The Epistemic Dimensions of

Ignorance

Rik Peels
Vrije Universiteit Amsterdam

Martijn Blaauw
Delft University of Technology
University Printing House, Cambridge CB2 8BS, United Kingdom

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781107175600
© Rik Peels and Martijn Blaauw 2016
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2016
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Peels, Rik, 1983– editor.
The epistemic dimensions of ignorance / [edited by] Rik Peels, Vrije Universiteit
Amsterdam, Martijn Blaauw, Delft University of Technology.
New York : Cambridge University Press, 2016.
LCCN 2016025916 | ISBN 9781107175600
LCSH: Ignorance (Theory of knowledge) | Knowledge, Theory of.
LCC BD221 .E65 2016 | DDC 121–dc23
LC record available at https://lccn.loc.gov/2016025916
ISBN 978-1-107-17560-0 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party Internet Web sites referred to in this publication
and does not guarantee that any content on such Web sites is, or will remain,
accurate or appropriate.
1 The Nature of Ignorance: Two Views

Pierre Le Morvan and Rik Peels

Introduction
Our purpose in this chapter is to explore the nature of ignorance.
When we ask about its nature, we ask what it is to be ignorant.
We address this question by considering two rival accounts that can
be found in the literature, each of which specifies a distinct set of
conditions taken to be individually necessary and jointly sufficient for
ignorance.
The two rival accounts have recently been developed in more detail
and defended on the basis of various arguments. In this chapter, we
spell out these two different views and provide an overview of the main
arguments for them. On the first view, called the Standard View,
ignorance is lack or absence of knowledge, whereas on the second
view, called the New View, ignorance is lack or absence of true belief.
Among the adherents of the Standard View are Lloyd Fields, Susan
Haack, Pierre Le Morvan, and Michael Zimmerman.1 The New View
is embraced, among others, by Alvin Goldman, Alexander Guerrero,
Rik Peels, and René van Woudenberg.2
The chapter is structured as follows. First, we make a few preliminary
comments by distinguishing various kinds of knowledge and explicating
in more detail what the difference between the Standard and New Views
amounts to (§ 2). Then, we provide a case for the Standard View (§ 3).
We spell out three arguments that one might provide in favor of this
conception: an argument from common usage (§ 3.1), an argument
from unifying theorizing about knowledge and ignorance (§ 3.2), and an
argument from ignorance of falsehoods (§ 3.3). Next, we provide a case
for the New View (§ 4) by laying out three arguments in favor of it: an

1
See Fields (1994, p. 403), Haack (2001, p. 25), Le Morvan (2011, 2012, 2013);
Zimmerman (1988, p. 75; 2008, ix).
2
See Goldman (1986, p. 26), Goldman and Olsson (2009, pp. 19–21), Guerrero (2007,
pp. 62–63), Peels (2010, 2011a, 2012, 2014), Van Woudenberg (2009, p. 375).

12
 The Nature of Ignorance: Two Views 13

argument from intuitions about cases of true belief that fall short of
knowledge (§ 4.1), an argument from ignorance as an excuse (§ 4.2),
and an argument from ignorance by acquaintance and procedural ignor-
ance (§ 4.3). The purpose of Sections 3 and 4 is not to convince the reader
of a particular view on ignorance, but merely to present arguments that
adherents of these two views might advance in favor of them. We leave it
up to the reader to decide which of the two views she or he finds more
convincing. We conclude with a couple of retrospective and prospective
remarks (§ 5).

2 Ignorance: Preliminaries
In this section we make two preliminary comments that will play an
important role in the two following sections.
First, since on the Standard View, ignorance is the lack or absence
of knowledge, it is important to note that it is widely thought that
there are three different kinds of knowledge.3 First, there is what is
often called factual or factive knowledge, that is, knowledge that some
specific proposition is true. Knowing that one’s wife is at her office,
that Abuja is the capital of Nigeria, and that 83 is a prime number
belong to this class of knowledge. Second, there is objectual knowl-
edge: knowing a certain object, where that object can but need not be
a person. One can know one’s friend, the taste of pineapples, one can
know cities such as Berlin, and one can know what it is to be fired or
to be in love. Third, there is procedural knowledge: knowledge of how
to do something, how to perform some task.4 Here are a few exam-
ples: knowing how to ride a bicycle, knowing how to open a wine
bottle, knowing how the play the oboe, and knowing how to forgive
someone who wronged one. We return to this threefold distinction
below.5

3
Many epistemologists have accepted the distinction between these three kinds of knowl-
edge. There are a few exceptions, though. Some philosophers contend that both objectual
and procedural knowledge are reducible to factual knowledge, or that they are a subspecies
of factual knowledge. For some tentative arguments in favor of this thesis, see Snowdon
(2004) and for an elaborate, mainly linguistic defense of it, see Stanley and Williamson
(2001). We find this view unconvincing, but cannot elaborate on this issue here; for a good
linguistic note on Stanley’s and Williamson’s article, see Rumfitt (2003).
4
We prefer to talk about procedural knowledge rather than knowledge-how, for, as Paul
Snowdon has convincingly argued, there are instances of knowledge-how that are not
instances of procedural knowledge. See Snowdon (2004, p. 7).
5
For an overview of these kinds of knowledge by one of those epistemologists, see Lehrer
(2000, p. 5). For an influential account of the distinction between factive and procedural
knowledge, see Ryle (1945, pp. 4–16; 1973, pp. 28–32, 40–41).
14 Pierre Le Morvan and Rik Peels

The second preliminary remark concerns only factual knowledge.

As we said, on the Standard View, ignorance is lack or absence of knowl-
edge. We should note, though, that there are five different ways in which
one can lack knowledge that some proposition p is true – whether or not
these five ways of lacking knowledge can be further divided into various
ways in which one can lack knowledge:
(i) p is false;
(ii) S disbelieves the true proposition p;
(iii) S suspends belief on the true proposition p;
(iv) p is true and S neither believes that p, nor disbelieves that p, nor
suspends belief on p;
(v) S believes the true proposition p, but S’s belief that p lacks warrant,
where warrant is that which turns true belief into knowledge.
The Standard and New Views agree that if one of the conditions (i) to (iv)
is met, then we have a case of factive ignorance. Let us explain. For ease of
exposition, we will start with (ii), (iii), and (iv), and then turn to, respec-
tively, (i) and (v).
As to (ii) and (iii), if someone disbelieves or suspends judgment on, say,
the true proposition that Napoleon lost the Battle of Waterloo in 1815,
she is ignorant that he did, because it is true that he did. As to (iv), if
someone neither believes that Napoleon lost the Battle of Waterloo in
1815, nor disbelieves that he did, nor suspends judgment on whether he
did, for instance, because she has never even heard of Napoleon, then,
surely, she is ignorant that he did.
Things are more complicated when it comes to (i). It seems clear that if
someone, say, falsely believes that Napoleon lost the battle of Waterloo in
1799, then that person lacks knowledge that this was the case. But one
might think that such a person is not ignorant that Napoleon lost the battle
of Waterloo in 1799, since he did not. As we shall see below, the Standard
and New Views differ on whether one can only be ignorant of truth.
We address this issue below.6
Whether or not instances of (v) also count as cases of ignorance is
a matter on which the Standard and New Views clearly disagree.
In other words, they disagree on whether one is ignorant that p if one
truly believes that p, but fails to know that p. On the Standard View, one is
ignorant in such cases, whereas on the New View, one is not. In the
following two sections, we will consider some arguments that might be
advanced in support of each of these two views.

6
As we later see, an intuition motivating the New View is that one can only be ignorant that
p if p is true; accordingly, one cannot be ignorant that Napoleon lost the battle of Waterloo
in 1799, because Napoleon lost that battle in 1815 rather than in 1799.
 The Nature of Ignorance: Two Views 15

3 Ignorance as the Lack of Knowledge

On the Standard View, ignorance is the absence or lack of knowledge,
with “ignorance” being an antonym of “knowledge.” On the prima facie
plausible assumption that the lack or absence of something x is the
complement or contradictory of x, ignorance is simply the complement
or contradictory of knowledge.7 Michael Zimmerman takes a position
that is at least close to the Standard View on ignorance when he says:
Ignorance . . . is a failure to know what is true. To know what is true, one must
believe it (something that involves having a certain level or degree of confidence in
it) and do so with adequate justification. Thus ignorance can come about in one of
two ways: either by way of failure to believe the truth or by way of believing it
without adequate justification.8

In this section, we attend to three arguments in favor of the Standard

View.

3.1 First Argument: Common Usage

As a reflection of how ignorance is ordinarily understood, the idea that
ignorance is lack of knowledge has considerable support from common
usage of the term “ignorance.” Of course, philosophical questions of the
nature of something x are rarely, if ever, conclusively settled merely by
considering common usage of a term for x. However, insofar as we seek to
understand what is ordinarily meant by a term, considering such common
usage has value, and philosophical critiques as to whether we really ought
to conceive of x as ordinarily understood requires of course understand-
ing how x is ordinarily understood in common usage.
In connection with such usage, consider the Oxford English Dictionary’s
definition 1a of the word “ignorance”: “The fact or condition of being
ignorant; want of knowledge (general or special).”9 The current meaning
of “ignorance” as an antonym of “knowledge” squares with its etymology
as the English term “ignorance” comes from the Middle English “ignor-
ance” or “ygnoraunce,” from the Old French “ignorance,” from the Latin
“ignōrāntia,” from the Latin “ignosco” derived from “in” (meaning: the
opposite of) and “gnosco” (meaning: know).
English is not unique in this regard, as definitions of cognates of
“ignorance” as antonyms of cognates of “knowledge” prove widespread.
In fact, in numerous languages, spanning several distinct linguistic
families, a cognate of “ignorance” is constructed as an antonym of
7 8
Thus also Le Morvan (2011, 2012, 2013). See Zimmerman (2008, ix).
9
“Ignorance” is also defined in terms of the lack of knowledge in the Longman Dictionary of
Contemporary English, the Merriam-Webster Dictionary, the Collins Dictionary, and others.
16 Pierre Le Morvan and Rik Peels

a cognate of “knowledge.”10 We find this phenomenon in many lan-

guages. Here are just a few examples:11
“Knowledge” Cognate “Ignorance” Cognate
Burmese: aasipanyar aasipanyar kainnmaehkyinn
Chinese: zhı̄shì wúzhı̄
Danish: viden uvidenhed
Finnish: tieto tietämättömyys
Hebrew: yediah i yediah
Hindi: jñāna ajñāna
Malagasy fahalalana tsy fahalalana
Russian: znaniya neznaniye
Turkish: bilgi bilgisizlik12
The common usage of “ignorance” thus provides considerable evidence
that it functions as an antonym of “knowledge” in English, and likewise
for cognates in numerous other languages. This in turn suggests that
taking ignorance to be the complement or contradictory of knowledge
reflects how we ordinarily conceive of the nature of ignorance and its
relationship to knowledge. Insofar as we maintain a presumption in favor
of our ordinary way of conceiving of something – such that we presume it
to be correct unless shown otherwise – we have a presumptive case in favor
of the Standard View, and a presumptive case against the New View
insofar as it denies the complementarity of knowledge and ignorance.13

3.2 Second Argument: Unifying Theorizing About Knowledge

and Ignorance
Taking ignorance to be the complement of knowledge unifies theorizing
about ignorance with theorizing about knowledge in that it seems that
insights into the nature of knowledge automatically yield corresponding
10
These linguistic families include the Austronesian, Dravidian, Finno-Ugric, Indo-
European, Japonic, Semitic, Sino-Tibetan, Tai-Kadai, and Uralo-Altaic.
11
Other examples of such languages are Azerbaijani, Basque, Bosnian, Bengali, Croatian,
Czech, Dutch, Esperanto, Estonian, German, Greek, Gujarati, Hungarian, Irish,
Japanese, Kannada, Korean, Latin, Latvian, Lao, Lithuanian, Macedonian, Malayalam,
Marathi, Serbian, Slovak, Tajik, Tamil, Telegu, Thai, Uzbek, Welsh, and Yoruba.
12
We are grateful to Pierre Le Morvan’s colleague, the linguist David Stillman, for help in
constructing this table. As Stillman has noted in correspondence, in the Germanic lan-
guages and the European Slavic languages, a term for “ignorance” is typically a calque
(loan translation) of the Latin “ignosco.” In the Romance languages, the Latin “ignoran-
tia” has become “ignorància” in Catalan, “ignorance” in French, “ignoranza” in Italian,
“ignorância” in Portuguese, “ignoranț ă” in Romanian, and “ignorancia” in Spanish.
13
The Standard View is maintained by linguists such as Stephen Levinson (2000, p. 208),
who notes that “not ignorant logically implies knows (because ignorance and knowledge
are contradictories).”
 The Nature of Ignorance: Two Views 17

putative insights into the nature of ignorance. This provides a conception

of ignorance on which each kind of ignorance corresponds to one of
the three kinds of knowledge that we discussed in the previous section:
factive knowledge, objectual knowledge, and procedural knowledge.
Accordingly, insights into the nature of knowledge automatically yield
corresponding putative insights into the nature of ignorance. This is
because, on the Standard View, ignorance has no substantive and positive
nature of its own. Being purely privative and negational, its nature is
completely determined by its contrast with the nature of knowledge.
So conceived, the relationship between ignorance and knowledge proves
analogous to the relationship between darkness and light inasmuch as
darkness is the absence or want of light. It also proves analogous to evil
understood in Augustinian terms as having no substantive or positive
nature of its own inasmuch as it is nothing more than the privation or
absence of good.14 If ignorance thus has nothing more than a privative or
negational nature relative to knowledge, then this nature can only be
properly understood in contrast with the latter. Thus, every theory or
conception of knowledge automatically yields by negation a theory or
conception of its complement ignorance, and theorizing about both is
thereby unified. To the extent that one finds such unification attractive, it
counts in favor of the Standard View and against the New View.
Assuming for the sake of argument that the Standard View is correct,
we exemplify below how insight into the nature of ignorance and its kinds
can be gained by considering them in contrast with the nature of know-
ledge in its three varieties. We consider in turn factive, objectual, and
procedural knowledge. First, the complement of factive knowledge is
factive ignorance understood as the absence or lack of factive knowledge.
Sam’s being ignorant that Caesar crossed the Rubicon or Pam’s being
ignorant that monotremes are egg-laying mammals provide examples of
such ignorance. Second, objectual ignorance is the absence or lack of
objectual knowledge. Sam’s being ignorant of this man’s character or
Pam’s being ignorant of the taste of mango, for instance, provide exam-
ples of such ignorance. Third, procedural ignorance is the absence or lack of
procedural knowledge. Sam’s being ignorant of how to operate a forklift
or Pam’s being ignorant of how to calm a crying baby provide cases in
point.
Factive knowledge has received the most attention in contemporary
epistemology. Therefore, we will focus primarily on it in addressing the
complementarity of knowledge and ignorance. In the literature, the most
extensively discussed basic conception of factive knowledge takes it to be

14
Augustine (2009, p. 43).
18 Pierre Le Morvan and Rik Peels

analyzable in terms of true justified belief with a codicil for Gettier-type

counterexamples.15 Let us attend to this conception in terms of its impli-
cations for understanding ignorance.
On this conception of factive knowledge, someone S’s knowledge that
p requires the satisfaction of the following four necessary conditions:
(i) a doxastic condition: S believes that p;
(ii) an alethic condition: p is true;
(iii) a justificatory condition: S believes that p with justification;
(iv) a Gettier-proofing condition: S’s justification for believing that p must
withstand Gettier-type counterexamples.
If ignorance that p16 is the complement of knowledge that p, then some-
one S’s failure to satisfy any of these putative necessary conditions for
such knowledge suffices for S’s being in the state of ignorance that p.17
Corresponding to each of these sufficient conditions is a kind of factive
ignorance that we may delineate as follows:
Ad (i): doxastic ignorance that p occurs when p is not believed. This can
happen in four principal ways. One way is for someone to lack the capacity
to form beliefs concerning p. Suppose, for instance, that Alex lacks the
learning and/or conceptual repertoire to grasp the following proposition:
(1) Platypuses are monotremes.
If so, then Alex fails to believe (1). Suppose, by contrast, that Barbara has
the capacity to believe (1), but withholds belief on it because she is doubtful
of its truth. Barbara in this case also fails to believe (1), but does so by not
exercising her capacity to believe it. Consider now Cindy who believes the
contradictory of (1), namely that it is false that platypuses are monotremes.
Assuming Cindy does not hold contradictory beliefs, Cindy does not
believe that (1) either. Take also Dan who, while having the capacity to
believe that (1), fails to believe this because it has never occurred to him.
Doxastic ignorance can thus arise from failing to believe a proposition
because (a) one is incapable of believing it, (b) one withholds belief even
15
We will later consider some other conceptions of factive knowledge in terms of their
implications for understanding the nature of factive ignorance.
16
René van Woudenberg has questioned whether it is proper English to say “S is ignorant
that p.” Here is why we think it is. According to the Oxford English Dictionary’s definition
2c of “ignorant,” it can be used in sentences with a subordinate clause. The OED gives
the following example: “I am ignorant that till now, I ever made you this offer.”
The construction has also been used by numerous philosophers. Here are two represen-
tative examples: (1) Ginet (1975, p. 16) writes: “it is conceivable that S should have been
in doubt or ignorant that p”; (2) Hyman (2006, p. 900) writes: “For a verb-phrase of the
form ‘is ignorant that p’ consists of a psychological verb followed by a ‘that’ clause.”
17
By contrast, on the New View, only if p is true does someone S’s failure to satisfy
condition (i) suffice for S’s being in the state of ignorance that p.
 The Nature of Ignorance: Two Views 19

though capable of believing it, (c) one believes its contradictory without
inconsistency, or (d) one fails to believe it because it has never occurred to
one to believe it.
Ad (ii): alethic ignorance occurs when a proposition is not true. Take the
following false proposition:
(2) Platypuses are native to Tanzania.
Since (2) is false, no one can know that (2) is true. Since, on the Standard
View, ignorance is the complement of knowledge, it follows that everyone
is alethically ignorant that (2) is true. As we shall see below, adherents of
the New View reject the possibility of alethic ignorance insofar as they
maintain that one can be ignorant that p only if p is true.
Ad (iii): justificational ignorance occurs when a proposition is believed
without justification. Suppose for instance that Alex believes (1) without
justification, that is, without any reason or ground. Lacking justification
for believing (1), Alex’s belief does not satisfy a necessary condition for
knowing that (1) is true, and thus satisfies a sufficient condition for
justificatory ignorance that Alex is ignorant that platypuses are
monotremes.18
Ad (iv): Gettier-type ignorance occurs when a proposition is true
and believed with justification, but is subject to Gettier-type
counterexamples.19 Imagine for example that Sam sees w, w is a genuine
Cartier watch, and Sam believes it to be a genuine Cartier watch because it
looks to him to be so. Suppose then that Sam’s belief is true and justified.
Suppose as well, however, that w is ensconced in a display of a hundred
counterfeit Cartier watches, and Sam is not able to distinguish w from the
counterfeits, and w is only in the display by accident. Many are inclined to
conclude that Sam does not know that w is a genuine Cartier watch even if
he has a justified true belief that it is. If this conclusion is correct and on the
Standard View’s supposition that failure to meet a necessary condition for
knowledge that p is a sufficient condition for ignorance that p, it follows
that Sam is in a state of Gettier-type ignorance.

18
Interestingly, different accounts of the nature of justification have a bearing on how to
conceive of the nature of justificatory ignorance. For instance, Foundationalism results in
a different account of justification than does Coherentism, and Externalism results in
a different account than Internalism. Thus, various accounts of Foundationalism,
Coherentism, Externalism, and Internalism will yield varying accounts of justificatory
ignorance. Space constraints preclude discussing here the various ways in which such
accounts can be developed.
19
Here, we understand Gettier-type counterexamples in a broader sense than the original
Gettier examples that involved inferences from false beliefs to justified true beliefs.
A Gettier-type counterexample in this broad sense is any counterexample to knowledge
understood as true justified belief.
20 Pierre Le Morvan and Rik Peels

In sum, if ignorance that p is the complement of knowledge that p as it is

on the Standard View, and if knowledge that p is true belief that p with
Gettier-proof justification, then ignorance that p occurs if (i) p is not
believed, and/or (ii) p is not true, and/or (iii) p is believed without justi-
fication, and/or (iv) the justified belief that p is subject to Gettier-type
counterexamples. Ignorance that p is doxastic in case of (i), alethic in case
of (ii), justificational in case of (iii), and Gettier-type in case of (iv).
We have so far considered factive ignorance as the complement of
factive knowledge where the latter is conceived of as true belief with
Gettier-proof justification. Worth noting however is that alternative con-
ceptions of factive knowledge lead to different complementary concep-
tions of factive ignorance. These alternative conceptions include taking
factive knowledge to be (i) analyzable as true belief alone, (ii) analyzable
as true belief with something other than justification, (iii) analyzable not
as a species of belief but rather as an ability, and (iv) unanalyzable. Let us
briefly consider each in terms of their implications for factive ignorance as
the complement of factive knowledge.
Suppose that factive knowledge is nothing more than mere true belief
as maintained for instance by Sartwell.20 If so, then factive ignorance as its
complement is nothing more than the absence or lack of true belief, and
someone S is factively ignorant that p if p is not true and/or S does not
believe that p. Thus, factive ignorance has only alethic and doxastic kinds.
Suppose that factive knowledge is true belief together with something
other than justification. Let us call this something other than justification
“O.” Alternative accounts of O include:
• the Nozickian truth-tracking condition: if p were not true, S would not
believe that p, and if p were true, S would believe that p;21
• the Plantingan warrant condition: S’s belief that p results from the
proper functioning of S’s cognitive equipment, i.e., functioning as it
was designed to function;22
• the Safety condition: S could not easily have believed that p is false.23

20
See Sartwell (1991, 1992). For critical discussion, see Le Morvan (2002). Others, such as
Goldman (2002a, 2002c) and Goldman and Olsson (2009), have argued that there is
a weak sense of knowledge according to which it is nothing more than true belief. For
critical discussion, see Le Morvan (2005, 2010).
21
See Nozick (1981). For critical discussion see DeRose (1995).
22
See Plantinga (1993a, 1993b). More fully stated: The Plantingan Warrant Condition
specifies that a belief B is warranted only if B is produced by a properly functioning
cognitive faculty in an environment fitting for that kind of cognitive faculty and is
governed by a designed plan aimed at the production of true beliefs and that there is
a high statistical probability that beliefs so produced will be true. For critical discussion,
see the essays in Kvanvig. (1996).
23
See Sosa (1999). For critical discussion, see Comesaña (2005).
 The Nature of Ignorance: Two Views 21

Accepting any of these alternative accounts of factive knowledge leads to

alternative complementary accounts of factive ignorance: in addition to
its alethic and doxastic kinds, factive ignorance could result from a failure
of truth-tracking, from a failure of proper functioning of cognitive equip-
ment, or from a failure to safely believe that p.
Suppose that factive knowledge that p is not a species of true belief but
rather an ability to “manifest various accurate representations of p” con-
cerning the “epistemic diaspora” of p such as being able to answer ques-
tions accurately concerning p, to reason accurately concerning p, and to
correctly use concepts related to p.24 Accepting such an account of factive
knowledge leads to a complementary account of factive ignorance accord-
ing to which such ignorance arises from the failure to manifest various
accurate representations of p concerning its epistemic diaspora.
Suppose that knowledge-first epistemology is correct and factive
knowledge that p is, therefore, not analyzable at all but rather the most
basic factive propositional attitude.25 If so, then the complementary
account of factive ignorance is simply that to be ignorant that p is to not
know that p where such knowledge is the most basic factive propositional
attitude.
In light of the considerations above, we have seen how the Standard
View yields a unified account of knowledge and ignorance; insofar as such
unification proves attractive, it counts in favor of the Standard View.
Worth noting in this context as well is that, by unifying theorizing about
knowledge and ignorance, the Standard View avoids what may strike
many as an implausible consequence of the New View, namely that
merely having a true belief that p suffices for not being factively ignorant
that p. Suppose for example that Sam in 2016 believes that an odd perfect
number exists (that is, an odd number that is half the sum of all of its
positive divisors including itself), and does so quite irrationally, or very
unjustifiably, or as a result of a highly unreliable doxastic process. At this
juncture in mathematical history, whether an odd perfect number exists
remains an unsolved problem in number theory; suppose though that it
is proven in 3016 that such a number exists. The New View entails that
Sam in 2016 is not factively ignorant that there is an odd perfect number,
a consequence liable to strike many as quite implausible.26 The Standard

24
See Heatherington (2011). For critical discussion, see Madison (2012).
25
See Williamson (2000). For critical discussion, see the essays in Greenough and
Pritchard (2009).
26
Of course, Sam in 2015 was not ignorant of the proposition (proven true in 3016) that there is
an odd perfect number, but as discussed in the next section, knowledge of a true proposition
is distinct from knowledge that it is true.
22 Pierre Le Morvan and Rik Peels

View (provided that knowledge that p is more than mere true belief that p)
carries no such entailment.

3.3 Third Argument: Ignorance of Falsehoods

Can there be ignorance of falsehoods? We consider in this section a case
for maintaining that there can be. The Standard View is fully compatible
with such ignorance, and insofar as the New View is not, this provides
further support for the Standard over the New View.
Propositions have truth-conditions. These truth-conditions can be
distinguished from their satisfaction.27 A proposition is true when its
truth-conditions are satisfied, and, assuming bivalence, false when not.
Consider the following three propositions:
(3) Damascus is north of Jerusalem.
(4) Mars is larger than Jupiter.
(5) An odd perfect number exists.
The first of these propositions is true. The second is false. The third is
presumably true or false, but, as noted in the previous section, at this
juncture in mathematical history we are not in a position to ascertain its
truth-value.
In order to be in a position to know that, believe that, desire that, or
doubt that a proposition is true (that is, that its truth-conditions are
satisfied), one cannot be ignorant of the proposition itself and its con-
comitant truth-conditions. Take, for instance, King Herod I (73 BCE–4
BCE), and consider the following two propositions:
(6) The most popular paid iOS app of 2014 was Heads Up!
(7) The most popular paid iOS app of 2014 was NOAA Hi-Def
Radar.
It turns out that (6) is true while (7) is false.28 In any event, being ignorant
of these propositions, King Herod, since he lived more than 2000 years
ago, was in no position to have any propositional attitude toward them.
As such, he was ignorant not just that (6)’s truth-conditions are satisfied
and that (7)’s truth-conditions are not, but in the even deeper sense of

27
Whatever else they may be, it is widely agreed that propositions have truth-conditions.
Whether they are nothing but their truth-conditions is more controversial, and not an
issue that can be addressed here. For discussion on the ontology of propositions in
relation to knowledge and ignorance of them, see Le Morvan (2015).
28
www.apptrigger.com/2014/12/28/best-2014-popular-paid-ios-apps/
 The Nature of Ignorance: Two Views 23

being ignorant of (6) and (7) themselves. King Herod I could not even
grasp these propositions.
This example involving Herod, the adherent of the Standard View
might suggest, illustrates how ignorance of propositions is not restricted
to those that are true, and exemplifies how one can be ignorant of false
ones too. In Herod’s case, he lacked the conceptual repertoire to grasp not
only true proposition (6) but also false proposition (7). Notice also that,
even if one has the conceptual repertoire for having an attitude toward
a proposition, one can still be ignorant of that proposition if one has not
deployed this repertoire. Suppose, for instance, that President Obama has
the conceptual repertoire to grasp (6) and (7), but has never deployed this
repertoire. If so, one might think in a certain sense he is ignorant of these
propositions.29
Turning now to knowledge of propositions as the complement of
ignorance of them, the following points merit attention. First, adherents
of the Standard View can argue that, although one can have knowledge of
true propositions – e.g., Pierre Le Morvan has knowledge of (1) – one can
also have knowledge of false ones – e.g., Pierre Le Morvan has knowledge
of (2). Second, one can have knowledge of propositions one does not
believe. For instance, Pierre Le Morvan has knowledge of (3), but does
not believe it. Thus, one might think that knowledge of a proposition
p does not share the putative necessary conditions of knowledge that p in
terms of believing that p, p’s being true, and having warrant for believing
that p.30
In light of the reasoning above and the standard three-fold distinction
that we made in Section 2 between factive, objectual, and procedural
knowledge, we can conclude that the complement of ignorance of
a proposition is not factive knowledge (knowledge that p). This leaves
us with objectual and procedural knowledge. It is far from clear how this
complement could be a form of procedural knowledge.31 So we are left
with objectual knowledge: the complement of ignorance of a proposition
is an acquaintance with or knowledge of an entity, where the entity in

29
Ignorance of a proposition can come in pre-conceptual and post-conceptual forms: if one
lacks the conceptual repertoire requisite for having an attitude regarding a proposition,
one is pre-conceptually ignorant of it, whereas one is post-conceptually ignorant of
a proposition if, though having this conceptual repertoire, one has not deployed it so as
to have such an attitude.
30
“Warrant” is meant in the neutral sense for whatever it is that turns true belief into
knowledge, and not in the particular sense of the specific account of warrant found in
Plantinga (1993a, 1993b).
31
A general point that extends beyond the question of the knowledge of propositions is that
procedural and objectual knowledge seem to be different kinds of knowledge, and it is not
evident that objectual knowledge is reducible to procedural knowledge.
24 Pierre Le Morvan and Rik Peels

question is a proposition. Such knowledge is not equivalent to knowledge

that p, for although the latter entails knowledge of p since knowledge of p is
a necessary condition for knowledge that p,32 knowledge of p does not
entail knowledge that p since knowledge of p is not a sufficient condition
for knowledge that p.33 Thus, knowledge that p should not be equated
with knowledge of p.34
Accordingly, on the Standard View, the complement of ignorance of
a proposition p is not knowledge that p. Its complement is rather knowl-
edge of p—an acquaintance with or knowledge of an entity, where the
entity in question is a proposition.35 Such objectual knowledge may be
occurrent (as when one is conscious of it) or dispositional (as when one
retains it in memory), and occurs only if the knower has the concepts to
grasp or comprehend the proposition. Knowledge of p is required to
have – and is therefore entailed by, and a precondition of – knowledge
that p, but also for having any propositional attitude concerning p such as
believing that p, considering that p, doubting that p, hoping that p, or
entertaining that p.
While the Standard View is fully compatible with the ignorance of
falsehoods, insofar as the New View is supported by the idea we can
only be ignorant of truths (and therefore not of falsehoods),36 the above
case for the ignorance of falsehoods supports the Standard over the New
View.
Allowing for the ignorance of falsehoods also enables the Standard
View to avoid a difficulty (at least an apparent one) the New View runs
into. This difficulty arises when we consider that the New View defines
factive ignorance as the lack or absence of true belief. Now a natural or
intuitive understanding of a lack or absence of something x is that it is
the complement of a presence of x; thus, if the presence of S’s true belief
that p is <S believes that p and p is true>, then its absence is understood as
~<S believes that p and p is true> and this in turn is logically equivalent to
<~S believes that p or ~p>. On this understanding of the absence of true
belief then, ~p (namely, p’s being false) is a sufficient condition for the
absence of true belief. Accordingly, if adherents of the New View were to
accept this understanding of the absence of true belief, they would also

32
Someone who is ignorant of p cannot know that p. See Le Morvan (2015).
33
Just because someone S knows of p, it does not follow that p is true, or that S believes that
p, or that S’s believing that p (if S does so) meets the anti-Gettier condition.
34
This non-equivalence holds even if p is true, for even in such a case, knowledge of p is
necessary but not sufficient for knowledge that p. Accordingly, knowledge that
a proposition is true should not be equated with knowledge of a proposition that is true.
35
See Le Morvan (2015) for a discussion of leading conceptions of the ontology of
propositions and their epistemic implications.
36
See Peels (2010, 2011a, 2012).
 The Nature of Ignorance: Two Views 25

have to accept that p’s being false is a sufficient condition for factive
ignorance, and this contradicts the idea that we cannot be ignorant that
p if p is false. To avoid this conclusion, adherents of the New View must
provide an account of the absence of true belief on which the absence of
true belief is not the complement of the presence of true belief, and this
may strike many as an intuitively odd or unnatural way of construing this
absence.37 On the Standard View, by contrast, the absence of true belief is
straightforwardly the complement of the presence of true belief, and so we
can be, and in fact are, ignorant that p if p false.
One final related point concerning falsehood. An intuition invoked in
support of the New View is that “ignorant that p” has the conversational
implicature that p is true.38 Thus, saying that S is ignorant that
p conversationally implies that p is true. Does this show that we cannot
be ignorant that p if p is false? Adherents of the Standard View can resist
this conclusion by pointing out that conversational implicature is
a matter of what is suggested or implied by an expression, not a matter
of what it strictly speaking expresses. Moreover, “does not know that p”
can also conversationally imply that p is true; for instance, to say “Pam
doesn’t know that platypuses are monotremes” can conversationally
imply that platypuses are monotremes. But just because “does not
know that p” can have this conversational implicature, it does not follow
that the following widely held epistemological tenet is false: that p’s
being false is a sufficient condition for not knowing that p. Mutatis
mutandis for “is ignorant that p.”39

4 Ignorance as the Lack of True Belief

In the previous section, we considered ignorance as the absence or lack of
knowledge. In this section, we consider an alternative conception of
ignorance, namely ignorance as the absence or lack of true belief.
If ignorance is the absence or lack of true belief, then none of the cases
that are cases of true belief that fall short of knowledge are cases of

37
Peels (2012, p. 743) has proposed that the absence of true belief can be understood
instead as <S does not believe that p and p is true>. Note that this entails that the absence
of true belief cannot be the complement of the presence of true belief, for <S does not
believe that p and p is true> is not logically equivalent to ~<S believes that p and p is true>.
Which of these ways of understanding the absence of true belief makes the most sense is
a question we leave to our readers.
38
See Peels (2010, 2011a, 2012).
39
A critic of the Standard View might counter that “is ignorant that p” always has the
conversational implicature that p is true, while “does not know that p” does not always
have it. Whether this claim is true cannot be settled here and would be a good subject for
linguistic inquiry and experimental philosophy.
26 Pierre Le Morvan and Rik Peels

ignorance – whether they are cases of Gettierized justified true belief,

mere justified true belief, or even mere true belief. Since, according to
adherents of the New View, one can be ignorant that p only if p is true, it
follows that there are three kinds of ignorance:
(i) Disbelieving ignorance: one disbelieves that p while p is true.
(ii) Suspending ignorance: one suspends belief and disbelief on p while p is
true.
(iii) Deep ignorance: one neither believes nor disbelieves nor suspends
belief and disbelief on p while p is true.
A similar view is adopted by René van Woudenberg:

S is ignorant with respect to p, when (iiia) S neither believes nor disbelieves p, even
though he has entertained p (rational ignorance). (iiib) S never so much as
entertained p and accordingly neither believes nor disbelieves p (deep ignorance).
(iv) S has the false belief that not-p. Each of these conditions is sufficient for
ignorance. There is a way to connect and summarize the three sufficient condi-
tions for ignorance by saying, as Alvin Goldman has done, that ignorance is “the
absence of true belief”; after all, each of these conditions entails the absence of
true belief.40

We should note that a fully spelled-out version of the New View will have to
add several caveats to a rough analysis of ignorance along these lines. For
one thing, it seems possible that one disbelieve the true proposition p, but
know that, epistemically speaking, one really ought to have the attitude of
belief toward p, but that, quite irrationally, one is unable to do so for
psychological reasons. It is not at all clear that such a case will count as
a case of ignorance that p, even though that is what the rough and ready
version of the New View as presented above implies. However, since the
aim of this paper is to sketch some important arguments for the Standard
and New Views, we will leave such details for another occasion.

4.1 First Argument: Intuitions About Cases of True Belief that Fall Short
of Knowledge
In order to sketch the first argument for the New View, let us consider
some of the ways in which one can believe truly that p and yet fail to know
that p and then consider whether they count as being ignorant that p.
Let us start with cases that just fall short of knowledge, such as Gettier-
cases. An adherent of the New View might suggest that they do not seem
to be cases of ignorance. Here is an example that can be used to illustrate
the point. Imagine that Sam enters his living room and that he looks at the
40
Van Woudenberg (2009, p. 375).
 The Nature of Ignorance: Two Views 27

clock. The clock tells him that it is 7 p.m., so Sam comes to believe that it
is 7 p.m. He knows that the clock normally works perfectly fine. However,
unbeknownst to him, the clock stopped working twenty-four hours ago.
Is Sam ignorant that it is 7 p.m.? It seems, the adherent of the New View
might suggest, that it is implausible that Sam is ignorant in such a case.
Of course, there are other propositions of whose truth Sam is clearly
ignorant, such as that the clock stopped working twenty-four hours ago
and that the clock is unreliable on this particular occasion. Of the truth of
the proposition that it is 7 p.m. itself, however, Sam does not seem, to
adherents of the New View, to be ignorant.
Next, they might suggest that even cases of mere true belief do not seem
cases of ignorance. Consider Alfred from Columbia, Missouri who
believes contrary to all the evidence that he is going to be the next
president of the United States. He thus comes to believe the proposition
q that the next president of the United States currently lives in Columbia,
Missouri. As it turns out, the next president is Ms. Howard, a female
member of Congress living in Columbia, Missouri, whom Alfred has
never even heard of. In this case Alfred believes truly, but does not
know that q. Is he ignorant that q is true? It seems to adherents of the
New View that he is not. Again, there are all sorts of truths in the
neighborhood that he is ignorant of and it is hard mentally to isolate
q from all those others truths, truths such as Ms Howard is going to be the
next president of the United States, Ms Howard lives in Columbia, and
The next president is currently a member of Congress. We may be inclined to
think that Alfred is ignorant that q is true because we know that he is
ignorant of all these other propositions. If we focus on q, however, it
seems – so the argument goes – that Alfred is not ignorant of q.
Now, let us assume with most epistemologists that knowledge is true
belief that satisfies some further conditions in order to provide an anti-
Gettier codicil. If both cases of true belief that just fall short of knowledge,
like Sam’s case, and cases of mere true belief, such as Alfred’s case, do not
count as cases of ignorance, then in-between cases will probably not
count as cases of ignorance either. Here is why. If, on the one hand,
such cases had a property that would make them cases of ignorance, then
it seems to adherents of the New View that knowledge would also have
that property and, therefore, be a case of ignorance. If, on the other hand,
such cases lacked a property that would make them cases of ignorance,
then it seems mere true belief would also lack that property and, therefore,
be a case of ignorance. Thus, if both cases of mere true belief and cases of
true belief that just fall short of knowledge are not cases of ignorance, then
we can safely assume that in-between cases are not cases of ignorance
either.
28 Pierre Le Morvan and Rik Peels

Note that each of the three steps of this argument needs to be successful
in order for this argument to provide a good reason to embrace the New
View, but that not all of them have to be successful in order for it to
provide a good reason to reject the Standard View. If, for instance, mere
true belief is a case of ignorance, but true belief that just falls short of
knowledge is not, then the Standard View is false: ignorance is not the lack
of knowledge. This means, of course, that there is a variety of potential
views regarding the nature of ignorance that one might be willing to
defend even if one thinks the Standard and New Views are mistaken,
such as the view that ignorance is lack of Gettierized true belief, that it is
lack of reliably produced true belief, that it is lack of true belief based on
sufficient evidence, and so forth. Or one might defend a particular version
of contextualism, arguing that which of these things ignorance is depends
on the context of the cognitive subject in question. For the sake of clarity,
in this paper we confine ourselves to the Standard and New Views.

4.2 Second Argument: Ignorance Excuses

Ever since Aristotle, it has been widely acknowledged among philoso-
phers that ignorance – at least as long as it is blameless and if it is ignorance
of the right kind of proposition – provides an excuse for wrong actions or
omissions for which one would otherwise be blameworthy.41 Here is an
example. Imagine that it is Claire’s birthday and that Sam decides to bake
a chocolate cake for her. When Claire is away from the kitchen for
a second, a jealous cousin poisons the cake. After Sam returns, he finishes
preparing the cake and offers it to Claire, entirely ignorant that the cake
has been poisoned. It seems clear that in such a case – again, as long as his
ignorance is blameless – Sam is excused for offering Claire the poisoned
chocolate cake and that it is his ignorance of the fact that the cake is
poisoned that excuses him.
In some cases, ignorance counts as a full excuse: it removes all blame-
worthiness. In other cases, it is merely a partial excuse: it reduces the
degree of one’s blameworthiness, but it does not block blameworthiness
altogether.42 If, for instance, Sam suspends judgment on whether the
chocolate cake is poisoned and still gives it to Claire, he is less blame-
worthy than if he is aware (believes truly) that the cake is poisoned, but

41
See Aristotle (2003, pp. 123–129; pp. 145–147; pp. 299–305) (NE III.i.13–27; v.7–12;
V.viii.3–12). For more recent examples, see Brandt (1969, p. 349), Fischer and Ravizza
(1998, pp. 12–13); Goldman (1970, p. 208), Rosen (2003, pp. 61–62); Smith (1983,
pp. 543–571), Zimmerman (2008, pp. 169–205).
42
For some examples, see Peels (2014).
 The Nature of Ignorance: Two Views 29

Sam is still blameworthy: in this case, he should not have offered Claire
the chocolate cake.
The point of the second argument in favor of the New View is this: any
kind of true belief that falls short of knowledge does not excuse. It does not
even provide a partial excuse. However, ignorance (as long as it is blame-
less) excuses. It follows, by a simple modus tollens, that ignorance cannot
be the lack of knowledge. The thesis that ignorance is lack of true belief, so
the argument goes, is more plausible than the view that ignorance is the
lack of knowledge, since each way in which one can lack a true belief –
disbelieving ignorance, suspending ignorance, and deep ignorance –
again, as long as it is blameless, seems to provide at least a partial excuse.
Let us elaborate on the earlier example to illustrate the point. Sam is in
a situation in which he has baked Claire a birthday chocolate cake and he
can give it to her or not. It seems to adherents of the New View that Sam’s
knowing that it is poisoned or his merely truly believing that it is poisoned
does not make any difference to the degree of his blameworthiness: in
both cases he is blameworthy to an equally high degree and he is not at all
excused. For whether he knows or rationally believes or merely believes
that the cake is poisoned does not make an important difference to his
phenomenology. In all these cases, he sincerely thinks that the cake is
poisoned; that is how reality appears to him.43
If, as many epistemologists believe, there are degrees of belief and if
degrees of belief are to be spelled out in terms of conviction, then maybe
one is more blameworthy if one is certain that the cake is poisoned than if
one is merely fairly convinced that the cake is poisoned. Notice, though,
that such varieties in degree of belief are not necessarily correlated with
whether one knows, believes on strong evidence, believes on weak evidence, or
believes without any evidence. One could in principle, quite irrationally, be
one hundred percent sure without having any reasons or evidence. Thus,
even though the degree to which one believes that, say, the chocolate cake
is poisoned may make a difference to the extent to which true belief
excuses, whether one knows or justifiedly believes, and so on, whether it
is poisoned does not make a difference to that.
Adherents of the New View would stress that the suggestion here is not
that a true belief that the cake is poisoned renders one blameworthy to the
highest degree possible. Maybe someone who believes truly that the choco-
late cake is poisoned and gives it to one’s friend in order to do wrong for
wrong’s sake is even more blameworthy than someone who does so
43
Given this argument, adherents of the Standard View need to develop a case for holding
that one’s being excused is a function of more than one’s phenomenology or degree of
conviction relative to a belief (or, as we point out below, argue that not all ignorance
excuses).
30 Pierre Le Morvan and Rik Peels

merely because she is scared of the poisoning cousin.44 In such cases,

however, it seems that one’s evil intention adds something to the degree of
one’s blameworthiness. Whether one believes or knows that the chocolate
cake is poisoned makes no difference to the degree of one’s blameworthi-
ness: in both cases, one is not excused at all, not even partially.
Since, as we said, ignorance is widely acknowledged to count as an
excuse, whereas it seems that true belief that fails to be knowledge does
not, ignorance, one might think, cannot be absence of knowledge. What
the discussion of ignorance in this section suggests is rather that ignorance
is the lack of true belief.
Of course, one could simply propose to revise the widespread view that
blameless ignorance excuses and say that most varieties of blameless
ignorance excuse, but that some varieties of ignorance, such as blameless
mere true belief and blameless mere justified true belief, do not. The New
View, though, provides a unifying account of ignorance as an excuse, for it
implies that all blameless ignorance provides at least a partial excuse and
this captures the intuitions about excusing ignorance that, it seems, are
widespread among philosophers.

4.3 There Is No Ignorance by Acquaintance or Procedural Ignorance

A third and final consideration in favor of the New View – one that cuts
against the idea that the Standard View provides a unified account of
ignorance – is the following. On the main rival view, the Standard View,
ignorance is lack of knowledge. Given that there is not only factive knowl-
edge, but also objectual knowledge and procedural knowledge, and given
that people lack objectual knowledge and procedural knowledge with
regard to many things, it would follow that people have objectual ignor-
ance and procedural ignorance with regard to countless (the vast majority
of) things. But that, one might think, seems problematic. Imagine that the
following sentence is true:
(8) Xavier knows Paris very well, because he has lived there for
more than twenty years.
This is obviously a case of objectual knowledge: Xavier is familiar with
Paris, since he has lived there for such a long time. If he has objectual
knowledge of Paris and if the Standard View of Ignorance is correct,
though, it would follow that he lacks objectual ignorance of Paris. But it
seems to adherents of the New View that it is not correct to say:

44
Thus, for instance, Beardsley (1979, p. 577).
 The Nature of Ignorance: Two Views 31

(8 0 ) Xavier is not at all ignorant of Paris, because he has lived

there for more than twenty years.
Something similar seems to apply to another kind of objectual knowledge,
namely knowledge of people. It is perfectly fine to say:
(9) She knows Albert since she moved to Oxford.
But again it seems to adherents of the New View that we would not say:
(9 0 ) She is not ignorant of Albert since she moved to Oxford.
When it comes to such things as cities or countries and persons, there
seems to adherents of the New View to be the absence or presence
of objectual knowledge, but no such thing as the absence or presence of
objectual ignorance. This is not to deny that we do use the expression
“ignorance of X” in at least some cases in which X is not a person, a city
a country or some such thing. We say, for instance:
(10) Marcel is ignorant of quantum physics.
(11) I was ignorant of their plans for the summer.
Adherents of the New View might suggest, though, that these are not cases
in which someone is claimed to suffer from objectual ignorance. What the
person in question seems to be ignorant of, one might think, rather seems
to be a set of propositions – those constitutive of quantum physics and
those that constitute a particular group of persons’ plans for the summer.
One might claim that something similar applies to expressions like “ignor-
ant as to how to φ” and “ignorant of how to φ,” such as they occur in the
following sentences:
(12) They are ignorant as to how to escape from that prison.
(13) Fred is ignorant of how change a car’s tire.
Arguing that the knowledge referred to in (12) and (13) is reducible to
factive knowledge, that the truth of these sentences does not require there
to be uniquely procedural knowledge, and to argue that something similar
applies to other cases would be needed to complete the argument, but is
beyond the scope of this paper. Again, if certain cases of objectual knowl-
edge do not have objectual ignorance as their contradictory or comple-
ment, the Standard View is in trouble, but it does not follow that the New
View is correct; more work would be needed for that, namely that every
case of seemingly objectual or procedural ignorance is in fact a case of factive
ignorance.
32 Pierre Le Morvan and Rik Peels

5 Epilogue
In this chapter, we have explored the nature of ignorance. In doing so, we
have discussed two main rival views of its nature, namely the Standard
and New Views, and three main arguments adduced in support of each.
As we shall see in some of the ensuing chapters, which view one takes on
the nature of ignorance can make a significant difference to how one
thinks of other philosophical issues. As Michael Blome-Tillmann points
out, the Standard View implies that if contextualism about knowledge is
correct, then so is contextualism about ignorance, whereas the New View
implies that even if contextualism about knowledge is correct, it does not
follow that contextualism about ignorance is correct. Justin McBrayer
argues in Chapter 8 that one’s understanding of the nature of ignorance
makes a significant difference to several issues in the philosophy of reli-
gion, such as whether or not a large number of people are ignorant of
God’s existence. Some of these chapters also provide considerations that
bear on the controversy between the Standard and New Views. Berit
Brogaard, for instance, argues in Chapter 3 that being ignorant of how
to do something is a special case of being ignorant of a fact – something
that an adherent of the New View might gladly embrace.
Since these are just a couple of examples, we conclude that more work
needs to be done, both on considerations in favor and against the
Standard and New Views, and on spelling out the ramifications of each
of these views in other areas of philosophy.45

45
We thank Nikolaj Nottelmann and René van Woudenberg for their helpful comments on
earlier versions of this paper.
7 Ignorance and Epistemic Value

Duncan Pritchard

1 Epistemic Value/Disvalue
Recent years have seen a huge upsurge of interest in the topic of epistemic
value, particularly with regard to the value of knowledge in contrast to
other positive epistemic standings such as justified belief and
understanding.1 The questions raised for positive epistemic standings
such as knowledge can, however, equally be posed with regard to negative
epistemic standings such as ignorance, which we will simply take to be the
lack of knowledge.2 Interestingly, as we will see, it does not follow from
the fact that ignorance is a negative epistemic standing that it is thereby
a disvaluable epistemic standing.3 For just as we can imagine positive
epistemic standings being sometimes disvaluable, so we can likewise
conceive of negative epistemic standing being valuable.
Before we get to this point, however, we first need to flag an ambiguity
in the very notion of epistemic value, one that is often overlooked but that
is, as we will see, very important to evaluating the putative epistemic value
of ignorance. The most natural way to understand the notion of epistemic
value is as picking out a particular kind of value that is distinctively
epistemic, just as we might suppose that aesthetic value picks out
a particular kind of value that is distinctively aesthetic. But there is also
a secondary usage of this notion in the literature, often not kept apart from
the first, whereby it is taken to also cover the value of a particular epis-
temic standing, whether that value is distinctively epistemic or otherwise.
1
As Riggs (2008) has put it, this reflects the so-called ‘value turn’ in epistemology. For some
key works on epistemic value, see Jones (1997), Kvanvig (2003), Zagzebski (2003), Greco
(2009) and Pritchard, Haddock and Millar (2010). See also Pritchard (2009, 2011, 2014,
forthcoming b). For two useful recent surveys of the literature on epistemic value, see
Pritchard (2007) and Pritchard and Turri (2011).
2
For a prominent recent defence of this account of ignorance, see Le Morvan (2011). For
a spirited defence of an alternative conception of ignorance, such that ignorance is lack of
true belief rather than knowledge, see Peels (2012).
3
Note that, for the sake of simplicity, I will here treat value and disvalue as exhaustive
categories. Thus, a lack of positive value will equate to a disvalue. This is, of course,
contentious, but I don’t think anything of substance hangs on this point for our purposes.

132
 Ignorance and Epistemic Value 133

It is common in the literature, for example, to explore the ‘epistemic

value’ of knowledge by appealing to its practical utility.4 But since no
one thinks that practical utility is a distinctively epistemic kind of value, it
is clear that we are here using the phrase ‘epistemic value’ in an impor-
tantly different way.5 In particular, ‘epistemic value’ here means not
a distinctively epistemic kind of value but rather instead the value of the
epistemic (which may itself be either distinctively epistemic or
otherwise).6 Henceforth, we will keep these two notions of epistemic
value apart, and do so by only using ‘epistemic value’ to refer to the
distinctively epistemic kind of value.
One reason why this distinction is so important to the debate about the
value of epistemic standings is that while it is very easy to conceive of
positive epistemic standings that sometimes have negative value – and,
indeed, negative epistemic standings which sometimes have positive
value – it is not so easy to conceive of positive epistemic standings that
sometimes have negative epistemic value (and, likewise, mutatis mutan-
dis, for negative epistemic standings). Take knowledge, for example.
Coming to know that one was adopted may cause one great unhappiness,
such that one wished one had never found out. In such a case, one might
reasonably regard this knowledge as highly disvaluable from the practical

4
A good example of this can be found in Goldman and Olsson (2009).
5
Of course, there are those in the literature who argue that there is no hard-and-fast
distinction to be drawn between practical and epistemic value, and there are also those
who maintain that practical factors can have a significant impact on epistemic standing,
but these views are some remove from the claim that practical value just is a variety of
distinctively epistemic value. For an example of the former, see Wright (2008), who offers
a conception of epistemic normativity that draws on both classically epistemic and pru-
dential considerations. For an example of the latter, see the literature on pragmatic
encroachment in epistemology. A good overview of this literature is offered in Fantl and
McGrath (2010).
6
I draw this distinction, and explain its import to the debate about the epistemic value, in
a number of places. See, for example, Pritchard, Millar and Haddock (2010, chapters 1–
4), and Pritchard (2011, 2014, 2015, forthcoming b). Note that there is a related distinc-
tion in the vicinity here – due to Geach (1956) – between what he calls ‘predicative’ as
opposed to ‘attributive’ expressions. Here is an example that he uses to illustrate the
distinction. To say that a fly is big is to say that it is big for a fly; it is not to say that it is both
a fly and big (i.e., big simpliciter). This is thus an attributive expression. In contrast, to say
that a book is red is not to say that it is red for a book, but rather to say that it is both red and
a book. This is a predicative expression. The contrast that Geach makes relates to the
distinction that we have drawn since in effect we can read the expression ‘epistemic value’
either attributively (i.e., as pertaining to a kind of value which is specifically epistemic) or
predicatively (i.e., as pertaining to something which is both epistemic and valuable). That
said, note that one should be wary about completely equating the distinction we have
drawn with Geach’s distinction, in that when we talk of the ‘value of the epistemic’ we
leave it open that it may be a particularly epistemic kind of value which is at issue – the
point is rather that this is not being demanded, as it is when ‘epistemic value’ is being read
attributively.
134 Duncan Pritchard

point of view of what promotes, or undermines, one’s own happiness. But

that this knowledge is practically disvaluable in this way does not entail
that it lacks epistemic value, as this is to evaluate that knowledge along an
entirely different axis of evaluation.
This issue is important to our discussion of ignorance, since if ignor-
ance is lack of knowledge, and knowledge can sometimes be disvaluable,
then it follows that ignorance can sometimes be valuable.7 In the case just
offered, for example, concerning the knowledge that one is adopted, it
would be more valuable to have not known that this was the case. Thus,
ignorance of this fact will be valuable. But since the kind of value in play
here is just practical value, this is not yet to say that ignorance is episte-
mically valuable. Accordingly, if we want to argue that ignorance can have
epistemic value, we will need to supply additional grounds.

2 The Epistemic Efficacy of Ignorance

Let us grant for the sake of argument that positive epistemic standings such
as knowledge are generally both epistemically valuable and valuable sim-
pliciter. There is of course a lively debate in epistemology about whether
this really the case, and in particular about the different relative value,
epistemic or otherwise, of particular epistemic standings (e.g., the value
of knowledge as opposed, say, to understanding). But it would take us too
far afield to get into these issues here.8 If the claim that knowledge is
generally both epistemically valuable and valuable simpliciter is true, then
ignorance, qua lack of knowledge, will generally be both epistemically
disvaluable and also disvaluable simpliciter. We have already noted, how-
ever, that the general value, epistemic or otherwise, of knowledge, does not
exclude cases in which knowledge is disvaluable, as in the case described in
the last section. And where knowledge is disvaluable, so ignorance will be
valuable. We noted too, however, that the kind of value/disvalue in play
here is not specifically epistemic. So if we want to defend the more inter-
esting claim that ignorance can have epistemic value, then we have further
work to do. On the plus side, we have identified one way of determining
such value, which is to look for cases in which knowledge is epistemically
7
Is ignorance just lack of knowledge? While I think this is broadly correct, there are some
difficult cases involving well-grounded true belief which does not amount to knowledge
(e.g., which has been Gettierized). Would we really class such a person as ignorant? I am
not so sure, though we clearly would class them as lacking knowledge. In any case, in order
to keep our discussion manageable I will set this complication to one side I what follows.
For a helpful recent discussion of these kinds of cases, and the problems they pose for the
view of ignorance as lack of knowledge, see Peels (2012).
8
I explore these issues at length in Pritchard, Millar and Haddock (2010, chapters 1–4). See
also Pritchard (2007) and Pritchard and Turri (2011).
 Ignorance and Epistemic Value 135

disvaluable, since they will be instances in which ignorance, qua lack of

knowledge, is epistemically valuable.
Before we can do that, however, we first need to say something about
what epistemic value is, specifically. For our purposes, we can character-
ise epistemic value in terms of the distinctive truth-directed goal char-
acteristic of good inquiry. In short, our distinctively epistemic goal is to
get at the truth, and so, in the broadest terms, what promotes this goal has
epistemic value. The acquisition of true belief, and the avoidance of false
belief, are thus two core epistemic goods which determine epistemic
value, in that whatever promotes truth in one’s beliefs, and the avoidance
of error, will have epistemic value. So, for example, this is why it is
epistemically valuable to have good evidence in support of one’s beliefs,
since good evidence is a guide to the truth (i.e., will lead to true beliefs),
and will generally steer you away from error (i.e., away from false beliefs).
More generally, whatever promotes not just true belief and the avoid-
ance of error, but also accuracy in one’s propositional attitudes in gen-
eral – where applicable anyway – will be epistemically valuable. So, for
example, consider the propositional attitude of accepting that p.
Accepting that p comes apart from believing that p at least to the extent
that one can accept a proposition without believing it.9 For example,
a scientist working in a highly controversial and theoretical domain
might accept a certain theory because she recognises that it is by far the
best current theory available even though she is sufficiently unsure of its
truth that she does not actually believe it. Just as we would want a good
inquirer’s beliefs to be responsive to the truth-relevant considerations
available to her, so we would want her acceptances to be similarly respon-
sive to these considerations. In this case, for example, the scientist should
not accept the theory unless the available evidence really does pick out
that theory as the best available.10 That said, henceforth in what follows
we will focus, for the sake of simplicity, on the goals of promoting true
belief and avoiding false beliefs.
We can now rephrase our question about whether there are cases in
which possessing knowledge is epistemically disvaluable in terms of
whether such an epistemic standing can either lead one to error or at

9
On most views, acceptance and belief are also different in that they come apart in the
other direction too. Since all that is important for present purposes is that they are not the
same propositional attitude, we can set this further difference to one side. For a classic
discussion of the notion of acceptance in contrast to belief, with specific reference to the
context of philosophy of science, see van Fraassen (1980). See also Cohen (1992).
10
Of course, one option here is to treat our epistemic evaluation of what the scientist
accepts as being derivative on the goal of promoting true belief and avoiding false belief,
in that epistemically good acceptances are precisely those acceptances which promote
these epistemic goals. Thanks to Rik Peels for pressing me on this point.
136 Duncan Pritchard

least prevent one from gaining true beliefs. So construed, there is one
straightforward type of case which fits the bill – misleading defeaters.
A defeater is a consideration that undermines one’s knowledge. So, for
example, finding out that one has recently ingested a hallucinogenic drug
can undermine one’s perceptual knowledge about one’s environment,
since it follows that the deliverances of one’s perceptual faculties are no
longer reliable. A misleading defeater is a specific kind of defeater which,
as the name suggests, points one away from the truth rather than towards
it. So, for example, being falsely told that one has recently ingested
a hallucinogenic drug will no less undermine one’s perceptual knowledge
about one’s environment than being truly told that this is the case.
Crucially, however, in the former case, since one has not in fact ingested
the hallucinogen, then it follows that one is not in fact forming one’s
perceptual beliefs unreliably. Defeaters, whether misleading or otherwise,
defeat one’s knowledge until they are in turn defeated by further evidence
(e.g., finding out that one hasn’t in fact ingested the hallucinogen, or else
discovering that while one has ingested this hallucinogen, it has been
ineffective in this case).
What is interesting about misleading defeaters is that while one cannot
rationally ignore them once one is made aware of them, there is a perfectly
good sense in which one is epistemically better off if one does not come into
contact with them. Consider, for example, two identical agents who occupy
otherwise identical environments except that only agent one is regularly
subject to misleading defeaters. Imagine, for example, that both agents are
plagued by an ‘epistemic joker’ who keeps ensuring that there are misleading
defeaters in play – who, for instance, puts it about that our agent has ingested
an hallucinogen when in fact she has not, and so forth. The difference
between the two agents, however, is that only agent one is in fact affected
by the epistemic joker, and hence has her knowledge defeated. In contrast,
agent two is never affected by the epistemic joker because she is in addition
protected by an ‘epistemic helper’ who, knowing that the epistemic joker is
planting these misleading defeaters, ensures that agent two never encounters
them. For example, she ensures that agent two doesn’t receive the mislead-
ing testimony that she has ingested a hallucinogen, and so continues to
(rightly) trust her perceptual faculties as before.
Here is the crux of the matter: agent two seems to be epistemically
better off than agent one, in that she knows lots of things that agent one,
who is subject to the misleading defeaters, is unable to know.11 Crucially,
11
Though note that in saying that agent two is epistemically better off than agent one, we
are in effect buying into a certain conception of epistemic value, as will be explained
below. After all, there is a sense in which agent one does know some propositions that
agent two doesn’t know (e.g., about the presence of the defeater). As explained below, the
 Ignorance and Epistemic Value 137

however, agent two knows more than agent one, all things considered, by
also in a certain sense knowing less – i.e., by being ignorant of the
misleading defeaters (on account of the fact that they have been neutra-
lised by the epistemic helper). The upshot is that having knowledge of
a misleading defeater can be epistemically disvaluable. This means, in
turn, that being ignorant of a misleading defeater can be epistemically
valuable. We have thus identified at least one plausible sense in which
there is an epistemic efficacy in being ignorant.12

3 Weighing Epistemic Value

The phenomenon of misleading defeaters thus offers one straightforward
way in which ignorance can be epistemically valuable. As we will see, this
phenomenon points towards a more general, and also more interesting,
way in which ignorance can be epistemically valuable, but in order to see
this we first need to think a little more about how we ‘weigh’ epistemic
value.
We noted above that epistemic value is concerned with what enables us
to get at the truth, and hence avoid error. Insofar as we focus on beliefs in
this regard, this means that epistemic goodness is about what promotes
the acquisition of true beliefs (and the avoidance of false beliefs). With
this point in mind, a very natural picture emerges of how to ‘weigh’
epistemic goodness. This is that we further our epistemic aims insofar
as we maximise the number of true beliefs that we hold while minimising
the number of our false beliefs. Accordingly, if belief system X has more
true beliefs than belief system Y, but they have an equal number of false
beliefs, then X is epistemically more valuable than Y.
Now one issue we might raise about this conception of how to weigh the
epistemic good is how to manage the trade-off between maximising the

point is that one should not ‘weigh’ epistemic value purely in terms of the number of true
propositions believed.
12
Note that it should not be inferred from the discussion of this case that merely being
ignorant of a misleading defeater suffices to ensure that it does not undermine one’s
knowledge. This is because there can be what are known as normative defeaters, which are
defeaters which one ought to be aware of (but which one might in practice not be aware
of). Accordingly, the mere fact that the epistemic helper has ensured that agent two does
not encounter the misleading defeater does not suffice to ensure that agent two retains her
knowledge, since it may nonetheless be the case that this is a defeater that agent two ought
to be aware of (e.g., perhaps she ought to be more attentive to the fact that there are
people in her environment who believe that she has ingested the hallucinogen). For our
purposes, we can set this complication to one side by stipulating that not only is agent two
unaware of the misleading defeater, but she is also quite rightly unaware of it. For a useful
recent discussion of normative defeaters, in this case with regard to the specific issues
raised by such defeaters in the epistemology of testimony, see Lackey (2005).
138 Duncan Pritchard

number of true beliefs and minimising the number of false beliefs. This is
clearly going to be a vexed question, with several competing strategies
available, depending on what kind of premium, if any, is placed on
accuracy over error. But this is not the issue that I want to engage with
here, interesting though it is. Instead, I want to suggest that this concep-
tion of how to weigh epistemic goodness is faulty in a fundamental
respect, regardless of how one settles the more parochial question, in
comparison, of how to manage the trade-off between maximising accu-
racy and minimising error within this general picture. As we will see,
realising that this conception of how to weigh epistemic goodness is
fundamentally mistaken will help us to recognise one important sense in
which ignorance can be epistemically valuable.
Here is the crux of the matter. In saying that epistemic goodness is
essentially about the acquisition of true rather than false beliefs, we are
not thereby saying that the epistemic good is equally served by the
acquisition of any particular true belief. We can bring this point out by
considering a fallacious kind of reasoning which is unfortunately quite
common in epistemology. Epistemologists often reason something like as
follows:
The Trivial Truths Problem
(P1) If acquiring true belief is always epistemically good, then
we should value all true beliefs, even the trivial ones.
(P2) We rightly do not value trivial true beliefs.
(C) So, acquiring true belief is not always epistemically good.
So, for example, if the epistemic good is furthered by maximising true
belief, then adding any additional true belief to our stock of true
beliefs should be an epistemically good thing to do, no matter how
trivial that belief might be – for example, even if this is a true belief
about the number of grains of sand on a beach. Clearly, however, we
do not think it is epistemically valuable to add additional trivial true
beliefs to our stock of true beliefs. So hence there must be more to our
evaluations of epistemic value than just a concern to acquire true
beliefs.13
Here is another way of putting this problem. Suppose one is faced with
a situation where we can acquire one true belief or another (but not both).
If we hold that acquiring true belief is always equally epistemically good,
no matter which proposition is in play, then from a purely epistemic point

13
Versions of this general line of argument abound in the contemporary epistemological
literature. For a sample of high-profile endorsements of this kind of reasoning, see
DePaul (2001), Sosa (2001) and Goldman (2002b). (The ‘sand’ example is due to Sosa.)
 Ignorance and Epistemic Value 139

of view we should be indifferent between which of these true beliefs that

we acquire. And yet it seems that if one of these beliefs is about a trivial
matter, such as how many grains of sand are on a given beach, while the
other belief is about something of great consequence, such as a truth of
fundamental physics, then we should not be indifferent between these
two beliefs. In particular, even from a specifically epistemic point of view
where we set aside any practical implications of acquiring the beliefs in
question – for example, the truth of fundamental physics, as it happens,
has no more practical utility than the truth about the number of grains of
sand on the beach – it still seems that as a good inquirer one should prefer
the ‘weightier’ truth about fundamental physics over the trivial truth
about the number of grains of sand on a beach. Hence it seems that it is
not, or at least is not just, the acquisition of true belief that we really care
about when it comes to epistemic value.
This is the so-called trivial truths problem, which on the face of it seems
to call for a fairly fundamental reappraisal of the notion of epistemic
value. While this style of reasoning is admittedly very appealing, it does
not stand up to close scrutiny. The fallacy in this reasoning is the idea that
if we epistemically value true belief then it follows that we should episte-
mically value all true belief equally. But the latter does not follow from the
former. As Nick Treanor (2013, 2014) has pointed out in some excellent
recent work on this problem, if this reasoning were sound then it would
follow from the fact that gold mining aims at acquiring gold that it there-
fore treats all gold that is acquired as being of equal value, whether it is
a small nugget of gold or a large block of the stuff. Since this obviously is
not the case, the upshot would be that gold mining is not aimed at
acquiring gold, which is of course absurd. The crux of the matter is that
just as two lumps of gold could be of different value, because one is larger
than another, so two true beliefs can be of different epistemic value,
because one of the beliefs has more content than the other, and hence
captures more of the truth.
This last point can look mysterious at first blush, but once one reflects
on the matter one can see that it actually reflects a fairly mundane point
about propositional content. If I am told (truly) that <something has
fallen down the cellar>, I am given far less of a grip on what has actually
happened than if I am told (truly) that <your brother has pushed your
mother and she has fallen down the cellar>. Even setting aside the greater
practical utility of the second true proposition (e.g., one knows in the
latter case that an ambulance needs to be called, and also that one needs
to be wary of one’s brother!), there is also a perfectly straightforward sense
in which the second true proposition gives one a more comprehensive grip
on reality than the first true proposition. If one’s epistemic goal in inquiry
140 Duncan Pritchard

is to get at the truth, then one ought to be motivated, ceteris paribus,

towards truths with more content over truths with less. Hence, aiming at
the truth does not entail epistemically valuing all true beliefs equally.14
Indeed, once we recognise this point, then we can start to see other
ways in which from a purely epistemic point of view we might value two
propositions very differently. Suppose one is building a comprehensive
theory regarding some subject matter, but is missing some crucial ingre-
dient, some fact which has yet to be unearthed. This truth, in isolation,
may seem relatively uninteresting, at least to the impartial observer any-
way, but given one’s current epistemic state it could well prove momen-
tous. Clearly, as a truth-seeker, one should prefer this truth over an
alternative truth that lacks this momentous import, even insofar as one
grants that the alternative truth is roughly equal in terms of its content.15
After all, this particular momentous truth, while perhaps not epistemi-
cally weighty in and of itself, is of great epistemic weight when it comes to
one’s own epistemic position, in that it will enable one to gain a far more
comprehensive grip on the nature of reality than would be otherwise
available to one.16
With these points in mind, we can see how the epistemic value of
ignorance when it comes to misleading defeaters is really a special case
of a more general phenomenon whereby one’s wide epistemic goals are
served by ignorance. Motivated only by purely epistemic concerns, a
good inquirer may nonetheless focus her attention on acquiring
a truth of substance – for example, a truth that will enable her to gain
a comprehensive understanding of a particular subject matter, say—as
opposed to focussing on even a large body of truths which don’t open up
such an epistemically valuable vista. This is why a good inquirer, whose
goal is to get at the truth, will tend to favour a set of true beliefs that
includes a true belief about fundamental physics rather than a true belief
about the number of grains of sand on a beach. To choose to know the

14
This solution to the problem of trivial truths is expertly discussed in Treanor (2013,
2014). I also further explore this proposal in Pritchard (2014).
15
How does one ‘measure’ content? I do not pretend to have a developed theory of this
(though the general idea that interests us is clearly to measure it along broadly information-
theoretic lines – that is, in terms of the amount of information contained within the content).
Fortunately, this does not matter for our purposes, since all that concerns us is whether the
subject regards these truths as carrying roughly equal amounts of content (for example, they
are not like the two statements about the object that fell down the cellar that we encountered
earlier, which were clearly containing very different amounts of information).
16
Note that this is not to say that our only epistemic interest is in gaining a scientific grip on
the nature of reality. Truths about the human condition might be of no less epistemic
interest, for example, and yet gain one no purchase at all on the nature of (physical)
reality.
 Ignorance and Epistemic Value 141

latter, rather than the former, would be an epistemically disvaluable

course of action. Accordingly, a good inquirer may well eschew the
possibility of acquiring knowledge of a large body of true beliefs about
grains of sand in favour of acquiring knowledge of a single truth about
fundamental physics. In doing so, our inquirer is not disavowing the
epistemic goal of truth, but rather better satisfying this goal. Moreover,
in pursuing this particular kind of inquiry she is also furthering the
epistemic good by being ignorant of certain truths. Indeed, our agent, in
pursuit of the epistemic good, is actively choosing to be ignorant of these
truths. We have thus captured an important sense in which ignorance can
be epistemically valuable.

4 Ignorance and Intellectual Humility

A final issue I want to explore in this regard is the epistemic utility of
a particular kind of attitude to one’s own epistemic standing and episte-
mic capacities, what we might term intellectual humility. By this I have in
mind a sensitivity to one’s own epistemic limitations, and hence
a standing disposition not to overestimate one’s epistemic grip on the
facts. Intellectual humility is often thought to be at least generally valu-
able, to the extent that it is a component part of a virtuous life of
flourishing.17 But is intellectual humility of epistemic value, specifically?
If so, then what is interesting about this from our perspective is that such
humility seems to involve awareness of one’s ignorance. To this extent,
then, the epistemic value of intellectual humility seems to suggest, at least
in a rather indirect way, a potentially new way in which ignorance can be
epistemically valuable.
We should note from the outset off that it is not at all obvious that
intellectual humility is epistemically valuable. It could be, for example,
that focussing on one’s epistemic limitations could undermine one’s
confidence in one’s judgements such that one’s inquiries are stymied,
and hence never lead to fruitful conclusions. Does not good inquiry
sometimes involve boldness rather than diffidence? Perhaps, but I think
the proponent of the idea of intellectual humility as epistemically valuable
has ways of accommodating this idea.
To begin with, note that boldness in one’s inquiries should be differ-
entiated from dogmatism. One might boldly put forward a scientific
conjecture that runs counter to current thinking and then proceed to
test that conjecture. If the conjecture is found to be true, one might
thereby come to gain a deep understanding of a subject matter that
17
See, for example, Roberts and Wood (2007) and Baehr (2011).
142 Duncan Pritchard

might not have otherwise have been available. Boldness in inquiry is here
seen to be delivering the epistemic good. But this bold inquiry, so
described, does not implicate any dogmatism – the bold conjecture was,
after all, confirmed by the tests run, and hence was eventually supported
by evidence. Suppose that, instead, the tests did not support the conjec-
ture, and that this is a stable feature of the tests run. To continue to press
the conjecture would now be to embrace dogmatism, in that one is now
inquiring without concern for the evidence, and hence for the truth. Here
we have gone beyond mere boldness of inquiry. Moreover, such dogma-
tism is clearly not serving the epistemic good, but leading one down an
intellectual dead end, and one would expect the intellectually humble
inquirer, who will not be prone to dogmatism, to not be lead astray in this
fashion. So long as we can keep boldness and dogmatism apart, then,
there seems to be a way of making the epistemic goodness of intellectual
humility compatible with bold inquiry.18
The point is that being intellectually humble does not require one to
avoid being bold. Rather, what it requires is that, in being aware of one’s
epistemic limitations – in being aware of the general level of one’s ignor-
ance – one is appropriately sensitive to the possibility of error, and so
revises one’s beliefs accordingly. Indeed, we can think of intellectual
humility as an intellectual virtue, and hence as a mean between two
epistemically viceful extremes. One of these extremes is the epistemic
vice of dogmatism that we just witnessed. The other extreme would be an
epistemic vice that we can call intellectual diffidence, where this involves
a complete lack of faith in one’s judgements, one that is entirely unre-
sponsive to the evidence to the contrary. The virtuous believer will have
the practical wisdom to steer between these two extremes and in the
process manifest her intellectual virtue. In doing so, she will appropriately
employ her knowledge of her general level of ignorance in order to better
track the epistemic good.
Now that we have set out what intellectual humility, qua intellectual
virtue, involves, however, then it becomes clear that it does not entail that
ignorance can be epistemically valuable. It is, after all, one’s knowledge of
one’s ignorance that is servicing the epistemic good when it comes to

18
Another point to bear in mind when we epistemically evaluate the potentially intellec-
tually virtuous behaviour of the intrepid scientist who proposes bold conjectures is that
we would not expect her commitment to these conjectures to be anything more than
provisional. In particular, we would not expect her to believe these propositions to be
true, given their conjectural nature. Thus the very character of her propositional commit-
ments in this regard is revealing a sensitivity to the evidence in play, and hence a lack of
dogmatism. (Indeed, it could well be that the scientist’s commitment to theoretical
claims in general is not one of belief, but rather a weaker propositional attitude, such as
acceptance. See, for example, van Fraassen [1980]).
 Ignorance and Epistemic Value 143

displays of intellectual humility, rather than the ignorance itself. Hence,

despite first appearances, intellectual humility does not offer us another
way, albeit a tangential one, in which ignorance could be epistemically
valuable.19

5 Concluding Remarks
Our focus here has been on the epistemic value of ignorance in a very
specific sense. In particular, our concern has not been the general value of
the epistemic standing of being ignorant, but rather the specifically epis-
temic value of ignorance. As we have seen, while it is much harder to
demonstrate that ignorance is valuable in the latter sense than the former
sense, there are some plausible candidates in this regard. The most
straightforward of these is misleading defeaters, as it seems that one is
epistemically better off in being ignorant of them.
A more interesting kind of case is exposed once we reflect on the best
way of ‘weighing’ epistemic value, since if we do not treat all true beliefs as
being of equal epistemic value, then we can accommodate a sense in
which one might be epistemically better off in believing some truths rather
than others, and thus capture a further way in which ignorance might be
epistemically valuable. Finally, we examined intellectual humility, and
the idea that this both serves the epistemic good while at the same time
incorporating an awareness of one’s ignorance. It thus seemed to poten-
tially capture a further, albeit rather indirect, sense in which ignorance
could be of epistemic value. As we saw, however, the epistemic value of
intellectual humility, properly understood, lies in one’s knowledge of
one’s ignorance rather than in the ignorance itself, and hence this is not
a further instance of how ignorance can be epistemically valuable.20

19
I am grateful to Rik Peels for discussion on this topic. I discuss intellectual virtue and
intellectual humility in more detail in Pritchard (forthcoming a), Carter and Pritchard
(forthcoming), and Kallestrup and Pritchard (forthcoming).
20
This paper was written as part of a research project hosted by the University of
Edinburgh’s Eidyn Philosophical Research Centre: the Templeton Foundation-funded
‘Virtue Epistemology, Epistemic Dependence and Intellectual Humility’ project. I am
grateful to the Templeton Foundation for their support of this research. Thanks to
J. Adam Carter, Jesper Kallestrup, and Nick Treanor for helpful discussion on related
topics. Special thanks to Rik Peels for detailed comments on an earlier version of this
paper.
8 Ignorance and the Religious Life

Justin McBrayer

For Life is a fire burning along a piece of string – or is it a fuse to a powder

keg which we call God? – and the string is what we don’t know, our
Ignorance, and the trail of ash, which, if a gust of wind does not come,
keeps the structure of the string, is History, man’s Knowledge . . .
~Robert Penn Warren, All the King’s Men1

1 Ignorance and Religion

There is debate over the nature of ignorance. Some philosophers argue
that ignorance is a lack of knowledge (e.g., Le Morvan 2011, 2012,
2013). This is the knowledge view of ignorance. On this view, ignorance
is compatible with true belief. Others argue that ignorance is a lack of true
belief (e.g., Peels 2010, 2011a, 2012). This is the truth view of ignorance.
On this view, ignorance is incompatible with true belief. Since a lack of
true belief entails a lack of knowledge, any mental state that counts as
ignorance on the truth view will also count as ignorance on the knowledge
view. But it is possible for a mental state to count as ignorance on the
knowledge view but not on the truth view. For example, one might have
a true belief but have arrived at the truth in a Gettier case.
Which of these two conceptions of ignorance we employ will affect our
thinking about the relation between religion and ignorance. For example,
on the truth view, if it is true that God exists, then it will follow that many
religious believers are not ignorant of that fact. Indeed, it will be the non-
theists who are ignorant for they shall lack true beliefs about divine reality.
These religious believers may not know that God exists, but if that’s not
required to avoid ignorance, then they can meet the test.

The ideas in this paper benefitted from several conversations with Dan Howard-Snyder and
Jon Kvanvig at the Nature of Faith conference in St. Louis, Missouri, in November of 2014.
That conference was generously funded by the John Templeton Foundation. Gijsbert van
den Brink, Dugald Owen, Rik Peels, and Martijn Blaauw provided very helpful comments
on an earlier draft of this paper.
1
Warren (1946, p. 151).

144
 Ignorance and the Religious Life 145

There is more likely to be a deep and pervasive religious ignorance on the

knowledge view. This is because there are more conditions that have to be
met to avoid ignorance. For this reason, this chapter assumes the more
traditional, knowledge view of ignorance. Ignorance is a lack of knowledge.
On this conception, it is at least initially plausible that many religious
people are ignorant: they believe a great many religious propositions that
they do not know. These propositions may be true. The religious likely
have faith that these propositions hold. They might even have some justi-
fication for the religious propositions in question. But, given the presence
of defeaters, the reliance on untrustworthy testimony, and so on, it is likely
that a great many of these beliefs will fall short of the higher standards
required for knowledge.
How does this ignorance affect the religious life? A careful investigation
would reveal the role played by ignorance in both religious theory and
religious practice. The point of the chapter to sketch a broad taxonomy of
the various roles that ignorance plays in the religious life. This is impor-
tant since the religious life is shaped as much by what we do not know as by
what we do.

2 Ignorance and Religious Theory

Plato’s cave is an image of ignorance. We live underground and confront
the mere shadows of reality. Enlightenment requires a painful and diffi-
cult struggle of breaking free and seeing the light of day for the first time.
The questions investigated by philosophy are not prone to easy answers,
leaving us in ignorance much of the time. This is even more so the case
with philosophy of religion. If there is a supernatural reality, it is plausible
that it would be far removed from our everyday experiences. As such, our
theoretical grasp on the divine is bound to come up short. For example, if
there is a God, it is plausible that we would be deeply ignorant of many of
his reasons for acting in the world. This section surveys the role that
ignorance has played in theorizing about the existence of a divine reality,
the nature of divine reality, and the relation between humans and divine
reality.

2.1 Ignorance and the Existence of Divine Reality

Is there a supernatural reality or is the natural world all that there is? Not
surprisingly, human ignorance plays a key dialectical role in the evalua-
tion of arguments both for and against the existence of the divine. This is
especially clear in the development of philosophy of religion in the West
where philosophers have been preoccupied with arguments for and
146 Justin McBrayer

Table 1 Utility Calculus

God does not

God exists exist

Believe that God exists Infinite gain Wash

Do not believe that God exists Infinite loss Wash

against the existence of God. Without making a claim to a comprehensive

survey, here are two examples of ignorance being invoked in favor of
theism or atheism and two examples of ignorance being invoked to
block arguments for theism or atheism.
Example 1: From ignorance to theism.
Historically, one of the most common ways of justifying the existence of the
divine has been through what has been derogatorily called the “God-of-
the-Gaps” strategy. This strategy finds some data that cannot be explained
(e.g., lightning, the seasons, the existence of life) and then posits God as the
explainer. It is obvious how ignorance plays a role in this sort of justification
for the divine. It is also obvious that this strategy will commit an illicit
appeal to ignorance unless the inference is framed in a careful way (more
about this below vis-a-vis fine-tuning arguments).
However, ignorance also plays a key role in “making room” for one of
the most contentious arguments for theistic belief, namely Pascal’s Wager
(e.g., Jordan 2006). Pascal’s Wager (and similar prudential arguments for
theistic belief) purports to show that it is in our best interest to believe in
God regardless of the available evidence. While there are many ways to
read the cryptic passages that give rise to the Wager, a standard rendering
of the argument is in the form of a utility calculus (see Table 1).
The idea is that if God exists, theists will receive all of the goods of the
afterlife whereas non-theists will not. On the other hand, if God does not
exist, there is no substantial difference between the goods secured by
a theist versus a non-theist. And even if we are ignorant of the existence
of God, as long as there is a finite chance that he exists and infinite
rewards to be gained, it makes sense to bet on God. So even in the absence
of evidence for the existence of God, there is still a powerful, prudential
reason to be a theist.
Example 2: From ignorance to atheism.
Some philosophers seem to think that human ignorance renders atheism
the dialectical status quo. In other words, they think that our ignorance is
a reasonable basis to deny the existence of the divine (e.g., Flew 1976;
 Ignorance and the Religious Life 147

Mion 2012). However, this kind of move would not be accepted in many
other contexts. For example, the status quo should not be to disbelieve
that there is life elsewhere in the universe until we get evidence otherwise.
In our ignorance, the most reasonable position is to be agnostic on
whether or not there is life elsewhere in the universe. For this reason,
most philosophers seem to think that agnosticism about the divine is
a more reasonable status quo and that the dialectical burden is on those
who would either affirm or deny the existence of the divine.
The argument from divine hiddenness is the most sophisticated argu-
ment that invokes human ignorance as a premise in an argument for
atheism (e.g., Schellenberg 1993). The gist of the argument is that the
absence of evidence for God constitutes evidence for the absence of God.
A very basic version of the argument proceeds as follows:
1. If there were a God, reasonable people would not be ignorant about his
existence.
2. But reasonable people ARE ignorant about the existence of God.
3. Therefore, there is not a God.
The defense of premise one rests on the idea that a relationship with God
would be among the very highest goods achievable and a relationship with
someone requires belief in that person’s existence. So given that God is both
perfectly loving and perfectly powerful, he would ensure that all reasonable
persons would have enough evidence so that belief in God is reasonable.
The defense of premise two appeals to the apparent fact that there are many
reasonable, thoughtful people who are capable of belief in God and yet are
ignorant about his existence due to a shortage of evidence. These people
may even have true belief concerning God, but this belief doesn’t amount to
knowledge and hence is ignorance. And according to the argument from
divine hiddenness, this ignorance about the existence of God provides us
with a reason to think that there is no such being.
Example 3: Using ignorance to block an argument for theism.

Not only can ignorance be invoked as a reason to be a theist or atheist, it can

also be used as a defensive maneuver in the face of arguments for theism or
atheism. One prominent example in the current literature concerns cosmic
versions of the argument from design (e.g., Swinburne 2004, chapter 8).
Sometimes called the fine-tuning argument, this argument starts from
a premise about the fine-tuning of the universe and concludes that
a powerful designer like God is the best explanation of the fine-tuning
data. Very roughly, the fine-tuning data is a collection of all of the facts
discovered by scientists working in cosmology regarding the life-permitting
parameters of the physical cosmos. It appears that there are many variables
148 Justin McBrayer

that, had they been any other way, would not have allowed for life in the
universe. Robin Collins (1999) provides a standard example of this sort of
data:
If the initial explosion of the big bang had differed in strength by as little as one part
in 1060, the universe would have either quickly collapsed back on itself, or expanded
too rapidly for stars to form. In either case, life would be impossible. (p. 49)
How can we explain the fine-tuning data? Only three possibilities are
salient: chance, necessity, or design. The fine-tuning argument attempts
to show that design is the best option.
However, design arguments of this sort are often countered by appeals to
human ignorance. The idea is that we know too little about the basic
structure of the universe or about the conditions for life to make any grand
pronouncements about how best to explain the fine-tuning data. Here are
two examples of this strategy. First, the design argument relies on the claim
that the parameters of the universe could have been different. But perhaps
this appearance of contingency is really just our ignorance of deep facts about
the parameters of the universe. If it turns out that the parameters of the
universe could not have been different, then the design hypothesis is
undermined. Second, the design argument relies on the claim that only
a handful of universe parameter combinations are life-permitting. But per-
haps we are not entitled to this assumption. It if turns out that radically
different life forms are possible, perhaps the range of universes that is life-
permitting grows substantially. In either case, our ignorance of the deep facts
about the universe and life offer avenues to undermine fine-tuning
arguments.
Example 4: Using ignorance to block an argument for atheism.
One of the most common arguments for atheism is the argument from evil
(e.g., van Inwagen 2006). Suppose the world is governed by a perfectly good
and perfectly powerful deity like God. It would be surprising to find that
world occupied by the amount, kind, and distribution of evil that we find in
the actual world. A very basic version of the argument proceeds as follows:
1. If there were a God, the world would not contain gratuitous evil.
2. But the world DOES contain gratuitous evil.
3. Therefore, there is not a God.
While the argument can be made more sophisticated, something very
much like this underlies the atheism of many people in the West.
However, theists have responded to the argument in a number of ways.
Relevant to our purposes here is a family of responses termed “skeptical
theism” (e.g., McBrayer 2010). In brief, the skeptical theist is a theist who
 Ignorance and the Religious Life 149

thinks that no one is justified in believing that any given instance of evil is
gratuitous. The idea is that we are simply ignorant of too much of moral
reality to draw an inference as to whether an evil is necessary for some
compensating good. Skeptical theists sometimes rely on analogies to
make this point: just as a novice chess player watching a match should
not assume that a particular chess move is worthless, so, too, humans
looking around at the world should not assume that a particular evil is
pointless. In both cases, if there is a master in control of the pieces, there
would be many moves/evils that seemed pointless that were not. And if
this general line of thought is correct, ignorance appears to undermine
the second premise of a general form of the argument from evil.

2.2 Ignorance and the Nature of Divine Reality

The previous section canvassed attempts to invoke ignorance in premises
for or rebuttals to arguments for divine reality, specifically God. Just as
ignorance plays a key role in arguments for the existence of a divine reality,
so too does ignorance play a key role in theorizing about the nature of that
divine reality. For example, in the Christian tradition, St. Paul acknowl-
edges our current ignorance of divine reality when he writes that “for now
we see through a glass, darkly; but then face to face . . .” (I Corinthians
13:12). This section explores two prominent examples of the role of
ignorance in shaping our thought about the nature of divine reality.
Example 1: The Apophatic Tradition.
Virtually every major world religion includes a strand of theology known
collectively as apophatic theology (apophanai is Greek for “to say no”).
The basic idea is that human concepts fall short of the true nature of the
divine. As a result, we cannot speak (or believe) truly about what the divine
IS. Instead, we are limited to saying what the divine is not. In short, our
ignorance of the true nature of the divine is utterly complete. The best we
could ever come to know is what the divine is not. However, just as dust
might show the form of an invisible man, so, too, might our claims about
what the divine is not illuminate the form of the divine. We know indirectly
through our ignorance.
Several Western religious thinkers turned toward apophatic theology in
the wake of what they saw as blatant anthropomorphism in the theology of
the Greek pantheon. Prominent examples of early church fathers include
Pseudo-Dionysius, Saint Cyril of Jerusalem, and the Cappadocian fathers.
For example, Gregory of Nazianzus wrote that “the only thing that could be
comprehended about the incomprehensible divine nature was its ‘bound-
lessness’ – what it was not rather than what it was” (Pelikan 1993, p. 42).
150 Justin McBrayer

The pendulum away from affirmative theology reached a peak in the

Medieval period where apophatic reasoning was termed the via negativa
(Latin for “the negative way”). While those working in the via affirmativa
tradition attempt to say what God is really like, thinkers in the via negativa
tradition demur and restrict themselves to claims about what God is not.
The idea is that God in his true form is ineffable, and the application of
finite concepts to an infinite being only results in confusion.2
This negative tradition claims a grounding in the Jewish scriptures at
least as early as Moses and the burning bush. When Moses asks who is
sending him back to Egypt, the answer is “I am who I am” and nothing
more. Maimonides, perhaps the greatest Jewish thinker in the medieval
period, writes that:
Know that the negative attributes of God are the true attributes: they do not
include any incorrect notions or any deficiency whatever in reference to God,
while positive attributes imply polytheism, and are inadequate . . . [w]e cannot
describe the Creator by any means except by negative attributes. (1956, p. 81)
And while many mystical strains of Christianity are obviously apophatic,
the via negativa can still be found in contemporary theology (e.g., Marcus
Borg’s The God We Never Knew). Despite this presence, it’s also probably
fair to say that it plays a larger role in Orthodox Christian theology as
compared to Roman Catholic or Protestant theology.
Apophaticism shows up in Eastern religions, too. Hindu philosophers
are cagey about their descriptions of Atman and Brahman. On the one
hand, Brahman is recognized as the ultimate reality. But what that reality
is like becomes hard to say. Those working in the Advaita Vedanta
tradition of Hinduism explicitly endorse our ignorance of divine reality
by naming reality Nirguna Brahman: reality indescribable by human pre-
dicates. Similarly, philosophers working in Taoist tradition insist that our
understanding of the divine is shrouded in ignorance. The first lines of the
Tao Te Ching are as follows:
The Tao that can be told of is not the eternal Tao;
The name that cannot be named is not the eternal name.
The Nameless is the origin of Heaven and Earth;
The Named is the mother of all things. (Chan 1963, p. 97)

As soon as you apply a human concept to the divine, you’re guaranteed to

have gotten it wrong. Ignorance is at the heart of our theorizing about the
nature of the divine.

2
There are, of course, theologians in the Medieval period who draw from both the via
affirmativa and via negativa traditions. One anonymous reviewer suggests that St. Thomas
is an example of this blend.
 Ignorance and the Religious Life 151

Example 2: Divine Reality as the Noumenon.

Kant famously argues that we cannot have unmediated metaphysical
knowledge. Thus we are ignorant of reality as it exists outside of our
own concepts. While we can know a great deal about how the world
appears to us (the phenomenal realm), we know next to nothing about
the world as it is in itself (the noumenal realm). In short, Kant insists that
a perceived situation owes as much to the perceiver as to the world outside
of the perceiver.
Several philosophers of religion have applied this insight to our under-
standing of divine reality. John Hick is the foremost contemporary exam-
ple to have exploited this comparison (Hick 1989). According to Hick,
. . . the realization that the world, as we consciously perceive it, is partly our own
construction leads directly to a differentiation between the world an sich, unper-
ceived by anyone, and the world as it appears to, that is as it is perceived by, us.
(p. 241)
As applied to divine reality, we might say that the divine being as it is exists
an sich may be different from the divine being as it appears to us.
The result is that “ . . . the great post-axial faiths constitute different
ways of experiencing, conceiving and living in relation to an ultimate
divine reality which transcends all our varied visions of it”
(pp. 235–236). Divine reality is the noumenon. The perceptions of the
divine are the phenomena. In each case, the Christian, Jew, Muslim,
Hindu, and Buddhist are responding to the same ultimate divine reality,
albeit this reality is perceived in radically different ways. So on the one
hand, our grasp of the divine is steeped in ignorance. Our knowledge of
the divine as it exists in itself is restricted to logical or formal properties like
“exists.” On the other hand, our grasp of the divine as it is perceived by
humans is much deeper, although obviously relative to different religious
communities. In this sense, our knowledge of the divine-as-perceived will
include the attribution of a range of substantive properties such as “is
good,” “created us.”

2.3 Ignorance and the Relation between Humans and Divine Reality
The last two sections canvassed major roles that ignorance plays in argu-
ments for the existence of a divine reality and arguments about the nature
of divine reality. This final section on ignorance in religious theory will
highlight two prominent cases in which ignorance plays a theoretical role
in the relation between humans and divine reality.
Example 1: Freedom and Foreknowledge Debate.
152 Justin McBrayer

At least in the West where the divine is portrayed as a person, the

question of human freedom looms large in the face of divine fore-
knowledge. If acting freely requires the possibility to do otherwise
and if infallible foreknowledge eliminates the possibility of acting
otherwise, how can humans ever act freely given the existence of an
omniscient being like God? Many philosophers, both historical and
contemporary, have agreed that foreknowledge threatens genuine
freedom (e.g., Hasker 1985).
While there are many purported solutions to the problem, one of
the most serious contenders in contemporary philosophy of religion
is a view known as open theism (e.g., Hasker 2004). According to
open theism, God is ignorant of our future actions that are not
necessitated by current states of affairs plus the laws of nature.
In other words, there are a great many things about the future that
God does not know, and since he does not foreknow these things, all
possibilities remain open for the human actors in question.
Of course, it would be a very good thing to know what would happen
in the future, so open theists owe an explanation for why a perfect,
omniscient being would fail to have this knowledge. One standard
explanation is that an omniscient being will know all that there is to
know. However, when it comes to future actions, these facts are
contingent upon human choices and so there is no fact yet to be
known. To use Aristotle’s example, it is neither true nor false that
there will be a sea battle tomorrow. But since knowledge implies
truth, there is nothing to be foreknown about the sea battle. Other
open theists grant that future contingents have a truth value but insist
that God remains ignorant of these truths in order to preserve the
possibility of free action. This is a perfect example of ignorance being
invoked to solve problems in religious theorizing.
Example 2: The Nature of Salvation.

While European explorers of the New World undoubtedly had many

different motivations, chief among them was a desire to evangelize the
native peoples. This religious drive continues to the present day. For
example, the Mormon (LDS) Church claims to have almost 100,000
missionaries in the field at any time.3 Why?
Theories of salvation vary widely among world religions, but histori-
cally a centerpiece in many of these religions is the view that ignorance is
inimical to salvation. To be truly saved, one must truly believe. As long as
one remains ignorant of divine reality, salvation is impossible. Thus the

3
www.mormonnewsroom.org/facts-and-stats
 Ignorance and the Religious Life 153

chief good that a missionary could deliver is the information needed to

reach salvation. Knowledge (or at least true belief) is a requirement for
avoiding damnation or achieving the beatific vision. And so ignorance
plays a major theoretical role in religious theories of the afterlife.
In Christianity this understanding is motivated in no small part by “the
Great Commission” in the gospel of Mark:
And [Jesus] said to them, “Go into all the world and preach the gospel to all
creation. He who has believed and has been baptized shall be saved; but he who
has disbelieved shall be condemned.” (Mark 16:15–16)

This verse, among others, has been used by Christians to argue that true
belief is necessary for salvation. This explains the preeminence of creeds
and the attendant importance of orthodoxy in many religious traditions:
heaven or hell hang in the balance.
However, just as some philosophers have defended the necessity of
true belief for salvation, others have criticized it. The core critique is
that praise and blame are typically reserved for features of one’s life
that are under one’s control. For example, it makes sense to praise
someone for his hard work but not for the color of his hair. Paired
with the uncontroversial claim that we do not exercise direct voluntary
control over our beliefs, this common sense moral principle implies
that we cannot be praised or blamed for our religious ignorance.
Of course, this is not to deny that we often have indirect control over
our beliefs. If I want to learn the capital of Kenya, I can consult an
authoritative source. But whether my experience of the source gives
rise to a belief about the capital of Kenya is beyond my control.
The same goes for religious propositions: perhaps we can be held
accountable for whether we carefully considered the evidence avail-
able. But whether these experiences give rise to religious belief is out
of our control. And so, according to some philosophers, it is proble-
matic to give ignorance a prominent role in determining afterlife
goods.
On a related issue, just as ignorance of certain propositions might bar
one from heaven, it might also be required to remain in heaven. For
example, on the Christian conception of the beatific vision, the saved
experience eternal bliss in the presence of God. But many philosophers of
religion have wondered whether this eternal bliss were possible in the face
of certain kinds of knowledge (e.g., Talbott 1990). For example, how
could one be happy and content in heaven on the assumption that many
of one’s former friends and relatives are suffering the torments of hell?
One popular response to this objection is that the saved are purposefully
kept ignorant about the fates of those consigned to hell (e.g., Craig 1991).
154 Justin McBrayer

And so while ignorance may play a theoretical role in determining where

one exists in the afterlife, it may also play a theoretical role in determining
the quality of that afterlife.

3 Ignorance and Religious Practice

We have seen that ignorance plays many key roles in religious theory.
How might ignorance affect religious practice? As with religious theory,
ignorance is important both for the communal lives of religious people
and also the lives of religious individuals as they live out their faiths.

3.1 Ignorance in Communal Religious Life

Religious communities have often employed the ignorance of others to
their advantage. In some cases, ignorance is essential for the survival of
the communities in question. Easy examples include the early Christian
church’s efforts to keep the Roman authorities in ignorance and
18th–20th century efforts by Jewish communities in Europe to stay
under the radar of anti-Semitic governments.
But in at least some cases, religious communities inculcate an ignor-
ance even among their own congregations. Again, the early Christian
church is an obvious example. Like something out of the Da Vinci Code,
early church leaders sought to keep both outsiders and newly initiated
Christians in the dark about many of the church’s rituals and practices.
This practice was much later termed the disciplina arcani: the arcane
disciplines. According to the Catholic encyclopedia New Advent, the
practice has its roots in the Christian scripture4: “Do not give dogs what
is sacred; do not throw your pearls to pigs. If you do, they may trample
them under their feet, and turn and tear you to pieces” (Matthew 7:6).
Some scholars have argued that this practice of concealing certain rituals
and services led rather naturally to the growth of Christian mysticism in
the early medieval period (e.g., Stroumsa 2005).

3.2 Ignorance in the Individual Religious Life

What role might ignorance play in the life of individual religious people?
It is initially plausible that ignorance is an impediment to living
a religiously virtuous life or a life of genuine religious faith. The virtuous
religious person is not a person of doubt. The faithful religious person, it
might be supposed, is a knower, not a doubter. But this initial

4
www.newadvent.org/cathen/05032a.htm
 Ignorance and the Religious Life 155

presumption is too quick. This final section canvasses the role of ignor-
ance in a life of religious virtue and a life of religious faith.
Regarding religious virtue, the pro tanto case against ignorance is based
on an isomorphic case against ignorance with regard to moral virtue.
In ethical theory, it is commonplace to assume that knowledge is
a prerequisite for exercising the moral virtues. Aristotle claims that,
“[the virtuous agent] must be in a certain condition when he does [the
virtuous action]; in the first place he must have knowledge; secondly he
must choose the acts, and choose them for their own sakes; and thirdly his
action must proceed from a firm and unchangeable character”
(Nicomachean Ethics II.4.1105a 31–34). So it appears that knowledge is
a necessary condition for virtuous action. Since religion is at least partly
a normative endeavor, it is inevitable that there are certain character traits
that function as virtues and others that function as vices. And just as
ignorance is bad for the practice of ethical virtues, so, too, is ignorance
bad for the practice of religious virtues.
But philosophers have challenged this picture of moral virtue as
overly-intellectualized. Julia Driver (2001) argues that not only is
knowledge not required for the exercise of all moral virtues but in fact
knowledge is inconsistent with certain moral virtues. She argues for an
entire class of virtues that she terms “virtues of ignorance” that include
things like genuine modesty, blind charity (roughly seeing the best in
others while ignoring their faults), impulsive courage, and a sort of
forgiveness. In each case, she claims that full disclosure would eliminate
the virtuous disposition. At least some recent experimental philosophy
backs up this picture (e.g., Feltz and Cokely 2012). It turns out that
ignorance does not rule out our attributions of virtue in many cases, and,
even more surprisingly, ignorance can boost the odds of ascriptions of
virtue in certain cases. Especially regarding cases of modesty and impul-
sive courage, respondents are somewhat more likely to attribute the
virtue in question to an actor when it is stipulated that the actor is
ignorant of certain facts.
Perhaps the same goes for religious virtues. As recounted in the Jewish
and Christian traditions, Adam and Eve were expelled from the Garden
of Eden for eating from the tree of the knowledge of good and evil. God
had created them with an ignorance that rendered them better off to at
least some extent. A loss of that ignorance left them shamed and home-
less. Is that true for religious virtues in general? A full examination of the
topic would require an account of religious virtue and vice and
a discussion of whether ignorance preempts the “activation” of such
virtues. For present purposes, a brief illustration is sufficient. In the
Christian tradition, there are three theological virtues that have
156 Justin McBrayer

traditionally been offered as primary religious virtues: faith, hope, and

love (sometimes termed “charity”). Since faith is considered separately
below, does ignorance impair the exercise of hope or love?
It is quite clear that religious belief can be a source of hope. For example,
empirical research suggests that religious belief in those who are ill boosts
a hope in recovery (e.g., McBrayer 2014). And there are famous examples
of religious believers whose belief sustained almost unbelievable efforts in
the face of dismal odds. As a singular example, recall the trials of Father
Trocme and his fellow villagers in the French town of Le Chambon as
they struggled to locate and save Jews from the Nazis during World War
Two. The townspeople managed to save literally thousands of lives over
the course of the war. Philip Hallie, the philosopher who chronicled the
effort, concludes that “in all certainty . . . [father] Trocme’s belief in God
was at the living centre of the rescue efforts of the village” (1979, p. xxi).
But can this sort of hope obtain in the absence of religious belief?
It seems so. Louis Pojman’s famous paper “Faith, Hope and Doubt,”
argues that faith does not require belief (see below) and further that
something like hope could serve as a substitute for belief. In his descrip-
tion of hope, he makes a persuasive case for the claim that we can hope for
things that we do not believe. As an easy example, I can hope that I win
the lottery even though I do not believe that I will win the lottery. There is
no apparent reason why religious hope would be any different than hope
simpliciter. And if so, then it makes perfect sense to conclude that religious
people might be able to exercise the virtue of religious hope even in the
presence of a deep and abiding religious ignorance. The religious indivi-
dual might have to believe that P in order to know that P, but she doesn’t
have to believe that P in order to hope that P.
The same might be said of charity. Take an example familiar to anyone
who has spent time in a large city: panhandling. When faced with
a destitute stranger asking for money, what would the virtuous person
do? There is ample evidence that in at least some cases the donated funds
are used for illicit purchases like drugs that harm the stranger’s welfare in
the long run. But this is not always the case. And so we are left in
ignorance about a great many things that would help us to make a wise
decision. Does that ignorance eliminate the possibility of true charity?
No. While the ignorance may function as an excuse that counters the duty
of beneficence, it need not do so. The truly charitable person may give
even in the presence of ignorance.
As applied to religious individuals, the amount of charity given by those
who practice a religion is staggering. Aside from the often significant
support that such religious individuals provide to their own religious
communities, the religious often give more to secular causes as well
 Ignorance and the Religious Life 157

(e.g., Brooks 2007). In many of these cases, the givers are ignorant not
only about the relevant religious propositions but also about the full uses
of the funds in question – how many supporters of the local mosque know
the local budget? So it is difficult to see how any sort of general ignorance
blocks the activation of religious virtues. Furthermore, the correlation
between religiosity and certain virtues suggests a causal connection.
Leaving the other virtues aside, what is the relation between faith and
ignorance? Many people are religiously faithful and practice their faith on
a daily basis. Does ignorance hamper their effort, aid their effort, or
neither? Answering this question requires a clear conception of the nature
of religious faith, and that is highly contested ground in which contem-
porary philosophers of religion are expending enormous amounts of
effort. Still, it is possible to give a quick survey of the field.
Some philosophers have argued that ignorance hampers faith. This
view has deep roots in the development of Christian thinkers in the
west. For example, John Calvin wrote that faith is “a firm and certain
knowledge of God’s benevolence towards us” (Institutes III, ii, 7, 551).
For Calvin, faith is not just identified with knowledge but with certainty!
Similarly, John Locke describes divine testimony as follows:
This is called by a peculiar Name, Revelation, and our Assent to it, Faith: which as
absolutely determines our Minds, and as perfectly excludes all wavering as our
Knowledge itself . . . so that Faith is a settled and sure Principle of Assent and
Assurance, and leaves no manner of room for Doubt or Hesitation. (An Essay
Concerning Human Understanding, Book IV, Chapter XVI and 14; Nidditch 1975,
p. 667)

And so for Locke, faith excludes doubt and ignorance in the same way
that knowledge does. At least some prominent contemporary philoso-
phers endorse this view that faith is either identical with or presupposes
knowledge (e.g., Plantinga 2000).
Other thinkers claim the very opposite: faith requires ignorance.
St. Thomas Aquinas writes that “you can’t know what you simulta-
neously put faith in, because knowledge sees and faith doesn’t” (Summa
Theologiae 1989, p. 329). Bishop (2010) makes a similar point by citing
Kant: “I have . . . found it necessary to deny knowledge, in order to make
room for faith” (preface to the second edition of The Critique of Pure
Reason). Additionally, William James’ work on what he calls the “will to
believe” presupposes a great deal of ignorance (e.g., James 1912). For
James, we can be within our intellectual rights in endorsing one hypoth-
esis over another even if we are ignorant of which of the two hypotheses
holds. And, in fact, it’s only in the face of such ignorance that the will to
believe “kicks in.” In other words, without ignorance, there is no will to
158 Justin McBrayer

believe. There are contemporary philosophers who endorse this incom-

patibility. Clegg (1979) opens with what he takes to be an obvious
admission: “. . . we all recognize that we can have faith only in what we
do not know – in an uncertain future, for instance, but not in an obvious
past” (p. 225). For philosophers in this camp, faith that X precludes
knowledge that X.
Finally, some philosophers reject both of these views. These philoso-
phers deny both that faith requires knowledge and that faith requires
ignorance. Faith is compatible with either. Many of these views of faith
are non-cognitive in the sense that they do not require the faithful to
endorse a specific cognitive content. While there are many such non-
cognitive views, what follows is a brief sample. According to Clegg
(1979), faith is the emotional inverse of fear. It is an affective state that
is manifested only in ignorance. Like other emotional states, “a faith may
be shaken, crushed, dashed, lost, or abandoned, but not falsified” (1979,
p. 232). Other philosophers think of faith as a kind of acceptance that is
different from belief (e.g., Alston 1996; Howard-Snyder 2016). The idea
is that one can accept a claim and act in accord with it even though one
does not yet believe the claim. The acceptance model of faith is similar to
a view sometimes described as a practical commitment model of faith
(e.g., Kvanvig 2013). On this view, having faith is having a disposition to
act in accord with an ideal even when one’s evidential situation falls short
of justifying a belief that the ideal will obtain. On any of these views of the
nature of faith, true faith is compatible with ignorance.
While perhaps at odds with everyday intuitions, the view that faith is
compatible with ignorance gains traction when we examine the lives of
role models in the faith. San Juan de la Cruz writes powerfully of his “dark
night of the soul” in which God deliberately pulls away from the con-
verted Christian:
. . . spiritual persons suffer great trials, by reason not so much of the aridities which
they suffer, as of the fear which they have of being lost on the road, thinking that all
spiritual blessing is over for them and that God has abandoned them . . .. For the
more a soul endeavours to find support in affection and knowledge, the more it
will feel a lack of these . . . (Peers 1951, pp. 71, 74)
During the dark night, ignorance reigns, even when the faithful follower
wishes it would go. A contemporary example is Mother Theresa who
confesses that her knowledge of God ebbed and flowed leaving her at
times with the challenge of living a faithful life in her ignorance. At one
point, Mother Theresa sets her goal “to live by faith and yet not to
believe” (Kolodiejchuk 2009, p. 248). While it is probably true that
Mother Theresa was a religious believer for much of her life, it seems
 Ignorance and the Religious Life 159

equally clear that she was not a religious believer for all of her life. She
wrestled with extreme periods of doubt, and hence she is an example of
someone living a life of faith not just in ignorance but even without
belief.

4 Conclusion
In the end, the relationship between the religious life and ignorance offers
a lesson of hope. Ignorance can be marshaled as evidence both for and
against theoretical conceptions of the divine, but it is not decisive in either
case. Furthermore, ignorance appears compatible with a life of religious
virtue and a life of religious faith. While it is surely worthwhile to improve
our knowledge of the world and whatever supernatural elements it may
contain, our current ignorance need not paralyze us. We can live, hope,
and even worship despite, and perhaps because of, our ignorance.
9 Epistemic Injustice and the Preservation of
Ignorance

Miranda Fricker

Ignorance is not always bad; far from it. Looking at the issue in its most
general aspect there is the obvious point that for finite beings, massive
ignorance is a precondition of having an epistemically functional life, for
cognitive overload is an epistemic liability. There is an indefinite, indeed
infinite, number of things that we do not have the slightest need to
know – the number of hairs on your head at midnight on your next
birthday, for instance. Furthermore, we actively need not to know most
of them (or not to spend time and energy investigating them) in order to
conserve cognitive capacity for those things that we do need to know.
Less abstractly, there is also the point that there are many things it would
be morally and/or prudentially bad to know – intimate details that are
none of our business; techniques of criminality; methods of rekindling
old ethnic hatreds in a population. These points are familiar from
debates about ‘the value question’ in relation to knowledge.1
Furthermore, as Cynthia Townley has argued, many forms of epistemic
cooperation and many of the dispositions involved in epistemic virtues
generally depend crucially upon our leaving some useless or harmful
things unknown by passively or actively preserving others’ ignorance of
things they need not or should not know (Townley 2011). In short, good
epistemic practice is necessarily highly selective in all sorts of ways.
What matters is that we know what we need to know, expanding out-
wards to the broader aim of knowing and telling what we should know
and tell, given our purposes and broadly ethical obligations all things
considered. Good epistemic conduct needs to be understood as the
maintenance of appropriate balances of knowledge and ignorance, in
oneself and also in relation to others.
This opening reflection on the epistemic value of ignorance and its
place in the epistemic economy directs our attention to the basic norma-

1
See Sosa’s example of ‘trivial’ knowledge (2002, p. 156); and Zagzebski’s examples of
prudentially and morally ‘bad’ knowledge (2003, p. 21).

160
 Epistemic Injustice and the Preservation of Ignorance 161

tive ambivalence in our use of the term.2 ‘Ignorance’ may refer simply to
an epistemically innocent absence of knowledge (this absence being
advantageous or disadvantageous, as the case may be, without any reflec-
tion on the conduct of the epistemic subject in question); or alternatively
it may refer to some kind of cognitive failure, which might be non-
culpable (perhaps the result of misleading evidence) or which might, on
the other hand, represent a blameworthy failure to put the requisite effort
or skill into knowing something one ought to know.
This paper will focus on those forms of culpable and non-culpable
ignorance that are created or preserved by one or another kind of epis-
temic injustice that I have elsewhere labelled ‘testimonial injustice’ and
‘hermeneutical injustice’.3 I shall discuss the first only briefly, for it is
more specifically in relation to hermeneutical injustice that new and
complex issues have recently been raised concerning various different
forms of ignorance that can be involved in this phenomenon.
In particular I hope to say something useful about the place of ‘wilful’4
or motivated ignorance, and to thereby contribute to recent debates in
which the phenomenon of hermeneutical injustice has been related to
what Charles Mills has termed ‘white ignorance’.5 Ultimately I shall
argue that the phenomenon Mills characterises on the whole picks out
a different kind of ignorance from any that is involved in hermeneutical
injustice. But I shall also argue that the two categories can overlap.

1 Preserving Patterns of Social Ignorance: Testimonial

Injustice and Hermeneutical Marginalisation
When the level of credibility attributed to a speaker’s word is reduced by
prejudice operative in the hearer’s judgement, the speaker suffers
a testimonial injustice. Despite the specific label, the speech act in which
his word is expressed need not be strictly that of testimony or telling, but
might equally be the airing of an opinion, suggestion or relevant possibi-
lity. Furthermore, as Christopher Hookway has suggested, it might even
be occasioned by the asking of a question that is designed to contribute to
some shared inquiry.6 The prejudice driving any case of testimonial
injustice may or may not be a belief, and it operates specifically in the
2
For debate about what ignorance is necessarily ignorance of, see for instance the exchange
between Pierre Le Morven (2010, 2011) and Rik Peels (2011a).
3
See Fricker (2007, 2013b).
4
The term ‘willful ignorance’ is from Gaile Pohlhaus (2012). 5
See Mills (2007, 2015).
6
I am grateful to Chris Hookway for this point that someone who puts a question as
a contribution to collective inquiry (perhaps in the classroom) might find her question
passed over due to prejudice (Hookway 2010). I hope I may ultimately be allowed this as
a limiting case of testimonial injustice, even though it concerns a speech act that is not an
162 Miranda Fricker

hearer’s judgement of credibility, where the judgement may be unreflective

and spontaneous – a matter of ingrained habit. (The trained quasi-
perceptual dispositions governing such judgements I have elsewhere
labelled the hearer’s ‘testimonial sensibility’.) The influence of prejudice
in judgements of credibility can make itself felt regardless of the hearer’s
beliefs, indeed in spite of them, for prejudice can operate unconsciously
or, as we have now learned to say, at the level of ‘implicit bias’.7
Testimonial injustice’s obvious connection to ignorance is that in cases
where the speaker knows that p and the prejudice operative in the hearer’s
credibility judgement prevents her learning that p from the speaker, other
things equal she thereby stays ignorant of p.
Testimonial injustice not only blocks the flow of knowledge, it also
blocks the flow of evidence, doubts, critical ideas and other epistemic
inputs that are conducive to knowledge. The free circulation of these
epistemic goods is conducive to knowledge not only in the direct sense
that ready-made items of knowledge may themselves be transmitted, but
also in the indirect sense that such items tend also to constitute reasons to
believe other things, so that they may have the epistemic power to convert
other of the hearers beliefs into knowledge. The obstructions that testi-
monial injustice introduces into the circulation of such epistemic items is
therefore not only bad for the person whose word is prejudicially down-
graded; it is epistemically bad for the hearer, and for the epistemic system
quite generally. An epistemic system characterised by testimonial injus-
tice is a system in which ignorance will repeatedly prevail over potentially
shared knowledge, despite speakers’ best efforts. Where a speaker knows
something the hearer does not (and where the level of credibility deficit is
such that the hearer does not accept what she is told) the hearer’s ignor-
ance is conserved. Alternatively, where the speaker is offering evidence
with a (positive or negative) bearing on something the hearer already

assertion. The label ‘testimonial injustice’ was always explicitly intended capaciously, to
include not only the broad class of tellings but also cases where a speaker ‘expresses
a personal opinion to a hearer, or airs a value judgement, or tries out a new idea or
hypothesis on a given audience’ (Fricker 2007, p. 60). Hookway’s case of the student’s
relevant question admittedly stretches my characterisation; but provided one can regard
the asking of such a question as potentially vulnerable to a prejudicial credibility deficit, then
it seems more or less to belong to the same category. I would certainly acknowledge that
this requires us to take ‘credibility’ in its everyday sense, as covering the wide range of
respects in which what someone says may be taken as more or less authoritative. Such
a colloquial construal is supported by the fact that the object of any credibility judgement
includes not only what is said but also the speaker. At any rate, I hope these considerations
provide enough commonality to keep the diverse possibilities sufficiently unified under the
category ‘testimonial injustice’.
7
There is a fast growing philosophical literature drawing upon empirical work in psychol-
ogy on implicit bias. See, for instance, Holroyd (2012), Saul (2013), Gendler (2014),
Nagel (2014), Leslie (forthcoming) and Brownstein and Saul (eds.) (2016).
 Epistemic Injustice and the Preservation of Ignorance 163

believes but does not know, then the hearer misses out on reasons which
(if positive) might render her belief knowledge or at least lend it greater
justificatory weight; or which (if negative) might disabuse her of a false
belief, or at least reveal it as less well supported than it had seemed. Either
way, an opportunity for epistemic improvement is lost, and ignorance
prevails.
A further, more buried, form of epistemic damage caused by testimo-
nial injustice is that, where it is persistent and socially patterned (as
anything driven by prejudice is likely to be), it will tend to create or
increase hermeneutical marginalisation. That is to say, it will tend to create
and sustain a situation in which some social groups have less than a fair
crack at contributing to the shared pool of concepts and interpretive
tropes that we use to make generally shareable sense of our social experi-
ences. We might gloss this idea of a pool of concepts and interpretive
tropes as ‘shared social meanings’, where the idea is that while this pool
will surely not exhaust all the various up and running sets of social mean-
ings that are being used locally by this or that group in a given society, the
shared pool (elsewhere I have called this the ‘collective hermeneutical
resource’) contains only meanings that just about anyone can draw
upon and expect those meanings to be understood across social space
by just about anyone else. The collective hermeneutical resource contains
those concepts and conceptualisations that are held in common.
This means that being a member of a social group that does not
contribute on an equal footing with other groups to that shared inter-
pretive resource (a position of hermeneutical marginalisation) puts one at
an unfairly increased risk of having social experiences that one needs,
perhaps urgently, to understand and/or communicate to certain powerful
social others—to a teacher, an employer, a police officer, a jury – but
which cannot be made mutual sense of in the shared terms available.
We are only now, for instance, entering a historical moment in the West at
which it is increasingly possible for a young person originally assigned as
‘male’ to be able to say to a parent, teacher or friend that he has always felt
himself to be a girl in ‘the wrong body’ and hope to be understood as
expressing an intelligible experience. Increasingly the various concepts
and conceptions of how sex, gender, sexual orientation and other deep
identity affiliations may be organised and reorganised in an individual’s
experienced identity – notably the concept of trans together with its less
established counterpart cis – are gradually entering the shared hermeneu-
tical resource instead of staying local to the trans community. Still now,
where a trans woman might attempt to describe her experience of gender
identity to a social other who does not share the relevant concepts, she is
unlikely to be able to make herself much understood, and this is where her
164 Miranda Fricker

remaining hermeneutical marginalisation will manifest itself in the unfair

deficit of intelligibility that constitutes a hermeneutical injustice. Like
testimonial injustices, this kind of hermeneutical injustice preserves
ignorance, for that which remains insufficiently intelligible to the relevant
social other cannot be passed on to them as knowledge.
Here we see how closely the two kinds of epistemic injustice are related:
testimonial injustice can create or sustain hermeneutical marginalisation
by blocking the flow of reports, ideas and perspectives that would help
generate richer and more diversified shared hermeneutical resources that
all can draw on in their social understandings, whether of their own or of
others’ experiences. Therefore the broad patterns of testimonial injus-
tice – most likely patterns created by the operation of negative identity
prejudices, inasmuch as these are the chief systematic prejudices – will
tend to reproduce themselves as patterns of hermeneutical marginalisa-
tion, and it is these that give rise to systematic hermeneutical injustices.
Thus we can see how the preservation of hearer-ignorance that is the
likely effect of any instance of testimonial injustice can contribute directly
to the hermeneutically marginalised position of the speaker. And
a hermeneutically marginalised speaker is vulnerable to hermeneutical
injustice. Charles Mills has noted this close connection between the two
kinds of epistemic injustice in respect of race:

Applying these concepts [of testimonial and hermeneutical injustice] to racial

domination, we could say that white ignorance is achieved and perpetuated
through both varieties working in tandem: a general scepticism about non-white
cognition and an exclusion from accepted discourse of non-white categories and
frameworks of analysis. Thus a double handicap will result—people of color will be
denied credibility and the alternative viewpoints that could be developed from taking their
perspective seriously will be rejected [italics added]. . . (Mills 2015 p. 222)

So the two kinds of ignorance that are preserved through the operation of
the two kinds of epistemic injustice are causally connected, and this
interconnection is part of why our subject-variant areas of ignorance—
especially our ignorance of different areas of our shared but dramatically
stratified social world—tend to display the patterns of social power.

2 Clarifying Hermeneutical Injustice: Spaces for Localised

Hermeneutical Practices
Hermeneutical injustice is internally diverse in various dimensions. One
internal differentiation we can usefully emphasise is between two sorts of
case. The difference is between a radical case where the person concerned
is at least temporarily unable to make full sense of her own experience
 Epistemic Injustice and the Preservation of Ignorance 165

even to herself; and a more moderate sort of case where she understands
the nature of her own experience perfectly well, and, furthermore, is able
to communicate it to members of a social group to which she belongs, and
yet she is unable to render it intelligible across social space to some
significant social other to whom she needs to convey it.
In Epistemic Injustice (Fricker 2007) I tried to bring out this distinction
by way of a contrast between what we might call a ‘maximal’ and
a ‘minimal’ case – that is, between a case where the individual was not
in a position to make proper sense of her own experience even to herself;
and, by contrast, a case where the individual could make perfect sense of
it, and could have communicated it to almost any social other except the
particular social others he specially needed to communicate it to. These
two opposite extremes were intended to imply a continuum of possibili-
ties in between – that is, a range of cases in which there is shared intellig-
ibility across an increasingly large group or groups. The maximal
example – drawn from Susan Brownmiller’s memoir of the US women’s
liberation movement (Brownmiller 1990) – was that of a woman in late-
1960s North America, Carmita Wood, who was being sexually harassed
at work but for whom extant hermeneutical resources did not enable her
to experience this lucidly for what it was, so that while she experienced it
as upsetting, intimidating, demeaning, confusing . . . somehow she was
also aware that these forms of understanding did not capture it.
As recounted by Brownmiller, Carmita Wood remained confused about
what it was she was experiencing, because there was an objective lack of
available concepts with which to make proper sense of it. Her achieve-
ment was to find a community of women who together created a safe
discursive space in which to explore their experiences and find a way of
interpreting them that rendered them more fully intelligible. Through
dialogue within the group they hit upon a critical composite label, ‘sexual
harassment’, and they overcame their hermeneutical marginalisation in
this regard by demanding that the term and the interpretation it expressed
become part of the wider shared vocabulary.
In The Epistemology of Resistance, José Medina emphasises that margin-
alised groups may often have perfectly functioning and sophisticated sets
of interpretive practices up and running within their social group or
community, which however do not work communicatively outside the
group – the non-sharedness of the requisite concepts and interpretations
reflecting the fact that the ‘privileged’ meanings held in common are
inadequate.8 This is indeed worth emphasising, and in this connection

8
Medina (2013). See also Medina (2012), Mason (2011), Pohlhaus (2011) and Dotson
(2012).
166 Miranda Fricker

I would reaffirm the idea that the concepts and meanings that are shared
by all are bound to reflect, in the broad view, the perspectives and
experiences of those groups with more social power generally, for the
reason that those with more social power are very likely to be over-
contributors to the shared hermeneutical resource. (That tendency, that
alliance of hermeneutical power with other kinds of social power, is
present in the very idea of hermeneutical marginalisation.) Accordingly,
the possibility of localised hermeneutical practices is built in to the picture
of how Carmita Wood and her fellow consciousness raisers overcame
hermeneutical injustice. The group was of course not a pre-existing com-
munity, but like other such groups it swiftly developed a voice of its own,
operating in a relation of dissonance and dissent as regards mainstream
understandings. If we jump forward a couple of years from the time of the
consciousness raising group’s first meetings, we would find a fully opera-
tive localised hermeneutical practice among feminists who readily named
sexual harassment for what it was, even while that concept had not yet
entered the shared hermeneutical resource, recalcitrant employers and
all, as later it came to do. This represents a localised well-functioning
hermeneutical practice that nonetheless leaves its practitioners suscepti-
ble to hermeneutical injustice whenever they should attempt to render the
experience intelligible across social space to others who are non-
conversant, perhaps resistant to, the requisite modes of interpretation.
Contrasting with the maximal case exemplified by Carmita Wood, the
minimal case of hermeneutical injustice presented in Epistemic Injustice
also, and more explicitly, illustrates the possibility of a fully functioning
yet insufficiently widely shared hermeneutical practice. It is already a case
of such a practice, though not this time on the part of anyone generally
lacking in social power. The example was that of Joe, the central character
in Ian McEwan’s novel Enduring Love (McEwan 1998). Joe is being
stalked by a religious fanatic who wants to convert him. Joe is an edu-
cated, white, middle-class man, whose hermeneutical marginalisation (if
any – it is the vanishingly minimal case) is highly specific, localised to the
particular matter in hand, and whose experience he himself has no diffi-
culty in understanding and would easily be able to communicate to
members of almost any social group.9 And yet when it comes to the
most important social body to which he needs to be able to communicate
it, namely the police, he finds they are not in a position to make proper
sense of it – there is quite literally no appropriate box to tick on their form.
9
That his hermeneutical marginalisation is so highly specific to one area of experience (that
of being non-violently stalked) renders Joe’s hermeneutical injustice, in the terminology
I used in Epistemic Injustice, a thoroughly ‘incidental’ rather than ‘systematic’ case of the
injustice (see Fricker 2007, chapter 1, p. 7).
 Epistemic Injustice and the Preservation of Ignorance 167

Thus I would argue that a commitment to the existence of localised

interpretive practices that may perfectly capture a given range of experi-
ences but whose meanings are not sufficiently shared across wider social
space is already present at the heart of the original account of hermeneu-
tical injustice. I gladly acknowledge, however, the importance of centre-
staging, as others have done,10 what I am here calling midway cases of
hermeneutical injustice – those situated somewhere between maximal
and minimal in virtue of the fact that they concern existing communities
who operate localised or in-group hermeneutical practices that are none-
theless not shared across further social space. These are cases in which
there are sophisticated interpretive practices, perhaps with their own
history of internal challenge and change, which are already functionally
entrenched for a given social group or groups, but not shared with at least
one out-group with whom communication is needed. Members of such
hermeneutically self-reliant groups are vulnerable to hermeneutical injus-
tices whose form does not involve any confused experiences whatever, but
only frustratingly failed attempts to communicate them to members of an
out-group. (In the next section we shall see this midway form of herme-
neutical injustice put to work in relation to a special case of white ignor-
ance.) Medina is right to emphasise that the intersectional ignorances
created by the possession and non-possession of this or that cluster of
interpretive concepts growing out of this or that area of social experience
tell a ‘polyphonic’ or multi-voiced story of power and resistance, societal
conceptual impoverishment and localised interpretive sophistication and
creativity. These opposing energies are present in both maximal and
minimal cases, but the creative and affirming energy involved in resisting
mainstream meanings and nurturing instead a set of more localised con-
cepts and interpretations is obviously more to the fore in those cases of
hermeneutical injustice that start from a situation in which a relatively
powerless group has developed well-entrenched but localised interpretive
practices of its own. In such cases, in-group intelligibility is doing just fine;
and any hermeneutical injustices that arise will be strictly a matter of
unfairly limited communicative intelligibility in relation to an out-group.
An illustration of such a midway case of hermeneutical injustice might
be drawn from the history of post-colonial race relations in the UK.
Drawing on an account of the experience of growing up in postwar
Britain as the children of Caribbean immigrants to ‘the mother country’ –
often symbolised by the Empire Windrush arriving at Tilbury Docks in
1948 – we find that the experience of integration into British life was not

10
See, for instance, Mason (2011), Pohlhaus (2011), Dotson (2012) and Medina (2012,
2013).
168 Miranda Fricker

structured in relation to the conceptual poles of ‘acceptance’ or ‘rejection’

to which the white perspective gave rise. Instead the black experience was
structured in relation to the concept of citizenship. Mike Phillips and
Trevor Phillips recount it as follows:

We observe that the overt declarations of racist hostility which were commonplace
in the fifties have, more or less, disappeared from public life in Britain. On the
other hand, it is clear that racial hostility and exclusion are a routine part of British
life, and few black British people can be in any doubt that the majority of their
fellow citizens take the colour of their skins to be a characteristic which defines
what they are and what they can do.
At the same time, paradoxically, among ourselves we never interpreted the
racial discrimination or hostility that we encountered as ‘rejection’, largely
because we never believed that ‘acceptance’ or ‘rejection’ was a choice available
to Britain. Far from it. Our instinct told us that such notions were merely part of
a racialised idiom, describing an identity which had long ago ceased to be relevant.
For us the issue was . . . about our status as citizens . . . (Phillips and Phillips
1998, p. 5)

What their instinct told them formed the cornerstone of their localised
conceptualisation of their situation – an understanding not supplied by the
shared hermeneutical resource dominated by the ‘racialised idiom’ that
characterised the perspective of white Britain – and it delivered a mode of
understanding which they were rightly concerned to insist on introducing
into the common pool of understanding: an idea of black colonial immigrants
as fellow British citizens. The Empire had told their parents that Britain was
their mother country and it seems that they had, in part at least, believed and
internalised this fact – many had signed up to fight in the war under the
identity it imposed – so that those arriving in Britain on ships like the
Windrush ‘regarded their Britishness as non-negotiable’ (Phillips and
Phillips 1998, p. 5). The mother country had made them British, and now
it was a matter of holding her to the full implications of that status. Ideas of
‘acceptance’ or ‘rejection’ may have structured white consciousness around
this immigration, but the immigrant population was living an independent
and novel conceptualisation according to which they were black British
citizens – a hermeneutical trope seemingly absent from the repertoire of the
white population. One could say the concept of a black British citizen had not
yet taken hold in white British consciousness, and white resistance to that
conceptual neologism was such that it would take some significant time to
do so.
In this example, it seems the sooner the new conceptualisation could
become widely entrenched in the shared hermeneutical resource the better.
But it is worth remembering that there can be cases in which it may not be in
the interests of an oppressed group to fight immediately for the introduction
 Epistemic Injustice and the Preservation of Ignorance 169

of local meanings into the wider collective hermeneutical resource. (This is

a point made by Mills, quoting ‘the black American folk poem, “Got one
mind for white folks to see / Another for what I know is me” (Mills 2007,
p. 18); and also emphasised by Medina.11) Sometimes there can be signifi-
cant advantage in keeping things local, perhaps so that there is more time in
a safe space to develop one’s dissenting forms of understanding, or perhaps
simply because the wider climate makes it pointless, or too dangerous to try
anything else. To take an example now from the history of race in the United
States, in a radio interview the writer Alice Walker describes aspects of her
upbringing under segregation in the American South in terms that indicate
the value of maintaining hermeneutical privacy. Confident in their own
interpretations of the social world, her parents inculcated in their children
a way of understanding racial oppression that might be read as incorporating
a certain security in on-going hermeneutical separation:
Lucky for us we lived very far in the country. We saw very few white people. And
when we went to town we followed rules about where we could go. And we just
followed our parents. They basically helped us to see white people as, you know,
very stunted. That was just the way they were. There was nothing you could do
about it, they were just like that. (Who knew why they were like that?) And that
was helpful. They were discussed as if they were the weather . . . Like ‘Oh well,
that’s how they are. You know, what we try to encourage in our children, they beat
it out of their children. They don’t want their children to be kind. They don’t want
their children to ever see a black person and think of them as human.12

Here the idea that white people were ‘very stunted’ captures a localised
hermeneutical practice that embodied a clear and confident knowledge
that black people were not as white people painted them, and moreover
that the racial attitudes of white people only showed them up as seriously
morally damaged. The moral knowledge at large in the black community
could not on the whole cross the segregated social space to find intelligible
expression in the white community. Thus the hermeneutical practices
that produced that moral knowledge in the black community was, judging
from Walker’s account, highly localised. The comment about ‘weather’ is
particularly resonant in this connection. Perhaps when the terms of
segregation mean that the normal ‘reactive attitudes’ of moral participa-
tion can only be a losing game, it is empowering to reach instead for what
P.F. Strawson identifies as the ‘objective’ attitude of non-engagement, so
that the agency of certain others is received as weather – meaningless (if
11
‘As many Latina feminists and colonial theorists have argued, colonized peoples have
a long tradition of exploiting the ignorance and hermeneutical limitations of the coloni-
zers to their advantage, which can be justified for the sake of their survival’ (Medina 2013,
p. 116).
12
Desert Island Discs. 2013. BBC Radio 4, May 24th.
170 Miranda Fricker

potentially dangerous) causal impacts to be managed, tolerated,

avoided.13 Keeping one’s hermeneutical practices localised in
a situation such as this might be a decision to leave the powerful to their
pitiful ignorance, safeguarding one’s localised forms of moral and social
understanding as a source of in-group solidarity and strength.

3 White Ignorance and Hermeneutical Injustice

Continuing with questions of race and the different forms of ignorance
that can be generated and preserved by the operation of epistemic injus-
tice, let me now relate our discussion to a different kind of ignorance:
what Charles Mills has named ‘white ignorance’ (Mills 2007). I would
like to offer an account of the boundaries of the two phenomena.14 I shall
argue that for the most part the ignorance that is produced and main-
tained by hermeneutical marginalisation, and made manifest in herme-
neutical injustice, is different in two key respects from the ignorance in
white ignorance. First, white ignorance is normally epistemically culp-
able; and, second, it does not generally involve any paucity of concepts on
anyone’s part. By contrast, in a case of hermeneutical injustice the
uncomprehending hearer is normally epistemically non-culpable; and
there is always, definitively, a paucity of shared concepts. However,
I hope to identify where the two phenomena overlap.
Most generally speaking, ‘white ignorance’ is a racialised form of ideo-
logical thinking. It names a certain kind of collective interested or motivated
cognitive bias in what social interpretations and/or evidence for such
interpretations a racially dominant group attends to and integrates into
the rest of their beliefs and deliberations. More specifically the label ‘white
ignorance’ names a motivated bias on the part of white people taken as
a group that leaves them ‘ignorant’ (in this special sense) of the situation
of their black compatriots taken as a group. We might say it names a form
of collective denial in the white community about some uncomfortable
truths.15 It therefore typically exhibits a culpable motivated irrationality.
Indeed in most cases of ‘white ignorance’ as that phenomenon is dis-
cussed it involves some self-serving epistemic fault on those inhabiting the
standpoint of whites – a conscious or unconscious resistance to accepting

13
See Strawson (1974).
14
See the substantial discussion of this issue in Medina (2013); and discussions in Mason
(2011), Pohlhaus (2012) and, in different terms, Dotson (2012).
15
In Mills’ list of elements it is clear that various forms of motivated irrationality, denial, or
other forms of epistemic culpability characterise the phenomenon. He says, for instance:
‘the dynamic role of white group interests needs to be recognized and acknowledged as
a central causal factor in generating and sustaining white ignorance’ (Mills 2007, p.34).
 Epistemic Injustice and the Preservation of Ignorance 171

or learning about the sources of their social advantage, for instance. Such
epistemic faults are generally culpable. As Rebecca Mason succinctly puts
it: ‘white ignorance is a kind of epistemically culpable and morally nox-
ious miscognition that facilitates the maintenance of the status quo’
(Mason 2011, p. 302).
Mills first discussed the phenomenon in the framework of the United
States, but more recently he has made clear that he considers the issue to
have global application. Referring back to his paper ‘White Ignorance’
(Mills 2007) he explains:
My discussion in the essay was focused mainly on the United States, but
I intended the application of the concept to be much broader. Insofar as the
modern world has been created by European colonialism and imperialism, and
insofar as racist assumptions/frameworks/norms were central to the theories
justifying white Western conquest and domination of that world, we would expect
white ignorance to be global. (Mills 2015, p. 217)

We might illustrate his point with another example drawing on British

colonial history, as pointed out by Mike and Trevor Phillips in their
discussion of the ignorance produced by the sheer absence of black
soldiers from the many British films about the war made in the post-
War period:
. . . it comes as a shock now to note the complete absence of black Caribbean or
African participants in the plethora of British films about the Second World War.
After all, the involvement of black colonials was a fact that was a part of our
experience . . . Our astonishment was, and still is, to do with the extent to which
they had disappeared, had been expurgated from the story, as if they had never
existed. (Phillips and Phillips 1998, p. 5)

Let us look closely at Mills’ characterisation of white ignorance in order to

see (a) whether all cases are epistemically culpable, and (b) whether any
involves the paucity of concepts that is definitive of hermeneutical injus-
tice. Mills presents two main forms of white ignorance, and they share
what he calls ‘racialized causality’ – that is, each involves the white
community failing to grasp certain facts or to hold certain truthful inter-
pretations of their social world where a significant part of the explanation
why not is race.
First, such racially caused ignorance might take the form of an indivi-
dual’s active racism blocking certain truths; or, second, it might be more
structural in form. Mills says in this connection:
[T]he racialized causality I am invoking needs to be expansive enough to include
both straightforward racist motivation and more impersonal social-structural
causation, which may be operative even if the cognizer in question is not
racist . . . But in both cases, racialized causality can give rise to what I am calling
172 Miranda Fricker

white ignorance, straightforwardly for a racist cognizer, but also indirectly for
a nonracist cognizer who may form mistaken beliefs (e.g., that after the abolition
of slavery in the United States, blacks generally had opportunities equal to whites)
because of the social suppression of the pertinent knowledge, though without
prejudice himself. (Mills 2007, p. 21)

In the case of the straightforward ‘racist cognizer’, the epistemic culp-

ability is clearly on display: depending on quite what form the racism
takes, such prejudiced cognizers are allowing some racist motive (perhaps
racial contempt, or some kind of racial self-aggrandisement) to distort
their perception of the social world and their place in it. Such motivated
irrationality is plainly epistemically culpable (though of course there can
be mitigating circumstances that reduce the degree of appropriate
blame). In cases of hermeneutical injustice, by contrast, neither speaker
nor hearer need be blameworthy for the failure of intelligibility. In itself
hermeneutical injustice is a purely structural phenomenon with no indi-
vidual perpetrator.16 In some cases the hearer would of course be blame-
worthy – for instance if she were self-interestedly to resist the meanings
being offered.17 But no such fault is a necessary feature of hermeneutical
injustice per se. Indeed part of the intrigue of the phenomenon is that it
can happen so widely without epistemic fault, which is why it calls not
only for increased individual virtue but also for structural remedy through
social policies and institutional arrangements that would increase equality
of hermeneutical participation.
What about the question of conceptual poverty – the requisite herme-
neutical gap? In the case of the straightforward racist cognizer’s white
ignorance there is no hermeneutical gap, indeed no poverty of concepts at
all, for the racist cognizer’s ignorance is not caused by any lack of
conceptual-interpretive resources. Let all the hermeneutical resources
stand available to him, what he lacks is the epistemic discipline to apply
the extant resources in an epistemically responsible way so as to achieve
cognitive contact with social reality. Given these features, the white
ignorance of the straightforward racist cognizer is clearly not any kind
of hermeneutical injustice. It is an independent phenomenon, played
out at the level of belief and (culpable) epistemic conduct.
16
Medina develops the point that individuals can however collude in hermeneutical injus-
tice by failing to be virtuous hearers (see Medina 2012, 2013, chapter 3). The point is well
taken, but I would resist his conclusion that this reveals that there are, after all, individual
‘perpetrators’ of hermeneutical injustices. Failures of virtue are bad in themselves, and
when we fail to be appropriately open to the perspectives of others we are doing some-
thing bad and may even be wronging them as individuals. But being culpable for one’s
part in a broader injustice makes one a perpetrator only of that part; it does not make one
a perpetrator of the broader injustice itself.
17
See Mason (2011), Pohlhaus (2011), Dotson (2012) and Medina (2013).
 Epistemic Injustice and the Preservation of Ignorance 173

Let us look now to the second sort of case that Mills gives us. This is the
‘more impersonal, social-structural’ case of the non-racist cognizer who
nonetheless ‘may form mistaken beliefs . . . because of the social suppres-
sion of the pertinent knowledge, though without prejudice himself’.
Perhaps such social suppression could be a matter of certain parts or
aspects of history not being taught at school; or perhaps another example
might be the cultural forgetting of the involvement of black Caribbean
and African soldiers in the Second World War, as noted by Mike and
Trevor Philipps in relation to British film. In most of these social-
structural cases of white ignorance, I take it, the individual remains
epistemically culpable to some significant degree inasmuch as it is likely
that she ought to be able to remain critically alive to at least some of the
ways in which the epistemic situation has been distorted. But equally one
can imagine (I emphasise, imagine) scenarios in which the individual is
not culpable, insofar as it is also possible that the epistemic fault driving
the ‘social suppression of the pertinent knowledge’ could be exclusively in
the collective, or in some sub-group who is manipulating collective
knowledge, in a manner that no individual could reasonably be expected
to detect.18 This in-principle possibility of individually non-culpable
white ignorance suggests that Mill’s social-structural kind of white ignor-
ance can in principle be non-culpable – which prompts one to ask whether
it might also constitute a case of hermeneutical injustice.
As before, however, we must also look for some kind of conceptual gap
caused by hermeneutical marginalisation, for it cannot be a hermeneuti-
cal injustice without at least some impoverishment in shared conceptual
resources. But in itself the ‘social suppression of the pertinent knowledge’
does not involve any loss of interpretive concepts or conceptions.
The white-ignorants19 in question might continue to have available to
them perfectly adequate conceptual resources for knowing that X, and yet
fail to know that X owing to the suppression of the requisite knowledge
itself – once again, a dysfunction at the level of belief and evidence rather
than the level of conceptual repertoire and intelligibility. White British
forgetfulness about the involvement of African and Caribbean soldiers in
the Second World War, for instance, involved no deficit of intelligibility,
for the shared hermeneutical repertoire was quite rich enough to have
supported the lost knowledge.

18
I have argued elsewhere that such a case might represent one of epistemic agent-regret
(Fricker 2016).
19
As Mills makes entirely clear, and by way of parallel with the phenomenon of false
consciousness on a Marxist picture, one does not have to be white to become embroiled
in white ignorance (Mills 2007, p. 22). But it helps.
174 Miranda Fricker

We might go on, however, to envisage a third, albeit non-standard,

case. One can imagine structural cases where the ‘social suppression of
the pertinent knowledge’ has included suppression of concepts requisite
for that knowledge. If this conceptual suppression is confined to the
privileged group, a genuine deficit in hermeneutical resources for the white
community would result, and thereby a deficit in the shared hermeneu-
tical resource. With the hermeneutical gap so envisaged, we are closer to
a case of white ignorance that is also one of hermeneutical injustice.
Given that the paucity of concepts in this case is all on the part of whites,
someone might wonder whether it was the white community that was
subject to the hermeneutical injustice. Not so; for in order for such a case
to constitute a hermeneutical injustice the deficit of concepts in the white
population would also have to be unfair to them in some way. It is true
enough that, as Medina emphasises (Medina 2013, p. 108), such
a hermeneutical deficit would clearly be bad for the white community in
a purely epistemic sense, for there is important social knowledge they
would be missing out on.20 (So it was for Carmita Wood’s harasser.)
It may well be morally bad for them too (as it was for the harasser, who
was prevented from grasping the ethical significance of his own behaviour,
and was to that degree alienated from the meaning of his own actions).21
But still, the disadvantage cannot be an injustice done to them, because ex
hypothesi this very epistemic disadvantage plays more generally to their
social advantage – that is the whole point: white people are represented as
having an interest in not knowing certain threatening facts, and if the very
concepts required for such unsettling knowledge have been suppressed,
then they are all the safer from having to confront it. Rather it is the black
community who suffers the hermeneutical injustice, for it is they who are
asymmetrically socially disadvantaged by the whites’ conceptual deficit
that entails the equivalent deficit in the shared hermeneutical resource.
What we have now arrived at in pursuing the overlap between herme-
neutical injustice and white ignorance is a form of hermeneutical injustice
that belongs in the range of cases identified in the previous section as
midway between maximal (Carmita Wood) and minimal (Joe) forms.
Such cases are those in which one group’s communicative attempts meet
with failure owing to a paucity of concepts on the part of an out-group and

20
Laura Beeby too has emphasised the importance of the purely epistemic disadvantage
suffered by the more powerful party in such cases (Beeby 2011).
21
Jason Stanley expresses a general version of this point in relation to legitimising myths:
‘false ideologies harm the elites in ways that cut deeper than material interest. The reason
that members of unjustly privileged groups are led to adopt legitimizing myths is that they
cannot confront the possibility that their actions are unjust. False ideologies blind even
those they seem to help, by making them “untrue to themselves”’ (Stanley 2015, p. 265).
 Epistemic Injustice and the Preservation of Ignorance 175

therefore in the shared hermeneutical resource. Among those cases, we can

locate the racially motivated concept-suppression scenario that we have
identified as a (non-standard) case of white ignorance. The motivated
concept suppression among the dominant white community means that
the hermeneutically marginalised black community nonetheless possesses
locally operative meanings that capture their experiences but which cannot
function properly in communicative attempts with social bodies that oper-
ate with the impoverished shared conceptual repertoire. However rich the
black community’s conceptual resources might be, these resources do not
get integrated into the shared resource, because the white community has
an interest in keeping them out. This, at last, is the overlap we have been
looking for: a white ignorance whose explanation is a conceptual deficit (on
the part of whites, and ipso facto a deficit in the shared hermeneutical
resource) that is significantly caused by the black community’s hermeneu-
tical marginalisation. In such a case, motivated conceptual poverty on the
part of a dominant racial group works to preserve their local ignorance of
a significant dimension of the social world and blocks another racial group
from making good that ignorance.
What about the question of epistemic culpability? In our earlier discus-
sion of Mills’ knowledge-suppression case I suggested that such cases
might normally be epistemically culpable, though we could imagine
scenarios where there was no epistemic culpability. The matter turned
on how far it was reasonable, in any given case, to expect the uncompre-
hending hearer to be alert to the distortions in the epistemic system.
The same goes for our concept-suppression example. Here the hermeneu-
tical marginalisation of the black community kettles their concepts,
thereby creating a conceptual lack in the shared hermeneutical resource,
and so preserving white ignorance by creating a barrier to the essential
conceptual means to their understanding expressions of the relevant black
experiences. The question of epistemic culpability in such cases will
depend, as it does in general, upon how far the uncomprehending hearer
could reasonably be expected to have been alert to the fact of her con-
ceptual impoverishment. If she could have known better, then she should
have known better.22 These issues of individual culpability and non-
culpability seem worth thinking about in principle, even if we are pessi-
mistic about how much individuals can really do.23 In cases of herme-
neutical injustice, the requisite structural remedy involves the reduction
of hermeneutical marginalisation; in cases of white ignorance, a whole
22
I have argued more fully for this view of the borders of culpable and non-culpable
ignorance more fully in Fricker (2010).
23
For this concern about the limitations of increased individual virtue, see Alcoff (2010),
Langton (2010) and Anderson (2012).
176 Miranda Fricker

range of structural remedies is no doubt called for.24 Such structural

changes are called for in addition to individual efforts – for, after all,
structural changes are often in significant part the upshot of individual
efforts.

4 Hermeneutical Injustice Is Not Necessarily a Face of

Oppression
I hope to have clarified and defended my original characterisation of herme-
neutical injustice by showing that its core notion of hermeneutical margin-
alisation allows for the sorts of midway and/or motivated cases of
hermeneutical injustice that other writers have rightly emphasised.25 If the
driving thought is that hermeneutical gaps are typically made rather than
found, then I agree. One group’s marginalisation is typically motivated by the
interests of another group whose purposes are served by the marginalisation.
It is therefore in the nature of any marginalisation that ideology, and other
kinds of privileged motivation, will be chief among its causes. Hermeneutical
injustice, like testimonial injustice, is typically a face of oppression – it tends
to preserve ignorance that serves the interests of dominant groups.26
However, I would also affirm that it is important we air possibilities of
hermeneutical marginalisation that are not themselves part of a pattern of
oppression. The category is broader than that, for there can be unfair forms
of hermeneutical marginalisation that are to be explained in terms of more de
facto forms of social powerlessness, or more fleeting kinds of ideological
struggle; and there can sometimes be hermeneutical gaps that are more like
unforeseen consequences of social flux, or of processes that do not particu-
larly reflect the long-term interests of one group over another. Perhaps an
example might be the kind of hermeneutical marginalisation against which
‘teenagers’ (itself a new concept at the time) in the early Sixties rebelled.
They did not get much of a look in to the processes of meaning-making
before that, but they found a noisy way of making new meanings among
themselves, interpreting and constructing their experiences accordingly.
If we imagine early-Sixties teenagers trying and failing to convey to their
parents what was so great about rock’n’roll and everything it stood for,
maybe we confront a case of hermeneutical injustice of the non-oppression
24
See Anderson (2012) for the proposal that racial integration is essential as a structural
feature of institutional epistemic justice. For her more general case for racial integration,
see Anderson (2010).
25
See Mason (2011), Pohlhaus (2011), Dotson (2012) and Medina (2012, 2013).
26
Using the terminology I employed in Epistemic Injustice (Fricker 2007), I would say that
testimonial and hermeneutical injustices in their ‘systematic’ (as opposed to ‘incidental’)
forms are the central cases of epistemic injustice, because it is these forms that reveal the
connection with other dimensions of social injustice.
 Epistemic Injustice and the Preservation of Ignorance 177

kind I aim to leave room for in my characterisation. It is of course political,

since it involves a struggle of power – the power of one generation over the
next – and even of competing ideologies. But we would not normally regard it
as a matter of oppression, for nobody is a teenager for very long, and this kind
of struggle represents a near inevitable process that is part and parcel of on-
going historical change, including ethical change. Such intergenerational
struggle might therefore play a role, even a desirable role, in any human
society.
For these reasons the teenage-culture case is not the kind of power
struggle we would ordinarily characterise as a fight against oppression.
It involves the hermeneutical marginalisation of the younger generation
for sure; but it would be a jaundiced view of the perennial struggle between
one generation and the next to insist that this marginalisation was funda-
mentally oppressive in nature. It is simply (and thankfully) in the nature of
young people to want to make their own world, and that involves a certain
overthrow of parental regime. Where that regime has hermeneutically
marginalised its young, hermeneutical injustices are bound to arise from
youthful attempts to express the new social ideas to the older generation.
Hermeneutical injustice can affect people’s lives in many different
ways. I believe it is most useful to have a theoretical framework that
makes room for all sorts of cases, so that the various degrees of wrongful
unintelligibility can be seen to run from maximal to minimal (from
Carmita Wood’s inarticulable outrage to Joe’s articulate yet ultimately
frustrated communicative attempts); and so that the forms of hermeneu-
tical marginalisation can be seen to run from actively oppressive moti-
vated ignorance (as per the case of motivated concept-suppression white
ignorance) to ordinary attempts by parents to shape a new generation
according to values they understand. The purpose of placing these differ-
ent formations in a single theoretical structure is to reveal the range of
possibilities in all their similarities and differences. Ignorance, as we
observed at the outset, is not always bad; but social ignorance that results
from hermeneutical marginalisation is intrinsically likely to be bad insofar
as it is likely to be conserving ignorance that sustains unequal social
relations. Those cases clearly are oppressive, and they preserve forms of
ignorance that demand to be made good.27

27
Earlier versions of some parts of this paper were first presented at a workshop on José
Medina’s book (Medina 2013) at the Autonomous University of Madrid. I am grateful to
all the participants there for helpful comments, and in particular to Linda Alcoff, Katharine
Jenkins, José Medina and Charles Mills for helpful informal discussions. I am also grateful
to Rik Peels for comments on a draft. Earlier versions of some parts were published online
as part of the Social Epistemology Review and Reply Collective (Fricker 2013a).
10 Ignorance and Racial Insensitivity

José Medina

Some insidious forms of racial oppression operate through patterns of

ignorance that contribute to the stigmatization of racial minority groups.
In this paper I offer an analysis of how racial ignorance produces systema-
tic distrust and miscommunication across racial lines. I explore ways in
which people could become aware of their racial ignorance and of their
complicity with racial injustices. I offer an account of racial insensitivity as
a kind of numbness, reflecting on the affective and cognitive aspects of
people’s inability to respond to racial injustices. I analyze insensitivity as
a form of narrow-mindedness that involves the incapacity to see the point of
view of the other and the resistance to acknowledge certain things that are
hard to live with (such as one’s complicity in the suffering of others).
I argue that epistemically responsible agency demands that we live up to
certain epistemic responsibilities with respect to oppressed groups: these
include our responsibilities for knowing and caring for the injustices they
suffer, and also our responsibilities to fight against their exclusion or
marginalization as communicators and fellow citizens. On my view, epis-
temic responsibility is the precondition of and the basis for other social
responsibilities we have; and until the epistemic responsibilities breached
in racial ignorance are repaired, complicity with racial injustices cannot
be uprooted. The work toward racial justice must begin with the acknowl-
edgment of racial ignorance and the epistemic limitations it creates for
social relationality.
In Section 1, I develop an account of racial insensitivity that begins with
classic accounts of racial ignorance as a form of blindness and continues
with more contemporary analyses of racial insensitivity as a form of
numbness that includes not only perceptual and cognitive elements, but
also conceptual and affective elements. In Section 2, I draw from my
epistemic interactionism and contextualism to articulate a robust notion
of epistemic responsibility. According to my polyphonic view, multiply
situated, overlapping and intersecting perspectives and sensibilities call

I am grateful to Rik Peels, Martijn Blaauw and Ben Ferguson for their critical feedback and
their suggestions for revisions on an earlier draft of this paper.

178
 Ignorance and Racial Insensitivity 179

for multiply situated, overlapping and intersecting forms of responsibility.

I argue that the accountability and responsivity required to address epis-
temic injustices can be of all sorts, but it must always start with the
acknowledgment of one’s epistemic positions and relations, and with the
acknowledgment of the epistemic privileges and epistemic limitations one
has. Finally, in Section 3, I discuss experiences of epistemic discomfort in
educational processes that offer the possibility of ethical growth and the
expansion of one’s epistemic sensibilities.

1 Insensitivity: From Blindness to Numbness

I am an invisible man . . . I am invisible, understand, simply because
people refuse to see me . . . it is as if I have been surrounded by mirrors of
hard, distorting glass. When they approach me they see only my sur-
roundings, themselves, or figments of their imagination – indeed, every-
thing and anything except me.
Prologue to Invisible Man, by Ralph Ellison

A great deal of social interaction happens in the dark, with people acting
“blindly” toward each other and exhibiting a stubborn resistance to
recognize crucial aspects of each other’s identities and lives. In classic
race theory, racial ignorance has been described as analogous to
a perceptual deficit, as a form of blindness.1 But although the traditional
account of racial ignorance as a form of blindness contains many rich and
powerful insights, there are at least three reasons why “numbness” is
a more apt term than “blindness” to describe the deficient epistemic
sensibilities involved in racial insensitivity.
In the first place, the problem of having been desensitized to certain
aspects of social relationality has a perceptual dimension, but the percep-
tual numbness involved is multidimensional and affects not only sight,
but all our senses simultaneously – in particular, it affects our capacity to
hear, to respond to voices and accents and to listen to them properly.
In the second place, the term “numbness” is more appropriate than
“blindness” because, although both are clearly related to our embodied
sentience, the former can be more easily extended to the non-perceptual,
and indeed the epistemic deficiencies involved in racial ignorance go
beyond our perceptual organs. There is an important disanalogy between
the failure of a sensory organ (such as sight) and the communicative and
interpretative failures involved in racial ignorance.
In the third place, the metaphor of blindness does not capture (in fact, it
hides) a crucial feature that characterizes racial insensitivity in its most
1
See, for example, Du Bois ([1903], [1994]).
180 José Medina

insidious forms: its self-effacing nature, its self-hiding and self-denying

mode of operation. Whereas the blind person is acutely aware that there
are things that escape her and she leads her life adjusting to this perceptual
deficit, the racially insensitive person is typically quite oblivious of there
being anything at all she is missing. This is what I have called the meta-
level aspect of racial insensitivity. Racial insensitivity becomes insidious
and recalcitrant when it contains meta-ignorance, that is, when the racially
insensitive person is ignorant of her own ignorance, unable to recognize
that there is anything she is missing concerning racial experiences and
meanings.
Racial insensitivity creates dysfunctional communicative dynamics.
And the very attempt to unmask the lack of attention and sensitivity to
racial issues often triggers communicative dysfunctions: complaints
about racial insensitivity are often answered with complaints about racial
oversensitivity. Especially when claims about racist attitudes and beha-
vior are voiced by members of racial minorities, these claims are often
neutralized and countered by the charge of being oversensitive. And thus
what was meant to be an objective claim about social interaction is
transformed into (or heard as) a purely subjective expression or an
emotional reaction—and a misplaced one at that! The same often hap-
pens when women make the claim that an attitude or action is sexist. Let
me say something brief about the gender case in order to capture some
important insights in the feminist literature, to then focus on the case of
race and the dysfunctional communicative dynamics in allegations of
racial insensitivity and racial oversensitivity.
As Naomi Scheman (1993) argues, the anger voiced by women as
a reaction to gender injustices is a rational response to facts about their
situation and a communicative act that demands recognition and action.
But the kind of social sensitivity that understands women’s anger in that
way and can enter in that non-dismissive communicative dynamic has not
been easy to achieve; it has been one of the achievements of the Women’s
Movement, the result of the consciousness-raising practices that started
in the 1970s; and of course a social achievement that is still partial and
unfinished. For as long as women’s emotional speech acts receive uptake
only as mere expressions of emotions, they are not taken to make any
epistemic or political demand on people: they are not taken to demand
that people revise their beliefs or that they take appropriate action to
respond to them. Because of their emotional nature and because of the
limited uptake they receive as a result, women’s speech acts of protesta-
tion—the communicative acts of denouncing and contesting gender
oppression—are neutralized, rendered devoid of normative content, and
in fact they are transformed into something else: a pure venting without
 Ignorance and Racial Insensitivity 181

any basis, a hypersensitive or hysterical reaction. This communicative

dysfunction amounts to an epistemic injustice:2 in these contexts women’s
discursive agency becomes constrained, unable to perform the speech
acts of contestation, and subject to unfair communicative expectations –
for it is unfair to expect from those who are oppressed to communicate
about their oppression without emotions such as anger and without
affective reactions that reveal how they are personally affected.
The communicative dysfunction in question involves an epistemic injustice
because it involves interpretative and testimonial disadvantages that
diminish women’s capacity to make sense of their experiences and to
use them as compelling reasons for their claims.
The same analysis applies also to claims about racism voiced by racial
minorities. These minority subjects denouncing racism are often depicted
as hypersensitive. They often encounter epistemic obstacles that result in
their unfair treatment as communicators and epistemic agents; that is,
they encounter epistemic injustices that makes it difficult for them to
communicate about their experiences of oppression and thus function
as defense mechanisms that hide and protect the relations of oppression.3
Following the literature on the epistemology of race, in this chapter
I develop an analysis of racial insensitivity in terms of numbness to the
perspectives of racial others and their experiences. In this analysis I want
to emphasize three things about racial insensitivity: (1) that racial insen-
sitivity involves epistemic labor and that in its most insidious form this
insensitivity protects itself through cognitive and affective mechanisms that
make people socially numbed to racial injustices; (2) that racial insensitiv-
ity becomes insidious and recalcitrant when it operates at two levels: at the
object level and at a meta-level; and (3) that racial insensitivity is a numbness
directed both outwards – to the social world, to others – and inwards – to
oneself, thus involving blind spots that result both in social ignorance and
in self-ignorance. The next three sections will discuss those three episte-
mic features of racial insensitivity. Sections 5 and 6 will sketch a notion of
epistemic responsibility in the light of this analysis. The focus of sections 5
and 6 will be the issue of complicity and shared responsibilities.

2
The notion of epistemic injustice has been defined by Miranda Fricker (2007) “as a kind of
injustice in which someone is wronged specifically in her capacity as a knower” (2007, p. 20).
Within the category of epistemic injustice Fricker has distinguished between testimonial
injustice, which involves unfair credibility assessments in testimonial exchanges, and
hermeneutical injustice, which involves unfair limitations in one’s capacity and resources
for making sense of and for communicating one’s perspective.
3
For a full account of racial epistemic injustices and the protective mechanisms inscribed in
racial ignorance, see Medina (2013) and Medina (forthcoming).
182 José Medina

2 Racial Insensitivity as Active (Self-Protecting) Ignorance

Describing racial insensitivity as being numbed captures well how it can
become an active form of epistemic withdrawal, the kind of ignorance that
protects and hides itself through a whole battery of defense mechanisms,
which include both cognitive and affective elements.
Racial insensitivity consists in being cognitively and affectively numbed.
This numbness involves epistemic deficits and vices. It can involve, for
example, a lack of openness to discuss racial problems, to take claims
about racism seriously as claims that make demands on all of us and
require a response. This lack of openness has an important affective
dimension. For example, it can take the form of hearing claims about
racism as personal attacks that call for defensive reactions, or as attempts
to be divisive that should be met with contempt. The kind of closed-
mindedness characteristic of racial insensitivity consists in epistemic
dysfunctions that limit the production and passing of knowledge in epis-
temic activities. Racial insensitivity is the kind of epistemic dysfunction
that involves not only epistemic lacunas and epistemic distortions
(absence of true beliefs or presence of false beliefs), but also the inability
to fill those lacunas or correct those distortions, that is, the inability to
learn. In its most insidious forms, racial insensitivity involves not just
a regular kind of ignorance about racial matters, but what is called in the
epistemology of race “active ignorance”:4 the kind of ignorance that
involves a whole battery of mechanisms of avoidance and resistance to
know and to learn; the kind of ignorance that is deeply invested in not
knowing. This kind of recalcitrant ignorance has to be distinguished from
the basic or plain kind of ignorance that involves nothing more than the
absence of true belief or the presence of false belief.5 Basic or plain
ignorance tends to be innocuous, for, when our ignorance is nothing
more than the absence of true belief and/or the presence of false belief,
learning is relatively easy (we just unmask false beliefs and inculcate true
ones), and education does not encounter psychological obstacles and
resistances. However, in the case of active (self-protecting) ignorance,
besides epistemic lacunas and distortions, there are also defense mechan-
isms that keep the ignorance in place by making it immune to criticisms
and neutralizing stimulations to learn what conflicts with it.
The following diagram offers a schematic illustration of the distinction
between basic and active ignorance and the defense mechanisms that
protect the latter:

4
See Mills (1998), Sullivan and Tuana (2007) and Medina (2012).
5
What I am calling basic or plain ignorance fits well with the New View of ignorance
articulated and defended by Peels (2013).
 Ignorance and Racial Insensitivity 183

Basic ignorance:
(1) absence of (true) belief
(2) presence of false belief
Active ignorance:
(3) cognitive resistances (e.g., prejudices, conceptual lacunas)
(4) affective resistances (e.g., apathy, interest in not knowing –
“the will not to believe”6)
(5) bodily resistances (e.g., feeling anxious, agitated, red in the
face)
(6) defense mechanisms and strategies (deflecting challenges,
shifting burden of proof, etc.)

3 Racial Insensitivity as Active Meta-Ignorance

The carefully cultivated disinterest in knowing (sometimes even the interest
in not knowing) that underlies racial insensitivity is socially orchestrated and
nurtured by social habits and dynamics, such as not talking about race in
certain contexts or in the presence of “mixed company.” But this closed-
mindedness is rarely explicit and consciously cultivated. Insidious forms of
insensitivity are forms of active ignorance that become invisible to the subject
despite his or her active participation in it. They are deeply entrenched blind
spots that remain unconscious. Racially insensitive people of this sort are not
only numbed to particular racial issues, they are also numbed to their own
numbness, that is, incapable of reacting to it or even of recognizing how they
have become numbed; they are insensitive to their own insensitivity. This is
what I call meta-ignorance, which consists in a pronounced difficulty in
realizing and appreciating the limitations of one’s social sensibility and
horizon of understanding.
Meta-ignorance involves meta-attitudes about one’s first-order proposi-
tional attitudes and cognitive repertoires, such as, for example, believing or
disbelieving that one has certain beliefs/disbeliefs, epistemic abilities/disabil-
ities, epistemic lacunas, etc. Meta-ignorance can often be encountered asso-
ciated with privilege, that is, as resulting from having lived a privileged and
sheltered life in which one does not encounter much epistemic friction7 in one’s
interactions, a life in which one is encouraged to be inattentive to certain
things, to disbelieve certain things, to trust people like oneself and distrust
people who are different, and to develop an inflated sense of one’s capacities
and epistemic contents.
6
For an account of this phenomenon, see Medina (2016).
7
For a full account of “epistemic friction”, see Medina (2012), especially chapter 1.
184 José Medina

According to the epistemic interactionism I have developed and

defended following Wittgenstein and pragmatists such as Jane Addams,
John Dewey and G.H. Mead, epistemic friction occurs in communicative
interactions when different perspectives or standpoints challenge and
contest each other; and this kind of friction is required for becoming
sensitive to the perspectives of those who are different from us and to
the limits of one’s own perspective. Those who live privileged and shel-
tered lives are less likely to encounter friction in their interactions; they are
less likely to run into communicative and epistemic obstacles that leave
their experiences, problems and even their entire lives at a disadvantage;
and as a result of not ever feeling severely constrained as speakers and
subjects of knowledge, privileged subjects tend to be more reluctant to
acknowledge the limitations of the horizon of understanding that they
inhabit; that is, they tend to be numbed to their own numbness, insensitive to
the blind spots that they have inherited and they recirculate in their
epistemic lives.
In some cases, meta-ignorance can be basic or plain and, therefore,
relatively innocuous. These are cases in which the subject simply does not
have meta-attitudes about their first-order epistemic attitudes, or have
incorrect meta-attitudes that she or he is willing to correct. For example,
we can find a case of basic meta-ignorance in a subject who is not aware of
ignoring important cultural facts about a minority group within her com-
munity, but she is willing to acknowledge this epistemic deficit and to take
steps to repair it when the opportunity arises. But there are also cases of
active meta-ignorance. Active meta-ignorance can be characterized as
resistance to epistemic friction – to acknowledge and engage epistemic view-
points that can challenge and contest one’s own. In fact, in active meta-
ignorance we find a double resistance to epistemic friction: the inability to
recognize alternatives (a first-order resistance to friction, a lack of open-
ness to epistemic counterpoints), plus the inability to recognize one’s
inability (a second-order resistance or meta-resistance, a resistance to
identify and acknowledge one’s lack of openness to epistemic counter-
points). This meta-ignorance blocks possible paths for fighting and
repairing epistemic injustices. If one is not even able to recognize one’s
own blind spots – if one is numbed to his/her own numbness – how is this
subject going to be able to embark on a journey to improve his/her
epistemic perspective and sensitivity? This is what I call the problem of
meta-ignorance, which becomes more intractable in cases of active meta-
ignorance in which the subject is not only incapable of recognizing her
epistemic deficits but resistant to do so when prompted by others.
Precisely because of the obstacles and defense mechanisms that operate
at the meta-level, the racially insensitive individual will need external
 Ignorance and Racial Insensitivity 185

help. In the case of the actively meta-insensitive person particular kinds of

epistemic interventions will be needed, for example, those interventions
that penetrate the resistances and the defense mechanisms of the subject
producing cognitive discomfort, which is sometimes expressed in affec-
tive and embodied ways (e.g., taking things too personally, getting angry,
getting red in the face). Sometimes these epistemic interventions may
come in the form of specific challenges raised in particular interpersonal
interactions: interlocutors raising challenges or provocations that make
the subject rethink what she thinks she knows or doesn’t know. But in
other cases, the epistemic interventions may take a more generalized
form. They may relate to changes in social practices and dynamics that
create new social pressures and expectations, and force subjects to explain
themselves when they say certain things or when they take certain things
for granted. An example could be found in recent shifts in practices of
telling jokes in which people are made accountable for what they find
humorous and/or offensive. The generalized and concerted forms of
epistemic interventions of the latter kind are particularly important
given that the hard cases of active meta-ignorance – such as in racial
insensitivity – concern not only particular individuals, but entire groups
and cultures. The actively meta-ignorant group or culture will need the
help of other groups or cultures, or of alternative viewpoints within them.8
But, of course, the more empowered the insensitive (and actively meta-
ignorant) individual, group or culture is, the more difficult it will be for
others to do the proper interventions and to set in motion the process of
transformation and cognitive-affective amelioration.

4 Racial Insensitivity as Involving Both Social Ignorance

and Self-Ignorance
The recalcitrant blind spots of racial insensitivity are both a form of social
ignorance and, at the same time, a form of self-ignorance. It is very
important to appreciate the connection that often exists between our
ignorance about others and our (typically more implicit and harder to
see) ignorance about ourselves. I have argued elsewhere that when an
individual is epistemically irresponsible with respect to others, it is very
often the case that his social ignorance involves self-ignorance: ignorance
about his own relationality with respect to those ignored others, and quite
possibly also ignorance about certain aspects of himself that he is unable

8
Luckily, social groups and cultures are not so homogeneous and monolithic that they
contain no discordant or dissenting voices, or at least their possibility. In my Speaking from
Elsewhere (2006) I have argued that this possibility is always there.
186 José Medina

to recognize, which prevents him from taking responsibility for his per-
spective and how it relates to the perspectives of others. As Charles Mills
(1997, 1998) suggests in his account of “white ignorance,” white subjects
often don’t know what it means to be white; they are ignorant about the
presuppositions and consequences of their own racial identity. More
recently, in Revealing Whiteness, Shannon Sullivan (2006) has offered an
analysis of how privileged white subjects maintain the ignorance of their
own racialization through well-entrenched cognitive habits that hide
themselves: whiteness is often rendered invisible for white subjects and
needs to be revealed.
As the analyses of white ignorance in race theory show, privileged white
subjects tend to lack the motivation and the opportunity to develop
expressive activities and interpretative tools to make sense of their own
social experiences of racialization and to understand how their lives have
been affected by racism and its legacy. And this self-ignorance, this
inability to interpret their social experiences on racial matters, under-
mines their testimonial and hermeneutical sensibilities in their commu-
nicative interaction with others. The phenomenon of the active ignorance
and interpretative impoverishment of the privileged has also been ana-
lyzed by epistemologists of ignorance with respect to gender and
sexuality.9 Feminist and queer theorists have argued that gender and
sexual experiences are particularly opaque to gender and sexual confor-
mists who, not having interrogated their own trajectories in these areas of
social life, become especially ill-equipped to understand their own gender
and sexuality, lacking interpretative tools and strategies specifically
designed to apply to their own case.10 This is why what passes for
obviousness or transparency in relation to masculinity, femininity and
heterosexuality typically hides a lack of awareness and sensitivity to
nuanced and plural gender and sexual meanings. As epistemologists of
ignorance have shown, the epistemic gaps that emerge from structures of
oppression and identity prejudices create bodies of active ignorance for
those subjects whose privileged positions are protected by the blind-spots
and insensitivities in question – racial, sexual or gender insensitivity; and
those bodies of ignorance include their lack of knowledge and their
inabilities to learn not only about racial, sexual and gender others, but
about themselves, their position in the world and the perspective that they
share with their social group. As a long tradition of philosophers from

9
See especially the pioneering work of Nancy Tuana (2004, 2006). See also Sullivan and
Tuana (2007).
10
On this point, see especially Scheman (1997).
 Ignorance and Racial Insensitivity 187

Strawson (1974) to Bilgrami (1998, 2012) and beyond11 have argued, in

order to become a responsible agent, a subject needs to have some mini-
mal self-knowledge; and this thesis can now be taken to entail that
epistemic responsibility requires eradicating the kind of self-ignorance
involved in racial insensitivity. This point will be part of my argumenta-
tion in the next section.

5 Racial Insensitivity and Epistemic Responsibility

Racial insensitivity is the kind of insensitivity that is indicative of moral
and political patterns that go well beyond the individual; it concerns
group behavior, clusters of individuals in relation to other clusters of
individuals (such as families, communities, racial groups and sometimes
even entire regions, nations and cultures). Racial insensitivity is misun-
derstood if it is conceived as a purely individualistic problem, for although
we can say that particular individuals are racially insensitive, the produc-
tion of insensitivity is a collective enterprise in which there are shared
responsibilities: the responsibilities of parents, educators, friends, neigh-
bors, citizens and so on. Accordingly, repairing racial insensitivity and
developing new forms of sensitivity that can make people more epistemi-
cally responsible require cooperative and collective efforts directed at
establishing new patterns of relationality and responsivity. The efforts at
overcoming insensitivity have to be oriented toward action and they have
to be sustained in particular contexts of action and interaction. They
cannot be carried out in a purely spectatorial and detached way.
Developing racial sensitivity is more than acquiring racial knowledge or
familiarity in an impersonal way; it requires an engaged perspective that
makes itself vulnerable to challenges and contestations from other per-
spectives in actual interactions with diverse others.12
11
For a full discussion of this thesis and the philosophical literature on it, see section 4.1 of
Medina (2012).
12
One worry here may be that the task of becoming epistemically responsible can become
an infinite and unattainable task given our epistemic obligations towards indefinitely many
minority groups. This is a serious worry. But I think my contextualized interactionism has
resources to handle it. As I have argued elsewhere, as long as we have strategies that
enable us to prioritize in any given context which obligations we should address first in
order to open ourselves to the epistemic friction of other perspectives, the task of
achieving epistemic responsibility can be embarked upon and we can assess whether we
are faring better or worse in that task even if we accept that it is a neverending task that will
never be completed once and for all, that is, even if at no point can one say: “Ok, I am
done with opening my mind, I have stretched my sensitivity fully and cannot possibly
become more open-minded” (an expression of epistemic arrogance which seems to
perform meta-blindness). For these purposes, I have proposed the Maxim of Eminent
Relevance (Medina 2012, pp. 156–157), according to which in any given context we have
to identify the eminently relevant others with whom we are connected and interact, and
188 José Medina

The account of racial insensitivity as active ignorance that I have

articulated above and elsewhere (Medina 2012, 2016) lends itself to an
expansive notion of epistemic responsibility that I will explore in this section.
In the first place, it is important to notice that the kind of responsibility
that attaches to various forms of racial insensitivity involves a broader
notion than the notion of doxastic responsibility. Although doxastic
responsibility13 is contained in my notion of epistemic responsibility, the
responsibility that we need to take for our epistemic positions and atti-
tudes does not only concern the beliefs that we hold, but also the absence
of belief and the epistemic mechanisms of avoidance and resistance that
protect doxastic lacunas and distortions (or patterns of false beliefs). And
in the second place, my notion of epistemic responsibility is broader than
standard notions of epistemic culpability or blameworthiness.14 In some
cases the issue can be whether racial insensitivity springs from culpable
ignorance or blameworthy beliefs, but in other cases, even when we find
doxastic attitudes or lacunas that can be said to be free from blame,15
there still remain normative issues concerning complicity and shared
responsibility for maintaining a body of ignorance in place. In other
words, even when we say of a racially insensitive subject that she or he
could not have known better and therefore she or he is not to blame for his or
her ignorance (any more than her or his peers, mentors, etc.), we may still
hold her or him partially responsible for not taking any steps to displace
the ignorance in question. Although broader and more diffused, my
expansive notion of epistemic responsibility (unlike standard notions of
culpability and blameworthiness) is particularly adept at handling cases of
widespread forms of harmful ignorance that no single individual (or
identifiable cluster of individuals) is solely responsible for and culpable
of. Let me illustrate what is most distinctive of my expansive notion of

their perspectives should be prioritized as we start expanding our epistemic sensitivities

and making them progressively more open to multiple others.
13
The debates around doxastic responsibility are not unrelated to my notion of epistemic
responsibility, but they only concern one aspect of it: the permissibility, praiseworthiness
or blameworthiness of the beliefs that we hold. See Booth and Peels (2010, 2014).
14
For a full account of doxastic blameworthiness, see Nottelmann (2007).
15
I fully agree with Peels and Booth’s (2014) minimalist position in the literature on
doxastic deontologism, which they call DDB. According to DDB, responsible belief is
blameless belief and not praiseworthy belief – a position they call DDP and which is more
epistemically demanding. For the cases they examine, being free from blame in the beliefs
one holds is sufficient to qualify as epistemically responsible and praiseworthiness is not
required. However, as I argue below, in cases of collective racial insensitivity, there is
a form of widely shared ignorance in which individuals participate without any fault of
their own (without choosing to ignore or being aware of their ignorance), and, none-
theless, they fall short of being epistemically responsible given their complacency and
complicity in maintaining doxastic lacunas and systematically distorted doxastic
attitudes.
 Ignorance and Racial Insensitivity 189

epistemic responsibility by summarizing a real-life example that I have

analyzed at greater length elsewhere.16
As reported by the local media, on Saturday, October 8, 2005, at the
Vanderbilt University campus, after a fraternity party in which a pig had
been roasted and eaten, an intoxicated frat boy walked across the street
with the pig’s head and left it at the doorsteps of the Ben Schulman Center
for Jewish Life. The following morning the Jewish community on campus
was outraged. The incident happened during the Jewish High Holy Days
that begin with Rosh Hashanah and end with Yom Kippur, and many
thought that “someone was sending Vanderbilt’s Jewish community
a chilling message during the holiest days of the year.”17 The student
who had dropped the pig’s head at the doorsteps of the Schulman Center
came forward and apologized, but he emphasized that he did not know
that the building in question was a Jewish Cultural Center and he also did
not know that pig’s parts had been used to stigmatize Jewish people and
attack them. What is interesting for our purposes is his exculpatory move
of appealing to ignorance in order to relieve himself of responsibility for
what had been taken to be an anti-Semitic act.18
If we accept the pig-head dropper’s account, the issue of epistemic
responsibility concerning ignorance and racial insensitivity that arises in
this case is a peculiar one. In the first place, if we believe that the student
did not know that the building in question was a Jewish Cultural Center
and that pig parts had been used to denigrate Jews, there is no harmful
false belief we can blame him for, but only a doxastic lacuna. And, in
the second place, the ignorance in question could reasonably be said not
to be culpable, for the student is free from blame if he has in fact not been
given opportunities to acquire the beliefs that would have made him
aware that there was an anti-Semitic layer to his act. This ignorance
does not seem to result from any specific irresponsible act that the subject

16
See Medina (2012, pp. 135–145). The next Section is adapted from that section of my
book.
17
Quote from p. 1 of www.nashvillescene.com/Stories/News/2005/10/20/Pig_Heads_at_
Vanderbilt
18
It is worth noting that the university administrators also appealed to ignorance in order to
minimize the significance of the incident. The Nashville Scene reported: “‘It was a bad
decision, a very bad decision,’ says Kristin Torrey, the school’s director of Greek Life,
who believes the student genuinely didn’t understand the religious overtones of his
action . . . the University has taken the position that its student may be dumb, but he’s
no bigot. ‘This incident happened out of stupidity and shockingly bad taste,’ says Michael
Schoenfeld, Vice Chancellor for Public Affairs. ‘But coming out of it is an opportunity to
advance understanding about how the university is diversifying, and about other tradi-
tions and cultures and religions and beliefs that may not necessarily be part of the
traditional mainstream at the university’” (pp. 2–3 in www.nashvillescene.com/Stories/
News/2005/10/20/Pig_Heads_at_Vanderbilt).
190 José Medina

had previously carried out and can be blamed for. It is not clear that the
student could have known better. And yet, there is still an important issue of
epistemic responsibility here that goes beyond the particular student who
desecrated the Jewish Cultural Center. The epistemic responsibility in
question here is not reducible to blameworthy beliefs or culpable
ignorance.19 Rather, it concerns the issue of taking collective responsibility
for the doxastic lacuna that left the student in the dark as to the symbolic
significance of the pig’s head and the cultural significance of the building
he vandalized: how can the community arrange its epistemic practices so
that they make sure that students will know better and will not maintain
such harmful bodies of ignorance concerning the cultural minority groups
with the university community? In an important sense, we could not
reasonably expect the pig-head dropper, considered in isolation, to have
known any better, and he is not more responsible for his ignorance about
Jewish history and anti-Semitic symbolism than his parents, his teachers,
his college peers, his fraternity friends and so on. So, aren’t we all
collectively responsible for sustaining an epistemic climate in which cer-
tain symbols and backgrounds (e.g., those that relate to Christianity) are
made familiar to everybody and expected to be known by all, whereas the
familiarity with other symbols and backgrounds (e.g., those that relate to
Judaism) is left to epistemic luck?
Cases of inherited ignorance raise a particularly challenging problem for
epistemic responsibility. On the one hand, we do not want to blame
individuals for a body of ignorance they have inherited from their social
milieu, without their choosing to do so and without doing anything in
particular to partake in the ignorance. But, on the other hand, we do not
want to dissolve the issue of responsibility entirely without acknowledging
the epistemic wrongs involved in it, and without issuing normative
demands that can correct the situation and result in epistemic ameliora-
tion. The epistemic responsibility that attaches to cases of collective and
inherited ignorance is a more diffused and expansive notion of responsi-
bility than the standard notion of culpable ignorance, one that is not
confined to the cognitive repertoire and functioning of particular indivi-
duals considered in isolation. But it is nonetheless a notion of responsibility
whose scope can be clearly identified (in terms of particular clusters of
subjects or communities), and whose demands can be concretized (in
terms of particular activities and dynamics that can be demanded of

19
As Rik Peels (2011b) puts it, a subject could be deemed culpably ignorant if “he
performed a culpable action [or omission] in the past which resulted in that culpable
ignorance” (2011b, p. 576). This standard notion of epistemic culpability is heavily
individualized and, as I will argue below, unfit to address the shared and diffused form
of epistemic responsibility involved in cases of collective insensitivity.
 Ignorance and Racial Insensitivity 191

particular groups of people). In the case of widespread racial insensitivity,

we have to take collective responsibility for the mechanisms of transmission
and inheritance of the epistemic deficits in question, that is, for the collec-
tive actions and omissions through which the ignorance and epistemic
dysfunctions are formed, propagated and passed on from generation to
generation. As illustrated by the pig-head incident, it is problematic to
assume that a subject can simply appeal to inherited ignorance and thus
shake off all responsibility for his action. For, even if it exculpates the
individual to some degree, it does not dissolve the issue of responsibility
altogether, but brings it to another level: the level of shared and collective
responsibility at which the relevant question is not simply whether or not
one is responsible for the ignorance in question, but rather, in what ways
and through which actions and omissions we partake in the ignorance and
share responsibility for it. By calling attention to the normative significance
of the alleged ignorance of the pig-head dropper, I am interested not only in
highlighting how the individual student falls short of achieving full episte-
mic responsibility in his thinking and action, but also and more importantly
in underscoring the responsibilities of the communities that shaped this
individual and made his ignorance possible, and the communities that
continue to influence his development and can do something to repair his
ignorance – including (and especially) the Vanderbilt University commu-
nity in which the incident took place (and of which I am also a member).
Although I want to underscore the collective side of epistemic respon-
sibility in cases of widespread insensitivity and shared ignorance, I also
want to call attention to the ways in which the individual and the collec-
tive levels of responsibility are inextricably intertwined, rather than simply
shifting from one to the other. Indeed, as I have argued elsewhere, “one
cannot hide oneself completely in the collective ignorance in which one
partakes, at least not for long. The collective ignorance may not be of
one’s choosing, but one cannot inhabit it comfortably and without mak-
ing any effort to combat it (even when opportunities to do so present
themselves), and legitimately use this inherited ignorance to excuse one’s
actions. Even if indirectly, by omission and inaction, one becomes an
active participant in one’s own ignorance if one lets it sit and grow, paying
no attention to its roots and its ramifications” (Medina 2012, p. 140).
One’s inattention to the ignorance one partakes in can become complicity
and active participation. One’s participation in the collective bodies of
ignorance one has inherited becomes active when and because one acts
on it and fails to act against it. As Shannon Sullivan (2006) has shown
convincingly in her transactional account of white habits in the United
States, racial ignorance recruits agents to act on it habitually in their daily
lives and it gets transmitted across generations through a complex array of
192 José Medina

(largely unconscious) habits and dispositions that we have to take respon-

sibility for, both individually and collectively.
The formation and transmission of epistemic sensitivities and insen-
sitivities with respect to a racial, ethnic or cultural group requires some
knowledge of that group, that is, minimal familiarity with their history,
their experiences and struggles, their symbols and aspirations, and so
on. When the experiences and perspectives of one particular group are
highly visible within a culture at the expense of the visibility of those of
other groups, this epistemic imbalance creates obstacles for developing
adequate epistemic sensibilities with respect to minority and/or mar-
ginalized groups. For example, in a context in which the symbols and
meaning of WASP Americans become mainstream and acquire default
status, the symbols and meanings of African Americans or of Jewish
Americans are less likely to be properly understood and to be given
adequate uptake. Subjects easily develop an epistemic sensitivity with
respect to the perspectives of those who live and speak like them, but it
is harder to become equally sensitive and open-minded to viewpoints,
lifestyles and forms of expressions that are unfamiliar or, worse yet, in
conflict with one’s perspective. Fricker (2007) and other scholars of
epistemic injustice have called attention to the hermeneutical obstacles
against the adequate understanding and interpretation of marginalized
groups and perspectives, and to the testimonial obstacles against giving
them adequate levels of trust and credibility. These scholars also
emphasize that the work toward epistemic justice involves removing
those obstacles and expanding our hermeneutical and testimonial sen-
sitivities. Becoming epistemically sensitive to other people who are very
different from us – becoming capable of understanding them, inter-
preting them, and trusting them adequately – involves more than
knowing facts about them; it involves knowing how to listen to them
in the light of those facts, knowing how to respond to their concerns and
aspirations, and learning how to take into account their viewpoints.
In other words, it involves the acquisition of an epistemic sensitivity
that can only be developed through communicative interactions and
engagements in which there is epistemic friction (i.e., challenge and
mutual contestation), the kind of friction that can make each perspec-
tive aware of itself and of its own limitations, and open and vulnerable
to other perspectives. This kind of epistemic sensitivity or openness is
not developed simply by acquiring factual knowledge about diversity.
It does not consist in knowledge that can remain purely detached or
external to us; rather, it is a practical knowledge – a know-how – that
deeply affects our sense of who we are in the social world and our
 Ignorance and Racial Insensitivity 193

understanding of our relationality with others. This is what my episte-

mic interactionism emphasizes.
By epistemic interaction with heterogeneous others I do not mean
simply being exposed to information about them, or even mere exposure
to the presence and the voices of those who are different from us, because
these exposures typically do very little to change the kinds of insensitivity
and meta-insensitivity underlying social injustices. Becoming epistemically
responsible involves experiencing epistemic friction in actual interactions
with heterogeneous others; it involves learning how to acknowledge the
limitations of our standpoint and the distance and differences between
our standpoint and alternative ones. As suggested by Stanley Cavell
(1969), acknowledging is a hybrid notion: it is an epistemic and ethical
notion that involves minimally knowing or recognizing the existence of
something and being able to avow it to oneself and to others, thus
accepting the normative implications of that knowledge as it positions
ourselves with respect to others. In some cases subjects may know some-
thing and they may not yet be ready to acknowledge it to themselves or to
others (e.g., that one has an addiction, that one is depressed, or that one
has internalized some racist prejudices). Acknowledging has the role of
normatively positioning oneself with respect to the knowledge in question
and taking responsibility for it by accepting the normative consequences
that it has in one’s life and in one’s relations to others. There are negative
forms of acknowledgment in which what one recognizes and avows is not
an epistemic content, but an epistemic absence, that is, the limits of one’s
knowledge: one’s ignorance and epistemic limitations. Recognizing and
avowing an epistemic limitation – acknowledging that one does not know
what something means, for example – can be a great epistemic and ethical
achievement. Acknowledging that a meaning or an experience lies outside
one’s horizon of understanding can be crucial for learning how to position
oneself with respect to that meaning or experience and for learning how to
interact with those who have knowledge of that meaning or experience.
Acknowledgment marks an epistemic and ethical relation that, through
cognitive and affective means, we establish with others (with their experi-
ences, problems, aspirations, values, etc.). Through the epistemic and
ethical relation of acknowledgment, we can learn to respect differences we
are not familiar with or cannot claim to understand, opening up the
possibility of epistemic and ethical growth and the formation of new
forms of sensitivity. Of course acknowledging one’s epistemic limitations
and the distance between one’s perspective and that of others, by itself,
will not repair insensitivity and the epistemic harms it can produce; but
acknowledgment of this kind can be a crucial first step toward instilling
epistemic virtues such as humility and open-mindedness and toward
194 José Medina

eradicating epistemic vices such as arrogance and closed-mindedness.20

Acknowledgment is a central notion for the ethics of knowing and ignor-
ing, for acknowledging is a crucial mechanism for establishing and
accepting epistemic responsibility: we have to take individual responsibility
for what we know and don’t know as well as collective responsibility for the
social production of knowledge and ignorance.21 In any given case, it is
a contextual issue whether the focus should be on individual or on
collective responsibility, that is, on the specific epistemic actions/omis-
sions that particular individuals are responsible for or on the widespread
patterns of action/omission and interaction that communities maintain
and individuals simply participate in.22
Developing a sense of epistemic responsibility in relation to the epis-
temic harms produced by racism (such as undeserved lack of credibility,
differential access to communicative and interpretative resources, etc.),
involves becoming more sensitive and responsive to racial epistemic
injustices as they appear in our daily activities, that is, more sensitive
and responsive to the testimonial and hermeneutical dysfunctions that
surface in our epistemic practices. Trying to fight racial epistemic injus-
tices is a very complex and difficult task; but, minimally, it should involve
being attentive to racially motivated epistemic mistreatments, acknowl-
edging dysfunctional epistemic dynamics that unfairly harm certain racial
groups, and resisting complicity with undeserved epistemic privileges and
disadvantages apportioned to members of different racial groups. Besides
being action-based and action-oriented, my notion of epistemic respon-
sibility (and of the obligation to resist contained in it) has two other key
features. Our responsibility as epistemic agents and communicators is (1)
situated and polyphonic – to be determined contextually, practice by

20
In The Epistemology of Resistance (2012) I have developed an account of these epistemic
virtues and vices as they appear under conditions of oppression and epistemic injustice.
See especially Chapter 1.
21
For a full discussion of the relation between individual and collective epistemic respon-
sibility, see Medina (2012) and Code (1987, 2014).
22
Although the individual and collective levels of epistemic responsibility typically go
together, there are cases in which they can be decoupled and even go in separate ways.
There are cases in which individuals take responsibility for racial insensitivity in their own
actions and omissions, while remaining unable to recognize and acknowledge their
involvement in widespread forms of racial insensitivity. For example, there are white
Americans who have become racially sensitive to the perspectives of African Americans in
their personal interactions, but nonetheless they still support policies and social arrange-
ments that stifle the participation of African Americans in public life (for example, the
recently proposed new requirements for voter registration and new boundaries of election
districts in the United States). There are also cases in which individuals are willing to
accept shared responsibility for widespread racial insensitivity and yet they remain unable
to recognize how such insensitivity surfaces in their personal actions and omissions in
daily life.
 Ignorance and Racial Insensitivity 195

practice, given one’s positions and relations within those practices; and
(2) it is essentially shared – it always requires cooperation with others and
collective efforts.
We all have a prima facie obligation to resist racial oppression and the
epistemic injustices associated with it whatever social locations we hap-
pen to occupy and whatever racial privileges or disadvantages (if any) we
happen to have. But of course what counts as our responsibility in any
given context will depend on how we are positioned in the practices
relevant in that situation and how we are related to others within those
practices. We are all responsible to resist epistemic injustices, but we do
not bear the same responsibilities. Piecemeal, practice by practice, and in
all corners of our life, we need to carry out the contextualist task of
figuring out what we can and should do given our position in the social
world and our relationality with others across practices. I claim that there
are special responsibilities that the racially privileged have in the fight against
racial epistemic injustices. While oppressed subjects are typically episte-
mically better equipped to resist, their agency is constrained and they do
not always have opportunities to make their resistance effective; and even
when they can resist effectively, there may be extremely high costs for
them in that resistance,23 and it would be unfair to demand it from them
as an obligation. On the other hand, privileged subjects, although they
may have more power and agency to speak up against racial epistemic
injustices and to change racial dynamics at a lesser cost for them, they
often are epistemically ill-equipped to fight epistemic injustices and they
have little motivation to do so (and often strong pressures not to make
that fight their own). This suggests that different groups must cooperate
and jointly discharge the shared obligation to remove racial ignorance and
insensitivity – oppressed groups providing the requisite knowledge,
insight and perspective while privileged groups providing support and
yielding the required power and agency.
If becoming epistemically responsible was an individual task, then it
would be utterly impossible for corrupted epistemic subjectivities to fulfill
epistemic demands and achieve responsibility: actively meta-ignorant
individuals by themselves are unable to detect their blind-spots and
recognize their insensitivities, and therefore they are incapable of becom-
ing epistemically responsible agents. But becoming epistemically sensi-
tive is not only or primarily an individual responsibility – it is not even
something that the individual can always accomplish fully by herself, in
23
The result for these subjects can be being perceived as non-compliant, as unruly, and this
can result in more exclusion and marginalization from the epistemic activities in question.
Think for example of classroom dynamics and how a student trying to change the
argumentative dynamics can be perceived by the other students and by the teachers.
196 José Medina

isolation. Rather, it is a shared responsibility that can only be discharged

cooperatively and collectively. In the first place, it can only be discharged
cooperatively because individuals have blind-spots and insensitivities that
they cannot overcome (sometime they cannot even detect) if left to their
own devices; they need the critical interventions, provocations, and chal-
lenges (the friction) of others. And, in the second place, since insidious
forms of epistemic insensitivity involve widely shared forms of ignorance
and inattention, the epistemic responsibility in question can only be
discharged collectively because it is the responsibility of groups, publics,
and institutions, and of individuals qua members of those social groups,
publics, and institutions.
Sometimes even well intentioned racially privileged subjects, willing to
resist epistemic injustices, are not yet capable of doing so by themselves.
In my view, this doesn’t mean that their obligations disappear—they are
not off the hook. Rather, what this means is that they will need the help and
guidance (as well as the provocations and pressures) of others. And besides
breaking their racial insensitivity with the help of others, privileged sub-
jects will also need help to find opportunities to resist and develop new
forms of sensitivity. Typically this means that the social world around
them needs to change: the institutional settings and social practices that
have supported their racial insensitivity need to undergo structural
changes, so that new social conditions can block insensitive attitudes,
produce costs and negative consequences for displays of insensitivity, and
create incentives for the interruption of insensitivity and the formation of
new forms of social sensitivity. Some of this can be done, for example, in
educational contexts and educational practices that shape our epistemic
sensibilities by instilling epistemic habits and by establishing dynamics of
epistemic interaction with others. To this area I now turn.

6 Becoming Epistemically Responsible: Toward an Ethics

and Pedagogy of Discomfort
I have argued that in order to be able to expand and meliorate our social
sensibilities, we need to start by exposing ourselves and making ourselves
vulnerable, by acknowledging our epistemic limitations and taking respon-
sibility for epistemic injustices, opening up our perspective to processes of
critical scrutiny and resistance. As I have argued elsewhere (2012), epis-
temically responsible social interaction requires resistance, and resistance
begins at home, that is, in the most intimate aspects of our cognitive-
affective functioning. Resistance has to begin within ourselves and in the
activities in which we feel at home. This is why I contend that for our
epistemic sensibilities to be truly open and responsive to differences, they
 Ignorance and Racial Insensitivity 197

must be self-questioning, rather than being defensive and self-protecting.

Self-questioning sensibilities can only be established when subjects make
themselves vulnerable to challenges and become exposed (either by
choice or by social design) to processes of self-questioning.24 Looking at
ourselves with fresh eyes through processes of self-questioning affords us
opportunities to interrogate what we find in the most intimate corners of
our perspective, and to recognize its limitations and the possibilities of
correction and improvement. In very different ways, Queer Theory,
Feminist Standpoint Theory and Critical Race Theory teach us the
importance of unmasking and undoing the processes of social construc-
tion of our perspective, of interrupting the flow of familiarity and obvious-
ness, making the familiar unfamiliar and the obvious bizarre. And this
critical exercise should not be thought of simply as the quaint activity of
some peculiar activists and intellectuals, but rather, as a crucial part of the
growth and development of critical subjects of knowledge, of subjects
who learn how to resist their cognitive-affective limitations and to
improve their sensibilities and capacities.
We all have a prima facie obligation to interrogate received attitudes
and habits, that is, to open ourselves to processes of self-questioning. If we
fail this obligation, the failure of other epistemic responsibilities will ensue
and possibilities of critique and resistance will be thwarted. Cultivating
this openness to being challenged about one’s own perspective involves
experiences of epistemic discomfort in which we feel disoriented, losing
our epistemic bearings as it were, not being able to rely on epistemic
norms and presuppositions that we have taken for granted, and therefore
being open to other perspectives in new ways. Undergoing processes of
self-questioning of this kind involves experiencing the kind of epistemic
discomfort that makes us re-evaluate our own perspective vis-à-vis alter-
native ones; and these experiences constitute invaluable learning oppor-
tunities, occasions for epistemic and ethical growth, for remaking our
epistemic habits and attitudes and reassessing the norms and presupposi-
tions of our epistemic practices. The ethics and pedagogy of discomfort that
I propose (2014) focuses on experiences of epistemic friction that make us

24
These processes of self-questioning can sometimes lead to processes of defamiliarization
and self-estrangement (Medina 2014, 2016). Challenging one’s own perspective to the
point that it becomes unfamiliar – to the point that one becomes a stranger to oneself – is
something that may or may not happen depending on the challenges, interrogations and
provocations available to oneself, and on how one processes them. All we can demand of
subjects is that they remain open to these processes, and we can hold them accountable
when they are not. We can also demand of communities (e.g., in relation to educational
policies) that they make these processes readily available to all subjects, and that they try
new techniques and strategies of interrogation and provocation if the ones in use prove to
be unsuccessful.
198 José Medina

re-evaluate the normative structure of our familiar practices and

dynamics. The kind of epistemic friction that triggers processes of self-
questioning is required for the formation of open-minded subjects and
open communities.
Building on my view, Lorraine Code (2014) has argued recently that an
“open society” and its members have epistemic responsibilities to live up
to: in particular, the responsibility to make available “accurate knowl-
edge/information” about all the problems and injustices that occur in
every corner of the social fabric, so that the plight of every member of
society becomes visible and no one is forced to live an invisible social
existence. Code rightly adds that the responsibility in question is primar-
ily “pedagogical: educators, investigative journalists, and public intellec-
tuals (among others) have a presumptive duty to know, address,
communicate, and debate these issues in their complexity; and respon-
sible citizens have some obligation to learn how to evaluate them, nega-
tively or positively. Yet assuming such responsibilities is, again, a fraught,
often frustrating task, and questions about where to confer trust are not
easily addressed” (Code 2014, pp. 672–673). Indeed lack of trust is one of
the characteristic features of the kinds of social insensitivity that exclude
and marginalize subjects; and building trust to repair those forms of
insensitivity and the injustices they keep in place is a difficult pedagogical
challenge. Should students trust educators? Should educators trust the
educational system within which they work? As Code emphasizes, “ped-
agogical responsibilities are multiply challenging”: to begin with, tea-
chers, journalists and public intellectuals who should educate the public
and promote less exclusionary forms of social sensibilities have them-
selves been socialized breathing the prejudices, biases and insensitivities
in question and should not be presumed to be free from them (Code
2014, p. 673).
The multiple challenges and social constraints of our pedagogical
responsibilities makes it clear that the expansion of social sensibilities
“could not be a purely individual effort” (Code 2014, p. 675), but it is
a challenge for the whole of society that requires the concerted efforts of
all communities and their members. It is a challenge that involves com-
munity building and the making and remaking of social networks.25 It is
crucial to foster solidarity with those who are marginalized in our epis-
temic practices and suffer epistemic injustices, being disproportionately
and unfairly disbelieved and misunderstood because of lack of trust, lack
of credibility and lack of access to epistemic and interpretative resources.
It is crucial that subjects resist their complicity with unfair and

25
See especially section 5.3 of Medina (2012, pp. 225–249).
 Ignorance and Racial Insensitivity 199

dysfunctional epistemic dynamics, and that they interrupt such dynamics

and make those who participate in them uncomfortable by questioning
habitual and familiar assumptions that skew communication and bias
assessments of intelligibility and credibility. In this sense, I have argued
(Medina forthcoming) that the micro-aggressions through which the
epistemic status and agency of racial minorities are routinely undermined
in epistemic activities should be countered with micro-practices of resistance
in which participants in epistemic practices (whatever their racial iden-
tity) express epistemic solidarity with those unfairly treated.
Examples of micro-practices of resistance against epistemic micro-
aggressions are the following: calling out and making explicit
a differential treatment – even if that means stopping the interaction
and/or making the participants uncomfortable; responding to a stare,
gesture or insinuation that calls into question someone’s competence
with a stare, gesture or insinuation that calls into question the aggressor’s
authority or ability to call into question other people’s competence; and
so on. Micro-resistance does not need to come from the person suffering
from the epistemic micro-aggression, but it can be produced effectively
(sometimes even more effectively) by others involved in the interaction
even though they were not targeted (including bystanders and even
eavesdroppers). For example, imagine passengers on a bus in a major
US city overhearing someone lecturing to some kids about blacks or
Hispanics being oversensitive when being questioned by the police;
instead of extricating oneself from the uncomfortable situation, every-
body present (and not just the kids involved) can and should feel respon-
sible to intervene and not let it stand, that is, not let the micro-aggression
go unquestioned and the micro-aggressor get away with it. Sometimes
racial micro-aggressions take the form of micro-invalidations such as
“you feel that way just because you are black,” “you are making it
a racial issue,” “you are overreacting and making something out of
nothing.” These micro-aggressions vitiate communicative dynamics by
biasing hearer’s receptivity to some racial perspectives and invalidating
the credibility or trust deserved by testimonial contributions from those
perspectives. Unfair micro-invalidations can be countered by anyone
present at the interaction with a who-are-you-to-invalidate response, or
with a way of deflecting or shifting the unfair argumentative burden being
posed. Micro-invalidations can also be countered through resistant
micro-validations, or alternative ways of validating and supporting sub-
jects who, in the given context, are likely to face challenges and obstacles
in their status of participants with full or equal epistemic agency. These
micro-practices of resistance that produce friction are at the core of my
ethics and pedagogy of discomfort, which recommends epistemic
200 José Medina

interventions and subversions that produce epistemic discomfort in dys-

functional dynamics in order to motivate agents to change their epistemic
attitudes and habits and to experiment with alternative dynamics.26
Epistemic responsibility is intertwined with ethical and pedagogical
responsibilities. Under conditions of epistemic injustice, when there are
dysfunctional epistemic dynamics, our ethical and pedagogical responsi-
bilities call for critical interventions that turn experiences of epistemic
discomfort into learning opportunities for individuals and communities
to become more sensitive to patterns of unfair epistemic treatment and
more active in producing epistemic solidarity and facilitating fair epistemic
cooperation. The ethics and pedagogy of discomfort that I propose targets
our complacency and complicity with epistemic injustices and demands
that we confront and live up to our epistemic responsibilities through
sustained pedagogical practices that disrupt epistemic habits and attitudes
that protect unfair privileges and disadvantages. “Sometimes people need
to be made uncomfortable so that they wake up from their numbness,
sometimes their familiar spaces and comfortable activities need to be
interrupted so that they become aware of their complicity and their motiva-
tional obstacles to pay attention to an injustice and to join the fight against
it. Part of what needs to happen to counter the protective mechanisms of
privilege is to call attention to the unfair consequences of keeping areas
of epistemic neglect unchecked and thus to make painfully visible the price
of epistemic comfort under conditions of oppression, so that people cannot
avoid the realization that the epistemic comfort of some comes at the cost of
the epistemic discomfort of others” (Medina 2014, p. 66), and that leaving
the epistemic appraisals of some uncontested means that the epistemic
appraisals of others are always in question. All of us (but those who occupy
positions of privilege especially) need to get out of our comfort zones and
familiar spaces and to open ourselves to the epistemic friction that can
make perspectives vulnerable and accountable to each other. Nothing short
of this complicated and concerted effort at epistemic resistance will be
effective in eradicating patterns of racial ignorance, uprooting racial insen-
sitivity and repairing the harms it produces.

7 Epilogue
I want to conclude my discussion of ignorance and racial insensitivity with
a caveat and a closing reflection on ongoing research and future work in

26
This section has been adapted from my essay “Epistemic Injustice and Epistemologies of
Ignorance” (Medina 2016), where the reader can find a fuller discussion of micro-
aggressions and micro-resistance.
 Ignorance and Racial Insensitivity 201

this area. First, the caveat. This paper has focused on racial ignorance and
insensitivity in privileged groups (such as white people in the United
States), but although privilege can be associated with particular dynamics
of racial ignorance and insensitivity, there is also racial ignorance and
insensitivity within and across oppressed groups, and not only against
privileged groups (e.g., anti-white biases among peoples of color in the
United States), but also against other oppressed groups and even against
members of their own group (e.g., the phenomenon of racial self-hatred).
A full account of racial insensitivity should include an analysis of how the
defense mechanisms of racial ignorance operate in differently situated
subjects and groups.
In closing, I want to acknowledge recently developed areas of empirical
research that have sparked theoretical discussions of racial ignorance and
have turned the epistemology of race into an area of vibrant interdisci-
plinary research. In particular, I want to highlight the significance and
fruitfulness of two areas of interdisciplinary research: implicit bias
(Brownstein and Saul 2016) and micro-aggressions (Sue 2010). On the
one hand, the paradigm of implicit bias and the empirical evidence
gathered within it illustrate well my conception of racial ignorance as
a form of self-hiding and unconscious insensitivity that is hard to displace
because it consists in well-entrenched habits of mind. On the other hand,
micro-aggressions are the perfect arena for the detailed analysis of how
racial insensitivity operates in situated interpersonal dynamics. I am con-
fident that these and other venues of empirical research will bear many
empirical and theoretical fruits for the discussion of racial ignorance.
At the same time, I also want to point out that there are cognitivist and
individualist biases in these areas of research that obscure (or at least
minimize) the affective and collective aspects of racial ignorance. While
a lot of research has been done on the cognitive mechanisms underlying
racial biases, the emotions and affective resistances that protect racial
ignorance have received little attention. And while a lot has been written
recently about the psychological mechanisms within the individual that
keep racial biases in place, there has been little discussion of racial
insensitivity as a collective pattern of shared ignorance, and about the
structural and institutional elements that are responsible for the propaga-
tion and perpetuation of racial ignorance at the collective level. I hope that
the insights and suggestions contained in this essay contribute to expand
the research on racial ignorance along those lines.