extent
and strength of this urge varies with the situation and with the certainty
with which I hold my belief. If I am teaching ethics and a student tells me
that J. J. Thomson is pro-life, the urge is strong. If I’m sitting on the bus and
I overhear someone claim that Yemen is east of Oman, the urge is easily
ignored.
What is the nature of this urge? Is it merely a compulsion borne of too
much philosophy training? Is it the result of a kind of intellectual arrogance?
Is it my responsibility as part of the epistemic economy?
Recent work on disagreement, conversation, and assertion explores the
idea that we sometimes have obligations to make our disagreements known
to our conversational partners. Accounts disagree on the details, but I have
elsewhere argued that agents have a pro tanto obligation to voice disagree-
ment.2 I need not voice my disagreements when eavesdropping on the bus,
as my obligation is outweighed by other considerations; however, my obli-
gation is likely not outweighed in my ethics classroom. And, I’ve argued,
that my obligation is epistemic in character rather than merely moral or
practical. My argument for this is based on four background theories any
of which one might reasonably accept about epistemology and epistemic
values. One involves epistemic justice and other intellectual virtues, another
is based on the norms of inquiry, a third rests on joint commitment to epis-
temic ends, and a fourth relies on the nature of justification itself. Taken
individually, these background theories generate obligations to voice dis-
agreement in overlapping but distinct sets of cases. These theories might
also be used in combination. And, as I say, others have argued for similar
obligations to voice disagreement, albeit for different reasons.
If it is true that we have epistemic obligations to voice disagreement, a
number of questions arise. For one, we might wonder when we are required
to voice our disagreements—that is, when aren’t the pro tanto obligations
outweighed? We might further wonder whether these background theories
obligate us to play the devil’s advocate, in addition to voicing our sincerely
held disagreements. Are we required, in other words, to pretend to or to
behave as if we disagree when in fact we do not? Are we obligated to play
devil’s advocate in the same ways we’re obligated to voice sincere disagree-
ment? These are the questions I will explore in this chapter.
I’ll start, in the next section, with a brief discussion of devil’s advocacy.
Then, in section 3, I’ll put our two questions to each background theory.
This will allow us to being to investigate the nature and extent of our obliga-
tion to voice disagreement.
Playing the Devil’s Advocate

A headline from The Onion in 2007 reads, “Man Who Plays Devil’s Advocate
Really Just Wants to be Asshole.”3 The satirical article describes a man asking
inappropriate questions while disclaiming sincere curiosity as to the answers.
This headline works as a joke, in part, because it gets something importantly
right about playing devil’s advocate. But what does this practice involve?
I won’t try to give necessary and sufficient conditions for devil’s advo-
cacy. This practice is familiar from our actual conversational lives, and so is
bound to be multifarious and slippery. Instead I’ll offer a characterization
of a central or paradigm case of devil’s advocacy, acknowledging that actual
cases of the practice are likely to vary from this central case in a variety of
ways and to a variety of degrees.
Paradigmatically, an agent who plays devil’s advocacy announces her
intention to defend a position she doesn’t hold, and then defends that posi-
tion in order to make progress on the issue at hand. So, when a group of
executives has decided to fire the northwest marketing team, Martha might
say, “look, I’m just playing devil’s advocate, but they do have some impor-
tant clients.” Martha expresses a view she does not hold, while flagging that
this is what she’s doing. This is admittedly a limited paradigm case, but it
seems to capture some important characteristics of devil’s advocacy.
The devil’s advocate in The Onion asks inappropriate questions; however,
questions are only one of the ways to play devil’s advocate. We can make
statements, defend claims, and create whole theories as devil’s advocate. It
is also clear that devil’s advocacy is only sometimes inappropriate. We play
devil’s advocate by asking questions in philosophical debates about the most
abstract issues—surely these questions aren’t (always) inappropriate. Mar-
tha’s assertion to the executives is not clearly inappropriate.
One hallmark of playing devil’s advocate is a lack of belief in the content
expressed. Martha’s point is to consider a position that neither she nor the
other executives holds. The man in The Onion disclaims sincere curiosity
about the answers to his questions, which allows him to claim he’s merely
playing devil’s advocate. If I say, “just to play devil’s advocate” and then
raise a point I sincerely believe, then I’m likely using the “devil’s advocacy”
For the Sake of Argument 99
language rhetorically (or to hide my sincere belief). I might have good rea-
son to do this, as when I’m expressing an unpopular or taboo position that
I sincerely hold, but this is a derivative use of the label of devil’s advo-
cate rather than the paradigm. Paradigmatically, the term devil’s advocate
describes someone who, given a certain point of view, takes a position she
does not necessarily agree with for the sake of debate or to explore the
thought further.4
While the practice is common in philosophy, the philosophical work on
devil’s advocacy is limited. Scott F. Aikin and J. Caleb Clanton in their paper
on group-deliberative virtue describe devil’s advocacy as a potentially virtu-
ous way of being overtly and openly uncooperative with other members of
your group during deliberation. Aikin and Clanton are interested in dev-
il’s advocacy in the context of group deliberation. This is because, as they
argue, groups can be more effective than individuals when they deliberate
virtuously, and virtuous devil’s advocacy can contribute to virtuous delib-
eration. According to Aikin and Clanton,
what the virtuous devil’s advocate does first is announce that she will
not be representing her own views, but that of some other person or
group whom the arguer takes as wrong or ripe for criticism. The dev-
il’s advocate, then, nevertheless tries to be true to those views in the
discussion.5
Devil’s advocacy is a familiar enough practice, but for the following dis-
cussion it is important to illuminate the differences between playing devil’s
advocacy and arguing for or defending one’s own sincere beliefs. In my pre-
vious work on obligations to voice disagreement, I only argued that agents
are obligated to voice sincere disagreements. One of the questions for the
next section, then, is whether or not the same sources that generate the
obligation to voice sincere disagreement also generate an obligation to play
devil’s advocate.
Four Sources of Epistemic Obligation

I have argued that belief in any of four epistemic theories should motivate
belief in an epistemic obligation to voice disagreement. In taking these theo-
ries on, in being committed to them, one is also committed to the epistemic
obligation that agents have to voice their sincere disagreements in at least
some contexts. Each of these background theories is inspired by the work
of a different philosopher: Miranda Fricker’s work on epistemic justice
and related topics inspired the first, which we’ll call the Frickerian theory.
The second, inspired by Catherine Z. Elgin’s work in the philosophy of sci-
ence, we’ll call the Elginian. Margaret Gilbert’s work on joint commitment
inspired the third, so it will be called the Gilbertian theory. And John Stuart
Mill inspired the fourth, so we will call it the Millian background theory.
In this section I will sketch each of these background theories and the
manner in which each generates an obligation to voice disagreement. I’ll
then examine each one using two questions: i) in what situations does the
obligation remain and when is it outweighed? ii) does the source in question
generate an obligation to play devil’s advocate?
Let’s begin with the Frickerian background theory. Fricker offers a theory
of epistemic justice and injustice.6 As part of this theory, she describes a
variety of features that are important for a knower’s epistemic well-being.
One component of epistemic well-being is that knowers are able to testify
and have their testimony taken seriously—they are able to participate in the
epistemic economy. One of the dangers of epistemic injustice is that the vic-
tims thereof lose confidence and internalize the credibility deficits they suffer
in the eyes of their prejudiced interlocutors. This makes them epistemically
worse off7—it harms them in their capacities as knowers.
Voicing disagreement, for some agents in some contexts, helps to miti-
gate epistemic injustice. This can happen in a number of ways. Consider, for
example, Stella, the only female mechanic working in an auto body shop.
Stella might be reluctant to voice her disagreements with the other mechanics
because of a history of her contributions being given a credibility deficit, or
because she feels stereotype threat, or because she’s internalized stereotypes
about female auto mechanics and so lacks confidence in her assessments. If she
voices disagreement despite this reluctance, she may be able to build her con-
fidence, may be able to participate in the exchange of ideas in the shop, and
may come to feel identity prejudice less. She is building the skills and contrib-
uting to a healthier epistemic economy. Because of the ways in which creatures
like us rely on one another to flourish epistemologically,8 participating in the
epistemic economy, and exercising one’s epistemic capacities, contributes to
that flourishing. So, Stella has a self-regarding duty to voice her disagreement.
Voicing disagreement can contribute to epistemic justice when it is voiced
in conversations with those who hold the disagreeable position, but it can
also be helpful when it is voiced in other conversations. We can see this in
contexts in which subordinated agents, or agents in an unpopular minor-
ity, practice voicing minority views in enclaves of like-minded interlocu-
tors. Doing so can help to amplify voices that would otherwise be silenced,
thereby potentially contributing to epistemic justice.9 A socially subordi-
nated knower might exercise her epistemic capacity to voice disagreement in
a safe (or safer) context before doing so in conversation with interlocutors
with whom she disagrees.
Of course, for voicing disagreement to do these things for Stella, sev-
eral factors must be in place—and here we turn to our first question. In
what kinds of contexts does Stella have a self-regarding obligation to voice
her disagreement? In other words, when should she, all things considered,
express her disagreement and thereby develop her epistemic skills?
The context in which Stella voices her disagreement must have a couple of
characteristics. First, Stella has to have a psychological disposition that will
allow her to benefit from voicing disagreement in that context. I’ll refrain
from speculating too much on the nature of the necessary psychology, but
I do feel confident that there are some (at least possible) agents for whom
voicing disagreement would undermine epistemic skill building because of
their psychology. If a student is deeply shy, for example, or an employee is
underprepared, voicing disagreement in some context could be epistemically
detrimental. This same person could, in some other context, nonetheless
benefit.
Second, it must be the case that Stella’s voiced disagreement in this context
has no lasting negative epistemic repercussions for her. If Stella’s voiced dis-
agreement results in her removal from discussion with her coworkers, i.e.,
because she is fired or physically removed, then she has lost access to that
epistemic resource. If Stella’s community is so intransigently biased against
women that they’re willing to fire her over disagreement, then it is not clear
that voicing disagreement will help Stella participate in the epistemic econ-
omy. It is unlikely that Stella would be able to find work as a mechanic
in such a community in the first place; however, in such a community her
voiced disagreement is unlikely to be epistemically beneficial. I think there
are likely to be many circumstances in which both of these factors are in
place, but if either fails to obtain, then voicing disagreement would not be
too risky and so Stella’s obligation to voice her disagreement is outweighed.
The second thing to explore regarding the Frickerian source is the question
of devil’s advocacy. If we have self-regarding duties to voice our disagree-
ments in order to develop our epistemic skills and flourish as an epistemic
agent, do we also have an obligation to play devil’s advocate? The answer
depends on whether devil’s advocacy contributes to our epistemic skills and
flourishing.
One can easily imagine contexts in which playing devil’s advocate could
contribute to an agent’s epistemic flourishing. In philosophy seminars, for
example, judicious devil’s advocacy is expected and can be an important
intellectual and philosophical skill. Similarly, a disinterested conflict media-
tor ought, sometimes, to take a position she does not necessarily agree with
for the sake of debate or to explore a thought further. In these contexts,
devil’s advocacy is an important skill. However, the context in which dev-
il’s advocacy contributes to epistemic flourishing are more specialized than
those in which voicing sincere disagreement contributes. Consider Stella’s
case. For Stella, voicing sincere disagreement contributed to her over coming
stereotype threat and participating in the epistemic economy. Her expres-
sion of a sincere belief is instrumental to those ends. It is possible that voic-
ing some imagined opponent’s position and advocating for it could be useful
for her. However, for her to fully participate in the epistemic economy, there
is something important about sincere expression of her own position. So,
even if we think that playing devil’s advocate is a useful skill for participa-
tion in the epistemic economy, the benefits of doing so are distinct from
those of sincere dissent.
The benefits of playing devil’s advocate are distinct from those garnered
by practicing voicing one’s own views. We should expect the skills built by
these practices to be distinct as well. Stella’s ability to identify and express
what she thinks is distinct from her taking on or parroting or even imag-
ining someone else’s position (which she can declaim). While both might
be important, sincere participation is riskier and more authentic. So, while
there is a self-regarding duty to play devil’s advocate, it is a distinct and more
specialized duty than the duty to voice sincere disagreement. The Frickerian
background theory generates both an obligation to voice disagreement and
an obligation to play devil’s advocate but these obligations are distinct and
are likely salient in different contexts.
Now we turn to our second background theory, inspired by work from
Catherine Z. Elgin. Elgin argues that, “Science requires collaboration. [And]
that collaboration requires trust.”10 Science (and other epistemic pursuits
like art and auto mechanics)11 requires collaboration and trust, she argues,
for the same reason it requires models, inference, and idealizations: the sub-
ject matter is too vast and complicated, and our findings are too fallible for
any scientist to make much progress on her own. If a scientist is to make
progress she must be able to trust her fellow scientists to let her know when
they believe she’s making a mistake, or following a bad line of inquiry. In
other words, she needs to be able to trust her fellow scientists to voice their
disagreements with her findings or her methods. And she needs to be able
to rely on their collaborative contributions. Participation in collaborative
epistemic pursuits, then, generates the epistemic duty to voice at least some
disagreements: collaboration and trust aren’t possible if agents keep all of
their disagreements to themselves.
To see this, consider an example. Theresa and Carla are art historians.
Carla is presenting some work at a casual faculty working group. As an
aside in the course of her presentation, Carla claims that, “Babylonian
architecture influenced Assyrian ceramic art.” Theresa, who has an inter-
est in Assyrian ceramic art, disagrees. She does not voice this disagreement
in the Q and A, nor does she bring it to Carla’s attention after the session
is over. The intuition in this case is that Theresa has failed to do as she
epistemically ought. She has an epistemic obligation to tell Carla about her
disagreement. Elgin’s theory predicts this because art history, as a kind of
inquiry, is of necessity collaborative. Collaboration requires trust and trust
requires collaborators to say when they disagree.
Turning to the first of our questions, then, in what contexts ought we to
voice disagreement according to an Elginian background theory? There are
two conditions under which we are obligated to voice our disagreement,
both of which are necessary for the obligation: first, we must be participat-
ing in a pursuit that requires collaboration, and second, we must be in a
context in which trust of the relevant kind is possible.
Examples of pursuits that require collaboration are widespread, including
science, art, history, and auto mechanics as Elgin highlights. Each of these
pursuits requires that we rely on information from others. For me to know
a carburetor from a piston requires some reliable source—someone in the
know needs to have told me.12 Indeed, I suspect that almost any pursuit that
involved epistemic ends requires collaboration at some stage or other. It
might be possible to imagine an epistemic situation in which all the relevant
information was available to the agent through individual investigation and
introspection. In a case like this, where an agent could proceed without
help, voicing disagreement would be neither useful nor obligatory. However,
this would be a highly unusual kind of epistemic pursuit. In most cases of
inquiry, collaborators have to have some degree of trust and so have an obli-
gation to make at least some of their disagreements known to one another.
It is, perhaps, more plausible that there would be cases of collaborative
epistemic pursuits wherein trust of the relevant kind is impossible. If my
computer is on the fritz and is not performing basic calculations correctly,
then I will not trust its verdict on more complicated cases. Similarly, if my
interlocutors and I cannot agree on basic premises, then I will not trust their
assertions at later stages of inquiry. In a case like this, or in a case in which
my interlocutors are universally or even frequently duplicitous, no trust is
possible. If no trust is possible, trust cannot obligate us to voice disagree-
ment. And agents engaged in this kind of inquiry are unlikely to make much
progress.
Turning now to the second of our questions, if collaboration and trust
obligate us to voice disagreements, do they also obligate us to play devil’s
advocate? I think they do not—or at least not for the same reasons that
obligate us to voice sincere disagreement. There may be a restricted set of
pursuits in which playing devil’s advocate contributes to the trust necessary
for collaboration, but this would be a very special circumstance. In general,
we do not fault our collaborators for failing to play devil’s advocacy the
way we do when they fail to tell us that they disagree. It would be unlikely
to erode my trust in an interlocutor if she never played devil’s advocate. If
she failed to tell me when she disagreed, however, I would likely not want to
work with her in any real collaborative way in the future,
This is not, despite appearances, contrary to the defense of devil’s advo-
cacy that Aikin and Clanton offer. Aikin and Clanton are interested in dev-
il’s advocacy as it contributes to good argument and careful deliberation,
rather than as it is necessary for trust. This is not to say that good argument
and careful deliberation don’t involve trust at all. Aikin and Clanton rightly
defend a sincerity condition on virtuous devil’s advocacy, and it is plausible
that sincerity in general is instrumental to developing trust. However, the
obligation to voice disagreement arises, on the Elginian view, more gener-
ally and for different reasons than even the obligation to be a sincere devil’s
advocate. This is because voicing sincere disagreement in at least some cases
is necessary for trust, while playing devil’s advocate, even sincerely, is not.
Let’s look now at the Gilbertian background theory and the obliga-
tion it generates to voice disagreement. On Gilbert’s view, some of our
commitments—in particular the commitments we make jointly with
others—generate obligations to those others. If Gilbert is right about this,
then when we’re jointly committed to discussing some topic or pursuing a
line of inquiry, this joint commitment generates an obligation to tell our con-
versational partners when we disagree. Or so I’ve argued in previous work.
The obligation to voice disagreement is, I think, central to the joint com-
mitment to pursue a line of inquiry. To see why, consider one of Gilbert’s
central examples: taking a walk together.13 Imagine that one day at lunch
time, you’re walking toward the coffee shop and I’m walking toward the
coffee shop. Our paths briefly converge. After a few minutes, I turn and
walk toward the library, leaving you to walk toward the coffee shop on
your own. There is, Gilbert points out, no sense that I’ve behaved badly in
this scenario. You have no claim against me, and I’ve violated no obliga-
tions to you. Now imagine that I invite you to walk to the coffee shop with
me. If you accept, then our walk to get coffee is something we do together,
rather than merely at the same time. This, according to Gilbert, is what it
takes for us to be jointly committed to walking. Because we’ve jointly com-
mitted to taking this walk together my turning toward the library would be
defecting—it would be violating my obligation to walk with you.
With that understanding of joint commitment in mind, imagine again
a case in which I’ve invited you to walk with me to the coffee shop. We
set off at lunch time toward the coffee shop, chatting pleasantly about the
weather. When we come to a fork in the path, I turn left, away from the
coffee shop. The geography of the path makes it the case that I’m either
taking a very circuitous route, or not going to the coffee shop at all. At this
point, you have a couple of options. You can turn back, defecting from our
joint commitment, you can proceed with me, accepting that we’re unlikely
to get coffee, or you can voice an objection to our route. And, if Gilbert is
to be believed, you are obligated to take the third option. We have jointly
committed to walking together to the coffee shop. Defecting without expla-
nation, or walking somewhere else, constitutes a failure to live up to our
commitment-generated obligations.
The same kind of obligation is generated when we’re jointly committed
to inquiry. When you and I find ourselves inquiring together, our commit-
ment generates obligations for us. If you find me pursing a line of inquiry
or proposing a claim that you find objectionable, then you’re obligated to
tell me so. Failure to do so constitutes acquiescence to my objectionable
claim, or defection from our joint epistemic activity. And so, the Gilbertian
background theory generates obligations to voice disagreement for agents
jointly engaged in inquiry.
Of course, if Jeff is not jointly committed to inquiring with Janie, then Jeff
is not obligated to voice disagreement if Janie’s proposal is objectionable to
him. However, Gilbert is clear that not all joint commitments are formed
on the basis of official or explicit invitation. Some come as a result of tacit
agreement or casual association. If we’re committed to doing some joint
activity, we’re jointly committed. Plausibly, participation in a scientific field
or an area of inquiry generates these joint commitments without an explicit
invitation or explicit undertaking of commitment.
For the Gilbertian, it is not at all clear how an obligation to play devil’s
advocate would arise, except in very specific cases. If we jointly committed
to considering objections to our views, as we are in the philosophy seminar
room, then that joint commitment obligates us to occasionally play devil’s
advocate. But notice, the obligation to voice disagreement arises in all cases
of joint commitment to inquiry. The philosophy seminar room is a restricted
subset of that kind of practice. And further, the obligation to play devil’s
advocate does not arise in virtue of objecting to a proposal or line of inquiry.
So, in answer to our second question, for the Gilbertian, the obligation to
play devil’s advocate and the obligation to voice disagreement are different,
differently generated, and will arise in different contexts.
We turn, finally, to the Millian background theory. Mill provides, in some
ways, the most direct case that we have an obligation to voice disagree-
ments. Mill says,
Complete liberty of contradicting and disproving our opinion is the

very condition which justifies us in assuming its truth for the purposes
of action; and on no other terms can a being with human faculties have
any rational assurance of being right . . . it has been [the wise person’s]
practice to listen to all that could be said against him; to profit by as
much of it as was just, and to expound to himself, and on upon occa-
sion to others, the fallacy of what was fallacious.14
According to the received reading of this quote, we have good epistemic rea-
son to listen to all objections that our interlocutors make to our views. This
is part of why, according to Mill, we can have no justification for restricting
the freedom of speech except in very specific contexts. Notice, though, that
Mill also claims that, for us to be wise we must also point out to others
when they’ve made a mistake. Because of what it takes to be justified, we
must sometimes voice our disagreement.
Mill does not say much more about the contexts in which voicing dis-
agreements is obligatory. He appears to believe that it is a very general kind
of obligation in cases where one seeks justified belief. He says that it is the
process of considering what can be said against a view that justifies that
agent in believing that view. And he argues that there are intellectual ben-
efits can be gained by “placing human beings in contact with persons dis-
similar to themselves, and with modes of thought and action unlike those
with which they are familiar.”15 So, there will be many familiar contexts in
which we’re obligated to voice sincere disagreement. Picking an ordinary
case, imagine that Anna believes that the Shop-for-less has better quality
peaches for sale than the Save’n’Buy. For Anna to be justified in that belief,
she has to listen when Bradley says he prefers the peaches at Save’n’Buy.
This is the case whether or not Anna has voiced her view. So far, this is
familiar. However, according to Mill, she must also tell Bradley that she
disagrees with him. In order to be justified in her belief, she has to voice
her disagreement with Bradley’s claim. And this is about as general of a
conversational context as we can get. We need not imagine any history or
background for Anna and Bradley to get the obligation going, according to
the Millian background theory. Because we are epistemically obligated to
hold justified beliefs, we are epistemically obligated to voice our disagree-
ments, at least some of the time.
What about devil’s advocacy? Does the Millian background theory gener-
ate an obligation to play devil’s advocate? Mill certainly appears to be in
favor of devil’s advocates. He even mentions the practice, remarking that
even in the Catholic Church (which he takes to be a close-minded institu-
tion), there is a place for a devil’s advocate in deliberation. However, every-
thing he says about devil’s advocates seems to oblige us to listening to them.
Are we similarly obligated to play devil’s advocate?
There are, I think, two ways of using a Millian background theory here.
One, which we might call the strong reading, holds that devil’s advocacy
is good for us epistemically, so we are sometimes obligated to play devil’s
advocate. There are certainly passages from Mill that suggest this read-
ing. One might think, for example, that “listening to all that might be said
against” a belief requires imagining hypothetical opponents—or sometimes
pretending to oppose a position or view you in fact believe. Certainly, Mill
warns that our favorite view, “however true it may be, if it is not fully,
frequently, and fearlessly discussed, it will be held as a dead dogma, not a
living truth.”16 This, too, suggests that we might need devil’s advocates who
will pretend to hold views they in fact do not. And if we need them, perhaps
we’re sometimes obligated to play them.
One problem with a Mill-inspired obligation to play devil’s advocate is
that it is implausible, given the other things that Mill claims, that any par-
ticular person should be responsible for playing the devil’s advocate on all
topics or even in any particular situation. Mill is clear that he advocates we
be free to express our opinions, and an obligation to assume one’s oppo-
nent’s views does not sit well with that. It is possible, however, to build from
a Millian background to a theory that specified the context and situations in
which a particular agent might be obligated to play devil’s advocate.
On the other hand, there is another reading of Mill’s background theory
available, which we can call the weak reading. Millian obligations to listen
to and occasionally voice disagreements seem to be predicated on the idea
that the positions being expressed are the actual opinions or beliefs of the
agents in question. Mill’s position assumes that each agent is doing their
best to believe truly—Mill suggests as much when he talks about the desire
to assume the truth of some claim, of disproving opinions, and of fallacious
beliefs. He also speaks of sincerity of belief as a hallmark of a well-meaning
interlocutor. This is highly suggestive that what Mill has in mind is sincere
expression of belief. And so, if we’re obligated to point out to people what
is fallacious in their beliefs or reasoning, this is an obligation to voice sincere
disagreement. So, we might read a Millian background theory as generating
a couple of different kinds of obligations. We are required to voice our sin-
cere disagreement. We are required to attend to or allow for the expression
of sincere disagreement. And we’re required to allow for, and be receptive of
those who play devil’s advocate if they so choose. This is the weak reading
of the Millian background theory and the obligations it generates.
There are a couple of ways we might proceed given the differences between
the weak and strong readings of the Millian background theory. One would
be to try to decide between the two by looking for historical clues and sug-
gestions as to what Mill thought. Another might be to run (or model) both
the weak and strong readings to see what kind of epistemic consequences
arose, to see if one reading is preferable.17 A third would be to bolster a Mil-
lian background view with one of the other views that clearly does generate
an obligation to play devil’s advocate. As I say, these background theories
can be complementary. I leave the adjudication between these options for
future work.
The goal of the current work has been to make progress in understand-
ing our epistemic obligations to voice disagreement. The details of the con-
texts in which we’re so obligated vary—the different background theories
obligate voiced disagreement for different reasons. And, regarding devil’s
advocacy, it has become clear that for all of our background theories the
obligation to play devil’s advocate is distinct from the obligation to voice
our sincerely held disagreements. According to all of our background theo-
ries, the obligations arise in different contexts and have different benefits.
These different obligations, however, are both epistemic insofar as they stem
from our background epistemic commitments.
Notes
1 I am grateful to the participants in the Voicing Dissent workshop at the Univer-
sity of Connecticut and to the participants of the Harms and Wrongs In Epis-
temic Practice workshop at Sheffield University for helpful discussions of these
ideas. I’m also grateful to Dana Miranda and Graham Priest for their help think-
ing about public disagreement and devil’s advocacy.
2 Jennifer Lackey (this volume), Johnson, Casey R. (unpublished), Goldberg
(2016)
3 www.theonion.com/man-who-plays-devils-advocate-really-just-wants-to-be-a-
1819568992
4 https://en.wikipedia.org/wiki/Devil%27s_advocate
5 (Aikin & Clanton, 2010)
6 (Fricker, 2007)
7 (Coady, 2010; Fricker, 2007, 2013; Hawley, 2011; Maitra, 2010; Medina, 2012;
Pohlhaus, 2013)
8 (Aikin & Talisse, 2013)
9 (Sunstein, 2000, 2002)
10 (Elgin, 2011)
11 Elgin is clear that she means to be ecumenical about what counts as science. She
argues that pursuits like arts and auto mechanics make similar requirements on
those who pursue them (Elgin, 1991).
12 Perhaps by way of a text book.
13 (Gilbert, 1999, 2006; Gilbert & Priest, 2013)
14 (Mill, 1869) Book 2 section 6–7.
15 (Mill, 1885, p. 389)
16 (Mill, 1869) Book 2 Section 21.
17 (Baumgaertner, 2014)
References
Aikin, S. F., & Clanton, J. C. (2010). Developing group-deliberative virtues. Journal
of Applied Philosophy, 27(4), 409–424.
Aikin, S. F., & Talisse, R. B. (2013). Why we argue (and how we should): A guide to
political disagreement. London and New York: Routledge.
Baumgaertner, B. (2014). Yes, no, maybe so: A veritistic approach to echo chambers
using a trichotomous belief model. Synthese, 191(11), 2549–2569.
Coady, D. (2010). Two concepts of epistemic injustice. Episteme, 7(2), 101–113.
Elgin, C. Z. (1991). Understanding: Art and science. Midwest Studies in Philosophy,
16(1), 196–208.
Elgin, C. Z. (2011). Science, ethics and education. Theory and Research in Educa-
tion, 9(3), 251–263.
Fricker, M. (2013). Epistemic justice as a condition of political freedom? Synthese,
190(7), 1317–1332. doi:10.1007/s11229-012-0227-3
Gilbert, M. (1999). Obligation and joint commitment. Utilitas, 11(2), 143–163.
Gilbert, M. (2006). Rationality in collective action. Philosophy of the Social Sci-
ences, 36(1), 3–17.
Gilbert, M., & Priest, M. (2013). Conversation and collective belief. In A. Capone,
F. Lo Piparo, & M. Carapezza (Eds.), Perspectives on pragmatics and philosophy
(pp. 1–34). Cham: Springer International Publishing.
Goldberg, S. (2016). Arrogance silence and silencing. Proceedings of the Aristotelian
Society, 90, 93–112.
Hawley, K. (2011). Knowing how and epistemic injustice. Knowing How: Essays on
Knowledge, Mind, and Action, 283–299.
Maitra, I. (2010). The nature of epistemic injustice. Philosophical Books, 51(4),
195–211.
Mill, J. S. (1869). On liberty. London: Longmans, Green, Reader, and Dyer.
Mill, J. S. (1885). Principles of political economy. New York: D. Appleton and
Company.
Pohlhaus, G. (2013). Discerning the primary epistemic harm in cases of testimonial
injustice. Social Epistemology, 1–16.
Sunstein, C. R. (2000). Deliberative trouble? Why groups go to extremes. Yale Law
Journal, 71–119.
Sunstein, C. R. (2002). The law of group polarization. Journal of Political Philoso-
phy, 10(2), 175–195.
7 Eloquent Silences
Silence and Dissent
Alessandra Tanesini
Introduction
A political dissident is interrogated. She is asked time and again to reveal
the names of other activists. In response, she remains silent. A black man
narrates his experiences of racial discrimination to a white audience. No
one interrupts him. When he finishes, his audience is silent, still, showing
no signs of emotion. Hundreds of students sit silent on the ground, motion-
less forming a path from the entrance of a University building. Through
them moves a woman, looking nervous, who is being followed by report-
ers.1 These, and many more, are things we do with silence. Some of them are
ways in which dissent can be “voiced.”
It is prima facie plausible to think that these silences are effective commu-
nicative acts and to take the silences they involve to be examples of illocu-
tions such as protesting, refusing, or dissenting.2,3 Further, it is commonplace
in linguistics and in political theory to accept that some silences are propo-
sitionally meaningful (e.g., Tannen & Saville-Troike, 1995; Ferguson, 2003;
Ephratt, 2008). Yet, there is, to my knowledge, no philosophical discussion
of this phenomenon. Whilst it is widely acknowledged in the philosophical
literature that speech acts can be performed silently, the examples which are
considered involve body language or other physical gestures. There is no
discussion of the communicative potential of silence itself.
This lacuna is particularly troubling given the recent prominence of silence
as a tool of dissent. Until the 1990s silence has often been equated with
powerlessness or acquiescence. The Act Up slogan “Silence = death” was
indicative of this approach (cf., Ferguson, 2003). In recent years, however,
uses of silence as a means of resistance or protest have proliferated. For
instance since 1996 in April every year the Gay, Lesbian and Straight Edu-
cation Network (GLSEN) in the U.S. holds a day of silence in schools to
express solidarity with gay and lesbian students and to raise “awareness
about the silencing effect of anti-LGBT bullying, harassment and discrim-
ination” (GLSEN, 2017; cf., Woolley, 2012). Silent demonstrations have
been held in Iran in 2009 protesting against perceived widespread elec-
toral fraud (Ranjbar, 2017); in Ecuador against the mining of rainforests
110 Alessandra Tanesini
(Fitz-Henry, 2016); and in Mexico against the violence that pervades the
lives of indigenous people (Carmona, 2014). In Turkey, a performance art-
ist began in 2013 a silent protest against Prime Minister Erdogan. He soon
became known as the standing man and hundreds joined him in silent pro-
test (Seymour, 2013).
There are several likely causes for the rise in popularity of silence as a
way of dissenting. Prominent among them is increased awareness that we
live in a society where individuals are on a daily basis mandated to disclose
personal information to authorities and institutions. Citizens are thus often
confronted with officials, impersonal forms or websites, whose task is to
make individuals produce speech about themselves, and that succeed in elic-
iting such speech whilst bypassing, undermining, or overriding the agent’s
will. In other words, we live in societies in which citizens frequently have
their speech extracted from them (McKinney, 2016).4 In addition, social
media such as Facebook or Twitter nudge their users to make public per-
sonal information about the minutiae of their daily lives. Within this context
it is hardly surprising that choosing silence becomes a way of withdrawing
and dissenting.5
This chapter has three main goals. The first is to fill the lacuna in the exist-
ing philosophical literature by showing that silence can be an illocution which
is used to communicate. In other words, silence can be eloquent.6 The second
is to argue that silence when eloquent is not usually expressive of acceptance.
Instead, it often signals that something is awry with the ongoing conversa-
tional exchange. The third is to identify some features of eloquent silences
that explain their effectiveness, in some contexts, in expressing dissent.
Eloquent Silences
The aim of this section is to provide an account of eloquent silences which
I define as silences that (i) are illocutions and (ii) are intended to commu-
nicate. I also show that silences of this kind are commonplace. I begin this
section by explaining what I mean by silence in order to exclude sign and
body language. Subsequently, I consider different kinds of silence before
focusing on some which seem intuitively to be eloquent. Using both Gricean
and one non-Gricean account of illocutions, I show that there are some
silences that neatly fit these accounts. I note that interlocutors’ recognition
of the communicative intentions of those who keep silent often depend in
part on the fact that silences are often adjacent to other speech acts to which
they respond such as a question, a greeting, or a request. Finally, I argue that
whilst some eloquent silences adhere to Grice’s Cooperative Principle, oth-
ers flout, violate or opt out of it.7
Before I can contrast eloquent silences, with which I am concerned, with
other silences that do not communicate, I need to provide a characterization
of silence in general. For my purposes here, some behaviors which are not
vocal are nevertheless not silence because they constitute linguistic or verbal
Eloquent Silences: Silence and Dissent 111
behavior which deploys non-acoustic means of communication. These
behaviors include: writing, sign-language, and communication by means of
gestures. Even though they can be carried out silently, these communica-
tions are verbal (cf., Saville-Troike, 1995, p. 5); they are not silence. Silence,
then, is non-acoustic behavior which is also non-verbal.
Even if we restrict our focus to silence so characterized, there are silences
that are not communicative. Some of these silences are extraneous to conver-
sations; others facilitate conversation without communicating anything them-
selves. First, silences that occur outside communicative exchanges include
the silence of a person who is asleep, or of strangers sitting side by side in
the reading room of a library. They also include silences caused by external
forces that prevent one from speaking such as the silence that follows a loud
noise, but also the silence of someone who is locutionarily silenced and thus
rendered unable to speak.8 These are all silences which are outside commu-
nicative events. Of course, whilst some may be innocuous, others—such as
those resulting from locutionary silencing—are generally harmful.
Second, there are also silences which are part of conversations, and enable
communication, but still have themselves no communicative function. These
include pauses in speech to swallow or breathe, to find the right words or to
decide what to say. Another kind of silence, which enables communication
but is not itself part of it, is the silence of listeners that allows the speakers’
words to be heard. Ritualized silences that serve to intensify the significance
of speech in religious or spiritual ceremonies such as Quakers’ meetings may
also be thought to enable communication by framing or foregrounding it
(Saville-Troike, 1995).9
Third, there are silences which can be taken as evidence of the mental
state of the silent person without being themselves communicative acts.
These are exemplified, for instance, by long hesitations which may be inter-
preted as evidence of indecision, or of thoughtfulness; they may also be read
as indicating that what one is about to say is painful or otherwise hard to
express. These silences are sources of information. Generally, however, they
are not communicative acts.10
These silences are not communicative because there is nothing that the
speaker means in being silent. Thus, a forteriori, there is no meaning that
they intend to communicate. That is, whilst it seems possible to illocute
without intending to have an effect on any audience (Davis, 1992), one can-
not purport to communicate without this intention. Yet, plausibly enough,
in all cases considered so far, the silent individual does not even mean any-
thing by her silence.
We can defend this claim in different ways depending on our preferred
account of speaker’s meaning. Given Grice’s account, speaker’s meaning
requires that one intends to have an effect on one’s audience (Grice, 1957).
This intention seems absent in many of the examples discussed above. In
others, such as the silence of the audience which allows the words of the
speaker to be heard, the silent individual intends to have an effect on the
audience, but they do not necessarily intend that their interlocutors recog-
nize their intentions. The claim also stands if we adopt a different account
of illocution. For instance, Mitchell Green (2007) has argued that speaker’s
meaning requires that one intends to make publicly manifest one’s commit-
ment to a content and also to make this intention publicly manifest. This
self-reflexive intention is absent in the examples of silence under consider-
ation. The person who hesitates, and whom we may interpret as pained,
may not intend to manifest any commitment.11
In this chapter, I do not discuss the kinds of silence that I have briefly
described so far. Instead, I am interested in silences that, on the face of it,
are eloquent because they communicate something to interlocutors. Such
silences are usually instances of elicited illocutions or example of resisting
elicitation.12 Silences can enact many different illocutions. Prominent among
these are refusals to be drawn into some conversations and announcements
that one is opting out of an existing one.
Students remaining silent in the face of a teacher’s invitation to speak
are examples of refusals to be drawn into a conversation. Such silences can
be expressions of boredom, lack of engagement, or absent-mindedness;13
often, they are also attempts to undermine teachers’ authority by chal-
lenging their entitlement to demand students’ participation. These silences
are aptly described by Perry Gilmore as non-submissive subordinate sulks
(1995, pp. 148–150). Individuals, who stop contributing to a conversation
by pointedly keeping silent, announce their opting out. Such silences, espe-
cially when prolonged over time and directed at particular individuals, are
an effective mean of ostracizing people by giving them the silent treatment.
Eloquent silences can also be moves within ongoing conversations. They
can express agreement as in the example of a congregation’s silence at a
Christian wedding ceremony after the minister has proclaimed that if any-
one knows of any impediment to the marriage, he or she must immediately
say so. They can function as answers to questions; for instance, when a per-
son keeps silent after a friend’s query: “Are you still mad at me?”14 They can
be invitations as exemplified by short silences marking turn-taking in con-
versation when a speaker gives the floor to a hearer inviting him to speak.
But they can also be acts of resistance, refusal, or protest. These include: the
silence of silent protesters; that of the person who resists or deflects inter-
rogation; and also, at least in some cases, white silence in response to black’s
autobiographical narratives of discrimination.15
In addition, some silences that make one’s emotional state manifest are
also eloquent moves within a conversation. I noted above that long silent
hesitations need not be communicative. However, if one noticeably hesitates
before speaking and that hesitation is unusually long, then it is plausible that
one’s silence is communicative.16 For instance, a person may hesitate with
the intention that her silence makes it publicly discernible that she is still
thinking about the answer, and also with the intention that it is publicly dis-
cernible that her intention is to make manifest that she is still thinking about
the answer. This hesitation is an action which is overt because it results from
the self-referential intention to make the agent’s commitment to a content
manifest and to make this very intention also publicly observable (Green,
2007, p. 66). In being silent in this way a speaker means that she is still
thinking about what to think on the topic but also, plausibly, she intends to
communicate a request that the interlocutor gives her time to formulate her
contribution to the conversation.
In the remainder of this section, I argue first that eloquent silences are
genuine illocutions in their own right. Second, I indicate that if we want
to understand the content and illocutionary force of these silences, it is
often useful to view them as one element within an adjacency pair of speech
acts. Finally, I argue that whilst there are cooperative silences, silence is
also an extremely effective means to communicate that one is not willing to
cooperate.
A speech act or illocution is commonly defined as an action that can, but
does not need to be, performed by saying that this is what one is doing.
Hence, for example, one can promise by saying “I promise,” assert by say-
ing “I assert” or order by saying “I order.” These are all illocutions (Austin,
1976, p. 137; Green, 2007, p. 70). Since, however, we cannot persuade
someone by saying “I persuade,” persuading is not an illocution but a perlo-
cutionary act. It is generally accepted that illocutions can be actions that do
not involve the use of spoken, written, or sign-language. However, examples
of silent illocutions in the existing literature usually involve gestures or so-
called body language.17 There is to my knowledge no philosophical litera-
ture on being silent as an illocution. Yet, without doubt being silent can
be an intentional action. One can intend to be silent, to hesitate or to give
someone the silent treatment. So, one may legitimately ask whether some
intentional silences are illocutions.
The answer, it would seem, is affirmative no matter which account of illo-
cution one adopts among some prominent contenders.18 One such approach
is inspired by Grice’s account of speaker’s meaning. In Grice’s view a speaker
means something by her actions only if she intends (i) to produce an effect
in her audience, and (ii) that her intention (i) is recognized by her audience,
but also (iii) that the effect in the audience is produced (at least in part) as
a result of the audience’s recognition of intention (i) (Grice, 1957). Illocu-
tions would be those actions which count as a speaker meaning something.
For example, I can assert that p by performing an action A (e.g., uttering p)
intending to make my interlocutors believe that p, and form that belief as
a result, at least in part, of their recognition of my intention to make them
believe that p.
Some silences exemplify this structure. A speaker may remain quiet after
having made some utterances intending (i) that her interlocutor believes that
he is invited to speak, and (ii) that her interlocutor recognizes that (i) is
her intention; but also (iii) that her interlocutor comes to believe that he is
being invited to speak partly at least because he recognizes that this is her
intention. Her silence is an invitation which communicates that it is the
interlocutor’s turn to speak.
Similarly, a person may keep silent for a while after a question intend-
ing that his audience believes that he is still thinking about the answer, and
intending them to believe this as a result (at least in part) of their recogni-
tion of his intention. Silences of this sort are often also requests for more
time before adding one’s contribution to the conversation. They are generally
recognized and respected as such by interlocutors who typically wait for the
other person to speak. Further, interlocutors’ silences are also illocutions.
They would seem to be a way in which one can grant the initial request for
thinking time.19 It is thus possible to have a silent dialogue in which one
person asks for more time before speaking and the other person accedes to
their request. This stretch of conversation can be carried out without words
and, at least among conversational partners who are comfortable with each
other, without the need for gestures or body language as ways of expressing
one’s intentions.
A similar interpretation readily offers itself for silent challenges of
authority. The students who publicly display non-submissive subordinate
sulks choose to remain silent. In being silent students challenge the teach-
er’s authority and intend that challenge to be recognized by the teacher
and to produce in the teacher the belief that his authority is being chal-
lenged (partly at least) through this recognition. It is natural to think of
this case as one in which students mean something with their silence and
successfully communicate what they mean. This content can be captured
with expletives or described as saying: “I am not doing what you want me
to do.”
Other eloquent silences, however, do not appear to require reflexive com-
municative intentions. That is to say, in some cases the intention, that the
effect be produced in the audience partly at least as a result of the audi-
ence’s recognition that the speaker intended to produce it, may be absent.
Grice explicitly discusses an example in which this intention is missing. He
classifies Herod’s action of presenting Salome with the head of St. John the
Baptist on a charger as an instance of letting someone know that something
is the case but not as an example of telling something. Herod intends to
make Salome believe that St. John the Baptist is dead. He also intends her
to recognize that this is his intention. Presumably, she forms the belief that
St. John the Baptist is dead. However, her belief does not require that she
recognizes his intention to get her to believe this since she can exclusively
rely on her own eyes (Grice, 1957, p. 382). For this reason, Grice takes this
not to be an instance of speaker’s meaning.
Many silences that I have categorized as eloquent have this structure.
Consider the person who keeps silent after a friend has asked whether she
is still mad at him. This person intends (i) that her friend believes that she is
still mad at him; she also intends (ii) that he recognizes that (i) is her inten-
tion. But she need not intend (iii) that he believes that she is still mad at him
through his recognition of her intention. She is happy enough if his belief is
based on the observation of her behavior rather than the recognition of her
intention. Hence, if Grice is right, these silences too would not be examples
of illocutions.
I do not need to take a stance on this issue here, since as I have shown
that some silences do involve reflexive communicative intentions. Neverthe-
less, it seems plausible to hold that not all illocutions require that one also
intends to have an effect on an audience.20 It is possible for someone to
mean something by an utterance without thereby intending to communicate
this meaning to anyone, not even herself.21 If this is right, Grice’s would be
best interpreted as an account of communicative acts rather than of illocu-
tions. In addition, his claim that showing or letting someone know is not
communicative because it is not an instance of telling is also open to serious
counter-examples. Suppose that, in response to an invitation to play squash,
I show you a heavily bandaged leg. I clearly mean by that gesture that I can-
not play. I also clearly communicate this fact to you. Yet in this case too your
belief that I cannot play may be wholly based on an observation of the state
of my leg (Schiffer, 1972, p. 56).
These considerations have prompted some to abandon Grice’s approach
in favor of other accounts of illocutions that do not even require that the
speaker intends to produce an effect in her audience. For instance, Green
has argued that speaker meaning is a matter of overtly showing one’s com-
mitment to a content under a given illocutionary force (2007, p. 74). More
precisely, a person S means that p with illocutionary force ϕ if and only if
S performs an action A, intending to manifest that one’s commitment to p
under force ϕ and that this intention is also manifest (ibid.)
The examples of eloquent silence that I have used so far to illustrate how
silence can be an illocution nicely fit this account. They are all examples in
which someone keeps silent intending to make publicly discernible one’s
commitment to some content such as “I need more time,” “you are not in
charge,” or “I am still mad” under the force of a request, a challenge, or an
assertion.22 This intention is self-referential since it is not just the intention
to make their commitments manifest, it is also the intention to make pub-
licly discernible that that is one’s intention.
I shall not try to adjudicate here between these different accounts of illo-
cution.23 What matters for my purposes here is that independently of which
account is to be preferred, my discussion shows that some silences are illo-
cutions. It also shows that the same action—being silent—can constitute a
plethora of different illocutions with very different contents. It is, therefore,
hard to fathom how silent agents can make publicly manifest one among
many different possible commitments. It is equally mysterious how audi-
ences are able to recognize agents’ intentions in keeping silent. Yet, silent
communication regularly succeeds. The most plausible explanation of this
achievement is based on the fact that eloquent silences are often either elic-
ited or a way of resisting elicitation.
The thought that silent illocutions are produced either in order to fulfill
at least some the intentions of the speaker whose speech elicits them, or to
resist such elicitation, is compelling, once we consider that by themselves
all silences are the same. This simple fact can be easily overlooked because
when we think of the role of silence in conversation we often think of it in
conjunction with distinctive gestures which are separate illocutions. If, as
I argued above, silence itself can be an illocution, one can only account for
its varied nature by thinking of it as an element within a structured sequence
of illocutions.
It is beyond the scope of this chapter to assess the prospects of theories
of conversation. But even those who, like Searle (1992), think there is lit-
tle prospect for success in the search for rules or regularities that would
structure conversations admit to the existence of notable exceptions: adja-
cency pairs. These are pairs of illocutions in which the second element is a
response to the first which precedes it. Searle briefly discusses three group-
ings: questions and answers; requests that one performs a speech act and
their responses; offers, proposals, or bets and the consequent acceptance or
refusals (Searle, 1992, pp. 8–10). It is striking that the examples of eloquent
silences I discussed above, all seem to belong to one such grouping.
One way to think about the relations between illocutions that form adja-
cency pairs is to think of the second as being elicited by the first. Following
McKinney (2016), we can think of an illocution of a speaker A as being
elicited by a speaker B’s illocution when it is produced by A to fulfill at least
some communicative and perlocutionary intentions of B towards A. For
example, when a person A keeps silent when a friend B asks whether she is
still mad at him, A fulfills B’s communicative and perlocutionary intention
that she answer his question. A can exploit this structure of elicitation to
make publicly manifest her commitment to a specific content under a given
force, whilst B’s knowledge of his own intentions and of A’s apparent com-
mitment to fulfilling them allows him to recognize A’s intentions.
Eloquent silences are also often acts of resisting elicitation. They are
therefore refusals to fulfill some communicative or perlocutionary inten-
tions directed toward the person who keeps silent in response. In this case
also the agent can rely on a common understanding of the structure of elici-
tation to intend to make manifest her commitments. In this case the nature
of A’s commitment is specified by the refusal to fulfill at least some of B’s
intentions; B is able to recognize A’s intentions because he has knowledge
of his own intentions and of A’s apparent commitment not to fulfill them.
Eloquent silences can be cooperative or non-cooperative.24 Interestingly,
they sometimes function as a cooperative way of announcing the end of
cooperation. Grice’s Co-operative Principle states that one should make
one’s conversational contribution “such as is required, at the stage at which
it occurs, by the accepted purpose or direction of the talk exchange in
which [one is] . . . engaged” (Grice, 1989, p. 26). Eloquent silences that
are moves within ongoing conversations usually satisfy the principle.25 The
congregation’s silence is maximally informative, relevant, and unambigu-
ous. Silently requesting time to think also satisfies the principle. One could
also ask for more time by means of words to that effect, but doing so in
most cases slows down the conversation even further. It is also not necessary
unless one is proposing to return to the topic at a later date. Even the silence
of the person who is still mad at one is cooperative since it unambiguously
and effectively communicates one’s annoyance.
Other eloquent silences flout the cooperative principle. None is clearer
than the silence of the dissident and of the protester in the face of attempts
to extract information, promises, or other commitments out of them.26 In
these cases the silent objector patently attempts not to make her contri-
bution informative and she resists the speakers’ attempts to place some
obligations on her. Non-subordinate submissive silences are examples of
intermediate cases since they are informative, relevant, and unambiguous
negative responses to the request or command made by a person in author-
ity that one performs an illocution. It does not seem implausible to identify
them as a way of marking cooperatively the end of one’s cooperation. These
are examples of individuals who respond to attempts to elicit their speech by
frustrating some of the perlocutionary intentions of those who intended to
make them speak. However, their silence is cooperative since it is relevant,
informative, and unambiguous.
Silence as assent and as acceptance

Recently, Philip Pettit (2002) and Sanford C. Goldberg (2016, 2018; MS)
have argued that, barring the presence of defeaters, speakers are entitled to
presume that silent interlocutors accept what speakers have said.27, 28 In this
section, I take issue with this view. Instead, I argue that even if we grant that
audiences have a defeasible obligation to make their dissent manifest, it does
not follow that such disagreement cannot be communicated by means of
silence. In particular, I show that at least when silence is eloquent, it cannot
be presumed by default to communicate assent.29
It may be instructive to consider first examples where silence is eloquent
and communicates assent. Goldberg mentions the so-called “tacit accep-
tance procedure” used by some committees when ratifying reports. Given
this procedure, documents are tacitly accepted unless someone explicitly
raises an objection before a pre-set deadline (MS, p. 7). Silence in these
cases may not be wholly freely chosen, but it communicates assent, as does
the silence of the congregation when it has been directed by the minister to
speak out against the marriage or forever acquiesce to its legitimacy. We
should not, however, draw any general conclusion based on these examples
alone since they essentially rely on speech acts known as directives.
The priest during the wedding ceremony instructs the audience either to
speak up or accept the marriage. The minister’s illocution is thus a directive
which is a speech act that tries to get the hearers to do something (Searle,
1976, p. 11) or that lays some obligation on them (see Alston, 2000,
pp. 97–98). It is, thus, plausible to conclude that the reason why silence in
this instance communicates assent is that members of congregation have
acquired a special responsibility to state presently any objections they may
harbor. The same considerations apply to tacit acceptance procedures. These
silences are forms of assent because they are responses to specific directives.
These examples of eloquent silences provide no evidence for the view that
silence following an assertion also communicates acceptance.
In addition, there are situations where silence following an assertion
overtly communicates disapproval.30 One is discussed by Grice in the con-
text of failures to fulfill the submaxim relating to making one’s contribu-
tion relevant to the conversation. Grice considers the case of a person
who, at a tea party, states that “Mrs. X is an old bag.” The claim is
followed by silence which Grice describes as “appalled.” Then one inter-
locutor continues the conversation by changing the topic to a discussion
of the weather (1989, p. 35). Grice focuses on the change of subject, but
the silence that precedes it is also a conversational move. The interlocu-
tor’s silence communicates her consternation about the speaker’s claim.
It conveys, by showing that one refuses to engage with it, that the speak-
er’s comment should not be dignified with a response (Ephratt, 2012,
pp. 65–66).
Further, it is not altogether clear that Grice is right to maintain that the
appalled interlocutor violates the Co-operative Principle by not making
her contribution relevant. Arguably, one may claim instead that the initial
remark is out of order. Rude comments are unacceptable given the pur-
poses of polite conversation at tea parties. Thus, it is the original speaker
who is uncooperative, whilst the interlocutor’s silence is cooperatively help-
ful by trying to steer the conversation back to its initially agreed purpose
(ibid., p. 66). Hence, at times, pointed silence is the clearest, most coherent,
and helpful way, of communicating to a speaker that some remark of his is
beyond the pale.
This example brings into relief some unwarranted assumptions in the
arguments in favor of the view that the default, albeit defeasible, interpreta-
tion of silence following an assertion is acceptance. First, these arguments
fail to distinguish clearly deliberate silence from a failure to remark on a
specific claim. Second, they appear to presume that disagreement can only
be conveyed by saying something. That is, they ignore the possibility that
silence may communicate dissent. Third, they overlook the possibility that
the initial speaker’s contribution may itself be uncooperative.
To substantiate these charges, I need to give a flavor of the position I wish
to oppose. According to this view, it is a feature of the practice of assertion
that any claim made in a conversation, which is not challenged, is accepted
by default; it, thus, becomes part of the background presuppositions of
that linguistic exchange. Therefore, it is incumbent on the interlocutors
to block the addition of a new assertion to the common ground or to the
conversational score of the conversation, if they object to it.31 If they fail to
act, the assertion is accepted.
Although I have reservations about it (Tanesini, 2016), I do not wish in
this chapter to challenge the thesis that assertions are accepted by default.
Instead, I want to challenge the claim that silence cannot be a rejection of
an assertion. Defenders of the view that silence indicates acceptance do not
address this issue head on. Instead, they ignore this possibility because its
existence is obscured by two unargued but related presumptions. One is
the assumption that silence may indicate something but says nothing. This
neglect of eloquent silences makes it difficult to recognize the difference
between keeping silent and not remarking upon a claim. Yet, whilst keeping
silent is an act, not remarking upon a claim is best thought of as an omis-
sion. This second assumption that silence is always an omission rather than
an act generates the tendency to conflate what are clearly distinct phenom-
ena: one is omitting to say anything about a claim, often by saying some-
thing which continues the conversation; the other is keeping silent when one
is invited to speak or is at least in a position to do so.
The interlocutor in the first case, by indicating his willingness to continue
the conversation as normal, indicates his assent. The person who is delib-
erately keeping silent, instead, indicates that something is amiss with the
conversation which, therefore, cannot continue as normal. It may be worth
noting in this regard how awkward silence often is in conversation. Silence
is uncomfortable because it often marks the fact that things are not going
well with the conversational exchange.32 Hence, it does not seem plausi-
ble that deliberate silence marks agreement. What may mark agreement is
behavior that indicates that all is well with the conversation. Such behavior
includes contributing to it without remarking upon, or drawing attention
to, previous contributions in critical or negative ways.
In addition, those who take silence defeasibly to indicate assent overlook
how frequently silence is used to express overtly, and make public, one’s dis-
sent. One of the reasons why the prevalence of dissenting silence might have
been overlooked is the focus on disagreement understood as having doubts
about the truth of a claim, or believing it to be false. But disagreement may
take different forms. One may object to an assertion because of the discrimi-
natory vocabulary it deploys. One’s censure of slurs or hate speech is not
primarily based on their falsity but on the harms they cause (and perhaps
constitute). Thus, when confronted with such speech one may want to con-
test the legitimacy of uttering the words, rather than challenge the veracity
of the assertion. That is, one may wish to reject that the oppressive speech
contributes to the conversation, rather than to respond to it as a legitimate
move which is epistemically defective.
In this context, there is a risk that trying to offer reasons for the unac-
ceptability of hate speech actually ends up giving it some legitimacy. If one
treats speech of this sort as requiring a reply rather than as deserving to
be ignored or dismissed, one can be plausibly taken to imply that it has
made a contribution to the conversation. Plausibly, however, when one finds
an assertion to be repugnant rather than merely false, pointed silence, that
conveys the thought that the speaker is out of order is an effective way of
distancing oneself from the speaker and his or her speech.
In sum, I have shown that one may be led to think that silence by default
signals acceptance only if one overlooks some plausible features of the use
of silence in communication. First, silence can be an illocution. Second,
silence as an illocution often, or even usually, marks that something is going
wrong rather than well in the conversation.33 Third, cooperation in conver-
sation is a two-way street. Sometimes when a speaker is not cooperative,
a helpful interlocutor will draw the speaker’s attention to the fact. Correc-
tions of this sort are akin to censure or disapproval rather than to critical
challenges. Silence can be an effective way to censure a speaker without
being uncooperative.
Dissident Silences
I have argued in the first section that silence can be eloquent. In the second
section I have shown that it is a mistake to think that silence defeasibly
indicates acceptance. In this section, I present a non-exhaustive and par-
tially overlapping taxonomy of eloquent silences as ways of expressing dis-
sent. Subsequently, I discuss two features of dissident silence that make it
a particularly effective tool in resisting discrimination and subordination.
First, silence is more effective than verbal criticism at expressing censure of
oppressive claims. Second, since silence often shows what it communicates,
its success as an illocution does not require that its target audience trusts
or is especially attentive to the intentions of silent individuals. Whilst the
speech of the dispossessed may be not be listened to, their silence is, at times,
harder to ignore.
Political theorists and linguists have offered different ways of categoriz-
ing silence (cf., Ephratt, 2008; Jungkunz, 2013), each suited for different
purposes. Here I propose a different taxonomy based on the idea that the
nature of silence is best understood in relation to those illocutions or stand-
ing conditions whose elicitation it resists. My taxonomy does not aim to be
exhaustive or to identify mutually exclusive possibilities. Instead, my goal
is to illustrate the variety of illocutions that can be performed by means of
silence.
The first family of eloquent silences concerns those silences in the face of
directives that demand that one speaks. These silences are refusals to com-
ply; they are ways of resisting the elicitation or even extraction of speech.34
They can constitute many illocutions such as defiance, refusal, resistance,
protest, or withdrawal. Within this category one may include students’
silence as a challenge to the authority of teachers by means of subordinate
non-submissive sulks (Gilmore, 1995); the resistance of the political activist
when remaining silent during an interrogation; the refusal of some athletes
to sing the national anthem in silent protest (Jungkunz, 2012, p. 140). It
also comprises many tactics adopted by welfare recipients to resist state
surveillance through failure to report some income, feigning ignorance,
and foot dragging (Jungkunz, 2012, p. 142).35 This latter kind of resistant
silence in the face of attempts to extract information has often been used
by the subordinated to make themselves unknown to people in positions of
power (Collins, 1991, p. 92).36
The second family includes those silences which are intentionally kept in
contexts in which speech is invited or expected within some ongoing con-
versation or debate. These silences are as varied as the speech that attempts
to elicit them. They include silence in response to questions or the appalled
silence of the person who thinks that what has just been said is out of order.
To this family also belong the silences of silent protesters who wish to com-
municate their refusal to engage with institutional explanations of some event
(Carmona, 2014), and of those who wish to draw attention to those voices
that are missing from the conversation in order to empower them (Jungkunz,
2012; Ranjbar, 2017, p. 618). Often these silences derive their significance
from the fact that they confound expectations and are a source of surprise.
This second family also includes silences in response to queries that are
intended to show that one should not be expected to be able to answer that
question. This is a strategy of survival that has sometimes been adopted by
subordinated individuals. It is, as I mentioned above, a way of making one-
self unknown to those who can cause one harm. In this instance, one may
keep silent to give the impression that one is slow-witted and has nothing
to say. Such silence may be a result of the psychological damage inflicted by
habitual silencing or it may be a conscious strategy to appear more deferen-
tial and less knowing than one actually is.37
Privileged individuals adopt a similar strategy to indicate that they regard
their ignorance of some issue to be unproblematic. Sometimes, the silence
of white people in the context of discussions of race belongs to this cate-
gory. White silence, in these instances, is not indicative of attentive listening.
Rather, it communicates that one thinks that one should not be expected
to know anything about the topic, and, therefore, to be able to make a
meaningful contribution. It thus implicates that white people are blame-
lessly ignorant about race (DiAngelo, 2012; Applebaum, 2016).
The third, and final, family of silences I wish to discuss concerns silence
on occasions where there is a standing and generic expectation that speech
will be present, although these have not been preceded by specific solicita-
tions that one speaks. These silences too can be a form of protest which
works by creating surprise, or discomfort, or by drawing attention to some
of the features of the usual speech that go unnoticed until it is replaced
with silence. Hence, for instance, silent protests in Ecuador were intended
to communicate, by being patently different from it, that dominant public
speech was toxic because of its reliance on strong men and alleged saviors
(Fitz-Henry, 2016, p. 11). Performances of stillness and quietness in public
spaces where bustling noise is generally expected also communicate similar
messages.38 The use of silence by teachers overtly to communicate to stu-
dents the acceptability of patterns of speech that include long silent pauses
is another example of silence which acquires its eloquence through defying
standing expectations.39
I hasten to add that the power of silence as a tool of dissent is not without
its limitations. It is not a coincidence that it is primarily used by those who
are powerless and as a weapon of last resort. It is often a strategy to deflect
violent reprisals by making it apparent that one’s protest is peaceful, and to
unify groups of demonstrators who may not share a common view of the
alternative to the status quo (cf., Ranjbar, 2017).
Nevertheless, silence can be a successful form of dissent. In what follows
I focus on two of its features that promote its effectiveness in the expression
of disapproval and disagreement. The first contributes to explaining why
silence can be used to communicate censure and disdain. The second may
be one of the reasons why silence is often chosen by those who are relatively
powerless in a given situation.
In the second section of this chapter I have noted that even if we grant that
audiences have a defeasible obligation to make manifest their dissent with
speakers, it does not follow that such disagreements must be verbally com-
municated. Here, I want briefly to press the point further by arguing that
silence on some occasions can be more powerful than speech in expressing
dissent. It has been noted that oppressive speech is often hard to reverse;
when appalling claims have been made salient, it is nearly impossible to
restore the conversation to the point before their utterance (McGowan,
2009; Simpson, 2013). Those who have lamented this situation have often
focused on verbal objections as a way of reversing the accommodation of
moves that are oppressive.
There are many factors that make it hard to cancel out the pernicious
influence on conversations of speech that subordinates. In some contexts,
as I have argued in the second section earlier, silence may be more effective
than rational dissent. The person who verbally objects to oppressive speech
because she engages in rational debate, indicates by her actions that the
objectionable speech at least deserves to be treated as a contribution to the
discussion. The person who, instead, keeps silent in response can succeed
in conveying that what has been uttered is beyond the pale. She may thus
force the initial speaker to take back what he said. Or, failing that, she may
compel him to ignore studiously the claim he just made, if he wants the con-
versation to continue. Silence is thus a way of censuring those whose speech
is perceived as being unacceptable. This kind of silent treatment can be used
to silence those whose views are at odds with those of the majority. How-
ever, as I indicated above, it can also work as a powerful way of canceling
out some of the effects of oppressive speech.
Finally, eloquent silences, I have argued, often communicate by show-
ing what one means by them. They communicate because they involve
the intention to have an effect on an audience as well as the intention to
make the agent’s commitment to a content manifest and to make this very
intention also publicly observable. Further, these silences are instances of
showing because they enact the contents that they communicate. So, the
person, who is silent meaning that she is still mad, shows her annoyance.
Her interlocutor can see that she is mad at him, as well as being aware
of her intention to make it manifest that she intends to make it publicly
accessible that she is still mad at him. Similarly, the person who keeps
silent to highlight the fact that he has been silenced shows what he means.
He intends to make public his protest that he has not been allowed to
speak, and to also make it manifest that this is his intention. He does this
by displaying the very silence to which he has been confined as an object
for others to attend to. Metaphorically speaking, he displays his silence on
the charger.
Showing is in some circumstances the most effective tool because, unlike
telling, it does not require that the target audience trusts the speaker. Ordi-
nary acts of telling require that the receiver of the testimony takes the
speaker to be both competent and sincere. However, members of subordi-
nated groups are often, because of prejudice, not credited with the amount
of credibility to which they are entitled (Fricker, 2007). In addition, members
of subordinated groups also often suffer from what Mary Kate McGowan
(2014) has identified as sincerity silencing. This occurs when speakers are
not recognized as being sincere. Women’s refusals of sexual advances are
often interpreted in this way. The person who makes the advance often
understands the woman’s “no” as a kind of refusal. But he does not think of
her as being serious. Instead, he may presume that she is coy or playing the
“hard to get” game. In so far as showing does not require that communica-
tive success is even in part based on a recognition of the speaker’s intention
or on an assessment of her sincerity, it is a possible way to bypass some of
the negative effects of prejudice that prevent effective communication.
In sum, whilst speaking out is in many contexts the most effective way of
expressing one’s dissent from, or disagreement with, prevailing views, there are
other circumstances in which silence is an effective way of protesting, refusing
or dissenting. In these instances, silence can speak louder than words.40
Notes
1 This protest was organized in 2011 at UC Davis by students objecting to the deci-
sions made by the University Chancellor to call the campus police in response to
previous protests (cf., Hatzisavvidou, 2015).
2 My focus is on literal silence. Some of the sociological literature on silence is
concerned with elephants in the room. These are taboo topics about which peo-
ple are silent whilst being extremely loquacious (cf., Zerubavel, 2006). I shall
ignore these cases.
3 Since silence does not use language, it is awkward to refer to some silences as
speech acts. In what follows I avoid this tension by using “illocutions” rather
than “speech acts” to refer to units of conversation. The two terms are used
interchangeably by others (e.g., Green, 2015, p. 6).
4 Some forms of speech extractions are unjust, whilst others are unproblematic.
Speech which is elicited by coercion or manipulation is paradigmatic of the first
kind.
5 Silence also stands in contrast to the noise characteristic of contemporary lives
in which we are bombarded with advertising, music, the web, television, and the
printed press.
6 I owe the expression to Michal Ephratt (2008).
7 Grice distinguishes four ways of failing to fulfill a conversational maxim and
thus to adhere to the Cooperative Principle. One may quietly violate the maxim
and thus mislead one’s interlocutors. One may opt out and indicate that one is
unwilling to cooperate. There may be a clash between different maxims so that
they cannot all be fulfilled. Finally, one may make a public display of failing to
fulfill a maxim and so flout it. This latter situation often generates implicatures
(Grice, 1989, p. 30).
8 A person is locutionarily silenced if she is literally prevented from carrying out
the locutionary act of uttering words; for example, if she is gagged (Langton,
1993).
9 Alternatively, one may think of these silences as being part of the communica-
tive exchanges. Nothing much hangs on this in my chapter as I shall not discuss
ceremonial or ritualized silences.
10 There may be cases when a person does communicate through a sustained hesi-
tation. My point here is that at least some hesitations are not instances of com-
munication, although they provide interlocutors with evidence of a person’s
mental state.
11 There are other instances where hesitations may be eloquent because the silent
person intends her silence to show her pain and to make that intention itself
manifest to her interlocutors.
12 Elicited speech is speech which one utters to fulfill the communicative and per-
locutionary intentions of another speaker which are directed towards oneself.
A communicative intention is an intention to change a person’s beliefs (partly
at least) through his recognition that this is one’s intention. A perlocutionary
intention is an intention to make another person do something partly through
his recognition of one’s intentions. See McKinney (2016, pp. 267–271) for this
notion of elicitation.
13 Sometimes being able to show disengagement without penalty is one of the privi-
leges of so-called white silence (DiAngelo, 2012, p. 7).
14 I borrow the example from Saville-Troike (1995, p. 9)
15 On white silence, see (DiAngelo, 2012; Applebaum, 2016).
16 What counts as unusually long varies from speech community to speech com-
munity (Sacks, Schegloff, & Jefferson, 1974).
17 For example, J. L. Austin noted that we can warn or protest by non-verbal means
and mentioned hurling a tomato as a way of protesting (Austin, 1976, p. 119).
18 There are many more which I am not considering here but none seem to preclude
silence being an illocution.
19 These requests can be misunderstood. However, interlocutors seem to assume
that a speaker who is silent is cooperative. Thus, they default to treating the
silence as a request which they cooperatively accept. If the silence continues for
too long, interlocutors may explore other possibilities such as that the speaker
may not have heard or may be refusing to continue the conversation.
0 Davis (1992) is often taken to have established this point conclusively.
2
21 Widely discussed examples are those of people speaking in the absence of an
audience or of the person proclaiming his innocence whilst knowing he will not
be believed (Green, 2007, pp. 59–63).
22 The nature of these contents is somewhat indeterminate.
23 However, I return to the idea that illocutions can show their contents in the third
section. Eloquent silences often have this feature. It is in my view key to their
effectiveness as acts of dissent.
24 Cooperation and the fulfillment of another’s communicative and perlocutionary
intentions are related but are not the same. A person may be cooperative and yet
frustrate another’s perlocutionary intentions.
25 Ephratt (2012) offers several examples of silences that satisfy the principle. He
is, however, employing a very capacious notion of silence including changing the
topic of conversation.
26 That said, one could argue that the principle should not apply to these cases
since these are not talk exchanges with agreed purpose or direction.
27 There is a growing literature on silencing following Langton’s now classic (1993).
The issue of silencing is, however, orthogonal to my concern here since it high-
lights how some people’s contributions to debate are blocked or undermined.
28 The notion of acceptance here is weaker than that of belief. The interlocutor
may accept a claim for the purpose of the ongoing conversation without believ-
ing it.
29 I shall not consider here the broader question whether other kinds of silences
which are not themselves communicative can be presumed to indicate assent.
30 Overtness requires a self-referential intention to make a commitment to a con-
tent manifest and to also make this very intention manifest.
31 Those who discuss this feature of the practice of assertion usually develop it
either in terms of Stalnaker’s notion of a common ground (1999, 2002) or David
Lewis’s notion of a conversational score (1979). These are different ways of
explaining kinematically changes in what is taken to be accepted or permitted in
an evolving conversation.
32 This is not always so, however. The ability to maintain silence without discom-
fort is often a sign of intimacy.
33 I say “usually” because silence in response to a directive may communicate assent.
34 Speech is extracted when the elicitation undermines, bypasses, or overrides the
agent’s will (McKinney, 2016).
35 These silences often are violations of the Co-operative Principle.
36 However, some of these self-chosen silences are forms of self-censure for the pur-
pose of self-preservation. Unlike cases of testimonial smothering when a speaker
curtails her testimony because of her knowledge that the audience is not capable
of genuinely hearing what she says (Dotson, 2011), in these examples a speaker
fears that her audience will use her words to harm her. Medina has argued that
these silences should be thought as a kind of hermeneutical injustice because the
victim, even though she has the expressive resources to communicate her experi-
ences, is coerced into silence (Medina, 2013, pp. 101–104).
37 Collins mentions some of these silences in her discussion of Afro-American
women who worked as house servants (1991, p. 53 and pp. 142–143).
38 Here belongs the silent reading of banned books in public spaces that took place
in Thailand in 2014 (Fitz-Henry, 2016, p. 2) as well as the standing man’s pro-
tests in Turkey.
39 See Montoya (2000) on the deliberate use of silence to give legitimacy to the
speech habits of some minority communities which value silence rather than the
quick fire responses adopted by other ethnic groups.
40 I would like to thank Casey Johnson for organizing a wonderful workshop on
the topic of this volume. I am also grateful to the Humanities Institute at the
University of Connecticut for hosting the event and for making my participation
possible. Finally, I am indebted to all workshop participants for their helpful
comments, insightful presentations, and joyous company.
References
Alston, W. P. (2000). Illocutionary acts and sentence meaning. Ithaca, NY: Cornell.
Applebaum, B. (2016). “Listening Silence” and its discursive effects. Educational
Theory, 66, 389–404. doi:10.1111/edth.12172.
Austin, J. L. (1976). How to do things with words. In J. O. Urmson & M. Sbisà
(Eds.), William James lectures (2nd ed.). Oxford and New York: Oxford Univer-
sity Press.
Carmona, A. (2014). A silence that speaks: Ayotzinapa and the politics of listening—
toward freedom. Toward Freedom. Retrieved November, 2014, from https://
towardfreedom.com/archives/activism/a-silence-that-speaks-ayotzinapa-and-the-
politics-of-listening/.
Collins, P. H. (1991). Black feminist thought: Knowledge, consciousness, and the
politics of empowerment. Perspectives on gender (2nd ed.). New York and Lon-
don: Routledge.
Davis, W. (1992). Speaker meaning. Linguistics and Philosophy. Kluwer Academic
Publishers, 15, 223–253. doi:10.1007/BF00627678.
Day of Silence (2018, April 27). GLSEN. Retrieved August 7, 2017, from www.
glsen.org/participate/programs/day-of-silence.
DiAngelo, R. (2012). Nothing to add: A Challenge to White Silence in Racial Dis-
cussions. Understanding & Dismantling Privilege, 2, 1–17.
Dotson, K. (2011). Tracking epistemic violence, tracking practices of silencing. Hypatia.
Blackwell Publishing Inc., 26, 236–257. doi:10.1111/j.1527-2001.2011.01177.x.
Ephratt, M. (2008). The functions of silence. Journal of Pragmatics, 40, 1909–1938.
doi:10.1016/j.pragma.2008.03.009.
Ephratt, M. (2012). “We try harder”—Silence and Grice’s cooperative principle,
maxims and implicatures. Language & Communication, 32, 62–79. doi:10.1016/j.
langcom.2011.09.001.
Ferguson, K. (2003). Silence: A politics. Contemporary Political Theory, 2, 49–65.
doi:10.1057/palgrave.cpt.9300054.
Fitz-Henry, E. (2016). The limits of the carnivalesque: Re-thinking silence as a mode
of social protest. Liminalities: A Journal of Performance Studies, 13, 1–16.
Fricker, M. (2007). Epistemic injustice. Oxford: Clarendon.
Gilmore, P. (1995). Silence and sulking: Emotional displays in the classroom. In
D. Tannen & M. Saville-Troike (Eds.), Perspectives on silence (2nd ed., pp. 139–
162). Norwood, NJ: Ablex Publishing Corporation.
Goldberg, S. C. (2016). Arrogance silence and silencing. Proceedings of the Aristo-
telian Society, 90, 93–112.
Goldberg, S. C. (2018). Dissent: Ethics and epistemology. In C. R. Johnson (Ed.),
Voicing dissent. London: Routledge.
Goldberg, S. C. Assertion, silence, and the norms of public reaction. Unpublished
manuscript.
Green, M. (2007). Self-expression. Oxford: Clarendon Press.
Green, M. (2015). Speech acts. In E. N. Zalta (Ed.), Stanford encyclopedia of phi-
losophy. Stanford: Metaphysics Research Lab, Stanford University
Grice, H. P. (1957). Meaning. The Philosophical Review. Duke University Press on
behalf of Philosophical Review, 66, 377–388. doi:10.2307/2182440.
Grice, P. (1989). Studies in the way of words. Cambridge, MA and London: Harvard
University Press. doi:10.1017/S0031819100064330.
Hatzisavvidou, S. (2015). Disturbing binaries in political thought: Silence as politi-
cal activism. Social Movement Studies. Routledge, 14, 509–522. doi:10.1080/147
42837.2015.1043989.
Jungkunz, V. (2012). The promise of democratic silences. New Political Science, 34,
127–150. doi:10.1080/07393148.2012.676393.
Jungkunz, V. (2013). Deliberate silences. Journal of Public Deliberation, 9, 1–32.
Affairs. Blackwell Publishing, 22, 293–330.
Lewis, D. (1979). Scorekeeping in a language game. Journal of Philosophical Logic,
8, 339–359. doi:10.1007/bf00258436.
McGowan, M. K. (2009). Oppressive speech. Australasian Journal of Philosophy.
Routledge, 87, 389–407. doi:10.1080/00048400802370334.
McGowan, M. K. (2014). Sincerity silencing. Hypatia, 29, 458–473. doi:10.1111/
hypa.12034.
McKinney, R. A. (2016). Extracted speech. Social Theory & Practice, 42, 258–284.
doi:10.5840/soctheorpract201642215.
epistemic injustice, and resistant imaginations. Oxford and New York: Oxford
University Press.
Montoya, M. E. (2000). Silence and silencing: Their centripetal and centrifugal
forces in legal communication, pedagogy and discourse. Univesity of Michigan
Journal of Law Reform, 33, 263–327.
Pettit, P. (2002). Enfranchising silence. In Rules, reasons, norms (pp. 367–378).
Oxford: Clarendon Press.
Ranjbar, M. A. (2017). Silence, silencing, and (in)visibility: The geopolitics of Teh-
ran’s silent protests. Hypatia, 32, 609–626. doi:10.1111/hypa.12343.
Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the
organization of turn-taking for conversation. Language. Linguistic Society of
America, 50, 696–735. doi:10.2307/412243.
Saville-Troike, M. (1995). The place of silence in an integrated theory of communi-
cation. In D. Tannen & M. Saville-Troike (Eds.), Perspectives on silence (2nd ed.,
pp. 3–18). Norwood, NJ: Ablex Publishing Corporation.
Schiffer, S. (1972). Meaning. Oxford: Clarendon Press.
Searle, J. R. (1976). A classification of illocutionary acts. Language in Society, 5, 1.
doi:10.1017/S0047404500006837.
Searle, J. R. (1992). Conversation. In H. Parret & J. Verschueren (Eds.), (On) Searle
on conversation (pp. 7–29). Amsterdam and Philadelphia, PA: John Benjamins
Publishing Company.
Seymour, R. (2013, June 18). Turkey’s “standing man” shows how passive resistance
can shake a state | Richard Seymour | Opinion | The Guardian. The Guardian.
Simpson, R. M. (2013). Un-ringing the bell: Mcgowan on oppressive speech and the
asymmetric pliability of conversations. Australasian Journal of Philosophy, 91,
555–575. doi:10.1080/00048402.2012.704053.
Stalnaker, R. (1999). Context and content: Essays on intentionality in speech and
thought. Oxford: Oxford University Press.
Stalnaker, R. (2002). Common ground. Linguistics and Philosophy, 25, 701–721.
doi:10.1023/A:1020867916902.
Tanesini, A. (2016). I—“Calm Down, Dear”: Intellectual arrogance, silencing and
ignorance. Aristotelian Society Supplementary Volume, 90, 71–92. doi:10.1093/
arisup/akw011.
Tannen, D., & Saville-Troike, M. (1995). Perspectives on silence (2nd ed.). Nor-
wood, NJ: Ablex Publishing Corporation.
Woolley, S. W. (2012). “The Silence Itself Is Enough of a Statement”: The day of
silence and LGBTQ awareness raising. Anthropology & Education Quarterly, 43,
271–288. doi:10.1111/j.1548-1492.2012.01180.x.
Zerubavel, E. (2006). The elephant in the room: Silence and denial in everyday life.
8 Epistemic Arrogance and the
Value of Political Dissent
Michael Patrick Lynch
To realize the epistemic powers of democracy, citizens must follow norms

that welcome or at least tolerate diversity and dissent, that recognize the
equality of participants in discussion by giving all a respectful hearing,
regardless of their social status, and that institute deliberation and reason-
giving, rather than threats and insults, as the basis of their communication
with one another.
—E. Anderson, “The Epistemology of Democracy”
Introduction
Dissent takes many forms: from marching down Pennsylvania Ave., to
refusing to stand for the national anthem at a football game. But it can also
take explicitly discursive forms, such as when we publicly voice our political
disagreements in speeches, rallies, and on social media platforms.
It seems clear that voiced explicit disagreement—online and off—has
political value; in democratic politics at least, its existence is typically taken
as a (defeasible) sign of a healthy republic. But over and above its purely
political value, disagreement can also increase our knowledge and bring to
light truths of political importance. It has epistemic value, in short.
In this essay, I examine four different reasons for thinking that political
dissent has epistemic value. The realization of this epistemic value hinges in
part on what I’ll loosely call the epistemic environment, or the environment
in which individuals come to believe, reason, inquire, and debate. In par-
ticular, to the degree that our social practices encourage and even embody
an attitude of epistemic arrogance, the epistemic value of dissent will be
difficult to realize. Ironically, it is precisely then that dissent is most often
needed.
Preliminary Definitions
“Dissent,” like all interesting and important philosophical words, is slippery.
For purposes of this chapter, I’ll understand it as publically expressed criticism
or disagreement directed at the (perceived) consensus view of a particular
130 Michael Patrick Lynch
group, or a view under consideration for adoption by decision-makers, for
the purposes of influencing belief and action relevant to that view.
Such a characterization is obviously meant to be inclusive, but it serves
to distinguish dissent from other kinds of criticism we may wield in public
discourse. Dissent is aimed, to put it bluntly, at those at least perceived to be
powerful, whether that power results from a consensus view or brute politi-
cal authority. Moreover, it is done with a particular intention. Criticism can
be voiced for a variety of reasons—to intimidate, to express one’s hostility
to others or solidarity with kindred spirits. Dissent, while it might do any
or all of these things, is meant to dissuade, persuade, or otherwise cause
a change in a particular policy. And finally, it is publically voiced. Critical
thought alone is not dissent in the sense intended here. Dissent is voiced or
otherwise expressed in a manner that is at least meant for the consumption
of others, whether or not it actually reaches their minds or hearts.
Nonetheless this definition leaves room, as is fitting, for a plethora of other
distinctions. For one thing, the publicity condition is a matter of degree—
my dissent to a departmental policy might only be voiced to a smattering of
my colleagues or it might be shouted from the pages of a local newspaper.
And dissent in the above sense could be either non-discursive (the upraised
fist) or it can be explicitly discursive (as in a U.S. State Department “Memo
of Dissent”). And it can happen in different realms. A scientist can dissent
from the consensus view on a particular topic, a historian from her profes-
sional association’s views, and a citizen can dissent from his government
policies or laws.
My focus in what follows will be on a particular kind of dissent: discur-
sive political dissent (or DPD where we need an acronym), that is, argumen-
tative dissent directed at influencing policy and policy beliefs. By a policy
belief, I mean beliefs about which policies are in a public’s best interest. And
by argumentative or discursive, I mean dissent that expresses an argument,
whether elliptically—as on a protest sign—or directly, by way of a speech,
or pamphlet, or op-ed. We’ll be interested in both particular acts of DPD as
those just mentioned and in various institutional social practices that sup-
port, directly or indirectly, such acts, including those found in journalism,
political parties, publishing, standards of academic freedom, and the law.
This is a familiar kind of dissent in liberal democracy. It is hardly the only
kind, as non-discursive actions like work slow-downs or boycotts make
clear. But it is the kind of dissent to which the question of epistemic value
applies most frankly and obviously. For insofar as it is directed at getting
people to see, that, e.g., the Trump administration’s immigration policies are
not in the nation’s best interest, it can be sensibly asked whether, how, and
in what way, dissent of this form can be epistemically, as well as politically,
valuable.
We can distinguish two kinds of epistemic value. One kind is consequen-
tialist. Something has epistemic value in this sense when it has good epis-
temic consequences, as when it increases our knowledge, or justified beliefs,
Epistemic Arrogance 131
or simply results in our believing more truths. The day-to-day practices
of normal scientific inquiry are presumably epistemically valuable in this
sense. Another kind of value is deontic. Something has epistemic value in
this sense when it in some way contributes to our becoming more respon-
sible epistemic agents or knowers—either by giving us more capacity to
better perform as such agents, or as a constitutive element of responsible
agency itself. Perhaps the most basic reasons to think that acts of DPD have
epistemic value are consequentialist in nature. We’ll take a look at two such
arguments next.
Two Initial Arguments

Let’s begin with what I’ll call the Millian Argument, which can be put like this:
Millian Argument: if there is dissent directed towards some policy P

then our justification for thinking that P is in the public’s interest will
be defeated, sustained, or possibly even strengthened. In any case, it is
likely that the stock of true beliefs increases, since justified beliefs are
precisely those that are more likely to be true. Therefore, dissent has
epistemic value.
The argument conceives of the epistemic value of dissent as emanating

from its critical function. Roughly speaking, it conceives of the space of
public discourse as a space of reasons, or more bluntly as (potentially) a
rational discussion. Just as the give and take of reasons is meant to make
our arguments better, to increase our evidence base, and to rule out some
possibilities and get us to consider others, so dissent, the argument implies,
can add to our stock of justified beliefs.
Simple as it is, there is merit in this argument. One value of public dissent—
perhaps derived from the value of free speech in general, although the
proponent of the Millian Argument, despite its name, need not make that
claim—is that it can act as a form of rational persuasion, as a way of laying
out the evidence for thinking that a policy is flawed (Mill, 1989). Indeed
that is precisely why so many revolutionaries and political critics and
dissidents—from Thomas Paine to Marx—have used the pamphlet, essay,
letter, book, or blogpost to lay out their case against the policies of the pow-
erful to those whom they hoped were still persuadable. The idea is that you
can educate via the practices of dissent.
But, of course, as many of these authors would be the first to tell you,
such practices often don’t work because what we might call the epistemic
environment is corrupted. The criticisms fall on deaf ears, or only on those
ears already ringing with the trumpets of the cause. And, of course, we
have assumed that the dissent is itself largely justified, informed, and on the
right track, where in reality the percentage of political dissent that is accu-
rate is roughly about the percentage of any criticism that is accurate. And
while unjustified or ill-informed criticism can teach us that it is so—and so
increase our knowledge, if only in the negative—the public must be capable
of recognizing that fact. Otherwise they may be persuaded by a less than
reasonable “argument by tweet.”
For all these reasons, the simple Millian argument, while it can help to
justify some acts of DPD and the institutional social practices that support
such acts, it cannot serve to justify all—even those we might intuitively
think have some epistemic value. More importantly, in order to realize the
good epistemic consequences of dissent forecast by this argument, one must
be in a good epistemic environment.
Let’s therefore turn to a very different reason for thinking that dissent
can have epistemic value. Like the previous argument, this one also justi-
fies dissent by way of good epistemic consequences. But unlike the previous
argument, it begins by recognizing that epistemic environments are often
less than ideal. In particular, this line of argument notes that information
relevant to forming reasonable policy beliefs is typically not distributed
evenly throughout the citizenry. This can be due to a number of factors: pro-
paganda (Stanley, 2015), fake news, unequal educational opportunities, or
conditions of hermeneutical or testimonial injustice (Fricker, 2007; Medina,
2013). Precisely because of such factors, dissent can be of particular epis-
temic value (Anderson, 2006). Consider, for example, a black woman who
is frequently told by white people that she is exaggerating racist incidents
or seeing bias where none exists, when in fact her experiences in this regard
are perfectly veridical. In such a context, it is extremely epistemically valu-
able for her to stick to her convictions, including by participating in acts of
DPD—simply because they provide information others should know.
Diversity of Information Argument: Public dissent can bring asymmetri-
cally distributed but policy-relevant information to light. Such information
can defeat or sustain our policy beliefs, and thus dissent that produces it can
have epistemic value.
The Diversity of Information argument, in short, points to examples like
the Black Lives Matter movement to note that dissent can have epistemic
value even in corrupted epistemic environments. Nonetheless, and as with
the Millian Argument, such value can be presumably be best achieved only
to the extent those environments are not completely corrupted. Where
information suppression and epistemic injustice is pervasive and legally pro-
tected, for example, it will be very difficult for DPD dissent—as opposed to
other forms of dissent—to have good epistemic consequences.
These two arguments don’t show, of course, that acts of DPD necessarily
increase all things considered epistemic value. That should not be surpris-
ing. Helping someone across the street can have moral value without neces-
sarily having all things considered moral value (for example, if one does so
only to obtain a reward, or if the person you are helping is only crossing the
street to harm someone else). What the arguments seem to show is that acts
of DPD can have epistemic value. Nonetheless, the contingency of DPD’s
epistemic value means that in order to be realized, certain other conditions
must be in place. The conditions we are interested in are those that must
obtain if acts of dissent are to realize their value.
Clearly, legal protections of free assembly, speech, and a free press are
vital. Moreover, each such protection arguably bears interesting and com-
plicated relationships to both DPD and epistemic value generally (Gold-
man, 1999). In what follows, however, I want to concentrate on a set
of extra-legal conditions. That there are such conditions seems clear. As
Dewey put it:
Merely legal guarantees of the civil liberties of free belief, free expres-
sion, free assembly are of little avail if in daily life freedom of com-
munication, the give and take of ideas, facts, experiences, is choked by
mutual suspicion, by abuse, by fear and hatred.
(Dewey, 1939)1
To achieve a democratic common space of reasons—one where, in par-

ticular, dissenting voices can be heard—our basic attitudes towards each
other, and ourselves, matter. It is clear that they must be attitudes conducive
of some degree of mutual trust and at least not reflective of open hostility.
Dewey’s point is that they also must also be conducive of inquiry—attitudes
that encourage us to actually participate in the give and take of reasons in a
landscape of plural and conflicting opinions. The thought, in short, is this:
if acts of DPD are going to realize at least many forms of epistemic value,
the epistemic environment in which those acts are carried out should not
encourage dogmatism, contempt, and close-minded dismissal of alternative
viewpoints. That is, it should not encourage epistemic arrogance.
Epistemic Arrogance
In order to defend this claim, let me first say what I take epistemic arrogance
to be. Then I’ll discuss why certain communicative practices encourage it
and therefore corrupt the epistemic environment in a way that blunts the
epistemic value of dissent.
As I understand it, epistemic arrogance is a complex social-psychological
attitude that contrasts with what is sometimes called open-mindedness and
with what others have called intellectual humility (Christen, Alfano, &
Robinson, 2014; Church, 2016; Tanesini, 2016; Whitcomb et al., 2015).2
Like other social attitudes, it is associated with certain characteristic behav-
iors and motivations. Thus, in order to appreciate its character, it can be
helpful to illustrate various features of the attitude via example. Consider:
Dogmatic Listener: you often have coffee with a colleague, and some-
times talk about politics. When you disagree, he does not interrupt, and
he politely responds to the points you make. But he never changes his
view or even admits that he might need to think of things from a different
perspective. Over time you realize he is not actively listening to what you
say. He is only waiting his turn to speak.
The Dogmatic Listener is civil, even pleasant, but he nonetheless displays
a key feature of epistemic arrogance. He fails to be “aware of [his] fal-
libility as a believer, and to be willing to acknowledge the possibility that
anytime [he] believes something, it is possible that [he] is wrong” (Riggs,
2010, p. 180).
The phenomenon I’m calling epistemic arrogance, however, has other fea-
tures. It extends beyond mere failures to recognize one’s fallibility. When
people adopt the attitude of epistemic arrogance, they also fail to assign
epistemic blame and credit appropriately. Two examples illustrate how this
can occur:
Blamer: your aunt is deeply biased on certain matters pertaining to race
and class. As a result, she often rushes to judgment faster than the evi-
dence warrants on issues on which such matters are relevant, resulting in
mistakes. She is willing to change her views when presented with reason-
able arguments or evidence that she is mistaken about the subject, but she
never admits those mistakes have to do with her own biases or limitations.
She always insists the blame lies with someone else, who gave her wrong
information.
Mansplainer: you are pressing a point on a professional acquaintance.
He argues back at first, but eventually says, “oh yes, now I see that you are
right,” and proceeds to lay out your point in his own terms, all the while
insisting that he has discovered the “real” flaw in his earlier position.
BLAMER is not just unaware of her own biases; she is not attentive to
the role they play in her cognition. Moreover, she assigns ultimate epistemic
blame elsewhere; what she attributes to credulity is really due to racial bias.
MANSPLAINER, on the other hand, is aware that he can be wrong,
admits mistakes, and may even own that they are due to his own limitations.
But he is (at least not in this case) willing to admit that his error was due
to evidence supplied by someone else. Put differently, he is listening, maybe
even learning; but he is does not regard himself as learning from you. He
sees his beliefs as being improved by his own genius, and improperly takes
epistemic credit.
The attitude I’m calling epistemic arrogance can be roughly described as
an unwillingness to regard your worldview (or some aspect of it) as capable
of epistemic improvement from other people’s knowledge or experience.
People who are epistemically arrogant are know-it-alls. That’s a motivat-
ing state—it motivates you to not only stop inquiring, but at its extremes,
to brook no objection to your views. To be epistemically arrogant is to be
dogmatic in the common sense of the term.
Epistemic arrogance is second personal (Darwall, 2006).3 That’s because
it is both a self-regarding and an other-regarding attitude.4 It is obviously
self-regarding by being an orientation towards one’s own worldview. But it
is other-regarding in that it is an attitude that, by definition, is towards that
worldview’s relation to other people. One can’t adopt the attitude of epis-
temic arrogance on an island so to speak. It is essentially a way of regarding
yourself in relation to others.
In this way it is analogous to contempt. Being contemptuous of religion,
for example, means more than believing (or being disposed to believe) that
any given religious belief is false. To be contemptuous in this way is to
regard religious worldviews as being unworthy in certain respects, and to
see those who have them as perhaps feeble-minded, or deluded, or both.
To have contempt towards religion (or a political ideology, like Marxism)
is to have a distinctive kind of negative orientation to forms of life that
embody it.
Of course, you can be epistemically arrogant with regard to some aspects
of your worldview and not others. Aspects are rough domains or pluralities
of commitments, such as your religious, political, moral, epistemic, culi-
nary, scientific, or yes, philosophical commitments. Aspects are not single
beliefs or propositions. It is not an aspect of your particular worldview that
you believe that two and two are four, nor is it something towards which
one (typically) has an orientation, just as it is rare for someone to be con-
temptuous of just a single religious or political belief. It can happen, but
more often, contempt spreads widely from commitment to commitment.
The same is the case with epistemic arrogance.
Epistemic Arrogance and Epistemic Agency

Recall that we are interested here in the potential for acts of DPD to obtain
epistemic value. Paradigms of such acts are speeches, press conferences,
essays, pamphlets, blog-posts, and Twitter comments. How might epistemic
arrogance undermine the potential for such acts to realize epistemic value?
Let’s begin with the most obvious. Acts of DPD are attempts to change
policy beliefs. They are typically expressions of an alternative viewpoint
from the one(s) held by some institution or decision-maker. In most cases,
however, they are not intended to directly change decision-makers’ minds.
In a democracy, acts of DPD are instead generally intended to have more
indirect causal effects. An essay written in opposition to President Trump’s
immigration policies, for example, needn’t be intended to change the Presi-
dent’s mind or even be read by him. But it might be used to supply relevant
information to voters in a region where there is an upcoming Congressional
election. Its intention is therefore both political and epistemic, but its target
is not the decision-maker himself. It is often a wider public or publics who
can indirectly influence policy. As a result, if the target public(s) are not
receptive—if, for example, they are epistemically arrogant about the mat-
ter in question by refusing to consider evidence, or by being prone to shut
down dissenting views—then the given act of DPD will be less prone to
realize epistemic value of the sort promoted by the either the Millian or the
Diversity of Information arguments.
Take the case of the Millian argument. It supports the idea that acts of
DPD can supply politically relevant information and knowledge by either
defeating evidence for believing that some policy is warranted, or sustaining
or strengthening that evidence. Should, however, those encountering this
act of dissent be epistemically arrogant with regard to the issue in question,
they will, to that extent, be less receptive to alternative viewpoints. Thus,
the dissenter’s testimony and arguments will be less epistemically effective.
Likewise with the Diversity of Information argument, as the reception of
the Black Lives Matter movement amongst certain white populations would
suggest. If a given majority public adopts an epistemically arrogant attitude
towards a minority population’s report of their experience, that will lower
the likelihood that the minority’s experience will be heard.
Protest movements, and the acts of DPD that tend to accompany them,
are typically well aware of this problem, however. As both the gay rights and
the civil rights movement demonstrate, dissent that fails to produce immedi-
ate or direct good epistemic consequences because of a corrupted epistemic
environment can still indirectly change people’s beliefs over time simply by
making certain biases and limitations part of the national conversation. This
in turn can have further downstream effects of the sort sometimes described
as “consciousness raising.” Arguably, being aware of the existence of cer-
tain prejudices and biases can, in some contexts, lead people to introspect
and become more sensitive of their own biases. Hence it can indirectly lessen
epistemic arrogance. We can put this down as a third reason for thinking
that political dissent can have epistemic value.
Changing Attitudes Argument: dissent can change commitments or atti-
tudes. In particular, dissent can cause people and/or institutions to be aware
of their own epistemic limitations and fallibility, and to be less dogmatic and
intellectually arrogant. Being aware of one’s prejudices and biases has prima
facie epistemic value; therefore political dissent can produce such value.
Still, whether or not any given act of DPD will be effective in changing
attitudes can itself depend on the epistemic environment. As a number of
recent researchers have argued, for example, our life online can itself pro-
mote epistemic arrogance in ways that can inhibit an exchange of political
reasons. Much of what we know (or think we know) now we “Google-
know,” in the sense that we acquire information online from websites and
social media (Lynch, 2016). The sheer availability of information, and its
perceived reliability, has been argued to actually increase people’s confi-
dence in their knowledge and make them think they know more than they
do. Thus, some researchers (Fisher, Goddu, & Keil, 2015) suggest that our
ability to access expert opinion so easily and immediately online can cause
us to “lose track of our own reliance on it and distorting how we view our
own abilities” (Fisher et al., 2015, p. 675). This conclusion, also supported
by recent work by Fernbach and Sloman (Fernbach et al., 2013; Sloman and
Fernbach, 2017), in turn indicates that we can over-estimate how much we
individually know because “externally accessible information is conflated
with knowledge ‘in the head’ ” (Fisher et al., 2015, p. 682). Some experi-
ments, for example, show that searching for explanations on the Internet
increases the likelihood that we think we know more than we do even about
an unrelated topic—even controlling for factors such as content and time.
The sheer accessibility of information itself can lead to increased epistemic
arrogance. Of course we are right, we think—just Google it!
Our use of information technology also provides other barriers to our
ability to inform others by way of political dissent. Our online life, particu-
larly for those who use social media, is deeply personalized. The ads we see
online and the news that comes across on our phones is the result of our
own information curation and management. It is tailored to satisfy our pre-
existing preferences and political viewpoints. Moreover, and as some have
noted (Lynch, 2016; Sunstein, 2009), our social lives online are increasingly
polarized; we live in isolated information bubbles. To the extent that this is
correct, it can blunt our ability to effectively dissent via social media. Plainly
put, it is difficult to count sharing a Facebook post to one’s friends as an act
of dissent at all if all of one’s friends share the same opinion. That’s venting,
not dissenting.
Ironically, political dissent becomes even more important the more wide-
spread epistemic arrogance becomes amongst decision-makers and the pub-
lic at large—as the Changing Attitude argument indicates. Yet while there
are numerous reasons to think that acts of DPD can fail to realize the good
epistemic consequences they are meant to achieve, there remains a final kind
of epistemic value in such dissent.
So far, we’ve concentrated on how acts of DPD done by an agent can
lead to an increase in politically relevant knowledge in others. But an act of
DPD can be something for which the agent too can receive credit, precisely
because the dissent is being done to disseminate knowledge. Take a profes-
sional football player who speaks out at press conferences on the dangers
and inequalities of police violence in his city. It is possible, as some might
allege, that he is motivated to do so only because he wishes to gain press
attention for himself. But if, as he contends (and is more likely the case), he
is speaking out so that more people will come to know about such violence,
then he is, to that extent, acting in an epistemically responsible manner.
Extrapolating from such example, we can present a fourth argument, build-
ing on the other three.
Responsibility Argument: to the extent that you engage in acts of DPD in
order to increase citizens’ political knowledge, or make them more aware
of their limitations, or to bring previously unappreciated politically relevant
information to light, you perform an epistemically responsible act.
Acts of political dissent—even acts of discursive dissent—are generally
acts of conviction and frequently of moral outrage. We engage in them to
speak out for what we believe is important, just, or true. Such acts often take
courage, and thus those who make them are generally deeply committed to
the cause. I’ve taken it as given that they therefore can have political value.
The above reflections suggest, moreover, that discursive public dissent can
have real epistemic value. Yet paradoxically, these acts of speaking out for
what we strongly believe can best realize that value only to the extent that
they happen against a receptive epistemic background or environment—one
which does not, in particular, encourage an attitude of epistemic arrogance.
Publically voicing our convictions works best when our social practices
encourage us to pay attention to one another and to not assume we know
it all already. Yet even when that fails to be the case—and sadly, it often
is—political dissent can be an expression of responsible epistemic agency.
Notes
1 Also in Dewey, John, “Creative Democracy: The Task Before Us,” in J. A. Boydston
(ed.), The Later Works of John Dewey, 1925–1953, vol. 14: 1939–1941, Essays
(Carbondale, IL: Southern Illinois University Press, 1981a): 224–230 at 227–228.
2 “Intellectual humility,” as I understand it, is a technical term whose meaning
and reference is still under negotiation within both philosophy and psychology.
Like many other technical terms, it has been introduced in the hope of more per-
fectly picking out assorted phenomena only imperfectly picked out by ordinary
language.
3 Note that treating epistemic arrogance as a second-personal attitude means it is
not a trait. Traits are stable dispositional qualities of a person; unlike traits, atti-
tudinal orientations need not be stable or even consistent.
4 For a related point about intellectual humility, see Priest, Maura, “Intellectual
Humility as an Interpersonal Virtue.” Ergo (Forthcoming).
References
Anderson, E. (2006). The epistemology of democracy. Episteme, 3(1–2), 8–22.
Christen, M., Alfano, M., & Robinson, B. (2014). The semantic neighborhood of
intellectual humility. In A. Herzig & E. Lorini (Eds.), Proceedings of the European
conference on social intelligence (pp. 40–49).
Church, I. M. (2016). The doxastic account of intellectual humility. Logos and Epis-
teme, 7(4), 413–433.
Darwall, S. L. (2006). The second-person standpoint: Morality, respect, and account-
ability. Cambridge, MA: Harvard University Press.
Dewey, J. (1939, October, 20). “Creative democracy—the task before us,” address
given at a dinner in honour of John Dewey. New York.
Fernbach, P. M., et al. (2013). Political extremism is supported by an illusion of
understanding. Psychological Science, 24(6), 939–946.
Fisher, M., Goddu, M. K., & Keil, F. C. (2015). Searching for explanations: How
the Internet inflates estimates of internal knowledge. Journal of Experimental Psy-
chology: General, 144(3), 674–687.
Goldman, A. I. (1999). Knowledge in a social world. Oxford: Oxford University
Press.
Lynch, M. P. (2016). The Internet of us: Knowing more and understanding less in the
age of big data. New York: WW. Norton/Liveright.
Mill, J. S. (1989). On liberty and other writings (S. Collini, Ed., pp. 8, 15–22). Cam-
bridge: Cambridge University Press.
Riggs, W. (2010). Open-mindedness. Metaphilosophy, 41(1), 172–188.
Sloman, S., & Fernbach, P. (2017). The knowledge illusion: Why we never think
alone. New York: Penguin.
Stanley, J. (2015). How propoganda works. Princeton, NJ: Princeton University
Press.
Sunstein, C. R. (2009). Republic. com 2.0. Princeton, NJ: Princeton University Press.
Tanesini, A. (2016). Intellectual humility as attitude., Philosophy and Phenomeno-
logical Research, 94(3).
Whitcomb, D., et al. (2015). Intellectual humility: Owning our limitations. Philoso-
phy and Phenomenological Research, 91(1).
9 Emancipatory Political Dissent
in Practice
Insights From Social Theory
Rachel Ann McKinney
Introductory Remarks
Here is a common picture of dissent (and it really is a picture—an image,
representation, or fantasy—rather than anything like a theory or model, let
alone a description of reality). A consensus that P seems to have emerged in
a public forum. P is not unanimous but there is an obvious majority opin-
ion, common view, or orthodoxy in the air that P. Deliberation has largely
closed—most people, for the most part, appear to have made up their minds
that P. This has practical consequences: people in the forum act on P or are
prepared to act on P—perhaps by putting P to a vote, or letting P guide their
actions, or using P as a ground for other beliefs or inferences. So there is,
generally, an equilibrium such that P is accepted.
Then, an individual (a dissenter, if you will) raises their hand to offer an
objection. P isn’t true, she says. Instead, not-P. And she gives some reasons
why not-P: reasons like Q, R, S. Our dissenter thereby offers arguments
against P, articulating claims and appeals to publicly accessible evidence
that suggest or entail not-P. Perhaps she also takes up what we might call a
campaign against P—in addition to developing a platform with compelling,
persuasive, and plausible arguments against P, she tries to get committed P
believers/voters/actors to consider whether not-P, she mobilizes those in the
forum who had no opinion on the “P or not-P?” question to the not-P side,
she organizes or codifies a bloc, party, or caucus around the platform that
not-P (because: Q, R, S). Maybe the campaign grows contentious, and the
P-ers move to negotiation or bargaining tactics.1 Perhaps they employ some
economic, political, or social pressure against P—demonstrations, boycotts,
strikes, blockades aimed at changing the equilibrium about P. Under such
conditions it seems as though the consensus around P has been disrupted.
At this point things may start to shift. If the other members of the forum
take Q, R, and S to be compelling, persuasive, and plausible, this may enact
a recalibration among participants in the forum regarding the acceptance of
P. People may become less sure of their original belief that P. Public opinion,
sentiment, and behavior adjust to accommodate the possibility that not-P.
People see that others they trust and find competent change their mind,
Emancipatory Political Dissent in Practice 141
giving them reason to change their own minds. Or perhaps members of the
forum do not change their minds but come to be motivated by incentives
or sanctions to shift from P—they switch sides as part of a negotiation,
say.2 And then there might be reaction and counter-reaction: proponents
of P might offer counter-arguments against Q, R, S, and further indepen-
dent new defenses of P. They might develop their own platforms, parties,
blocs, or their own set of more contentious maneuvers: counter-recruitment,
counter-demonstrations, and so on.
This is a fairly standard, if idealized, story about the nature and origins of
disagreement in deliberative democracy. Elements of it can be found in Mill,
as well as more contemporary theorists including Habermas and McCarthy,
Stanley, Haslanger, and Fricker.3 Call it the 12 Angry Men theory of political
dissent. Like the play by the same name, dissent here seems to originate in open
challenge—when the speaker raises her hand to say: not-P (and here’s why).4
Once the dissenter speaks openly, the real struggle can begin—arguments to
and fro, political alignments, emotional appeals, bloc voting, escalation to
negotiation/bargaining, and so on. Dissent on this picture is a move that initi-
ates contention within a public cooperative deliberative context, structured as
a single conversation, where there was previously consensus.
Parts of the story are under-described. A forum is never a mere aggregate
of individuals and their attitudes: a group is always structured by back-
ground conditions—e.g., prior interactions, physical relations, and geogra-
phy; norms and conventions, social roles, institutions, modes of production,
power, and so on. Further, a forum rarely consists of complete strangers
who do not or cannot communicate other than via their participation at the
time of the forum itself. We talk to each other, share projects, and inhabit
space alongside one another—we share a civil society.
Nevertheless, parts of the 12 Angry Men story surely ring true. For some
in the forum (the P voters), it surely seemed as though contention started
with the public voicing of dissent.5 It probably seemed as though there was
consensus and then suddenly, out of nowhere, disagreement. And it surely
seemed as though collective action against P happened after the public voic-
ing of not-P: that’s when we saw the formation of aboveground coordination
and expression in the form of the development of a platform, campaigns,
bloc voting, demonstrations, threats of boycotts and exit, and so on.
But it’s likely that the not-P voters would tell the story differently. It’s plau-
sible that they would reply: first, P was never actually consensus. It may have
appeared so, but this appearance is misleading: people can and do disagree
privately with all sorts of views with which they are not seen as openly dis-
agreeing.6 Second, it’s plausible that many of the not-P voters knew P wasn’t
actually consensus from the beginning: after all, if I am a not-P voter, I know
of myself that I don’t believe P—that is enough for me to know that P is not
unanimous. And third, given that members of a forum likely know each
other, talk to each other, share some assumptions and a lifeworld, it’s plau-
sible that many of the not-P voters had communicated with each other about
142 Rachel Ann McKinney
the status and nature of P before the forum convened. Perhaps this commu-
nication took subtle forms: eye-rolling when P is asserted, grumbling when
flyers presupposing P are posted around town, bathroom graffiti proclaiming
“NOT-P!!”. Or perhaps the expression was more articulated: dinner table
fights about the evidence for not-P; off-the-record appeals to Q, R, and S;
hushed complaints about the stupidity of P and the obviousness of not-P
given Q, R, S.7 But the basic claim here is the same: just because dissent isn’t
openly visible doesn’t mean it isn’t there in the world, with a life of its own.
In this chapter I’m going to try to uncover this latter story—the origins
of dissent before it becomes “voiced”—that is, before it is public, collec-
tive, and recognized, especially by those in power. I’m going to suggest that,
among other things, the above description of the story gets the order of
operations wrong. Rather than initiating contention, open collective public
dissent comes late in the game of social struggle. The speaker in our story
did not emerge ex nihilo, ready and able to publicly articulate a claim and
defend it with a set of arguments—let alone ready and able to align with
others in defense of collective action on the basis of a shared commitment
to it. Under actual conditions, open collective public dissent is prefigured
by a field of thought, discourse, and activity less visible, identifiable, and
overt that prepares agents at the bottom of a social hierarchy for the tasks
of speaking and acting against the conditions (symbolic and otherwise)
that they find themselves in. Such activity is both discursive and practical.
It includes discourses like gossip, rumor, graffiti, anonymous or pseud-
onymous publication, myth, and folklore, as well as practical activity like
foot-dragging, pilfering, desertion, evasion, shirking responsibilities, false
compliance, feigned ignorance, and sabotage. In this chapter I will provide a
review of some of the relevant social theory for philosophers who might not
be familiar, and suggest that in order to understand this thing—“dissent”—
we must understand the conditions that prefigure it, set its parameters, and
make it possible. I describe emancipatory political dissent as “speaking truth
to power” that emerges from a field of activity that is paradoxically unspo-
ken, untrue, and unaddressed to power. From this field of activity, agents
are able to develop ways of challenging and upending the unjust systems in
which they find themselves.
What Is Dissent? Some Considerations

It may help to start with some brief discussion of the conceptual lay of the
land. Dissent seems straightforwardly contrary to consensus. It’s only intel-
ligible against the background of something like widespread agreement or
acceptance (or orthodoxy or hegemony or. . .). To dissent on P presupposed
that P is generally treated as true, or is seen as being generally treated as
true. Similarly, dissent cannot be affirmation of some content that is already
widely accepted: “everyone accepts P, but I dissent. I think P” is nonsensical.
So dissent is oppositional: it is offered in the spirit of negating, disagreeing,
or challenging something widely accepted.8 And this presumes some collec-
tive status: when one dissents one opposes something that is (thought to be)
common ground or mutual knowledge.
Common ground and mutual knowledge are familiar concepts for phi-
losophers of language and might present an appealing possibility for under-
standing political consensus and dissent.9 For instance, it might be tempting
to operationalize consensus in terms of mutual knowledge or common
ground—to define “consensus that P” as “an equilibrium such that P is
widely accepted (and all participants know that it is so accepted).” Then
one could, say, model all public discourse about P as one big, diffuse con-
versation among a polity about the status of P.10 Participants can then be
said to enter and exit this “conversation” as they discuss P.11 This is broadly
what the above vignette supposed. Something like this—a society-level
conversation-like discourse object—is similar in spirit to Miranda Fricker’s
description of the social imaginary, Sally Haslanger’s analysis of the func-
tion of generics in discourse, and Jason Stanley’s theory of ideology.12
One hazard of this approach in the political context is that it exaggerates
the hegemonic dimensions of mutual knowledge at the expense of oppo-
sitional, local, and counter-hegemonic knowledge. As we shall see, what
appears as consensus varies in interaction, and sometimes is the mere mani-
festation of strategic ambivalence on the part of political agents. We must be
careful not to overstate the case of what appears as consensus.
The problem is particularly acute for Jason Stanley. In his How Pro-
paganda Works it at times becomes mysterious how disagreement and
opposition could even emerge in the first place against hegemonic political
ideologies. In the context of developing a view from Weber, Stanley says:
There is a simple argument from the premise that belief is not under our
direct voluntary control to the conclusion that ‘negatively privileged
groups’ will acquire the flawed ideological beliefs of the ‘positively priv-
ileged group.’ The positively privileged group will control the dominant
narrative. If the positive privileged group controls the dominant nar-
rative, then the testimonial evidence of authorities will be the ideology
of the positively privileged group. That is the mechanism by which the
flawed ideology of the positively privileged group comes to be held by
the negatively privileged groups. The negatively privileged groups are
not exposed to an alternative ideology.13
Stanley then gives examples of elites duping the public—the Iraq War
is a paradigm here. Stanley claims that, “the psychology of persons is not
naturally suited to pursue autonomy, either of belief or action,” and as such
people are easily manipulated by those in authority. Stanley continues:
There are not many multiply replicable results in social psychology. But
one result that stands out in terms of both its robustness and its bearing
on the issue at hand shows that the psychology of persons is not natu-
rally suited to pursue autonomy, either of belief or action. These are
the well-known results of Yale social psychologist Stanley Milgram in
the 1960s. Milgram’s large body of experimental work on obedience
to authority provides strong evidence that the assertions of authority
figures are accepted even in the face of explicit counter evidence.
The challenge in explaining why negatively privileged groups would
accept the ideology of the highly privileged group is that negatively
privileged groups have clear evidence that the ideology is false. They
see around them instances of social injustice that are caused by that
ideology. But we see in the Milgram experiments that the true experi-
mental subjects, those administering the shocks, were also given ample
evidence throughout that their actions were wrong . . . by a large
majority, the real subjects nevertheless persisted in applying the electri-
cal shocks. They did so because of their uncritical acceptance of the
scientist in charge’s claims of the moral acceptability of their actions,
despite clear evidence to the contrary. . . . Milgram’s work supports the
view that the ‘uncritical acceptance’ of the claims of authority figures,
especially when delivered in the language of pseudoscientific expertise,
is the norm rather than the exception. Thus Milgram’s work provides
some evidence of at least many groups’ susceptibility to a technicist
ideology.14
This is a striking interpretation of the Milgram experiments. One won-

ders how, if compliance meant volunteers “uncritically accepted” the “moral
acceptability” of what they were told to do, Stanley would explain the dis-
tress, anxiety, and guilt felt by the volunteers involved. Nevertheless, Stanley
claims that these results can be generalized outside of the lab—members of
the public, including and maybe especially those at the bottom of a social
hierarchy, will tend to uncritically accept the dominant narratives, values,
beliefs, and preferences that they are provided by those at the top. This ren-
ders them incapable (!) of opposition. Stanley again:
Suppose we accept that members of negatively privileged groups tend

to adopt the dominant narrative. If so, they will be incapable of acting
against the very system that oppresses them. Members of the negatively
privileged group will fling themselves against the high barriers erected
against them, only to blame themselves for their failure to scale them.
The few who do manage the feat will be convinced that they did so out
of their own individual merit. They then will be used as pawns in a pro-
paganda game of legitimization of the dominant ideology. The situation
will seem deceitful to no one.15
The claims made here bear repeating. Again for emphasis: the psychology
of persons “is not naturally suited to pursue autonomy, either of belief or
action,” those at the bottom of social hierarchies “are not exposed to an
alternative ideology” (let alone agents who develop their own alternatives
in the form of, e.g., class consciousness), in experimental settings people
“uncritically accept” the moral permissibility of cruelty and domination on
the basis of authority in the face of explicit counter-evidence, and members
of “negatively privileged groups” are “incapable of acting against the very
system that oppressing them.”
This analysis of political ideology is deeply fatalistic. How, in such a struc-
ture, are agents able to acquire the knowledge necessary for understanding
the social world, let alone acting together to change it? How is emancipa-
tory political dissent possible?
Luckily, the history of social struggle gives us reason to be skeptical of
such a picture. Pursuing autonomy of belief or action—thinking for your-
self, acting on your own will—is not contrary to the nature of persons,
even and especially subaltern persons. Indeed, it is what makes us moral
and political agents in the first place, and what makes the working class the
engine of history. People at the bottom are not quarantined from alternative
ideologies. Under the right conditions they are the ones who create alter-
native ideologies! Oppressed people do not uncritically accept the moral
claims of authorities—they don’t blindly trust the elite’s interpretations of
the world. They are instead ambivalent about the moral claims the powerful
make, sometimes performing strategic ambiguity while developing and act-
ing on their own oppositional interpretations. Finally: people at the bottom
are not “incapable of acting against the very system that oppresses them.”
They may not do so (or may not do so overtly, as we see below), but this is
not because they lack the capacity—it’s because they haven’t seen a useful
opportunity, or haven’t been able to coordinate well enough with others,
or the risks of doing so publicly are too high at the moment. We shouldn’t
mistake the practical challenges of resistance for epistemic deficits.16
This is all just to say that we should be suspicious of attempts to ground
consensus and dissent in some equilibrium state of mass acceptance that
mimics acceptance of a proposition for the purposes of a cooperative con-
versation. Modeling consensus as mutual knowledge or common ground
runs the risk of misdescribing the lay of the land. and obfuscating the very
political phenomena we’re interested in understanding in the first place. We
need a story of the origins of dissent that can help make sense of how people
actually do change the world around them.
Luckily, social theory is here to help. Here I will review some of the lit-
erature on the complex field of relations, actions, spaces, and discourses
that constitute the background conditions of consensus and dissent. I will
focus on emancipatory political dissent—dissent that aims at changing a
social structure to more readily manifest conditions of justice. Such politi-
cal dissent is typically but not exclusively addressed by agents to states,
regimes, institutions, or policies/practices thereof. Below I hope to give some
resources for helping us understand the nature of such interaction.
Emancipatory Political Dissent
Against the preceding picture that posits dissent as a relation to a diffuse
discursive object (“consensus” or “the social imagination” or “ideology”),
I suggest we think of political dissent as a relation between discourse
communities—in particular, between a public and a counterpublic. In this
chapter I will focus on emancipatory political dissent as a characterization of
social or political opposition that develops in a counterpublic. In the prac-
tical realm such opposition is not (just) rejection of some widely accepted
propositional content (e.g., P in favor of not-P) but challenge to a practi-
cal state of affairs—e.g., a structure of conventions, norms, expectations,
institutions, or regimes that constrain and enable thought and action. Such
thought and action includes but is not reducible to propositional content.17
To this end, within the political realm, dissent is a practice of both voice
and influence. It is a practice of voice insofar as it constitutes expression:
e.g., a speaker making her point of view known.18 This is obvious in the
parlance of a common idiom of the left: dissent is about speaking truth to
power. This idiom will be a useful heuristic for us going forward. Political
dissent is also a practice of influence because it is not merely expressive—it
also seeks to enact social or political change. Such change might involve
persuasion—e.g., changing hearts/minds—but also changing laws, norms,
policies, institutions, habits, expectations, behavior, and so on.19 Because
under non-ideal conditions mere words (“voice”) are unlikely to motivate
agents and institutions to change, material, social, and economic pressure
is often used to back up expression of dissent on practical states of affairs
(e.g., “influence”).20
In a democracy, we are likely to think of political dissent as a feature
of the public sphere: as overt speech or action done in the open against
something like orthodoxy, consensus, dominant ideology, or popular opin-
ion.21 Familiar examples that draw out the “voice” dimension of dissent in
the public sphere include letters to the editor, public speech at a town hall
meeting, flyering, pamphleting, displaying signs or flags, publishing essays
or books. Public examples that draw out the “influence” dimension of dis-
sent include voting, campaigns, petitions, legislation, advocacy, calling con-
gresspeople, and court filings. Examples of public dissent aimed at influence
outside the mechanisms of the state may involve activity that is a little row-
dier: e.g., demonstrations, teach-ins, civil disobedience, strikes, boycotts,
and blockades.22 As can be seen from these examples, influence often occurs
via “people power”—through the formation of a group, party, association,
coalition, or movement willing and able to create incentives and sanctions
against a regime (e.g., a state or legal structure) and its actors through col-
lective action.
At this point we reach an ambiguity. Some dissent—much of it famil-
iar from the history of democracy—is emancipatory. The English term dis-
sent arises, for instance, with the English Dissenters during the Protestant
reformation: dissent “from below” against unjust authoritative orthodoxy.
This usage further finds resonance in the history of egalitarian thought and
action—e.g., the history of the American, French, and Haitian revolutions,
the labor movement, the abolitionist movement. But while dissent is a mech-
anism for achieving social change, such change need not be emancipatory. It
need not aim to make conditions freer or more equal, just, humane. It may
in fact aim for the opposite.
Reactionary and revanchist movements often also appear as dissent “from
below” against a state or legal regime—they are superficially similar to and
often contemporaneous with the liberatory movements described above.
The Louis Napoleon of Marx’s 18th Brumaire is a good example. Donald
Trump is another. Such movements evolve to produce claims and practical
action against the powers that be toward political goals of restored empire
or monarchy, expanded or repossessed territory, religious conquest, absolut-
ist monarchy, social domination. Trumpism in 2017 America, for instance,
opposes the professed commitments to racial egalitarianism of both the
state and mainstream liberal society—e.g., residential and school integra-
tion, immigration, ameliorative projects like affirmative action, as well as
“p.c.” cultural expressions like television and movies featuring people of
color—and seeks to alter the social structure to more thoroughly recognize
and defend the “superiority” and “sovereignty” of whites. Trumpism is, in
this way, dissent against a social structure with the aim of making it less just,
equal, humane, or free.
Both liberatory and reactionary social movements have the formal
characteristics of collective, public, often contentious voice and influence
against a state regime. It is therefore tempting to group reactionary move-
ments together with liberatory movements. I reject this move. In this chapter
I will ask us to distinguish between liberatory and reactionary social move-
ments, and thus to distinguish emancipatory political dissent from reaction-
ary political dissent. The object of analysis here is the former: voice and
influence by a liberatory social movement against an unjust social structure.
A moralized conception of political dissent like the one on offer provides
the following proposal. Political dissent is a social mechanism of voice and
influence with liberatory democratic potential: it is a means by which a
public can give feedback to a regime toward realizing human interests in,
e.g., justice, liberty, equality, and welfare that a social structure (e.g., a state
or legal regime) does not yet embody.23 On this conception political dis-
sent is a mechanism for the (possible, not inevitable) realization of moral
progress. Reactionary political dissent, by contrast, is a malfunction of such
a mechanism.24 A full defense of this position is outside the scope of this
chapter, but the very concept of moral progress requires us to be able to draw
such distinctions in the first place.25 Following this line of thought, I hope to
sidestep the question of how reactionary and liberatory social movements
differ and relate, and adopt a straightforwardly moralized interpretation of
emancipatory political dissent. I’ve described “speaking truth to power” as
a bumper sticker—and it is in the nature of bumper stickers to be moralized.
With that in mind, I leave the analysis of reactionary social movements for
another time.
Publics and Counter Publics

As I try to stress below, such practices of voice and influence aren’t just
embedded in a public sphere but also construct a public sphere.26 Publics
and their properties (e.g., what is widely accepted in them) do not just
exist—they are made.27
Michael Warner offers a useful analysis of publics and their properties.28
A public, on Warner’s account, is a relation between strangers built partially
through discourse. That is, people come to relate to each other as members
of the same public, and do so at least partially by being subject to, addressed
by, and participating in the same ideas, symbols, vocabulary, representa-
tions, images, texts, and so on.29 Discourse provides a lexicon, a set of asso-
ciations and references, and a body of mutual knowledge that members
of a group who are otherwise strangers can treat as common ground. So
an accurate answer to the question, What makes us members of the same
public? will include—at least in part—appeal to facts like: it’s likely that
we watch the same shows, read the same newspapers, see the same bill-
boards, and so on. Publics on this model are prior to nation-states, are self-
organizing, arise via attention rather than deliberation, are phenomena that
constrain and enable intertextual reference and reflexivity, and are explicitly
materially, historically, and temporally located.30
The background of discourse that becomes shared among members of
a public circulates, evolves, and is adapted by particular speakers, agents,
and groups relative to their beliefs, needs, interests, tastes, and so on.
These beliefs, needs, interests, and tastes themselves also shift in response
to a background of discourse: agents adapt to the world around them.31
Importantly, members of a public do not relate to each other as mere fel-
low members of an audience (in the way that two strangers sitting next to
each other watching a ballet performance might), but as agents and speak-
ers in their own right. Members of a public add to the discourses of which
they are a part. This structure is what makes a public participatory.32 Prior
discourse—that tweet, that podcast, that graffiti—comes to be an object of
attitudes, orientations, and responses: prior public objects of the discourse
are referred to, built on, critiqued, ridiculed, echoed by members of a pub-
lic.33 Thus: YouTube comment sections, Facebook threads, retweets.
Where there is a public there is also often a counterpublic: a space of
discourse that is not spoken with everyone in mind, but rather finds bound-
aries in what a general or mass public finds unacceptable.34 Mass audiences
self-select out of discourse that assumes, e.g., deviant identities, interests,
norms, or values. Thus subcultures (e.g., queers, punks, nudists, trekkies,
and so on) can evolve “in plain sight” without pressure to accommodate
to more mainstream audiences. They take on a life of their own: what is
obvious, taken for granted, mutually accepted, or common knowledge in
a counterpublic is often completely unnoticed and unheard of in the more
mainstream public.35
Some but not all such counterpublics are subaltern: they are created by
those at the bottom of a social hierarchy. Such subaltern counterpublics
are a critical discursive space for the possible development of novel, alter-
native, and oppositional texts, meanings, narratives, habits, scripts, and
expectations.36 For instance, hip-hop as it came to be constructed among
marginalized and ghettoized black communities in the U.S. created just such
a subaltern counterpublic.37 As Nancy Fraser puts it:
History records that members of subordinated social groups—women,

workers, peoples of color, and gays and lesbians—have repeatedly
found it advantageous to constitute alternative publics. I propose to
call these subaltern counterpublics in order to signal that they are par-
allel discursive arenas where members of subordinated social groups
invent and circulate counterdiscourses, which in turn permit them to
formulate oppositional interpretations of their identities, interests, and
needs. Perhaps the most striking example is the late-twentieth century
U.S. feminist subaltern counterpublic, with its variegated array of jour-
nals, bookstores, publishing companies, film and video distribution net-
works, lecture series, research centers, academic programs, conferences,
conventions, festivals, and local meeting places. In this public sphere,
feminist women have invented new terms for describing social reality,
including “sexism,” “the double shift,” sexual harassment,” and “mari-
tal, date, and acquaintance rape.” Armed with such language, we have
recast our needs and identities, thereby reducing, although not elimi-
nating, the extent of our disadvantage in official public spheres. . . . In
principle, assumptions that were previously exempt from contestation
will now have to be publicly argued out. In general, the proliferation of
subaltern counterpublics means a widening of discursive contestation,
and that is a good thing in stratified societies.”38
This brief discussion provides us with some scaffolding for understanding

the discursive and relational nature of emancipatory political dissent. Dis-
sent is oppositional voice and influence. It does not appear out of a vacuum
via the insights of a lone agent. Rather, it emerges from the intertextual,
relational, material, and variable background conditions that constitute the
social world: what I’ve called (after Michael Warner and Nancy Fraser) the
public and counterpublic spheres.
Counterpublics and Social Movements

Some politicized counterpublics may eventually become social movements.
Social movements are a formation of political agents unified, loosely, by
their attempt to change some aspect of a social structure “from the ground
up.” Social movements are the context for much of the public, collective
voice and influence of dissent (indeed, we even describe some social move-
ments as dissident social movements—e.g., pro-democracy movements in
China and the Soviet Union).
Sociologists describe the expressive action of such social movements as
contention—the
episodic, public, collective interaction among makers of claims and

their objects when (a) at least one government is a claimant, an object
of claims, or a party to the claims and (b) the claims would, if realized,
affect the interests of at least one of the claimants. Roughly translated,
the definition refers to collective political struggle.39
The contentious politics framework developed by Sidney Tarrow, Charles

Tilly, Elizabeth Anderson, and others involves a set of articulated claims (by,
e.g., members of a social movement) and a particular mode of claim-making.
For instance, a movement may involve demands such as lower grain prices,
a disgraced leader leaving office, an 8-hour work day, or an end to civil war,
as well as a justificatory framework on behalf of such demands. So, grain
prices should be lower because they are a result of unfair price gouging by
a corrupt monopoly; the leader should leave office because he has broken
the law, lost public support, and can’t be trusted to be in charge; an 8-hour
work day should be instituted because it provides a standard against which
workers have the opportunity to exercise the fruits of their labor (to which
they are entitled in the first place); a civil war should end because it violates
the basic security and freedom of all members of a polity. A movement may
engage in a variety of modes of advocating on behalf of the claims that
justify such goals. These modes of advocating—or “repertoires of conten-
tion”—involve actions that are both practical and discursive, and that are
culturally intelligible within the particular context in which they arise.40
Take for instance the Women’s Peace Movement in Liberia. This movement
consisted of both Christian and Muslim Liberian women advocating and
agitating for an end to the Liberian Civil War. Their advocacy and agitation
addressed state actors, militias, as well as the “rank and file” men in their
communities. It involved practices such as singing in fish markets, wearing
all white, sex strikes, and the strategic use of curses and taboos to shame
(male) politicians back to the negotiating table.41
Given the voice/influence and contentious politics frameworks as
described, we can revise our bumper sticker. Political dissent (in particu-
lar, the emancipatory political dissent under discussion here) is not merely
speaking truth to power. Rather it is something close to what sociologists
mean by contention—that is, the historically and culturally embedded pro-
cess of public claim-making and collective action to change an institution,
state, or regime.
Contention Before Publicity
Now I will claim that this thing called “political dissent”—as we’ve described
it, public claim-making and collective action against an institution, state, or
regime in pursuit of change—comes quite late in the game of social struggle.
Before the outbreak of contentious politics (the public articulation of claims
and the manifestation of collective action on their behalf), an entire field of
oppositional expression and action is mobilized just below the surface of
public voice and influence.
The picture I develop will be this. People in unjust regimes push against
the system at the point at which they find themselves long before they are
able to articulate such action in terms of public claims for justice (e.g., “grain
prices should be lower because they are a result of unfair price gouging
by a corrupt monopoly”), or to act collectively in pursuit of such claims
(e.g., grain riots). Such prior activity is not idle or apolitical. Instead, it
constitutes the discursive, practical, social, and psychological preconditions
necessary for public collective claim-making, and sets the terms, parameters,
and limits of the claims that end up being made. Analytic philosophers have
largely ignored this field of activity, but understanding the phenomenon
called “emancipatory political dissent” requires understanding the activity
that sets the stage for it.
Remember our bumper sticker: dissent is speaking truth to power—in
particular, publicly, collectively speaking truth to power in pursuit of social
change. I’m going to argue that this speaking truth to power comes some-
what paradoxically from a field of activity that is unspoken, un-true, and
un-addressed to power. Successful liberatory social movements involve
counterpublics “bootstrapping up” from this field of discourse and activity
into a social formation capable of publicly making claims and acting col-
lectively in their defense (e.g., a movement).
The field of activity I’m going to describe constitutes the practical, dis-
cursive, and relational field against which public dissent and contention are
formed. I describe this prior field of activity as unspoken, untrue, and unad-
dressed to power. It is unspoken because much of the relevant activity is
noncommunicative—it is (often) done for instrumental reasons of securing
food, shelter, autonomy, or sanity without expressive content and without
any “audience” in mind. It is un-true because, when it is communicative, it
(often) takes the form of discourse that is not truth-apt: e.g., gossip, rumor,
myth, superstition, curses, religion, and ritual. It is unaddressed to power
because it (often) circulates within the subaltern counterpublic only—a
space to which those in power are not interested or welcome—and when it
is performed “in the teeth of power,” it is done covertly, anonymously, or
ambiguously. Yet somehow from this field of the unspoken, the untrue, and
the unaddressed to power, dissident social movements emerge to publicly
speak truth to power and change the conditions under which they exist.
One useful account of this prior field of activity comes to us from politi-
cal scientist James C. Scott. Scott, influenced by scholars in the tradition
of Subaltern Studies who focus largely on the lives of the poor and dispos-
sessed in understanding colonial history, trains his attention on the split
between “on-stage” and “off-stage” discourse and action.42 Interlocutors
of a given social group speak differently when they believe they are “alone
among themselves.” For example, the words of the master in front of “his”
slaves is very different from the words of the master alone with other planta-
tion owners, both of which vary dramatically from the words of the slaves
among themselves.43
Consider first the unspoken aspect of such counterpublics.44 Scott argues
that the public transcript—the front-facing discourse, action, relations, and
social meanings of slaves as they interact with masters and masters with
slaves—is never hegemonic. Instead, it is a negotiation between political
agents premised on the social, political, and economic needs of those in
power to save face and those under their thumb to stay safe and sane.45
As such, the public transcript displays an incredible amount of intentional
ambiguity, euphemism, pretense, deniability, deflection, and insincerity as
agents in both positions navigate displays of legitimacy, respect, affront,
and subservience.46 For example, a slave’s flattery toward the master helps
secure the master’s image of himself as kind, generous, and forgiving, while
also allowing the slave to secure for themselves better treatment and per-
haps a bit of autonomy (“It’s so generous of you to allow us to keep some of
the harvest for ourselves”). The act of playing dumb serves as a mechanism
for getting out of onerous tasks and avoiding accountability for mistakes or
inattention without being visibly insubordinate (“I had no idea that was off-
limits. If I did I would never have gone.”) Foot dragging—the intentional
or deliberate delay or slowdown of compliance with a directive—secures
a wider breadth of time for tasks and allows one to work at one’s own
pace, thereby manifesting some control over one’s own labor. Shirking—
avoiding or looking the other way regarding one’s responsibilities—secures
autonomy from onerous tasks and expectations. Desertion—the abandon-
ment of a post or role entirely—is a means of exiting as much as possible the
relations into which one has been nonvoluntarily pressed. Scott defines these
practices as infrapolitics—the low-level activity of covert noncompliance.47
Such practices are as much a matter of what people at the bottom can
get away with within the “public transcript ” of acknowledged mutually
understood meanings as they are whatever first-order material resources
such agents aim to secure in the first place. As such it is important that
infrapolitics happen within the bounds of plausible deniability.48
Suppose a tenant farmer is allowed to take home low-quality harvest
grain from the fields owned by a lord. Suppose also that the tenant farmer
is himself the one tasked with evaluating or “grading” the grain. In this
scenario, it is rational for the tenant farmer to grade as much grain as “low-
quality” as he can get away with without arousing suspicion. The last part
is crucial—the tenant farmer only benefits from this ad hoc arrangement as
long as the high-grade grain he turns over to the lord is within the range
of plausible harvest weight—that is, only so far as the lord does not notice
it’s lighter than it would otherwise be. If the farmer grades too much of the
grain as low-quality such as to arouse the lord’s suspicions that he is pilfer-
ing, the jig is up. Hence the importance of the preservation of plausible deni-
ability. As Scott is careful to note, the public transcript always betrays such
moments of ambiguity precisely because it is in the interest of people on the
bottom to manufacture it and people on the top to ignore it: such plausible
deniability provides cover for those at the bottom to achieve some condi-
tions of sanity, autonomy, and better material conditions “in the teeth of
power” while also saving face and preserving the appearance of legitimacy
for those at the top.49
Spirituals, hymns, and songs provided a more symbolically mediated kind
of ambiguity: on one interpretation, slaves sang of the good of the kingdom
to come—on another, they sang of the promise of a new life in the north.
Such public expressions make use of euphemism when in earshot of the
powers that be. As Scott says:
The use of euphemism as disguise is most striking in the pattern of folk-

tales and folk culture generally among powerless groups. . . . Euphe-
misms continually test the linguistic boundary of what is permissible
and that often they depend for their intended effect on their being
understood by powerholders. Slaves in Georgetown, South Carolina,
apparently crossed that linguistic boundary when they were arrested for
singing the following hymn at the beginning of the Civil War:
We’ll soon be free [repeated three times] When the Lord will call us
home.
My brudder, how long [repeated three times] ‘Fore we done suffering
here?
It won’t be long [repeated three times] ‘Fore the Lord call us home.
We’ll soon be free [repeated three times] When Jesus sets me free.
We’ll fight for liberty [repeated three times] When the Lord will call
us home.43
Slave owners took the references to “the Lord” and “Jesus” and “home”
to be too thinly veiled references to the Yankees and the North. Had their
gospel hymn not been found seditious the slave worshippers would have
had the satisfaction of having gotten away with an oblique cry for freedom
in the public transcript.50
Such subterfuge (euphemism, pilfering, playing dumb, and so on) is criti-
cal precisely because under slavery, feudalism, and caste hierarchies open
defiance is not merely risky but often suicidal. To openly defy one’s master
would be to invite retaliation and punishment. Unless one has high credence
that both (i) similarly situated others are ready and willing to also risk their
own well-being to participate in confrontation and (ii) such confrontation
is likely to be successful, any attempt at overt rebellion is at least ostensi-
bly irrational.51 Hence, the complex, ambiguous, deniable nature of public
transcripts.
Prehistory of Dissent
Let’s now look at manifestations that reveal the untrue aspect of this prior
field. Anthropologist Aihwa Ong describes the dislocations, conflicts, and
disturbances of female Malay workers in Japanese-owned Free Trade Zone
factories in the 1970s and 1980s. Industrialization and changes in land ten-
ure caused significant rural dislocation as many Malay peasants migrated to
urban centers looking for work. Many of these peasants—especially young
unmarried women—took jobs in Japanese-owned production factories,
often run by Chinese foremen. The values, expectations, and tempo of this
new cultural context contrasted sharply with their earlier village life: not
only their rigid discipline into capitalist regimentation, pace of work, and
supervision, but also into Western constructions of sexuality, consumer cul-
ture, and modernity.
The process of proletarization and urbanization were not seamless transi-
tions; the Malay women who moved for work were not the docile bodies
and spirits that capital expected. Indeed, in the 1970s a number of factories
experienced significant disruption as a result of hauntings, ghost sightings,
and spirit possession. The background to such disruptions was a Malay
folk culture that includes many spirits—both benign and malevolent—who
exist at the interstices of the human and spiritual worlds. Some communi-
ties of Malay believe that women who are not sufficiently spiritually vigilant
will “[b]ecome possessed by angry spirits (kena hantu) when they wander
unsuspectingly onto the sacred dwelling places of spirits.”52 Lo and behold,
according to the newly urbanized Malay peasants, many of the Japanese
factories in the newly decreed Free Trade Zones happened to have been built
on top of ancient grave sites. And when the sacred dwelling place of spirits
becomes a factory, entire floors staffed by young “spiritually risky” Malay
women are opened to disruption and chaos. Says Ong:
The late 1970s produced a flurry of newspaper reports on “mass hys-

teria” in free trade zones. In 1975, forty Malay operators were seized
by spirits in a large American electronics plant based in Sungai Way.
A second large-scale incident in 1978 involved some 120 operators in
the microscope sections. The factory had to be shut down for three
days and a spirit-healer (bomoh) was hired to slaughter a goat on the
premises.53
Spirit possession manifested in multiple ways: as screaming fits, seizures,

fainting spells, and physical violence against those who came to the young
women’s aid. Factory owners and foremen expressed frustration with the
provinciality and superstition of the peasant women workers. One Ameri-
can director “wondered how he was to explain to corporate headquarters
that ‘8,000 hours of production were lost because someone saw a ghost.’ ”54
Meanwhile women’s accounts of the phenomenon describe being over-
taken by floating headless apparitions, were-tigers, and tall men with long
tongues, all while they were at their most vulnerable: in the bathroom, the
prayer room, or taken unawares on the factory floor.
Ong’s case studies are dramatic. The waves of disruption described
brought entire factory floors screeching to a halt for days at a time, required
factory owners to bring in outside support from the workers’ home villages,
and necessitated the organization of feasts, sacrifices, and ritual purifica-
tions on the workers’ behalf. The lesson before us is clear: a ghost can have
the same effects as a well-coordinated strike.
Yet the women in Ong’s examples don’t see their experiences in politi-
cal terms. Rather, their normative frameworks are ones of personal moral
failure (the unfairness of this rule, the cruelty of that foreman) and spiritual
vigilance or intemperance (the pollution of the factory built on a graveyard,
the ritual carelessness or cosmic weakness of one’s fellow workers). We can
easily imagine a trajectory whereby the spirit possessions and hauntings
come to be politicized—why would the factory be haunted? Because of the
bad boss. Why is the bad boss so bad? Because he’s cruel—but this was not
the case in this instance.
Along these lines, Scott offers another example of the emerging politi-
cal significance of the untrue. In the months before the Haitian revolution
a wave of rumors swept across the slave communities and plantations of
Sainte Domingue. Word had it that the French crown had outlawed the use
of the whip and mandated the observance of a three-day weekend, but that
plantation owners on the island were recalcitrant in acting in accordance
with these decrees. Similar rumors spread in the British context as well:
In the slave rebellions in the late eighteenth and early nineteenth centu-
ries . . . there was a fairly consistent belief that the king or British offi-
cials had set slaves free and that the whites were keeping the news from
them. Barbadan slaves in 1815 came to expect they would be freed on
New Year’s Day and took steps to prepare for that freedom. . . . Slaves
treated the supposed decree as an accomplished fact, and incidents of
insubordination and resistance to work routines increased, leading
within a short time to the revolution that would culminate in Haiti’s
independence.”55
Other examples abound. Patricia Hill Collins describes the significance

of Biblical parables in moral testimony; Peter Leeson describes the func-
tion of maledictions and curses in building social norms against brigand-
age; researchers in anthropology and ecology describe the role of folklore in
indigenous land management.56 These discourses may not be truth-apt when
taken literally but neither are they epistemically inert—they are invaluable
for the culturally embedded uses to which they are put, and serve as leverage
on the attitudes and behavior of others, especially those in power.
While these examples may seem remote or fanciful, they are neverthe-
less instructive. Superstition, myth, and gossip serve as tools in the hands
of subaltern communities for developing the oppositional narratives and
values necessary to the eventual formation of articulated, truth-apt claims
for justice—as well as the collective power to see them met. They constitute
the discourse, vocabulary, and values against which agents can, eventually,
publicly and collectively speak and act against an unjust regime, institution,
or state. These “untrue” discourses provide the raw materials of resistance
prior to acknowledged, open, public, and collective displays of defiance,
insubordination, and noncooperation.
Dissent’s Transition to Publicity, to Claim-making,

and to Collectivity
How does a social movement bootstrap up from the practices described by
Scott and Ong into something like contention in Tilly and Anderson’s sense?
How do people get from that which is un-spoken, un-true, and un-addressed
to power to that which constitutes speaking truth to power? While there
may always be moral heroes, martyrs, and brave leaders who set examples
for others, the courage to openly speak and act against the powers that be
does not arise ex nihilo. Instead it requires an oppositional culture, what
I’ve referred to above (following Michael Warner) as a counterpublic—a
discursive space created by participants who do not speak with an entire
public in mind but rather some smaller subset of it—a subset to which the
rest of the public does not or would not want to be a part.
The development of such a culture oriented against a regime is jointly
discursive and practical. It involves the development of oppositional mean-
ings and frames of reference that contest the regime (manifested in practices
like narrative, myth, gossip, rumors, curses, complaints, stories, songs, and
prayers) as well as practical action against the operation of the regime—often
through activities of small-scale disobedience, refusal, and noncompliance
like I have described above—e.g., shirking, pilfering, poaching, or desertion.
Successful social movements are able to “bootstrap” up from the covert,
deniable, and subjunctive expressions of an oppositional culture or coun-
terpublic into rational claims of fairness, right, and justice by identifying,
articulating, clarifying, and translating the moral insights embedded in the
underlying folk discourse for a (more) public audience and solving the col-
lective action problems associated with risky joint action under unjust condi-
tions. Space does not permit a full examination of this process, including the
development of oppositional consciousness or a politically informed stand-
point in this process of bootstrapping, but a telling example is warranted.57
Earlier, I described ambiguous public transcripts between, e.g., slaves and
masters. Now let’s turn to “hidden” transcripts among slaves themselves.
This is a manifestation of what I describe as activity that is “unaddressed”
to power. Scott highlights discursive practices such as gossip, rumor, song,
and folklore as prime examples of such hidden transcripts. These hidden
transcripts occur in what we have described earlier as a kind of counter-
public: a space of relations and discourse removed from popular attention
that set the conditions for the development of alternative and oppositional
interpretations, assumptions, presuppositions, values, themes, associations,
and meanings. In this case the counterpublic is also subaltern—it is a rela-
tion among people at the bottom of sustained social, political, and economic
hierarchies that allows them to speak and act among themselves outside
surveillance by those with power over them.
The religious practices of black slaves in the American south provide a
good case study in such subaltern counterpublics and how they develop
as sites of oppositional consciousness and resistance. “Hush arbors” were
meetings of slaves after dark, in whatever place they could find—fields or
basements or clearings—for the purposes of worship, religious teaching,
community, and kinship.58 Such spaces were predicated on practices of con-
cealment and subterfuge—sneaking off after dark, meeting secretly with
slaves at other nearby plantations, singing into upturned chamber pots to
muffle song.59 These spaces had to be organized covertly precisely because
they were so dangerous—slaveowners had enormous interest in preventing
association among slaves capable of coordination: e.g., mass desertion, mar-
ronage, sabotage, mutiny, and revolt.
Plantation owners were right to be wary of slave religion. While white
Christian teaching of slaves emphasized New Testament doctrines of meek-
ness, subservience, and humility, slaves among themselves found resources
in the Christian tradition for imagining emancipation: the Hebrews’ flight
from Egypt, David and Goliath, the parable of the Good Samaritan, and
so on. Sermons and teachings favored by slaves when among themselves
included themes like the weak prevailing over the strong, the poor over the
rich, and the righteous over the cruel. Christian thought introduced con-
cepts such as oppression, liberation, justice, desert, charity, and courage via
Biblical narratives and parables that emphasize the independence of moral
truth from a context’s prevailing social relations.60 Pharaohs, kings, and
emperors are often positioned as villains, not heroes. The Bible is replete
with confrontations against the cruelty, corruption, and injustice of the
powers that be. The discourses of the early black church—sermons, spiri-
tuals, scripture, parables—provided a moral compass independent of the
desires and interests of the ruling class. This framework provided much of
the vocabulary and concepts for articulating claims of moral clarity and
justice in the struggle for emancipation.61 Thus, the hidden transcripts of
hush arbors provided participants with a moral compass not dictated by the
terms of the prevailing hegemony of chattel slavery.62
Such collectives—and the abolitionist movement that was informed by
and emerged from it—is a prime example of Fraser’s subaltern counterpub-
lic. The counterpublic of the hush arbor supplies not only propositional
content (e.g., information pooled via testimony), but other epistemically and
politically crucial resources—concepts like “oppression,” stories of previous
victory against the powerful, rich, and cruel, a set of values (cunning, soli-
darity, mutual aid) independent of prevailing powers, as well as the social
relations and networks necessary for the circulation and dissemination of
such knowledge, values, and narratives—among slaves on a plantation as
well as between plantations.
Conclusions
We are now in a position to see what is missing from Stanley’s hegemonic,
mutual knowledge theory of political ideology and the 12 Angry Men
theory of dissent. Stanley’s picture of political ideology left it mysterious
how agents could ever come to dissent in the first place. If dissent requires
opposition, and if agents are “naturally” suited to compliance and uncritical
acceptance, we should expect people to never push back against the system.
Why? Because they are never “exposed” to the epistemic tools necessary for
doing so. As the brief discussion above suggests, this grossly misstates the
trajectory of emancipatory political dissent. Subaltern peoples themselves
develop the epistemic tools necessary for understanding, responding to, and
changing the conditions of their lives—in the off-stage spaces of a counter-
public like a hush arbor they are able to develop the narratives, concepts,
and values necessary for rejecting the background ideological structure
against which they are positioned. The development of oppositional values
fuels both small-scale noncompliance like infrapolitics (pilfering, sabotage,
evasion) and, eventually, collective noncompliance in the form of a social
movement’s contentious politics. While it is undoubtedly true that this pro-
cess is difficult and there are no promises to its outcomes, subaltern people
are not doomed to an epistemic fate of misunderstanding and uncritical
acceptance. Instead, they are creative and perceptive agents collectively
capable of moving history.
The 12 Angry Men theory of dissent is likewise inadequate as a story of
the social life of emancipatory political dissent. Emancipatory dissent does
not arise atomically from a single individual who then builds support in
the form of rationally persuading others to change their beliefs, opinions,
attitudes, or behavior. Instead, dissent emerges through prior collectivities:
in the to-and-fro space of a counterpublic, with cultural tools like supersti-
tion, myth, and gossip as scaffolding for and leverage on collective action.
Articulate, truth-apt, publicly reasonable political claims come very late in
the game of social struggle: after a counterpublic has become politicized and
organized enough to address a state or regime on its own terms. Such politi-
cal claims are prefigured by a field of activity that is ambiguous: actions
and discourses that are less visible, identifiable, and overt but that prepare
agents at the bottom of a social hierarchy for the tasks of speaking and
acting against the conditions (symbolic and otherwise) in which they find
themselves.
Luckily, these are not the only models available. What I have tried to do
in this chapter is provide an alternate reconstruction of the origins and trajec-
tory of emancipatory political dissent that makes explicit how such forma-
tions are possible in the first place. I have described some of the conditions
that prefigure public collective dissent: the development of off-stage coun-
terpublics and oppositional narratives, values, and interpretations; the stra-
tegic ambiguity of practical noncompliance prior to its emergence as open
insubordination and resistance; and the role of folk culture such as gossip,
myth, and superstition in leveraging local control of resources. Social theory
has been adept at providing both historical analyses and case studies of
the emergence of social movements and the variety of methods they use
to change history. Philosophers investigating emancipatory political dissent
would do well to tune in.
Notes
1 McKinney, “Threats from Below,” unpublished manuscript.
2 Goodin & Brennan (2001).
3 Habermas and McCarthy (1984); Stanley (2016); Fricker (2011); Haslanger (2010).
4 Twelve Angry Men (1957).
5 This is surely when the P voters came to know that not everyone actually agreed
that P, after all!
6 See Tanesini (this volume) for more on this theme.
7 No thinker thinks alone. It’s likely that Q, R, S were circulating (in some protean
form) among not-P voters before their public expression by our “lone” dissenter.
8 A statement like “I accept P but I’m registering my dissent on P” has a Moore’s
Paradoxical flavor—the best interpretation of it is as a move that is either meta-
linguistic or performative: I accept P but nevertheless it ought to be the case that
someone is on the record as disagreeing with P.
9 Lewis (2013); Stalnaker (1999); Stanley (2016); Haslanger (2010).
10 Compare Rorty (1980).
11 See for instance Haslanger: “Implicatures and presuppositions of this sort

become part of the common ground, often in ways that are hard to notice and
hard to combat, and they become the background for our conversations and our
practices. Once the assumption of, e.g., women’s submissive nature has been
inserted into the cultural common ground, it is extremely difficult and disruptive
to dislodge it” (Haslanger, 2010, p. 27).
12 Fricker (2011); Stanley (2016), Haslanger (2010).
13 Stanley (2016, p. 237), emphasis mine.
14 Stanley (2016, pp. 246–248), emphasis mine.
15 Stanley (2016, p. 250), emphasis mine.
16 See also Celikates (2006).
17 Bourdieu and Nice (2015); Haslanger (2007).
18 See Green (2007).
19 See Lessig (1995, 1996). Compare Fraser (1990) on independence of public from
state spheres.
0 See Allen & Light (2015).
2
21 Recall again our bumper sticker: dissent as speaking truth to power.
22 In nondemocratic regimes influence might involve local resistance and pressure
on clerks, landlords, tax collectors, military conscriptors, and the like. And
again, more contentiously in nondemocratic regimes: food riots, pulling down
poor houses, mutiny, arson. See Tilly (2005).
23 See also Sharp (1980); Elizabeth Anderson (2014).
24 Marx and McLellen (2000); Trotsky (1962); Benjamin, Eiland, and Jennings
(2006).
25 It’s critical that such a distinction be principled and not merely stipulative.
Anderson (2014) supplies a useful set of reflexive standards of evaluation for
determining whether (proposed) social changes constitute moral progress. See
also Appiah (2011).
26 recall again my critique of the ‘social discourse object’ picture of consensus.
The idea I would like to convey here is that to understand something like public
opinion one has to understand the nature of a public.
27 Dewey, Hickman, and Alexander (1998); Fraser (1990); Warner (2002); Ander-
son (2016).
28 Warner (2002).
29 Warner (2002), p. 414.
30 Warner (2002), pp. 420–421.
31 The relationship between publics and their discourses is therefore an instance of
the kind of looping effect discussed by Hacking (2004); Haslanger (2013).
32 See, e.g., Youth and Participatory Politics Research Network.
33 This process is called entextualization. See Silverstein (2003).
34 Warner (2002, pp. 423–424). See also Rubin (2012).
35 Note again that such spaces are not always or even usually liberatory. See,
e.g., Nagle (2017).
36 Fraser (1990). See also Felski (1989).
37 Shelby (2016).
38 Nancy Fraser (1990, p. 67).
39 McAdam, Tarrow, & Tilly (2008, p. 5). See also Sharp and Hatfield (1980);
Chenoweth (2013).
40 McAdam et al. (2008).
41 See Gbowee and Mithers (2013); Pray the Devil Back to Hell (2008); See also
Patterson (1996).
42 Scott (1990).
43 See also Eckert & McConnell-Ginet (1992).
44 Again, see also Tanesini (this volume)
45 Scott (1990, Chapters 1 and 2).
46 Scott (1990, Chapter 2); see also Brown & Levinson (2013).
47 Scott (1990, Chapters 6 and 7).
48 For a philosopher of language’s take on plausible deniability, see Camp (2018).
49 Scott (1990, Chapter 6).
50 Scott (1990 p. 153, quoting Raboteau, Slave Religion, p. 245).
51 Further, crucially, all participants know that this is the state of play: the slaves
know that the master is vulnerable, but that open resistance would be met with
likely deadly force. The master knows that he is outnumbered, but that open
revolt is unlikely. There is mutual knowledge among all participants of the risks
associated with overt confrontation.
52 Ong (1987, p. 203).
53 Ong (1987, p. 204).
54 Ong (1987, p. 204).
55 Scott (1990, p. 147).
6 Collins (2008, p. 258); Leeson (2012); Berkes, Colding, & Folke (2000).
5
57 See, e.g., Mansbridge and Morris (2001); Nancy Hartsock (2003).
58 Raboteau, Slave Religion, cited in Scott (1990, p. 115); Yetman (2000) cited in
Scott (1990, p.121).
59 Raboteau, Slave Religion, Chapters. 4, 5, quoted in Scott, 1990, p. 115.
60 For more on the epistemic value of Biblical stories, see Collins (2008, Chapter 7).
61 For instance, slave narratives are replete with lessons from Christian thought
on the dignity of the human person, the righteousness of justice, and themes of
resistance to persecution and bondage. Frederick Douglass, for instance, explic-
itly says he does not think he would have been able to articulate his worth and
dignity without faith in God (Douglass, 2003).
62 Yet, importantly, still firmly rooted in accepted (thereby nonthreatening) Chris-
tian doctrine.
Works Cited
12 Angry men. (1957). Dir. Sidney Lumet. United Artists. DVD.
Allen, D. S., & Light, J. S. (2015). From voice to influence: Understanding citizen-
ship in a digital age. N.p.: University of Chicago. Print.
Anderson, B. R. O’G. (2016). Imagined communities: Reflections on the origin and
spread of nationalism. N.p.: Verso. Print.
Anderson, E. (2014). Social movements, experiments in living, and moral progress:
Case studies from Britain’s abolition of slavery. Lindley Lecture Series, 52, n. pag.
Appiah, A. (2011). The honor code: How moral revolutions happen. N.p.: Norton.
Print.
Benjamin, W., Eiland, H., & Jennings, M. W. (2006). The work of art in the age of
mechanical reproduction. In Walter Benjamin: Selected writings (N. pag.). N.p.:
Belknap. Print.
Bourdieu, P., & Nice, R. (2015). Outline of a theory of practice. Cambridge: Cam-
bridge University Press.
Brown, P., & Levinson, S. C. (2013). Politeness: Some universals in language usage.
Cambridge: Cambridge University Press. Print.
Camp, E. (2018). Insinuation, common ground, and the conversational record. In
M. Moss, D. Harris, & D. Fogel (Eds.), New work on speech acts (N. pag.). N.p.:
Oxford University Press. Print.
Celikates, R. (2006). From critical social theory to a social theory of critique: On the
critique of ideology after the pragmatic turn. Constellations, 13(1), 21–40. Web.
Chenoweth, E. (2013). Why civil resistance works: The strategic logic of nonviolent
conflict. Columbia: Columbia University Press. Print.
politics of empowerment. N.p.: Routledge. Print.
Dewey, J., Hickman, L. A., & Alexander, T. M. (1998). The public and its problems.
In The essential Dewey (N. pag.). N.p.: Indiana University Press. Print.
Douglass, F. (2003). My bondage and my freedom. New Haven, CT: Yale University
Press. Print.
Eckert, P., & Mcconnell-Ginet, S. (1992). Think practically and look locally: Lan-
guage and gender as community-based practice. Annual Review of Anthropology,
21(1), 461–488. Web.
Felski, R. (1989). Beyond feminist aesthetics: Feminist literature and social change.
Cambridge, MA: Harvard University Press. Print.
Fraser, N. (1990). Rethinking the public sphere: A contribution to the critique of
actually existing democracy. Social Text, 25/26, 56. Web.
Gbowee, L., & Mithers, C. 2013). Mighty be our powers: How sisterhood, prayer,
and sex changed a nation at war: A memoir. N.p.: Beast. Print.
Goodin, R., & Brennan, G. (2001). Bargaining over beliefs. Ethics, 111(2), 256–277.
Green, M. S. (2007). Self-expression. N.p.: Clarendon. Print.
Habermas, J., & MacCarthy, T. A. (1984). The theory of communicative action.
N.p.: Beacon. Print.
Hacking, I. (2004). Historical ontology. Cambridge, MA: Harvard University Press.
Print.
Hartsock, N. C. M. (2003). The feminist standpoint: Developing the ground for a
specifically feminist historical materialism. Discovering Reality, 283–310. Web.
Haslanger, S. (2007). “But mom, crop-tops are cute!” Social knowledge, social
structure and ideology critique. Philosophical Issues, 17(1), 70–91. Web.
Haslanger, S. (2010). Ideology, generics, and common ground. Feminist Metaphys-
ics, 179–207. Web.
Haslanger, S. (2013). Resisting reality social construction and social critique.
Oxford: Oxford University Press. Print.
Leeson, P. T. (2012). “God damn”: The law and economics of monastic malediction.
Journal of Law, Economics, and Organization, 30(1), 193–216. Web.
Lessig, L. (1995). The regulation of social meaning. The University of Chicago Law
Review, 62(3), 943. Web.
Lessig, L. (1996). Social meaning and social norms. University of Pennsylvania Law
Review, 144(5), 2181. Web.
Lewis, D. (2013). Convention: A philosophical study. N.p.: Wiley. Print.
Mansbridge, J. J., & Morris, A. D. (2001). Oppositional consciousness: The subjec-
tive roots of social protest. Chicago: University of Chicago. Print.
Marx, K., & McLellan, D. (2000). 18th Brumaire of Louis Bonaparte. In Selected
writings (N. pag.). Oxford: Oxford University Press. Print.
McAdam, D., Tarrow, S., & Tilly, C. (2008). Dynamics of contention. Cambridge:
Cambridge University. Print.
McKinney, R. Threats from Below. Manuscript
Nagle, A. (2017). Kill all normies: Online culture wars from 4chan and tumblr to
trump and the alt-right. N.p.: Zero. Print.
Ong, A. (1987). Spirits of resistance and capitalist discipline: Factory women in
Malaysia (SUNY series in the anthropology of Work). New York: State University
of New York. Print.
Patterson, O. (1996). Freedom. N.p.: Basic. Print.
Pray the Devil Back to Hell. (2008). Dir. Gini Reticker. Fork Films. Web.
Raboteau, A. J. (2004). Slave religion: The invisible institution in the Antebellum
South. Oxford: Oxford University Press, 2004. Print.
Rorty, R. (1980). Philosophy and the mirror of nature. N.p.: Blackwell. Print.
Rubin, G. (2012). Deviations: A Gayle Rubin reader. Raleigh, NC: Duke University
Press.
Scott, J. C. (1990). Domination and the arts of resistance: Hidden transcripts. New
Haven, CT: Yale University.
Sharp, G. (1980). The politics of nonviolent action. N.p.: Porter Sargent.
Sharp, G., & Hatfield, M. O. (1980). Social power and political freedom. N.p.:
Porter Sargent.
Shelby, T. (2016). Impure dissent. In Dark Ghettos. Cambridge: Harvard University
Press.
Silverstein, M. (2003). Indexical order and the dialectics of sociolinguistic life. Lan-
guage & Communication, 23(3–4), 193–229.
Stalnaker, R. C. (1999). Assertion. Context and Content, 78–95. Web.
Stanley, J. (2016). How propaganda works. Princeton, NJ: Princeton University
Press. Print.
Tilly, C. (2005). Popular contention in Great Britain: 1758–1834. N.p.: Paradigm
Publ. Print.
Trotsky, L. (1962). Fascism: What it is, how to fight it. N.p.: Pioneer. Print.
Warner, M. (2002). Publics and counterpublics (abbreviated version). Quarterly
Journal of Speech, 88(4), 413–425. Web.
Yetman, N. R. (2000). Voices from slavery: 100 Authentic slave narratives. N.p.:
Dover Publications. Print.
10 Speaking and Listening to Acts
of Political Dissent
Matthew Chrisman and Graham Hubbs
Introduction: Two Views of Political Speech

When a person voices dissent, she performs a particular sort of speech
act. In this chapter, we consider a specific sort of dissent: political dissent.
These speech acts belong to the class of acts that comprises political speech
more generally. We will say more about what makes a speech act politi-
cal as we proceed, but as a first pass, consider those communicative acts
that are performed in a public context for the purpose of influencing how
we live together in civic society.1 Paradigmatic examples include leading
a rally, testifying before an oversight committee, voting, petitioning, suing
the government, and protesting. These are paradigms of political speech,
in part, because they are acts that are commonly performed with the self-
conscious goal of influencing how we live together in civic society.2 There
are at least two distinct ways to understand the role of these speech acts in
a basically just democracy,3 which we will distinguish by the labels liberal
and Rousseauean.4
First, the liberal understanding: because a just society is one that strives to
peacefully secure and fairly distribute the advantages of mutual cooperation
while protecting the rights of each individual, a common approach amongst
those influenced by, inter alia, Hobbes, Locke, and Rawls, is to conceive
of citizens as autonomous and equal participants in political compromise
about how to live together. Citizens, according to this approach, have a
strong but defeasible obligation to follow the laws of their society, at least
when the laws are basically just. Hobbes argues that this obligation derives
from one’s self-interest; Locke argues that it derives from natural law; in an
early paper, Rawls rather boldly “assume[s], as requiring no argument, that
in a society such as ours, [there is] a moral obligation to obey the law.”5 We
are not concerned here with the source of the obligation; rather, we want
only to note that within the liberal tradition such an obligation is typically
taken to exist. The obligation can hold even if those citizens disagree with
a particular law or view it as unjust. Roughly speaking, this is because the
body of law in force in a society is seen as the result of a basically fair pro-
cedure for compromising between the competing interests of the citizenry,
Speaking and Listening to Acts of Political Dissent 165
and such compromises are necessary for all (or at least most) citizens to
enjoy the vast benefits of social cooperation compared to any alternative
that lacks the rule of law.
Proponents of this liberal approach would not, however, regard the laws
of society as indefeasible or beyond reproach. One of the core purposes of
political society is continued discussion, debate, refinement, and enhance-
ment of the laws, which represent the compromise currently in force for liv-
ing together in society. Given this, political speech might be conceived of as
the means by which freestanding citizens participate in a continual negotia-
tion about which general norms will govern all participants in the collective
so that the benefits of cooperation get divided fairly, given the private inter-
ests of individuals. It’s not only that such negotiation aims to improve the
compromise in force, moving closer and closer to some ideal. The terms of
any previous compromise will need to be updated as the material and tech-
nological situation of the society evolves and the citizenry’s diverse interests
change. Political speech, according to the liberal, is the principal means by
which this improving and updating of the social contract transpires. The
point of such speech from the liberal perspective is for individuals in the
group to have a forum to express their beliefs, desires, preferences, and
goals as a way of getting their voice heard. A right to free speech is a right
to voice one’s position in collective negotiation.6
The Rousseauean will demur at the liberal’s account of the way indi-
viduals are thought to stand to the collective in a polity.7 According to
this approach, citizens are not “freestanding,” for the notion of being an
autonomous participant in social negotiations about how to live together is
dependent on, and not prior to, the notion of a political community. Ver-
sions of this thought are expressed by Rousseau, Hegel, and Habermas; they
seem to be at least part of what Aristotle has in mind when he asserts that
the polis is prior to the individual.8 These philosophers emphasize the ways
in which citizens’ interests are inexorably wrapped up with their place in
society. Because of this, Rousseaueans do not present political speech as the
means by which freestanding citizens with diverse private interests engage
in collective decision making. Good collective decision making is not, on
this view, a matter of everyone pursuing their own private interests within
certain constraints.
According to the Rousseauean view, we achieve our autonomy not by
being protected from and by the majority in our pursuit of our own pri-
vate interests but by participating in civil society. Ideally this gives us what
Rousseau calls “civil freedom,” a kind of freedom that is impossible out-
side of civil society.9 This is the freedom, Rousseau claims, of self-mastery,
freedom from base impulsion, “for impulsion of mere appetite is slavery,
and obedience to the law one has prescribed to oneself is freedom.”10 This
self that prescribes this law is the collective self, the polity to which one
belongs as a part. In this Rousseauean framework, public speech isn’t so
much staking out one’s own private (freestanding) negotiating position but
166 Matthew Chrisman and Graham Hubbs
rather something more like playing one’s role in brainstorming a solution to
a collective action problem from the perspective of the lawmaking, freedom-
constructing collective.
When a people gives itself laws to structure its collective activity, it has
given itself, in Rousseau’s idiom, a general will. This will coordinates and
structures the activity of a collective much as a play-call coordinates and
structures the activity of a basketball team acting as a unified whole. When
a person participates in the formation of a general will, she thinks about
what rule or goal or course of action will benefit the collective, considered
as such, rather than think about what would benefit her, considered as an
individual. If a people gives itself a general will, individuals belonging to the
collective who act on that will do so in a self-interested way, but they do so
out of collective self-interest, not individual self-interest. Again, to act on
the general will is to exercise one’s agency, to overcome the shackles of one’s
individual inclinations to exercise one’s civil freedom.11
In this theoretical context, in contrast with the liberal view, citizens have
an obligation to obey the law in order to be free, insofar as obeying the law
is partly constitutive of having civil freedom and exercising autonomy. Still,
of course, we shouldn’t think of the laws of any society as indefeasible or
beyond reproach. Freedom comes in degrees, and—for the Rousseauean—
one of the core purposes of political society remains continued discussion,
debate, refinement, and enhancement of the laws and so the autonomy of
members of the collective. On this alternative view, however, the law rep-
resents more than the compromise that happens to be currently in force
for living together in society so that each of us privately shares in the ben-
efits of mutual cooperation. Instead, the law represents the collective will of
the society, a will which can and often is changed. Given this, it becomes
more natural to view political speech as the principal means by which the
collective will is expressed and changed. Rather than seeking a negotiated
solution in light of competing private interests, this alternative view treats
political speech as part of continual collective deliberation amongst mutu-
ally dependent citizens about which norms will govern its collective actions.
As Meiklejohn sums up the view of political speech according to this per-
spective: “The point of ultimate interest is not the words of the speakers, but
the mind of the hearers.”12 On this view, any right citizens have to engage in
political speech is grounded in the interest of the hearers; it is not, in Meikle-
john’s words, a right to “unregulated talkativeness.”13
This essay will explore some of the ways these views differ on the norms
that govern politically justifiable dissent in a basically just democracy. In
speaking of an act of dissent being “politically justified,” we mean to abstract
away from the moral quality of the content of what dissenting speakers say.
A major reason to want to understand dissent in a political context is to
figure out what we as a collective should do when we as individuals have
deep moral disagreements with each other about matters that influence how
we live together. Even in contexts with deep moral disagreement, we can
seek an overlapping consensus that forms a conceptual space within which
robust moral issues can be bracketed in order to query matters of common
concern, including the justifiability of acts of dissent. Our notion of politi-
cal justifiability belongs to this space. We will investigate how the liberal
and Rousseauean views differ on the political justifiability of acts of voicing
dissent, and we will also look at their explanations of when and why acts
of dissent give an audience reason to listen to and take seriously what the
dissenters are saying. Although we think there is much to learn from the
liberal and the Rousseauean on these matters, we argue that the conceptual
resources on which each view relies to explain when we should listen to
protest are, in a sense we shall clarify, insufficiently morally neutral. Point-
ing out this common shortcoming will pave the way for what we want to
contribute to the discussion, viz., a speech-act analysis of dissent. The tools
of speech-act theory, we hope, will further enrich our understanding of the
norms surrounding political acts of voicing dissent.
Topic: Voicing Dissent

There are various circumstances in which one might find oneself disagreeing
with one’s government about some issue. When one expresses this disagree-
ment, one voices dissent against one’s government. By “voicing dissent”
we have in mind the multifarious ways one might give voice to this dis-
agreement through political speech. Examples include voting against the
government or its policies, campaigning for the opposition, challenging a
sitting politician, marching in protest, breaking laws in civil disobedience,
and whistleblowing within the government.
These examples are not uniform. Some are more extreme than others.
Some are more constructive than others. Some open one up to legal prosecu-
tion more than others. Within each example, we can also imagine a plethora
of different instances of its particular form of voicing dissent. Voting against
the government in a situation where one also disagrees with the main oppo-
sition has a different character from voting for an opposition in which one
truly believes. Marching with a few fanatics in support of a niche issue feels
different from coordinated marching with millions across the country. Some
ways of whistleblowing responsibly protect government agents whose lives
would otherwise be in danger; other whistleblowing exposes everything to
the Internet.
Any basically fair democracy, we claim, needs to have certain institu-
tionalized channels for competing viewpoints to be voiced and discussed
regularly. These include regular elections, a press that is free to criticize the
government, and civic fora that are explicitly constituted for the purpose of
publicly voicing views intended to influence policy. For any speech act per-
formed in one of these institutions, it will be easy enough to say whether the
act is justified—if the rules are followed, the act is justified. To be a bit more
precise, we should say that such an act would be institutionally justified.
The act might have been morally unjustifiable, or prudentially unjustifiable,
but as long as the rules were followed, the act will be institutionally justified
for playing by the rules. Maybe these sorts of institutions are all that would
be needed for voicing dissent in an ideal democracy. However, real-world
politics illustrates how most existing democracies have alternative, less insti-
tutionalized channels where competing viewpoints are voiced. Civil rights
campaigners boycott busses, students occupy administrative buildings at
their university, famous football players refuse to pledge allegiance to their
nation’s flag, much-followed tweeters go on rants, etc. At the limit, acts of
civil disobedience, where one breaks the law in protest of some law or gov-
ernmental policy, are either minimally or completely non-institutionalized
acts of dissent.14 Such acts, nevertheless, may be politically justifiable even
though they cannot be institutionally justified; sometimes illegal dissent
deserves to be heard.
Although we think that acts of dissent come in varying degrees, we
will focus, for ease of presentation, just on acts that are legally performed
through institutionalized channels and acts of civil disobedience. The next
two sections will treat each of these sorts of acts in turn. In each section,
we will, as indicated in the introduction, explore two normative issues sur-
rounding these act-types. The first issue concerns how acts of dissent may
be politically justified. As we shall see, the liberal view and the Rousseauean
view offer competing accounts of what justifies acts of dissent. The second
issue concerns when and why an act of dissent gives its audience reason to
listen to and take seriously what the dissenters are saying. Here, the liberal
and the Rousseauean differ when it comes to listening to institutionalized
dissent, but they share a general account of when and why civil disobedience
should be heard. We will eventually argue that shortcomings of both views
should lead us to seek an alternative, but let us learn from them first before
criticizing them.
Legal Dissent
To home in on the normative issues at hand, consider the following pair of
cases. Imagine a group of students from the Black Lives Matter movement
who have assembled to voice dissent against the names of certain buildings
on campus. The buildings in question are named after deceased slavehold-
ers, and the students have been granted permission by university adminis-
tration to conduct their protest. Compare this case to one in which a group
of Alt-Right students are granted permission to protest affirmative-action
admissions policies. On many moral views (e.g., those that include a robust
conception of historic justice), one of these groups is voicing dissent against
something worth protesting, and the other is voicing dissent against some-
thing worth preserving. Even if only one of these positions is justifiable,
it nevertheless may be true that both acts of dissent are justifiable. As the
examples have been presented, these acts of voicing dissent are institutionally
justified, because they have been permitted by the relevant authorities. Still,
even in this institutionally justified context, questions remain: first and fore-
most, should a public institution, such as a public university, permit either
or both of these dissenting speech acts?
The liberal view of political speech gestured at earlier would understand
the Black Lives Matter and Alt-Right protests as an attempts to express
positions as part of the ongoing negotiation about the norms by which we
live together in civic society. Liberals would argue that we should respect
the speakers’ autonomy (as a liberal will understand it) and so their right to
their own opinion; moreover, they would argue that we should respect the
speakers’ right to voice that opinion in a public forum with the intention of
influencing policy making. In general, the liberal views civil rights as protec-
tions of the private interests of individuals, especially for cases where these
interests are challenged by what’s expedient for the collective. So liberals
believe in a general right to free speech, one which protects the ability to
advocate for any view about how we should live in society together, no mat-
ter how discriminatory or even anti-liberal the view might be. They do so
precisely because they see political speech as the arena in which competing
viewpoints about how we should live together can be negotiated so that the
resulting view counts as a fair compromise.
By contrast, the Rousseauean view of political dissent gestured at ear-
lier would understand the Black Lives Matter and Alt-Right protests as
attempts to express and to form the general will. According to the Rous-
seauean, these groups have a right to voice their dissent publicly if—but
only if—the opinions and arguments they express are not already on the
table and might plausibly become a part of the general will. If the opinions
and arguments are already on the table, then for the sake of forming the
general will, there is no need to let, e.g., a Black Lives Matter or Alt-Right
protest repeat the case for its side. Knowing whether a speech act is com-
pletely repetitive, however, or whether it is contributing something novel to
the conversation is rarely straightforward. We often are not in a position
to know whether every relevant argument for a view has been put forward,
so there will be a general presumption on the Rousseauean view to allow
speakers to engage in political speech, even if the topic under discussion has
been broadly considered.
There are other reasons, however, that on Rousseauean grounds the
Alt-Right protesters might not have the right to speak. For example, the
Rousseauean might exclude old views that history has already rejected—
e.g., slavery or Nazism. We can think of history as the process of collective
deliberation and settling on these matters. That doesn’t mean that histori-
cally settled topics can’t be reopened, but following the principle of stare
decisis, there is a general presumption against reconsidering views that are
already settled. Also, certain views might be so beyond the pale as not to
merit serious consideration. If the Alt-Right supporter wants to advocate
that everyone opposed to his view give him all of their money and then
willfully jump off of a cliff, the Rousseauean can say that he has no right
to advocate for this view, as it is one whose potential to become part of the
general will is totally implausible.
The liberal thus cleaves tighter than does the Rousseauean to a right to
free speech, which should be protected from government intervention.15
These differences between the liberal and Rousseauean approaches can
show up in which rules come to govern specific institutions. For example, in
the U.S., the Supreme Court has made a number of decisions that interpret
the law in a liberal vein. For example, in Brandenburg v. Ohio, the Court
ruled that unless a speech act could be shown to lead to “imminent lawless
action,” the speaker has a right to express their view. The case concerned a
Ku Klux Klan rally which dissented to the “suppression” of the Caucasian
race and called for a march on Washington. The Court ruled that the rally
was permissible as a just exercise of free speech, thereby clarifying the scope
of the First Amendment. Had the Court reasoned on more Rousseauean
lines, it might have interpreted the case differently. The Rousseauean might
note the historical influence of white racism in fracturing the general will,
pointing out that white racism for centuries resulted in laws for the collec-
tive will of white males, not of the polity as a whole. She might go on to
paraphrase Dr. King, saying that the arc of history bends towards justice,
and racist speech no longer has any place in forming or maintaining the
general will—the matter is settled, and the topic is off the table. No listener
benefits from hearing such speech, so, pace the Supreme Court’s ruling on
the case, no one has the right to express such racist views. An institution
embodying this line of thought would differ from the institution of free
speech as it presently exists in the U.S.
The difference between liberal and Rousseauean here also shows up in
their respective views of when and why an audience has a reason to listen
to and consider dissent.16 For the liberal, the right to speak entails no cor-
responding right to be heard. Indeed, just the opposite: a given member of
the audience has the right to consider or to ignore the speaker’s dissent as
that member chooses. An audience member has a reason to listen to dissent,
according to the liberal, just in case doing so either is or appears relevant to
the individual interests of that member. By contrast, the Rousseauean claims
that an audience member has a reason to listen to any dissent that is relevant
to the project of collective decision making, even if the topic of dissent is
not relevant to that member’s individual interests. The Rousseauean, again,
countenances a right to speak only when this is plausibly viewed as an
attempted expression of the general will. When it is, there is a correspond-
ing right to be heard by the general public. For the Rousseauean, the value
of speech rests in the minds of the hearer, not the mind of the speaker, so if a
given speech act is politically valuable, it has a right to be heard. Because the
speaker of such an act has a right to be heard, the members of the speaker’s
community have a reason to listen to this act of voicing dissent.
Civil Disobedience
Although these views differ on when and why to listen to acts of legal dis-
sent in an institutionalized context, they agree, at a general level, on when
and why to listen to acts of civil disobedience. To see why, let us start by
slightly modifying the cases from the previous section. As in the previous
section, let one case concern Black Lives Matter students who dissent to the
names of certain campus buildings, and let the other concern Alt-Right stu-
dents who dissent to affirmative-action admissions policies. In both cases,
however, let the act of dissent not be a permitted rally but rather the illegal
occupation of the university president’s office. What could justify either of
these illegal acts of dissent?
The liberal and Rousseauean agree that such acts can be justified, and
that when they are, the justification is exculpatory, in that the justification
exculpates citizens for violating their default obligation to follow the law.
Consider the liberal view. Although liberals hold that one has a default right
to free speech, this default ends with illegality. Nevertheless, liberals gener-
ally think some civil disobedience is justified as a form of dissent. However,
they’re going to say that it shouldn’t be promoted, and its protection should
only occur as an exception to the law (e.g., through pardon or judicial dis-
cretion of sentencing) in light of its overall promotion of liberal ideals (espe-
cially rights). The basic idea is that civil disobedience could be justified when
going to the extreme of breaking the law is justified by how bad the object
of one’s dissent is. More precisely, liberals will say that civil disobedience
is justified when it is done in the face of a gross violation of civil rights (or
something of comparable liberal value), when one cannot reasonably expect
this to be corrected through more institutionalized political processes, and
when this dissent will tend to strengthen rather than undermine the overall
rule of law in the society.17
On the Rousseauean view, by contrast, illegal acts of political dissent
are justified if breaking the law reasonably appears the only way for some
neglected possible component of the general will to be considered. More pre-
cisely, the Rousseauean claims that civil disobedience is justified when there
is a “democratic deficit,” i.e., a procedural shortcoming in the formation of
the general will.18 There are many ways that normal democratic processes
can give rise to these deficits, including strategic compromises between large
factions, the marginalization of issues in the name of procedural efficiency,
special interest manipulation of normal deliberative processes, or the “iner-
tia” of matters that have been decided as law. So, to determine whether the
act of civil disobedience is justified, Rousseaueans will ask whether the dis-
senter’s voice, or even simply the message that she would have delivered, is
being or has been illegitimately ignored in “brainstorming” about how to
live together in civil society. This could happen through overt silencing, but
there are other subtler democratic deficits (e.g., campaign laws that prevent
some candidates from getting on the ballot, voting systems that corner a
perpetual minority, corporate control of media). These deficits provide plau-
sible grounds for some paradigm examples of civil disobedience (e.g., the
anti-global WTO protests).
The views differ, then, on what grounds the justification of illegal acts of
voicing dissent. They agree, however, on the following: whenever a speaker
is justified in illegally voicing dissent, the members of the speaker’s com-
munity have a reason to listen to what the dissenter is saying. On the liberal
view, if a rights violation is severe enough to justify civil disobedience, then
that fact gives members of the relevant community a reason to listen to those
who protest about the rights violation. Similarly, on the Rousseauean view,
if a democratic deficit is gross enough to justify civil disobedience, then that
fact gives the members of the relevant community a reason to listen to the
content of that dissent. In spite of their differences regarding the underly-
ing justification of civil disobedience, both views treat civil disobedience as
exceptional and thus in need of a special justification. If that justification
can be provided, that fact gives the audiences of these illegal acts of voicing
dissent a reason to listen to and consider what the dissenter is saying.
We think that studying the liberal and Rousseauean approaches to voic-
ing dissent provides valuable insights into the normative issues surround-
ing both institutionalized dissent and civil disobedience. These views falter,
however, by relying on substantive normative views to explain when audi-
ences have a reason to listen to and consider what dissenters are saying. To
see how the liberal view falters, consider again the activities of the Black
Lives Matter protesters who occupy the university president’s office. A sym-
pathizer might argue that we should listen to illegally voiced dissent in this
case because these protestors are reacting to the gross historical injustice of
slavery and that, moreover, they would be reasonable to expect that this will
not be corrected through more institutionalized political processes and that
this action would tend to strengthen rather than to undermine the overall
rule of law. But is that right? We suspect that those on the other side of the
debate would say that the names of university buildings are no matter of
consequence and that the protesters should just get over it and quit com-
plaining about slavery. With no agreement about whether there is a serious
rights violation, there can be no agreement about whether the audience has
a reason to listen to the dissenters.
A similar disagreement is predictable between the Alt-Right dissenters and
those who disagree with them. The Alt-Right dissenters claim that the rights
of whites to equal opportunity are violated by affirmative-action policies;
their opponents respond that this is a misunderstanding of equal opportu-
nity, and even if there are such rights violations, these are far outweighed by
the historical injustice of slavery, including its long-term consequences. So,
should the audience listen to and consider what these dissenters are saying?
We think the liberal account is going to have a hard time adjudicating on
these issues without deploying a theory of rights that already has the answer
“baked in.” A liberal justification of one or the other examples of illegal dis-
sent won’t have any prospect of convincing someone on the other side of the
debate that the act is politically justifiable and so should be listened to. If the
disagreements are ultimately about whether there has been a gross violation
of rights, appealing to rights in the justification of illegal dissent is likely to
be question-begging or bump-in-the-rug displacing.
The Rousseauean is no better off than the liberal on this score, for those
who disagree about whether a policy merits illegal dissent will also tend to
disagree about whether there has been a democratic deficit. Returning to
the examples: a Rousseauean sympathizer might argue that these protestors
should be listened to because they are giving voice to marginalized points
of view in our collective brainstorming about how to live together. In either
case, the Rousseauean justification would go, the act of civil disobedience is
worth listening to because it expresses something important to the forma-
tion and maintenance of the general will. But a critic in each case may simply
deny that there is any “democratic deficit” in the sense relevant to justifying
dissent. Against the Black Lives Matter protesters, an opponent might say
that the names of buildings are too trivial a matter to the general will to jus-
tify breaking the law; against the Alt-Right protesters, an opponent might
say that white people already have a strong, well-heard voice, so their claim
that their voice hasn’t been heard is unjustifiable. A Rousseauean justifica-
tion of one or the other examples of illegal dissent won’t have any prospect
of convincing an opponent that the act is politically justifiable and thus one
that should be listened to and taken seriously by the relevant audience. If
the disagreements are ultimately about whether there is a democratic deficit,
then, in a manner that parallels the problem on the liberal view of appealing
to rights, appealing to democratic deficits to justify an act of illegal dissent
is likely to be question-begging or bump-in-the-rug displacing.
The common problem here, we think, is that both views about political
speech are committed to relaxing the abstraction from content too soon
to provide a satisfying treatment of dissent in a context of robust moral
disagreement. If a theory of political speech is going to make sense of when
there is reason to listen to dissenters and consider what they are saying in
the context of such disagreement, then it is going to have to carve out a
space where those who disagree can still recognize and respect each other’s
rights to voice their views and have them taken seriously by the relevant
audience as part of our collective decision making. The liberal and Rous-
seauean accounts fail to carve out this space, for each attempts to settle
justificatory issues by appealing to its preferred robust moral view (about
rights or democratic deficits). In many cases of disagreement, this collapses
the space where those who disagree about the relevant robust moral issues
can still recognize and respect each other’s rights to voice their views. To
preserve this space, a theory of dissent needs to articulate some normative
constraint on political speech that does not turn on robust moral views too
soon in the chain of justification.
We say “too soon” because ultimately the ideas that people have rights
and that collective decisions should be made democratically are moral
views, as are the ideas that people deserve respect and that peaceful discus-
sion of differences is preferable to violent exercises of power. Weak moral
assumptions like these are going to be in the background of any justifica-
tion of dissent in the context of political philosophy, and so every view
about the justifiability of various forms of dissent is going to be ultimately
grounded in morality. However, by making the question of justifiability of
illegal dissent turn on who has which rights, a liberal justification will tend
to lose many who would agree with these weak moral assumptions but still
disagree about who has the relevant rights. Similarly, by making the ques-
tion of justifiability of illegal dissent turn on what counts as a democratic
deficit, a Rousseauean justification will tend to lose many who would agree
with these weak moral assumptions but still disagree about when there
has been a democratic deficit. This is why we think the liberal and Rous-
seauean views about political speech fail to provide theoretical tools that
are neutral enough to make sense of the political justification of dissent in
the face of robust moral disagreement. None of this means that the liberal
and Rousseauean views don’t illuminate particular cases of voicing dissent:
we just want some more neutral tools for articulating a general normative
constraint on political speech that does not turn as soon on robust moral
views in the chain of justification.
The Speech-Act View

We think speech-act theory can supply these tools. To apply speech-act
theory to voicing dissent, note first that all verbal acts of dissent have an
evaluative element and that most have a corresponding prescriptive element.
All verbal acts of dissent evaluate something as bad or wrong in some way,
and most correspondingly demand change to rectify the badness or wrong-
ness in question.19 Focus here on the standard case, in which disapproval
is expressed and some corresponding change is demanded. For any such
speech act, we want to suggest that sincerity in disapproval and good faith
in making the demand are two of its felicity conditions. This means that
one engaging in dissenting political speech should sincerely disapprove of
that to which they dissent, and the way they demand change should reflect
a good faith commitment to the norms on which these changes are based.
(These are not the only felicity conditions we acknowledge—more on this
as follows.) That is to say, the norms “be sincere” and “demand in good
faith” are partly constitutive of political dissent being the sort of speech act
that it is. In something approaching a slogan, if you’re going to say contrar-
ian things in the context of negotiating the social contract, then you should
genuinely disapprove of what you’re against and be willing to try to live up
to the alternative norms you are for.
To repeat, we’re not claiming that a speech act counts as political dissent
just in case it is sincere and done in good faith. Rather, we’re claiming that
these standards provide constitutive norms for the speech act of political
dissent; if these norms are flouted, the result is, in Austin’s terminology, an
abuse.20 To get clearer on the way in which these norms can be thought of as
felicity conditions, let us consider, for the sake of comparison, what it means
to say that truth is a constitutive norm of assertion that partially determines
its felicity conditions. An assertion can be bad—that is, infelicitous—if it is
difficult to understand as seriously aiming at the truth. If, for example, a
speaker asserts “I am a pumpkin,” the speech act is bad not only because
it is false, but also because it is hard to take seriously as even aiming at the
truth. Speech acts of dissent can go similarly bad, we claim, if they are car-
ried out in ways that indicate either insincerity or bad faith. Suppose, for
example, that someone in a town hall meeting proposes to enforce rules on
how long one may speak at the meeting, arguing that a standing but usually
ignored one-minute-per-speech-rule should be enforced, as meetings are tak-
ing too long. If it turns out the dissenter dissents not because she really dis-
approves of longer slots for each speaker but because someone else paid her
to say this, we’d think her insincere. Moreover, that’d make her dissenting
speech act bad qua political speech in an important way: it would at least
partially undermine the audience’s reason for listening to the speaker.21 Simi-
larly, if she were to demand that the rules be enforced by giving a ten-minute
speech, it would become difficult to take her to be seriously engaged in good
faith political discussion, as she is actively violating the change in policy she
purports to be demanding. Unless she is so absent-minded that she com-
pletely loses track of time, she will seem hypocritical. On our analysis, this
makes the person’s speech a bad qua act of voicing dissent, not because of
the particular policy that is being advocated for, or just because the dissent
is unlikely to affect change. It is a bad speech act because it is conducted in
a way that is self-undermining and so is infelicitous.
So far, this speech-act analysis of dissent may seem fairly anemic. Sure,
political dissent should—because of the sort of speech act it is—be sincere
and in good faith. But that is arguably true of most evaluative and prescrip-
tive speech performed in a cooperative context. What does this have to do
with the political justifiability of dissent and when audiences have political
reason to listen to what dissenters are saying?
We think it is already interesting to notice that indications of insincerity
and bad faith would serve to undermine dissenters’ claim to any right to be
listened to by their audience. That provides a kind of positive guidance to
would-be dissenters: if you want your dissent to be heard and taken seri-
ously by your political community, aim to dissent in such a way that avoids
any suspicion of insincerity or bad faith. However, we also think there is a
third constitutive norm on the speech act of political dissent that brings us
closer to the distinctively political realm. We also want to suggest that acts
of political dissent, because of the kind of speech act it is, should be based
(at least implicitly, but recognizably) on considerations of justice.
We intend “justice” to be understood broadly and generically here: fol-
lowing Aristotle, we do not think of justice merely as distributive justice, but
also as rectificatory, reciprocal, and political.22 If the audience of a speech
act of political dissent can’t tell that the speaker is addressing, at least in
part, what he thinks is and/or isn’t just, then the audience will have a hard
time interpreting the speaker as engaged in political dissent. Of course, the
speaker’s reasons may be based in justice even if his audience can’t see that,
and of course, a speaker and his audience may disagree on whether a pur-
ported injustice is, in fact, an injustice. However, if it is not clear that justice
is at issue at all, a speech act of voicing dissent will likely be seen as a sort
of personal complaining—it is not likely to be comprehended as the sort of
political dissent that has been our topic here. Attempts to engage in political
dissent that do not recognizably engage considerations of justice will, we
would expect, be infelicitous.
Having noted this third felicity condition on political speech, let us apply
the speech-act view to an analysis of political dissent. The view allows for an
account of the justifiability of dissent both in legal institutionalized contexts
and for illegal civil disobedience, an account that does not, as the liberal and
Rousseauean accounts do, bring in substantive moral views too soon. Those
on opposing sides of the political spectrum may disagree whether the name
of a building constitutes a substantive rights violation or whether there is a
democratic deficit in marginalizing the voice of whites, but they might still
agree that self-undermining speech acts of dissent fail to be good political
speech. They can agree that whatever the dissenters are saying, if they are
saying it in a way that comes across as insincere, in bad faith, or completely
unconcerned about justice, then it will be hard for an audience to recognize
it as political speech worth listening to. We are not claiming here that rights
and democratic deficits are irrelevant to justifying dissent. Rather, we think
that one gets to appeal to those only (i) if both sides of a disagreement agree
about them, or (ii) if more neutral considerations attaching to the speech act
as such haven’t settled the issue about the justifiability of the dissent. Many
interesting real-world cases won’t be like (i), at least not in the moment. And
when we get to (ii), we’ve basically given up hope on convincing those with
whom we disagree—and we shouldn’t do that if we’re still engaged in politi-
cal speech with them (which is consistent with thinking that some people
are beyond the pale).
We also think our speech-act view provides a more basic and—for that
reason—more attractive view of when an audience has a reason to listen to
and consider an act of dissent. If dissent is voiced in a sincere, good faith
way and appeals to recognizable considerations of justice, that gives the
community members to which the dissenters belong a reason to listen to and
consider what the dissenters are saying. To return to our examples: if the
Black Lives Matter students are sincere, protest in good faith, and base their
complaints in recognizable considerations of justice, the university commu-
nity has a reason to listen to them. The same holds for the Alt-Right protest-
ers. Now, much will turn on how one fills in the details of “appealing to a
recognizable conception of justice,” but we think this generic consideration
allows us to catch the variety of cases in which there is reason to listen to
dissent, a variety that is broader, we think, than the more specific liberal and
Rousseauean views can capture.
Moreover, we think the generality of the speech-act view gives it diagnos-
tic potential that the liberal and Rousseauean views lack. Unlike the other
views, ours is equipped to explain why an act of dissent fails when it does
so. Some people fail to listen to Black Lives Matter protesters because they
are unconvinced that the injustices voiced in their dissent are genuine injus-
tices. Others fail to listen to Alt-Right protesters because they hypocritically
rely on the government programs they decry. We are not claiming that every
failure of dissent can be explained by the norms we have outlined here, but
we do not think that counts against our approach—rather, it points to ways
to develop it. We hypothesize that when an act of voicing dissent fails qua
speech act, it usually does so because some felicity condition, which can be
formulated as a constitutive norm, has been flouted. For failures that cannot
be explained by the norms we have presented, we predict there is some fur-
ther felicity condition and further norm that, when added to the speech-act
view of voicing dissent, will explain the failure. Proceeding in this way, we
can enrich our understanding of the speech act of dissent.
Conclusion: Reflections on Silencing

Our essay has examined some of the political justifications for voicing dissent
and some of the reasons one might have for listening to acts of dissent. In
the last section, we drew upon the resources of speech-act theory to provide
an account of the norms surrounding voicing dissent that is, we think, more
basic than what the liberal or Rousseauean has to offer. We will close with
some reflections on what our view has to say about acts that silence speech.23
By “silencing,” we mean acts that make it very difficult, if not impossible,
to hear what a speaker intends to say. The limiting cases are those in which
a speaker is literally silenced, by, e.g., being knocked unconscious or by
being gagged, but a speaker may be silenced even though she can still speak,
perhaps even loudly. If someone who wants to hear the speaker cannot do
so because of the obstructive actions of others, those others, as we intend
the notion, have effectively silenced the speaker. Of interest to the present
discussion are cases in which a speaker is silenced by the speech acts of oth-
ers; we shall call this phenomenon shouting down. What, if anything, does
our view have to say about silencing that is accomplished by shouting down
a speaker? On our view, does this ever count as political dissent?
Based on the arguments of the previous section, if a speaker speaks sin-
cerely and in good faith, and if she recognizably appeals to considerations of
justice, her audience has a reason to listen to her. If her audience has a rea-
son to listen to her, then they have a reason not to silence her, and so not to
shout her down. This does not mean that the audience cannot be justified in
dissenting to her, but there is a difference between an audience voicing dis-
sent and a crowd silencing with noise. If a crowd uses the amplitude of voice
to shout down a speaker, it is not voicing dissent. The content of the crowd’s
speech is merely incidental to the goal being pursued; the crowd may just
as easily achieve its goals by making non-speech noise, e.g., by blowing air
horns. To defend acts of shouting down by appealing to free speech, we
think, is misguided, or disingenuous, or cynical, but not justifiable.
This is not to say that there cannot be other justifications for silencing
speakers. The Antifa movement, for example, argues that the pernicious
effects of allowing fascist speech sufficiently outweigh any considerations—
including liberal and Rousseauean—to the contrary. There is thus every rea-
son to toe the line, to give “not one inch,” as they say, to the views of fascist
speakers.24 Our view does not contradict the Antifa position. We claim that
the Antifas have a reason to listen to fascist speech, if that speech is sincere,
made in good faith, and recognizably appeals to considerations of justice,
but that does not preclude that reason being overruled by competing con-
siderations. One such consideration might be the pernicious effects of such
speech (although to make this argument compelling, the Antifa will need to
persuade her opponent that these effects outweigh the negative effects both
of silencing fascist speech and of silencing speech in general). Another might
be a Rousseauean argument mentioned above, viz., that the principle of
stare decisis should be applied to fascist views, so, having been historically
repudiated, they have no place in discussions of the general will. These are
competing, potentially decisive reasons to silence fascists, but they outweigh
the reason to listen to the fascist—they do not dissolve that reason.
That is, however, only if the fascist speaks in a way that is sincere and
in good faith and that recognizably appeals to considerations of justice. If
fascism itself simply cannot appeal to any such conception, because it is in
its nature an unjust ideology, then there is no reason on our view to listen
to the fascist. If there is no such reason, our view does not oppose silencing
fascists. The possibility or lack thereof of just fascism is beyond the scope of
this essay, even if, sadly, it presently merits at least as much attention as the
norms surrounding verbal acts of dissent.
Notes
1 In his contribution to this volume, Michael Lynch defines political speech as
speech aimed at influencing policy. We suspect that this is simply a different way
of characterizing speech that is aimed at influencing how we live together in civic
society. By focusing on policy, Lynch draws attention to the general rules and
norms that organize civil society; we think that any speech that aims to influence
collective living will have implications for general policies.
2 A speech act can be political, however, even when it is done primarily for other
reasons. A newspaper, for instance, might publish stories primarily because
they are of interest to its readership and will sell copies, not explicitly because
it wants to influence how we live together in civil society. Nevertheless,
many editorial decisions, it seems to us, count as political speech of a more
subtle sort.
3 By “basically just democracy,” we mean, roughly, a society with a relatively
stable rule of law and equality opportunity for all citizens to vote for legislation
and/or representatives in free and fair elections. We don’t mean a society that is
mostly just, as we take it that even basically just democracies can harbor serious
and systemic injustices.
4 This distinction is inspired by Daniel Markovits (2005). Markovits uses the term
“republican” to label what we are calling the Rousseauean view. In contempo-
rary political philosophy, “republican” may call to mind Philip Pettit’s account
of non-domination, which is not obviously Markovits’s or our topic. We hope
that calling the view “Rousseauean” will prevent any confusion with Pettit’s
view. Also, we do not claim that these are the only two ways of understanding
political speech in a basically just democracy, but they are the two that will
interest us here.
5 See, respectively, Hobbes (1994), ch. 1–13, Rawls (1964), “Legal Obligation and
the Duty of Fair Play,” in S. Hook (ed.), Law and Philosophy (New York: New
York University Press): 3–18 (the quote is from p. 3). To be sure, Hobbes also
thinks there are laws of nature, and self-interest underwrites Locke’s conception
of natural law, but their emphases are sufficiently different to warrant the char-
acterizations above.
6 We use “right” here as non-commitally as is possible; we take no stand on the
metaphysics of rights. For worries about such non-committal usage of the term,
see O’Neill, Onora (2009).
7 As is common when one draws on an historical figure as the basis for a view, we
do not claim here to be representing Rousseau’s precise view, nor do we claim
that ours is the only plausible interpretation of his view. We are, however, partial
to the reading of Rousseau that Joshua Cohen offers in Rousseau: A Free Com-
munity of Equals (2010). When we do draw on the historical Rousseau, we do
so from Victor Gourevitch’s translation of The Social Contract (1997).
8 Aristotle makes this remark at Politics I.2.1253a19 and I.2.1253a26. For Hegel’s
view, see Elements of the Philosophy of Right, (1820/1991); for Habermas’s,
(1981/1984).
9 Social Contract, I.8.2.
10 Social Contract, I.8.3.
11 For more on the idea of the general will and its application to contemporary
political issues, see Hubbs, Graham, “Transparency, Corruption, and Demo-
cratic Institutions,” Les ateliers de léthique / The Ethics Forum 9(1) (2014):
65–83.
12 Meiklejohn, Alexander (1948): 28.
13 Meiklejohn, Alexander (1948): 28.
14 It won’t matter for our purposes whether civil disobedience is understood as
non-institutionalized, because it involves law-breaking, or whether it is under-
stood as minimally institutionalized, because at least some cases are recog-
nized as democratically justifiable means of pursuing political objectives. Also,
we understand civil disobedience as anti-revolutionary, in the following sense:
whereas revolution aims at the total replacement of existing governmental insti-
tutions, civil disobedience seeks to reform certain institutions while preserving
most of a government’s core institutions. We suspect that the distinction between
revolution and civil disobedience is a matter of degree, not kind. We don’t think
our arguments turn on any specific view of civil disobedience, so these rough
remarks should suffice for our purposes.
15 The liberal view is thus compatible with the “more speech” doctrine popularized
by Justice Brandeis in Whitney v. California. For a critique of this doctrine, see
Mary Kate McGowan’s contribution to this volume.
16 We aim here to take no substantive view on the nature of reasons. For a review
of the various views one might take, see Weaver, Bryan and Kevin Scharp
(forthcoming).
17 The locus classicus for this view of civil disobedience is Rawls (1999): 319–343.
Other liberal views include Dworkin, Ronald (1999): 104–116, and Lefkowitz,
David (2007): 202–233. Kimberlee Brownlee’s account in Conscience and Con-
viction: The Case for Civil Disobedience (2012) also belongs to this tradition;
see esp. chs. 1–4. Brownlee advises that we understand justifications for civil
disobedience as the sort of thing a defense attorney might give; on this, see chap-
ters 5–6 of Conscience and Conviction.
18 The source of this idea is Markovitz; see his “Democratic Disobedience.” Wil-
liam Smith argues for a deliberative democratic view of civil disobedience that
has anti-liberal affinities with Markovits’s position; see his Civil Disobedience
(2011): 145–166.
19 Note, then, that some acts of direct action will fall outside of the scope of our
analysis. Consider an animal rights activist who breaks into a psychology labo-
ratory and liberates the lab mice. This activist is not demanding change; he is
enacting change directly. His act is not primarily communicative; indeed, if he
leaves no trace behind, the public may be left to guess at his motive, in which
case his act has no obvious assertive element. We have no gripe with anyone who
wants to characterize direct action as dissent, but we will presently bracket the
phenomenon.
20 There is another kind of speech-act norm we won’t discuss here. The relevant
norms do not determine good and bad instances of their corresponding acts but
enable something to count as a given act in the first place. For example, to marry
someone in the Catholic Church, one must be a priest and the people being
married must be male and female. It’s not that a Catholic marriage violating
these conditions is somehow bad or not as it ought to be qua Catholic marriage;
rather, it’s not a Catholic marriage at all. Austin labels these sorts of infelicities
“misfires” (1962): 16.
21 Explaining precisely why this reason would be undermined is a complex matter.
For more on this, see Richard Moran’s work on testimony (2005) and (2013):
115–135.
22 Aristotle discusses these sorts of justice in Nichomachean Ethics, V.1–5.
23 Our topic here is different from the “eloquent silences” discussed in Alessandra
Tanesini’s contribution to this volume.
24 For more on the contemporary Antifa movement, see Mark Bray (2017).
References
Aristotle. (1998). Politics (C. D. C. Reeve, Ed.). Indianapolis, IN: Hackett.
Austin, J. L. (1962). How to do things with words. Cambridge, MA: Harvard Uni-
versity Press.
Bray, M. (2017). Antifa: The anti-fascist handbook. Brooklyn, NY: Melville House.
Brownlee, K. (2012). Conscience and conviction: The case for civil disobedience.
New York: Oxford University Press.
Cohen, J. (2010). Rousseau: A free community of equals. New York: Oxford Uni-
versity Press.
Dworkin, R. (1999). A matter of principle. Cambridge, MA: Harvard University Press.
Habermas, J. (1981). The theory of communicative action, vol. I: Reason and the
rationalization of society (T. McCarthy, Ed.). Boston, MA: Beacon Press.
Hegel, G. W. F. (1991). Elements of the philosophy of right (H. B. Nisbet, Ed.). New
York: Cambridge University Press.
Hobbes, T. (1994). Leviathan (E. Curley, Ed.). Indianapolis, IN: Hackett.
Hubbs, G. (2014). Transparency, corruption, and democratic institutions. Les Ate-
liers de L’éthique, 9(1), 65–83. https://doi.org/10.7202/1024295ar
Lefkowitz, D. (2007). On a moral right to civil disobedience on a moral right to civil
disobedience. Ethics, 117(2), 202–233. https://doi.org/10.1086/510694
Locke, J. (1980). Second treatise of government (C. B. Macpherson, Ed.). India-
napolis, IN: Hackett.
Markovits, D. (2005). Democratic disobedience. Yale Law Journal, 114(8), 1897–
1952.
Meiklejohn, A. (1948). Free speech and its relation to self-government. New York:
Harper Brothers Publishers.
Moran, R. (2005). Getting told and being believed. Philosophers’ Imprint, 5(5),
1–29. Retrieved from https://quod.lib.umich.edu/p/phimp/3521354.0005.005/1
Moran, R. (2013). Testimony, illocution and the second person. Aristotelian Soci-
ety Supplementary Volume, 87(1), 115–135. https://doi.org/10.1111/j.1467-
8349.2013.00222.x
O’Neill, O. (2009, October). The dark side of human rights. Contemporary Debates
in Political Philosophy, 1, 423–436. https://doi.org/10.1002/9781444310399.ch23
Rawls, J. (1964). Legal obligation and the duty of fair play. In S. Hook (Ed.), Law
and philosophy (pp. 3–18). New York: New York University Press.
Rawls, J. (1999). A theory of justice (Revised). Cambridge, MA: Harvard University
Press.
Rousseau, J.-J. (1997). The social contract and other later political writings
(V. Gourevitch, Ed.). New York: Cambridge University Press.
Smith, W. (2011). Civil disobedience and the public sphere. Journal of Political Phi-
losophy, 19(2), 145–166. https://doi.org/10.1111/j.1467-9760.2010.00365.x
Weaver, B., & Scharp, K. (forthcoming). Reasons and semantics. New York: Oxford
University Press.
11 Responding to Harmful Speech
The More Speech Response,
Counter Speech, and the
Complexity of Language Use1
Mary Kate McGowan
Speech can be harmful. It can harass, demean, and incite. It can inflict
emotional and psychological harm and it can cause others to participate
in harmful practices. A widespread view in the free speech literature is that
the proper response to harmful speech is more speech. The idea is that
rather than regulate the harmful speech, we ought to remedy the harms of
the harmful speech with further speech. Speech, not speech regulation, is
alleged to be the answer. This has come to be known as the “more speech”
response. In this chapter, I will investigate this response. I will clarify which
sort of “more speech response” concerns me and I will argue, using work in
the philosophy of language and the kinematics of conversation, that things
are considerably more complex than the “more speech response” appears
to assume. I am concerned with the complexities associated with speech
that allegedly undo or prevent the effects of other speech; I am not here
concerned with issues of regulation.2
The chapter proceeds as follows. In § 1, I motivate and clarify the “more
speech” response. Then in § 2, I show that speech can cause a variety of
different harms and it can do so in a variety of different ways. In § 3, the
scorekeeping framework for tracking conversational context is presented
and distinguished from the common ground framework. Reasons are also
given for preferring the scorekeeping framework. In § 4, I show that counter
speech cannot simply undo the harmful speech. Complete reversals are not
an option. In § 5, I argue that counter speech can be risky and it can be risky
in a variety of ways. Then, in § 6, I argue that saying nothing can count as
doing something and it can also do something harmful. Finally, in § 7, the
good news: I argue that it is possible for speech, even speech by a relatively
powerless person, to have profound positive effects.
The “More Speech” Response

It is a widespread view in the free speech literature that the proper remedy
for harmful speech is “more speech.” Speech itself is alleged to be the best
way to remedy the harms of harmful speech. The most famous and most
Responding to Harmful Speech 183
cited example of this view is that expressed by Justice Brandeis in Whitney
v. California where he said:
If there be time to expose through discussion the falsehood and falla-

cies, to avert the evil by the processes of education, the remedy to be
applied is more speech, not enforced silence.3
Although Brandeis is here concerned with how the law ought to respond
to speech that might incite violence against the state, the “more speech”
response, that his words are taken to support, has taken on a life of its
own and it is often treated as a general remedy for various kinds of harm-
ful speech. In particular, it is claimed that more speech, or counter speech,
is the proper remedy for face-to-face instances of allegedly harmful speech
(e.g., racist hate speech or sexist jokes). In what follows, I shall focus on
this face-to-face version of the more speech response. Again, I am not here
concerned with regulation.
Technically, “more speech” can take many forms. It can take the form
of a general education campaign. One can inform the population about a
certain topic so that people are more or less receptive to certain kinds of
messages or utterances. Here, however, I shall focus on counter speech in
conversational contexts. I am concerned with direct and fairly immediate
responses to allegedly harmful utterances that are made by the addressee,
a hearer, or even an over-hearer. A person yells racial epithets to a women
riding her bike and this face-to-face version of the “more speech” response
has it that the proper remedy for this sort of harmful utterance is for the
cyclist or a bystander to respond with speech that blocks, objects, mitigates,
reverses, or somehow undoes the harms of this use of the racist epithet.
A man tells a sexist joke to a group of colleagues and the face-to-face ver-
sion of the “more speech” response has it that the proper remedy is a fairly
immediate conversational counter-move.
Criticisms of the “More Speech” Response

This “more speech” response has been criticized in a number of ways. First,
it has been pointed out that it places the burden of the remedy on those
targeted (and potentially harmed) by the allegedly harmful speech.4 This
is unfair. Moreover, the targets of such speech are typically members of
already marginalized groups.5
Second, if speech can be disabled and if targets of harmful speech have
their speech disabled, then this undermines the efficacy of (not just the fair-
ness of) the more speech response. To lend credence to this possibility, it
is important to point out that speech can be disabled in a wide variety of
ways. One form of speech disablement (often called silencing) involves rec-
ognition failure during attempted communication. If there are systematic
184 Mary Kate McGowan
reasons for the recognition failure in question, then the speaker is silenced
and her attempted communication is thwarted.
Suppose that an addressee fails to recognize what sort of speech act a
speaker is intending to perform. Hornsby and Langton argue that when
this happens, the speaker is silenced and her speech act fails.6 I have argued
elsewhere that there are other sorts of recognition failures (e.g., failure to
recognize that the speaker has and is exercising the requisite authority to
perform the speech act in question, failure to recognize that the speaker is
sincere) that can constitute other types of silencing.7
Silencing need not involve recognition failure; it can involve group-based
credibility deficits. If a speaker is (even sometimes) given less credibility than
the speaker ought to be given and if this happens because the speaker is a
member of a socially marginalized group, then the speaker is silenced. Such
testimonial quieting (also called testimonial injustice) systematically under-
mines the epistemic impact of these speakers.8
Speech can also be disabled when the shared conceptual resources are
impoverished. In extreme cases, a person cannot make sense of her own
experience. Before the concept of hostile work environment sexual harass-
ment was a socially shared concept; for example, women who experienced
such harassment had difficulty recognizing it as such.9 As feminist philoso-
phers of language have stressed, the meaning of terms and the concepts
available in a language tend to reflect the interests and experiences of those
in power.10 This means that language is ill equipped to express the experi-
ences and interests of marginalized persons. Such hermeneutical injustice is
a matter of degree.11 In less extreme cases, marginalized persons communi-
cate perfectly well with one another but face challenges in communicating
their experiences with members of the privileged group.12 Such limitations
in socially shared linguistic meaning can certainly undermine the efficacy of
the “more speech” response.
Finally, not all forms of silencing actually involve speech. Sometimes a
speaker decides against speaking and thus self-silences. Whether a decision
to refrain from speaking counts as silencing depends on the reasons for
doing so.13 If a speaker decides to remain silent because she knows that
her audience will get something wrong and harm will result, then this is
a case of what Dotson calls testimonial smothering and it is a form of self
silencing.14 If harm prevention is a reason against speaking for marginalized
persons, then the phenomenon of testimonial smothering also undermines
the efficacy of the “more speech” response.
Third, empirical work shows that targets of hostile speech in public places
tend not to respond to it.15 Targets of such speech offer different reasons for
remaining silent. Many expressed concern for their safety; others refused to
take on responsibility for the speaker’s ignorance and many others viewed
such counter speech as a pointless waste of energy.16 That, as a matter of
fact, targets of hostile public speech tend to refrain from responding cer-
tainly tells against counter speech as an effective remedy.
Fourth, the “more speech” response appears to make other false empiri-
cal assumptions. For starters, it appears to assume a level playing field for
all speakers but, as we all know, who is able to do what with her words is
a function of social power. Some speech acts require the exercise of speaker
authority; who has an audience, is recognized as credible, has perceived
expertise, or the ability to disseminate her views are all also strongly affected
by one’s social position. By assuming a level playing field then, the “more
speech” response overlooks the crucial role of social power in effective lan-
guage use.
Fifth and relatedly, the “more speech” response operates with a false view
of language use; it seems to completely overlook the fact that language use
is a social practice and it overlooks the important fact that language use is a
social practice deeply embedded within other social practices. My primary
focus in this chapter is to elaborate on the importance of the latter claim, but
first I shall say a bit more about the former claim.
Different versions of the “more speech” response operate with different
assumptions about language use. One such view of language use might be
called the mini-philosophy-paper view of speech. On this view, all speech
acts are expressions of a view along perhaps with explicitly stated reasons
in support of that view. Some have called this the argument view of speech.17
This mini-philosophy-paper view of language use is false. Hate speech is
not a mere assertion of a point of view; it is not an invitation to discuss mat-
ters of social justice. It is an act of demeaning a person and doing so in vir-
tue of that person’s group membership. As Austin has so effectively shown,
language is not merely in the business of making true or false claims about
the world.18 It also does things. Speech can be used to perform illocutionary
acts (e.g., promising, scolding, apologizing) and it has causal consequences
(e.g., persuading, exciting, scaring) in virtue of functioning as speech. As
I shall argue later in the chapter, though, there is considerably more com-
plexity to even this Austinian picture of language use.
There is also a version of the “more speech” response that operates with
a different and less naïve view of language use. On this version, it is recog-
nized that speech does considerably more than merely express content, but
it is tacitly assumed that whatever speech can do counter speech can undo.
As we shall see, in § 4, however, even this is false.
In the meantime, I turn to the task of identifying various ways that speech
can bring about harm. There are many different potential connections
between speech and its associated harms. Moreover, recognizing the various
options will highlight some of the all too often overlooked complexities of
language use.
Ways That Speech Can Cause Harm

Speech can cause harm in a variety of ways. Persuasion is one such way.
If I manage to persuade some people that red-headed persons are evil and
sub-human, those convinced by what I say may well behave in ways that
are harmful towards red-headed persons. In this case, I express a view and
reasons in support of that view. When speech harms via persuasion, it fits
the mini-philosophy-paper model of speech presupposed by the more naïve
version of the “more speech” response. If persuasion were the only way for
speech to be harmful, then the more speech response might be an adequate
general solution for the problem of harmful speech. As it stands, however,
there are very many other ways for speech to cause harm.
Notice first that, in this case, the view and the reasons for the view are
explicitly stated (or said) but coming out and saying it is only one way to
communicate content. As a matter of fact, we tend not to say what we
mean. Instead, we tend to say something else that enables our interlocu-
tors to infer what we mean.19 There are many ways that content can be
communicated without actually being said. It can be presupposed, impli-
cated, insinuated, or taken as background for an abductive inference.20
Although such indirectly communicated content can certainly factor into a
mini-philosophy-paper model of speech, it is worth highlighting that what
is communicated indirectly is much less susceptible to scrutiny. Thus, even
on the mini-philosophy-paper model, speech can express content and alter
beliefs in sneaky, and thus difficult to identify, ways thereby complicating
any attempts to reverse the effects via counter speech.
Another way that speech can cause harm is merely by making things rel-
evant. Referring to me with the pronoun “she,” for example, is sufficient to
make it salient that I am a woman and, in some contexts, this is enough to
trigger stereotype threat, implicit biases, and harmful associative schemas.21
That speakers are causally responsible for such effects certainly does not
mean that they are either morally or legally responsibility for them. Further-
more, speakers typically do not intend such results. Rather, they come about
because of the interaction between speech and the complex social world in
which we use it.
Yet another way that speech can cause harm is by enacting norms that
prescribe a harmful practice. In such a case, we say that the speech in ques-
tion constitutes the harm. An example might help. Suppose that the C.E.O.
of a company verbally enacts a new hiring policy when he says, “From now
on, we don’t hire women. They’re too irrational to make good decisions.”
Suppose further that, with this policy in place, a discriminatory hiring prac-
tice ensues. Because the C.E.O.’s utterance enacts a policy that prescribes
a discriminatory (and thus harmful) practice, it constitutes the harm of
discrimination. Notice that the harm (of discrimination) is causally down-
stream from the speech that constitutes it. Clearly, “constitute” is here being
used in a special technical sense. Notice also that the policy is enacted via
an exercise of speaker authority; this is the standard way for speech to enact
norms and thus to enact harmful norms.
Elsewhere I have argued that there is another way (besides the exercise of
speaker authority) for speech to enact norms and thus for speech to constitute
harm by enacting harmful norms.22 Contributing to a conversation (or any
other norm-governed activity) enacts norms for that very activity. In this
way, norm-enactment is very widespread and this opens up the possibil-
ity that we are often and even unwittingly enacting norms. Moreover, since
some of those enacted norms might prescribe harmful practices, speech that
constitutes harm might actually be quite common. I will have more to say
about this (sneaky mechanism of norm-enactment) in the following section.
Focusing on this phenomenon of harm constitution matters for at least
two reasons. First, the difference between constituting harm and causing it
other ways matters to the law. Because the C.E.O.’s utterance enacts a dis-
criminatory policy that results in a discriminatory practice, it is an unlawful
(and thus actionable) act of discrimination. The C.E.O.’s utterance is also
outside the scope of a free speech principle.23 Second, the clearer we are on
the connection between harmful speech and the associated harms, the better
placed we are to redress those harms. When harm is constituted by speech,
for example, the relevant norms are an appropriate focus of attention.
As one can see, speech can bring about harm is a wide variety of ways.
Although the mini-philosophy-paper model of speech can recognize one
way that speech can cause harm (via persuasion), it is unclear that it can
recognize any of these other ways. Language use is considerably more com-
plex than the “more speech” response would have us believe; it is not a
mere matter of negotiating the credence of stated contents. It communicates
content in a variety of indirect ways; it involves and shapes social power; it
changes relevance facts and it enacts norms. Moreover, each of these things
can be harmful.
The Scorekeeping Framework

I work within the conversational score framework. The conversational score
is a way of tracking the conversational context; it tracks that which is rel-
evant to the proper development and assessment of the conversation.24 My
notion of score is thus (maximally) inclusive; it tracks everything relevant.25
It tracks the topics, the salience facts, the presuppositions, the standards of
accuracy, what is shared, said, heard, misheard, believed, accepted for the
purposes of the conversation, metalinguistic facts about turn-taking, and so
forth. In working with such an inclusive notion of score, I take myself to be
following Lewis.26
Common ground is another way to track conversational context but it
tracks less than the score does. It tracks only psychological states of a cer-
tain sort (it tracks belief-like states of acceptance) and it tracks only those
belief-like states that are shared by all participants (accepted by each and
recognized by each to be accepted by each). For the purposes of analyzing
the phenomenon of pragmatic presupposition, common ground is a fine
choice.27 For my purposes, however, score is preferable. It tracks everything
the common ground does and then some.
In particular, the scorekeeping framework helps to highlight an important
but previously overlooked mechanism of norm-enactment. When one con-
tributes to a conversation, one enacts (a variety of) score changes. Moreover,
since what is permissible in a conversation depends on the norms governing
it and what has happened thus far in that conversation (as captured by the
score), changing the score thereby changes what is subsequently permissible
in that particular conversation.28 In this way, conversational contributions
thus routinely enact conversation-specific permissibility facts and are thus
what I call conversational exercitives.29
An example may help to illuminate the phenomenon. Suppose that Sarah
and I are talking about her dog Chloe’s coloration and I then say “Our
dog Braun has the cutest coloring above his eyebrows; he is just the sweet-
est!” My conversational contribution enacts many score changes but one
of them concerns which dog is the most salient dog in the context of this
particular conversation. By bringing up my dog Braun, I thereby make him
the most salient dog; I thus enact a score change to the salient-dog compo-
nent of the conversational score and this has permissibility consequences:
Braun (and not Chloe) is the proper referent of the expression ‘the dog.’ It is
conversationally inappropriate to try to refer to any other dog with the
expression ‘the dog’ until of course that salience fact changes again. These
conversation-specific permissibility facts are called s-norms and they are to
be distinguished from the more general norms that govern all conversations
(which are called g-norms).30
Elsewhere I have also argued that this rather hidden mechanism of norm-
enactment is not peculiar to conversation; it is an aspect of participation in
any norm-governed activity.31 When one makes a move in a norm-governed
activity, one enacts score changes that then enact permissibility facts for that
token instance of that norm-governed activity. When I say something kind,
for example, I am making a move in a conversation but I am also making
a move in the more general norm-governed activity of social interaction.
I enact s-norms for the conversation (via enacting changes to the conversa-
tional score) but I also simultaneously enact s-norms for that token instance
of social interaction and I do so by enacting changes to the score for that
particular social interaction.
I recognize that this summary of work done elsewhere is a mere gloss
but I hope that it is sufficient to show that our actions and our utterances
involve a lot of norm enacting and this is so both as a matter of course and
even though we are hardly aware of it. We are routinely enacting norms for
the conversations to which we contribute and for the broader social prac-
tices in which we participate.
Counter Speech Cannot Undo

On the mini-philosophy-paper model of speech, counter speech can—when
perfectly effective—completely reverse the effects of the speech to which it
responds. Suppose, for example, that Sam argues for Q with reasons L but
Steve objects by demonstrating that L affords no support whatsoever for Q
and Sam then recognizes this. On the mini-philosophy-paper model, Steve’s
contribution completely undoes Sam’s. Tracking only content and argumen-
tative support, it is possible for counter speech to completely reverse those
effects.
Clearly, speech does considerably more than merely communicate con-
tent and reasons for belief. On this more complex and nuanced picture of
language use, counter speech cannot simply reverse all of the effects of the
speech to which it responds. Some effects cannot be undone. One cannot
change the fact that the speaker spoke, that she said what she said, that
certain topics have been introduced or that particular salience facts have
changed. Of course, some score changes are fairly easy to change back.
Whose turn it is to speak or which dog is the most conversationally salient
one are each components of the score than can be changed rather easily.
The real issue, when it comes to responding to allegedly harmful speech,
is whether the particular components of conversational score (that would
lead to harm) are easy to change in ways that prevent or mitigate that harm.
Given the complex causal connections between speech and broader social
practices, it is difficult to control or even predict all of the causal conse-
quences of one’s utterances. We are unwittingly contributing to an array of
social practices whenever we speak and that inevitably involves enacting
s-norms for those practices. Elsewhere, I have suggested that sexist remarks
are especially difficult to undo and that attempts to reverse their effects are
akin to trying to “unring a bell.”32 “What do you mean chicks can’t do
math . . . and please don’t call me a chick!” It seems that the damage is done
and attempts to undo it are bound to fail.
I still think so. Bob Simpson has argued in support of this claim by
appealing to empirical work on implicit associations.33 As he points out,
sexist comments trigger a whole host of implicit associations (say, e.g., an
association between femaleness and irrationality) and that even attempts to
undermine those associations end up reinforcing them. Once these associa-
tions and schemas are made relevant (by the sexist remark), it is very dif-
ficult to make them un-relevant. For this reason, it is especially difficult to
undo or prevent their harm. Such comments set in motion a complex causal
process that affects our cognition in ways of which we are mostly unaware.
In addition to such unintended purely causal effects, our utterances are
also routinely yet unwittingly enacting norms—both for the conversation
to which we contribute and to the broader social practices in which we are
participating. Although we are not consciously intending to enact conversa-
tional permissibility facts when we add to a conversation, we are at least able
to identify those consequences on reflection. We are able to do this because
the g-norms of conversation (e.g., Grice’s cooperative principle) are fairly
well understood.34 The same cannot be said, however, about the precise con-
tent and force of the g-norms governing broader extra-conversational social
practices. Those g-norms are significantly less well understood. Because
the extra-conversational normative consequences are a function of those
g-norms, they are harder to identify and thus prevent or otherwise remedy.
Counter Speech Can Be Risky

Given the complexity of language use and the role that not-very-well-understood
broader social practices play, it can be very difficult to predict or control
the consequences of one’s utterance. Language use is not a mere matter of
information sharing or content communication; it is also a way of treating
people and it is rife with normative consequences. What follows are just
some of the possibilities.
Counter speech can be dangerous and there are different sorts of dan-
gers and risks involved. If a trans woman verbally objects to being mis-
pronounced in a way that highlights her being a trans woman, her remark
can lead to transphobic violence. Other dangers are less physical and more
social. Correcting your boss for her racially insensitive remarks can be pro-
fessionally damaging. Taking issue with a microaggression at a formal din-
ner party can be extremely socially awkward. I grant that there are better
and worse ways to respond and all of this is extremely context sensitive, but
it is also clear that there are real risks sometimes involved in so doing.
Counter speech can also backfire. When an ally speaks out in defense of
another, even with the best of intentions, it can perpetuate the harm of the
privileged speaking on behalf of or even over the less privileged; it can also
serve to reinforce the belief that all of this is just an overreaction by the pre-
ciously oversensitive. Speaking up can undermine one’s credibility or social
standing in the eyes of one’s interlocutors and, if one is pegged as a sancti-
monious ass, for example, this can render one considerably less effective in
counteracting harmful speech in the future.
Moreover, anticipating ways that one’s utterance could backfire, because,
say, of mistakes that would be made by (incompetent) audience members,
can lead one to decide against speaking. Such self silencing, dubbed “tes-
timonial smothering” by Dotson, can itself be harmful.35 Suppose, for
example, that an African American woman decides against objecting to a
racist joke at work because she knows that doing so will only perpetuate
the stereotype of the slightly unhinged “angry black woman.” Discussion of
this self-silencing phenomenon has a long history in Black Feminist scholar-
ship.36 In the following section, I explore some of the complexities of not
speaking.
Saying Nothing Can Be Harmful

Silence is complex. In conversational contexts, it is often meaningful. Fur-
thermore, as Tanesini stresses (this volume), silence can be a communi-
cative act.37 Saying nothing in response to an interlocutor’s assertion can
communicate one’s complete agreement with that assertion. Saying nothing
can also communicate one’s complete and utter lack of agreement! Con-
sider, for example, an incredulous stare in response to a person’s claims
about her abduction by aliens.
Silence can be meaningful even when the silent person does not have com-
municative intentions. Silence, or failure to object, can be causally responsi-
ble for problematic content being added to the common ground (and for the
relevant score changes persisting for longer). Suppose, for example, that my
interlocutor says, “Even a girl could pass that math entrance exam; it was
so easy, it was ridiculous!”38 This utterance presupposes that girls/women
are not good at math and failing to object to it allows that presupposition
to both have an enduring effect on the score and be added to the common
ground. Recall that allowing the introduction of this presupposition into
the conversation enacts norms not just for the conversation but also for this
particular social interaction and those s-norms are likely to have a nega-
tive impact on females. One might think that any harm resulting from the
presupposition-introducing utterance is on the speaker but, since conversa-
tion is a joint cooperative activity, the presence of that presupposition in
the common ground and the persistence of the relevant score changes also
depend on the hearer’s failure to object. In conversational contexts then,
silence is often an act of acceptance and should not be regarded as simply
doing nothing.39
It now looks like some speakers face a double bind. Counter speech can
backfire and be harmful but failing to speak can also be harmful. Either
way, it seems, those speakers risk harm. It is probably no surprise that mem-
bers of marginalized social groups face this sort of dilemma; such double
binds are symptomatic of oppression.40 The situation may seem hopeless,
but I think that there is room for optimism: the same sneaky mechanisms
of norm-enactment involved in harmful speech can enact more egalitarian
s-norms. To this more uplifting possibility, I now turn.
The Possibility of Positive Counter Speech

In the same way that one can do more harm than one intends or has the
individual power/authority to accomplish because of the social practices to
which one is perhaps unwittingly contributing, so too one can do more good
than one intends or has the individual power to do by tapping into broader
social practices. Again, what we say does not just affect what is appropriate
in the conversation; it also affects the norms of the social space in which
that conversation is taking place. It is possible to say things that enact more
egalitarian norms in our social spaces.
Imagine an American philosophy department that is collegial but hier-
archical. The full professors are above the associate professors who are
above the assistant professors who are above the instructors who are above
the adjunct faculty members. This ranking is institutionalized via salary,
benefits, job responsibilities, and possibilities for advancement but it is also
reflected in the culture of the department. Adjuncts do not attend depart-
ment meetings; instructors are neither consulted about scheduling nor
included in the social life of the department. You get the idea.
Consider the following exchange that takes place during a department
meeting.41 Peter is a full professor and the chair. Ralph is a longtime perma-
nent instructor and the only person of color in the department; Jane is the
only junior faculty member and the only woman. Being chair is a mere chore
to the senior men and its associated stipend negligible. To Ralph, however,
being chair would be a much valued signal of professional recognition and
inclusion and the stipend it would bring would be a real windfall for him
and his family.
PETER: “We really need to figure this out. I am stepping down as chair at the
end of this year and someone really needs to step up. Nobody wants to
do it but someone has to.”
[Awkward silence.]
JANE: “Well, I know that I am not eligible to be chair. I just got here and
I don’t have tenure but is Ralph eligible?”
[Long awkward silence.]
JANE: [with her hands in the air] “I am just trying to be clear on the eligi-
bility conditions so we can be sure that we are considering all of our
options. I am not sure that this would be of interest [turns to Ralph
sitting alongside wall] to you, Ralph, but if it does and we could make
it happen, it would be good for everyone—especially given that no one
else seems to want to do it!”
[Many slightly reluctant nods in the room]
PETER: Ahh, well we’ll come back to this.
Let’s focus on Jane’s contributions. Jane’s comments enact many changes

to the conversational score but the primary change of interest is that she
expands the scope of the relevant quantifier in order to include Ralph. By
raising the possibility of Ralph being chair, Jane made it subsequently con-
versationally inappropriate for others to continue to ignore that possibility.
Jane rendered Ralph conversationally relevant.
Of course, Jane’s comments are not just moves in a conversation, they are
also moves in a social interaction; they are ways of treating people; they are
ways of treating Ralph. There are many things going on here but one of the
main things Jane did was to treat Ralph like a valued and equal member of
the department. She acted as if the department was not hierarchical and she
did so in the hopes that doing so would come to make the department more
egalitarian.
Suppose it works. Suppose, for example, that the senior faculty only
needed to see someone treat Ralph like a valued colleague in order to realize
that they really hadn’t been. Suppose that going forward they invite Ralph
to lunch; they consult him in the hallways about pedagogy and they advo-
cate for him with the administration. In such a case, Jane’s comments rely
on egalitarian norms being operative in the department and doing so sets
in motion a process that results in egalitarian norms being operative in the
department.
Several points are worth stressing. First, what Jane does here is risky. It
could have turned out badly for everyone. It could be that Jane angers her
senior colleagues and effectively destroys any chance she has of getting ten-
ure. It could be that Ralph resents her intervention and feels humiliated by
her drawing attention to his subordinate status. It could be that her senior
colleagues resent her interference and take it out on both Jane and Ralph.
Unfavorable consequences are real possibilities; Jane’s actions are genuinely
risky.
Second, when it comes to socially constructed facts (that ontologically
depend on a complex form of collective recognition), acting as if that fact
obtains is one way to make it obtain. As we have just seen, acting as if the
department is egalitarian can come to make it so. Acting as if I am confident
can cause people to treat me in ways that actually make me confident. Act-
ing as if I have authority can, so long as the relevant others play along, come
to make it the case that I actually have authority in a certain realm.42 Acting
as if you are my friend will help to make it the case that you are my friend.43
Third, Jane chose to be indirect when she verbally challenged the culture
of the department and her doing so was strategic. Feigning ignorance of the
import of her comments allowed herself and others to act as if her remarks
were really only about the requirements for being chair. Being indirect in
this way allows participants, Jane included, to save face. Suppose, instead,
that Jane had been explicit and direct in her challenge. Suppose that she
said, “You all think you are so decent but actually you are really crappy and
you perpetually step on our only instructor and our only person of color!
Get a grip!” Had she said this, it is unlikely to have gone well for anyone.44
Indirection can be face-saving.
Fourth, Jane manages to enact egalitarian s-norms for this social interac-
tion and her doing so causes the g-norms of the department to become more
egalitarian and she does this even though Jane is a junior woman with rela-
tively little power. Enacting norms does not require authority. When one’s
utterances are moves in norm-governed activities, one enacts score changes
and thus s-norms for the activity in question. Because Jane’s remarks abide
by egalitarian g-norms, she brings those egalitarian social norms to bear in
that micro-context and she does this despite her relative lack of power.
Fifth, this example illustrates one way to change social norms. It can
often seem that they cannot be changed. After all, abiding by such norms
certainly does not change them; it perpetuates them. Moreover, violating
the norms triggers their enforcement and thus also reinforces and perpet-
uates them. This grim and hopeless picture of our social world is some-
times attributed to Butler.45 As Bicchieri and Mercier stress, however, the
changing of social norms is greatly facilitated when participants are mutu-
ally aware of alternatives to those social norms.46 In Jane’s case, she acted
as if (everyone-is-an-equally-valuable-member-of-the-team-and-the-team-
has-a-shared-goal-and-purpose) norms were operative and these norms are
already familiar. That they are already familiar to all department members
greatly facilitates the process of reshaping the department’s norms. Possibili-
ties like this one give me hope.
Sixth and finally, I think of Jane’s remarks as counter speech but her
kind of counter speech works differently than the way counter speech is
standardly understood. Notice that Jane isn’t primarily taking issue with
the content of the speech she is responding to; she is taking issue with the
(hierarchical) norms functioning in the background. As I see it, Jane’s most
important contribution is the egalitarian nature of the norms that she brings
to bear with her comments. Her speech is not so much counter to the con-
tent of what Peter said than it is counter to the hierarchical norms implicit
behind what Peter said. Counter speech like’s Jane’s is an overlooked but
potentially powerful and empowering option; it’s the right kind of “more
speech.”
Conclusion
The “more speech” response operates with an impoverished view of lan-
guage use. Our utterances are not mini-philosophy papers; we do consid-
erably more with our words than merely express content and reasons in
support of that content. Even on more sophisticated versions of the “more
speech” response that acknowledge illocutionary and perlocutionary force,
not everything that speech does can be undone. Once we realize that lan-
guage use is a social practice and it is a social practice deeply embedded
in broader social practices, we see that language use is infused with norm-
enactment and it is difficult to predict or control the effects of our utter-
ances. What is done by speech is not easily tracked and/or undone. That
said, the news is not all bad: in the same covert way that speech can enact
harmful norms and sneakily perpetuate unjust social practices, so too it can
enact egalitarian norms and surreptitiously reshape the social world for the
better.
Notes
1 I thank audiences at the Voicing Dissent Workshop at the University of Con-
necticut Humanities Institute and the free speech seminar of the Republiques des
Savoirs research group at the Ecole Normale Superieure for extremely helpful
feedback on this material.
2 Elsewhere, I am concerned with regulation. See, for example, McGowan, Mary
Kate, “On ‘Whites Only’ Signs and Racist Hate Speech: Verbal Acts of Racial
Discrimination” (2012): 121–147; and McGowan, Mary Kate (forthcoming).
3 Whitney v. California: 380.
4 See, for example, Schauer, Frederick.
5 Some theorists shift the burden away from those targeted but this involves a
different kind of counter speech. For an example of this approach, see Gelber,
Katherine (2002).
6 Hornsby, Jennifer (1993); Langton, Rae (1993); Hornsby, Jennifer and Rae
Langton (1998).
7 For authority silencing, see McGowan, Mary Kate. For sincerity silencing, see
McGowan, Mary Kate. For a survey of various types of silencing, see McGowan,
Mary Kate (2017).
8 Dotson coined the expression “testimonial quieting” in order to draw attention
to a long history of black feminist scholarship on the phenomenon of testimonial
quieting/injustice. See, for example, Collins, Patricia Hill (2000) and Dotson,
Kristie (2011). The expression “testimonial injustice” belongs to Fricker and it
has caught on in analytic circles. See Fricker, Miranda (2007).
9 Fricker discusses this example. Fricker, Miranda (2007): 153.
10 Spender, for example, discusses the male bias in the shared social meaning of the
concept of motherhood (1980): 54–58.
11 “Hermenuetical injustice” is Fricker’s term (2007).
12 Fricker’s view of hermeneutical injustice is more nuanced in her “Epistemic Injus-
tice and the Preservation of Ignorance,” in Rik Peels and Martijn Blaauw (Eds.)
(forthcoming). In part, she is responding to critics. See, for example, Medina,
José (2013).
13 For a discussion of this, see Dotson, Kristie (2011) and McGowan, Mary Kate
(forthcoming).
14 Dotson, Kristie (2011).
15 Nielsen, Laura Beth (2006).
16 Nielsen, Laura Beth (2006).
17 See, for example, Langton, Rae (2009): 72–93. She attributes the argument
model of speech to Ronald Dworkin. See his “Two Concepts of Liberty” (1991):
100–109.
18 Austin, J. L. (1962).
19 Much of what we communicate is done so via conversational implicature. Grice,
H. Paul (1989).
20 For a discussion of logical or semantic presupposition, see van Fraassen, Bas
(1968). For pragmatic presupposition, see Stalnaker, Robert (1973): 77–96. For
both conventional and conversational implicature, see Grice, H. Paul, Studies in
the Ways of Words (1989), and Davis, Wayne, Inplicature (1998). For insinua-
tion, see Camp, Elisabeth (forthcoming).
21 On stereotype threat, see Steele, Claude (1997); On implicit bias, see Banaji, M.,
C. Hardin, and A. Rothman (1997). See also, Simpson, Robert Mark.
22 McGowan, Mary Kate; McGowan, Mary Kate (2004); McGowan, Mary Kate
(2012); McGowan, Mary Kate (forthcoming).
23 Schauer, Frederick (1982); Maitra, Ishani and Mary Kate McGowan (2009).
24 Lewis, David (1983).
25 There are different ways to characterize score and they are not all as inclusive
as mine. For a less inclusive notion of score, see Camp, Elisabeth (forthcoming).
For Camp, score tracks inter-subjective obligations and thus only the on-record
contributions.
26 Lewis, David (1983).
27 Stalnaker, Robert, “Presupposition” (1974); Stalnaker, Robert (2002). Arguably,
the common ground literature arose from Lewis’s notion of common knowledge.
See Lewis (1969). For an especially detailed and empirically informed discussion
of common ground, see Clark (1996). Stalnaker credits Grice’s William James
Lectures with the introduction of the notion of common ground. See Stalnaker
(2002) and Grice (1989).
8 McGowan, Mary Kate; McGowan, Mary Kate (2004).
2
29 Exercitives enact permissibility facts in some realm. The term comes from Austin
who regarded exercitives as a type of speech act. Austin, J. L. (1962). I do not
regard exercitive force as a type of speech act, but rather as something an utter-
ance does along with everything else that it does.
30 McGowan, Mary Kate (2009a); McGowan, Mary Kate (2012); McGowan,
Mary Kate (forthcoming).
31 McGowan, Mary Kate; McGowan, Mary Kate (2012); McGowan, Mary Kate.
32 McGowan, Mary Kate (2009a).
33 Simpson, Robert Mark (2013).
34 Grice, H. P. (1989). In conversation, one ought to be cooperative and give only
information that is relevant and for which one has adequate evidence. One
should aim to enable your interlocutor to figure out what you mean by what you
say. These are some of the g-norms governing conversation generally.
35 Dotson, Kristie (2011). For a detailed discussion of the complex intersectional
issues at play with the disinclination to report about violence in non-white com-
munities, see Crenshaw, Kimberle (1991).
36 Collins, Patricia Hll (2000).
37 Tanesini, Alessandra, “Eloquent Silence,” this volume.
38 Jokes, and our responses to them, are even more complex than presupposition-
introducing utterances. For a discussion of the linguistic functioning of jokes,
see Horisk, The Pragmatics of Demeaning Jokes, manuscript. What silence (or
laughter) means in response to a joke is even more complex and sensitive to
context.
39 Some argue for an obligation to object. Johnson argues for a defeasible epis-
temic obligation to object. See her “Just Say ‘No’: Obligations to Voice Disa-
greement,” (manuscript). Ayala and Vasilyeva argue for a moral obligation to
object. See Ayala, Saray and Nadya Vasilyeve (2016). Horisk also argues for
a moral responsibility to object. See her The Pragmatics of Demeaning Jokes,
manuscript.
40 Frye, Marilyn, “Oppression,” (2003).
41 This example is discussed in more detail in my Just Words: Speech and the Cov-
ert Constitution of Harm (forthcoming).
42 Maitra calls this licensing. Maitra, Ishani (2012). The phenomenon was identi-
fied earlier in Thomason, Richard (1990).
43 In arguing against Clifford, James points out that faith in a fact can help to make
that fact obtain (or open one up to otherwise unavailable evidence of that fact).
James uses the friendship example. See James, William (1979), first published in
1897.
44 That said, I grant that sometimes being blunt and confrontational is the best
choice—all things considered.
45 Nussbaum attributes this view to Butler (1999). Posted November 2000. www.
tnr.com/archive/0299/022299/nussbaum022299.html.
46 Bicchieri, Cristina and Hugo Mercier (forthcoming).
References
Austin, J. L. (1962). How to do things with words. Oxford: Oxford University Press.
Ayala, S., & Vasilyeve, N. (2016). Responsibility for silence. Journal of Social Phi-
losophy, 47(3), 256–272.
Banaji, M., Hardin, C., & Rothman, A. (1993). Implicit stereotyping in person judg-
ment. Journal of Personality and Social Psychology, 65(2), 272–281.
Bicchieri, C., & Mercier, H. (Forthcoming). Norm and beliefs: How change occurs.
In B. Edmonds (Ed.), The dynamic view of norms. Cambridge: Cambridge Uni-
versity Press.
Camp, E. (Forthcoming). Insinuation, indirection, and the conversational score. In
Daniel Fogal, Matt Moss, & Daniel Harris (Eds.), New directions in speech act
theory. Oxford: Oxford University Press.
Clark, H. (1996). Using language. Cambridge: Cambridge University Press.
politics of empowerment (2nd ed.). New York: Routledge.
Crenshaw, K. (1991). Mapping the margins: Intersectionality, identity politics, and
violence against women of color. Stanford Law Review, 43(6), 1241–1299.
Davis, W. (1998). Implicature. Cambridge: Cambridge University Press.
Dotson, K. (2011). Tracking epistemic violence, tracking practices of silencing.
Hypatia, 26(2), 236–257.
Dworkin, R. (1991). Two concepts of liberty. In Edna Ullman-Margalit & Avishai
Margalit (Eds.), Isaiah Berlin: A celebration (pp. 100–109). London: Hogarth
Press.
Fricker, M. (Forthcoming). Epistemic injustice and the preservation of ignorance. In
R. Peels & M. Blaauw (Eds.), The epistemic dimensions of ignorance. Cambridge:
Cambridge University Press.
Frye, M. (2003). Oppression. In The politics of reality: Essays in feminist theory
(pp. 1–16). Freedom, CA: The Crossing Press.
Gelber, K. (2002). Speaking back: The free speech versus hate speech debate. Amster-
dam: John Benjamins Ltd.
Grice, H. P. (1989a). Logic and conversation. In Studies in the ways of words
(pp. 22–40). Cambridge, MA: Harvard University Press.
Grice, H. P. (1989b). Studies in the ways of words. Cambridge, MA: Harvard Uni-
versity Press.
Horisk, C. The pragmatics of demeaning jokes, unpublished manuscript.
Hornsby, J. (1993). Speech acts and pornography. Women’s Philosophy Review, 10,
38–45.
Hornsby, J., & Langton, R. (1998). Free speech and illocution. Legal Theory, 4, 21–37.
James, W. (1979). The will to believe. In The will to believe and other essays in
popular philosophy. Cambridge, MA and London: Harvard University Press; first
published in 1897.
Johnson, C. Just say ‘No’: Obligations to voice disagreement, unpublished
manuscript.
Affairs, 22, 293–330.
Langton, R. (2009). Beyond belief: Pragmatics in hate speech and pornography. In
In. Maitra and M. K. McGowan (Eds.), Speech and harm: Controversies over free
speech (pp. 72–93). Oxford: Oxford University Press.
Lewis, D. (1969). Convention: A philosophical study. Cambridge, MA: Harvard
University Press.
Lewis, D. (1983). Scorekeeping in a language game. In Philosophical papers Volume
I (pp. 233–249). New York: Oxford University Press.
Maitra, I. (2012). Subordinating speech. In Maitra and McGowan (Eds.), Speech
and harm: Controversies of over free speech (pp. 94–120). Oxford: Oxford Uni-
versity Press.
Maitra, I., & McGowan, M. K. (2009). On racist hate speech and the scope of a free
speech principle. Canadian Journal of Law and Jurisprudence, 23(2), 343–372.
McGowan, M. K. (2004). Conversational exercitives: Something else we do with
our words. Linguistics and Philosophy, 27(1), 93–111.
McGowan, M. K. (2012). On ‘Whites Only’ signs and racist hate speech: Verbal acts
of racial discrimination. In I. Maitra & M. K. McGowan (Eds.), Speech and harm:
Controversies over free speech (pp. 121–147). Oxford: Oxford University Press.
McGowan, M. K. (2017). On multiple types of silencing. In M. Mikkola (Ed.),
Beyond speech: Pornography and analytic feminist philosophy (pp. 39–58).
McGowan, M. K. (2003). Conversational exercitives and the force of pornography.
Philosophy & Public Affairs, 31(2), 155–189.
McGowan, M. K. (Forthcoming a). Just words: Speech and the covert constitution
of harm. Oxford: Oxford University Press.
McGowan, M. K. (Forthcoming b). On political correctness, microaggressions, and
silencing in the academy. In J. Lackey (Ed.), Essays on academic freedom. Oxford:
McGowan, M. K. (2014). Sincerity silencing. Hypatia, 29(2), 458–473.
McGowan, M. K. (2009a). Oppressive speech. Australasian Journal of Philosophy,
87(3), 389–407.
McGowan, M. K. (2009b). On silencing and sexual refusal. Journal of Political
Philosophy, 17(4), 487–494.
Medina, J. (2013). Epistemologies of resistance: Gender and racial oppression, epis-
temic injustice, and resistant imaginations. Oxford: Oxford University Press.
Nielsen, L. B. (2006). License to harass: Law, hierarchy, and public offensive speech.
Princeton, NJ: Princeton University Press.
Nussbaum, M. (1999, February). The professor of parody. The New Republic Online.
Posted November 2000. Retrieved from www.tnr.com/archive/0299/022299/
nussbaum022299.html.
Schauer, F. (1982). Free speech: A philosophical enquiry. Cambridge: Cambridge
University Press.
Schauer, F. (1992). Uncoupling free speech. Columbia Law Review, 92(6),
1321–1357.
Simpson, R. M. (2013). Unringing the bell: McGowan on oppressive speech and the
asymmetric pliability of conversation. Australasian Journal of Philosophy, 91(3),
555–575.
Spender, D. (1980). Man made language (pp. 54–58). London: Routledge & Kegan
Paul.
Stalnaker, R. (1974a). Pragmatic presupposition. In M. Munitz & P. Uner (Eds.),
Semantics and philosophy (pp. 197–213). New York: New York University Press.
Stalnaker, R. (1974b). Presupposition. Journal of Philosophical Logic, 2, 77–96.
Stalnaker, R. (1998). On the representation of context. Journal of Logic, Language
and Information, 7(1), 3–19.
Steele, C. (1997). A threat in the air: How stereotypes shape intellectual identity and
performance. American Psychologist, 52(6), 613–629.
Tanesini, A. Eloquent silence, this volume.
Thomason, R. (1990). Accommodation, meaning, and implicature: Interdisciplinary
foundations for pragmatics. In Cohen, Morgan, & Pollack (Eds.), Intentions in
communication (pp. 325–363 at 342–343). Cambridge, MA: MIT Press.
van Fraassen, B. (1968). Presupposition, implication and self-reference. Journal of
Philosophy, 65(5), 136–152.
Whitney v. California. 380.
Contributors
Matthew Chrisman
University of Edinburgh
Catherine Z. Elgin
Harvard University
Sanford C. Goldberg
Northwestern University
Graham Hubbs
University of Idaho
University of Idaho
Klemens Kappel
University of Copenhagen
Jennifer Lackey
Northwestern University
Michael Patrick Lynch
University of Connecticut
Mary Kate McGowan
Wellesley College
Rachel Ann McKinney
University of Pittsburgh
Duncan Pritchard
University of Edinburgh
Alessandra Tanesini
Cardiff University
Index
acceptability 15, 17 – 18, 56, 69, hate speech 1, 2, 119, 183, 185

122, 144 hinge commitment 6, 22 – 3, 31 – 4
Austin, J. L. 175, 185
ideal theory 82 – 4, 86, 91, 94
behavioral regularity 44 – 5, 48 – 9,
50 – 4, 57 James, William 17 – 18
belief; agnosticism 24 – 5, 30; conceptual justice 50 – 1, 68 – 9, 70, 83 – 4, 145, 147,
connection 25; transparency 27 – 8, 151, 156 – 7, 168, 170, 176 – 8, 185;
89, 91 injustice 84 – 5, 90, 144, 157, 172,
176 – 7, 184 – 5
conciliation 10, 12 – 14
cooperative conversation view 6, 82, Mill, John Stuart 99, 105 – 7, 131 – 2,
84 – 6, 89, 90 – 1, 94, 145 136, 141
counter speech 8, 182 – 6, 188 – 9, Mills, Charles 82 – 4, 86
190 – 1, 194 mutual knowledge 143, 145, 148, 158
devil’s advocate 4, 7, 26, 98 – 9, non-ideal theory 6, 82, 84, 91 – 2

101 – 7
dialectical posturing 5, 6, 23, 27 peer disagreement 1, 2, 5, 10 – 11, 15,
dissent; bad 61 – 3, 65 – 6; good 61, 23, 25 – 6
65; reasonable 61 – 2, 67, 70, 71 – 9; pressure 41 – 3, 45, 53; normative
unreasonable 6, 61 – 2, 66, 68, 72 – 4, pressure 49, 50 – 5, 58, 85, 92
76, 78 – 9
dogmatism 14, 133 regulate 182 – 3
Rawls, John 68 – 9, 73, 82
epistemic agency 135 – 8 Rousseau, J.-J. 7, 164 – 9, 170 – 9
epistemic arrogance 7, 129, 133 – 8
epistemic asset 18 – 20 Searle, John 116, 118
epistemic blamelessness 71 signaling 6, 43 – 4, 49, 58
epistemic environment 129, 131 – 3, 136 speech acts 174, 176 – 7; illocution 7,
epistemic justice 97, 99, 100; epistemic 109, 110 – 118, 120 – 1, 185, 194
injustice 2, 97, 99, 100, 132 Stanley, Jason 143
epistemic peer 2, 11, 13 – 18, 22, 25 – 6 steadfastness 10 – 12, 14, 64
epistemic responsibility 14
epistemic vulnerability 10, 21 testify 100, 164
testimony 1, 2, 30, 33, 92, 100, 123,
false belief 47 – 9, 51 – 5, 57, 63 – 4, 66, 136, 155, 158
85, 93
Fricker, Miranda 99, 100 – 2, 123, 141, 143 Wittgenstein, Ludwig 31 – 3

(Routledge Studies in Contemporary Philosophy) Casey Rebecca Johnson (Ed.) - Voicing Dissent - The Ethics and Epistemology of Making Disagreement Public-Routledge (2018)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Routledge Studies in Contemporary Philosophy) Casey Rebecca Johnson (Ed.) - Voicing Dissent - The Ethics and Epistemology of Making Disagreement Public-Routledge (2018)

Uploaded by

Copyright:

Available Formats

Voicing Dissent

Casey Rebecca Johnson is an assistant professor of philosophy in the Politics

The Legacy of Kant in Sellars and Meillassoux

Subjectivity and the Political

Aspect Perception after Wittgenstein

Nature and Normativity

Formal Epistemology and Cartesian Skepticism

Epistemic Rationality and Epistemic Normativity

From Rules to Meanings

Toleration and Freedom from Harm

For a full list of titles in this series, please visit www.routledge.com

Edited by Casey Rebecca Johnson

2 Disagreements, of Beliefs and Otherwise 22

3 Dissent: Ethics and Epistemology 40

4 Dissent: Good, Bad, and Reasonable 61

5 Silence and Objecting 82

6 For the Sake of Argument: The Nature and Extent of Our

7 Eloquent Silences: Silence and Dissent 109

8 Epistemic Arrogance and the Value of Political Dissent 129

10 Speaking and Listening to Acts of Political Dissent 164

11 Responding to Harmful Speech: The More Speech Response,

In the spring of 2016, I found that I wanted to have a particular conversa-

Dissent and Disagreement

The Value of Dissent

List of Articles in This Volume

Normally, we take it that we know rather a lot. We take ourselves to know

A. Be steadfast: hold on to your convictions, disregarding disagreement.

If S knows that p, then p is true.

This argument is perfectly general. It applies to first-order evidence as well

Much of the work on the epistemology of disagreement tends, not unrea-

Disagreement and Belief

Conviction and Disagreement

Overview: The Pressure to Conform in Signaling Practices

Two Social Practices and a Question

FLYING THE “W”

In this scenario, magenta-on-Monday is a hyper-reliable signal of member-

The Appearance of Pressure to Conform to Ongoing

APTC Given a familiar social practice involving signaling, where each

The Ethical and Epistemic Basis of Pressure to Conform

RAAH If S reasonably anticipates that φing will have harmful effects on

I regard RAAH as highly plausible. It merely speaks of S’s having “some

RTH If S recognizes that A needs help, then S has some reason to

HARM To knowingly do something that one can anticipate will have

Moving on to the good/help connection, we might formulate the “Goodness-

HELP To knowingly do something that one can anticipate will have

EBFB False beliefs are bad from an epistemic point of view.

EVID If S has evidence E that n% of the members of a population

(Here the “corresponding credence” is a credence of n/100 in the hypoth-

RAAHR If S reasonably anticipates that φing risks having harmful

RAAHR postulates practical reasons9 to avoid doing what one anticipates

Is There Pressure to Conform When the Social

Toward an Understanding of the Pressure

Normative Pressure to Speak Out in Doubt or Disagreement

DD Whenever anyone harbors doubts regarding an assertion made in

Imagine that this regularity constitutes a strict empirical generalization:

SD If audience A observes an assertion and does not publicly dissent,

SA If audience A observes an assertion and does not publicly dissent,

GK In general, people—speakers and audience members alike—are

AA If an audience A accepts an assertion, then this is some evidence that

Extending Our Result

DD Whenever anyone harbors doubts regarding an assertion made in

On the contrary, people respond to assertions made in their presence in all

DDWeak It sometime happens that when one harbors doubts regarding an

2 Disagreements, of Beliefs and Otherwise 22

3 Dissent: Ethics and Epistemology 40

4 Dissent: Good, Bad, and Reasonable 61

5 Silence and Objecting 82

7 Eloquent Silences: Silence and Dissent 109

8 Epistemic Arrogance and the Value of Political Dissent 129

10 Speaking and Listening to Acts of Political Dissent 164