University of Southern California Dissertations and Theses

Beliefs that wrong

(USC Thesis Other)

Beliefs that wrong

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Beliefs That Wrong
by
Rima Basu
A dissertation submitted in partial fulfillment
of the requirements for the degree of Doctor of Philosophy (Philosophy)
to the faculty of the USC Graduate School at the University of Southern California
August 2018
Doctoral Committee:
Ralph Wedgwood, University of Southern California (Chair)
Mark Schroeder, University of Southern California
Stephen Finlay, University of Southern California
Kenny Easwaran, Texas A&M University
Aimee Bender, University of Southern California (External)
c
Rima Basu 2018
All Rights Reserved
I dedicate this dissertation to my parents. I wish I could’ve been the daughter you
deserved. To my wife, you are more than I deserve. And to my brother, you owe me.
i
ACKNOWLEDGEMENTS
I’ve failed, a lot. To say I did horribly on the GRE would be an understatement. The
first time I applied to graduate school, I was rejected by every school. An end-of-year
report after my first year at USC reads: “[at] this point in her career, it is not yet clear
to me whether Rima is going to be able to be a successful student in our program.” In
my second year, I improved, but an end-of-year report reminds me that I would “need to
continue on a steep curve of improvement.” After I finally started a dissertation project,
I ended up scrapping it and starting again on a new topic, and at some point I stopped
counting rejections.
We don’t often talk about our failures, and I know what you’re thinking, an acknowl-
edgements section seems an odd place to bring them up. But, when we don’t acknowledge
our failures, it can be easy to forget that these stories of failure and rejection are not unique.
Everyone has these stories. And if not for the unwavering support of everyone who has
guided me, supported me, and most importantly, believed in me, any one of these failures
might have derailed me. I have been surrounded by more love and support than I deserve,
and I have a lot of people to thank.
I want to start by thanking my professors at the University of Toronto Mississauga
(UTM). When I chose my undergraduate institution, I didn’t put much thought into it.
UTM was a 10-minute drive from my parents’ house; it was convenient. Little did I know
that that undergraduate institution had one of the best collections of philosophy faculty
in the world. I owe everyone there a lot. Amy Mullin was chair at the time and respon-
sible in large part for the great undergraduate experience I had. Likewise, Jennifer Nagel,
Marleen Rozemond, Bernard Katz, Jonathan Weisberg, Mohan Matthen, Diana Raffman,
and Jacqueline Brunning all played a large role. In particular, I must say a special thanks
to Sergio Tenenbaum, Paul Franks, Phil Clark, Gurpreet Rattan, and Andrew Sepielli.
Not only did they write my letters for grad school, but they are also responsible in large
part for the kind of philosopher I am today, as well as the kind of philosopher I hope to
be.
I remember losing hours in Sergio’s office talking about material we’d covered in class
(and material we hadn’t). Any Kantian sympathies that emerge in my work are because
ii
of seeds that Sergio planted. I’m also grateful to both Sergio and Phil for being at my
very first SLACRR and introducing this petrified 2nd-year grad student to famous people
whose names she’d only ever seen on paper. Phil is also responsible for my love and
admiration of Philippa Foot (whose portrait, painted by Renee Bolinger, hangs above my
desk). Paul’s mentorship and patience while I was an undergraduate fellow at the Jackman
Humanities Institute is a model I can only hope to emulate one day. The paper I wrote
for Andrew’s class (after much revision and input from everyone at UTM) got me into
graduate school. And Gurpreet was there for me with words of encouragement when I
was at my lowest. Gurpreet convinced me that I should take that extra year Toronto was
offering in the MA program and just get better and prove that I could do work at the
graduate level. Second time applying to grad school things did go better, and I’m forever
indebted to everyone at UTM and the University of Toronto generally for all the support
and guidance they offered.
When I got to USC I entered firm in the knowledge of three things. First, I was going
to work at the intersection of philosophy of language and metaethics. Second, I resolved
to never write about reasons. Third, Los Angeles is the best city in the world. Any
interest I had to continue working in philosophy of language evaporated after our first-
year proseminar. And although I ended up writing about reasons (for a bit), my love for
Los Angeles has never wavered. So, one outta three ain’t bad.
There are many people at USC I have to thank. First, there is my cohort—Erik En-
carnacion, Caleb Perl, and Steve Bero. In addition, Beth Snyder might as well have been
part of our cohort. Beth’s knack for storytelling is really quite something. If you ever
have the opportunity to meet her, you should ask her about that time she got a haircut.
Erik, Caleb, and Steve made me feel welcome. At first I was intimidated by them, but
then Caleb got lost on the way back to his car, Erik confessed his fears, and Beth dished
all the dirt on Steve. I’m grateful that they were my cohort and consider them the older
siblings I never had. To this day I also remain jealous of Steve’s salt and pepper hair.
In addition to my cohort, there was a wonderful group of older graduate students
who immediately made me feel welcome and part of a community. I had only just ar-
rived in Los Angeles when I was invited to come out for dinner and Shakespeare in the
Park. To Matt Lutz, Ben Lennertz, Justin Snedegar and Emmy Feamster, Marina Fo-
lescu and Alex Radulescu, Justin and Amanda Dallmann, Johannes Schmitt, Jen Liderth,
Aness and Myles Webster, Shyam Nair, Julia Staffel and Brian Talbot, Alida Liberman
and Joshua Crabill, Ara Astorian, Kenneth Silver, Matthew Babb, Abelard Podgorski,
Indrek Reiland, Sam Shpall, Lewis Powell, and Keith Hall: thank you. I’ve also bene-
fited greatly from discussing my work with various post-docs that have visited USC, from
iii
Bryan Roberts, Alex Sarch, Leif Hancox-Li, to Monica Solomon: thank you as well.
Now, as lovely as Erik, Caleb, Steve, Beth, the older graduate students and the post-
docs were, there is am undeniable generational divide between us. None of them grew up
playing Pokemon or arguing about who would play which Spice Girl during recess. In my
second year, a new batch of first-years entered. There were now people with which I could
debate the finer points of being a 90s kid. Weekly trivia nights with Mike and Amanda
Ashfield, Jonathan Wright, Renee Bolinger, and Maegan Fairchild were what kept me go-
ing through the pressures of second year and the ever-looming second-year review process,
which would determine if I was worthy enough to continue in USC’s program. Every
Thursday night we would meet at Barney’s Beanery in Oldtown Pasadena, and we’d just
unwind. And although we were really bad at trivia, we were really good at the free raffle
that accompanied trivia. I cannot stress enough how important those nights were to me
and how important those friendships became. But then, Mike and Amanda decided to
have a kid, and I fell in love with a girl that lived in Santa Monica. Trivia nights weren’t
ever quite the same after that.
But before I move on from them, let me add that Mike and Amanda are two of the
kindest people I know and they’re raising two amazing kids. I also couldn’t have asked
for a better teaching partner than Mike. Mike approaches every topic he comes across as
though it is inherently interesting, and that’s a quality I greatly admire. Renee is stubborn
and determined and has a single-mindedness that is admirable. Jonathan’s got that gift for
getting people out of their shell and making them feel welcome, and Maegan’s not only
extremely kindhearted, she’s also got that talent of explaining your own research back
to you in a way that’s better than you ever could’ve stated yourself. Together, with my
cohort, I couldn’t have asked for a better group of people.
I also owe an additional debt of gratitude to Tanya Kostochka. Tanya is one of the few
people who has ever witnessed truly cynical Rima. I am surrounded by a lot of optimists,
a lot of folks who just want to believe the best of others. I am grateful for all of them,
but I’m also grateful for Tanya as she’s is also someone that can cut through the noise and
see people for who they are and approaches interpersonal relationships with a healthy
dose of cynicism. I also have to thank August Gorman and Nathan Howard for long
conversations about philosophy and beds that I’ve crashed on. I also regret not getting to
know Nicola Kemp, Jesse Wilson, and Christa Peterson better.
Another magical event that not only created deeper friendships, but also resulted in
the formation of many dissertations, is John Hawthorne’s seminar. The seminar was
supposed to be about publishing. We were supposed to bring to that class an idea or a
paper that we hoped to turn into a publishable paper. Each week we would meet, present,
iv
give feedback, and then get drinks. And over the course of those fifteen weeks, we each
discovered that the small paper idea we had, well, to do it justice it’d have to be more
than one paper. That class is responsible for generating at least four dissertations at USC,
and I’m forever grateful to John Hawthorne and my fellow classmates Tanya Kostochka,
Mike Ashfield, Maegan Fairchild, Renee Bolinger, Matt Babb, and Matt Leonard for the
feedback they offered over that semester.
It is not only USC graduate students I have to thank, but also UCLA graduate stu-
dents. From hosting me as a prospective to long conversations about the nature of belief,
I am grateful to call Greg Antill a friend. In addition, talking philosophy with Laura
Gillespie (also the best wedding officiant Gabby and I could’ve asked for), David Friedell,
Kim Johnston, Eric Tracy, Will Reckner, Sarah Beach, Ally Peabody, and Olufemi Taiwo
has always been a delight. I guess I should also thank Kevin Lande, but that’s more so for
being a real person and not making me feel like we have to talk philosophy all the time.
In addition, there is The Compound: a set of duplexes right on the edge of West
LA and Santa Monica where Gabbrielle Johnson, Gabriel Dupre, Ayana Samuel, Am-
ber Kavka-Warren, Eleonore Neufeld, Andrew Stewart, Noah Gordan, Jasmine Gunkel,
and I live. And where Isaac Khalaf and Laura Heida used to live. The conversations I’ve
had with everyone at the compound, and the friendships that have been made over the
years, have been an important part of my growth as a person and a philosopher, and to
all my compound roommates: I am forever grateful. In particular, Gabe is secretly a big
softie that genuinely cares about others and would do whatever he needed to do to make
things work out for them. Amber holds herself to too high a standard and will succeed at
anything she sets her mind to (except parking). Ayana is everyone’s loveable grump, and
Elli her foil.
Finally, we get to my committee. The friends we choose and the people we surround
ourselves with are a reflection of the kind of person we are and the kind of person we hope
to be. And with that thought in mind, I found myself in the lucky position of being able
to work with not only excellent philosophers, but philosophers of excellent character. I
cannot stress enough that I made a very conscious decision with regard to who I chose to
work with. I chose each with the following thought in mind: if one day someone were
to say that I remind them of any member of my committee I would not have to qualify
or feel ashamed, e.g., “although x would never read anything...”. Each member of my
committee sets a standard for mentorship I not only look up to and hope to emulate, but
they’re also genuinely good people.
In Ralph Wedgwood, one finds patience and a deep well of intellectual curiousity.
Talking with Ralph can be a lot like researching a topic on Wikipedia. You start with
v
rational requirements and you end up discussing the Agrarian Revolution of the mid-17th
century (or the philosophical equivalent). Ralph joined USC in my second year with
Ralph I found the freedom to explore whatever topics I was interested in and a patient
conversational partner. That freedom was crucial to me finally landing upon this topic
that has consumed me, and that I expect will continue to consume me. It can be easy to
develop a narrow mind when working on narrow topics, but Ralph will always remind
you of just how much larger everything is. I cannot say enough nice things about Ralph
and I look forward to a long friendship.
In Mark Schroeder, one finds structure and foresight. Mark was one of the key reasons
I chose USC. When I visited USC as a prospective student, we talked for an hour in his
office and then for over an hour while crawling along the 10. My writing sample had also
been on expressivism and epistemic modals, so of course it made sense that I wanted to
work with Mark. However, my interest in expressivism waned and I did so horribly in
my first class with Mark that I was afraid to keep working with him because surely he
thought I was irredeemably stupid. The two quotes at the start of the acknowledgements
are from end of year reports written by Mark. But, for some reason, Mark continued to
believe in me. He continued to ask to read my writing and provide feedback, and once I
was finally able to articulate the problem of the supposedly rational racist, we both were
hooked. None of this yet speaks to the qualities I mentioned at the start of the paragraph:
structure and foresight. Whereas with Ralph I had the freedom to just explore, Mark is
who I turned to when I needed to really nail down an idea. Mark can’t see the future, but
he can see exactly where an argument leads. It’s not a skill I’ve figured out yet, but until
I do, I know that I’ll always still be able to send Mark any thoughts I have.
In Kenny Easwaran, one finds one of the most genuinely good and kind person to
have ever existed. I do not think there is a bad bone in his body and very few people have
come close to being as genuine (only other Canadians come to mind). I thought I hated
epistemology until I took a class with Kenny and found out that epistemology wasn’t just
about knowledge, but also about belief, the degrees those beliefs come in, justification,
evidence, etc. Kenny opened up a whole area of philosophy to me, but most importantly
I have never left a conversation with Kenny feeling as though I can’t do philosophy, no
matter how hard the topic. Finally, not only does Kenny have the gift of being able to
explain complex topics in a simple way, he and Ralph share the following characteristic:
they know a lot about a lot of stuff that’s not philosophy. As a result, I’ve learnt not
only a lot about philosophy from Kenny, but also about city planning and bike lanes.
Losing Kenny to Texas A&M was a huge loss for USC, but I’m lucky to have been able
to continue to work with him.
vi
In Steve Finlay, one finds an uncanny ability to inhabit differing perspectives when
reading a piece of work to figure out all the ways in which a sentence is likely to be misread.
This careful attention and broad perspective is one that I hope to be able to offer for my
own students one day. Steve is also part of the camp that recognizes that to do philosophy,
one must read more than just philosophy. I’m also grateful for our discussions about
Doctor Who. On his advice I decided to give the show a second chance (after becoming
frustrated with Stephen Moffat), and it turns out that Season 10 is actually quite good.
Thanks, Steve!
Due to “understood” rules that weren’t explicitly listed, Pamela Hieronymi was not
able to be an official member of my committee. She did, however, play the part in ev-
ery way. In Pamela Hieronymi, one finds a standard of precision and rigour to aspire to
(and that is characteristic of UCLA’s reputation—Gabby made me add that). Pamela is
someone who understands that there is something we have to get right when we’re doing
philosophy. We’re not just asking questions and clumsily positing answers, there’s hard
work to be done and not everyone’s up to the task. Over the course of writing this dis-
sertation, Pamela has been my hardest critic, and for that I am deeply grateful. Pamela
has provided that useful outsiders perspective. If I could convince Pamela of a point, then
I was golden. As to whether I’ve succeeded... luckily I’ve got a few more years in the
profession to keep trying.
Altogether, I couldn’t have asked for a better committee. There are also the outside
members I should thank. I’m grateful to Roumyana Pancheva for playing the role of the
outside member for my qualifying defense, and to Aimee Bender for stepping in on such
short notice for the dissertation defense.
There are also faculty that I didn’t work closely with, but who’ve contributed to any
successes I’ve had at USC: Shieva Kleinschmidt, Robin Jeshion, Janet Levin, and Gabriel
Uzquiano. In particular, I want to say a bit more about both Shieva and Robin. Shieva
provided friendship, guidance, and a buddy to watch superhero movies with. I regret the
way in which we drifted apart once I moved to west of the 405, but I hope that there
will be more ice cream and Marvel in our future. In addition, Robin Jeshion is the best
teaching mentor—and mentor in the profession more generally—I could have asked for.
In addition, the office staff during my time at USC have been absolutely central to
any successes I’ve had: Cynthia Lugo, Corey Resnick, Barrington Smith-Seetachitt, J.N.
Nikolai, Natalie Schaad, and Rujuta Parikh. In particular, Cynthia, Barrington, and Na-
talie have saved me from many crises, and Natalie and Rujuta were the steady calming
voice of reason and competence I needed during the job market. I can’t say enough kind
things about the amazing office staff that we’ve been blessed with. And although I haven’t
vii
had a chance to interact with the newest member of our office staff yet, Michele Root did
process a reimbursement for me in less than an hour so I know she’s going to be fantastic
and we’re lucky to have her.
Further, parts of this dissertation have been presented at the New Insights and Di-
rections for Religious Epistemology Workshop at Oxford University, the University of
Toronto Mississauga, Athena in Action at Princeton University and the Princeton Work-
shop in Social Philosophy, the Penn-Rutgers-Princeton Social Epistemology Conference
at the University of Pennsylvania, the 2017 Central Division Meeting of the American
Philosophical Association, Duke University, the University of British Columbia, Wayne
State University, McMaster University, Georgetown University, Claremont McKenna
College, Queen’s University, Simon Fraser University, University of Missouri, and the
2018 Pacific Division Meeting of the American Philosophical Association. I’ve benefitted
tremendously from feedback that I’ve received every time I’ve presented my work, and I
have tried to thank individuals for their contributions to my thinking on these topics in
footnotes throughout the dissertation. Inevitably I’ll have missed someone, for that I am
deeply sorry.
And finally, I’ve saved the two most important people for last: Gabbrielle Johnson
and Kripke (the dog). I don’t know where I’d be if not for Gabby. Gabby has been a
constant source of love, support, friendship, guidance, every virtuous trait you can name.
Gabby has been my copy editor, my sounding board, someone I can run any idea by and
in just talking it through she’ll make it better. Gabby is such a demanding interlocutor
that for the sake of our relationship we tend not to talk philosophy, but I will never send
something off until she’s given it the once-over. She has made me a better person and a
better philosopher, and I cannot thank her enough. I look forward to our living together
for a hundred autumns.
Last, but not least, Kripke is the best dog that anyone could ask for. She is truthfully
the only reason I work as hard as I do (when I do). One day I hope to be the dog-parent
she sees in me already, and to return the favour of her many years of friendship with the
giant yard she’s always deserved.
viii
TABLE OF CONTENTS
DEDICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
CHAPTERS
I. The Wrongs of Racist Beliefs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 The Motivating Challenge . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 The Supposedly Rational Racist . . . . . . . . . . . . . . . 2
1.1.2 An Assumption and A Constraint . . . . . . . . . . . . . 4
1.1.3 The Shape of Things to Come . . . . . . . . . . . . . . . . 6
1.2 Beliefs Aren’t Racist: Locating the Wrong Downstream from Belief 8
1.3 Locating the Wrongs Upstream From Belief . . . . . . . . . . . . . . 13
1.4 A Restricted Account of When Beliefs Can Wrong . . . . . . . . . 17
1.5 Another Hurdle: A Worry About Control . . . . . . . . . . . . . . 22
II. What We Epistemically Owe To Each Other . . . . . . . . . . . . . . . . . 25
2.1 Examining Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1.1 Mistaken Identity . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1.2 Wounded By Belief . . . . . . . . . . . . . . . . . . . . . . . 29
2.1.3 The Racist Hermit and The Security Guard . . . . . . . 30
2.1.4 Why Everyone Hates Sherlock Holmes . . . . . . . . . . 32
2.1.5 Reflections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2 On How We Relate to Each Other . . . . . . . . . . . . . . . . . . . . 35
2.2.1 The Involved and The Participant Stance . . . . . . . . . 35
2.2.2 The Cognitive Demands of The Moral Standpoint . . . 38
2.2.3 The Objective, The Theoretical, The Naturalizing Stance 45
2.3 When We Epistemically Owe More to Some Than To Others . . 50
2.4 Recognizing Our Epistemic Obligations . . . . . . . . . . . . . . . . 52
ix
III. The Moral Stakes of Our Beliefs . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.1 Rethinking the Evidentialist Assumptions . . . . . . . . . . . . . . . 59
3.2 Varieties of Pragmatic Encroachment . . . . . . . . . . . . . . . . . . 64
3.2.1 The Contextualist Roots of Pragmatic Encroachment . 66
3.2.2 Explaining Encroachment: Knowledge and Action . . 67
3.2.3 Explaining Encroachment: Knowledge-Justification and
Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.2.4 Explaining Encroachment: Belief and Action . . . . . . 72
3.2.5 Explaining Encroachment: Epistemic6= Evidential . . 74
3.2.6 The General Framework of Pragmatic Encroachment . 76
3.3 Extending Pragmatic Encroachment to Moral Encroachment . . 78
3.4 The Moral Stakes of Our Beliefs . . . . . . . . . . . . . . . . . . . . . 81
3.4.1 Low-Stakes variants . . . . . . . . . . . . . . . . . . . . . . . 82
3.4.2 What are moral stakes? . . . . . . . . . . . . . . . . . . . . . 87
3.5 Three Further Worries . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.5.1 The Incompleteness of Moral Encroachment . . . . . . 90
3.5.2 Perverse Consequences . . . . . . . . . . . . . . . . . . . . . 91
3.5.3 Too Demanding . . . . . . . . . . . . . . . . . . . . . . . . . 92
IV. Assessing Alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
4.1 Conflict Creation and Conflict Resolution Strategies . . . . . . . . 96
4.2 The All Things Considered Ought, Ought Simpliciter, or Just
Plain Ought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
4.2.1 Ought Simpliciter as Either the Moral Ought or the
Epistemic Ought . . . . . . . . . . . . . . . . . . . . . . . . . 102
4.2.2 The Free and Unsubscripted Sense of Ought . . . . . . . 104
4.2.3 Ought Simpliciter vs Moral Encroachment . . . . . . . . 106
4.3 Skepticism about Ought Simpliciter . . . . . . . . . . . . . . . . . . . 107
4.3.1 Three Arguments for Skepticism . . . . . . . . . . . . . . 107
4.3.2 Genuine Dilemmas . . . . . . . . . . . . . . . . . . . . . . . 111
4.3.3 Permissivism and Pluralism . . . . . . . . . . . . . . . . . . 113
4.4 No Positive Duties to Believe . . . . . . . . . . . . . . . . . . . . . . . 116
4.4.1 The Account . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.4.2 The Drawbacks . . . . . . . . . . . . . . . . . . . . . . . . . 120
4.5 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
x
INTRODUCTION
The dissertation begins with the example of Spencer, the supposedly rational racist.
Spencer argues that although it might be “unpopular” or “politically incorrect” to say
this, the evidence supports believing that the black diner in his section will tip poorly.
He insists that the facts don’t lie. The facts aren’t racist. If you were to deny his claim and
believe otherwise, he’d challenge that it is you who engages in wishful thinking. It is you
who believes against the evidence. You, not Spencer, are epistemically irrational.
An unfortunate fact (amongst many) is that we live in a society that has been shaped
by racist attitudes and institutions. Given the effects of structural racism, Spencer’s belief
could be supported by the evidence. Spencer notes that it might make him unpopular, but
he cares about the truth and he is willing to believe the unpopular thing. But, Spencer’s
belief seems racist. Spencer asks, however, how could his belief be racist if his beliefs reflect
reality and are rationally justified? Moreover, how could he wrong anyone by believing
what he epistemically ought to believe given the evidence?
The overarching goal of the dissertation is to give an account of the moral-epistemic
norms governing belief that will help us answer Spencer and the challenge he poses. In
answer, I argue that beliefs can wrong. At least one of the distinctive wrongs commit-
ted by a racist—and of central importance to this dissertation, the wrong committed by
Spencer—plausibly lies in what they believe about another human being.
To bolster support for this intuitive thought, there are many cases we can turn to
as prima facie evidence that we can wrong someone by what we believe of them. For
example, one common formulation of the Christian Eucharistic confession, “we have
sinned against you in thought, word, and deed”, appeals to the idea that we can sin against
God in thought, as well as in word and in deed. When loved ones believe the worst of
us, it is tempting to think that we can demand an apology for the beliefs they hold, and
not just their actions. Consider, for example, how common it is to hear the phrase, “You
shouldn’t have believed that of me!”. Further, many people also think that we can wrong
not only the living but also the dead when we believe the worst of them. As I will show
in this dissertation, the idea that we can wrong someone by what we believe reveals itself
in many places. But, you might counter, perhaps we simply mistaken when we talk this
way.
xi
I grant that this idea that beliefs can wrong is philosophically puzzling. The norms
that properly govern belief are plausibly epistemic norms such as truth, accuracy, and
evidence. Moral and prudential norms are taken to play no role in settling the question
of whether to believe p, and they are irrelevant to answering the question of what you
should believe. It would follow, then, that the person who believes of Barack Obama that
he’s more likely to be a valet than the President of the United States need only attend to
the relevant statistical likelihoods before settling their belief. As we will see argued by the
antagonist of this dissertation, Spencer, that they commit an injustice to Barack Obama
when they mistake him for a valet ought to be irrelevant in assessing the belief itself (as
opposed to, say, its consequences). This leaves us with several puzzling questions: can we
wrong one another by virtue of what we believe about each other, i.e., can beliefs wrong?
If not, why not? What then is going on in these cases? If so, how? What then does this
tell us about the content and nature of the norms governing belief?
In this dissertation I explore this perceived conflict between the intuitive idea that we
can wrong one another by what we believe about each other, and the philosophically or-
thodox idea that the only norms that relate directly to belief concern truth-related factors
such as evidence. Motivated by the recognition that our epistemic practices exist in an
unjust and non-ideal world, I argue that there are moral and social constraints on our
epistemic practices. Specifically beliefs can wrong and the epistemic justification of our
beliefs can be affected by the moral demands of our environment. That is, when it comes
to what we should believe, morality is not voiceless: there is a moral dimension to the
way in which our beliefs are justified. This thesis has important consequences for not
only how we think about moral questions, e.g., what we owe to each other, but also how
we think about epistemic rationality. My dissertation is thus divided into two parts. In
the first, I make the case for the thesis that beliefs can wrong and I explain why, when,
and who our beliefs can wrong. Then, in the second half I show how these moral consid-
erations affect the epistemic rationality of Spencer’s belief by developing an account of
moral encroachment.
Proceeding chapter by chapter, the task of Chapter 1 is to explain why there’s a wrong
that should be located in belief. Many try to explain away the appearance that beliefs can
wrong by locating the wrong elsewhere, e.g., in the actions of the agent. However, these
alternative accounts only give us a diagnosis of Spencer. These accounts fail to both engage
with Spencer and fully explain the moral wrong in these cases. To give a diagnosis that
also speaks to Spencer, we must appeal to the surprising thesis that people can wrong other
people in virtue of what they believe about each other, and not just in virtue of what they
do. The task of Chapter 1, in short, is to make a case for why we should think there is a
xii
wrong that is located in what an agent believes of another person as opposed to what they
say or do.
In Chapter 2 I show that the phenomenon of wronging beliefs is more general than
just Spencer’s racially charged beliefs. Although you might not be harmed when someone
believes something of you on the basis of statistical evidence, you are nonetheless wronged
because a moral demand has not been met: the demand to be related to in a way that is
characteristic of being an agent rather than an object. Despite the fact this thesis seems
to make an implausibly wide range of things morally problematic, I suggest we switch
from considering Spencer as our central example to considering what it feels like for the
agent who is mistaken for a waiter on the basis of their race or presumed to be a bad
tipper on the basis of their race. In many contexts if your melanin levels more closely
match that of the waitstaff than that of the diners, it seems the belief that you are a waiter
is supported by the weight of the evidence. Granting that there is nothing wrong with
being a waiter, why then does it hurt when someone mistakes you for a waiter in this
way? In this chapter I argue that this way of forming beliefs about others fails to meet the
moral demand of recognizing others as agents, rather than objects. We fail, in short, in
what we epistemically owe to each other.
We now arrive at a new problem. Although the evidentially rational belief might be
morally wrong, the belief still seems like it might be epistemically fine. This suggests the
following dilemma: given the evidence, you ought believe; but, given the moral consider-
ations, you shouldn’t. After all, in many real-world scenarios, Spencer seems right. That
a diner is black can provide a substantial amount of evidential support for the belief that
the diner will not tip well. If Chapter 1 and 2 are correct, it is morally impermissible to
believe that he will not tip well. On the other hand, the belief seems epistemically justi-
fied. If one had more hands, one might go further and add that it would be epistemically
irrational not to believe that he won’t tip well. To refrain from believing seems to involve
a wilful disregard of the evidence. So, what should you believe?
Many conclude that in such cases we face an irresolvable conflict between what moral-
ity requires and what we epistemically ought to believe. I disagree. In Chapter 3, I argue
that we can resolve this conflict by appealing to a form of moral encroachment. That is,
the epistemic justification of our beliefs can be determined, in part, by the moral demands
of our situation, i.e., the moral stakes. As an upshot of this account, the epistemic consid-
erations one weighs when trying to settle the question of what to believe include moral
considerations. That is, moral and epistemic considerations need not stand at odds with
one another. I argue that we must rethink the assumption that gives rise to these con-
flicts: the evidentialist assumption that only evidence matters for justification. Evidence
xiii
alone is not sufficient. I argue that non-evidential considerations—in particular, the moral
stakes of a belief—must enter into the picture here. This is the phenomenon of moral en-
croachment. What is sufficient evidence does not supervene on what your evidence is.
The moral stakes can affect whether a belief is justified.
Finally, in Chapter 4, I canvas reasons for dissatisfaction with the traditional answers
to this supposed dilemma between moral and epistemic demands, e.g., that there is an all-
things-considered ought that adjudicates conflicts. Given that the epistemic justification
of our beliefs can be affected by the moral demands of our situation, I argue that we
can prescind from questions about whether there is an all-things-considered ought that
adjudicates the conflict. Spencer’s belief is both morally objectionable and epistemically
irrational. Further, it is a mistake to think that our moral lives and our epistemic lives are
separate realms governed by conflicting obligations. As a final take away, to be moral we
must not only do better, we must believe better as well.
xiv
CHAPTER I
The Wrongs of Racist Beliefs
We care not only about how people treat us, but also what they believe of us. Suppose
I believe that you’re a bad tipper given your race. Have I wronged you? What if you are
a bad tipper? It is commonly argued that the way racist beliefs wrong is that the racist
believer either misrepresents reality, organizes facts in a misleading way that distorts the
truth, or engages in fallacious reasoning. In this opening chapter I begin by presenting a
case that challenges this orthodoxy: the case of the supposedly rational racist. We live in a
world that has been, and continues to be, structured by racist attitudes and institutions.
As a result, the evidence might be stacked in favour of racist beliefs. But, if there are racist
beliefs that reflect reality and are rationally justified, what could be wrong with them?
Moreover, how do I wrong you by believing what I epistemically ought believe given the
evidence?
To address this challenge, I argue that we must recognize that there are not only epis-
temic norms governing belief, but moral ones as well. This view, however, is at odds with
the assumption that moral obligation requires a kind of voluntary control that we lack
with regard to our beliefs. This background assumption motivates many philosophers
to try to explain away the appearance that beliefs can wrong by locating the wrong else-
where, e.g., in an agent’s action. Further, even accounts that accept the thesis that racist
beliefs can wrong restrict the class of beliefs that wrong to beliefs that are either false or the
result of hot irrationality, e.g., the racist belief is a result of ill-will. In this chapter I argue
that although these accounts will capture many of the wrongs associated with racist be-
liefs, they will at best be only partial explanations because they cannot explain the wrong
committed by the supposedly rational racist.
The challenge posed by the supposedly rational racist concerns our epistemic practices
in a non-ideal world. The world is an unjust place, and there may be many morally objec-
tionable beliefs it justifies. I argue that to address the challenge posed by the supposedly
rational racist we must seriously consider the thesis that people wrong others in virtue of
1
what they believe about them, and not just in virtue of what they do. In short, my goal in
this chapter is to show that we miss something when we ignore the possibility of doxastic
wronging. As for what that something is, that I will explain in Chapter 2.
1.1 The Motivating Challenge
1.1.1 The Supposedly Rational Racist
It is not up for debate that we live in a world that has been shaped by, and continues
to be shaped by, racist attitudes and racist institutions. From the transatlantic slave trade,
to anti-miscegenation laws, lynchings, redlining, and voter identification laws that “target
African-Americans with almost surgical precision”, racism is an unfortunate part of the
fabric of our world.
1
It should not be surprising, then, that as a result of structural racism,
there may be morally objectionable beliefs that are well-supported by the evidence. Fur-
ther, some of the morally objectionable beliefs could be paradigmatic examples of racist
beliefs. Although my focus in this chapter are racist beliefs and the ways they can wrong,
what I say can also be easily extended to cover the possibility of rationally held sexist be-
liefs, homophobic beliefs, and other morally objectionable beliefs of this kind. The world
is an unjust place and there may be many morally objectionable beliefs that it justifies.
Consider, for example, Miranda Fricker (2007, pp. 88)’s description of the following
case. In The Talented Mr. Ripley, the character Marge is constantly being construed as
hysterical by the males in the novel. Their dismissive attitude towards her and her testi-
mony builds up to her eventual breakdown and having to be physically restrained. She
becomes what she has been constructed as: a hysterical female that expresses herself in
semi-contradictions who disregards the facts and is unable to keep a grip on her emo-
tions. This fictional case illustrates how it’s not unreasonable to think that if women are
constantly told that they are hysterical, then they might start behaving in that manner.
As a result, the sexist belief that women are hysterical may become well-supported by the
evidence.
2
Continuing this theme, let us consider the following case that is the motivating puzzle
of this dissertation.
The Supposedly Rational Racist. You shouldn’t have done it. But you did.
You scrolled down to the comments section of an article concerning the state
1
Regarding North Carolina’s discriminatory voter ID law, see:
https://www.nytimes.com/2017/05/15/us/politics/voter-id-laws-supreme-court-north-carolina.html?
2
For more on this point see Haslanger (2012) who draws on Hacking (1995, 1999)’s work on the “looping
effects” of social kinds and structures. These social kinds can be self-fulfilling and self-reinforcing. Thanks
to Fran Fairbairn and Jessie Munton for bringing this to my attention.
2
of race relations in America, and you are now reading the comments. The
comments on such articles tend to be predictable, but there is one comment
that catches your eye. Amongst the slurs, the get-rich-quick schemes, and the
threats of physical violence, there is the following comment: “Although it
might be ‘unpopular’ or ‘politically incorrect’ to say this, I’m tired of con-
stantly being called a racist just because I rightly believe of the black diners
seated in my section that they will tip worse than white diners in my sec-
tion.” The user posting the comment, Spencer, argues that the facts don’t lie,
and he helpfully reproduces those facts. For example, he links to studies that
show that on average black diners tip substantially less than white diners. The
facts, he insists, aren’t racist. If you were to deny his claims and were to be-
lieve otherwise, it would be you who is engaging in wishful thinking. It would
be you who believes against the evidence. It would be you, not Spencer, who
is epistemically irrational.
If we could dismiss Spencer’s comment on the grounds that he is making a mistake, there
would be no need to continue reading. The challenge posed by Spencer is that we cannot
so easily dismiss his comment. From a study by Michael Lynn (2006), there is evidence
of racial disparities in tipping practices that support Spencer’s belief that black diners
tip worse than white diners. Perhaps Spencer is, as he suggests, rationally believing in
accordance with the evidence. Although his belief might seem racist and as such irrational,
perhaps he is not guilty of making any kind of epistemic fault. That is, perhaps he is not
misinterpreting the evidence or cherry-picking statistics to justify his pre-held prejudices.
Perhaps the belief isn’t the result of ill-will or hostility. Perhaps Spencer is right in his
claim that it is not his fault that as a result of structural racism the evidence is stacked in
favour of racist beliefs. Given that our world is racist, some of our beliefs will also be
racist.
A rational racist, however, sounds like an oxymoron. Racism is often regarded as defi-
nitionally involving standard epistemic flaws such as false beliefs, obscuring or concealing
or disregarding relevant information, organizing facts in a misleading way, and/or engag-
ing in fallacious reasoning. For example, as Tommie Shelby (2016, pp. 22) argues, “in its
most basic form, racism is an ideology: a widely held set of associated beliefs and implicit
judgments that misrepresent significant social realities and function, through this distor-
tion, to bring about or perpetuate unjust social relations.”
3
The problem thus posed by
the supposedly rational racist is that Spencer does not seem to be committing one of these
3
This is a commonly shared view. See for example, Appiah (1990), Ikuenobe (2011), Lengbeyer (2004),
Clough and Loges (2008), etc. I will explore similar accounts in more detail in Section 1.4.
3
standard epistemic flaws, but his belief nonetheless seems racist, and by believing what he
does—i.e., that the black individual in his section will tip poorly—he seems to commit a
wrong. But, if he is not guilty of an epistemic flaw, then it seems that we are at a loss to
explain what wrong he commits.
Perhaps the reason we are at a loss to explain what wrong Spencer commits is simple:
Spencer is not committing a wrong. Sometimes there are unpleasant things that we ought,
at the very least, are permitted, to believe. For example, it might be unpleasant for me to
believe that one day soon my favourite television show will come to an end (as many of
my past favourite television shows have done, RIP Battlestar Galactica); nonetheless, it is
something I must accept and it would be irrational for me to believe otherwise. This is
precisely the line that Spencer advocates: there is at least some value to having accurate
beliefs, even when those beliefs are unpleasant. The beliefs that Spencer holds, however,
are not merely unpleasant. The beliefs that Spencer holds seem racist. But, again, one
might counter that although the beliefs seem racist, it would be wrong to call them racist
given that they reflect reality.
The challenge, then, is to explain why we should call Spencer’s belief racist if his beliefs
reflect reality and seem to be rationally justified. Perhaps by calling such a belief racist we
are contributing to a conceptual inflation of the term and thereby making it less powerful
for calling out things that are actually racist. Perhaps it is not racist to believe a black diner
won’t tip well. This is the challenge posed by the supposedly rational racist.
1.1.2 An Assumption and A Constraint
This is a troubling conclusion. I am not as willing as Spencer to give up on the com-
pelling idea that there is something wrong with what he believes even when the belief
might accurately reflect reality. Assuming the reader shares this hesitancy, this leaves us
with the task of figuring out what Spencer does wrong.
As we move forward with this task, there is one assumption and one constraint I ask
the reader to take on. First, I will be assuming that by believing what he does, Spencer
commits a wrong.If we take it as given that Spencer commits a wrong, our task becomes
explaining what wrong Spencer commits. To reject this assumption is to reject that there
is a challenge posed by Spencer. That is, if Spencer does not commit a wrong, there is no
puzzle that needs to be explained. The case of the supposedly rational racist, however,
is puzzling because of how compelling this assumption is, i.e., that Spencer commits a
wrong by believing what he does.
4
4
In this chapter I will argue that Spencer commits a wrong and the wrong is properly located in what
he believes, and then in Chapter 2 I will give an account of why Spencer’s belief wrongs.
4
Further, I also ask the reader to take on the following constraint on the range of ac-
ceptable answers to the case of the supposedly rational racist: when giving an explanation
of the wrong committed by Spencer, we should, as much as possible, do our best to grant
Spencer his conception of himself.
5
Spencer is not merely a philosopher’s concoction to
counterexample or test the limits of a proposed theory. Spencer exists. Spencer might be
our brother sharing infographics of “black-on-black crime”; Spencer might be our aunt
that everyone politely tolerates at Thanksgiving dinner when she complains about immi-
gration reform given her belief about the correlation between undocumented immigrants
and crime rates. Your aunt is always careful to choose her words so as to avoid the mis-
conception that she’s being racist, to present herself as though she is simply compelled
by the facts. We could easily write off our brother or our aunt as making some kind of
mistake, but just as I suggest we try to engage with Spencer and not believe the worst of
him, we certainly do not want to believe the worst of our loved ones.
We could tell our brother that the fact that he’s constantly sharing infographics about
black-on-black crime reveals his motivated irrationality, e.g., his ill-will towards black
people. His belief that 90% of African-American homicides are committed by African-
Americans might not be false, but that he claims to be justified in believing that his black
neighbour is dangerous is the result of both a kind of motivated irrationality and a misun-
derstanding of statistics. That is, our brother is clearly willing to believe anything about
his black neighbour as long as it is negative. We could also say to our aunt that her belief
about a correlation between crime rates and undocumented immigrants goes in the op-
posite direction, i.e., that undocumented immigrants commit less crime than American
citizens. I grant that these descriptions are often accurate, and that encountering some-
one who truly is as Spencer envisions himself is rare. Nonetheless, the case of Spencer
stands apart from our brother and our aunt. I ask that we take on the following con-
straint regarding our diagnosis of Spencer: we should try to describe him in a way that he
would accept. Ideally, we wish not only to explain the wrong/s that Spencer commits in
believing what he does, we also do not want to rule out the possibility of engaging with
Spencer. Engagement, however, requires working with this constraint and not speculat-
ing about the nature of Spencer’s psychology or making claims about various character
5
You might worry here that this constraint conflicts with the assumption that Spencer commits a wrong.
After all, surely Spencer does not accept that he commits a wrong. Hence, I only claim that we should, as
much as possible, do our best to grant Spencer his conception of himself. Crucial to Spencer’s self-conception
is that he is epistemically rational. To show my hand, I plan to show over the course of this dissertation that
it is not at odds with epistemic rationality to accept that others can place demands upon on how we respond
to evidence. If Spencer grants that for his friends, then it is a short step to granting that for vulnerable
minorities like the black man in his section. It is not, I argue, at odds with the demands of epistemic
rationality to exhibit epistemic partiality.
5
flaws he might exhibit. To engage, ideally we should strive for a response to Spencer that
would show that we can grant his conception of himself and still, by his own lights, he
should not believe what he believes.
1.1.3 The Shape of Things to Come
Returning to the task of explaining how Spencer commits a wrong by believing what
he does, the first problem we face is explaining why Spencer’s belief should be called
‘racist’. The first problem is that the belief seems to accurately reflect the evidence. Sec-
ond, perhaps the term ‘racism’ should be reserved for morally problematic false beliefs.
Further, when it comes to what the term ‘racism’ should be reserved for, perhaps it
shouldn’t apply to beliefs at all. Perhaps the term should be reserved for the actions that
follow from a belief.
In Section 1.2 I consider this strategy of locating the wrong downstream from the
belief. According to this strategy, Spencer ought believe what he does, but where he
would go wrong is in acting on the belief. Against this proposal, I argue that if we restrict
ourselves to locating the objectionability of racist beliefs in the downstream effects of the
beliefs, we fail to capture a number of other ways in which racist beliefs can wrong. We
care not only what people do to us, but also about the attitudes and beliefs they hold
about us. This strategy, then, leaves us with only a partial explanation of the wrong of
racist beliefs. For a more complete explanation, we must recognize that there are other
places the wrong can be located.
In Section 1.3 I consider accounts that locate the wrong of racist beliefs upstream from
the belief, i.e., the many processes that contribute to the agent forming the belief. For
example, the wrong might lie in an agent’s cognitive limitations or predispositions. The
problem with these accounts is twofold. First, they violate the working constraint that we
grant Spencer his conception of himself; second, the accounts also run into trouble with
the assumption that deontic notions do not apply to beliefs because we lack sufficient
voluntary control over our beliefs. This second assumption is part of the motivation of
locating the wrong elsewhere. Although you could avoid endorsing the thesis that beliefs
can wrong by locating the wrong upstream from belief, what little control we exercise
over what we believe is surely more than the control we exercise over any upstream feature
of our cognitive architecture. The irony for upstream accounts is that the motivation for
locating the wrong upstream is also an objection against attempts to locate the wrong
upstream. If beliefs can’t wrong because we lack sufficient control over beliefs, then it’s
not clear how cognitive predispositions could wrong either. I conclude this section by
noting that we could nonetheless say that Spencer’s wrong is that he ignores the risk of
6
creating a biased cognitive system when he believes that his black diner, Jamal, will tip less
well than his white diner, James. However, this requires endorsing the thesis that beliefs
can wrong.
In Section 1.4 I turn to accounts that locate the wrong in what an agent believes. I call
these accounts restricted accounts of the wrong because they specify two conditions for
racist beliefs that wrong. First, the belief is false, i.e., the false condition. Second, the belief
is a reflection of a character flaw such as ill will or racial disregard, i.e., the hot condition.
Spencer does not satisfy the first condition—his belief seems rationally justified and it is
not false. Further, the second condition requires denying the working constraint that we
grant Spencer his conception of himself. I grant in the vast majority of cases people who
claim to be like Spencer satisfy either the false condition or the hot condition. However,
the case of Spencer illustrates that they needn’t. Thus, these accounts are incomplete.
This brings us to the final section. In Section 1.5, I defend the intuitive plausibility of
the thesis that beliefs can wrong. As we will have seen in the previous sections, the stan-
dard explanations for the wrong committed by Spencer do not work. I suggest, then, we
rethink the assumptions that both stand in the way of the thesis that beliefs can wrong,
and prevent us from positing moral-epistemic norms governing belief. Doing so allows
us to preserve the working assumption and working constraint, i.e., that by believing
what he does Spencer commits a wrong, and in explaining the wrong we should describe
Spencer in a way that he would accept. As I have been noting, however, the thesis that
beliefs can wrong conflicts with a commonly accepted background assumption regarding
the nature of belief and moral disapprobation: beliefs lack the kind of voluntary control
needed for our beliefs to wrong. My argument against this final challenge is indirect: hav-
ing ruled out all plausible alternatives for locating the wrong committed by Spencer, we
arrive at the conclusion that it is Spencer’s belief itself that wrongs. The final consider-
ation, then, is a choice between Spencer not wronging at all or Spencer having wronged
through what he believes. It seems intuitively obvious to me that the supposedly rational
racist does wrong. And as I shall show in this final section, this is more obvious, in fact,
than the theoretical normative posit that voluntary control is always required to wrong.
To give Spencer a pass on the basis of a theoretical technicality does a disservice to the
common folk-intuition that something has gone wrong in the case of Spencer. It is thus
worth considering how it could be true that beliefs can wrong.
7
1.2 Beliefs Aren’t Racist: Locating the Wrong Downstream from Be-
lief
Our task is to explain the wrong Spencer commits. I have already showed my hand by
noting that the answer I find most compelling is that by believing what he does Spencer
does something wrong. This is not, however, the most obvious answer. To get to this
answer, we must first work our way through other accounts that have been offered and
show their shortcomings.
In this section, my focus is on downstream accounts of the wrong. According to
downstream accounts of the wrong, it is not the belief itself that wrongs: rather, it is
the downstream effects of the beliefs that wrong, e.g., the agent’s actions. Spencer only
does something wrong given what he believes insofar as he acts on what he believes. Here
is the intuitive thought behind such accounts: we have been assuming that what Spencer
believes is an accurate reflection of the evidence. Now we can ask, how might Spencer
wrong a black diner by simply believing that they are going to tip less well than a white
diner? Surely he doesn’t wrong the black diner unless he treats the black diner differently
than the white diner. For example, it is clear that he would wrong the black diner if he
were to provide worse service on the basis of the diner’s race. But, if he simply believes?
What’s the harm (or wrong) in that? If no racist actions follow from the belief, then he
does not do anything wrong. The downstream theorist, thus, can accept a restricted ver-
sion of my claim that people wrong each other in virtue of what they believe about each
other. People wrong each other, this theorist claims, in virtue of what they believe about
each other, but only insofar as how they treat each other is a reflection of what they believe
about each other.
As we can see in this formulation, it is implicitly assumed that our beliefs toward one
another are beyond the purview of what’s regarded as proper or improper treatment.
There are two reasons for this restriction. The first is an assumption about control and
responsibility: moral obligation requires a kind of voluntary control that we lack with
regard to our beliefs. If you accept this assumption, then you would be driven to locate the
wrong downstream from belief in something that we do have voluntary control over, i.e.,
what we say and what we do. The second reason stems from a worry about the conceptual
inflation of the term ‘racism’. If you are concerned that the terms ‘racism’ and ‘racist’ are
being over-applied and losing their force as terms of moral reproach, you might be driven
to reserve the term ‘racism’ and ‘racist’ for the worst race-based harms, and be disinclined
to apply the terms to what an agent believes. Although distinct reasons, both reasons
can be summed up with the following slogan: beliefs don’t wrong people, people wrong
8
people. Alternatively, people wrong people through their words and through their deeds,
not simply what they believe.
I turn now to considering the case for downstream accounts of the wrong in more
detail. As I’ll show, such an account will be too narrow in scope. It will fail to capture
the more subtle, insidious, and more prevalent forms of racism. Further, there is intuitive
reason to think that the demand for proper treatment extends to what people believe of
us, not simply how they act towards us or what they say of us. Although this would put
us at odds with the assumption about control and responsibility, there might be reason
to think that that assumption may not be as strong as it is commonly assumed to be. I
leave the argument for that point, however, to Section 1.5. Let us begin, then, with the
concern about the conceptual inflation of the term of ‘racism’ and why that leads us to
locate the wrong of racist beliefs solely in the actions of an agent.
A prominent example of an argument for why the wrong of racism is located only in
action comes from J. Angelo Corlett (2005). Corlett argues that the only kinds of ethnic
prejudice and discrimination that rise to the level of racism, what he calls “racism proper”,
are overt negative actions or inactions that are harmful and hurtful to victims. Further,
these actions or inactions must be egregious enough to be punished by law. His grounds
for arguing this come from the belief that “it does little good (though I suppose some
good, on some occasions) to call something racist when the law cannot and should not
effectively deal with it” (pp. 578-9).
There are two parts to Corlett’s view that we need to keep separate. The first claim is
that “racism proper” is located only in action, and the second is that actions and inactions
that the law cannot and should not deal with are rightly excluded from “racism proper”.
The second claim is Corlett’s grounds for the first. There are, however, good reasons
for rejecting the second assumption. But first, why might we want to delineate a special
case of racist wrongs as rising to the level of racism proper? For this argument, we must
turn to Lawrence Blum (2002)’s influential distinction between, on the one hand, racist
beliefs and racist actions and, on the other, and beliefs and actions that are merely racial
in nature.
Blum argues that the term ‘racism’ has been the victim of a conceptual inflation. The
term has been overused and, in serving as a term of moral reproach, it has joined other
vices such as dishonesty, cruelty, hypocrisy, etc. Blum argues that we need a more varied
and nuanced moral vocabulary in order to capture the sense in which not every instance
of racial conflict, insensitivity, discomfort, miscommunication, exclusion, injustice, or
ignorance is “racist”. He argues that insensitivity, ignorance, discomfort, etc. in actions
and beliefs are lesser ills, and the term ‘racism’ should be reserved for the worst race-based
9
wrongs, rather than used to categorize and condemn all race-related ills. Applying this
account to our case of Spencer, Corlett and Blum would say that Spencer’s belief might
be morally bad in many ways, but it is not racist.
However, this policing of when the term ‘racism’ can be employed is problematic.
Racism takes various forms and admits of degrees of wrong. Lynching, redlining, and the
transatlantic slave trade are all particularly egregious examples of racism. Although being
thought of as a bad tipper on the basis of your race is not as bad as not being allowed to
marry outside of your race, it does not follow that the former is not racist. Blum makes
the distinction because he is concerned with how dialogue about racism might be impeded
by charges of “racism”. And Corlett makes this distinction between racism and “racism
proper” because he is concerned with legal remedies in a racist society. But, as Shelby
(2003, pp. 125) argues, these approaches downplay the value of using the term.
6
Many people in the US now think that racism is largely a problem of the past
and are thus little concerned with racial issues. The expansive conception of
racism with its strong negative valence could encourage those who are com-
placent to be more vigilant and circumspect in the racial realm. Perhaps they
should accept the comparatively small burdens of hyperbolic uses of “racism”
as the cost of achieving racial justice and of publicly affirming the humanity
of those who are racially stigmatized.
[...]
Moreover, though many of Blum’s “non-racist” race-related ills are lesser moral
evils taken singly, their cumulative and wide-ranging impact on the life-chances
of subordinate “races” creates a heavy burden that is properly thought of as
a form of oppression. Calling these lesser wrongs “racist” reminds us of the
seriousness of these sources of disadvantage, even if their perpetrators are not
moral monsters.
These considerations allow us to reject attempts to distinguish between different degrees
of the wrong, e.g. racial wrong, racism, and racism proper, and argue that there is good
reason for calling all the ills that fall on that spectrum racist or instances of racism. And
although Corlett’s argument for locating the wrong of racism solely in action or inaction
rested on the assumption that there was a special class of wrong, i.e. racism proper, that
could only be located in action or inaction, we have not yet undermined the intuitive
6
Blum later changes his mind in Blum (2004) and argues that there is a plurality of racial ills that could
all be regarded as racist.
10
case for thinking that the wrong of racism resides only in action. So, let me now turn to
providing this intuitive case and then challenging it.
Ferguson. Consider Officers Stella and Stanley who are told in their morning
briefing that the latest statistics available suggest that 92% of black residents
in their neighbourhood have open arrest warrants. Later, they go on their
separate patrols and each see a black resident. They both form the belief that
the black resident has an open arrest warrant. Stella uses this belief to stop
the resident and run their name through the database to see if they should be
taken in to the precinct. Stanley, too believes that the resident he encounters
has an open arrest warrant: however, he chooses not to stop the resident for
other reasons. Stanley knows that the justice system is corrupt and that his
other police officers have been using warrants and fines on the residents as
a way of padding the municipal budget. So, although he believes that this
resident has an open arrest warrant, he doesn’t act on the belief.
7
The difference in moral blame between Stella and Stanley suggests that it is not just what
they believe that is morally problematic. If it were the case that the one belief that some
resident has an open arrest warrant were morally problematic, then both Stella and Stan-
ley would be just as bad as one another. But, Stanley seems commendable whereas Stella
does not. Thus, it seems that what distinguishes them is how they act on the basis of
their beliefs. But, if we focus on just the acts, however, we’ll fail to capture other ways in
which we can wrong one another. In contrast with the case given above, now consider
the hermit in the woods who holds racist beliefs.
The Racist Hermit. The racist hermit in the woods will never interact with
the disadvantaged person he believes something negative of, he will never in-
teract with or contribute to the institutional structures of racism. He may
be a product of these institutional structures, but we are hard pressed to say
that he contributes to them given his isolation. But, suppose that he comes
to believe that Sanjeev smells like curry. How did he come to form this be-
lief given his isolation from society? Let us just suppose that he discovered
7
This statistic is loosely based on the Department of Justice’s investigation of the Ferguson police de-
partment which found that 62% of residents had open arrest warrants, and of those residents, 92% were
black. On Aug 25, 2015, a municipal court judge in Ferguson issued an order to withdraw all arrest war-
rants issued in Ferguson before Dec 31, 2014 due to the Ferguson police and the city’s municipal court
pattern and practice of discrimination against African-Americans. If you are worried that this case seems
contrived, I note that in neighboring predominantly black municipalities, the ratios are even worse. For
example, in Country Club Hills there are 33,102 active warrants for a population of 1,381. Similarly, in
Wellston there are 15,000 outstanding warrants for a population of 2,460.
11
some trash on the ground which happened to be an alumni newsletter from
Sanjeev’s university that included a picture of him. Upon seeing that picture,
the hermit believes of the pictured person—Sanjeev—that he smells of curry.
Now, suppose also that Sanjeev happens to have recently made curry so in
this instance the hermit’s belief is true—Sanjeev does smell of curry. Has the
hermit wronged Sanjeev?
According to the downstream theorist, the hermit does not and cannot wrong Sanjeev be-
cause Sanjeev will never learn of the hermit’s belief. However, neither harm nor knowl-
edge are a prerequisite for being wronged. As I will argue further in Chapter 2, the wrong
here appears to be relational. For example, if Sanjeev’s partner were to cheat on him but
he never found out, it is a commonly accepted intuition that Sanjeev would still have been
wronged. Following Thomas Nagel (1970), wrongs, such as betrayal, needn’t be appre-
hended by the agent that is wronged. To continue pressing this point that we care not
only how people act towards us and what they say about us, but also the attitudes they
hold of us, consider the following related case.
The Security Guard. Jake is a security guard at a fancy department store. He
hates the company he works for, and he couldn’t care less if people shoplift
and cost the company money. One day, as Jada leaves the department store
Jake works at, he believes that Jada shoplifted the purse she’s carrying. But,
given his contempt for his company, he chooses not to intervene. He acts
exactly as he would act if he had believed that that was Jada’s purse and that
she hadn’t stolen it. Has Jada been wronged by what Jake believes of her?
I want to press the intuition that Jada has been wronged by what Jake believes of her. If we
put ourselves in Jada’s shoes, we don’t want people to believe of us that we stole the purse.
We care what people believe about us.
8
The problem, then, with downstream accounts
of the wrong is that they are at best partial explanations of the wrongs of racist beliefs. It
8
If you are having difficulty with this intuition, consider the following analogous case from the movie
Pretty Woman. Vivian Ward, played by Julia Roberts, is an escort who has been hired by Richard Gere’s
character, Edward Lewis. Edward gives Vivian money for a new wardrobe. Vivian visits a store in Beverly
Hills, and the first shot when she enters the store is the reaction from the workers in the store. Although
she reminds them she has money to spend in the store, the clerks refuse to believe her and ask her to
leave. Although it seems like the wrong committed in this case concerns how the store clerks acted towards
Vivian, you need only ask almost anyone you know with a darker tint what it feels like in a store where
people believe you will shoplift. They may not always follow you around, they may not always act any
differently towards you, but that they believe you will steal, or that they believe you can’t afford anything
in the store, etc. hurts. See also Williams (1992, pp. 44-46)’s retelling of her experience in a Benetton store.
I discuss this in more detail in Chapter 2.
12
is right that one way people wrong each other in virtue of how what they believe about
each other is that how they treat each other is a reflection of what they believe about each
other. However, the ways in which we can treat each other poorly include not only our
words and our deeds, but also what we believe of each other.
Earlier I had summed up the reasons in favour of locating the wrongs of racist be-
liefs downstream from belief with the following slogan: beliefs don’t wrong people, peo-
ple wrong people. We can still preserve some of the intuitive force of this slogan while
nonetheless recognizing that in the cases given above you are wronged by what someone
believes of you. We must simply recognize that one of the ways in which people wrong
people is by having beliefs about people, not simply what they do or say.
1.3 Locating the Wrongs Upstream From Belief
I have now shown the shortcomings of an account that locates the wrong of racist be-
liefs solely in the downstream effects of the belief. However, those wanting to resist the
thesis that beliefs wrong have the option to locate the wrong elsewhere still, namely up-
stream. In this section, I turn my attention to accounts that attempt to locate the wrong
of racist beliefs upstream from belief. Upstream features of a racist belief include the many
processes that contribute to the agent forming the belief. On such a view, we could re-
sist the thesis that beliefs can wrong, and instead say that what goes wrong in the case
of Spencer is the result of his cognitive predispositions or limitations. Alternatively, per-
haps the ways he gathers and evaluates evidence, i.e., his epistemic practices, are flawed.
Alternatively, still, the flaws lay in problematic motivations and affective attitudes that are
guiding his belief formation. On such an account, beliefs aren’t themselves wrong, rather
they are but a symptom of a set of states and processes that result in the belief. Further,
although this set of states and processes that result in the belief might itself include other
beliefs, the problematic beliefs are merely symptoms of a defect in the cognitive system.
We can illustrate this distinction between locating the wrong downstream from belief and
locating the wrong upstream from belief with the following diagram.
9
9
Thanks to Jessie Munton for first illustrating this distinction in a similar way when presenting com-
ments on this chapter at Athena in Action.
13
Fig. 1.
We can see an example of this upstream strategy in the case of stereotypes. Stereotypes
are helpful in guiding us through the world. As McGarty et al. (2002) note, stereotypes
are not only aids to explanation, they are also energy-saving devices or heuristics. Not
only are stereotypes devices that we form to help us understand the world, but also they
also help us understand the world by saving us time and effort. After all, by treating
people as members of a group rather than individuals, we can save energy by ignoring
the diverse and detailed information about each. However, because we live in a complex
social environment, by taking shortcuts we often adopt biased and erroneous perceptions
of the world. That is, rather than being aids to understanding, a common thought is that
stereotypes are aids to misunderstanding more often than they aid in understanding. So,
although stereotypes can assist with explanation, they problematically produce falsehoods
and distortions due to upstream features of our cognitive system.
10
Further, stereotyping not only problematically produces falsehoods and distortions, it
does so in a way that is so systematic that it cannot be trusted. Sarah-Jane Leslie (2015), for
example, argues that stereotypes are more likely to form about some group if that group
is perceived to share an underlying essence. For Leslie (2015) these “striking-property
generalizations”—e.g., sharks attack bathers, mosquitos carry West Nile—are formed be-
cause they treat a group as having an underlying essence that predisposes them to some
activity or character trait. Further, our cognitive system is such that we more readily
form and endorse these generalizations when they are negative. In sum, the stereotype
itself may be unproblematic. There could, perhaps, be creatures who could unproblem-
atically believe stereotypes concerning groups.
11
But given our own cognitive limitations
and predispositions, stereotypes wrong. Further, it is not the belief in the stereotype that
10
This is a simplified discussion of stereotypes and stereotyping. For a more nuanced and careful discus-
sion, see Beeghly (2015) and Antony (2016).
11
As a note of clarification, the beliefs that are my focus are beliefs about individuals. These beliefs might
be formed on the basis of stereotypes. But, I take a belief in a stereotype to be a belief about a group, not a
belief about an individual.
14
wrongs, rather, the wrong lies upstream from the belief.
For another example, perhaps the belief itself does not wrong but given our cognitive
predispositions and limitations, we can reasonably expect that the belief will interact with
other frame-working beliefs, cognitive biases, etc. in highly systematic and structuring
ways. This kind of wrong is tied closely to the the essential functional role of belief.
Our beliefs prepare us for more beliefs of the same kind. That is, our beliefs guide us
in assessing new evidence as it comes in. This is how priors work. If Spencer has good
reason to think that his priors are going to lead him to update on evidence in morally
problematic ways, then his failure to appreciate this risk is the wrong he commits. What
he believes about black diners in his section is what we could call a risky amplifying belief.
The belief itself is of neutral value, but once it enters a prejudiced cognitive system, it will
have negative effects. Further, the fault lies not with the belief, but rather the prejudiced
cognitive system.
12
You don’t have to be explicitly racist or sexist or any other kind of –ist to see this
amplifying effect. We live in a world that has been shaped by and continues to be shaped
by racist attitudes and racist institutions. So, no matter how well intentioned we are when
we add the belief, we may end up with a racist amplified belief set. To see this, consider the
cognitive saint—someone who doesn’t have any racist background beliefs, assumptions,
or biases—and as such can add the belief this black diner will not tip well or Sanjeev smells
of curry without any untoward consequences, and perhaps ought to. The worry is that
the mere addition of the belief could cause the entire framework to become biased. The
addition of the belief might itself mark the beginning of an implicit association of the sort
“black=poor” or “black=bad”. As IAT studies have shown us (see Banaji and Greenwald
2013), many people in the US have negative associations of this type towards African
Americans. Although, there is still a lot of work to be done to figure out the exact nature
and role that these biases play in the cognitive framework, it is not unreasonable to suspect
that increased exposure to these sorts of beliefs can create piggybacking associations of the
form that I’m concerned about.
13
In light of these examples, we might want to say that what we believe isn’t bad on its
own. But, given the way these cognitive biases work they’re often invisible to us. So, we
have reason to be skeptical about whether we can add the belief unproblematically. So,
we should exercise extreme caution (up to and including avoidance) in adding these sorts
12
Note that a risky amplifying belief could also be a stereotype, but not all risky amplifying beliefs are
stereotypes. I owe the term “risky amplifying belief” to a conversation with Renee Bolinger.
13
See for example Mandelbaum (2016). For some of the controversy surrounding the IAT see Oswald
et al. (2013) Oswald et al. (2015), Greenwald and Banaji (2015), and Madva and Brownstein (2016). Thanks
to Gabbrielle Johnson for many conversations about implicit bias.
15
of things to our own belief set. A problem, however, with this account is that it’s not
clear we can clearly demarcate the wrong upstream or whether the wrong is in what we
might call the whole stream. That is, the wrong lies in the evidence that sparks the belief
and how that belief interacts with upstream features of an agent’s cognitive framework.
A further problem is that one of the reasons that the thesis that beliefs can wrong
seems so unintuitive is because it clashes with the following commonly accepted thesis
about control, responsibility, and belief: moral obligation requires a kind of voluntary
control that we lack with regard to our beliefs. As I’ll show in Section 1.5, although this
assumption is often treated as a settled matter of orthodoxy it is up for debate whether
we have sufficient control over our beliefs (or whether control is even need) for deontic
notions to apply to belief. However, if we take this intuition about control and respon-
sibility seriously, then although it commonly stands in the way of the plausibility of the
thesis that beliefs can wrong, here it actually pushes us towards locating the wrong in
belief rather than in upstream features that causally impact the belief.
Presumably we have less control over the upstream features than what we lack with
regard to our beliefs. As noted earlier, the way in which our cognitive predispositions,
limitations, biases, etc. work are often invisible to us. Believing that Jamal will tip less well
than James may lead us on a path that makes it easier to form negative beliefs about black
individuals we interact with. We have less control over what happens in our cognitive
system after we believe than we do before we settle the belief. The upstream theorist
would say that the bad thing is whatever cognitive framework allows Spencer to (on the
basis of adopting the belief) make a bunch of other terrible decisions. But, the bit that
wrongs is the framework, not the other bad beliefs or decisions that result on the basis of
it. But if the motivation for locating the wrong in the framework rather than the belief
is that beliefs don’t seem like the kind of things that wrong, the same can be said for an
agent’s cognitive framework. How does a biased framework, by itself with no beliefs
entering as inputs, wrong? This upstream account, thus has only a partial explanation of
the wrong committed by Spencer. The wrong is not only in his cognitive limitations and
predispositions, but it is also in what he believes.
14
14
For more on locating the wrong of racist and prejudiced beliefs in upstream features of an agent’s cog-
nitive architecture see Begby (2013) and Munton (MS). Begby argues that the reason why prejudices are
epistemically insidious concerns how they become interalized as background beliefs that are recalcitrant to
empirical counterevidence. Munton argues that the flaw in well-founded statistically-accurate beliefs about
certain demographics concerns how we tacitly ascribe the domain of the statistic in question. That is, there
are often flaws in our implicit representation of the generality of statistics concerning the rates of violent
crime, etc. As noted, I am in agreement that these accounts all get something right, they simply lack the re-
sources to address the character of Spencer, and the peculiarities of the character of Spencer are my primary
concern in this chapter and this dissertation as a whole.
16
1.4 A Restricted Account of When Beliefs Can Wrong
According to a standard way of thinking about the moral wrong of racism, when the
wrong is located in belief there are two conditions that the wronging belief must satisfy.
The first is the false condition—the belief is false—and the second is the hot condition—
the false belief is accompanied by other negative qualities that the belief exhibits, e.g.
being held in bad faith, being self-deceptive, being indicative of ill-will on the part of the
believer etc. We can already see how such an account will not be able to explain what
is morally objectionable about the supposedly rational racist. Although Spencer’s belief
seems objectionable, it is not false. Spencer’s belief appears to be an accurate reflection of
the weight of the evidence. Further, Spencer claims to harbour no feelings of ill will.
I grant that in the vast majority of cases people who claim to be like Spencer satisfy
either the false condition or the hot condition. The problem, then, with the accounts
discussed in this section is not that they are incorrect. Rather, these accounts are incom-
plete. They foreclose the possibility of a character like the supposedly rational racist. But,
as I argued in Section 1.1.2, there is good reason to think that there could be people like
Spencer. Thus, my goal in this section is not to argue against these accounts. Rather, my
goal is to show the limitations of these accounts when they are faced with the particular
challenges raised by a character like Spencer.
Let us begin with a prominent account of the wrong of racist beliefs from Kwame An-
thony Appiah. Appiah (1990, 1995) argues that racism consists of two elements: racial-
ism and racial prejudice. Racialism is a proposition about the existence of different races
or racial essences. Racial prejudice is a disposition about the moral significance of a race
because of some morally relevant racial properties. For example, one might think that
some races are more primitive or barbaric than others. The belief rests on a false belief
about the existence of different races or racial essences, and it also involves thinking that
there are morally relevant racial properties, e.g. being more primitive or barbaric. Ac-
cording to this account, racial prejudice requires accepting a racialist belief, so one cannot
have racial prejudice without accepting racialist beliefs.
15
The wrong of racism, then, is
that it places unjustified moral significance on race. That is, the racist uses race as a basis
for making moral distinctions, e.g., the racist uses race illegitimately as the sole basis for
15
Although Appiah himself does not think that these racial essences are biological, this account does
seem to paint the racist as a biological essentialist. But, that is not necessary. For example, imagine a
racist who doesn’t believe that the differences are biological, merely cultural. For example, the racist who
refuses to hire a Korean dogwalker because of the false belief that Koreans eat dogs. Again, we have a racial
proposition Koreans eat dogs, which is false, and the disposition to treat their race (or culture) as being
of moral significance because it indicates some morally relevant racial property. Alternatively, a clearer
example of the cultural racist can be seen in cases involving resentment towards new immigrants from a
particular country from the descendants of immigrants from that same country.
17
treating some people morally. Further, akin to the upstream accounts discussed in Sec-
tion 1.3, part of the wrong of racism comes from the racist disposition that imbues in the
believer a tendency to accept, morally and theoretically, false propositions or beliefs that
are hurtful, discriminatory, bigoted, etc. towards one race and not the other. This moral
wrong leads to other bad epistemic practices such as accepting racist beliefs in the face of
overwhelming evidence that should have led to giving up on those beliefs. In this regard,
the wrong of racism is located in the beliefs of the agent, not their actions. Racism, thus,
is a cognitive incapacity (see Mills 1997).
Similarly, Lewis Gordon (1995, 2000) argues that racism is a character flaw, but specif-
ically it is a flaw in one’s beliefs. Gordon’s account, then, is a bridge from the purely
cognitive incapacity account of racism—where at its core the wrong of racism lies in a
false belief and how that belief interacts with other beliefs—to an account that encom-
passes the other attitudes of the agent. For Gordon, racism is a character flaw because it is
a reflection of one’s defective psychology, personality, or character. After all, if it were a
mere matter of a false belief, then we wouldn’t be able to distinguish the merely ignorant
racist who does not deserve moral approbation from the racist that does. Racism must be
accompanied by some negative non-cognitive attitudes. This leads us to J. L. A. Garcia
(1996, 1997, 1999)’s suggestion that racism is a personal phenomenon of ill-will. This ill-
will is often manifested in actions and perhaps helps create and sustain social institutions
that contribute to the wrong of racism. The wrong of racism, thus, is this racial disregard.
Here, we also hear echoes of Arpaly and Schroeder (2014)’s conception of the wrong of
racism: those who have prejudices display some distinctive kind of epistemic irrational-
ity that can be explained by a deficiency of good will. That is, they are ready to believe
anything about a particular group as long as it is bad. The racist is guilty of a kind of hot
irrationality.
To summarize, according to accounts that locate the wrong of racism in what an agent
believes, there are two distinct but closely related suggestions. First, the wrong is that the
belief is false. Second, the wrong is not a mere false belief, but rather the false belief
accompanied by ill-will, hatred, antipathy, etc. I call the first the false condition, and the
second the hot condition. We have already seen that Spencer is not guilty of the false
condition. We have been granting that Spencer’s belief is true. We have been granting
that it seems to be rational for Spencer to believe that the black diner in his section will
tip less well than the white diner in his section. Perhaps, however, Spencer is guilty of
the hot condition. Perhaps his belief is epistemically irrational because it is the result of
negative non-cognitive attitudes or ill-will. It is this diagnosis that I turn to considering
now.
18
Following Nomy Arpaly (2004, pp. 102), let us start with the following case.
Boko Fittleworth (a character in a P.G. Wodehouse novel) overpowers and
traps a man whom he spots hiding in his would-be-father-in-law’s garden shed
at midnight, because he believes this man to be a burglar. In fact, the man is
not a burglar but a business tycoon whose presence in the shed is part of
a secret, unlikely, and harmless plot in which the future in-law is a willing
participant.
In most circumstances, if all that a person gets wrong are the facts, then her false beliefs—
i.e. her ignorance—should excuse her action. Although Boko owes the businessman an
apology, he is not blameworthy for his action because he has the excuse “But I thought
you were a burglar.” The racist, however, does not presumably have a similar excuse.
Consider, for example, the anti-semite who has the false belief that Jews are members
of a worldwide conspiracy set on world domination. Arpaly argues that the anti-semite
does not have the analogous excuse “But I thought you were a member of worldwide
conspiracy set on world domination.” This is an asymmetry in need of explanation. That
is, we need to explain why some false beliefs excuse actions, whereas other false beliefs do
not.
Arpaly argues that the answer has to do with a difference in the epistemic rationality
of Boko’s belief as opposed to the epistemic irrationality exhibited by anyone’s belief, in
this day and age, that Jews are part of worldwide conspiracy set on world domination.
Whereas it is epistemically rational to believe of a man hiding in your shed at night and
who refuses to identify himself that he is up to no good, she argues that unless you have
just arrived on Earth from another planet with a seriously flawed travel guide, it is diffi-
cult to reach the belief that all Jews are involved in a worldwide conspiracy set on world
domination. Whereas the first is an honest mistake, it is difficult for the latter belief to be
an honest mistake. This is because many of the people who hold such beliefs have met
Jewish people, know of historical events motivated by such beliefs, and are able to see
the unlikely nature of such a conspiracy. So, rather than being an honest mistake, their
beliefs seem to be the result of motivated irrationality.
16
16
She reiterates this point in Arpaly and Schroeder (2014, pp. 234) when she asks, so, what’s the difference
between the ordinary racist, i.e. the one who deserves condemnation, and the alien racist who does not?
One difference that springs to mind is the fact that the Earthly racist holds his belief in, e.g.
the supremacy of the majority Chinese ethnicity against plentiful evidence that is readily
available to him. With a typical level of intelligence and with the information that is available
to rather uneducated people in the first world, he would probably not have developed his
racist beliefs if there were not something amiss with him epistemically. The run-of-the-mill
racist is epistemically irrational, as are other run-of-the-mill prejudiced people.
19
This, however, is still not enough. Although some irrational beliefs are morally objec-
tionable, Arpaly recognizes that this does not hold in general. She argues that the person
who irrationally believes of their lottery ticket that it will win is not being morally vicious
in the same way as the anti-semite.
17
If irrationality in a belief is not sufficient for moral
condemnation, what is? Arpaly and Schroeder (2014, pp. 235) argue that the difference is
that the bigot’s irrationality is “hot” irrationality.
We agree with Kwame Anthony Appiah that the racist’s irrationality is mo-
tivated irrationality. His thinking that the Jews are a conspiratorial people is
more likely to be caused by his hatred of them, or of people who are ostenta-
tiously “other,” or to be caused by his resentment of his low social station, or
the like. And the fact that the racist’s belief appears to be caused by a desire
opens the possibility that, however involuntary, the belief might have some-
thing to do with ill will or moral indifference.
According to this account, prejudiced beliefs are morally vicious when they are held as
a result of moral indifference. Returning to our challenger, Spencer, he might not think
that he is a racist or a bigot because he is just believing in accordance with the evidence,
but that is not why he believes. If that is what he reports, he is guilty of bad faith or
motivated irrationality. He is just ready to believe anything of a particular group as long
as it is bad.
This is a strong claim to make about the psychology of such persons. It puts us in a
difficult position when arguing with the supposedly rational racist. Firstly, it makes for an
uncharitable exchange. Second, when they inevitably deny the description, we’re led to a
sort of foot-stomping on which neither can convince the other. We would be in a stronger
dialectical position if we were to grant that they are not believing in bad faith, or due to
motivated irrationality, and still point out how their belief wrongs. This, recall, is the
motivation for the working constraint on acceptable answers to the supposedly rational
racist, i.e., that we grant Spencer his conception of himself.
Not only would we be in a dialectically advantageous position in response to the sup-
posedly rational racist if we meet this constraint, but also there is another problem with re-
stricting the class of beliefs that can wrong to beliefs that exhibit hot irrationality. An over
reliance on the hot condition can obscure other cases of beliefs that wrong. To show this,
I demonstrate that there can be an agent with a good will and who has well-intentioned
beliefs who nevertheless makes a distinctive kind of mistake when settling a belief, which
17
Of course, we could add more to the case in which they would be morally vicious.
20
doesn’t seem excused and is morally objectionable. For such a case, consider the following
scenario.
The Questioning and Unquestioning Scientists. Sierra and Tango are conduct-
ing a study to see if there is a relationship between race and IQ. According to
their results, not only is there a 15-point difference between the average IQs
of white and black test takers, but IQ is 60% inheritable within the White
population. Tango doesn’t like the conclusion, but as a scientist he resolves
that facts are facts, and he thus unquestionably accepts the conclusion. Sierra
similarly does not like this result, but on the other hand is willing to question
the result. So, she does some extra research and notices another hypothesis
that they did not consider: that the environmental conditions could account
(and do a better job accounting) for most of the disparity rather than genetic
factors.
18
Is the badness of Tango’s belief due to a deficiency of good will? Arpaly and Schroeder
(2014) would say that it is, but I do not find that a convincing gloss on this scenario. It
does appear in this case that it is Sierra who is the one with scientific integrity, whereas
Tango’s acceptance of the result may be due to his prior negative beliefs about IQ in the
black population. However, if we alter the scenario slightly, we’ll see why that cannot be
the whole story.
Let’s now set the example in the heyday of scientific racism. That is, scientific exper-
iments are being run in order to justify the racist practices of a society. Now, suppose
that the result that Tango* and Sierra* arrive at is the correct result—that there is no re-
lationship between race and IQ. Now, again, Tango* doesn’t like the conclusion, but as
a scientist he resolves that facts are facts and he accepts the conclusion. Sierra* also does
not like the result, but she is willing to question it. So, she does some extra research in
order to find a relationship between race and IQ. In this case it is Tango* who exhibits
scientific integrity and it is Sierra* that is engaged in motivated reasoning.
19
I present this pair of cases to illustrate how problematic it is to assume that Tango’s ac-
ceptance of the results is due to bad faith or motivated irrationality. This is also the point
that Spencer is trying to make when he insists that he is not reasoning in bad faith nor
18
This case tracks the debate between Herrnstein and Murray (1996) and Block (1996).
19
Further, as I will discuss in Chapter 3, if the amount of evidential support is the crucial factor to deter-
mining whether a belief is justified, we cannot distinguish Sierra and Sierra*. The difference between them
is that one has an impermissibly low threshold for justification, one that is not matched to the moral stakes
of her environment. To explain this, I will argue that we must reject another standard piece of orthodoxy:
that sufficiency of evidence supervenes on what your evidence is. As I argue in Chapter 3, the moral stakes
can affect whether you have sufficient evidence and whether a belief is justified.
21
does he have any ill will towards black diners. It is unfair to assume of the supposedly ra-
tional racist that he believes what he does because he is ready to believe anything negative
about some group, rather than he believes what he does because it is what the evidence
suggests. What the case of Tango* shows is why it’s wrong to believe of Tango that he
only accepts the result due to his prior negative beliefs about IQ in the black population.
The supposedly rational racist sees himself as Tango (and in turn as Tango*). He is will-
ing to accept unpopular conclusions if that’s what the evidence shows. When it comes to
responding to the evidence, Tango and Tango* do better than Sierra and Sierra*. In con-
trast, our moral intuitions suggest that Sierra is better than Tango, but also that Tango*
is better than Sierra*.
Further, it is now well-documented that although it is tempting to think that the racist
suffers from ill will or a deficiency of good will, racial injustice can survive and even thrive
in the absence of such negative non-cognitive attitudes. Racism can come in cold varieties
as well as hot varieties. So, the hot condition is not necessary for identifying the moral
wrong of racism.
20
This is why it is not only a dialectical advantage if we can provide an
account that doesn’t undermine Spencer’s conception of himself, this really ought be a
constraint on any acceptable answer to what Spencer does wrong when he believes what
he believes.
1.5 Another Hurdle: A Worry About Control
My aim in this chapter has been to present a case—the case of the supposedly rational
racist—that pushes us to seriously rethink why we reject the naïve thesis that beliefs can
wrong. I’ve argued that this case forces us to reconsider why we are so resistant to the
idea that I could wrong you by believing something of you that I seem to be justified in
believing. The challenge posed by the supposedly rational racist is a challenge concerning
our epistemic practices in a non-ideal world. A common refrain throughout this chapter
has been the following: the world is an unjust place and there may be many morally
objectionable beliefs that it justifies. As a result, the evidence might be stacked in favor of
racist beliefs. The world we inhabit is a racist one, so it is no surprise then that some of
our beliefs are racist as well.
If our best accounts of the way racist beliefs can wrong require the believer to always
act on the belief before we call the belief wrong or require the believer to be distorting
the truth, misrepresenting reality, organizing facts in a misleading way, or engaging in
20
See for example Valian (2005)’s account of the gender disparity in philosophy, which she calls a “cold”
social-cognitive account given that "it is purely cognitive rather than emotional or motivational" (pp. 1).
22
fallacious reasoning, then these accounts will fail to meet the challenge posed by the sup-
posedly rational racist. Spencer appears to be responding to the evidence in a way that
seems epistemically rational. This is why I suggest that we stop looking elsewhere for
what wrong Spencer commits. The question I am interested in is what happens when we
take seriously the thought that Spencer wrongs the black diner in his section in virtue
of what he believes of the diner. This is the question that will motivate the rest of this
dissertation.
There is, however, another hurdle to clear to establish the plausibility of the thesis that
beliefs can wrong. This hurdle is the assumption that moral obligation requires a kind of
voluntary control that we lack with regard to our beliefs. As we have seen, this back-
ground assumption is the motivating force behind the attempts to locate the wrong com-
mitted by Spencer either in the downstream effects of the belief or somewhere causally
upstream from the belief itself. My argument against this assumption is going to be short
and indirect. As I shall show, there is more intuitive plausibility to the thought that the
supposedly rational racist wrongs given what he believes than there is to the thought that
voluntary control is required to wrong.
The problem of control can be presented as follows. Although we may talk in every-
day discourse about what people should believe, deontic notions such as ought and should
and moral notions such as praise and blame do not apply in the domain of epistemology.
The argument goes as follows.
1. For deontological concepts such as obligation or duty to apply in the domain of
epistemology, agents need to have voluntary control over their beliefs.
2. It is not the case that we have sufficient voluntary control over our beliefs for deon-
tological concepts such as obligation or duty to apply in the domain of epistemology
3. Therefore, deontological concepts such as obligation or duty do not apply in the
domain of epistemology.
To briefly canvas various arguments against the idea that we either lack control or that
we even need control for these deontic notions to apply, Nishi Shah (2002) rejects (2) on
the grounds that we exert control over our beliefs in the following way: through our
appreciation of the evidence we are agents with regard to our beliefs and are capable of
regulating our beliefs. Matthias Steup (2000) similarly argues that our beliefs are respon-
sive to epistemic reasons and in this way we exercise control over our beliefs. Further,
Brian Weatherson (2008) has persuasively argued that this argument relies on too narrow
a conception of voluntary, and Amy Flowerree (2016) presents the following reductio to
23
establish that we are agents with respect to our beliefs: if we are not agents with respect
to our beliefs, then we are not agents with respect to our intentions. Others have also
argued that (1) is false, i.e. for deontological concepts to apply, control is not needed. For
example, Pamela Hieronymi (2006, 2008) argues that it is not troubling to say that we can
no more intend at will than believe at will. She further rejects that voluntary control is
required for deontic notions to apply. Connor McHugh (2012) similarly argues that vol-
untariness is not a central condition. In general, these strategies have taken the following
form: they show that for deontological concepts to apply in the domain of epistemology,
either we have sufficient control or there is no control needed for deontological concepts
to apply.
21
What we are left with is the following choice. Should we give Spencer a pass on the ba-
sis of a theoretical technicality that is both (a) not as intuitively obvious as it is commonly
regarded, and (b) does a disservice to the common folk-intuition that something has gone
wrong in the case of Spencer? Our choice is between saying that Spencer is not wronging
at all or that Spencer wrongs through what he believes. The main consideration prevent-
ing us from saying the latter is this theoretical normative posit that voluntary control is
always required to wrong. However, this intuition is more esoteric and has less claim to
be either the intuition of the folk or share widespread acceptance amongst philosophers
than the intuition that I have been defending: that Spencer does something wrong.
To conclude, I hope to have convinced the reader that we should take seriously the
thought that people wrong people in virtue of what they believe about each other, and
not just in virtue of what they do. To further show how beliefs can wrong, is the next
task, and to answer that we turn to the next chapter.
21
In Basu and Schroeder (2018), we also consider another way of resisting the siren’s call that is the prob-
lem of control. We reject the following underlying assumption that motivates the problem of control: the
thesis that our beliefs are at the mercy of the evidence. I will also explore this is in more detail in Chapter
4.
24
CHAPTER II
What We Epistemically Owe To Each Other
At the end of the previous chapter we ended on the following note: Spencer does
something wrong and that whatever that wrong is, that wrong is properly located in what
he believes of the diner in his section. The task of this chapter is explain what it is that
Spencer does that is wrong. That is, to explain how, why, and when beliefs—beliefs of the
sort Spencer holds of the diner in his section, i.e., beliefs about another person–wrong.
Notice that I have not said that the task is to explain what beliefs wrong. As I shall show,
the wronging is not a matter of the content of the belief, but rather concerns a morally
problematic way of relating to others through what we believe of them. As a result, the
title of this dissertation is misleading. You might expect that a dissertation titled “Beliefs
That Wrong” would tell you which beliefs wrong. Instead, as we’ll see over the course of
this chapter, a better name would’ve been “Doxastic Morality”. My goal in this chapter
is to show how moral demands that are constitutive of the relationships we stand in with
one another give rise to moral demands not only on our words and deeds, but also our
beliefs.
In addition, another goal of this chapter is to show that it is not only racist beliefs that
can wrong in this way, but that many of the beliefs that we form about one another can
wrong. The particular beliefs held by the supposedly rational racist that were the focus
of the previous chapter are merely an instantiation of a wider class of beliefs that wrong.
As I shall show, when we relate to others and form beliefs about them in the same way we
relate to objects and causal phenomenon and form beliefs about tables and the weather, we
fail to give others what they are owed: recognizing them as agents rather than objects. In
general, we wish to be seen as individuals, not as objects to be predicted and managed. This
failure to relate as we ought is a morally criticizable failure that stems from a kind of moral
indifference, a kind of moral disregard, for which we can be held accountable. Already
you may object that this thesis seems to make an implausibly wide range of things morally
problematic. So, another task of this chapter is to show that it’s not an implausibly wide
25
range of things that are morally problematic, rather, it’s perfectly plausible that the wide
range of things this chapter discusses are in fact morally problematic. Thus, the three tasks
of this chapter are (i) give an explanation of why Spencer’s beliefs wrongs, (ii) show that
Spencer’s moral failing is an instantiation of a more general way in which we can wrong
others through what we believe, and (iii) address the worry that this proposal makes a
wide range of things morally problematic.
Finally, in way of set-up, I should note that this chapter will be primarily case driven.
In Section 2.1 I will present a number of cases that suggest that there is an overlooked
aspect—the cognitive or epistemic aspect—of the moral demand we place on one another
to be treated well. These cases—Mistaken Identity, Wounded By Belief, The Racist Hermit,
The Security Guard, and Sherlock Holmes—all suggest that there is a moral demand, a de-
mand to be related to in a way that is characteristic of being an agent rather than an object,
and this moral demand has an epistemic dimension. We fail to relate as we ought when
we try to make predictions of other agents as though they are objects whose acts are de-
termined by causal laws, e.g., beliefs about someone’s status as a diner or a waiter on the
basis of their skin color. In Section 2.2, I turn to presenting a picture of morality that
can explain how moral considerations matter to the question of how we settle belief. I
will argue that in virtue of our relationships and in virtue of moral features of our en-
vironment, e.g., structural injustice, we fail in what we epistemically owe to each other
when we observe and form beliefs about other people in the same way we observe and
form beliefs about planets. In Section 2.3, I note that although there is this general moral
concern we all have in virtue of being agents—the wish to be seen as individuals, not as
objects to be predicted, anymore than managed—for some groups of people, in particular
marginalized groups, this epistemic-moral demand is particularly strong.
Finally, in Section 2.4 I attempt to provide some guidelines for recognizing the condi-
tions under which there is a greater risk that we will fall short of our epistemic obligations
to one another. That is, I attempt to articulate some general principles for when beliefs
wrong. As we’ll see, however, this is no easy task. The diversity of cases that push us
seriously considering the thesis that beliefs can wrong also makes it difficult to articulate
any principle (or set of principles) that can provide necessary and sufficient conditions
for identifying when beliefs wrong. To avoid wronging an agent must not only be aware
of certain moral facts, but also many contingent social, structural, and historical facts as
well that determine when those moral considerations are applicable. In short, the agent
must exhibit a moral-epistemic virtue that is akin to being woke. The account of moral
encroachment to be defended in Chapter 3 will be an attempt to further flesh out what
this moral-epistemic virtue looks like. But first, let us begin by considering some cases.
26
2.1 Examining Cases
2.1.1 Mistaken Identity
Mistaken Identity. The conference has ended, and the organizers have had the
forethought to book a number of tables at a nearby restaurant so that con-
versation can continue over dinner. You’re having a good time at dinner and,
after a few drinks, you get up to use the restroom. As you return to your
table, one of the diners, Jim, reaches out to grab your arm and says, “I asked
for a refill fifteen minutes ago.” For a moment you’re confused, then it dawns
on both of you what mistake has been made. Most philosophers don’t look
like you. With regard to melanin levels, you share more in common with
the wait staff than your fellow diners. Given your skin colour, the likelihood
that you are a member of the staff rather than a fellow diner was high enough
to seemingly make it rational for Jim to assume that you were a waiter, not a
fellow diner. The belief that Jim had—and in turn his actions—might amount
to a social faux pas, but, given that the belief was well-supported by the ev-
idence, the belief—and in turn his act—was reasonable. He’s not a bad guy;
he just made an honest mistake. Of course, in the moment you don’t reason
through all of that. In the moment, you quickly laugh it off, eager to return
to your table and dinner conversation.
Later, you notice that you can’t stop thinking about that interaction, and you wonder
why such a small thing is still bothering you. It’s not just the act that’s bothering you:
something about your colleague’s belief that you were a waiter is keeping you up at night.
So, you lay in your bed and you run the incident over and over again in your head. You
think to yourself, well, some epistemic mistake was made—the belief was false—the belief
was well-supported by the evidence. Presumably, given that the belief was well-supported
by the evidence, the belief was also rationally permissible. If all of the of dinner guests,
except you, are white whereas all the staff members share your skin complexion, then
based on race alone, it seems that anyone in the position of Jim—needing to flag down a
waiter for their drink order—should similarly believe that you are a staff member. That
is the doxastic attitude that is seemingly supported by the evidence.
But, you also think to yourself, isn’t it also the case that mistaking a person of color
for a member of the staff is one of the paradigmatic examples of racism? In Slumdog
Millionaire, Jamal is mistaken for a tour guide while he himself waits for a tour of the Taj
Mahal. In Miss Saigon, Kim is mistaken for a maid when she enters a hotel room. Barack
Obama has often been asked to get coffee even when dressed in a tuxedo at a black-tie
27
dinner.
1
The examples are numerous. Given how strong the association is between non-
white and staff, it’s not surprising that when a young Whoopi Goldberg saw Lt. Uhura—
the communications officer on Star Trek played by Nichelle Nichols—she ran around the
house screaming “there’s a black lady on television and she ain’t no maid!”
2
In light of
this, let us reconsider the opening case.
Returning to our dinner companion, we can imagine Jim offering the following apol-
ogy: “My evidence supported that you were a waiter, not a conference participant, so
I had the attitude I epistemically ought to have had towards you being a staff member
as supported by my evidence." Further, he might add, “When I believed you were a staff
member I did not actively feel any ill will towards you, nor exhibit indifferent disregard.”
3
This excuse, however, does not assuage the feeling of resentment one bit. You are still mad
at him. And it is not just his action—flagging you down as a waiter— that you are mad
at, it’s that he believed that you were a waiter. Imagine if things had gone differently, and
Jim still had the belief but didn’t act on that belief. After all, a common response to these
cases is that Jim should believe, but also that he should not act on his belief. Even in such a
case, there seems to be something upsetting about the way in which he so readily believed
that you are a waiter. Consider an analogous case. Suppose, for example, someone saying
to you that they believe the stereotype that women are bad drivers, but that they would
never act on that stereotype. That does not make the fact that they believe that women are
bad drivers any less insulting. The belief remains insulting even when the communicative
act is removed.
It would be all too easy to dismiss these feelings as irrational. I will use this case,
and others, to argue that it is not irrational to be upset. Jim has done something wrong,
and that one has these feelings of hurt, anger, indignation, and resentment when these
events happen should be taken seriously. I will argue that we should take these feelings
seriously, and my argument follows the Strawsonian thought that our reactive attitudes
such as resentment are often good indicators that some moral demand hasn’t been met
and the Lordeian thought that our feelings are a source of evidence and information.
4
1
As Obama notes, “there’s no black male [his] age, who’s a professional, who hasn’t come out
of a restaurant and is waiting for their car and somebody didn’t hand them their car keys.” See:
http://people.com/celebrity/the-obamas-how-we-deal-with-our-own-racist-experiences/
2
https://www.yahoo.com/tv/blogs/tv-news/exclusive-video-whoopi-goldberg-star-trek-
213710247.html
3
Regular folk don’t talk in this way, but remember, we’re imagining this case taking place at a post-
conference dinner.
4
There is already a lot of ground to cover in this chapter, so to restrict it’s scope, let me note that I will just
be assuming that we can rely on our feelings in this way. One might object that anger is not the appropriate
response because anger is not virtuous. It might be reasonable or excused to feel anger, after all, it can be
morally exhausting to be magnanimous or forgiving or to laugh off the fifth time you’ve been pulled over
28
2.1.2 Wounded By Belief
Wounded By Belief. Suppose that Mark has an alcohol problem and has been
sober for eight months. Tonight there’s a departmental colloquium for a vis-
iting speaker, and throughout the reception, he withstands the temptation to
have a drink. But, when he gets home his partner, Maria, smells the wine that
the speaker spilled on his sleeve, and Mark can tell from the way Maria looks
at him that she thinks he’s fallen off the wagon. Although the evidence sug-
gests that Mark has fallen off the wagon, would it be unreasonable for Mark
to seek an apology for what Maria believes of him? He is, after all, wounded
by what his partner believes of him.
5
I contend that cases like Wounded By Belief are common place and such cases point us to
a folk practice of holding people accountable for their beliefs, for thinking that it is the
beliefs themselves that wrong, and for being upset with others for what they believe about
us. Consider, for example, how common it is to say the following, “You shouldn’t have
believed that of me."
One may try to respond by objecting that Mark is not wronged until he has clarified
that he hasn’t had a drink. Perhaps Mark has only been wronged if after he has clarified
that he hasn’t had a drink Maria continues to believe that he has had a drink. Further,
the reason why Mark deserves an apology is because of Maria’s act of discounting his
testimony, not merely her having the belief. I grant that this could be a partial explanation
of the wrong. However, we can imagine that even if Maria does not discount Mark’s
testimony, a wrong still occurs. Mark should be able to demand that his partner think
better of him and not immediately settle on the belief that he must’ve had a drink—even
when the evidence strongly suggests that he has. He would be further hurt if she continues
to believe that he’s had a drink after he explains that he hasn’t, but there is still the initial
hurt before any exchange occurs.
As I will argue in Section 2.2, Mark has been wronged because his partner has taken
an agent-neutral stance towards forming beliefs about him. Further, that some people
do weaponize and abuse this move in relationships—demanding that their partner be-
lieve better of them, disregard or discount evidence, etc.— provides additional support
by police or mistaken for a staff member. Nussbaum (2016) puts forward such an argument. But, as her
critics note, anger plays an important role as a response to injustice in the world and provides importance
information and evidence about injustice (see Bell 2009, Srinivasan 2016, 2017, Bommarito 2017, and of
course Lorde 1984). For the purposes of this chapter I will just follow the latter’s assumptions about the
role of emotions in guiding inquiry into injustice.
5
See Basu and Schroeder (2018).
29
for thinking that they are only able to effectively to do so if it is in fact a practice that we
expect within a relationship.
6
One might further object to this case that what we are reacting to in this case is not
what Maria believes of Mark, rather, we are reacting to a violation of trust. Through her
belief Maria reveals that she doesn’t trust Mark to not have had a drink. What Mark then
seeks an apology for, what he wants to be corrected, is that trust that Maria has violated.
That is where the wrong is located; the wrong is not in the belief, the wrong is in the
violation of trust.
7
I grant that the belief may be an indication of a lack of trust, and it is
that lack of trust that Mark, and in turn if the case were about us, we would be wounded
by. But, it does not follow from that that we are not also wounded by the belief. Here is
an analogy to press this point. A slap in the face may be indicative of someone’s ill will
towards you. One thing we desire in our relationships with others is that they not feel
that ill will towards us. Nonetheless, we are wounded by the slap, and can rightly seek an
apology for the slap. Similarly, the belief may be indicative of a lack of trust, and what
we want is that there be trust in our relationship. Nonetheless, we are wounded by the
belief, and can rightly seek an apology for the belief.
I grant that one may remain unconvinced. But to see how deep this folk practice goes,
let us consider more cases.
2.1.3 The Racist Hermit and The Security Guard
Previously, in Chapter 1, I presented the following two cases: The Racist Hermit and
The Security Guard. Whereas in both Mistaken Identity and Wounded By Belief the belief
in question seems epistemically justified but false, in these two cases the beliefs are not
epistemically justified but in one case the belief is true and the other the belief is false.
Nonetheless, in both cases the beliefs seem to wrong. In Section 2.1.4 we will consider
a case that parallels that of Spencer, the supposedly rational racist: a case in which the
belief seems epistemically justified and the belief is true yet the belief seems to wrong.
The purpose of these cases is to illustrate the diversity of circumstances in which we seem
to wrong others given what we believe. Sometimes we wrong when the belief is false,
sometimes when it’s true, sometimes when we’re justified and sometimes when we’re
not. The task after these cases are presented is to then determine what is common to all
the cases. But first, let us consider The Racist Hermit and The Security Guard.
6
Thanks to Elle Benjamin for pressing me on this point. See also Stroud (2006)’s related work on the
moral demand on friends that they believe better of their friends. I will discuss Stroud’s work in more detail
in section Section 2.2.2.
7
Thanks to Laura Gillespie for pressing me on this point.
30
The Racist Hermit. The racist hermit in the woods will never interact with
the disadvantaged person he believes something negative of, he will never in-
teract with or contribute to the institutional structures of racism. He may
be a product of these institutional structures, but we are hard pressed to say
that he contributes to them given his isolation. But, suppose that he comes
to believe that Sanjeev smells like curry. How did he come to form this be-
lief given his isolation from society? Let us just suppose that he discovered
some trash on the ground which happened to be an alumni newsletter from
Sanjeev’s university that included a picture of him. Upon seeing that picture,
the hermit believes of the pictured person—Sanjeev—that he smells of curry.
Now, suppose also that Sanjeev happens to have recently made curry so in
this instance the hermit’s belief is true—Sanjeev does smell of curry. Has the
hermit wronged Sanjeev?
In this case the racist hermit’s belief is true, and it in no way affects Sanjeev. Nonetheless,
I suggest that we have an intuition here that the hermit is still doing something wrong.
If you think that only actions or words said to a person can wrong, then in this case
you’ll have to say that the hermit does not (nor cannot) wrong Sanjeev because Sanjeev
will never learn of the hermit’s belief. But, as I noted earlier, when it comes to relational
harms neither harm nor knowledge of the harm or wrong are required to be wronged.
For example, if Sanjeev’s partner were to cheat on him but he never found out, it is a
commonly accepted intuition that Sanjeev would still have been wronged. For another
case to continue to warm the reader up to the idea that we really care what people believe
of us and can be wronged by what they believe of us, consider the following case.
The Security Guard. Jake is a security guard at a fancy department store. He
hates the company he works for, and he couldn’t care less if people shoplift
and cost the company money. One day, as Jada leaves the department store
Jake works at, he believes that Jada shoplifted the purse she’s carrying. But,
given his contempt for his company, he chooses not to intervene. He acts
exactly as he would act if he had believed that that was Jada’s purse and that
she hadn’t stolen it. Has Jada been wronged by what Jake believes of her?
I want to press the intuition that Jada has been wronged by what Jake believes of her. If
we put ourselves in Jada’s shoes, we don’t want people to believe of us that we stole the
purse. We care what people believe about us. The ways in which we can treat each other
poorly include not only our words and our deeds, but also what we believe of each other.
31
Finally, before one dies of exhaustion from cases, one may be unconvinced by Wounded
By Belief because you might think that if Mark had been drinking, he wouldn’t have been
wronged by Maria’s belief. Hence, what is doing the work in getting the intuition that
Mark has been wronged is that Maria falsely believes that Mark has been drinking. Note,
however, that in The Racist Hermit the belief is true, nonetheless there is the intuition that
Sanjeev has been wronged. To further press this point, consider a case that is analgous to
the motivating case of the dissertation, i.e., Spencer, the supposedly rational racist. Recall
that Spencer’s belief about the diner in his section seems to be both epistemically justified
(and we can assume for this purpose) a true belief. To see an analogous case, let us consider
Sherlock Holmes.
2.1.4 Why Everyone Hates Sherlock Holmes
Within the fiction people are constantly annoyed with and get mad at Sherlock Holmes
for various reasons. Of particular interest to this chapter, Sherlock Holmes is an excellent
example of someone who engages in this kind of morally objectionable statistical reason-
ing that we find in the previous cases. Further, he engages in this kind of reasoning with
regard to every person he meets. Further, without fail, the people he encounters find the
beliefs he forms about them to be insulting. Sometimes, for one in Victorian England,
it’s because it is a negative belief, e.g., that someone is sexually promiscuous on the basis
of mannerisms that increase the likelihood that they are sexually promiscuous. But other
times, people also respond negatively to his belief about what they ate for breakfast. What
remains in common is the manner in which Sherlock Holmes forms beliefs about people.
He observes everyone as objects to be studied, predicted, and managed. I suggest that
the negative way in which people respond to what Sherlock Holmes believes about them,
whether it concerns their culpability in a crime or what shoe they put on first when get-
ting dressed in the morning, can be explained by feeling wronged when people look at us
the way that Sherlock Holmes looks at us.
As noted, within the fiction people are constantly annoyed with and get mad at Sher-
lock Holmes for various reasons. Of particular interest for this chapter and this disserta-
tion, often the annoyance stems from some yet-unarticulated moral demand with regard
to forming beliefs about others that Sherlock Holmes consistently fails to meet. Sherlock
Holmes, as a character, presents us with a really suggestive case in support of the claim
there’s a way of viewing people—observing people in the same manner we observe natural
phenomena like planets, the weather, etc.—that wrongs the people viewed in such a way.
Consider, for example, for the following scene from the 1985 movie, The Young Sherlock
Holmes.
32
SH: Wait–let me. Your name is James Watson. You’re from the North, your
father’s a doctor, you spend much time writing, and you’re fond of custard
tarts. Am I correct?
JW: My name isn’t James, it’s John.
[...]
SH: Very well, so your name is John. How did I do on the others?
JW: You were correct. On every count. How is it done? Is it some sort of
magic trick?
SH: No magic, Watson. Pure and simple deduction. The name-tag on your
mattress reads “J Watson”. I selected the most common name with “J”. “John”
was my second choice. Your shoes aren’t made in the city. I’ve seen them be-
fore when visiting the north of England. Your left middle finger has a callus,
the trademark of a writer. You were carrying a medical book not available to
the general public, only to physicians. Since you can’t have been to medical
school, it was given to you by an older person, someone who is concerned for
your health: Your father, the doctor.
JW: And the custard tarts?
SH: Simple. There’s a stain of yellow custard used in making tarts on your
lapel, and your shape convinced me you’ve eaten many before.
JW: There’s no need to be rude.
Now Sherlock, of course, engages in this kind of reasoning with regard to every person he
meets. And, without fail, the people he encounters find the beliefs he forms about them
to be insulting. This negative way in which people respond to what Sherlock Holmes
believes about them, whether it concerns their culpability in a crime or what shoe they
put on first, illustrates how we feel wronged when people look at us the way that Sherlock
Holmes looks at us, how we feel wronged by the way in which Sherlock Holmes forms
beliefs about us.
8
8
A worry one might have here is that this wrong that Sherlock commits, this failure to relate to others
as he ought, is something that is impossible for him to correct. That is, folks who are neurotypical may be
able to relate to others in morally unproblematic ways that are impossible for Sherlock Holmes. Perhaps
Sherlock cannot but relate to others as though they are objects or causal phenomena, and as a result, a moral
theory that concludes that Sherlock is constantly wronging others seems perverse, it compounds Sherlock’s
plight. This objection is part of a larger problem in moral philosophy concerning neurodiversity, disability,
and the demands of morality. For the time being I must set this issue aside, but I want to thank Regina Rini
for pressing me to think more about this issue.
33
2.1.5 Reflections
What we have now seen is a wide array of cases. As noted earlier, in some we wrong
when the belief is false, in some when the belief is true, in some the agent seems epistem-
ically justified and in others the agent doesn’t. See, for example, the following table.
Ep. Justified? True? Negative Bel? Moral Wrong?
Mistaken Identity Yes* No Yes Yes
Wounded By Belief Yes* No Yes Yes
Racist Hermit No Yes Yes Yes
Security Guard No No Yes Yes
Sherlock Holmes Yes* Yes No Yes
Spencer Yes* Yes Yes Yes
Table 1.
9
What is common to all of these cases is this nagging intuition that we can be wronged by
what others believe of us. Further, the the beliefs we are wronged by needn’t be negatively
charged beliefs, i.e., something that most people would not want believed of them. The
beliefs could be mere “observation[s] of trifles.”
10
Over the course of this chapter I will be arguing that a commonality in these cases
is that when someone looks at us not as a person but as though we are an object that
is determined by causal laws, as something whose behaviour is to be predicted, that is
to step back from seeing us as a person. Naturally, that is upsetting. You might now
question whether we can move from the mere fact that these scenarios are upsetting to the
conclusion that someone has been wronged. As I will argue, I think we can, and my goal
is to show how this could be true. My goal, now, moving forward into Section 2.2 is to
show that it is not idiosyncratic to think that moral considerations matter for belief. After
all, our relationships are partly constituted by how we form beliefs about one another.
Although when we think of ethics, we think of moral theories that concern maximizing
the good, or willing into being universal laws, and the such, there is a tradition at the
heart of moral philosophy that doesn’t limit the scope of what we care about to just how
people act towards us and say of us, but also what others think of us.
9
I have put a ‘*’ by the “Yes”’s because, as I’ll explain in Chapter 3, the beliefs aren’t really epistemically
justified. They could potentially be epistemically justified, but the wrong in these cases is that the threshold
for justification does not correctly match the moral stakes of the environment.
10
See Doyle, The Boscombe Valley Mystery.
34
2.2 On How We Relate to Each Other
That the way we ought relate to each other is importantly different from the way we
relate to objects is not a new idea. It can be found in the Kantian idea that we need to
recognize each other as legislators in a kingdom of ends acting under the idea of freedom,
the Sartrean idea that due to our consciousness we transcend the facticity that is our situa-
tion and are thus more than our situation, the Darwallian idea that by standing in relevant
authority relations with each other we can give each other second-personal reasons or the
Strawsonian idea that we just happen to be the kind of beings that care deeply about the
attitudes that we hold about each other. Through weaving together a story about stances
and standpoints that draws on these various accounts and philosophical traditions con-
cerning how we ought relate to one another, I aim to demonstrate that there is a cognitive
dimension to how we ought relate to one another.
As I stated earlier, my goal is to show that this idea is not idiosyncratic, i.e., that all
these various accounts and philosophical traditions are intelligibly gesturing at the same
moral ideal with regard to how we relate to one another. Further, that whatever the story
about why we owe what we owe to each other, there’s a cognitive element to what we owe.
So, over the course of this section I develop an account of what I call the moral standpoint
that draws on these various accounts. Importantly, the account need not commit itself to
additional claims concerning the nature of free will, the human being as a being-for-itself,
etc. As I’ll show, the moral standpoint encompasses not only what we owe each other in
word and deed, but also what we to owe each other in thought.
2.2.1 The Involved and The Participant Stance
Let us start with a Kantian idea; as Rae Langton (1992) explicates it there is an im-
portant way in which we relate to other people that is characteristically different from
how we relate to objects. This idea stems from the following picture of the world and our
place in it. We, as human beings, find ourselves in a world that consists of things—tables,
chairs, corn, cotton—and we try our best to understand how this world works, e.g., why
plants grow when watered, why dogs only give birth to other dogs and never cats, etc. To
understand the world we turn to science as a way of discovering these patterns in nature.
But, we not only try to understand how things work, we think also about how to use
them, i.e., how they can be used as a resource, as a means for human ends. Crucially, ac-
cording to this Kantian picture, each of these things can be priced, and each is essentially
35
replaceable.
11
But, the world consists not only of things; in it there are also people. It is
then, “in our dealings with people,[that] we have a different way of going on, though it
is hard to capture just what that is" (Langton 1992, pp. 486). It is in an effort to capture
just what that is that Langton turns to Strawson to find the following idea:
We don’t simply observe people as we might observe planets, we don’t simply
treat them as things to be sought out when they can be of use to us, and avoid
when they are a nuisance. We are, as Strawson says, involved. (Langton 1992,
pp. 486)
What is it, then, to be involved? P. F. Strawson (1962) argues that insofar as we stand
in relationships with one another, we are susceptible to a wide range of responses from
resentment to love to shame, etc. This is characteristic of the participant stance. Of these
responses, a narrow class—closely associated with responsibility and blame—are often re-
ferred to as the ‘reactive attitudes’. These reactive attitudes include resentment, indigna-
tion, and guilt. These reactive attitudes are characteristic of being involved as they are
reactions to the attitudes and intentions people have towards us, and they concern a con-
nection between the attitudes and moral demands.
12
Further, a significant part of being
so involved concerns according a certain importance to others’ attitudes and intentions
towards ourselves and being cognizant of those demands with regard to our treatment of
others. Strawson argues that we ought take an involved stance towards others and rec-
ognize that, just as our own self-understanding relies on the attitudes that others take
towards us, so too do theirs. This is what we owe to each other.
Some theorists (such as Wallace 1994, 2014) attempt to construe this idea of being
involved in a relationship quite narrowly to include only the paradigmatic cases such as
spouses, friends, lovers, family, etc. But, as we can see in the following quote, Strawson
has a much broader notion of relationships:
We should think of the many different kinds of relationship which we can
have with other people—as sharers of a common interest; as members of the
same family; as colleagues; as friends; as lovers; as chance parties to an enor-
mous range of transactions and encounters. Then we should think, in each of
these connections in turn, and in others, of the kind of importance we attach
11
Although I am sympathetic to the general Kantian picture, there are many things to dislike about the
Kantian picture. For example, it is not the case that all things in the world in virtue of not being people are
thereby essentially replaceable. For example, the environment, natural wonders and monuments, etc.
12
These reactive attitudes "are essentially reactions to the quality of others’ wills toward us, as manifested
in their behaviour: to their good or ill will or indifference or lack of concern" (Strawson 1962, pp. 194).
36
to these attitudes and intentions towards us of those who stand in these rela-
tionships to us, and of the kinds of reactive attitudes and feelings to which we
ourselves are prone. (Strawson 1962, pp. 187)
We can also see a similar sentiment expressed when Tim Scanlon (1998, pp. 168) notes
that morality
...requires us to be moved by (indeed to give priority to) the thought of our
relation to a large number of people, most of whom we will never have any
contact with at all. This may seem bizarre. But if the alternative is to say
that people count for nothing if I will never come in contact with them, then
surely this is bizarre as well.
The participant stance consists in a particular way of caring about or being invested in the
attitudes of another person. Further, this is simply what we owe to others in virtue of
standing in relationships with them.
13
Further still, these relationships we stand in with
one another create the possibility of wrongs that wouldn’t exist otherwise.
14
Moreover, our feelings are indicators that such moral demands haven’t been met. Con-
sider, for example, the anger you feel when you call out to a person to hold the elevator
door for you and they ignore your calls, the disappointment you feel when you tell a
friend a secret and they share it with someone else, or the irritation you feel when some-
one cuts you off in traffic. We can easily recognize that you have been wronged in some
way in each of these cases. As I retell the stories, I start to feel the disappointment, the
anger, the indignation, etc. And I am sure that you start to feel the same way. When moral
demands or these expectations, such as trust, fair regard, respect, etc. are not met, a sign
that they are not met is these reactive attitudes.
15
So, in the spirit of Strawson (and Lang-
ton’s Kant), I suggest that we take these feelings as indicators that some moral demand
has not been met and that we have been wronged. Given that these attitudes are present
in the above cases, we can infer that some demand hasn’t been met. The question now is
what moral demand hasn’t been met.
16
13
I owe much of my thinking on this topic to Steve Bero and am grateful to him for his discussion of
Strawson (1962) in Bero (MS).
14
This is easiest to see in the case of close relationships, such as the demand for doxastic partiality in
friendship. I will address this point in more detail in Section 2.2.
15
Even if you are the kind of person that strives to not have these negative emotions, they are often an
almost automatic response to having been wronged. A drink is spilled on you and immediately you feel
anger, though it may subside for various reasons, e.g. you realize it was a child. More on this in Section 2.3.
16
Marusic and White (2018) develop a similar account to what I will be calling the moral standpoint by
(i) extending some Strawsonian thoughts in moral philosophy to the realm of epistemology, and (ii) some
Burgean apparatus. They argue that we wrong in belief the same way we wrong in action. We wrong, in
both cases, when we fall short of justified normative expectations, for example, when we fail to be genuinely
37
Consider again the person that doesn’t hold the elevator, the person that cuts you
off, the person that shouts “Gross!” when you sneeze. Their attitudes and their actions
convey an indifference and lack of moral care with regard to how they affect. Now, look-
ing back to the opening cases, you may object that the examples that have been used to
illustrate the Kantian idea of an involved stance or the Strawsonian idea of the partici-
pant stance all concern action. The motivating cases of this chapter, however, concerns
thought, in particular, what is believed of you. Perhaps, for example, in Mistaken Iden-
tity when the diner pulled on your sleeve mistaking you for a waiter, they were showing
this criticizable kind of indifference in their action, but not in their belief. We find this
same resistance, this same attempt to limit what we owe each other to exclude epistemic
demands, in Stephen Darwall (2006). So, I turn now to addressing this Darwallian worry
and arguing for an extension of moral care to the realm of thought.
2.2.2 The Cognitive Demands of The Moral Standpoint
Darwall (2006)—in a manner similar to both Kant and Strawson—argues for what he
calls the second-personal standpoint, i.e., the perspective you and I take up when we make
and acknowledge claims on one another’s conduct and will. In speech, we can see this per-
spective in performatives such as demanding, apologizing, etc. In thought, we see this per-
spective in Strawsonian reactive attitudes like resentment and guilt. The reasons that stem
from this standpoint—namely, second-personal reasons—derive from our relations with
one another. Further, he argues that our concept of moral obligation is second-personal
as it concerns a commitment to the idea that those to whom we are morally accountable
have the authority to demand that we do what they ask of us. In addition, when we make
demands on others we do not always do so as individuals, we do so as representatives of
the moral community.
Darwall distinguishes this second-personal stance from the third-person perspective,
i.e., where we regard each other “objectively" or “agent-neutrally". Interestingly, for our
purposes, Darwall explicitly denies there is a cognitive element to the second-personal
stance. Although he argues that the second-person standpoint is the perspective we take
up whenever we make valid claims or demands on someone else (whether explicitly in
speech or implicitly in thought), he also argues that this epistemic authority is third-
personal as opposed to second-personal. Epistemic authority, unlike the relevant author-
cooperative. Belief, they argue (as I will too), is part of the participant stance. hey differ, however, by argu-
ing that instead of adopting evidential policies (as Paul and Morton 2018 argue) or adjusting our thresholds
(as I will argue in Chapter 3, and as Basu and Schroeder 2018 and Schroeder 2018b argue), what is required
of us within our relationships and our shared activities of reasoning is an epistemic permission or a default
entitlement to believe others.
38
ity for second-personal reasons,
...depends fundamentally on a person’s relations to facts and evidence as they
are anyway, not on her relations to other rational cognizers. Even in cases of
testimony where we take someone’s word for something, this second-personal
authority can be defeated by deficiencies of epistemic authority of the ordi-
nary third-personal kind. If we have reason to distrust her beliefs or judg-
ment, we also have reason to reject her second-personal epistemic claims.
(Darwall 2006, pp. 12)
Darwall (2006, pp. 56-7) further goes on to say that,
[o]f course, what reasons people have to believe things about the world de-
pend in many ways on where they stand in relation to it. But ultimately their
reasons must be grounded in something that is independent of their stance,
namely, what is the case believer-neutrally. Our beliefs are simply the world
(including our place in it) as seen (committedly) from our perspective; what
we should believe depends ultimately on the world as it actually is.
There is something right to Darwall’s claim here, but there is also something that he
misses. What he misses regards how the relationships we stand in with each other change
the stance we ought take towards the reasons that ground our beliefs. But first, let me
start with what Darwall gets right.
Ordinarily, when it comes to settling the question of whether to believe p, all that you
should take into consideration is whether p is true, and correspondingly, reasons that are
likely to get us onto the truth of p. In this way, reasons for belief are, as Darwall suggests,
grounded in something that is independent of our stance. The role of our beliefs is to
accurately capture the world as it actually is, not how we would like it to be. However,
even if we accept that we ought believe in accordance with our evidence, we can still ask
the question of when the evidence is sufficient to justify belief. We recognize that in some
cases we need more, and stronger, evidence than in other cases. For example, we require
more evidence when passing judgement in a criminal case than when settling a playground
dispute. Although Darwall does note that what reasons we have to believe things depends
on the ways and where we stand in relation to it: the reasons we have to believe things
depend also on the stance we take towards others. Our standards for justification are not
stance-independent. Crucially, in order to meet the moral demands in our relationships,
these standards must be stance-dependent.
I suggest that what Darwall misses is that our relationships change how we should
respond to evidence. Our reasons for belief with regard to other people must not be
39
grounded in something independent of the stance we ought take towards others. This
point is initially hard to see because it runs counter to our usual intuitions about belief
and our belief-forming practices. So, let me illustrate the point with some cases that speak
to the intuition behind this point, i.e., the point that we can fail to meet the demands of
our relationships precisely by taking this believer-neutral stance.
Consider, for example, the moral requirement that we believe better of our friends.
Sarah Stroud (2006) argues that we have special responsibilities toward our friends which
we don’t have toward strangers that extend beyond our actions towards our friends ver-
sus strangers. Being a good friend also requires a kind of epistemic partiality. What we
ought believe of our friends crucially cannot be something that is independent of our
stance towards them. That is, we should not be neutral with regard to how we respond
to evidence about our friends. For example, suppose you receive testimony that Arjun
cut someone off in traffic. Ordinarily, that testimony would be sufficient to justify your
belief that Arjun cut someone off in traffic. If Arjun is a close friend, however, your
threshold for when the evidence is sufficient to justify believing poorly of your friend is
much higher. So, although testimony would ordinarily be sufficient to justify the belief,
because you stand in this relationship with Arjun you require more evidence before you
will be justified in believing poorly of your friend. So, you have a reason for belief that is
independent of your stance, but whether that is sufficient to believe is something that is
stance-dependent.
Similarly, Paul and Morton (2018) ask us to to consider a person who adopts the goal of
finishing a doctoral dissertation. As I find myself in the same position, let’s just consider
the question of whether I will succeed in finishing my doctoral dissertation. What should
you believe about whether I’ll succeed? The seemingly obvious answer is whatever the
total accessible evidence supports. The unobvious answer, they argue, is that it depends on
your relationship to me. Normally, when beliefs are influenced by personal relationships,
those beliefs are our paradigmatic cases of epistemic irrationality. After all, reasons arising
from our relationships are irrelevant to the truth of those beliefs. Perhaps we have reasons
arising from our relationships to act as though we’re more confident in our loved one’s
ability to succeed, but we shouldn’t believe that they are more likely to succeed. Paul and
Morton argue that although there are not practical or ethical reasons for belief, there are
practical or ethical influences on the standards by which we reason about what to believe.
To establish this claim, they note that evidential policies govern the way we adjust our
evidential thresholds in different contexts. However, there is no uniquely best evidential
policy to have. There are multiple evidential policies that are rationally permissible for
a given thinker to have from the point of view of purely epistemic considerations. Fur-
40
ther, if that’s the case, then practical and ethical considerations can and should play a role
in deciding between epistemically permissible policies. Further, these evidential policies
can also be shaped by our relationships with others. Within the context of significant
relationships, Paul and Morton argue that the default perspective we should take towards
one another is the rational perspective. This perspective is contrasted with the diagnostic
stance in which you view another person as an object whose behavior can be explained
and predicted like a mechanism. The rational perspective, on the other hand, requires
viewing others as rational beings to be engaged with on rational terms. We see that this
distinction tracks the earlier Strawsonian distinction between the objective and the par-
ticipant stance, and I will argue for a moral standpoint akin to Paul and Morton’s rational
perspective. To briefly summarize then, when we doubt our loved ones, we can wrong
them if that doubt is the product of an overly low evidential threshold for adopting the
diagnostic perspective. In short, what we believe of others is not stance-independent.
Finally, Berislav Marusic (2015) also notes that when it comes to beliefs about what
we will do in the future, the stance we take to such beliefs is importantly different from
the stance we take towards other beliefs. When it comes to what you will do in the future,
you should believe against the evidence because whether we succeed is up to you. Further,
those who are partial to us such as our friends, lovers, our spouse, etc. would be wronging
us if they didn’t also believe against the evidence. In virtue of being close to you, they
should exhibit a similar kind of doxastic partiality to what we discussed above. As Marusic
(2015, pp. 2) notes,
[w]e cannot be scientific observers of ourselves without distorting our agency.
In particular, when something is up to us, we distort our view of what we
will do if we seek to predict what we will do—even if our predictions are
based on excellent evidence. Similarly, we cannot be scientific observers of
others without distorting our relationship to them. In particular, if we are
close to them, we distort our relationship if we seek to predict what they
will do—even if our predictions are based on excellent evidence. Agency and
partiality require a view of ourselves and others that is incompatible with a
fully naturalistic view[...].
17
What we see here is epistemic authority having a second-personal nature. The reasons we
have to believe can derive from our relations with one another.
17
It is worth noting here that Marusic (2015)’s view is also heavily influenced by Kant, Sartre, and Straw-
son. His positive account introduces Sartre to develop what he calls a Sartrean response—approaching the
question of what to do as a practical question, not as a theoretical question—as another way of develop-
ing the Kantian dictum that we act under the idea of freedom, as well as Strawson’s claim that we take a
participant point of view toward ourselves.
41
Before continuing, let me also briefly note the following. What Marusic (2015) says
here seems overly strong. Sometimes part of what is required in close relationships is to
predict things about our partners. In virtue of being in a close relationship we are better
predictors of our friends, partner, colleagues, etc. Consider the following case. You’ve
been talking with your partner on the phone and they seem curt with you. Do you wrong
your partner, do you distort your relationship with them, if you suspect they’ve been
having a bad day so on the way home you stop and buy them something that shows you
care?
18
Sometimes the way we show our thoughtfulness is by being good predictors of the
people we care about. So, to simply say that the way in which beliefs can wrong is due to
a kind of moral indifference that stems of a lack of care would not get us the full picture.
Sometimes showing the amount of care that is demanded by the relationships requires
treating one another as objects to be managed and predicted. We can accommodate this,
however, by simply noting that not all objectification is bad. Sometimes it is licensed,
excused, or even invited.
19
However, if these moral demands emerge from the relationships we stand in with
others, we not only have close personal relationships, we also plausibly have relationships
with strangers in virtue of being part of the same moral community. Recall the following
Scanlon (1998, pp. 168) quote from earlier.
[Morality] requires us to be moved by (indeed to give priority to) the thought
of our relation to a large number of people, most of whom we will never have
any contact with at all. This may seem bizarre. But if the alternative is to say
that people count for nothing if I will never come in contact with them, then
surely this is bizarre as well.
Now consider the following complicating case. Suppose that Maria is riding the bus and a
complete stranger gets on the bus smelling of alcohol. Ought she believe that they’ve been
drinking, or should she withhold judgment? I suggest that although the nature of Maria’s
relationship with the stranger on the bus is different from the nature of her relationship
with Mark, there is still something owed in this case. It shouldn’t be surprising that we
also owe strangers something like moral care when forming beliefs about them. After all,
just as we are invested in what they think about us, they are also invested in what we think
18
Perhaps it’s flowers, chocolate, a DVD of a terrible movie, etc. For me, nothing makes me feel better
on a bad day than Mexican coke and a bag of XXtra Flamin’ Hot Crunchy Cheetos.
19
Perhaps the Kantian will be helpful again here. When you use your partner’s stomach as a pillow, when
you manage their emotions when they’re hungry, etc. you are not merely relating to them as an object. As
for the Kantian I can still respect my taxi driver as an end-in-themselves although I am using them as a means
of transport between point A and B, perhaps too in these cases when we make predictions of others we can
do so in a way that still respects their agency.
42
about them. The difference, however, between Wounded By Belief and the stranger on
the bus is not that in one case Maria wrongs and the other she doesn’t, rather it’s a point
about a difference in degree of the wrong. Maria owes more to Mark than to the stranger
on the bus. We owe more to some people in virtue of our relationships with them.
20
To briefly recap before continuing, what I have done so far is provide support for
the claim that our reasons for belief, in particular reasons for believing things of others,
stem from and are constrained by moral demands that emerge from the relationships we
stand in with one another. With regard to believing things of other people, we can wrong
them through what we believe if our beliefs are not responsive viewing them as agents as
opposed to objects to be predicted or managed. This suggests that a key part of the wrong
that is done concerns a kind of objectification and alienation. I turn now to Jean-Paul
Sartre (1943) to expand upon this idea further.
The basic Sartrean idea is that the human condition is both being-in-itself and being-
for-itself. We have both facticity—what we are by our nature, the objective facts about
us—and transcendence—an ability to surpass what we are by our nature, to act against
it, our consciousness. With regard to the relationships we stand in—being-with-others—
we encounter others through the “look". We become aware of ourselves when we are
confronted by the gaze of the other. That is, it is when we are watched that we are aware
of ourselves (as an object), and this gaze is objectifying. Others look at us and define us
as a thing, as a being-in-itself. When seen this way we are objectified. The other robs us
of our freedom and our understanding of ourselves as a being-for-itself. This gaze, Sartre
argues, is alienating. We seek in our relationships to be seen as we wish to be seen; hence
this objectifying gaze of the other unsettles our sense of self.
21
Combining this with the previous accounts, we can see why it hurts to be seen as
a waiter even though there is nothing wrong with being a waiter. Your fellow dinner
patron—a person whom you consider an equal, a colleague, someone with whom you
20
Further, as I shall show in Section 2.3, we also owe more to some people rather than others in virtue of
moral features of our environment such as structural injustice. The nature of our relationships, the moral
environment, whether the belief is negative, whether the belief involves taking the third-person perspective
or the objective stance, etc. are all conditions that play a role in determining whether a belief wrongs and
whether we’ve fulfilled our epistemic obligations to one another.
21
This distinction also plays an important role in de Beauvoir (1949)’s explanation of women being seen
as the other, as a sex object whose being is to be interpreted in relation to men’s desires and choices, rather
than being seen as a free consciousness. Women’s actions are seen as causally determined and women are
seen as governed by their facticity by their biology and their hormones. Part of the struggle, then, of
existing as a woman is that you are forced to see yourself as both an unfree object and as free. This is a
kind of double-consciousness that we see articulated in Du Bois (1903, pp. 2). I will expand on this point in
Section 2.3 where I explain why although we in general owe everyone moral care with regard to forming
beliefs about them, we also owe more moral care with regard to beliefs about members of marginalized
groups.
43
stand in a special relationship—fails to see you. Look again at the apology that’s offered:
“My evidence supported that you were a waiter, not a conference participant. So, I had
the attitude I epistemically ought to have had towards you being a waiter as supported
by my evidence. This belief wasn’t motivated by any feeling of ill-will toward you, I was
just believing in accordance with the evidence." We see in the apology that he takes this
theoretical perspective. He is responding as though he is a scientific observer explaining
casual phenomenon in the world. But, that is precisely what he should be apologizing
for. He should be apologizing for failing to take the participant attitude. For failing
to see why you would be invested in the attitudes he holds of you. Similarly, Maria in
Wounded by Belief fails to recognize that when forming beliefs about Mark in virtue of
their relationship she owes him more than an ordinary person on the street (though of
course, some is still owed there, just less).
This, I suggest, can be captured under the demands of the moral standpoint. We wish
to be related to as we are, as we see ourselves, not as we are expected to be on the basis
of our race, gender, sexual orientation, class, etc. The moral care we must exercise in our
beliefs, then, is a reflection of the fact that we should not be as quick to settle our beliefs
about other people as we are with our beliefs about natural phenomena. The relationship
we stand in with others is importantly different from the one we stand in with objects.
This may be due to recognizing each other’s humanity, recognizing that we we act under
the idea of freedom, recognizing that we are not a being-in-itself but also that we each
are a being-for-itself, etc. I suggest that it can also just simply be that we are the kind of
beings for whom it matters what people think of us. We wish to be related to as we are,
as we see ourselves, not as we are expected to be on the basis of our race, gender, sexual
orientation, class, etc. There need not be a deeper story than that to ground what we owe
to each other.
This is the thought that lies behind the wrong that’s committed to Jamal in Slumdog
Millionaire when he’s mistaken for a tour guide, to Kim in Miss Saigon when she’s mis-
taken for a maid, to Barack Obama when he is mistaken for a waiter or a valet. If the world
is structured such that the evidence suggests you are more likely to be a staff member on
the basis of race so that is what people believe of you, we can now see why it mattered so
much to Whoopi Goldberg that she could finally see a black woman on television who
was playing an officer on a spaceship, not a maid. To repeat the slogan of this chapter,
you should not observe another person, form beliefs about another person, in the same
way we observe and form beliefs about planets. This is the cognitive dimension of what
we owe each other. This is the cognitive demand of the moral standpoint.
44
2.2.3 The Objective, The Theoretical, The Naturalizing Stance
There is, however, another objection that can be raised here. We are not always re-
quired to take the participant stance. How, after all, is sociological work possible if we
must always take the participant stance towards others? Sometimes, the perspective of
the scientific observer is the view we should take towards others. So, perhaps some of
the opening cases are similar, perhaps those are also cases in which we are excused from
taking the participant stance.
To understand the objective stance, let us turn back to Strawson. Strawson (1962, pp.
190) notes that to adopt the objective attitude, in contrast to the participant attitude, is to
see another person as
an object of social policy; as a subject for what, in a wide range of sense, might
be called treatment; as something certainly to be taken account, perhaps pre-
cautionary account, of; to be managed or handled or cured or trained; perhaps
simply to be avoided.
We adopt this objective attitude when the agent is excluded from ordinary adult human
relationships, for example, if they are a child. It can also be available to us as a resource
when aiding in policy or to use out of intellectual curiousity. Although it is often inap-
propriate to take the objective attitude, sometimes we move between the participant and
the objective attitude. For example, the therapist ought to take the objective stance to-
wards their patient, my doctor ought treat me as an object to be managed with regard to
risk factors for disease and ailments, a social scientist ought treat her subjects as objects to
be studied, etc. So, if the opening case is sufficiently similar to cases in which we ought
adopt the objective attitude, then we can disregard the hurt feelings described in the case.
For example, when my doctor informs me of my greater chance of heart disease given
my racial background, I might be upset at the news, but that is not a criticism of my doc-
tor’s judgment. Similarly, if Jim believes that I am a waiter given my racial background,
although I might be upset about being mistaken for a waiter, perhaps it is not a criticism
of his judgement.
We should note, however, that there are nonetheless still moral constraints on how my
doctor is allowed to treat me, how they convey their beliefs about me, and even what they
believe of me. Suppose a more serious case than merely increased risk of heart disease.
Suppose my doctor informs me I have cancer. It would be inappropriate if they laughed
at me while telling me that, or if they announced the news without the seriousness it
deserves. Further, although it is appropriate for my doctor to believe that I have a greater
chance of heart disease because of my race, it would be inappropriate for them to believe
45
that I was a bad tipper on the basis of my race—even if the weight of the statistical evidence
were the same in both cases.
Consider, for example, a case that Keisha Ray (2017) describes. She is suffering from
high blood pressure, so the doctor takes into account her racial background and prescribes
a drug that is shown to be effective for black patients. Although it is appropriate to take
into account her racial background for that purpose, note the inappropriateness of the
doctor adding the following, “Plus it’s cheap, so it’s good for poor black people.” On the
basis of the same statistical racial information, one belief was fine—the belief about which
drug to prescribe—but the other belief was not—the belief about Ray’s financial status.
As Ray further notes,
[e]xperiences like this have made me hyperaware of my blackness when I en-
ter a doctor’s office. Often, patients complain of feeling like just another
number to their doctors, but when I go to the doctor I feel like just another
black person. My doctors don’t see me. They don’t see that I am a black
person who is also a woman, a professional philosopher, a friend, a runner, a
candy aficionado and so forth. They just see a black person.
Recall that this is the wrong that I articulated in Section 2.2. We don’t want to be seen
just as a waiter, just as a black person, just as a shade of brown.
For now, I will just note that taking the objective attitude doesn’t free you entirely
from moral demands. Here’s a general principle that doctors nonetheless follow: serious
matters deserve more moral care. Further, if my doctor is licensed to form beliefs about
me as a scientific observer might, that is due to the relationship we stand in with one
another. The doctor-patient relationship is different from the relationship I stand in with
others. For example, whereas it is appropriate for my doctor to believe that I’m at a
higher-risk for heart disease given my racial background, it would be inappropriate for
a person on the street to believe that. Similarly, in the context of a doctor’s office it
appropriate to be asked to fill out a form listing your past medical history and your current
ailments. It would, however, be inappropriate if an employer asked you to do so.
22
To
repeat, occupying the objective attitude does not free one from all moral demands.
23
22
Consider the episode “Health Care” from Season 1 of The Office in which Dwight Schrute is tasked with
picking a new health care plan for the employees. Initially he chooses an incredibly cheap plan with little
coverage. In response to backlash from the employees, he then distributes forms that ask all the employees
to list the ailments and illnesses they have that they wish to be covered.
23
As a clarificatory point, I should note that part of the reason why the objective attitude does not com-
pletely free us from the moral demands that seem more appropriate for the participant stance is due to the
fact that we are still participants in these relationships and ought recognize each other as other agents even
when occupying this objective attitude. It may be possible that in some cases, such as the doctor-patient
46
There is another question here that remains unanswered. I had said that if my doc-
tor is licensed to form beliefs about me as a scientific observer might, that is due to the
relationship we stand in with one another. However, with regard to doctor-patient rela-
tionship, is it only the relationship that creates the privilege of taking the objective stance?
That is, is it that the relationship changes things such that it’s consistent with the engaged
stance to view the evidence in this detached theoretical way? Alternatively, is it the subject
matter—your body and how it operates, as opposed to social factors concerning poverty,
crime rates, etc.—that changes what is owed? That is, perhaps forming beliefs about oth-
ers on the basis of some statistical evidence, e.g., medical diagnoses that attend to purely
biological factors (for now we are excluding the role that social factors have on people’s
health), is morally fine, whereas other forms of statistical evidence, e.g., social facts, are
problematic. In the abstract, I think there is something to this idea. Perhaps there are
morally neutral reference classes we can invoke and form beliefs about individuals on the
basis of such reference classes. However, our world is messy. As I noted in my earlier
parentheses, a doctor who attends only to biological markers and ignores a patient’s so-
cial setting, family and generational history, etc. will have an incomplete profile of their
patient’s health. This is why I think the approach of trying to find morally unproblematic
reference classes that we can reason with to be a non-starter.
There are still two more complications that arise out of what I’ve said so far. Previ-
ously I noted that whereas it is appropriate for my doctor to believe that I’m at a higher-
risk for heart disease given my racial background, it would be inappropriate for a person
on the street to believe that. However, there are at least two additional cases in which
taking the objective stance may be what’s morally required. First, sometimes we must
recognize objective facts about others. For example, I am a visible minority in more than
one regard in philosophy. As a result, you should recognize that I’ve likely had certain ex-
periences in the profession that others who do not share my demographics have not had.
If you were to disregard that, you would fail to understand me. Consider, for example,
the following poem by Pat Parker (1978), “For the white person who wants to know how
to be my friend".
The first thing you do is to forget that I’m Black.
Second, you must never forget that I’m Black.
To understand each other, we must not forget that human beings aren’t just free to define
relationship, you could take a completely objective attitude towards another person and thus not feel com-
pelled by moral demands. But, to take such a stance would be to stop seeing the other person as human. To
stop relating to another person and recognizing them as a human agent is to wrong them.
47
themselves as they wish.
24
Sartrean bad faith, after all, involves either acting as though
you are totally free to define yourself despite how others see and interpret you or acting
as though you are the role you’ve decided to play. Human beings are complicated, and
how we ought relate to each other is similarly complicated. What we must hedge here is
what Schroeder (2018a, pp. 5) calls the Kantian Gambit.
If and when you interpret someone’s behavior in causal terms, this explana-
tion goes, you take the objective stance toward them, and if and when you
take the objective stance toward someone, you thereby preclude the partici-
pant stance
As Schroeder goes on to note, sometimes treating someone as a person is compatible with
treating them as a thing, and sometimes treating someone as the thing that they are is
how we successfully engage with and understand each other. As I will argue in Section
2.4 and take up in more detail in Chapter 3, perhaps given these complications there are
neither necessary nor sufficient conditions we can state for when beliefs wrong. Instead,
the best we can do is develop our epistemic character such that it responds correctly to
the morally relevant features of our environment in a virtuous manner. As I’ll be noting
later, we already have reason to think there exists a moral-epistemic virtue of this kind:
being woke.
For now, let us turn to one final complication. That is, another case in which per-
haps what we morally ought do is take the objective stance not only towards others, but
also towards ourselves. And I warn the reader now that this discussion will concern both
rape and torture. Strawson himself notes that sometimes we can take occupy the objec-
tive stance as a respite from the strain of human engagement. In addition, we can imagine
cases in which, for reasons of self-protection, perhaps we ought not relate to the best inter-
pretation of another’s behaviour. For example, during traumatic events, depersonalizing
the other (or even oneself) might be a matter of self-preservation. For a less-morally-laden
example, imagine that going to the dentist is an anxiety-inducing prospect. In such a case,
it can be helpful to not think of the dentist as another person or to even consider yourself
as a person. Rather, you are an object being operated on. In cases where a much deeper
moral wrong is being committed to you or against you, e.g., cases of torture or rape,
relating to the other as required by the participant stance would compound the wrong.
As David Sussman (2005) has argued, part of what makes torture a distinctive moral
wrong concerns the interpersonal relationship it enacts. Effective torture pits the victim
24
Thanks to Wendy Salkin, Serene Khader, and Aarthy Vaidyanathan for pressing me to say more on
this.
48
against herself. The victim, Sussman (2005, pp. 25) argues, “is forced into a position where
she must try to anticipate and understand every little mood and quirk of her torturers.
Despite herself, she finds herself trying to grasp her torturer’s interests, anticipate his
demands, and present herself in a way that might evoke pity or satisfaction from him."
Similarly, consider Offred and The Commander in Margaret Atwood’s novel The
Handmaid’s Tale. Every month there is a ceremony, that is, a monthly rape of Offred
by The Commander. The ceremony is bearable only if Offred pretends that it is not hap-
pening to her, if she is able to be detached from both The Commander and his wife. But
soon The Commander begins inviting Offred to play scrabble with him because the im-
personalness of “the ceremony” bothers him too and he wishes to befriend her. He tries
to earn her trust, he gives her gifts, he allows her to read in his presence (an act that is
forbidden). These acts of taking the participant stance, of becoming engaged with and
relating to others as persons now makes the ceremony an even more troubling case of
wronging. As Offred remarks.
When the night for the Ceremony came round again, two or three weeks
later, I found that things were changed. There was an awkwardness now that
there hadn’t been before. Before, I’d treated it as a job, an unpleasant job to
be gone through as fast as possible so it could be over with. Steel yourself,
my mother used to say, before examinations I didn’t want to take or swims in
cold water. I never thought much at the time about what the phrase meant,
but it had something to do with metal, with armor, and that’s what I would
do, I would steel myself. I would pretend not to be present, not in the flesh.
This state of absence, of existing apart from the body, had been true of the
Commander too, I knew now. Probably he thought about other things the
whole time he was with me; with us, for of course Serena Joy was there on
those evenings also. He might have been thinking about what he did during
the day, or about playing golf, or about what he’d had for dinner. The sexual
act, although he performed it in a perfunctory way, must have been largely
unconscious, for him, like scratching himself.
But that night, the first since the beginning of whatever this new arrangement
was between us—I had no name for it—I felt shy of him. I felt, for one thing,
that he was actually looking at me, and I didn’t like it. The lights were on, as
usual, since Serena Joy always avoided anything that would have created an
aura of romance or eroticism, however slight: overhead lights, harsh despite
the canopy. It was like being on an operating table, in the full glare; like
49
being on a stage. I was conscious that my legs were hairy, in the straggly way
of legs that have once been shaved but have grown back; I was conscious of
my armpits too, although of course he couldn’t see them. I felt uncouth. This
act of copulation, fertilization perhaps, which should have been no more to
me than a bee is to a flower, had become for me indecorous, an embarrassing
breach of propriety, which it hadn’t been before.
He was no longer a thing to me. That was the problem. I realized it that
night, and the realization has stayed with me. It complicates. (pp. 160-161)
These cases can be taken in two ways. First, they could provide more evidence for the
moral demand that we relate to others as persons, that we occupy the participant stance,
the rationalist perspective, the moral standpoint, etc. That is, without this requirement
that we relate to others in a particular way, we wouldn’t be able to explain this way in
which torture is a distinctive kind of moral wrong or how when The Commander and
Offred begin to recognize each other as people, no longer as objects, things become com-
plicated. But, it also suggests that there may be times when it is morally required not to
take such a stance towards others. After all, if a moral theory requires that I relate to my
torturers and believe the best of them, that is a strike against the moral theory.
In short, when we are morally required to take the objective stance is a complicated
matter. It is not clear that we’ll be able to articulate a general principle for when we
can recuse ourselves from the demands of engagement that the participant stance and the
moral standpoint require. Human life is complicated, there are cases in which we ought
not relate to others as persons, but instead see them and engage with them as things. In
general, however, the point still stands that our default attitude towards others should be
that of the moral standpoint. The cases discussed in this section should certainly not be
regarded as the norm.
2.3 When We Epistemically Owe More to Some Than To Others
I have focussed on how our relationships can change what we epistemically owe. And
as I suggested at the beginning of this chapter, it is not only our relationships that cre-
ate moral demands on what we ought believe of others, so too do moral features of our
environment such as structural injustice.
To articulate why there is this demand and the differing degrees it comes in, let me
turn to a very general observation about the human condition that lays behind Strawson’s
thoughts about the participant stance.
25
We each conceive of ourselves under a variety of
25
Here I am indebted again to Bero (MS).
50
more or less stable conditions, from marginal ones such as being born on a Friday, to
quite central ones such as being a philosopher, being a spouse, etc. The more central
self-descriptions are important to our sense of self-worth, to our self-understanding, and
constitute our sense of identity. They are the way in which we understand who we are and
where we stand in the world. When these central self-descriptions are ignored in favour
of what we are expected to be on the basis of our race, on the basis of our gender, on
the basis of our sexual orientation, etc. we see ourselves being reduced to this Sartrean
being-in-itself. This is a particularly severe wrong in the case of non-dominantly situated
groups.
For example, imagine mistaking a white man at a Beyoncé concert for a staff mem-
ber rather than a concert-goer. A mistake has been made; a wrong has been done. You
failed to relate to him as he sees himself, and instead formed a belief on the basis of well-
founded statistical evidence concerning the racial demographics of concert-goers and staff
at a Beyoncé concert.
26
I suggest that although there is a wrong in this case, it is much
less severe than the wrong to marginalized groups. The difference depends on the way
in which members of marginalized groups are dispositionally vulnerable with respect to
their self-descriptions while members of dominantly-situated groups are not. Members
of marginialized groups are more dependent on this Sartrean idea of “the look" or “the
gaze" than members of dominantly-situated groups are.
As Desirée Melton (2009) notes, to be dispositionally vulnerable involves an awareness
of self dependence on others for understanding and respect. Yes, we are all invested in how
other people see us, but in the case of racial and gender oppression, of homophobia, of
being disabled, you are more dependent than if you are not a member of a marginalized
group. Your sense of self is more fragile. We owe everyone moral care, but we owe some
people more than others due to the way they are more dependent on others for their
self-descriptions. To help illustrate this point, consider W. E˙ B. Du Bois’s concept of
double-consciousness. Du Bois (1903, pp. 5) notes the following.
It is a peculiar sensation, this double-consciousness, this sense of always look-
ing at one’s self through the eyes of others, of measuring one’s soul by the
tape of a world that looks on in amused contempt and pity. One ever feels his
two-ness,—an American, a Negro; two souls, two thoughts, two unreconciled
strivings; two warring ideals in one dark body, whose dogged strength alone
keeps it from being torn asunder.
When you are not a member of a dominantly-situated group, you must see yourself through
26
Thanks to Gabbrielle Johnson for suggesting this case.
51
two self-conceptions. First you see yourself how the dominantly-situated group sees you;
second you see yourself how you wish to see yourself. Importantly, one of these concep-
tions is maintained structurally, i.e., it matters not whether any individual believes that
you are lazy, there exists this structurally maintained conception of members of your
racial group. An inner conflict arises when those two perspectives inevitably clash. Re-
call the quote from Ray (2017) earlier. Ray sees herself as a professional philosopher, a
friend, a candy aficionado, etc. All that her doctor sees is a black patient. Further, all
her doctor sees is a patient, who in virtue of being black, is likely also poor. This is the
difference between the white man at the Beyoncé concert and the black man at a fancy
restaurant. The first person might be mistaken for a staff member, but given the structure
of our society, he does not suffer from this feeling of double-consciousness. As Simone
de de Beauvoir (1949) notes, the white male perspective is considered the universal per-
spective. For the second person—for Barack Obama, for Whoopi Goldberg—when they
are mistaken for staff members, that deeply challenges their central self-descriptions from
which they draw their sense of self-worth.
To conclude in a manner similar to Kant, Strawson, Sartre, Marusic, et al., I suggest
that in many of the circumstances we find ourselves in we ought not be scientific observers
towards other people. When we are, we distort our relationship to them. Fulfilling what
we owe to others involves an epistemic dimension. This moral demand is the demand to
be related to in a way that is characteristic of being an agent, rather than an object. And for
some groups of people, in particular non-dominantly situated folks, this moral demand
is particularly strong. We ought not relate to other agents in the same way we relate to
determined causal phenomena. We all wish to be seen as individuals, not as objects to be
predicted and managed. This is how morality can make demands on belief. This is how
beliefs wrong. This is the epistemic dimension of what we owe to each other.
2.4 Recognizing Our Epistemic Obligations
My goal has been to make plausible the claim that fulfilling what we owe to others
involves not only being sensitive to how we act towards each other and what we say of each
other, but also what we believe of each other. Along with that theoretical task of clearing
space for this kind of view I favour because it makes sense of our intuitions about the
cases in Section 2.1.1-2.1.5, I wish to conclude by leaving the reader with some practical
advice with regard to recognizing when we’re most at risk of failing to meet our epistemic
obligations to one another.
In many of the circumstances we find ourselves in we ought not be scientific observers
52
towards other people. I have noted, however, that although there are circumstances in
which we ought, the nature of the relationship and the moral demands of the environ-
ment determines when that’s the case. To determine when this is the case, I suggest the
following guidelines.
In general, a belief p wrongs a subject S when some or more of the following
conditions are met,
(i) the belief is a threat to their moral standing,
(ii) the belief is a negatively charged belief,
(iii) the belief coheres with an existing social stereotype that is being invoked,
and/or
(iv) the belief involves viewing someone as an object, as something to be
predicted and determined, unless otherwise licensed by the relationship.
...but not always
With regard to (i), I’ve suggested that part of why we owe marginalized groups more
turns on the way that what we believe of them can be a threat to their moral standing
insofar as we understand their sense of self as a key part of their moral standing (as I think
we should). With regard to (ii), however we want to cash out that condition, it should
be clear that seemingly positive beliefs can also wrong, e.g., supposedly “positive stereo-
types” such as African-Americans excel at athletic activities and Asian-Americans excel
academically can wrong the individuals believed about because their agency is reduced to
statistical generalizations concerning their group membership.
27
With regard to (iii), in
many of these cases where we fail in what we epistemically owe there exists a social stereo-
type that is being invoked, and the belief falls in line with that stereotype. Perhaps it is a
greater threat to one’s standing when that happens. But, there need not necessarily exist
a social stereotype in the background for the belief to wrong. And finally with regard to
(iv), although it is often the case that viewing someone as an object, as something to be
predicted and determined, wrongs them, I’ve also shown that that is not always the case.
One way, then, of determining the degree to which a belief wrongs is by seeing how
many of the conditions I have outlined hold with regard to the belief. But, no condition
27
Perhaps we should instead consider what Schroeder (2018b) calls diminishment. Schroeder argues that
beliefs wrong only when they diminish you, i.e., when they bring you down. To bring someone down,
Schroeder argues, is to interpret them in a way that makes their contribution out to be less. To be dimin-
ished, then, is to be seen as less than, to not been seen as a full agent. Consider again the positive stereotype
that South Asians excel academically. I may have benefitted from this stereotype, but there is also a sense in
which when people believe that I will excel academically because of my ethnicity, my successes are dimin-
ished. My successes are no longer a reflection of me.
53
by itself settles the degree to which the belief wrongs, or even in some cases whether the
belief wrongs. But, they’re still helpful guidelines. Racist beliefs, for example, satisfy
all of these conditions. As a result, that simply suggests that racist beliefs—the kinds of
beliefs that cohere with an existing social stereotype, the kinds of beliefs that view another
person as an object, the kinds of beliefs that are negative, the kinds of beliefs that are a
threat to a person’s moral standing—are the worst kinds of beliefs. Such beliefs are the
most morally bad of the beliefs you can hold of another person. That is in fact what we
think of racist beliefs.
Note, however, that calling out someone for their white privilege might also satisfy
all four of these conditions.
28
In her discussion of white fragility, Robin DiAngelo (2011,
pp. 54) presents the following case.
[She has] just presented a definition of racism that includes the acknowledg-
ment that whites hold social and institutional power over people of color. A
white man is pounding his fist on the table. His face is red and he is furious.
As he pounds he yells, “White people have been discriminated against for 25
years! A white person canÕt get a job anymore!”
We see that he is reacting in anger. Acknowledging one’s white privilege is a threat to one’s
moral standing. If being good, kind, compassionate, egalitarian, etc. is central to your
conception of your self, then being told that you are a racist and that you are complicit
in racist structure is a threat to that understanding of yourself. But if anger is a guide to
moral wronging, in what way has he been wronged here? Are all threats to one’s moral
standing, one’s sense of self, problematic? Similarly, consider the following case. A black
kid is walking the streets of Chicago when he sees a white male police officer. The kid
then forms the belief that the police officer is a bad cop who will (or has) probably engaged
in police brutality against other black residents in Chicago.
29
Here the cop could invoke
all four conditions. Suppose he is a good cop that has never engaged in police brutality
and he wishes that others wouldn’t believe that he was a bad cop. Again, it is central to
his conception of himself.
To answer these cases I turn back to the discussion of dispositional vulnerability from
the previous section. The white participant in Robin DiAngelo’s anti-racism training
seminar, the white male police officer, they are not dispositionally vulnerable. Their sense
of self-worth will be reaffirmed for them through structural features of our environment.
Both need only turn on the TV or watch a movie to see countless positive portrayals of
28
Thanks for Gabbrielle Johnson for raising this point.
29
Thanks to Briana Toole for suggesting this case.
54
white males (and police officers). This is not something that exists for non-dominantly
situated groups.
In general, this problem brings out another problem concerning the attempt to present
necessary and sufficient conditions for when beliefs wrong. A challenge for accounts of
norms on belief (or the epistemic processes that lead to the production of particular be-
liefs) is precisely this way in which standardly accepted norms of reasoning can be manip-
ulated to wrong others. The police officer, the angry man in DiAngelo’s seminar, they all
attempt a move that Kristie Dotson (2018) refers to as a seize of epistemic power. Another
term, to capture the way in which this move resembles emotional abuse and manipulation
would be epistemic manipulation. For example, emotional abusers make use of norms that
are part of a healthy relationship, and they use those norms to manipulate their partners.
Part of why it is so difficult to recognize emotional manipulation as a form of abuse is
how closely it resembles healthy behaviour. As the counterpart to emotional abuse and
emotional manipulation, it can be similarly difficult to identify epistemic manipulation
because of how closely it resembles epistemic norms we standardly accept as good epis-
temic norms. These very tools we can try to identify and articulate as ways of making
sense of ways in which non-dominantly situated groups can be wronged by what others
believe of them can be used and turned against them as tools of the oppressors. This is
one of the many reasons I’m suspicious of attempts to give necessary and sufficient condi-
tions for when beliefs wrong. I am generally suspicious of universal rules that apply in all
contexts, environments and situations. Instead, as I’ll suggest in the next chapter, perhaps
there are general epistemic strategies we can employ when we recognize the moral stakes
of a situation. However, recognizing the moral stakes, adjusting our epistemic behaviour
to meet the stakes, etc., involves the development of a virtuous character that is sensitive to
these issues as opposed to application of laws or rules. Employing the epistemic strategy
alone will not be enough. The epistemic strategy must be applied with a certain amount
of moral sensitivity. So again, we see, that the epistemic and moral cannot be kept entirely
apart. It is a mistake to think that our moral lives and our epistemic lives are separate
realms governed by conflicting obligations. To be moral we must not only do better, we
must believe better as well.
To finally bring this chapter to a close, part of the point I have been trying to make
is that there are many examples where we’ve been wronged by what someone believes
of us that suggest some moral demand hasn’t been met, and these examples all suggest
that there is a pattern for determining when this is the case. However, the diversity of
cases also suggests that there is no simple principle we can articulate. Although this is
not an entirely satisfactory answer, as a general rule of thumb answers to difficult moral
55
problems are not going to be tidy like the solution to a math problem. If they are, then
they’re either solutions for some other kind of creature than a human being or solutions
for a world that is quite different from the unjust one we live in. To fully understand our
epistemic obligations with regard to what we believe of each other, we must understand
what friendship requires of us, what being in committed relationships require of us, what
being conscientiousness requires of us, etc. The final takeaway, however, that I hope to
have made a case for is that fulfilling what we owe to others involves not only being sen-
sitive to how we act towards each other and what we say of each other, but also what we
believe of each other. There are moral demands on the beliefs we form and hold about
each other.
56
CHAPTER III
The Moral Stakes of Our Beliefs
My goal in this dissertation has been to argue that even if racist beliefs were to reflect
reality and thereby seem rationally justified, they are nonetheless wrong. Given the nebu-
lous nature of determining which beliefs are racist, I have focussed on what I take to be the
hardest case in which to see a moral wrong: beliefs that seem justified given the evidence.
With regard to these beliefs, I have previously argued that the way in which they wrong
is by neglecting an important moral difference between our beliefs about objects and nat-
ural events in the world and beliefs about other people. That is, how we morally ought
relate to others affects how we ought respond to evidence concerning them before we set-
tle belief. My first goal in this chapter is to expand upon the question of how we should
respond to the evidence before settling belief. My other goal is to show that this moral
requirement that emerges out of our relationships to others and in response to structural
injustice reveals a moral dimension to the way in which our beliefs are epistemically jus-
tified. Accordingly, the question of what individuals should believe is, in part, a moral
one.
To begin, let me introduce yet another case in which, as a result of the racist world
we live in, the evidence seems stacked in favor of seemingly racist beliefs. Consider the
following case from Gendler (2011) in which a social club’s discriminatory membership
practices have resulted in a stark racial divide between club members and staff members:
only a fraction of the club’s members are black, whereas all of the club’s staff members
are black.
Social Club. Agnes and Esther are members of a swanky D.C. social club
with a strict dress code of tuxedos for both male guests and staff members, and
dresses for female guests and staff members. While preparing for their evening
walk, the two women head toward the coat check to collect their coats. As
they approach the coat check, they both look around for a staff member. As
57
Agnes looks around she notices a well-dressed black man standing off to the
side and tells Esther, “There’s a staff member. We can give our coat check
ticket to him.”
In this case, it seems that Agnes’s belief is justified. That the man standing in the lobby
is black provides substantial evidential support for the belief that he is a staff member.
Such a case, and others like it—recall that in supposedly rational racist we were assuming
that that the diner is black provides substantial evidential support for Spencer’s belief that
the diner won’t tip will—raise the troubling possibility that by simply attending to our
epistemic obligations, we may contribute to the harms and wrongs of racism. On one
hand, as we’ve seen discussed in Chapter 2, assuming that someone is a staff member on
the basis of their race is a paradigmatic example of racism and is morally impermissible.
On the other hand, the belief seems epistemically justified. If one had more hands, one
might go further and add that it would be epistemically irrational not to believe that they
are a staff member. After all, to refrain from believing in this case seems to involve a wilful
disregard of the evidence.
In light of such cases, one might be tempted to follow Gendler (2011) and conclude that
we face an irresolvable conflict between our moral obligations and what we epistemically
ought to believe. I disagree. I suggest that these cases push us to rethink the assumptions
that pit moral considerations and epistemic considerations against each other: namely,
the evidentialist assumption that the only considerations relevant to whether you should
believe are evidential considerations. That is, according to the evidentialist, moral con-
siderations are neither here nor there when it comes to the question of whether a belief is
justified given one’s evidence.
In order to call into question whether Agnes or Esther are justified in believing what
they believe on the basis of their evidence, we need first to understand what it is for your
evidence to support believing p. In Section 3.1 I turn to looking closely at what eviden-
tialism says, what it doesn’t say, and to what degree various claims it makes are ones that
are worth keeping in light of the troubling possibility that by simply attending to our
epistemic obligations we may contribute to the harms and wrongs of racism. From this
analysis, we will see that evidentialism leaves open both questions of how much evidence
is enough evidence and the question of what determines how much evidence is enough
evidence. As I’ll show in Section 3.2 these open questions are where stakes-based en-
croachment accounts in epistemology gain traction: non-evidential considerations plau-
sibly play a role in settling the question of how much evidence is enough evidence.
I am not the first to note this point. Defenders of pragmatic encroachment in episte-
mology have argued that non-paradigmatically epistemic features of a subject’s environ-
58
ment can make a difference to what that subject is justified in believing or what they know.
Consider the following example, if you are deathly allergic to nuts, the threshold your ev-
idence must pass for you to be justified in believing that a dish does not contain nuts is
higher than for someone lacking such an allergy. In Sections 3.2.1-3.2.6 I canvas various
accounts of pragmatic encroachment that have been offered in the literature and show
how they challenge and weaken the evidentialist thesis. In so doing, I will suggest that the
motivating cases for pragmatic encroachment will also help us motivate moral encroach-
ment. Whereas traditional stakes-based encroachment accounts are articulated in terms
of practical stakes, such accounts leave open the question of which non-paradigmatically
epistemic features of a subject’s environment can make a difference to what that subject
is justified in believing or what they know. This leaves open the possibility that features
that matter morally can affect whether a subject is justified or whether a subject knows.
Similar arguments can be found in Fritz (2018), Pace (2011), Moss (2018), and Schroeder
(2018b) and I will expand upon these ways in which others have similarly motivated moral
encroachment in Section 3.3.
Once the stage is set for an account of moral encroachment, in Section 3.4 I address
three questions that arise for such an account. First, what would the low-stakes variants
of the traditional pragmatic cases be; second, what is it that makes these cases high stakes;
and third, is what makes these cases high stakes the same in both the pragmatic and the
moral domains. I argue that our epistemic practices are constrained by a kind of moral
sensitivity. This moral sensitivity amounts to not only a mere recognition of an unjust
environment, but also an adjustment of one’s threshold for justification to meet the moral
stakes of one’s environment. Finally, in Section 3.5 I consider three objections to moral
encroachment. First, that moral encroachment is an incomplete account of the ways in
which beliefs can wrong, i.e., of doxastic morality. Second, that moral encroachment has
perverse consequences. Third, that moral encroachment is too demanding.
3.1 Rethinking the Evidentialist Assumptions
Let us start with another case to get the intuitive feel for evidentialism. Suppose your
friend Veronica tells you that she knows the identity of the Zodiac Killer. Intrigued, you
follow up. It turns out that Veronica believes that Ted Cruz is the Zodiac Killer. Not
wanting to dismiss her out of hand, presumably you’d go on to ask her for what evidence
she has in support of her belief. If you found out that she hopes that Ted Cruz is the Zodiac
Killer, because it would be funny, that would be irrelevant to the question of whether in
fact Ted Cruz is the Zodiac Killer. After all, that Veronica hopes Ted Cruz is the Zodiac
59
Killer has no bearing on whether or not Ted Cruz is the Zodiac Killer.
The purpose of this vignette is to illustrate the simple point that spelling out the re-
lationship between evidence and justification is not a straightforward matter. This point
complicates the standard evidentialist picture since, although the evidentialist is correct
to note that there is a connection between justification and evidence and that we should
believe in accordance with the evidence, this picture leaves open two distinct questions
about what settling in accordance with the question amounts to. How much evidence is
enough evidence? What determines how much evidence is enough evidence? In what fol-
lows, I use this example of Veronica to demonstrate that spelling out what it is to believe
in accordance with the evidence first requires settling these two questions. Further, these
questions are sensitive to non-evidential considerations.
Returning to Veronica, she might have some evidence that Ted Cruz is the Zodiac
Killer. For example, Ted Cruz somewhat resembles the police sketch of the Zodiac Killer.
This evidence, however, is not sufficient to make Veronica’s belief that Ted Cruz is the
Zodiac Killer justified.
1
For the belief to be justified, she would need more evidence. This
simple case illustrates that when we try to determine whether a belief is justified, we look
not only for evidence—reliable signs, symptoms, or marks of that which it is evidence of,
e.g., reasons that raise the probability of p being true—we also consider the strength of the
evidence. In short, whether a belief is justified is a matter of both what the evidence is
and second, whether the evidence is sufficient to justify the belief. We can articulate this
minimal intuitive gloss on evidentialism as follows.
Support. For each proposition p and body of evidence E, there is some thresh-
old for justification t, such that it is permissible to believe p on the basis of E
only if the pr(pjE) t.
Here pr(pjE) corresponds to some way of understanding the strength of the evidence E
for the proposition under consideration, and the threshold t corresponds to a sufficiency
condition the strength of one’s evidence must either meet or surpass to be justified. For
example, consider an easy case, such as whether you are justified in believing that the ket-
tle has boiled. The sound of a kettle whistling raises the probability of it being true that
the kettle has boiled. Further, it raises the probability higher than if you had just glanced
at your watch and seen that the kettle has been on for two minutes. Both make it more
likely than not that the kettle has boiled, but only the sound of the whistling raises the
1
Stepping away from cases such as the identity of the Zodiac Killer, we see this constraint on our beliefs
everywhere in our daily life. If you are trying to decide whether to pack an umbrella, you might look
outside or check the weather report to gather evidence before settling your belief. Your friend, on the
other hand, who believes that it will rain because they want it to rain believes irrationally.
60
probability to a degree that is sufficient for being justified in believing that the kettle has
boiled. However, that threshold need not be the same for every belief. Presumably, in
some circumstances you would want stronger evidence, i.e., you would want the thresh-
old to be higher. In others, you might only require weaker evidence, i.e., for the threshold
to be lower. For example, in adjudicating someone’s guilt in a criminal case, the thresh-
old is higher than when adjudicating a playground dispute. All of this is consistent with
thinking that there is some threshold that the strength of one’s evidence must cross in
order to justify belief.
In line with this minimal gloss on the evidentialist intuition, in Section 3.4 I argue that
many factors, including the moral demands of our environment, affect where we place
the threshold. For example, if you are in a morally demanding situation, e.g., a context in
which you risk contributing to the harms of racism given what you believe, the threshold
is higher than when you are not in such situations, e.g., when you are deciding whether to
bring an umbrella to work. However, many evidentialists reject the idea that the threshold
for justification can be shifty in this way because it conflicts with some more restrictive
accounts of evidentialism.
Many evidentialists are not satisfied with only Support. This dissatisfaction leads to
the endorsement of restrictions on justification that are designed to restrict the shiftiness
of the threshold. One commonly endorsed principle that is considered central to eviden-
tialism is the following.
Supervenience. For any two subjects S and S’, necessarily, if S and S’ have the
same evidence for/against p, then S is justified in believing that p iff S’ is, too.
2
Supervenience is designed to capture the thought that, for example, two jurors charged
with adjudicating someone’s guilt shouldn’t have differing thresholds for justification. If
they both have the same evidence for a guilty verdict, then if one is justified in believing
that the defendant is guilty, so too is the other juror.
3
A further restrictive version of evidentialism suggests that there is only ever one belief
that is justified given the evidence—Uniqueness.
2
See Fantl and McGrath (2002) and Conee and Feldman (2004). Fantl and McGrath call this principle
‘purism’.
3
Note that the subjective Bayesian would reject this formulation of Supervenience, but they would accept
the following version: for any two subjects S and S’, if S’s evidence and S”s evidence supports p to the same
degree, if S is justified, S’ is too. Note that the difference is that is not merely a matter of having the same
evidence, the evidence must support the belief to the same degree for both agents. Pragmatic encroachment
is usually taken to be incompatible with both. The account of moral encroachment that I will develop in
this chapter gives us counterexamples to this second form of Supervenience, but not to the first. In part
because of the difficulties of creating cases contrasting pairs of high-low cases.
61
Uniqueness. A body of evidence justifies at most one proposition out of a
competing set of propositions (e.g., one theory out of a bunch of exclusive
alternatives) and that it justifies at most one attitude toward any particular
proposition.
4
Returning to the cases that motivated us at the beginning of the chapter, according to the
restrictive version of evidentialism that is jointly committed to Support and Supervenience,
anyone who finds themselves in the same situation as Agnes and Esther is similarly jus-
tified in believing that the black man standing in the lobby is a staff member. That is, if
Agnes and Esther are justified, so too are they; and if Agnes and Esther are not justified,
neither are they. If we add Uniqueness, then that is the only belief that is justified given the
evidence. That is, it is not only the case that one ought believe that he is a staff member,
but also the rest of us are irrational if we don’t believe that he is a staff member.
Although both Supervenience and Uniqueness have been taken to be central compo-
nents of evidentialism, both are stronger than the initial gloss we gave of evidentialism
in terms of Support. It is open for debate whether our intuitive practices of requiring
evidence in support of our beliefs requires all three theses—Support, Supervenience, and
Uniqueness. I suggest that the seemingly irresolvable nature of the conflict at the heart
of this dissertation—the conflict between what we ought morally to believe and what we
ought epistemically to believe—only arises because of the more restrictive accounts of ev-
identialism masquerading as intuitive glosses of the more general evidentialist intuition
that one ought believe in accordance with the evidence.
An important upshot of the account I develop in this chapter is that although stakes-
based encroachment views are often considered rejections of evidentialism, they can re-
main compatible with the most intuitive form of evidentialism. To embrace moral en-
croachment, then, does not require denying an important connection between justifica-
tion and evidence. All that moral encroachment asks for is the recognition that what
counts as enough evidence to justify belief can vary according to practical factors and the
moral stakes of the agent’s situation.
Moving forward my strategy is to show how we can resolve this conflict between our
moral and our epistemic obligations while preserving the basic evidentialist intuition un-
derstood as a commitment to Support.
As a matter of framing, you might respond that this is not so much an issue for eviden-
4
From Feldman (2007, pp. 205). Alternatively, White (2005, pp. 445) presents uniqueness as follows:
“Given one’s total evidence, there is a unique rational doxastic attitude that one can take to any proposi-
tion.” Further, as before with Supervenience, there is also a weaker version of Uniqueness which restricts
impermissivism to just one person at a particular time. That is, for each agent at each time, the evidence
justifies only one proposition.
62
tialism, rather it is a challenge for the moral requirement to occupy the moral standpoint
with regard to what we believe of each other. That this requirement conflicts with a seem-
ingly orthodox view of belief and justification of belief is a strong reason against thinking
that there could be such a requirement. Thus, the burden is not on evidentialism to ac-
commodate this moral-epistemic requirement.
Rather, the burden is on those that defend the distinction to do one of five things.
First, we could explain away the intuitive plausibility of evidentialism insofar as it seems
responsible for generating this kind of conflict. Alternatively, we could show that there is
in fact no conflict between the moral and epistemic ought by rethinking and questioning
the assumptions that seem to pit them together (as I will be suggesting). Third, perhaps we
could show that although there is a conflict between epistemic and moral demands, there
is a way of adjudicating between them such that we can say there is something Spencer and
Agnes ought all-things-considered believe. Fourth, we could deny that there’s anything
that Spencer or Agnes ought all-things-considered believe while also denying that that
means we must resort to the pessimistic conclusion that this conflict is irresolvable. That
is, when oughts seem to conflict there is no conflict, there is one sense in which you should
do x and a different sense in which you should do y. Finally, fifth, we could just endorse the
pessimistic conclusion that there is no way to determine what to do in cases of conflicts.
My goal in this chapter is the second approach—show that there is in fact no conflict
between the moral and epistemic ought by rethinking and questioning the assumptions
that seem to pit them together. I leave the task of canvassing reasons for preferring this
approach to the third, fourth, and fifth approach for Chapter 4. With regard to the first
alternative, I will briefly note that although it might be tempting given the many chal-
lenges to evidentialism despite it being considered the orthodoxy, the standard ways of
challenging the intuitive plausibility of evidentialism will not help us in this case. For
certain classes of belief, evidentialism fails as an explanation of their justificatory status.
For example, reasons to believe mathematical truths or a priori truths do not rest on one’s
evidence. Nor do cases of self-knowledge require that an agent believe in accordance with
her evidence. For empirical beliefs, on the other hand, like your friend’s belief that Ted
Cruz is the Zodiac Killer, evidence does seem like what we look for to determine whether
the belief is rational and justified. Beliefs about others are also empirical in this way. So,
there is reason to think that evidentialism is a requirement on such beliefs. My goal, then,
is to show that the requirement to occupy the moral standpoint, and for moral consider-
ations to affect the epistemic justification of our beliefs, is not at odds with a weak version
of evidentialism. To do so, we find friends amongst the pragmatic encroachers.
63
3.2 Varieties of Pragmatic Encroachment
Suppose you are not allergic to nuts but simply have a general dislike for them. In
such a case, a waiter’s assurance that they’re pretty sure there aren’t nuts in any of the
dishes would be sufficient for you to justifiably believe that there aren’t nuts. However, if
your dining partner Sarai is deathly allergic to nuts, then the waiter’s assurance no longer
suffices. Sarai, given the stakes, should not believe that none of the dishes contain nuts.
Both you and Sarai, however, have the very same evidence that supports the belief none
of the dishes contain nuts to the very same degree. That is, if we suppose the evidence—the
waiter’s testimony—supports that the dish doesn’t contain nuts to degree .7 for you, it
also supports that the dish doesn’t contain nuts to degree .7 for Sarai. However, whereas
you are justified in believing that none of the dishes contain nuts, we tend to be resistant
to the claim that Sarai is similarly justified in believing that none of the dishes contain
nuts. As I have been pressing, the evidential situation for both you and Sarai is identi-
cal. But, there is a difference in our intuitive judgements concerning justification. This
suggests that whatever difference there is between our judgments about justification must
come down to something other than the evidence. By simply changing a practical fea-
ture of the scenario—what is at stake for Sarai—we have also thereby changed what she is
epistemically justified in believing. Although the evidentialist might reject practical con-
siderations as irrelevant to the question of justification, our intuitions suggest otherwise.
Defenders of pragmatic encroachment accounts in epistemology use cases like the one
above to argue that practical features of one’s environment, not evidence alone, can change
the epistemic status of a believer—e.g., one can change from being justified to not being
justified, from having to knowledge to lacking knowledge. Pragmatic encroachment, de-
spite having historical roots in Locke (1975)/1689, is often considered a relatively recent
philosophical thesis that offers an invariantist alternative to contextualist treatments of
‘knows’.
5
Understood in this way, it can initially be difficult to see how pragmatic en-
croachment might offer a structural analogue to the moral challenge I raise against evi-
dentialism. The challenge I raised concerns the rational permissibility of belief formation
in light of moral concerns. But if pragmatic encroachment only concerns the semantics
5
From Grimm (2016):
According to Locke, for instance, “it is very wrong and irrational way of proceeding, to
venture a greater good for a less, upon uncertain guesses and before a due examination be
made, proportionable to the weightiness of the matter, and the concernment it is to us not
to mistake” (2.21.66); and again, “where the assent one way or the other is of no importance
to the interest of anyone ... there ‘tis not strange that the mind should give itself up to the
common opinion, or render itself to the first comer’ (1975: 717)
64
of ‘knows’ and the question of whether knowledge ascriptions are context-sensitive, what
does that have to do with the rationality of belief? As I’ll show, by canvassing the vari-
ous versions of pragmatic encroachment, pragmatic encroachment is not merely about
knowledge and knowledge ascriptions.
To start, our concept of knowledge involves various components that can be given a
probabilistic gloss. For example, knowledge entails justified belief. When you know that
p, that is partly constituted by your believing that p. So, when we attribute knowledge to
someone we are also attributing a belief to them. And when we attribute a belief to some-
one, we can ask how sure, how confident, and how strongly they believe that p. Similarly,
for that belief to count as knowledge it must be justified. But as we saw in the discussion
of evidentialism, we can ask how much evidential support is required to justify belief.
In light of this, as I have been suggesting, it is a misconception to think that pragmatic
encroachment only concerns knowledge.
6
As I’ll show, some pragmatic encroachment
proposals argue that pragmatic considerations do not apply directly to knowledge, but
rather they enter into our assessments of knowledge through pragmatic constraints on
belief or pragmatic constraints on justification. Whereas other defenses of pragmatic en-
croachment emerge from a muddying of the distinction between practical and theoretical
rationality with regard to how to balance the competing goals of believing the truth and
avoiding error, and further questioning whether epistemic rationality is exhausted by ev-
idential considerations. The phenomenon of pragmatic encroachment raises the general
question of whether there are any epistemic notions that aren’t sensitive to an agent’s
interests.
Following the presentation of the varieties of pragmatic encroachment, I then turn to
sketching the most innocuous and least-committal version of pragmatic encroachment I
can. Doing so allows us to explore the challenge pragmatic encroachment raises for evi-
dentialism while taking on as few other commitments as possible and thus being neutral
between the various theories. Although still controversial, my hope is that by presenting
the least-committal version of the basic pragmatic encroachment intuition we can see the
shape of the challenge pragmatic encroachment offers against evidentialism, and how it
modifies evidentialism in light of the response. And then in Section 3.3 we can explore
if the same strategy can be extended to accommodate moral considerations. This way,
we need not commit ourselves to any particular theory of pragmatic encroachment in
presenting a story about moral encroachment.
6
We can also see this limited version of pragmatic encroachment expressed in the Stanford Encyclope-
dia of Philosophy definition of pragmatic encroachment as merely the thesis that a difference in pragmatic
circumstances can constitute a difference in knowledge.
65
3.2.1 The Contextualist Roots of Pragmatic Encroachment
As noted earlier, pragmatic encroachment, despite having historical roots in Locke, is
often considered a relatively recently philosophical thesis that offers an invariantist alter-
native to contextualist treatments of ‘knows’. So to understand the pragmatic encroach-
ment thesis, it will be useful to start with the view that it is often contrasted against:
contextualism. Contextualism about ‘knows’ is motivated by the following pair of cases
from DeRose (1992).
Bank Case A. My wife and I are driving home on a Friday afternoon. We plan
to stop at the bank on the way home to deposit our paychecks. But as we drive
past the bank, we notice that the lines inside are very long, as they often are
on Friday afternoons. Although we generally like to deposit our paychecks
as soon as possible, it is not especially important in this case that they be de-
posited right away, so I suggest that we drive straight home and deposit our
paychecks on Saturday morning. My wife says, “Maybe the bank won’t be
open tomorrow. Lots of banks are closed on Saturdays." I reply, “No, I know
it’ll be open, I was just there two ago on Saturday. It’s open until noon."
Bank Case B. My wife and I drive past the bank on a Friday afternoon, as in
Case A, and notice the long lines. I again suggest that we deposit our pay-
checks on Saturday morning, explaining that I was at the bank on Saturday
morning only two weeks ago and discovered that it was open until noon. But
in this case, we have just written a very large and important check. If our pay-
checks are not deposited into our checking account before Monday morning,
the important check we wrote will bounce, leaving us in a very bad situation.
And, of course, the bank is not open on Sunday. My wife reminds me of
these facts. She then says, “Banks do change their hours. Do you know the
bank will be open tomorrow?" Remaining as confident as I was before that
the bank will be open then, still, I reply, “Well, no. I’d better go in and make
sure."
That I know that the bank will be open on Saturday in Bank Case A seems true, but I
also seem to be saying something true in Bank Case B when I concede that I don’t know
that the bank will be open on Saturday. However, the evidence supports my belief to the
same degree, and it is not as though I am in any better epistemic position in A than in B,
so following Supervenience, if I know in Bank Case A, I must also know in Bank Case B.
66
In light of these cases, epistemic contextualists argue that the content of ‘S knows that
p’ depends upon the context of utterance (i.e. the attributer’s context). A purported
benefit to this explanation is that it does not undermine the following thesis.
Intellectualism. The factors in virtue of which a true belief amounts to knowl-
edge are exclusively truth-conducive, in the sense that their existence makes
the belief more likely to be true.
In the discussion of evidentialism from Section 3.1 we captured something similar to In-
tellectualism in our definition of evidence as that which raises the probability p of being
true. We can see the evidentialist as committed to a more restricted version of intellec-
tualism in which the factors in virtue of which a true belief amounts to knowledge is
determined by the strength of the evidence alone. After all, the evidentialist defines ev-
idence as anything that raises the probability of p being true. So, the evidentialist take
on Intellectualism would be that the factors in virtue of which a true belief amounts to
knowledge are exclusively evidential.
Hawthorne (2003), however, argues that an unattractive feature of contextualism is
that it cannot maintain the intuitive connections between knowledge, assertion, and prac-
tical reasoning. In response, he develops an account of pragmatic encroachment as an
invariantist alternative to contextualism. Relatedly, Stanley (2005) also argues that con-
textualism is counter-intuitive and that the counter-intuitive results that it delivers on vari-
ants of the bank cases also displays the way in which Intellectualism stands in conflict with
an intuitive connection between knowledge and action. I turn now to presenting in more
detail Stanley (2005)’s account of pragmatic encroachment. At the same time that Stanley
was developing his view, Fantl and McGrath (2002, 2007, 2009) over a series of papers and
a book were developing a similar account but from different motivations. What remains
central to all three accounts, and later the combined Hawthorne and Stanley (2008) ac-
count, is the importance of preserving the intuitive connection between knowledge and
action.
3.2.2 Explaining Encroachment: Knowledge and Action
Stanley (2005)’s motivating thought is that our intuitions about whether someone
knows that p seem to vary according to non-epistemic features of that person’s situation.
For example, if someone is in a risky situation we feel that she must be in a particularly
good epistemic position. Whereas we are more permissive when she is not in a risky
situation. Recall that when not much was at stake, it is true that I know that the bank
will be open on Saturday. However, when I’m in a riskier situation with regard to my
67
finances, it no longer seems that I know that the bank will be open on Saturday. Further,
not only do our intuitions change based on the risk, they also seem to change based on
what it is salient to either the knower or ourselves. Stanley argues that in a variation of
Bank Case B in which I am ignorant of the high stakes if the bank’s closed and I don’t get
a chance to deposit my paycheck, ‘I know the bank will be open on Saturday’ is not true,
whereas according to the contextualist it would be true.
In short, Stanley’s argument for pragmatic encroachment proceeds by creating similar
low-stakes/high-stakes pairs of cases to the Bank Cases that motivated contextualism and
showing how our intuitive responses suggest an alternative to contextualism. The basic
formula for creating these cases is to first take an agent and put them in a scenario where
it would be reasonable to think that they know. The contrasting case, then involves a
hypothetical parallel in which the cost of being wrong no longer makes it reasonable
to think that they know. We can also see how the formula for creating cases is a direct
challenge to the Supervenience component of evidentialism. Given Supervenience, our
judgements about knowledge should not shift merely because of a difference with regard
to the cost of being wrong. It is worth making this formula explicit because it is common
to all accounts of pragmatic encroachment. Distinctive of Stanley’s approach, however,
is the weight he places on our intuitive responses to these cases.
Stanley suggests that what our intuitions are tracking in such cases is a conceptual
link between knowledge and action. That is, to know that p is related to being rationally
permitted to act upon the belief that p. He suggests that given that the latter is a function
of practical facts, it wouldn’t be a surprise if the former was as well. His suggestion, then,
is that given this connection with action, knowledge is an interest-relative notion. That
is, whether someone knows that p at a time t depends at least in part upon practical facts.
Although his argument proceeds through a polling of intuitions, he defends this approach
by noting the following.
The role of these intuitions is not akin to the role of observational data for
scientific theory. The intuitions are instead intended to reveal the powerful
intuitive sway of the thesis that knowledge is the basis of action. Someone
who denies that we have many of these intuitions is denying the pull of the
link between knowledge to action. (pp. 12)
So, what explains the intuitions we have in the various variants of the bank case that
Stanley presents is a plausible conceptual connection between knowledge and action? To
capture these intuitions, Stanley puts forward the following view as an alternative to con-
textualism.
68
Interest-relative Invariantism. Whether or not someone knows that p may be
determined in part by practical facts about the subject’s environment. (pp.
86)
In so doing, Stanley straightforwardly rejects intellectualism and instead endorses the
view that whether a belief constitutes knowledge depends in part on non-truth-conducive
practical matters such as the cost of being wrong.
Now, despite the discussion thus far being couched in terms of knowledge it can be ex-
tended to evidence given that both Stanley and Hawthorne endorse Williamson (2000)’s
claim that E=K, that is S’s evidence is S’s knowledge. What one knows can be identified
with a person’s total evidence. So, in the last pages of his book, Stanley begins to ques-
tion whether all epistemic notions are interest-relative. After all, if knowledge is interest-
relative and knowledge just is an agent’s total evidence, perhaps evidence is interest-relative
as well. He asks,
If all epistemic notions are interest-relative, then evidence is interest-relative as
well. If evidence shares the interest-relativity of knowledge, then two people
who do not share the same practical situation will not in general have the
same evidence. (pp. 124)
We see that pragmatic encroachment is not only a theory about knowledge, it is also a
theory about evidence, and in general about the boundaries of what counts as ‘epistemic’.
To gesture ahead, as we’ll see in discussion of Schroeder (2012a)’s account of pragmatic
encroachment, if we think that the only reasons that determine the rational permissibility
of belief are epistemic reasons, given the phenomena that motivates pragmatic encroach-
ment we cannot presume that that claim is identical with the claim that the only evidence
(or strength of evidence alone) determines the rational permissibility of belief. So, we see
another way in which pragmatic encroachment challenges traditional evidentialist theses.
But, before we turn to Schroeder (2012a)’s account of pragmatic encroachment, we
must first consider an account of pragmatic encroachment more similar in spirit to that
of Stanley. That is, an account similarly motivated by a desire to preserve the relation be-
tween knowledge and action, a distrust of Intellectualism, and a rejection of Supervenience.
This is the account offered by Fantl and McGrath (2002, 2009).
3.2.3 Explaining Encroachment: Knowledge-Justification and Action
Fantl and McGrath (2002) start with a familiar pair of cases, this time concerning
trains.
69
Train Case 1. You’re at Back Bay Station in Boston preparing to take the com-
muter rail to Providence. You’re going to see friends. It will be a relaxing
vacation. You’ve been in a rather boring conversation with a guy standing
beside you. He, too, is going to visit friends in Providence. As the train
rolls into the station, you continue the conversation by asking, “Does this
train make all those little stops, in Foxboro, Attleboro, etc." It doesn’t matter
much to you whether the train is the “Express" or not, though you’d mildly
prefer it was. He answers, “Yeah, this one makes all those little stops. They
told me when I bought the ticket." Nothing about him seems particularly un-
trustworthy. You believe what he says.
Train Case 2. You absolutely need to be in Foxboro, the sooner the better.
Your career depends on it. You’ve got tickets for a southbound train that
leaves in two hours and gets into Foxboro in the nick of time. You overhear
a conversation like that in Train Case 1 concerning the train that just rolled
into the station and leaves in 15 minutes. You think, “That guy’s information
might be wrong. What’s it to him whether the train stops in Foxboro? Maybe
the ticket-seller misunderstood his question. Maybe he misunderstood the
answer. Who knows when he bought the ticket? I don’t want to be wrong
about this. I’d better go check it out myself."
As before in the bank cases, in the first case we have a different judgement than in the
second case. When not much is at stake, you have good enough evidence to know that the
train stops in Foxboro. However, when more is at stake, you do not have good enough
evidence to know that the train stops in Foxboro. Fantl and McGrath (2002) use this case
to suggest that epistemic justification—the kind of justification that entails knowledge—is
not simply a matter of the evidence that one has. In both train cases, you’re supposed
to have the same evidence. But in one, you are justified. And in the other, you are not.
So, not only do we have a violation of Supervenience (which Fantl and McGrath (2009)
refer to as Purism), we also see the way in which the discussion concerning pragmatic
encroachment is not only about knowledge, it is also about justification.
In their paper, Fantl and McGrath (2002) use our judgment in this case to defend the
following pragmatic condition on epistemic justification.
(PC) S is justified in believing that p only if it is rational for S to prefer as if P.
This principle follows from the following principle connecting knowledge to action. Sup-
pose S knows that p—e.g., the door is locked—and assume that S knows that if p, then A is
70
the thing to do—e.g., if the door is locked, don’t double back to check the door is locked.
If I know that p, and if I know that if p, then A, then I’m rational in doing A. The train
cases illustrate that knowledge shifts because what it is rational to do shifts.
Later, in their book, Fantl and McGrath (2009) build their account of pragmatic en-
croachment is around the following principle, Knowledge-Justification.
(KJ) If you know that p, then p is warranted enough to justify you in-ing,
for any.
But, why suppose that (KJ) is true? On similar grounds to their earlier paper and to
Stanley and Hawthorne, Fantl and McGrath argue that (KJ) makes sense of our habits
of citing knowledge to criticize or justify action. When you know that p, you can count
it in action as well as in belief.
7
Whereas Fantl and McGrath (2002) proceed with some
intuition marshalling, Fantl and McGrath (2009)’s argument for pragmatic encroachment
is supposed to follow directly from our understanding of knowledge (and the related con-
cept of justification) to action.
They start with the following two intuitive theses about knowledge.
(Fallibilism) You can know that p even if there is a non-zero epistemic chance
for you that not-p.
(Purism) If two subjects are just alike in their strength of epistemic position
with respect to p, then they are just alike with respect to whether they are in
a position to know that p.
8
7
Similarly, Hawthorne and Stanley (2008) agree. They note that our our ordinary folk appraisals of the
behaviour of others suggest that the concept of knowledge is intimately intertwined with the rationality of
action, and they suggest the following principle connecting action to knowledge.
The Action-Knowledge Principle. Treat the proposition that p as a reason for acting only if
you know that p.
This principle is offered as a way of making sense of our use of ‘know’ to criticize the actions of others.
Further, in their defence of this principle, they also note that
an analogous principle seems plausible for reasons for belief, viz.: Treat the proposition that
p as a reason for believing q only if one knows that p. One attractive feature of the Action-
Knowledge Principle is that it unifies the practical and theoretical domain of reasons: if it is
correct, then proper reasons for belief and reasons for action have a uniform nature.
So we see the endorsement of something similar to Fantl and McGrath’s unity thesis which says that if p is
warranted enough to be a reason you have to believe, p is also warranted enough to be a reason you have to
act (or anything else).
8
We have been calling this thesis Supervenience. Also, implicit in Purism/Supervenience is a commit-
ment to intellectualism. As Fantl and McGrath explain, when the Purist talks of “strength of epistemic
position” they only mean truth-relevant considerations. How much is at stake for you in whether p is true,
71
Purism does have an intuitive appeal because whether you know is a matter of your rela-
tion to the truth. Should purism turn out to be false, then an unattractive consequence
would be that you could lose knowledge not by losing information or forgetting some-
thing, but simply because of a change in your stakes. However, we cannot preserve both
purism and fallibilism. The price of purism is giving up fallibilism. But fallibilism is also
compelling because it seems to be required in order for us to know much of anything im-
portant and because many of things it seems, intuitively, we know are such that there is a
non-zero chance that they are false. Fantl and McGrath suggest that of these two theses, it
is purism that must go, and it follows from the intuitive plausibility of KJ and Fallibilism.
Their argument goes as follows.
Let’s start by assuming that fallibilism is true. Now suppose that I fallibly know p
where p is that the door is locked. Following KJ, p is warranted enough to justify me in
not doubling back to check that the door is locked. But, because of fallibilism, there’s a
chance that my door is not locked. So, there’s a hypothetical case in which that the door
is locked isn’t warranted enough to justify me in not doubling back to make sure the door
is locked. In the hypothetical case, I must double back. So, in the hypothetical case I’m
not in a position to know that the door is locked, even though in the hypothetical case
I’m in the same strength of epistemic position—alternatively, I have the same strength of
evidence—with respect to the proposition that the door is locked as I am in the actual case
in which I know the door is locked. Therefore, purism must be false.
This is very quick sketch, but we can see the general picture of the approach and can
again set aside the misconception that pragmatic encroachment is only about knowledge.
Knowledge is practically sensitive, according to Fantl and McGrath, because justification
is practically sensitive. So again, we see a broadening of interest-relativity to the epistemic
concepts related to knowledge, and not only to knowledge itself. This again, should be
unsurprising given that knowledge entails justified belief.
9
3.2.4 Explaining Encroachment: Belief and Action
In contrast with the accounts presented above, Ganson (2008), Weatherson (2005), and
Ross and Schroeder (2012) argue that pragmatic considerations get their foot in the door
in virtue of a pragmatic constraint on belief.
whether p would make you happy, etc. are not part of the strength of your epistemic position. If we again
understand evidence for p as broadly meaning anything that raises the probability that p is true, then we can
substitute ‘strength of evidence’ for ‘strength of epistemic position’. But, as we’ll see in Schroeder (2012a),
there is reason to be suspicious of such a substitution.
9
Further, Fantl and McGrath (2009) extend this general framework to argue that there’s pragmatic en-
croachment on justification. That is, we can argue that purism is not only false for knowledge ascriptions,
but so for justification.
72
The arguments for pragmatic encroachment presented by Hawthorne, Stanley, and
Fantl and McGrath are presented in the language of belief and disbelief. As such, they do
not take into account credences or degrees of belief. When we take into account the dif-
ference between outright belief and degrees of belief, Ganson (2008), Weatherson (2005),
and Ross and Schroeder (2012), despite disagreeing about the precise nature of belief, argue
that there is no pragmatic condition on justification, rather, there is a pragmatic condition
on belief.
To expand, knowledge-action accounts, as their name suggest, posit connections be-
tween knowledge and action. The belief-action accounts, however, argue that pragmatic
considerations are relevant only to the question of whether acting as if p is reasonable,
so that means pragmatic considerations are relevant to whether the agent is willing to act
as if p. The question of whether an agent is willing to act as if p is one that’s settled by
whether she has an outright belief. So, there’s a pragmatic constraint on the justification
of outright believing that p. Weatherson (2005) uses this argument to suggest that there
is no pragmatic encroachment in epistemology, rather, pragmatic encroachment is a thesis
for philosophy of mind. He notes,
interests matter not because they affect the degree of confidence that an agent
can reasonably have in a proposition’s truth. (That is, not because they matter
to epistemology,) Rather, interests matter because they affect whether those
reasonable degrees of confidence amount to belief. (That is, because they mat-
ter to philosophy of mind.) There is no reason here to let pragmatic concerns
into epistemology. (pp. 435-6)
Weatherson (2012), however, changes his mind and now argues that the interest-relativity
of knowledge ascriptions goes beyond the interest-relativity of corresponding belief as-
criptions. He motivates this change in light of the important role that knowledge plays
in decision theory, and thus can also be classed in the same group as Hawthorne, Stanley,
and Fantl and McGrath given the connection between decision theory and action. The
difference between Weatherson and Fantl and McGrath is that whereas Fantl and Mc-
Grath argue that knowledge is practically sensitive because the justification condition, or
the rational permissibility condition, on belief is practically sensitive, Weatherson argues
that practical considerations impact knowledge directly.
The takeaway, however, is simply that we can see here again that pragmatic encroach-
ment is not merely a thesis about knowledge. Rather, it can be construed more broadly
as the thesis that non-evidential considerations, such as one’s practical interests or one’s
willingness to act on p, determine the rationality of or justificatory status of one’s belief.
73
3.2.5 Explaining Encroachment: Epistemic6= Evidential
As I have been stressing, construing pragmatic encroachment as merely a thesis about
knowledge ascriptions misses the complexity of the challenges and the questions raised
by the motivating intuition guiding the various pragmatic encroachment proposals offer.
One such question is whether by epistemic considerations we mean only evidential con-
siderations. As noted in my representation of evidentialism, it is plausible to think of
evidence for p as anything that raises the probability that p is true. Further, if we consider
epistemic rationality is the kind of rationality entailed by knowledge and thus constrained
by truth-conducive reasons, then perhaps epistemic considerations are merely evidential
considerations.
The motivating intuition guiding the various pragmatic encroachment proposals, how-
ever, calls this assumption into question. If practical interests can affect knowledge, and
if epistemic rationality is the kind of rationality that is entailed by knowledge, then prac-
tical interests can affect epistemic rationality. But epistemic rationality is more than just
about knowledge, it is about what it is rational and reasonable to believe. So, we see
encroachment encroaching deeper. Practical interests can affect what it is reasonable to
believe. So, when it comes to the rational permissibility of belief, that is determined by
more than strength of evidence alone. Thus, we again seem to challenge the evidentialist
assumption that only strength of evidence determines the rational permissibility of belief.
This line of argument is presented by Schroeder (2012a). Schroeder, following Stanley
and Fantl and McGrath, starts with a pair of cases modelled off of the earlier bank case
from DeRose. However, he removes the knowledge ascriptions to focus on our judge-
ments about what it is reasonable to believe.
Bank-Low. Hannah and her wife Sarah are driving home on a Friday after-
noon. They plan to stop at the bank on the way home to deposit their pay
checks. It is not important that they do so, as they have no impending bills.
But as they drive past the bank, they notice that the lines inside are very long,
as they often are on Friday afternoons. Hannah remembers the bank being
open on Saturday morning a few weeks ago, so she says, ‘Fortunately, it will
be open tomorrow, so we can just come back.’ In fact, Hannah is right—the
bank will be open on Saturday.
Bank-High. Hannah and her wife Sarah are driving home on a Friday after-
noon. They plan to stop at the bank on the way home to deposit their pay
checks. Since their mortgage payment is due on Sunday, they have very little
74
in their account, and they are on the brink of foreclosure, it is very impor-
tant that they deposit their pay checks by Saturday. But as they drive past
the bank, they notice that the lines inside are very long, as they often are
on Friday afternoons. Hannah remembers the bank being open on Saturday
morning a few weeks ago, so she says, ‘Fortunately, it will be open tomorrow,
so we can just come back.’ In fact, Hannah is right—the bank will be open on
Saturday.
Whereas it seems reasonable to believe in the first case, it does not seem reasonable to
believe in the second. To see how this could be the case, consider the following analogue
to Support that Schroeder (2012a, pp. 268) offers.
Support*. It is epistemically rational for S to believe p, just in case p is ade-
quately supported by S’s evidence.
Both Support and Support* leave open the question of what determines the threshold for
adequate support. That is, when is the evidence such that a belief is adequately supported,
and what determines when the evidence is strong enough to adequately support p? It is
here, in this question of what it takes for a proposition to be adequately supported by the
evidence, that non-evidential considerations come into play.
Schroeder argues that in addition to reasons to believe and reasons to not believe,
our intuitions in the cases that motivate pragmatic encroachment suggest another class
of reasons: reasons to withhold. When we look at the cases used to motivate pragmatic
encroachment, we see that they tend to involve some kind of disadvantage that might arise
in forming the belief, in making up your mind. They involve reasons to withhold. Costs
of error are reasons to withhold and these are epistemic reasons that are not exhausted by
the evidence. Why?
Reasons to withhold cannot be evidence because evidence is exhausted by
evidence which supports p and evidence which supports:p. But the evidence
which supports p is a reason to believe p, and the evidence which supports
:p is reason to believe:p. Consequently the reasons to withhold must come
from somewhere else. So they cannot be evidence. (pp. 276)
Finally, it is also here that we see a suggestion for how the pragmatic encroachment chal-
lenge can be seen as not being in conflict with a version of weak evidentialism. As Schroeder
(2012a, pp. 282) suggests,
all of this is consistent with the thesis that it is epistemically rational for S
to believe p just in case S has adequate evidence for p. Reasons to withhold
75
simply raise the bar on how good the evidence needs to be, in order to be
adequate.
On one final note, pragmatic encroachment not only gains its plausibility from these in-
tuitive cases that conflict with the prescriptions of evidentialism, it also gains plausibility
with regard to how it accommodates the two goals we attempt to balance when forming
beliefs—believing truly and avoiding error. As James (1896) notes, our beliefs are governed
by two intellectual goals that often pull in two different directions.
Believe truth! Shun error!—these, we see, are two materially different laws;
and by choosing between them we may end by coloring differently our whole
intellectual life. We may regard the chase for truth as paramount, and the
avoidance of error as secondary; or we may, on the other hand, treat the avoid-
ance of error as more imperative, and let truth take its chance.
10
On the one hand we try to avoid believing what is false, and on the other we try to believe
what is true.
11
These goals, however, can pull us in different directions. In the cases we’ve
been considering that motivate pragmatic encroachment, given the risk of being wrong,
we put more weight on avoiding error than on believing truly. However, sometimes,
given the stakes, we should put less weight on avoiding error. For example, whether or
not climate change is real has high stakes. If it’s real, then we must act immediately to
cut back on carbon emissions and fossil fuel extraction. Doing so, however, would have
a large economic cost. So, perhaps when the studies first surfaced in the 1980s we did
not have sufficient evidence for climate change. However, given the high human costs to
not acting—disastrous hurricanes, monsoons, flooding, droughts, etc.—we should have
put less weight on the goal of avoiding error. Given the need to act immediately to stem
the disastrous effects of climate change we do not need to be 100% certain. So, this cost
and risk to being wrong shifts the threshold for justification. Although in most cases it
seems to push the threshold up, in some cases it can also push the threshold down. So,
we should be careful not to conflate ‘high stakes’ with a higher threshold in every case.
3.2.6 The General Framework of Pragmatic Encroachment
The takeaway of the intuitive plausibility of pragmatic encroachment is that our prac-
tical stakes can determine how good our evidence must be in order to make our beliefs
10
As we’ll see, Pace (2011) develops his account of moral encroachment by drawing on these Jamesian
themes.
11
Schroeder (2012a)’s discussion of balancing type-1 error and type-2 error can be read as a way to attempt
balancing these two competing goals.
76
epistemically justified. Pragmatic encroachment remains, however, controversial. Just as
evidentialism faces the challenge of explaining our intuitive judgements in the cases that
motivate pragmatic encroachment, pragmatic encroachment shares a fair number of chal-
lenges as well. My goal here is not a straightforward defence of pragmatic encroachment.
Rather, I present a new challenge for evidentialism and argue that pragmatic encroach-
ment raises a similar challenge to evidentialism. So, pragmatic encroachment can give us
a framework that can be extended to capture an important aspect of the way in which we
try to hold each other accountable for our beliefs about others.
All of the cases of pragmatic encroachment have focused on trains and banks and how
an agent’s personal preferences can affect what it is epistemically rational for them to
believe. In that regard, all of the cases that have been used to motivate pragmatic en-
croachment focus on beliefs about objects in the world and only concern the risk of being
wrong. However, we do not only have beliefs about objects. We also have beliefs about
each other. When it comes to beliefs about others, we need to be cognizant not only of
the risk of being wrong, but also of the risk of wronging. This is a challenge that the ev-
identialist cannot accommodate. I suggest, however, by developing an account of moral
encroachment following the same general structure as pragmatic encroachment, we can
propose a view that preserves the intuitive plausibility of a weak version of evidentialism
while also being able to say that morality places a constraint on what it is epistemically
rationally permissible to believe about others.
My goal in this section is to now turn to sketching the most innocuous and least-
committal version of pragmatic encroachment I can. Doing so allows us to explore the
challenge pragmatic encroachment raises for evidentialism while taking on as few other
commitments as possible and thus being neutral between the various theories. Although
still controversial, my hope is that by presenting the least-committal version of the basic
pragmatic encroachment intuition we can see the shape of the challenge pragmatic en-
croachment offers against evidentialism, and how it modifies evidentialism in light of the
response. And then in the next section we can explore if the same strategy can be ex-
tended to accommodate moral considerations. This way, we need not commit ourselves
to any particular theory of pragmatic encroachment in presenting a story about moral
encroachment.
As we’ve seen, the pragmatic encroachment challenge to the more restricted accounts
of evidentialism tends to begin with a case in which a subject reasonably seems to know,
be justified in believing, can rationally rely on their belief, etc. For example, consider the
belief that the bank will be open on Saturday. You’d like to deposit your paycheck, but it’s
not time sensitive that you do so. The pragmatic encroacher then presents a contrasting
77
case in which the evidential position of the subject is preserved, but the practical stakes
have changed. That is, in the contrasting case there is a severe cost to being wrong. For
example, it is currently Friday and if you don’t deposit your paycheck before Monday
your account will be overdrawn. In this contrasting case, you’re not sure that the bank
will open on Saturday, but the evidence you have that it will be open on Saturday is the
same as the evidence you have in the low stakes case in which it is not time sensitive
when you deposit your paycheck. Defenders of pragmatic encroachment argue that this
cost to being wrong leads us to the intuition that in a low stakes case you can reasonably
believe that p—e.g., that the bank will be open on Saturday, but that in contrasting high
stakes case—despite no change in the evidential position of the subject—it is no longer
permissible to believe p—that the bank will be open on Saturday. These intuitions clash
with Supervenience and suggest that Supervenience is not as intuitive as initially thought.
Recall Supervenience says that if the two agents have the same evidence, then if one is
justified, so too is the other. In this case, however, our intuitions suggest that the two
agents with the same evidence can differ with respect to whether they are justified.
Note, however, that evidentialism cashed out merely in terms of Support is compatible
with the view that the threshold t is determined, in part, by practical interests. If you have
a strong interest in the truth of p—e.g., there is a large cost to being wrong about p—then
the threshold your evidence must either meet or surpass is higher than when that risk
of being wrong about p is absent. As we’ve seen, in general there is a question of what
determines how much evidence is enough evidence, and all the work in this section has
been to show the simple point that practical interests play a role in settling that question.
We can now begin to see the shape of an answer to the dilemma posed by Agnes and
Spencer; there’s some feature of the scenario—the moral stakes—that changes what she
and he are epistemically justified in believing. To make this answer plausible, we must
first make plausible an extension from pragmatic encroachment to moral encroachment.
3.3 Extending Pragmatic Encroachment to Moral Encroachment
I am not the first to argue that extension can be made. For example, Fritz (2018) argues
that insofar as there are good arguments for pragmatic encroachment, there are also good
arguments for moral encroachment. Fritz (2018, pp. 657-8) notes,
Our beliefs are often insufficiently warranted to allow us to rationally take
certain high-stakes gambles. But our beliefs are also often insufficiently war-
ranted for us to permissibly act on them when something very morally im-
portant hangs in the balance. To the extent that the former phenomenon
78
illustrates pragmatic encroachment on knowledge, the latter phenomenon il-
lustrates moral encroachment on knowledge.
However, there are two issues we’d have to confront if we went the same route as Fritz.
First, the examples Fritz uses are direct analogues to the traditional pragmatic encroach-
ment cases found in Fantl and McGrath et al., and to rest a defense of moral encroachment
on the plausibility of any particular account of pragmatic encroachment risks the plau-
sibility of moral encroachment being undercut by less plausible versions of pragmatic
encroachment. Further, such accounts tend to be motivated by a connection between
knowledge and action (or justification and action). Notice, however, that I have been
careful to motivate the intuitions suggestive of an account of moral encroachment from
general observations about the nature of evidence and justification. After all, we’ve been
noting that beliefs can wrong independent of their connection to action.
12
Pace (2011), on the other hand, offers an argument for moral encroachment on epis-
temic justification that attempts to systematize the Jamesian dictums that we must balance
believing truth and shunning error. Following James (1896), Pace suggests that there is
no purely epistemic rationale for setting the threshold for justification in one place rather
than another. Rather, the evidential standards one adopts, i.e., how one determines the
threshold, is determined by and vary according to pragmatic factors. Similarly, although
Paul and Morton (2018) do not call their account an account of moral encroachment,
they do note that although evidential policies govern the way we adjust our evidential
thresholds in different contexts there is no uniquely best evidential policy to have. There
are multiple evidential policies that are rationally permissible for a given thinker to have
from the point of view of purely epistemic considerations. Further, if that’s the case,
then practical and ethical considerations can and should play a role in deciding between
epistemically permissible policies. Consider, for example, their primary case: grit. A ca-
pacity for grit can be advantageous for an agent. Further, that it would be advantageous
to be resilient to incoming evidence that one will likely fail can bear on the question of
which evidential policies one should reason with. A gritty agent requires more evidence
before she believes she’ll fail than a non-gritty agent; that is, the evidential threshold for
12
To expand, in these standard accounts of pragmatic encroachment practical considerations encroach
because of an intuitive connection between knowledge and/or justification and action. The case that I have
been using to motivate moral encroachment, however, does not concern action. The case concerns our
beliefs about others. This raises worries about whether the account I will be proposing will be a straight-
forward application of pragmatic encroachment, and these are worries that I am sensitive to. Moral stakes,
we will see in Section 3.4.2, encroach in a manner dissimilar to practical stakes. Nonetheless, this remains
a stakes-based encroachment account. It just so happens that these observations are also part of the moti-
vation for accounts of pragmatic encroachment.
79
the question “will I succeed at-ing if I continue to try?Ó is higher.
13
Further, in a manner similar to how I will proceed, Schroeder (2018b) argues that that
moral encroachment is just a special case of pragmatic encroachment. Moral considera-
tions raise the stakes for how much evidence is required to epistemically justify one’s be-
lief. These stakes-related reasons, Schroeder argues, arise from the cost of error. Whereas
traditional accounts of pragmatic encroachment have focussed on practical costs, an in-
sight of accounts of moral encroachment is that moral costs of error are a special case of
costs that aren’t tied directly to action. In the standard diet of cases offered for pragmatic
encroachment, the costs of error that count against belief are the consequences of what
you will do if you rely on the belief. As we’ve seen, if you have a severe peanut allergy
and rely on the belief that your sandwich is almond butter on the basis of insufficient
evidence, you risk a terrible turn of events. In these cases, the primary wrong is one of ac-
tion. However, there can be some moral costs that beliefs carry in and of themselves, i.e.,
independently of their actual or risked consequences. The argument for the way moral
considerations raise the stakes stems from a familiar theme from Chapter 2: interpersonal
relationships are constituted, in part, by what we believe of each other. The effects our
beliefs can have on our relationships is not mediated by the effects of our beliefs on our
actions. The difference, however, between the account I develop in this chapter and the
account put forward by Schroeder concerns whether beliefs wrong even when the belief
is true. I will say yes, Schroeder no. I will argue that what matters is the agent’s belief
forming practices and whether those practices, and the basis for their belief, is matched
appropriately to the stakes of the situation. Sometimes there can be a morally problem-
atic mismatch when the belief is true, and that the belief is true does not let the believer
morally off the hook.
Finally, another account of moral encroachment is offered by Moss (2018, MS). Moss
argues that when we use statistical reasoning about a group to conclude something about
an individual member of that group, our failure to keep in mind the salient possibility
that the individual may be an exception to the statistical generalization is both a moral
and an epistemic failing. However, Moss’s account concerns beliefs formed on the basis of
statistical generalizations and we might worry that there is something objectionable about
the statistical nature of the evidence on which the belief in question is based. One might
argue that what makes the moral stakes high is itself the statistical nature of the evidence.
14
13
Similarly, in defense of epistemic partiality as epistemically rational, Preston-Roedder (2018) argues that
because people are opaque and the evidence we have about them is at best partial and ambiguous, there are
many epistemically permissible responses to evidence. Thus, we are permitted to form judgements about
others that cast them in a more or less favorable light.
14
See, for example, Munton (MS) for a critique of this kind.
80
Although this is another way to present an account of moral encroachment, if we go
this route, we face two questions. First, the question of why statistical generalizations
are uniquely subject to moral considerations, whereas forming beliefs on other kinds of
evidence isn’t.
15
Second, we also face the question of why some beliefs formed on the basis
of statistical evidence seem fine whereas others don’t, i.e., the contrast between profiling
pit bulls on the basis of statistical evidence as opposed to people.
16
For these reasons, the
cases I’m concerned with extend beyond mere statistical generalizations. Instead, I focus
on beliefs that are formed without attention to the moral demands of one’s environment
and spelling out what that means.
Moving forward I will suggest the following intuitive grounding for moral encroach-
ment. I do not expect all readers to share this intuition, but I do expect that most will
recognize the demand, in colloquial terms, to stay woke. To be woke is to be aware of the
moral demands of one’s environment. With regard to our epistemic practices, it is the
demand to be aware of the moral stakes of our beliefs about one another. It is the demand
to be aware of the background against which our epistemic practices exist, i.e., the un-
just world we inhabit, and to ensure that our epistemic practices are not only responsive
to unjust features of our environment but that they also do not themselves contribute
to those unjust features of our environment. Motivated by these considerations, I argue
that the epistemic justification of our beliefs is determined, in part, by the moral stakes.
An upshot of this account is that the epistemic considerations one weighs when trying to
settle the question of what to believe include moral considerations.
3.4 The Moral Stakes of Our Beliefs
To get moral encroachment off the ground, one might be tempted by the following
strategy. Begin with a case in which a subject reasonably seems to know, be justified in
believing, can rationally rely on their belief, etc. In short, a case with low moral stakes.
Then contrast that low stakes case with a morally high stakes case in which the subject
15
Within legal scholarship this is known as the problem of naked statistical evidence. See, for example,
Schauer (2006) and Buchak (2014). The common example goes as follows: we know that a bus has hit an
individual, but the individual cannot identify the bus. 80% of the buses in the city are blue buses operated
by the Blue Bus Company, whereas the other 20% are red and operated by the Red Bus Company. Given
the balance of the probabilities, the individual was probably hit by a blue bus. However, this merely statis-
tical evidence is not permissible in the courtroom. Further, if we had an eyewitness who was only reliable
80% of the time, that testimony would be allowed. Thus, we face a problem of explaining why the testi-
mony is allowed, but the statistical evidence is not allowed. I set aside the issues as they appear in the legal
scholarship, as my focus is not whether there is something objectionable about beliefs formed on the basis
of statistical evidence. Rather, as I’ve stated, my focus is on beliefs that are formed without attention to the
moral demands of one’s environment.
16
To answer this, I do think Moss will need the moral account from Chapter 2.
81
has a belief, but they are failing to respond with sufficient sensitivity to some morally
relevant feature of the environment. Similar to the low stakes case, the belief is supported
by the evidence to the same degree. To explain the difference one is forced to conclude
that the moral features of the agent’s environment are responsible for the change in our
intuitive assessment of whether the belief is reasonable or justified. This is despite the fact
that both subjects in both cases stand in similar evidential positions with regard to the
strength of their evidence. Thus, you get a familiar clash with Supervenience. The moral
features of one’s environment, according to Supervenience, shouldn’t make a difference to
whether it is reasonable to believe. But, just as pragmatic encroachment was compatible
with the general form of evidentialism consisting only of Support, moral encroachment
is similarly compatible. After all, when we are trying to determine how much evidence
is enough evidence with respect to a morally weighty matter such as whether you would
wrong someone through your belief, such a moral consideration can raise the threshold
for the strength of your evidence. Things are not, however, going to be so simple.
3.4.1 Low-Stakes variants
To see this, let us consider Social Club again. Plausibly, Social Club is a high stakes
case. To build a contrasting case, then, we must ask three questions. First, what would
the low-stakes variants of the traditional pragmatic cases be; second, what is it that makes
these cases high stakes; and third, is what makes these cases high stakes the same in both
the pragmatic and the moral domains. To answer the first question, we need a low-stakes
case in which an agent is permitted to believe p and a high-stakes case where the agent is
not permitted to believe p. The difficulty of building minimally contrasting low-stakes
variants for moral considerations will in turn also help us answer the second and third
questions. That is enough in the way of suggestive remarks, let us turn to considering the
following attempt to construct a low-stakes case and a contrasting high-stakes case.
The Enlightened Northerner. Nathaniel is an enlightened visitor from the
North who likes to keep track of how well-staffed social clubs are. He has
found that in the South, staff are predominantly black. During his visit to
the Cosmos Club, he mentally makes a note of each black person he sees as a
staff member, and at the end of day, he leaves for the rest of his journey with
the belief that there was one staff member for every ten guests. One of those
people he counts as a staff member is John Hope Franklin, the club’s first
black club member.
The Racist Club Member. Duke is a racist member of the Cosmos Club. Ser-
82
vice has been bad today, so he wants to reprimand a staff member in front
of a lot of people in order to call attention to his dissatisfaction with the ser-
vice. Furthermore, he wants the race of the staff member to be a major focal
point of his complaint. Duke sees a well-dressed black man standing off to the
side, and, being a racist, assumes on the basis of his race that the staff member
must be lazy and deserving of punishment. Duke then proceeds to confront
the man in an attempt to make an example of him.
Despite the acts (and beliefs) of Duke being much worse than Nathaniel, I argue that it is
still wrong of Nathaniel to so easily believe of every black person in the club that they are a
staff member. Nathaniel is ignoring how his environment is shaped by the South’s racism,
which here serves as the relevant moral feature of his environment that he is insensitive
to. This additionally makes him ignorant of the way he wrongs others by forming beliefs
about them on the basis of facts that are due to racism. It is for this reason that both Duke
and Nathaniel would still fail to be justified even if the person they picked out is in fact
a staff member. Both fail to show the moral care demanded by the unjust nature of the
environment they find themselves in. This is not to say that there is no scenario in which
they are justified in believing that the person picked out is a staff member. Only that the
threshold for justification is higher for that belief in the environment that is the social
club than it is in other environments. The reason for that being the history of racism that
has resulted in race being a reliable tracker for staff members.
To reiterate, both Nathaniel and Duke fail to show the moral care demanded by the
unjust nature of the environment they find themselves in. Whereas the facts may not
themselves be racist, the facts that we often rely upon can be the result of racism and racist
institutions and policies. Thus, when forming beliefs on the basis evidence that is a result
of our racist history, it seems appropriate to ask for more moral care. But what, precisely,
does this moral care amount to? It cannot simply be recognizing that the environment is
racist. That doesn’t seem to be enough—for example, you might think that I can recognize
that fact, but then use that very fact as evidence that the black person I’m seeing in the
club is a staff member. So, being sensitive to it isn’t merely a matter of knowing that we’re
in a racist environment. Moral sensitivity amounts to not only a mere recognition of an
unjust environment, but also an adjustment of one’s threshold for justification to meet the
moral stakes of one’s environment. This is not a fully satisfactory answer, in part, because
it turns on the deeper question of what being a moral person requires of us. Nonetheless,
we see the following constraint in operation: given the background against which our
epistemic practices exist, i.e., the unjust world we inhabit, our epistemic practices should
be responsive to unjust features of our environment.
83
To illustrate with another analogy, consider the following case. You study African-
American history and you are putting together a book on poverty in 1930s America.
Your focus is specifically on poverty in black communities, so you are visiting archives
and collecting photos. The 1930s were a difficult time, and you have very good reason
to think that almost any picture you come across of an African-American family will be
a picture of a family living in poverty. However, one of the pictures you include is of
an African-American family not living in poverty. Contrast this case with one in which
you are a photographer putting together a book on North Dakota ranches and one of the
pictures included in the book is mistakenly a picture of a South Dakota ranch. There is
a wrong in the former case that is different from the latter. It is not merely a matter of
presenting things accurately. The problem in the first case is that there exists a neglected
demand for extra care and diligence when believing things of others. This same demand
is not present when believing things of ranches. I suggest that what is going to separate a
high-stakes case from a low-stakes one is that the wrong that’s being done has to do with
the added layer of a moral injustice—in this case taking the form of a racial injustice.
In answering the first question—what would the low-stakes variants of the traditional
pragmatic cases be—the difficulty we face in creating minimal contrasting low-stakes cases
for moral encroachment is due to the very features that makes these cases high stakes
cases. This, in turn, leads us to an answer to the second question I set forward at the start
of this section—what is it that makes these cases high stakes. Underrepresented groups
are more often mistaken for employees because of the color of their skin and the racist
institutions that make their skin color such a big factor in their inability to gain access
to more prestigious employment opportunities. Being mistaken in this context, namely
one in which you’ve historically been excluded, is a greater harm and wrong than being
mistaken in a space where that historical disadvantage is lacking. It is this social and moral
fact that we’ve been neglecting. It is this social and moral fact that makes all the cases so
far seem like high-stakes cases. I conjecture that this is because of how deeply our social
environments are steeped in and shaped by a history of racism. So, whereas we can toggle
high stakes for agents in traditional pragmatic encroachment cases by stipulating that they
have mortgage payments due, we cannot simply toggle a history of racism on or off to
make simple high/low pairs of cases.
To envision such an environment for a low-stakes variant would require invoking
something like a twin-earth type scenario where our society hasn’t been shaped by the
racist events and practices. That instead, this imagined world is such that skin color is as
irrelevant to one’s social status as[fill in the blank]. Note that ‘[fill in the blank]’ is not an
editing error. I challenge the reader to fill in the blank themselves and consider whether
84
such a society isn’t similarly shaped by racist events and practices. For example, suppose
we were to fill in the blank with hair colour, height, the ability to roll one’s tongue, the
length of one’s index finger, etc.
17
Plausibly, in such a world any of these options could
be relevant to one’s social status and the opportunities that one is afforded. Not only am
I suspicious of conclusions drawn from highly artificial twin-earth type thought experi-
ments, but also I think to imagine such a world to create a low-stakes variant is to so widen
the distance between it and the actual world that the comparison would ultimately have
no substantive relevance, thus failing to constitute anything like a “minimal pair”.
Bolinger (2017), however, offers an alternative proposal. Consider the following case.
Target-Low Stakes
Agnes and Esther are shopping at Target. Esther wants to purchase the newest
iPod for his nephew, and so she needs to find a staff member to unlock the
case containing the iPods. They reluctantly wait for a free staff member until
Agnes points to a black man dressed in a red shirt and khakis and tells Esther,
“There’s a staff member that can unlock the iPod case.”
Note that the evidence that Agnes relies on in Target-Low Stakes is importantly different
from the evidence that Agnes relies on in Social Club in the following regard: wearing
khakis and a red shirt is an easily avoidable signal whereas one’s skin colour is not. In
order to not be mistaken for a staff member in Target, one simply needs to not wear
khakis and a red shirt.
18
Bolinger argues that if John Hope Franklin were to wear khakis
and a red shirt to Target, his
voluntary performance of an avoidable behaviour whose social meaning is to
signal that[Agnes] is permitted to assume that[John Hope Franklin] effec-
tively waives[his] standing to complain against[Agnes] accepting that[he is
a staff member].
Similarly, if I were to go to the Cosmos Club dressed in an attendant’s uniform, if someone
were to see me and assume I was a staff member, Bolinger argues that the epistemic and
17
Perhaps it would be wrong to call these ‘racist’ events and and practices in the imagined world. Perhaps
they are instead, hair colorist, heightist, etc. practices. But, that is beside the point.
18
Of course, it is not that simple. For example, Michelle Obama retells in a People magazine interview,
I tell this story—I mean, even as the first lady—during that wonderfully publicized trip I took
to Target, not highly disguised, the only person who came up to me in the store was a woman
who asked me to help her take something off a shelf. Because she didn’t see me as the first
lady, she saw me as someone who could help her. Those kinds of things happen in life. So it
isn’t anything new.
85
moral wrongs of the motivating cases would be absent. Although Agnes’s beliefs are still
based on a generalization, they’re based on generalizations regarding permissible signals
(e.g. people who wear staff uniforms) as opposed to morally impermissible signals (e.g.
one’s skin colour). What is crucial to the alternative diagnosis is that wearing a uniform
is a public and avoidable signal that one is a staff member. One’s membership in a racial
group, however, is not a permissible signal. So, in such a case as the following, any moral
or epistemic complaint is waived.
Lark. For a lark, John Hope Franklin attends the Cosmos Club dressed in an
attendant’s uniform. Seeing him, a woman assumes he is a staff member and,
handing him her coat check, demands her coat.
Although I would be happy to accept the Bolinger point regarding public and avoidable
signals affecting the degree of the wrong, it is important to recognize that the historical and
social features of an environment in addition to whether the signal is public and avoidable
play a role in determining whether someone is wronged by the belief.
Further, although in Lark there may be less of a moral wrong than in the other Social
Club cases, there is still some moral wrong here. We cannot ignore the social environment
of the Cosmos Club. For example, imagine instead that it is Esther that puts on a staff
member’s uniform in order to play a prank on Agnes. If she succeeds in being so mistaken,
she will not be wronged to the same degree that John Hope Franklin would be wronged
if he partook in the same prank. We have good reason to suspect that even if she were
wearing a staff uniform and John Hope Franklin were not, in the context of the Cosmos
Club someone is more likely to ask John Hope Franklin for their coat than they are to
ask Agnes.
19
Bolinger’s proposal, then, can help us determine the degree of the wrongs, but it must
still be supplemented with a story about the morally relevant features of the environment
and how those features affect the orientation we ought to take towards the evidence when
forming beliefs about groups that have suffered historical and ongoing injustices in those
environments. Supplementing the story in this way also helps with the story regarding
which signals are morally permissible. For example, although it is tempting to think that
clothing is an avoidable signal so you can make predictions about someone on the basis of
their clothing, there are worries about the racialized nature of certain articles of clothing.
As I have argued previously in Chapter 2, we owe everyone the right to not see them as
objects to be predicted on the basis of statistical information about them, whether it’s their
clothing, skin colour, sexual orientation, etc. However, in certain practical environments
19
Recall Michelle Obama’s story of being mistaken for an employee in Target.
86
we might owe more than in others. That is why in Target we owe less than we owe in
Cosmos Club. It is the situation, not the avoidability or public nature of the signal, that
triggers the duty and the strength of the duty. What we’ve seen is that there are many
different ways in which our beliefs can wrong. Some beliefs wrong in virtue of not taking
the moral standpoint, some beliefs wrong in virtue of ignoring the risk of being wrong,
some beliefs wrong in virtue of being based on non-avoidable signals.
3.4.2 What are moral stakes?
One might wonder at this point whether these moral stakes are simply practical stakes.
If so, it might seem that moral encroachment is a mere extension of pragmatic encroach-
ment, rather than a view that brings with it anything novel, surprising, or distinctive of its
own. This brings us to the third question that I set out at the start of this section: is what
makes these cases high-stakes the same in both the pragmatic and the moral domains? Ul-
timately I will reject the idea that moral encroachment is a mere extension of pragmatic
encroachment. However, for now, it will be helpful to explore this possibility. One way
to conceive of moral encroachment as a version of pragmatic encroachment is to consider
it an extension of practical stakes to others rather than ourselves. For example, consider
the following first-pass distinctions between the two versions of the view:
Practical Stakes. A risk of being wrong you impose on yourself when you
believe on the basis of insufficient evidence.
Moral Stakes (First Pass). A risk of being wrong you impose on others when
you believe on the basis of insufficient evidence.
Practical stakes concern the agent’s preferences, whereas moral stakes take into account
other people’s preferences. This is a point that has already been recognized. For example,
consider the following pair of cases from Ross and Schroeder (2012, pp. 261).
Almond Butter-Low. Five minutes ago, Hannah made three sandwiches and
placed them in the refrigerator. She told Sarah that she placed the peanut but-
ter sandwich on the left, the tuna sandwich in the middle, and the almond
butter sandwich on the right. Hannah then departed just as Sarah’s friend
Almira arrived for lunch. Sarah knows that Almira has no allergies. Almira
says: “I’d love an almond butter sandwich.” And so Sarah opens the refriger-
ator door, points to the sandwich on the right, and says: “The sandwich on
the right is an almond butter sandwich. You can have it.”
87
Almond Butter-High. This case is just like Low, except here it is Sarah’s nephew
Algernon who is visiting for lunch, and he has a severe peanut allergy. He asks
Sarah for a sandwich. Sarah knows that the peanut butter sandwich would be
fatal to Algernon, but that the almond butter sandwich would be harmless.
She also knows that he would slightly prefer the almond butter sandwich to
the tuna sandwich. When Sarah goes to the fridge, she can tell by visual in-
spection which is the tuna sandwich, but she cannot tell by visual inspection
which is the peanut butter sandwich and which is the almond butter sand-
wich. So she gives him the tuna sandwich.
The contrasting cases above serve as illustrations of how it is not only your own stakes
that can change whether you are justified in believing p, but also whether someone else is
in a high stakes situation with regard to p can change your justification as well.
In both cases, Sarah has the same evidence and the evidence supports the proposition
that The sandwich on the right is an almond butter sandwich to the same degree. If we
were still fond of more restrictive accounts of evidentialism we might be tempted to say
that if Sarah is justified in believing that the sandwich on the right is an almond butter
sandwich in Almond Butter-Low, she is also justified in believing that the sandwich on
the right is an almond butter sandwich in Almond Butter-High (as we’ve seen, this follows
from Supervenience). However, given the risk involved with acting on that belief not for
herself but for Algernon, our intuitive response is that Sarah is not justified in believing
that the sandwich on the right is an almond butter sandwich. So, despite being in the same
evidential situation Sarah is justified in Almond Butter-Low and not justified in Almond
Butter-High. As a result, it appears that not only do the stakes for the believer make a
difference to whether she is justified, but the stakes for those who would be affected by
her acting on the belief make a difference to whether she is justified. But if this is all that
moral stakes are, this is uninteresting.
Instead, as I have been suggesting, what is morally at stake in the cases we began
with is an added insensitivity to background features of one’s environment. This in-
volves a broadening of our understanding of how our epistemic practices are environment-
sensitive. There is a morally and epistemically problematic form of insensitivity that I
have outlined earlier. This sensitivity is distinctive to our moral agency and it is what
separates cases of high practical stakes concerning other people from high moral stakes
concerning others. Thus, it’s critical that we capture this potential for insensitivity in our
gloss of moral stakes. Here’s a second-pass attempt to capture this point:
Moral Stakes (Second Pass). The risk one imposes on another person by failing
88
to respond with sufficient sensitivity to the morally relevant features of the
environment when settling a belief.
This gloss on moral stakes focuses attention on what the non-evidential features relevant
to settling belief in these cases are. Namely, the relevant features are moral facts that
you should be sensitive to. Saying this, however, is pretty empty. It is like answering the
question “when do we behave immorally?” by saying “we behave immorally when we fail
to respond with sufficient sensitivity to the morally relevant features of the environment
before acting.” This is not a helpful answer. Obviously what one wants to know are what
the morally relevant features are that we should be responding to. This is where the work
of Chapter 2 will help us. The work of Chapter 2 was an attempt to articulate the morally
relevant features we should be sensitive to.
Now, although this bears some similarity to the idea of practical stakes, it cannot be
subsumed under our understanding of practical stakes. Recognizing that our environ-
ment can place moral demands on our epistemic practices results in a more robust form
of pragmatism than is commonly defended. As a result, we should not think of moral en-
croachment as merely a straightforward application of pragmatic encroachment; rather,
it builds upon the foundations laid by pragmatic encroachment and extends our under-
standing of the non-evidential considerations that matter for justification.
20
20
There is a point to be made here about how the cases that are traditionally used to motivate pragmatic
encroachment rely on A-stakes (see Worsnip 2015). A-stakes track what is at stake with respect to whether
an agent believes (or relies) on the proposition. On the other hand, there are W-stakes which track what
is at stake respect to whether the proposition is true or false. That is, the relevant cost is the cost of the
proposition of the potential belief being true versus being false as opposed to the cost of relying on the be-
lief should it be false. Worsnip presents the following challenge for accounts of stakes-based encroachment:
only a W-stakes reading preserves intuitions about knowledge attributes, whereas only an A-stakes reading
preserve the putative link between knowledge and practical reasoning that motivates many pragmatic en-
croachment accounts. But as we see in light of the cases of interest to this chapter, not only does it seem
that in these cases it is the A-stakes that seem relevant thus complicating Worsnip’s claim that W-stakes are
the only stakes that genuinely relevant for knowledge, but also moral stakes complicate the relationship
between W-Stakes and A-stakes.
Perhaps in addition to W-stakes and A-stakes there are also M-stakes that don’t quite operate in the same
way as either W-stakes or A-stakes. As noted above, W-stakes are the stakes as to whether the proposition
is true, whereas A-stakes are the stakes as to whether the subject ought reliantly believe or withhold from
believing p. M-stakes are neither quite W-stakes or A-stakes. After all, in Social Club, it doesn’t make sense
to say that the W-stakes are high or low. Partly why this is difficult is because Worsnip only presents A-
stakes and W-stakes as they pertain to you, not to others. For example, you might think that we perhaps
could capture the stakes for others in terms the W-stakes. But, Worsnip’s W-stakes only consider how the
world turning out a certain way rather than another would be bad for the believer. Hence, it is unclear
to me what the mapping between A-stakes, W-stakes, and moral stakes would be. A conclusion we can
draw, however, is that the matter of stakes and what kind of stakes there are, or more difficult than initially
thought. There are not only plausibly two kinds of stakes, but potentially three. Thanks to Jessie Munton
for pressing me on this point.
89
3.5 Three Further Worries
Finally, I turn to three objections that can be raised against this account of moral
encroachment. First, that moral encroachment is an incomplete account of the ways in
which beliefs can wrong, i.e., of doxastic morality. Second, that moral encroachment has
perverse consequences. Third, that moral encroachment is too demanding.
3.5.1 The Incompleteness of Moral Encroachment
As Begby (2018) has argued, encroachment-style views are premised on the idea that
we can resolve this conflict between epistemic and moral considerations by showing how
moral considerations can affect an agent’s evidential policies. However, even if the moral
stakes or practical considerations determine the evidential threshold a belief must meet,
that is a constraint on epistemic processing, not a constraint that operates directly on be-
lief itself. As a result, an agent could adopt the morally required evidential policy, adjust
their threshold to meet the moral stakes of their situation appropriately, and upon re-
ceiving enough evidence, be justified in forming the problematic prejudiced beliefs. In
that case, the encroachment theorist must either say that the morally objectionable belief
is no longer morally objectionable or, if it is still morally objectionable, it doesn’t mat-
ter because the threshold has been surpassed. Further, as Marusic and White (2018) have
also noted, there is a redundancy problem here. If the belief is genuinely wrong, then
the moral questions should remain on the table after the epistemic questions have been
settled. Encroachment-style views, then, are incomplete accounts of doxastic wronging.
I have noted previously, however, that there are many different ways in which our
beliefs can wrong. Some beliefs wrong in virtue of not taking the moral standpoint, some
beliefs wrong in virtue of ignoring the risk of being wrong, some beliefs wrong in virtue
of being based on non-avoidable signals. My goal, as I have repeatedly noted, has not been
to give necessary and sufficient conditions of when beliefs wrong or a taxonomy of beliefs
that wrong. The goal of this dissertation has been more limited, it has been to engage in
a particular dialectic with the supposedly rational racist. The goal has been to show how
Spencer, by his own lights, fails to live up to the demands of epistemic rationality. This
account of moral encroachment is one that Spencer should be able to accept. Spencer
should recognize that when it comes to beliefs about his friends, his evidential threshold
for believing negatively of his friends is higher than for strangers. Given that it is rational
to have higher thresholds in some contexts depending on what you morally owe to the
folks in that context, it is epistemically rational for him to set a higher threshold for beliefs
about non-dominantly situated groups. Just as he wrongs his friend when he believes on
90
insufficient evidence that his friend cut someone off in traffic, so too he wrongs the black
diner in his section when he believes that the diner won’t tip well.
I do leave it open, however, that this threshold can be met, and when the threshold
is met the belief no longer wrongs. But, I do not think it then follows that the morally
objectionable belief is no longer morally objectionable. There is the belief formed on the
basis of insufficient evidence and the belief formed on the basis of sufficient evidence. One
wrongs, the other doesn’t. To determine which belief wrongs is a matter of how the agent
responds to the moral stakes of their situation. The moral concerns are not redundant,
they settle the question of what is sufficient.
3.5.2 Perverse Consequences
Here are two seemingly perverse consequences that this account of moral encroach-
ment might have. First, consider people who grow up in deeply prejudiced social settings
with no access to contrary evidence. As Begby (2018) argues, these people should be
counted as victims. If you had the bad luck of growing up in a severely constrained socio-
epistemic environment, then holding you responsible and morally blameworthy for the
predictable consequences of your limited epistemic opportunities would be perverse. It
would compound your plight.
Second, consider what this account of moral encroachment would recommend when
it comes to believing things of any member of non-dominantly situated group. You must
walk on epistemic eggshells.
21
However, by only walking on epistemic eggshells around
members of non-dominantly situated groups and being more free in your epistemic prac-
tices around members of dominantly situated groups you may end up compounding a
wrong that’s done to them. Relatedly, whereas I have focussed on cases in which we wrong
by believing, when it comes to members of non-dominantly situated groups it seems we
more often wrong them when we refrain from believing or when we fail to believe.
To address the second challenge first let us consider an example: when we fail to believe
on the basis of a woman’s testimony that she has been sexually harassed. It would be
inappropriate to say that it’s a case in which the stakes are high and so the threshold for
settling belief is also higher. What this suggests, however, is that the moral stakes can not
only raise the threshold for evidence, but also lower it. I am skeptical, however, that we’ll
be able to articulate a general principle to determine in every case whether the threshold
goes up or whether the threshold ought go down. The very same class of evidence—fact
patterns that are the basis or result of an injustice—and the very same way of relating
to others—taking the moral standpoint—can sometimes make the threshold go up and
21
Thanks to Adrienne Martin for suggesting this phrase.
91
sometimes down. For example, it is easier (and perhaps ought be) easier to believe well of
our friends. The threshold is lower. Similarly, perhaps to address the epistemic eggshells
concern, when it comes to some situations we should be more ready to believe things of
vulnerable groups than otherwise.
Now, to return to the first version of the perverse consequences worry, let me start
by noting that although this is a serious challenge I think it brings to light a promising
feature of taking the thesis that beliefs can wrong seriously. To see this, let me start with
a biographic note. I did not learn about residential schools until I attended university. In
no history textbook, in no class or conversation was the fact that First Nation, Inuit
and Métis children were removed from their homes and communities in the name of
“assimilation” ever taught to me. The way in which I was taught, the way in which all
Canadian children were taught, was (and continues to be) constrained by a socio-epistemic
environment that deeply wrongs First Nation, Inuit and Métis peoples (see Mills 2007).
An account of doxastic morality that holds people responsible and morally blame-
worthy for the predictable consequences of our limited epistemic opportunities might be
perverse. But, these wrongs at the level of the believer must be balanced against the fact
that we must also do justice to this ongoing wrong that occurs when history is erased or
forgotten. The wrong to First Nation, Inuit and Métis children is not just a wrong in
how the Canadian government acted. If we do not accept responsibility and blame for
our beliefs that are the result of forgetting or miseducation, we continue to contribute
to a terrifying epistemological reality for the First Nation, Inuit and Métis peoples: we
devalue their lives and experiences (see also Dotson 2018).
3.5.3 Too Demanding
Finally, let us turn to the demandingness worry. If one will fail to be justified in virtue
of failing to appreciate the burden and risks that they impose on another, then almost all
beliefs about other people—especially any belief about a person on the basis of their race or
another protected class—are going to be high stakes. This seems overly demanding because
it requires moral agents to be fairly sophisticated in recognizing when they should occupy
this kind of moral standpoint. As a result, if this level of sophistication is beyond the reach
of most, then most (maybe all) of our beliefs about individuals from non-dominantly
situated groups are unjustified. After all, it is not clear whether most people would ever
find themselves in a position to meet such a requirement. I will address these concerns
about sophistication and demandingness by first motivating that it makes sense why such
scenarios ought be considered high-stakes and next say a bit addressing why this isn’t
problematic. To do so, we will find a friend in one of the classic evidentialists: Clifford
92
(1877).
Clifford is a standard example of an evidentialist, but he is an evidentialist who rec-
ognizes the very same puzzling interaction between beliefs and stakes that defenders of
pragmatic encroachment attempt to capture. Clifford argues that the rightness or wrong-
ness of a belief turns on not solely whether it is true or false, but rather whether the agent
has a right to believe on the evidence before her. To motivate his thesis he gives us the
example of a shipowner who casts aside doubts about the seaworthiness of his ship. The
shipowner convinces himself that because the ship has returned home safely so many
times, he need not worry about whether it would return safely home from this trip as
well. But, supposing that the ship does go down, what do we say of him? Clifford sug-
gests that the shipowner is guilty of the death of people on his ship. Further, even if he
sincerely and with conviction believed that the ship was seaworthy, “he had no right to
believe on such evidence as was before him.” Given the high stakes—the possible deaths
of the passengers should the ship not be seaworthy—he was not justified in believing that
the ship was seaworthy. Further, Clifford argues that even if the ship were seaworthy and
the ship had made its voyage safely, that would not diminish the guilt of the shipowner
“one jot.”
22
Clifford understands belief as occupying a special social role. It is never the case that
any belief is merely trivial or insignificant.
23
He famously remarks that it is wrong always,
everywhere, and for anyone, to believe anything upon insufficient evidence. This is a duty
we have to all of humanity. Further, if you believe irresponsibly, i.e., without sufficient
evidence, then you have sinned against humanity. This moral concern about the social
nature of belief leads to an epistemic constraint on how beliefs ought to be formed. It
is not just important to the goal of believing the truth that we believe on the basis of
sufficient evidence, but also it is our moral duty to do so.
24
Clifford’s view, given how extreme it is, has become an object of ridicule. Famously,
James (1896) referred to Clifford as that delicious enfant terrible. James accuses Clifford
for taking “Shun error!” to be of utmost importance because Clifford is afraid of looking
like a dupe if he believes a falsehood. Perhaps James is right, but that is a question for
James-Clifford scholars. For the purposes of this chapter, we can latch onto the following
thought that we also find echoed by Clifford: there are moral costs associated with be-
lieving that should drive us to be more cautious. Given the picture we’ve been working
with, we can see how the moral demands we place on one another extend beyond sim-
22
We can see Clifford as one of the early proponents of Uniqueness with this idea.
23
Clifford is rightly mocked for this extreme feature of his view. After all, surely the belief that there are
an odd number of leaves on the tree outside my window is a trivial or insignificant belief.
24
See Feldman (2004) and Mitova (2008) for more on this point.
93
ply how we expect others to treat us, but also what others believe of us. All that moral
encroachment requires is a certain level of epistemic humility when it comes to beliefs
about others.
At this point, you might be willing to agree that this demanding requirement is un-
warranted, but still worry that it requires moral agents to be too sophisticated. Thus, it
is demanding in an additionally problematic way: the standard is too difficult to meet.
However, given that moral encroachment brings moral considerations to bear on our
epistemic practices, this should not be surprising. Morality is demanding. Recall that if
we understand moral encroachment as a systematic treatment of the imperative to stay
woke, staying woke is demanding. One does not simply wake up and stay woke. Rather,
it is something that one has to work at everyday. Morality is much the same.
It should not be surprising then that a moral constraint on our epistemic practices
would be similarly demanding. In our everyday lives and our day-to-day beliefs we may
often fall short of the moral and epistemic ideal. The ideal, however, exists as a standard
we ought to strive to meet nonetheless. Being a good person is a moral project that is never
complete. Hence an upshot of this account of moral encroachment is that the question of
what individuals should believe is, in part, a moral one. The promise of moral encroach-
ment is a promise of reconciliation. Our moral projects and our epistemic projects do
not stand at odds with each other. Sensitivity to the moral demands of one’s situation not
only makes an agent a better moral agent, but a better believer as well.
94
CHAPTER IV
Assessing Alternatives
Our stalking horse has been the supposedly rational racist, Spencer. Spencer wants
to know what’s wrong with believing that the black diner in his section won’t tip well.
The troubling question raised by Spencer, the question that has been the motivation of all
the work in this dissertation so far, is the following. If seemingly racist beliefs just reflect
reality, what’s the way in which they could be either morally or epistemically wrong?
The task of Chapter 1 and Chapter 2 was to answer the first disjunct. That is, the task
was to show that racist beliefs—even when seemingly supported by the evidence—could
morally wrong. Spencer, however, could concede this point and note that although he
morally shouldn’t believe, his belief is nonetheless epistemically rational. Thus, there is
still a sense in which he should believe the morally objectionable belief. Further, as we’ve
been presenting Spencer, what Spencer claims to care most about is being epistemically
rational. To answer this, the task of Chapter 3 was to show that Spencer’s belief is not
even okay in the respect he cares most about, i.e., the belief is not epistemically justified.
As I argued in the previous chapter, if there is good reason to think that non-evidential
considerations can affect the threshold we set for when a belief counts as justified, it is also
plausible to think that moral considerations are amongst those non-evidential considera-
tions. According to this account of moral encroachment, in ignoring the moral demands of
his situation, i.e., the relevant non-evidential considerations that affect where we place the
threshold for when a belief counts as justified, Spencer fails to see that there is no answer
to the question of what makes his evidence sufficient to believe on which he has suffi-
cient evidence to believe. Thus, his belief is both morally objectionable and epistemically
irrational.
Moral encroachment, however, is not the only way to answer the challenge raised by
Spencer. The task of defending moral encroachment is incomplete until I show why it
was worth developing that account rather than any other plausible account of what to do
in cases of normative conflicts—i.e., conflicts between oughts. The tasks of this chapter,
95
then, are primarily organizational and clarificatory. With regard to the organizational
task, I will present a number of alternative paths we could have taken when presented with
what appear to be apparent conflicts between the moral ought and the epistemic ought. I
will assess each alternatives’ strengths and weaknesses, and consider the answer it offers to
the problem of the supposedly rational racist. Doing so also fulfills the clarificatory task
of this chapter. In presenting, assessing, and comparing the alternatives with the answer
I give in Chapter 3, my goal is to make clearer the motivations and reasons for preferring
the moral encroachment view to the alternatives. So, by the end of this chapter we’ll
have not only a satisfactory answer to the supposedly rational racist, we’ll also see why
the answer given by moral encroachment—that there is no real conflict between the moral
and the epistemic ought when it comes to the question of what one should believe—ought
be preferred to the alternatives.
4.1 Conflict Creation and Conflict Resolution Strategies
The two theses and an assumption that set us up with a conflict are the following.
First, according to the evidentialist—whom we discussed at length in the previous
chapter—all there is to the rationality of belief, i.e., epistemic rationality, is ensuring that
you have the attitude that you ought have towards p. The attitude you ought have is
one that is determined by whether p is supported by the evidence. Intuitively, this is a
compelling account of what we mean when we say you ought believe something. For
example, if I report that there’s a 92% chance it’ll be sunny tomorrow, there is a sense in
which you ought believe that it will be sunny tomorrow.
A related assumption that the traditional evidentialist smuggles into their account con-
cerns what we mean by ‘epistemic’. That is, if we were more careful in the earlier presen-
tation of the evidential thesis, the evidentialist thesis is not simply that the attitude you
ought have is one that is determined by whether p is supported by the evidence but rather
that the attitude you epistemically ought have is one that is determined by whether p is sup-
ported by the evidence. This assumption is smuggled in because it doesn’t ordinarily need
to be defended. If we were to step back and think about epistemology as it is traditionally
conceived, i.e., as the study of knowledge, then the subject matter of epistemology is the
factors in virtue of which a belief is an instance of knowledge.
As Stanley (2005) notes, typically these factors are those that are truth-conducive, i.e.,
factors that make the belief more likely to be true. If we return now to the question of
what you should believe and the rationality of your beliefs, i.e. epistemic rationality, what
we look for are truth-conducive factors to answer the question of what one should believe.
96
This shouldn’t be surprising if we understand epistemic rationality as the kind of ratio-
nality that is entailed by knowledge. If it is truth-conducive factors that are relevant to
whether a belief is an instance of knowledge, then similarly those same truth-conducive
factors are relevant to whether a belief is epistemically rational. As a result, if moral
encroachment argues that what makes a true belief into knowledge (or justified) is not en-
tirely a matter of truth-conducive factors but rather a matter of moral and practical factors
as well, then on this conception of ‘epistemic’ it sounds like what the moral encroacher is
committed to is rejecting this assumption. That is, what makes true belief into knowledge
is not entirely an epistemic matter.
But, there is more than one sense in which we can use the term ‘epistemic’. As Stanley
(2005) notes, there are two senses in which epistemologistÕs use the term ‘epistemic’. The
first use we saw has to do with factors relevant to whether a true belief is knowledge. Call
this ‘epistemic
K
’. Alternatively, ‘epistemic’ can also be used to denote truth-conducive
factors that make the belief more likely to be true. Call this ‘epistemic
T
’. So, when we
use the term ‘epistemic’ we must be careful not to conflate epistemic
T
and epistemic
K
.
Recall that I previously noted that the subject matter of epistemology is the factors in
virtue of which a belief is an instance of knowledge and that typically these factors are
those that are truth-conducive. This leaves it open that what makes a true belief into
knowledge or makes a true belief rational or justified is not entirely an epistemic
T
matter.
Returning to the evidentialist, in the previous chapter we saw that the evidentialist defines
evidence in support of p as reliable signs, symptoms, or marks of that which it is evidence
of, e.g., reasons that raise the probability of p being true. So, we see that the traditional
evidentialist restricts epistemic rationality, i.e., the rationality of our beliefs, to being an
epistemic
T
matter. Moral encroachment, however, is able to recognize a role for both uses
of epistemic.
The second thesis that sets us up for a conflict is the thesis that beliefs can wrong.
If a belief wrongs, then there is a sense in which you ought not believe. In general you
shouldn’t violate the demands of morality and the sense in which you ought not believe
is this moral ought. It follows, then, that if believing that someone won’t tip well given
the colour of their skin would violate a moral demand, then Spencer, our supposedly
rational racist, shouldn’t believe that the diner in his section won’t tip well. So, we find
ourselves facing the following normative conflict: Spencer ought (epistemically) believe p,
but Spencer ought (morally) not believe p. When the specter of normative conflict looms,
what do we do?
In light of this conflict, in the previous chapter I noted that there are five ways one
could go. First, we could explain away the intuitive plausibility of evidentialism insofar
97
as it seems responsible for generating this kind of conflict. Alternatively, we could show
that there is in fact no conflict between the moral and epistemic ought by rethinking and
questioning the assumptions that seem to pit them together (as I argued in Chapter 3).
Third, perhaps we could show that although there is a conflict between epistemic and
moral demands, there is a way of adjudicating between them such that we can say there is
something Spencer and Agnes ought all-things-considered believe. Fourth, we could deny
that there’s anything that Spencer or Agnes ought all-things-considered believe while also
denying that that means we must resort to the pessimistic conclusion that this conflict is
irresolvable. That is, when oughts seem to conflict there is no real conflict, there is one
sense in which you should do x and a different sense in which you should do y. Finally,
fifth, we could just endorse the pessimistic conclusion that there is way to determine what
to do in cases of conflicts.
Previously I noted that although the first option might be tempting given the many
challenges to evidentialism despite it being considered the orthodoxy, the standard ways of
challenging the intuitive plausibility of evidentialism will not help us in this case. My goal
in Chapter 3 was the second approach—show that there is in fact no conflict between the
moral and epistemic ought by rethinking and questioning the assumptions that seem to pit
them together. The task now is to show why it was worth doing the work of the previous
chapter by now exploring the alternatives to the encroachment picture I defended. To
illustrate this space of alternatives, consider the following flowchart.
98
Fig. 2.
As we see in this flowchart, moral encroachment is not even the only plausible account
within the space I have been describing under the second option—showing that there is
in fact no conflict between the moral and epistemic ought by rethinking and questioning
the assumptions that seem to pit them together. That is, there is more than one way to
answer “no” to Q1.
Further, as we see in this flowchart, there is a wide range of alternatives to moral
encroachment that have thus far gone unconsidered—i.e, accounts that answer “yes” to
Q1. These accounts can further be distinguished into two broad categories given how
they answer Q2. First, there are the accounts that answer “yes” to Q2, accounts that
argue that although there is a conflict between epistemic and moral demands, there is a
way of adjudicating between them such that we can say there is something Spencer and
Agnes ought simpliciter believe. These accounts can further be distinguished by whether
99
what Spencer and Agnes ought simpliciter believe is merely a matter of the moral ought
beating out the epistemic (or vice versa) on the one hand, or whether there is an all-things-
considered ought that adjudicates such conflicts.
Second, there are accounts that answer “no” to Q2, accounts that deny that there’s
anything that Spencer or Agnes ought simpliciter believe. These accounts can then be
distinguished further by how they answer Q4. Either when oughts conflict there is no
conflict—that is, there is one sense in which you should do x and a different sense in which
you should do y—or when oughts conflict there is just no way to determine what to do
in cases of conflicts.
I turn now to assessing these alternative accounts, starting with the accounts that an-
swer “yes” to Q1 and Q2.
4.2 The All Things Considered Ought, Ought Simpliciter, or Just
Plain Ought
Let us look again at the cases that seem to either set up a normative conflict or are
examples of a normative conflicts, i.e., a conflict between different oughts.
The Supposedly Rational Racist. You shouldn’t have done it. But you did.
You scrolled down to the comments section of an article concerning the state
of race relations in America, and you are now reading the comments. The
comments on such articles tend to be predictable, but there is one comment
that catches your eye. Amongst the slurs, the get-rich-quick schemes, and the
threats of physical violence, there is the following comment: “Although it
might be ‘unpopular’ or ‘politically incorrect’ to say this, I’m tired of con-
stantly being called a racist just because I rightly believe of the black diners
seated in my section that they will tip worse than white diners in my sec-
tion.” The user posting the comment, Spencer, argues that the facts don’t lie,
and he helpfully reproduces those facts. For example, he links to studies that
show that on average black diners tip substantially less than white diners. The
facts, he insists, aren’t racist. If you were to deny his claims and were to be-
lieve otherwise, it would be you who is engaging in wishful thinking. It would
be you who believes against the evidence. It would be you, not Spencer, who
is epistemically irrational.
Mistaken Identity. The conference has ended, and the organizers have had the
forethought to book a number of tables at a nearby restaurant so that con-
100
versation can continue over dinner. You’re having a good time at dinner and,
after a few drinks, you get up to use the restroom. As you return to your
table, one of the diners, Jim, reaches out to grab your arm and says, “I asked
for a refill fifteen minutes ago.” For a moment you’re confused, then it dawns
on both of you what mistake has been made. Most philosophers don’t look
like you. With regard to melanin levels, you share more in common with
the wait staff than your fellow diners. Given your skin colour, the likelihood
that you are a member of the staff rather than a fellow diner was high enough
to seemingly make it rational for Jim to assume that you were a waiter, not a
fellow diner. The belief that Jim had—and in turn his actions—might amount
to a social faux pas, but, given that the belief was well-supported by the ev-
idence, the belief—and in turn his act—was reasonable. He’s not a bad guy;
he just made an honest mistake. Of course, in the moment you don’t reason
through all of that. In the moment, you quickly laugh it off, eager to return
to your table and dinner conversation.
Wounded By Belief. Suppose that Mark has an alcohol problem and has been
sober for eight months. Tonight there’s a departmental colloquium for a vis-
iting speaker, and throughout the reception, he withstands the temptation to
have a drink. But, when he gets home his partner, Maria, smells the wine that
the speaker spilled on his sleeve, and Mark can tell from the way Maria looks
at him that she thinks he’s fallen off the wagon. Although the evidence sug-
gests that Mark has fallen off the wagon, would it be unreasonable for Mark
to seek an apology for what Maria believes of him? He is, after all, wounded
by what his partner believes of him.
Social Club. Agnes and Esther are members of a swanky D.C. social club
with a strict dress code of tuxedos for both male guests and staff members,
and dresses for female guests and staff members. They are about to go on
their evening walk so they head towards the coat check to collect their coats.
As they approach the coat check, Esther looks around for a staff member.
All of the club’s staff members are black, whereas only a small number of the
club members are black. As Agnes looks around she notices a well-dressed
black man standing off to the side and tells Esther, “There’s a staff member.
We can give our coat check ticket to him.”
A way to interpret these cases is as though they set up the following kind of conflict:
101
you ought (epistemically) believe p, but you ought (morally) not believe p.
1
If we take
seriously the thought that there really is a deep conflict here, we can see why it would be
natural for someone to ask “well, what ought I really do?” or "what ought I do simpliciter?”
I turn now to exploring the alternative accounts that take seriously this idea that there is
something that you ought simpliciter do. That is, accounts that answer “yes” to Q2: when
oughts conflict, is there anything that the agent ought simpliciter do?
4.2.1 Ought Simpliciter as Either the Moral Ought or the Epistemic Ought
When faced with such cases, it might be tempting to think that what you ought sim-
pliciter do is just what you morally ought do. For a case in which this seems compelling,
consider the examples we’ve seen from Stroud (2006) in which the moral demands of
friendship seem to come into conflict with what you epistemically ought believe. Sup-
pose someone tells you a damning story about a friend of yours, e.g., your friend cut
them off in traffic. As a good friend, how should you respond to this information? Or-
dinarily, if you were to receive testimony that some person a cut someone else, b, off in
traffic, that would be sufficient to justify believing that a cut off b. Stroud (2006, pp.
504), however, argues that we owe our friends “something other than an impartial and
disinterested review of the evidence where they are concerned.”
In this case of a conflict between the moral ought and the epistemic ought, Stroud
suggests that the moral ought trumps the epistemic ought. As you consider what you
ought simpliciter do, you should follow the demands of friendship which require being
loyal to your friend and discounting testimony that suggests they behaved badly. As this
is an account in which morality wins, let us call it moral trumping. If we return to the
flowchart, this account not only answers “yes” to Q2, but also “yes” to Q3 (i.e., is what
the agent ought simpliciter do identical with either what they ought morally do or what
they ought epistemically do? According to moral trumping there is something Spencer
ought simpliciter do, it is just whatever morality recommends. The moral ought is the
privileged ought.
On the other hand, another way of responding to these cases of conflicts is to argue
that it is not what morality recommends rather it is what epistemic rationality recom-
mends. As this is an account in which epistemic rationality wins, let us call this epistemic
trumping. The cases in which this alternative seems most compelling take a slightly dif-
ferent form from what we’ve been considering. Whereas we’ve been considering cases in
1
I am careful to hedge my language here because as the reader knows by now, I do not take these cases
to be setting up a real conflict between the epistemic ought and the moral ought. The interpretation that
there is a conflict assumes that epistemic considerations are completely distinct and nonresponsive to moral
considerations, and I have argued previously that this is incorrect.
102
which the conflict has the form you ought (epistemically) believe but you ought (morally)
not believe; the cases that seem most compelling for epistemic trumping are cases with the
following form: you ought (epistemically) not believe but you ought (morally) believe.
In the literature these appear as cases in which moral reasons to believe seem like the
wrong kind of reason to believe.
2
For example, suppose an evil demon were to threaten
to destroy the world unless you believed that 2+2=5. Morally, you ought believe that
2+2=5, but epistemically you ought not believe that 2+2=5. With regard to these cases,
fans of epistemic trumping argue that moral reasons are not the right kind of reasons for
belief. There is something Spencer ought simpliciter do, it is just whatever epistemic ra-
tionality recommends. The epistemic ought is the privileged ought.
However, with regard to this direction of the conflict—you ought (epistemically) not
believe but you ought (morally) believe—there are also cases that seem to make moral
trumping more compelling than epistemic trumping. For example, Preston-Roedder (2013,
2017) argues that civic trust—the kind of trust that is is an essential part of living with
others in the sort of harmony that characterizes morally permissible interaction—requires
a kind of faith in humanity that is epistemically irrational. Maintaining such an attitude,
nonetheless, is an important moral virtue. Further, Marusic (2012, 2015) argues that to
make sincere promises we sometimes have to believe against the evidence.
In light of these various cases pulling us in different directions, one might be tempted
to posit a separate ought—one that is not identical with either the moral or the epistemic
ought, i.e., ought as opposed to ought
mo ral
or ought
e pi s t e mi c
. That is, a separate perspec-
tive from which to adjudicate what Spencer ought simpliciter do. That is, you might be
tempted to answer positively Q1, positively to Q2, but negatively to Q3 (i.e., is what the
agent ought simpliciter do identical with either what they ought morally do or what they
ought epistemically do?) This brings us a defense of the all-things-considered ought, i.e., a
free and unsubscripted sense of ought.
3
I should note that this gloss on both moral trumping and epistemic trumping is very
brief. To fully flesh out both accounts and the pros and the cons for either would require
more space. My goal, however, in presenting the range of alternatives that are to the left
of moral encroachment (as they appear on the flowchart) is simply to show that there
is a range of thorny issues that must be settled and tackled for a full defense if any are
to offer a promising alternative to moral encroachment. One key advantage of moral
encroachment concerns how moral encroachment can sidestep many of the problems that
2
See Rabinowicz and Ronnow-Rasmussen (2004); Hieronymi (2005); Schroeder (2012b) et al.
3
The terminology here is borrowed from Foot (1972), but I should also note Foot as a critic of whether
there is a coherent unobscured notion of this kind of ought.
103
plague accounts that answer “yes” to Q1. As will be shown, the debate to the left of
moral encroachment has a troubling circular structure. Troubling objections to the ought
simpliciter accounts motivate answering “no” to Q2. However, troubling objections also
arise for those accounts, and those objections then lead one back to positing an ought
simpliciter. But the previous problems remain which push you back to answering “no”
to Q2. Moral encroachment offers a way to side-step these problems. But first, let me
continue with the presentation of accounts that posit an ought simpliciter
4.2.2 The Free and Unsubscripted Sense of Ought
Let us begin by considering the plain ought that is distinct from the moral trumping
and the epistemic trumping views. In general, defenders of the ought simpliciter, the just
plain ought, the all things considered ought, whatever you wish to call it, do not identify
this ought with the moral ought or the epistemic ought. Partly it is for the reasons we
have already seen, i.e., in some cases it seems that it is the moral ought that wins and in
others it’s the epistemic ought, but sometimes what we want to know is what we really
ought to do. Another reason is that there are not only moral and epistemic oughts, there
are also oughts of etiquette, legal oughts, aesthetic oughts, etc. The all-things-considered
ought considers all of these oughts before delivering a judgment. If the ought simpliciter
were just identified with one of these oughts, e.g., the moral ought, a consequence would
be that anytime we asked about what we ought do we would be asking a moral question.
As Wedgwood (MS, pp. 6) notes regarding the all things considered ought.
This all-things-considered ‘ought’ is not the same as any distinctively moral
‘ought’. If there is a distinctively moral ‘ought’, judgments involving this
moral ‘ought’ would be normative or evaluative judgments of a special kind—
judgments based on considerations such as rights and duties, the importance
of being fair and helpful towards others, and so on. But not every judgment
about what one “ought” to do is a judgment of this kind. For example, if I
judge that I ought to buy a new pair of shoes, this need not be a moral judg-
ment. I need not be violating anyone’s rights, or neglecting any of my duties,
or failing to be sufficiently fair or helpful towards others if I didn’t buy a new
pair of shoes. Perhaps no one else would be entitled to blame me if I didn’t.
But it can still be true that I ought buy a new pair of shoes.
In a manner similar to Wedgwood, McLeod (2001, pp. 272-3) notes that when we find
ourselves in the difficult situation of trying to adjudicate between conflicts
104
...you might ask—or rather exclaim—“What ought I to do?” That might be a
natural question, given the circumstances, but it would be puzzling as well.
After all, it seems you already know everything you ought to do: morally
you ought to do A, prudentially you ought to do B, legally you ought to do
C, and so on. Why then, in the situation described, would you ask what you
should do? [...] One reason for thinking this is that, in the situation described
at the outset, the question “What ought I to do?” seems reasonable, even
though you already know all of your relative duties. Your question would
not be reasonable if JPO [just plain ought] were simply identical to one of
the relative “oughts.” Thus, JPO is not identical to any of them.
Although your moral duty, prudential duty, etc. might, in some cases, also be what you
just plain ought to do, McLeod and Wedgwood’s point is that this concept of a just plain
ought is distinct from these relative oughts.
Now, why should we go in for an account of an ought simpliciter, of a just plain
ought? One consideration in favor of such an account is that it offers us a unified answer.
Another reason is that the other oughts are partial or incomplete collections of relevant
considerations. A just plain ought is comprehensive, it is based on all the considerations
that weigh either in favour or against. The moral ought, and other oughts, do this to a
degree. For example, when choosing between what you morally ought do and what’s in
your self-interest, morality isn’t blind to your self-interest, it just assigns it less weight.
4
A benefit to the all things considered ought is that it is a kind of comprehensive value
(see Chang 2003). Further, this more comprehensive value comes with a kind of special
authority that the just plain ought has the other oughts don’t. For morality, epistemic
rationality, etiquette, prudence, etc. we can always ask “Why should I be moral", “Why
should I take my hat off when I enter a church?", “Why should I care about self-interest?",
etc. The reason we can ask these questions of these oughts is because we are implicitly
granting the authority of a more comprehensive ought, of a third perspective from which
to answer these should questions (see Joyce 2001; Velleman 2005).
This view can reasonably be held to be the view of common sense. But as we’ll see,
there will be many reasons to reject it. But before we turn to those objections, there are
still some further points to be made in its favor. As we’ve seen, there are many uses of
‘ought’ where ought does not occur in a narrowly moral sense, a narrowly epistemic sense,
a narrowly prudential sense, etc. Rather, ‘ought’ occurs in a more general normative sense
(see Wedgwood 2007, pp. 24). As Wedgwood (MS, pp. 6-7) notes,
4
For a dissenting view, see Finlay (2007).
105
For this reason, a statement of the form ‘A ought to’ need not mean either
that A is morally required to, or that it would best serve A’s interests or
purposes for A to; it can mean that A ought to all things considered—
i.e., given all relevant considerations (which might include both moral and
non-moral considerations), A ought to.
To get an intuitive grasp on this idea, the all things considered ought is the ought that
we deploy when we offer advice to one another, it is the kind of ought that would issue
from a wise guru who has weighed all the relevant considerations. Consider the following
case from (Thomson, 2001, pp. 46) that suggests that there is an unsubscripted all-things-
considered sense of ought that could tell Spencer what he ought simpliciter to believe.
Suppose that Alfred is ill, and that only a dose of a certain medicine will cure
him. It tastes truly awful, however. Alfred asks us “Ought I really take it?”
It is a wildly implausible idea that we can reply only: “Well, your taking it
would be very unpleasant, so in one sense of ‘ought,’ it’s not the case that
you ought to take it, namely the ‘ought
en j oyab l e
’ sense of ‘ought.’ But your
taking it would be good for you, so in another sense of ‘ought’, you ought
to take it, namely the ‘ought
g ood ne s s f o rAl f r ed
’ sense of ‘ought.”’ It is likely
that Alfred will repeat his question: “But ought I take it?” It surely won’t do
to reply: “Are you deaf? I just told you that in one sense you ought to and in
another sense it is not the case that you ought to, and that’s all the advice that
anyone can give you.”
Intuitively, Alfred’s question is intelligible. Alfred isn’t asking about some relativized
ought, he knows what he ought do relativized to each ought. Alfred’s also not asking
what he should do by his own lights, i.e. what he subjectively ought do, as how would
we know better than Alfred as to what he subjectively ought do? These kinds of cases,
and those given above, all seem to present a prima facie case for an ought simpliciter that
is distinct from the various relativized oughts such as the moral ought, epistemic ought,
etc.
There are, however, numerous reasons to be skeptical of this picture. I will consider
some of those reasons in Section 4.3.1, but for now, I turn to considering whether this
account should be preferred to that of moral encroachment.
4.2.3 Ought Simpliciter vs Moral Encroachment
According to moral encroachment the oughts do not recommend incompatible acts
because both oughts speak to different questions—what the evidence is, and whether the
106
evidence is sufficient to justify belief—and as such are not at odds with each other when
it comes to the question of whether to believe p. It is compatible with this account that
there is an all-things-considered ought that also delivers a verdict about what one should
believe. It is not clear to me, however, why we must posit such an ought. There is nothing,
however, in moral encroachment that commits us to the ought simpliciter view. If you
like the ought simpliciter view, however, moral encroachment is compatible with it. I,
however, am in general skeptical of the idea that there is an ought simpliciter. Despite the
reasons given thus far, I find the idea obscure and mysterious for reasons that I will discuss
in the next section.
Another reason in favour of moral encroachment concerns a version of the location
challenge that will arise for oughts that are not the moral ought and the epistemic ought.
The all-things-considered ought, after all, takes into account self-interested reasons, eti-
quette reasons, athletic reasons, aesthetic reasons, legal reasons, etc. I spent Chapter 1
arguing that moral reasons are relevant to the question of what one should believe. I ar-
gued for this against the objection that moral reasons are irrelevant and that whatever
wrong occurs is located somewhere other than in the belief. This objection, which I had
answered for moral considerations, can now be raised against the other reasons that the
all-things-considered ought takes into account before delivering a verdict about what one
ought simpliciter believe. That is, reasons stemming from etiquette seem irrelevant to the
question of what one should believe. As do athletic reasons, aesthetic reasons, etc. How-
ever, we could also imagine cases in which they might bear on what one should believe. If
the all-things-considered account says that what Spencer ought believe is determined by
taking into account all the reasons for or against belief, it must answer its own version
of the location challenge: why do legal oughts, aesthetic oughts, etc. bear on question
of what Spencer should believe. Alternatively, why would those considerations have any
weight at all?
5
4.3 Skepticism about Ought Simpliciter
4.3.1 Three Arguments for Skepticism
You may now worry whether there is a coherent notion of this ought simpliciter and
thus answer negatively to Q2 on the flowchart (when oughts conflict, is there anything
that the agent ought simpliciter do?). I confess, this is a skepticism that I share. Perhaps
there are just facts about what you morally ought do, what you epistemically ought do,
5
Finlay (2014), critic of the all-things-considered view, also argues that some of these ‘oughts’ do not
have any weight that can be added up. See Chapter 6.
107
what you politely ought do, what you athletically ought do, what you legally ought do,
what you aesthetically ought do, etc. But, there is no ought that adjudicates conflicts
between these other oughts. I turn now to considering three reasons for being skeptical
about the ought simpliciter. That is, three reasons that push us to answer “no” to Q2.
First, you could worry that these distinctive kinds of reasons or oughts don’t share a
common justificatory source. For example, Tiffany (2007) argues that there is no com-
mon scale on which these various considerations—moral, epistemic, aesthetic, legal, etc.—
can be compared. He argues that however we combine these competing reasons is de-
termined by facts about our psychology (see also Finlay 2006, 2009). That is, different
agents will weigh considerations differently. There is no further normative perspective
from which we can both ask and answer the question of how these reasons should be com-
bined. Tiffany argues that any proposed account of weighings would be guilty of some
degree of arbitrariness. On this point, this arbitrariness might be due to the incommen-
surable nature of these oughts. That is, the oughts are incommensurable, and that is why
there is no common scale on which they can be weighed.
6
For example, Kelly (2003, pp.
619) argues:
In cases in which what it is epistemically rational to believe clearly diverges
from what it is practically advantageous to believe, there is simply no gen-
uine question about what one should believe: Although we can ask what
one should believe from the epistemic perspective, and we can ask what one
should believe from the practical perspective, there is no third question: what
one should believe, all things considered. In any case in which epistemic and
practical considerations pull in opposite directions, there is simply nothing
to be said about what one should believe all things considered.
This is merely a statement of the skepticism, it is not yet an argument for skepticism. To
fill in that argument, we can turn to Feldman (2004, pp. 692) who presents the following
two examples to argue that there is no sense to be made of a plain ought that “somehow en-
compasses moral considerations, epistemic considerations, and perhaps others, and then
weighs them against one another to come up with an overall assessment.” First, consider
a child comparing two dolls—where one is short and square, the other tall and thin—how
do we determine which one is bigger? Similarly, imagine trying to give an all-things- con-
sidered judgement about who is the all-things-considered best billionaire by comparing
billionaires according to their wealth and their strength. Feldman argues that it doesn’t
6
I should briefly note that this same worry might push you to posit an all-things-considered ought pre-
cisely so that we can weigh these incommensurable oughts. See Chang (2003).
108
make sense to ask whether height trumps width, or whether volume trumps height, etc. in
the case of the dolls. Similarly, there is no independent concept of strealth for which there
is some correct combination or weighing of the two components—wealth and strength.
There is no sense in which one doll is just plain bigger than the other, there is no sense
in which one billionaire is just plain strealthier than the other.
7
We could think that this
is also the case for the ought simplicter. As strealth is the combination of strength and
wealth, the ought simpliciter is supposed to be a combination of moral considerations,
epistemic considerations and more. So as there is no sense to be made of strealth, perhaps
there is also no sense to be made of the ought simpliciter.
Second, you might worry that the ought simpliciter doesn’t make sense because it
requires there to be some standard that is the most normatively important standard and
either way you cash out this idea is incoherent. Copp (1997) argues that whatever way we
try to explain the authority that this most normatively important standard has will force
us to embrace a contradiction. Copp (1997, pp. 101) notes:
...the claim that the candidate[standard] S has the property of supremacy is
the claim that it is normatively more important than any other standpoint, as
assessed from a relevant authoritative standpoint. That is, if S is normatively
the most important, then there is some authoritative standard R that yields
the verdict that S is normatively the most important standpoint.
The challenge we now face is answering whether R is identical to S. If R is identical to S,
then R cannot play the role of the more important normative standard that establishes
the normative supremacy of S. That kind of self-endorsement is characteristic of all the
normative standards we’ve been considering and as such it is unimpressive, e.g., morality
tells you to listen to morality and self-interest tells you to listen to self-interest. So, R
cannot be identical to S.
But, were R to a be a standard other than S, that is similarly unimpressive. As Copp
(1997, pp. 102) further notes:
We don’t want to know whether S is most important, according to some ar-
bitrary standard R. We want to know if S is most important full stop. But
that means R would have to be the normatively most important standard.
7
I should note that Thomson (2008) has a response to this worry. She argues that normative evaluations
are not ambiguous. The best option to choose is which is the most choice worthy option in a particular
situation. So in a case like that presented by Feldman (2004) in which are trying to determine who’s the best
billionaire, that question does not make sense in the abstract. Rather, if we ask for a purpose, e.g., who’s
the best billionaire to choose for a game of baseball, there is no longer an ambiguity.
109
But then S is not the most important standard. This contradicts our initial
assumption.
So, Copp concludes, there is no coherent way to cash out the idea that there is a norma-
tively most important standard that is the standard of what we ought simpliciter do.
Third, you might worry, as Baker (MS) does, that if the initial conflict resulted from
too many oughts, “how does adding another ought help, instead of multiplying the num-
ber of conflicts?” In order for the ought simpliciter to add anything and resolve the
dilemma it must possess some special property of greater normative authority than the
other oughts. Baker (MS) summarizes:
The characterization of that special property is typically metaphorical or oth-
erwise hopelessly vague and ambiguous. Attempts to eliminate this vagueness
face a dilemma: either they characterize the special property in more famil-
iar normative terms, leading to vicious circularity, or they are psychological
characterizations that seem to change the subject.
On this point regarding the vagaries of the ought simpliciter, Baker (MS) argues that the
ought simpliciter not only lacks any sort of a link to a more familiar theoretical notion—
i.e., it is often merely gestured at through the use of evocative phrases—it also lacks the-
oretical utility. Baker (MS) argues that there is no theoretical problem that the ought
simpliciter solves. In the case of the supposedly rational racist, Spencer ought morally re-
frain from belief, but he also ought epistemically believe. Baker (MS) argues that “[t]here
is no mystery here calling out for explanation.” Whatever problem there is here is a prac-
tical dilemma, and there is no reason to assume that practical problems admit of solution.
The problem of the supposedly rational racist, perhaps, is just a practical problem we
face given that we live in a world that is structured by racism and racist institutions, and
perhaps it doesn’t admit of solution.
These reasons for skepticism can lead us in a few directions. We might worry whether
denying that there’s anything Spencer ought simpliciter do amounts to endorsing the pes-
simistic conclusion that the conflict is irresolvable and he’s damned if he believes and
damned if he doesn’t. In Section 4.3.2 I’ll consider the account that answers yes—i.e., that
Spencer is damned no matter what he does—and then in Section 4.3.3 I’ll consider the ac-
counts that answer no. All of these accounts, I suggest, emerge from a general skepticism
of the idea that there is an ought simpliciter.
110
4.3.2 Genuine Dilemmas
The case for genuine dilemmas—a position that, following Thomson (2008),
8
I will
call the Dilemmist—is most clearly seen in cases of homogeneous dilemmas, dilemmas in
which both oughts are of the same normative kind. The primary example of homoge-
neous dilemmas are moral dilemmas. In this discussion of the dilemmist—who answers
the flowchart as follows, Q1: Y, Q2, N, Q4, Y—I will begin by briefly sketching the ap-
peal of the dilemmist position with regard to homogeneous dilemmas and then we can
see if it carries over to the case of heterogeneous dilemmas, i.e., dilemmas in which the
oughts are not of the same normative kind.
Consider an agent, Ginny, who is torn between two incompatible options. She has
promised Fred she’d see the new Marvel movie with him and only him (A), but she also
promised George that she’d see the new DC movie with him and only him (B). Looking
at her busy schedule she notices that she can only either keep her promise to Fred and
do A or she can keep her promise to George and do B. That is, there is no way for her
to do both. Ginny finds herself torn between two incompatible options—A and B—and
asks “What ought I do? A or B?” Let us imagine that she shouts out this question to the
universe, and while she’s shouting a dilemmist happens to be walking by. The dilemmist,
noting Ginny’s dilemma, helpfully answers, “Well, you ought to do A and you ought to
B”. We could reasonably expect Ginny to reply reminding the dilemmist that she can’t do
both and she wants to know which she ought to do. If the dilemmist were again to simply
reply that she ought do A, and she ought do B. Perhaps she shouldn’t do either until she
flips a coin.
But, why agree with the dilemmist? Following a line presented by Williams (1965),
Thomson (2008) notes that in the case of moral dilemmas you might think that whatever
Ginny does, she’ll feel regret at not having done the other. If she breaks her promise to
George and goes to the movies with Fred, she’ll feel the need to make it up to George.
Seeing that both incompatible acts are required does justice to why we would feel this
regret. If the dilemma could be solved by a flip of a coin, we would fail to do justice to
the fact of this regret.
9
So now let us reconsider the heterogenous dilemma that has been our focus and see
whether this fact of regret carries over. As Gendler (2011) introduces one of the cases
that has been our focus, Social Club, let us reconsider what she says. Gendler suggests
8
Although I am using Thomson (2008) to present the dilemmist position, Thomson herself rejects the
dilemmist position.
9
Thomson (2008) contrasts this kind of hard dilemmist with a soft dilemmist who argues that what Ginny
ought to do is ambiguous. This would be compatible the pluralist and permissivist accounts will be laid
out.
111
that given that we live in a society structured by racial categories we simply face a tragic
irresolvable dilemma. She presents the case of Social Club as one in which we face the
following dilemma: we must either (a) pay the epistemic cost of failing to encode certain
sorts of base-rate or background information about cultural categories (i.e., encoding the
information that a minuscule fraction of club members are black whereas all the staff
members are black) or (b) we must actively believe against the evidence and work at reg-
ulating the inevitable associations which that information gives us. In other words, we
must choose between doing what we epistemically ought (encode the base-rate informa-
tion about the race of the staff members) and what we morally ought (not use someone’s
race as a determiner for whether they are staff even when it’s the determiner that is highly
correlated with correctly picking out a staff member). This, she argues, places us in a
tragic irresolvable dilemma. Not only can we not do both what morally ought and what
we epistemically ought, there is no way to resolve this conflict.
Now, is it the case that as with Ginny’s dilemma a simple coin flip could settle for
Agnes or Spencer what they ought believe? I do not think so. Imagine Agnes approaching
John Hope Franklin as though he were a staff member, being corrected, and then replying
“Well I was torn between what I morally ought do and what I epistemically ought do, so
I flipped a coin and went with what I epistemically ought do.” From Chapter 2, it seems
that John Hope Franklin can rightly feel wronged by that, and in general, when you’ve
wronged someone you generally feel regret about it. So, this fact of regret carries over to
the case of heterogeneous dilemmas. That there is this feeling of regret, this feature than
an agent would be wronged if one simply flipped a coin to make a decision concerning
whether they are a staff member, suggests that perhaps we a face a genuine dilemma in
these cases. If either option was permissible and the dilemma could be settled by flipping
a coin, then John Hope Franklin wouldn’t be in a position to feel wronged. Notice that
we can avoid these grounds for regret by refraining from belief (as suggested by moral
encroachment). The dilemmist, however, seems committed to saying that you should
both believe and not believe.
As we’ll see, however, there is a defusing move that is available if we take a slightly
different track on the flowchart. The pluralist, whose account I will expand upon in Sec-
tion 4.3.3, notes that given the plurality of oughts, these oughts are constantly in conflict.
There is nothing special to lose sleep over with regard to conflicts between moral oughts
and epistemic oughts any more than we should lose sleep over conflicts between athletic
oughts and aesthetic oughts. No standard is more normatively important. As we’ll fur-
ther see in Section 4.3, all of these accounts that emerge from a skepticism concerning
ought simpliciter result in this kind of pessimistic conclusion. Spencer, our supposedly
112
rational racist, will always end up doing something right, even if he also does something
else wrong. This is unsatisfying, and I will explain why.
4.3.3 Permissivism and Pluralism
First, this skepticism about an ought simpliciter does not require endorsing that these
conflicts are irresolvable. That is, if we return to the flowchart, it is possible to answer
“yes” to Q1 and “no” to both Q2 and Q4. Instead of the dilemmist picture we saw artic-
ulated earlier, one could endorse a kind of permissivism about the oughts in these cases of
normative conflict. It is not the case that Spencer ought simpliciter believe that his neigh-
bour has an open arrest warrant (call this option A), nor is it the case that Spencer ought
simpliciter not believe that his neighbour has an open arrest warrant (call this option B).
Rather, there is just some sense in which he is permitted to do either A or B. Both in-
compatible options are permitted (see Horty 2003 and Brink 1994). However, what this
account seems to get wrong in the case of the supposedly rational racist is that it is not
that Spencer is permitted to believe or not believe, but there’s a real sense in which he gets
something wrong when he reasons in this way. But, if you don’t feel the force of that you
likely don’t feel the force of the thesis that beliefs can wrong and there is nothing more
that I can say to convince you if you are drawn to this permissivist line.
Second, one could deny that both options are permitted, but insist that there are a
plurality of oughts and from the perspective of each ought you simply ought do what
it prescribes. As Case (2016, pp. 2-3) presents pluralism, pluralism is committed to the
following three claims.
1. Source pluralism: There exists an irreducible plurality of normative domains, which
issue oughts of distinct kinds.
2. Conflict: It is possible for two or more of these domains to issue conflicting recom-
mendations on what to do (i.e., “X-ly one ought to, Y-ly one ought not to,"
where X and Y are two normative domains.)
3. No authoritative adjudication: There is no “all things considered” domain capable
of issuing recommendations about what one ought simpliciter to do; hence, no au-
thority exists that is capable of adjudicating disputes between normative domains.
According to pluralism there is an irreducible plurality of normative domains, and al-
though these domains sometimes issue conflicting recommendations, there is nothing that
one ought simpliciter to do. Things bottom out at what each respective domain says you
113
ought to do. So, in the case of Spencer, he ought morally not believe and he ought epis-
temically believe. It is not, however, as though Spencer is damned whatever he decides
to do. There is simply nothing more to say. The accounts that I listed under skepticism
about ought simpliciter are all pluralist accounts, and I turn now to expanding further on
the accounts defended by Tiffany (2007); Copp (1997); Baker (MS); Finlay (2014) et al.
Pluralism comes in two varieties: unrestricted and restricted. As Case (2016) notes
in his arguments against pluralism, according to unrestricted pluralism all coherent stan-
dards are reason-generating normative domains, whereas according to restricted pluralism
only some standards are reason-generating normative domains. For example, suppose you
are a member of the Knights Who Say Ni. One of the standards governing Knights Who
Say Ni is to shout “Ni!” until your demands are met. For the unrestricted pluralist, this
standard is reason-generating for you. So, in the same way that as an agent that cares
about morality you have moral reasons not to murder, insofar as you are a Knight Who
Says Ni you have reason to shout “Ni!” until your demands are met. According to a stan-
dard example of an unrestricted pluralist, Tiffany (2007, pp. 255), any standard can be a
reason-generating standard as long as:
There is some well-defined aim (e.g., intrinsic-desire-satisfaction), institutional
framework (e.g., positive law), or standard of value (e.g., aesthetic value) rel-
ative to which considerations may be judged as standing in some normative
relation to action (e.g., favouring, defeating, enabling)
In contrast restricted pluralists, such as Copp (1997) and Stroud (1998), argue that these
Knights-Who-Say-Ni reasons shouldn’t be on par with reasons of morality, epistemic ra-
tionality, prudence, etc. Stroud (1998) argues that reasons like the Knights-Who-Say-Ni
reasons are domain-relative and they have no normative force on their own.
10
These rea-
sons might have instrumental value, e.g., social-climbing reasons are reasons that are in-
strumental to the goal of social climbing, but they are only D-reasons, i.e. domain-relative
reasons. These reasons are not genuine reasons for action.
The unrestricted pluralist could reply that the reason it seems as though there are not
genuine Knights-Who-Say-Ni reasons is because most people aren’t partisan toward that
domain. Our partisanship might affect what we take to be normative reasons. As Tiffany
(2007, pp. 244-5) notes.
Just as one may be a partisan supporter of the Canucks over the Maple Leafs—
perhaps even seeing support for the Leafs as a character flaw, admittedly non-
10
For other examples, Dorsey (2013) gives the example of reasons to follow the Constitution of the Sa-
tanic Grave Robbers Society, and Copp (1997) gives the example of moon-love reasons.
114
culpable for those raised in greater Toronto—without thinking that there is
some deep metaphysical truth backing up one’s partisanship; so too can one
be similarly partisan toward, e.g., morality, prudence, or authenticity.
A problem here for this unrestricted version of pluralism is that this view seems to de-
volve into nihilism about practical reason and a skepticism about normative importance.
As Case (2016) notes, for every coherent standard (and there are many according to unre-
stricted pluralism), there is an equal and opposite standard that is also reason-generating.
As a result, we will have an infinite number of reasons generated by an infinite number
of standards and we will constantly be pulled in every direction. Case (2016) argues that
if all competing standards are rightly thought to be normative domains, then the result is
that none are. Case (2016, pp. 6) notes:
A normative domain’s recommendations must deserve our respect, and must
possess authority over the recommendations of competing standards at least
some of the time. I cannot make sense of the claim that all standards are
normative domains any more than I can make sense of the claim that everyone
is properly addressed as Thakin, if this is truly a title of distinction. The only
sense I can make of the suggestion that all standards are normative domains is
that no standards are really normative domains; hence unrestricted pluralism,
so understood, collapses into nihilism.
If the pluralist attempts to avoid this collapse into nihilism by arguing that partisanship
makes a normative difference, i.e., that the only domains that are normatively significant
for me are ones toward which I am partisan, then pluralism devolves into subjectivism
about practical reason. Both nihilism and subjectivism about practical reason are incom-
patible with pluralism. Nihilism about practical reason is inconsistent with source plural-
ism, i.e., that there are a plurality of normative domains (as according to nihilism there
are no normative domains). Subjectivism is inconsistent with no authoritative adjudica-
tion because one’s partisanship would determine what standards are authoritative for one.
This might lead you to reject unrestricted pluralism for restricted pluralism, but there
are issues for restricted pluralism as well. Recall Stroud (1998) argues for domain-relative
reasons that contrast with genuine reasons for action. The restricted pluralist now faces
the challenge of explaining why some standards are normative and deserve to be respected,
whereas others do not. Further, the restricted pluralist must also explain why our intu-
itions in cases involving a small good in one domain at the expense of a large cost in an-
other domain suggest that there is something you plain ought to do. For example, if you
are walking by a lake and see a drowning child you could either save the child or you could
115
not. From the perspective of self-interest, perhaps you shouldn’t save the child—after all,
you would get your clothes wet and from the perspective of self-interest you prefer hav-
ing dry clothes. From the perspective of morality, you should rescue the drowning child.
The pluralist in this case must simply say that what you morally ought do is rescue the
child, but self-interestedly you ought not rescue the child. This seems like the wrong con-
clusion. What we want to say in this case is that you just ought to do what you morally
ought do. Thus, we are pushed back towards positing an ought simpliciter.
11
This intuition we have is also the intuition we have in the case of the supposedly ratio-
nal racist. Just as it doesn’t seem right to say of the person that didn’t save the drowning
child for self-interested reasons that at least they did what they self-interestedly ought
to have done, it similarly seems wrong to say of Spencer that he at least did what he
epistemically
T
ought to have done.
Recall the apology that Jim offers to the conference goer he mistakes for a diner from
Chapter 2: “My evidence supported that you were a waiter, not a conference participant.
So, I had the attitude I epistemically ought to have had towards you being a waiter as
supported by my evidence. This belief wasn’t motivated by any feeling of ill-will toward
you, I was just believing in accordance with the evidence." As I argued, in this apology
Jim problematically takes the theoretical perspective towards the agent wronged by his
belief. In his apology he is responding as though he is a scientific observer explaining
casual phenomenon in the world. But, as I argued, that is precisely what he should be
apologizing for. He should be apologizing for failing to take the participant attitude. For
failing to see why you would be invested in the attitudes he holds of you. For failing to
be sensitive to the moral considerations and seeing how those moral considerations bear
on what he should believe. The pluralist offers a similar response to the one offered by
Jim. According to the pluralist, Jim is off the hook for any criticism of the form “you
shouldn’t have believed that of the diner" as long as he does what he epistemically ought.
We could criticize him for not doing what he morally ought, but epistemically he’s off
the hook. The advantage of moral encroachment is that it can explain why Jim isn’t off
the hook epistemically either.
4.4 No Positive Duties to Believe
Finally, let us return to the first question on the flowchart—do the oughts recommend
incompatible acts?—and consider another view that answers “no” to that first question,
11
This “argument from notable-nominal comparisons” can be found in Chang (1997); Parfit (2011);
Dorsey (2013) and Scanlon (1998). Case (2016) also discusses this problem and raises two further objec-
tions to restricted normative pluralism: the sorting problem and the concurrent-case argument.
116
but nonetheless can be distinguished from moral encroachment by how it answers another
question. That is, the question of whether there is anything that the epistemic ought rec-
ommends (Q5). Moral encroachment answers “yes” to Q5 because the epistemic ought
concerns the strength of one’s evidence. That is, we have epistemic obligations to at-
tend to the evidence. However, attending to the evidence is not sufficient to settle the
question of what one ought believe. Nonetheless, there is something the epistemic ought
recommends according to moral encroachment, the Cliffordian charge “to attend to the
evidence!”.
However, another way to proceed is to answer “no” to Q5. That is, to deny that there
is a conflict in these cases because the epistemic ought does not recommend or require
anything. One such account is provided by Nelson (2010). Nelson argues that we have
no positive epistemic duties to believe, that is, there are only negative epistemic duties.
Nelson argues that the most authority the epistemic ought has is in saying what we are
permitted to believe. What settles whether to believe or might require one to believe,
are non-epistemic considerations. This, so far, sounds a lot like the moral encroachment
picture that I have been defending. As we’ll see this picture is compatible with moral
moral encroachment, but there will be one important point of difference.
4.4.1 The Account
Nelson argues that there is nothing we ought believe on purely epistemic grounds.
His argument starts by contrasting ethics and epistemology. In ethics, he argues, it is
hard to deny that there are both positive duties—things we ought to do—and negative
duties—things we ought not do. There is more debate, however, in epistemology regard-
ing whether there are things that we ought to believe and things that we ought not be-
lieve. A reason for initially thinking that there are positive epistemic duties stems from
the many parallels between ethics and epistemology. First, both ethics and epistemol-
ogy are normative. That is, both disciplines concern what we ought do. Further, many
of the meta-frameworks, normative theories, and accounts of structure we employ in
ethics have counterparts in epistemology as well, e.g., non-cognitivism, reductionism,
non-naturalism, nihilism, consequentialism, deontology, virtue theories, foundational-
ism, coherentism, etc. As a result, we could suppose that “whatever is true in ethics, about
action, is also true, mutatis mutandis, in epistemology, about belief” (Nelson, 2010, pp.
84).
In short, the basic thought for there being positive epistemic duties is that there are
positive duties in ethics and perhaps those positive duties extend into epistemology as
well. However, as Nelson (2010) persuasively argues, the idea that we have positive epis-
117
temic duties seems demanding and unrealistic. Evidence might give us a reason to believe
something, but to say that it follows that I have a duty to believe everything for which I
have evidence leads to a conclusion we ought reject: that I am required to believe an in-
finite number of things. This is what Nelson calls “the infinite justificational ‘fecundity’
of perceptual and propositional evidence.”
This, Nelson (2010, pp. 87) argues, is suggestive of a permissive epistemic theory in
which “first-order normative epistemic principles concern what we are permitted to be-
lieve, given our epistemic circumstances—not what we are obligated to believe.” Accord-
ing to this permissive epistemic theory certain beliefs are licensed by our epistemic situa-
tion, i.e., the evidence, but what we should believe is constrained by other non-epistemic
considerations. Note that this is entirely compatible with moral encroachment accord-
ing to which evidential considerations set limits on what what we are epistemically per-
mitted to believe—i.e., the strength of the evidence determine what you are permitted to
believe—while the non-evidential considerations, such as moral considerations, determine
what you should believe.
12
Before comparing these accounts, however, let us turn to how Nelson establishes his
conclusion. For consider the following case.
Given the appearance of some distinctive dark, winged shapes, moving across
my visual field, what should I believe? That visual evidence, joined with other
factors, may license me to believe propositions such as:
(1) There are things moving through the air in front of me
(2) There are birds flying in front of me
(3) There are jackdaws flying in front of me
(4) At least three jackdaws exist (Nelson, 2010, pp. 87)
Of these, which proposition I believe will depend on many things. For example, it could
depend on “how my perceptual abilities have developed (e.g. have I learned to discriminate
different kinds of bird on the wing?); the background information I happen to have (e.g.
do I know what a jackdaw is?); and my particular interests at that moment (e.g. what do
I want to know or do now?).”
12
We have also seen similar remarks made by Paul and Morton (2018) where they note that there are
multiple evidential policies that are rationally permissible for a given thinker to have from the point of
view of purely epistemic considerations. Similarly, Preston-Roedder (2018) argues that because people are
opaque and the evidence we have about them is at best partial and ambiguous, there are many epistemically
permissible responses to evidence.
118
When we ask which proposition I should believe, the answer depends only partly on
my epistemic situation, i.e., the evidence. The answer also depends partly on my needs
and interests. For example, if I am interested in launching a model airplane, then I should
believe (1). But, if I am afraid of birds and wish to avoid them, then (2). If I am conducting
a species survey, then depending on the specificity of the survey I should believe either
(3) or (4). However, if my interest is to hail a cab, I need not believe either (1)-(4).
13
Thus,
the difference between ethics and epistemology, is that although there are often things
you positively ought do given the morally relevant features of your environment, there is
nothing you positively ought believe given the totality of epistemically relevant features
of your environment.
I should note here that Nelson is using “epistemically relevant features of your envi-
ronment” in a way that stands at odds with the claims I make regarding moral encroach-
ment. That is, Nelson is using the restricted ‘epistemic
T
’ version of ‘epistemic’. Accord-
ing to moral encroachment, the epistemically relevant features of your environment in-
clude moral features of your environment, not the merely truth-conducive features of
your environment. However, if we make this kind of change, then Nelson’s view just be-
comes the moral encroachment view. As I have argued, I do not think there is any good
reason to exclude non-evidential considerations from the list of considerations that are
epistemically relevant. Once we include those considerations, however, there will be pos-
itive epistemic duties to believe. To clarify terminology once more, by positive epistemic
duty here, I will mean a reason to believe. According to moral encroachment both moral
and epistemic reasons, insofar as both can count as reasons to believe, will be classed as
positive epistemic duties. And this again is compatible with Nelson’s view as the evidence
alone does not settle what you should believe, to have positive duties the non-evidential
considerations must enter into the picture.
Returning now to Nelson’s picture, what explains this difference between ethics and
epistemology according to Nelson? The answer brings us back to the infinite justifica-
tional fecundity of evidence. Nelson (2010, pp. 96) notes that the infinite justificational
fecundity of evidence is
13
Further, although this example concerns perceptual beliefs, the general point extends to inferential
beliefs as well. Nelson (2010, pp. 87-8) notes:
what conclusion (epistemically) ought I to draw if I believe ‘p’ and ‘if p then q’ (and grasp
the relevant logical rules etc.)? It is impossible to say in advance. It may be ‘q’, of course,
but depending on my needs or interests it may equally be ‘q or r’, or ‘p’, or ‘p and if p then
q’—or, if I am looking to hail a taxi, nothing at all. The premisses license all of these and
more, they constrain me from believing anything incompatible with the licensed beliefs, but
they require me to draw no conclusion on any topic that does not concern me.
119
...the fact that every single bit of evidence, whether experiential or proposi-
tional, potentially epistemically justifies an infinitely large array of different
beliefs. By itself, this may not appear terribly significant. When combined
with positive epistemic duties, however, it takes on a different appearance,
because together they entail the duty to believe everything that is justified for
us in our epistemic circumstances.
According to this result, the positive epistemic duties thesis—you ought add every propo-
sitional belief that the evidence epistemically justifies for you—and the fecundity thesis—
the evidence justifies infinitely many propositions—you ought now add infinitely many
beliefs.
14
But as Nelson (2010, pp. 98) notes:
we cannot have a duty to add an infinite number of beliefs at any given mo-
ment, because (given similarly reasonable assumptions about our limited psy-
chological capacities) this is not humanly possible, and we do not have a duty
to do the impossible. Hence, we have no positive epistemic duties.
It turns out that belief, unlike ethical action, is optional.
15
4.4.2 The Drawbacks
One might think that an advantage of this Nelson picture over that of moral encroach-
ment is that it keeps the distinction between epistemic considerations and moral con-
siderations. According to moral encroachment, there is no conflict because part of the
epistemic considerations—where by ‘epistemic’ we mean whatever considerations are rel-
evant to whether a belief is epistemically justified—are the moral considerations. The
moral considerations play a role in setting the threshold the evidence must pass for a be-
lief to be to justified. That is why Spencer’s belief is not epistemically justified in the way
that he thinks that it is. On Nelson’s picture, however, Spencer is epistemically justified,
he’s just not required to believe what he thinks he’s required to believe, he’s merely per-
mitted to believe it. If Spencer were to rest his defence of believing that the diner in his
section won’t tip well on the grounds that he’s not doing anything wrong because he’s do-
ing what he’s epistemically permitted to do, on Nelson’s account we can question Spencer
14
An earlier version of it can also be found in Feldman (2004, pp. 678): “Suppose that a person has
evidence that conclusively establishes some proposition, q. There are then a huge number, perhaps an
infinite number, of obvious logical consequences of q that are also supported by this evidence.”
15
I should also note here that there are other ways to avoid Nelson’s conclusion here. You could argue
that it’s not the case that every belief we have must be represented in the mind somewhere. Many beliefs
can be implicit, can be forgotten, etc
120
about why he chooses to believe that of his neighbour, but not believe all the other be-
liefs he’s permitted to believe. What he chooses to believe, then, could be a reflection of
a morally criticizable character akin to the story that Arpaly (2004) and co. tell about the
supposedly rational racist back in Chapter 1—i.e., Spencer is likely engaging in motivated
reasoning. As I previously noted in Chapter 1, a disadvantage of this approach is that we
must say that Spencer’s self-conception of himself—as someone earnestly just believing in
accordance with the evidence—is false. This may in fact be the case for many people who
claim to be like Spencer, but there’s something more powerful about an account that is
able to preserve Spencer’s self-conception of himself and still say why what he is doing is
not right even by his own lights.
A further drawback of Nelson’s account is that no matter how strong the evidence you
are never required to believe. This, however, seems counterintuitive and seems to license
a kind of epistemic behaviour we might call wishful denial. Wishful thinking is when an
agent believes that something is the case despite evidence to the contrary because believing
would make their life go better in some way. For example, I might wishfully believe that
I am not as behind on my work as I in fact am because that belief helps me avoid crippling
anxiety. On the flipside, wishful denial would be avoiding believing something you ought
believe. For example, I ought believe that I am in fact behind on my work, but I can use
Nelson’s argument to show that I’m not required to believe that I am behind on my work.
For such a trivial case, perhaps that is fine. But now consider more pressing cases that we
previously considered in Chapter 3, Section 3.5.2. Consider, for example, the tobacco
industry insisting for years that contrary to the evidence we had no reason to believe that
there’s a connection between smoking and cancer. Or business and financial interests
insisting that contrary to the evidence we have no reason to believe that climate change
is happening. Or when a woman comes forward with an account of sexual harassment
saying that we have no duty to believe her. In such cases, it seems compelling to say
that someone who fails to believe that there’s a link between smoking and cancer, or a
link between the increased carbon emissions from burning coal and the sea levels rising,
or who refuses to believe testimony concerning sexual harassment is doing something
epistemically wrong. In this cases it seems that they ought believe.
Nelson might reply that that is because of the moral stakes of the beliefs, so again it is
something non-evidential that is responsible for the ought. But I suggest that even if the
moral stakes were absent, the sheer strength of the evidence would be sufficient to force an
agent to believe. To stubbornly refuse to believe in the face of irrefutable evidence strikes
me to be an epistemic fault. The benefit of the stakes picture is twofold. First, in contrast
with Nelson, although moral considerations can shift the threshold once that threshold
121
is crossed you ought believe. Second, the moral stakes can not only raise the threshold for
evidence, but also lower it.
4.5 Concluding Remarks
My goal in this last chapter has been to complete the task of defending moral encroach-
ment. The goal of Chapter 1 and 2 was to show that beliefs can wrong, and the goal of
Chapter 3 and 4 was to show that the beliefs that wrong are also epistemically irrational.
Although I developed an account of moral encroachment in Chapter 3, that account was
incomplete until I could show why moral encroachment should be preferred to the tra-
ditional ways of answering conflicts that arise between different oughts. As I hope to
have shown in this chapter, a key advantage of moral encroachment is that it allows us to
prescind from questions about whether there is an all-things-considered ought that adju-
dicates conflicts and thereby avoid many of the challenges that arise for defenders of the
all-things-considered ought and those skeptical of it.
122
CONCLUSION
A central theme in this dissertation has been that when it comes to what we should
believe, morality is not voiceless. Our epistemic practices exist in an unjust and non-ideal
world and there are moral and social constraints on our epistemic practices. Specifically,
I have argued that beliefs can wrong and the epistemic justification of our beliefs can be
affected by the moral demands of our environment. There is much, however, that still
needs to be said and this conclusion is not so much a final word on the topic, but rather a
confession of everything else I wish I could have addressed.
For example, the focus of this dissertation has been restricted to directed wrongs, i.e.,
wrongs done to specific individuals. This leaves open, however, large questions concern-
ing when the wrongs done to an individual extend to the group the individual belongs
to. For example, when you believe of an individual that they are dangerous given their
race, that wrong plausibly extends to the group. As Kendrick Lamar raps in Fear, “How
they look at me reflect on myself, my family, my city”. This may not, however, apply to
other groups, e.g., employees and their place of work. This is a difference in need of an
explanation and sadly it’s not something I had time to take up in this dissertation.
I have argued that we can wrong each other simply in virtue of the beliefs we hold
about each other, regardless of how or even whether those beliefs manifest in our actions.
To what extent, however, can I legitimately place a claim on you to believe a particular
thing about me? A challenge for this account that I have not yet addressed concerns giving
an adequate explanation of what we can legitimately require of othersÕ beliefs about us
(see also Begby 2018 for this line of objection). For example, I may want you to believe
well of my philosophical ability, but what right do I have to claim that of you? Do I
really have a moral claim upon you to believe that I am an excellent philosopher? For a
more troubling example, consider Srinivasan (2018)’s (2018) discussion of Elliot Rodger.
Quoting from his manifesto, she notes that he claims to have been “cast out and rejected,
forced to endure an existence of loneliness and insignificance, all because the females of
the human species were incapable of seeing the value in[him].” What right does Elliot
Rodger have to demand that the females he interacts with see value in him? In particular,
notice that where Elliot Rodger speaks of value, what he means is being sexually desired by
“hot, beautiful blonde girls”. Certainly he has no right to that. In short, when do we make
morally permissible demands on the cognitive lives of others? I owe the reader an answer
to this question, but I have yet to offer one. In part, I think an answer to this question
will turn on how we settle another thorny set of issues: what we are licensed to believe of
each other. I admit this appears much like avoiding answering one difficult question by
123
suggesting that the answer turns on how we answer a different difficult question. Further,
there is yet another difficult question we must answer before we can even address this
challenge.
Who is it that we wrong and what kind of relationship do we need to stand in with
them? Consider, for example, our beliefs about people in our distant past. When I see
images of Roman graffiti and come to believe that whoever drew that was childish, have
I come to wrong that Roman? Consider, also, Ea-Nasir. Ea-Nasir is a merchant from
ancient Babylonia. We know about Ea-Nasir because a customer, Nanni, sent a letter—
i.e., a clay tablet—complaining of the substandard copper that Ea-Nasir had sold him.
16
Must I take the participant stance towards Ea-Nasir (and Nanni)? The answer is clear in
the case of deep personal relationships to each other, but when the relationship is thinner,
indirect, or non-existent, to what extent can others have moral claims on our cognitive
lives?
Further this dissertation has focussed only on the question of outright beliefs and how
outright beliefs can wrong. But beliefs are not the only attitudes we hold towards each
other. Further, neither are they the most common attitudes we hold towards each other.
We not only believe things of others, we also hope, imagine, assume, expect, fantasize
things about others. There is a whole range of belief-like attitudes we hold, and as a result,
perhaps an even wider range of ways in which we might wrong each other. In some sense,
I made the task for myself in this dissertation harder than it needed to be by focussing
on belief. Instead, there could be an easier route to the thesis that beliefs can wrong by
attending to the many belief-like attitudes we hold and how they wrong. For example,
parents’ expectations of their children and how their hopes for their children can wrong
them. And if belief-like attitudes can wrong, perhaps it’s a short step from the thesis that
belief-like attitudes can wrong to the thesis that beliefs can wrong.
17
Further, our mental
life is even richer than just belief and belief-like attitudes: these attitudes can also come
in differing degrees of strength, i.e., credences. Can credences wrong in the same way
outright beliefs wrong?
Finally, it seems that we not only wrong when we believe, we can also wrong when
we refrain from belief and we can wrong when we undermine belief. Consider, for exam-
ple, the Holocaust denier, the tobacco lobbyist insisting that the evidence is insufficient
to support a link between smoking and lung cancer, the politician who denies that the ev-
idence is sufficient to support anthropogenic climate change, the parent who denies that
the evidence is sufficient to support the recommended vaccination schedule, etc. These
16
See: https://en.wikipedia.org/wiki/Complaint_tablet_to_Ea-nasir
17
Thanks to Amy Kind for pressing me on this point.
124
examples, I think, point to a broader phenomenon than what I have addressed in this
dissertation: issues of epistemic injustice (see Fricker 2007).
Partly why I have not attended to the issue of epistemic injustice in this dissertation
is because much of the current research on the topic of epistemic injustice has focused on
cases involving testimony and questions of credibility assignment (Dotson 2011; Davis
2016; Pohlhaus 2017). In contrast, the kind of wronging I have been drawing attention to
does not fall neatly into any of the discussion of epistemic injustice. That is, this kind of
wronging involved in refraining from belief or undermining belief concerns not how an
individual is wronged in their capacity as a knower, but in their capacity as a believer. I
think this points to a phenomenon I very briefly gestured towards at the end of Chapter
3, epistemic manipulation. The Holocaust denier and the climate change skeptic both
illustrate the way in which doubt can be harmful when it is unreasonable. However, just
as it has been difficult to articulate necessary and sufficient conditions for when beliefs
wrong, it can also be difficult to identify when doubt is unreasonable. This difficulty,
however, it central to what makes the wrong so problematic.
Epistemic manipulation, as the counterpart to emotional abuse and emotional manip-
ulation, can be difficult to identify because of how closely it resembles epistemic norms
we standardly accept as good epistemic norms. For example, emotional abusers make use
of norms that are part of a healthy relationship and they use those norms to manipulate
their partners. Part of why it is so difficult to recognize emotional manipulation as a form
of abuse is how closely it resembles healthy behaviour. Consider the following norm of
a healthy relationship: you should trust your partner. An emotionally abusive partner,
however, will manipulate that trust to harm their significant other.
An example of epistemic manipulation that has garnered some attention recently is
gaslighting (see Abramson 2014; McKinnon 2017). Gaslighting is a form of emotional and
epistemic manipulation that abuses the following standardly accepted epistemic norm:
when someone you respect disagrees with you, many argue that you should revise your
confidence in your initial belief. Consider the following case from the 1944 film, Gaslight.
You believe that the lights have been flickering, but someone you respect repeatedly insists
that they don’t flicker. You are never able to prove that the lights flicker while this person
is around. In normal circumstances, many argue that you should lower your confidence
in the initial belief (Christensen 2007; Elga 2007; Feldman 2007). My hope is that the
story I have told thus far in the dissertation can be extended to show how this case it is the
epistemic norms themselves that are being used to wrong you. You are being epistemically
manipulated.
In short, although I believe I have shown what I set out to show—that beliefs can
125
wrong and that when it comes to what we should (epistemically) believe morality is not
voiceless—much more needs to be said about how the moral considerations of social in-
justice and epistemic norms interrelate. You might think, as Foot (2001) thinks about her
own work at the end of Natural Goodness, that “[t]he proper reply is that in a way nothing
is settled, but everything is left as it was.” I agree that nothing is settled, but I also hope
that it’s not the case that everything is left as it was.
126
BIBLIOGRAPHY
127
BIBLIOGRAPHY
Abramson, K. (2014). Turning up the lights on gaslighting. Philosophical Perspectives,
28(1):1–30.
Antony, L. M. (2016). Bias: Friend or foe. In Brownstein, M. and Saul, J., editors, Implicit
Bias and Philosophy, volume 1. Oxford University Press.
Appiah, K. A. (1990). Racisms. In Goldberg, D., editor, Anatomy of Racism, pages 3–17.
University of Minnesota Press.
Appiah, K. A. (1995). The uncompleted argument: Dubois and the illusion of race. In
Mosley, A. G., editor, African Philosophy: Selected Readings, pages 199–215. Englewood
Cliffs.
Arpaly, N. (2004). Unprincipled Virtue: An Inquiry into Moral Agency. Oxford University
Press.
Arpaly, N. and Schroeder, T. (2014). In Praise of Desire. Oxford University Press.
Atwood, M. (1998). The Handmaid’s Tale. Anchor Books.
Baker, D. (MS). Skepticism about ought simplicter. In progress.
Banaji, M. and Greenwald, A. (2013). Blindspot: Hidden Biases of Good People. Delacorte
Press.
Basu, R. and Schroeder, M. (2018). Doxastic wrongings. In Kim, B. and McGrath, M.,
editors, Pragmatic Encroachment in Epistemology. Routledge.
Beeghly, E. (2015). What is a stereotype? what is stereotyping? Hypatia, 30(4):675–691.
Begby, E. (2013). The epistemology of prejudice. Thought, pages 90–99.
Begby, E. (2018). Doxastic morality: A moderately skeptical perspective. Philosophical
Topics.
Bell, M. (2009). Anger, virtue and oppression. In Tessman, L., editor, Feminist Ethics and
Social and Political Philosophy: Theorizing the Non-Ideal. Springer.
Bero, S. (MS). Relationships and reactive attitudes. in progress.
Block, N. (1996). How heritability misleads about race. The Boston Review, 20(6):30–35.
128
Blum, L. (2002). Racism: What it is and what it isn’t. Studies in Philosophy and Education,
21(3):204–218.
Blum, L. (2004). What do accounts of “racism” do? In Levine, M. and Pataki, T., edi-
tors, Racism in Mind: Philosophical Explanations of Racism and Its Implications. Cornell
University Press.
Bolinger, R. (2017). Mistaken Defense and Normative Conventions. PhD thesis.
Bommarito, N. (2017). Virtuous and vicious anger. Journal of Ethics and Social Philosophy,
11(3).
Brink, D. (1994). Moral conflict and its structure. The Philosophical Review, 103:215–247.
Buchak, L. (2014). Belief, credence, and norms. Philosophical Studies, 169(2):285–311.
Burge, T. (1993). Content preservation. The Philosophical Review, 102.
Case, S. (2016). Normative pluralism worthy of the name is false. Journal of Ethics and
Social Philosophy, 11(1).
Chang, R. (1997). Incommensurability, Incompatibility and Practical Reason. Harvard
University Press.
Chang, R. (2003). All things considered. Philosophical Perspectives, 18(1):1–22.
Christensen, D. (2007). Epistemology of disagreement: The good news. Philosophical
Review, 116(2):187–217.
Clifford, W. (1877). The ethics of belief. Contemporary Review.
Clough, S. and Loges, W. (2008). Racist value judgments as objectively false beliefs: A
philosophical and social-psychological analysis. Journal of Social Philosophy, 39(1):77–
95.
Conee, E. and Feldman, R. (2004). Evidentialism. Oxford University Press.
Copp, D. (1997). The ring of gyges: Overridingness and the unity of reason. Social
Philosophy and Policy, 14(1):86–101.
Corlett, J. A. (2005). Race, racism, and reparation. Journal of Social Philosophy, 36(4):568–
585.
Darwall, S. (2006). The Second-Person Standpoint: Morality, Respect, and Accountability.
Harvard University Press.
Davis, E. (2016). Typecasts, tokens, and spokepersons: A case for credibility excess as
testimonial injustice. Hypatia, 31(3):485–501.
de Beauvoir, S. (1949). The Second Sex, volume 2010. Random House.
129
DeRose, K. (1992). Contextualism and knowledge attributions. Philosophy and Phe-
nomenological Research, 52:912–29.
DiAngelo, R. (2011). White fragility. International Journal of Critical Pedagogy, 3(3):54–
70.
Dorsey, D. (2013). Two dualisms of practical reason. In Shafer-Landau, R., editor, Oxford
Studies in Metaethics, volume 8. Oxford University Press.
Dotson, K. (2011). Tracking epistemic violence, tracking practices of silencing. Hypatia,
26(2):236–257.
Dotson, K. (2018). Accumulating epistemic power: A problem with epistemology. Philo-
sophical Topics.
Doyle, A. C. (1887). A Study in Scarlet.
Doyle, A. C. (1892). The boscombe valley mystery.
Du Bois, W. E. B. (1903). The Souls of Black Folk, volume 1994. Dover.
Elga, A. (2007). Reflection and disagreement. Nous, 41(3):478–502.
Fantl and McGrath (2009). Knowledge in an Uncertain World. Oxford University Press.
Fantl, J. and McGrath, M. (2002). Evidence, pragmatics, and justification. The Philosoph-
ical Review, 111(1):67–94.
Fantl, J. and McGrath, M. (2007). On pragmatic encroachment in epistemology. Philoso-
phy and Phenomenological Research, 75(3):558–589.
Feldman, R. (2004). The ethics of belef. In Conee, E. and Feldman, R., editors, Eviden-
tialism: Essays in Epistemology, pages 166–195. Oxford University Press.
Feldman, R. (2007). Reasonable religious disagreements. In Antony, L., editor, Philoso-
phers without Gods. Oxford University Press.
Finlay, S. (2006). The reasons that matter. Australasian Journal of Philosophy, 84(2):1–20.
Finlay, S. (2007). Too much morality. In Bloomfield, P., editor, Morality and Self-Interest.
Oxford University Press.
Finlay, S. (2009). Oughts and ends. Philosophical Studies, 143(3):315–340.
Finlay, S. (2014). Confusion of Tongues: A Theory of Normative Language. Oxford Univer-
sity Press.
Flowerree, A. (2016). Agency of belief and intention. Synthese.
Foot, P. (1972). Morality as a system of hypothetical imperatives. The Philosophical Re-
view, 81(3):305–316.
130
Foot, P. (2001). Natural Goodness. Oxford University Press.
Fricker, M. (2007). Epistemic Injustice: Power and the Ethics of Knowing. Oxford Univer-
sity Press.
Fritz, J. (2018). From pragmatic encroachment to moral encroachment. Pacific Philosoph-
ical Quarterly.
Ganson, D. (2008). Evidentialism and pragmatic constraints on outright belief. Philosoph-
ical Studies, 139(3):441–458.
Garcia, J. L. A. (1996). The heart of racism. The Journal of Social Philosophy, 27(1):5–45.
Garcia, J. L. A. (1997). Current conceptions of racism: A critical examination of some
recent social philosophy. Journal of Social Philosophy, 28(2):5–42.
Garcia, J. L. A. (1999). Philosophical analysis and the moral concept of racism. Philosophy
and Social Criticism, 25(5):1–32.
Gendler, T. S. (2011). On the epistemic costs of implicit bias. Philosophical Studies, 156:33–
63.
Gordon, L. (1995). Bad Faith and Antiblack Racism. Humanities Press.
Gordon, L. (2000). Existentia Africana: Understanding Africana Existential Thought.
Routledge.
Greenwald, A. and Banaji, M. (2015). Statistically small effects of the implicit associa-
tion test can have societally large effects. Journal of Personality and Social Psychology,
108(4):553–561.
Grimm, S. R. (2016). Knowledge, practical interests, and rising tides. In Greco, J. and
Henderson, D., editors, Epistemic Evaluation: Purposeful Epistemology. Oxford Univer-
sity Press.
Hacking, I. (1995). The looping effects of human kinds. In Dan Sperber, D. P. and
Premack, A. J., editors, Causal cognition: A multidisplinary debate, pages 351–394. Ox-
ford University Press.
Hacking, I. (1999). The social construction of what? Harvard University Press.
Haslanger, S. (2012). Resisting Reality: Social Construction and Social Critique. Oxford
University Press.
Hawthorne, J. (2003). Knowledge and Lotteries. Oxford University Press.
Hawthorne, J. and Stanley, J. (2008). Knowledge and action. Journal of Philosophy,
105(10):571–590.
Herrnstein, R. and Murray, C. (1996). Bell Curve: Intelligence and Class Structure in
American Life. Free Press.
131
Hieronymi, P. (2005). The wrong kind of reason. Journal of Philosophy, 102(9).
Hieronymi, P. (2006). Controlling attitudes. Pacific Philosophical Quarterly, 87:45–74.
Hieronymi, P. (2008). Responsibility for believing. Synthese, 161:357–373.
Horty, J. F. (2003). Reasoning with moral conflicts. Nous, 37(4):557–605.
Ikuenobe, P. (2011). Conceptualizing racism and its subtle forms. Journal for the Theory
of Social Behaviour, 41(2).
James, W. (1896). The will to believe. The New World.
Joyce, R. (2001). The Myth of Morality. Cambridge University Press.
Kant, I. (1785). Groundwork of the Metaphysics of Morals, volume 2012 of Cambridge Texts
in the History of Philosophy. Cambridge University Press.
Kelly, T. (2003). Epistemic rationality as instrumental rationality: A critique. Philosophy
and Phenomenological Research, 66(3).
Langton, R. (1992). Duty and desolation. Philosophy, 67.
Lengbeyer, L. (2004). Racism and impure hearts. In Levine, M. and Pataki, T., editors,
Racism in Mind: Philosophical Explanations of Racism and Its Implications. Cornell Uni-
versity Press.
Leslie, S.-J. (2015). The original sin of cognition: Fear, prejudice, and generalization.
Journal of Philosophy.
Locke, J. (1975). An Essay Concerning Human Understanding. Oxford University Press.
Lorde, A. (1984). Eye to eye: Black women, hatred, and anger. In Sister Outsider, pages
145–75. Crossing Press.
Lynn, M. (2006). Race differences in restaurant tipping: A literature review and discussion
of practical implications. Journal of Foodservice Business Research, 9(4).
Madva, A. and Brownstein, M. (2016). Stereotypes, prejudice, and the taxonomy of the
implicit social mind. Nous.
Mandelbaum, E. (2016). Attitude, inference, association: On the propositional structure
of implicit bias. Nous, 50(3):629–658.
Marusic, B. (2012). Belief and difficult action. Philosopher’s Imprint, 12(18).
Marusic, B. (2015). Evidence and Agency. Oxford University Press.
Marusic, B. and White, S. (2018). How beliefs can wrong—a strawsonian epistemology.
Philosophical Topics.
132
McGarty, C., Yzerbyt, V., and Spears, R. (2002). Social, cultural and cognitive factors in
stereotype formation. In McGarty, C., Yzerbyt, V., and Spears, R., editors, Stereotypes
as Explanations. Cambridge University Press.
McHugh, C. (2012). Epistemic deontology and voluntariness. Erkenntnis, 77:65–94.
McKinnon, R. (2017). Allies behaving badly: Gaslighting as epistemic injustice. In The
Routledge Handbook of Epistemic Injustice. Routledge.
McLeod, O. (2001). Just plain ’ought’. The Journal of Ethics, 5:269–91.
Melton, D. H. (2009). The vulnerable self: Enabling the recognition of racial inequality.
In Tessman, L., editor, Feminist Ethics and Social and Political Philosophy: Theorizing
the Non-Ideal, pages 149–164. Springer.
Mills, C. (2007). White ignorance. In Sullivam, S. and Tuana, N., editors, Race and
Epistemologies of Ignorance. SUNY Press.
Mills, C. W. (1997). The Racial Contract. Cornell University Press.
Mitova, V. (2008). Why w. k. clifford was a closet pragmatist. Philosophical Papers,
37(3):471–489.
Moss, S. (2018). Probabilistic Knowledge. Oxford University Press.
Moss, S. (MS). Moral encroachment. in progres.
Munton, J. (MS). The epistemic flaw with accurate statistical generalizations. in progress.
Nagel, T. (1970). Death. Nous.
Nelson, M. T. (2010). We have no positive epistemic duties. Mind, 119(473).
Nussbaum, M. C. (2016). Anger and Forgiveness. Oxford University Press.
Oswald, F., Mitchell, G., Blanton, H., Jaccard, J., and Tetlock, P. (2013). Predicting ethnic
and racial discrimination: A meta-analysis of iat criterion studies. Journal of Personality
and Social Psychology, 105(2):171–192.
Oswald, F., Mitchell, G., Blanton, H., and Tetlock, P. (2015). Using the iat to predict
ethnic and racial discrimination: Small effect sizes of unknown societal significance.
Journal of Personality and Social Psychology, 108(4):562–571.
Pace, M. (2011). The epistemic value of moral considerations: Justification, moral en-
croachment, and james’ ‘will to believe’. Nous, 45(2):239–268.
Parfit, D. (2011). On What Matters. Oxford University Press.
Parker, P. (1978). Movement in Black, volume
http://condor.depaul.edu/mwilson/multicult/patparker.htm, chapter For the
white person who wants to know how to be my friend.
133
Paul, S. and Morton, J. (2018). Believing in others. Philosophical Topics, 46(1).
Pohlhaus, G. (2017). Varieties of epistemic injustice. In The Routledge Handbook of Epis-
temic Injustice. Routledge.
Preston-Roedder, R. (2013). Faith in humanity. Philosophy and Phenomenological Research,
87(3):664–687.
Preston-Roedder, R. (2017). Civic trust. Philosophers’ Imprint, 17(4).
Preston-Roedder, R. (2018). Three varieties of faith. Philosophical Topics.
Rabinowicz, W. and Ronnow-Rasmussen, T. (2004). The strike of the demon: On fitting
pro-attitudes and value. Ethics, 114.
Ray, K. (2017). Intersectionality and the dangers of white empathy when treating black pa-
tients. http://www.bioethics.net/2017/03/intersectionality-and-the-dangers-of-white-
empathy-when-treating-black-patients/.
Ross, J. and Schroeder, M. (2012). Belief, credence, and pragmatic encroachment. Philos-
ophy and Phenomenological Research.
Sartre, J.-P. (1943). Being and Nothingness: An Essay on Phenomenological Ontology, vol-
ume 2013. Routledge.
Scanlon, T. (1998). What We Owe to Each Other. Harvard University Press.
Schauer, F. F. (2006). Profiles, probabilities, and stereotypes. Harvard University Press.
Schroeder, M. (2012a). Stakes, withholding, and pragmatic encroachment on knowledge.
Philosophical Studies, 160(2).
Schroeder, M. (2012b). The ubiquity of state-given reasons. Ethics, 122(3):457–488.
Schroeder, M. (2018a). Persons as things. In Oxford Studies in Normative Ethics. Oxford
University Press.
Schroeder, M. (2018b). When beliefs wrong. Philosophical Topics, 46(1).
Shah, N. (2002). Clearing space for doxastic voluntarism. The Monist, 85(3):436–445.
Shelby, T. (2003). Review of "i’m not racist, but ...": The moral quandary of race by
lawrence blum. The Philosophical Review, 112(1):124–6.
Shelby, T. (2016). Dark Ghettos. Harvard University Press.
Srinivasan, A. (2016). Would politics be better off without anger.
https://www.thenation.com/article/a-righteous-fury/.
Srinivasan, A. (2017). The aptness of anger. Journal of Political Philosophy.
134
Srinivasan, A. (2018). Does anyone have the right to sex? London Review
of Books, 40(6):https://www.lrb.co.uk/v40/n06/amia–srinivasan/does–anyone–have–
the–right–to–sex.
Stanley, J. (2005). Knowledge and Practical Interests. Oxford University Press.
Steup, M. (2000). Doxastic voluntarism and epistemic deontology. Acta Analytica,
15(1):25–56.
Strawson, P. (1962). Freedom and resentment. In Freedom and Resentment and Other
Essays, volume 2008. Routledge.
Stroud, S. (1998). Moral overridingness and moral theory. Pacific Philosophical Quarterly,
79(2):170–89.
Stroud, S. (2006). Epistemic partiality in friendship. Ethics, 116(3).
Sussman, D. (2005). What’s wrong with torture? Philosophy Public Affairs, 33(1).
Thomson, J. J. (2001). Goodness and Advice. Princeton University Press.
Thomson, J. J. (2008). Normativity. Open Court.
Tiffany, E. (2007). Deflationary normative pluralism. Canadian Journal of Philosophy,
37(5):231–62.
Valian, V. (2005). Beyond gender schemas: Improving the advancement of women in
academia. Hypatia, 20(3).
Velleman, J. D. (2005). A brief introduction to kantian ethics. In Self to Self: Selected
Essays, pages 16–44. Cambridge University Press.
Wallace, R. J. (1994). Freedom and the Moral Sentiments.
Wallace, R. J. (2014). Emotions and relationships: On a theme from strawson. In Oxford
Studies in Agency and Responsibility, volume 2, pages 119–42.
Weatherson, B. (2005). Can we do without pragmatic encroachment? Philosophical Per-
spectives, 19(1):417–443.
Weatherson, B. (2008). Deontology and descartes’ demon. Journal of Philosophy, 105:540–
569.
Weatherson, B. (2012). Knowledge, bets, and interests. In Brown, J. and Gerken, M.,
editors, Knowledge Ascriptions, pages 75–103. Oxford University Press.
Wedgwood, R. (2007). The Nature of Normativity. Oxford University Press.
Wedgwood, R. (MS). The unity of normativity. in progress.
White, R. (2005). Epistemic permissiveness. Philosophical Perspectives, 24:573–615.
135
Williams, B. (1965). Ethical consistency. Proceedings of the Aristotelian Society, 39.
Williams, P. J. (1992). The Alchemy of Race and Rights. Harvard University Press.
Williamson, T. (2000). Knowledge and its Limits. Oxford University Press.
Worsnip, A. (2015). Two kinds of stakes. Pacific Philosophical Quarterly, 96:307–324.
136

Abstract (if available)

Abstract You shouldn’t have done it. But you did. Against your better judgment you scrolled to the end of an article concerning the state of race relations in America and you are now reading the comments. Amongst the slurs, the get-rich-quick schemes, and the threats of physical violence, there is one comment that catches your eye. Spencer argues that although it might be “unpopular” or “politically incorrect” to say this, the evidence supports believing that the black diner in his section will tip poorly. He insists that the facts don’t lie. The facts aren’t racist. In denying his claim and in believing otherwise, it is you who engages in wishful thinking. It is you who believes against the evidence. You, not Spencer, are epistemically irrational. My dissertation gives an account of the moral-epistemic norms governing belief that will help us answer Spencer and the challenge he poses. We live in a society that has been shaped by racist attitudes and institutions. Given the effects of structural racism, Spencer’s belief could have considerable evidential support. Spencer notes that it might make him unpopular, but he cares about the truth and he is willing to believe the unpopular thing. But, Spencer’s belief seems racist. Spencer asks, however, how could his belief be racist if his beliefs reflect reality and are rationally justified? Moreover, how could he wrong anyone by believing what he epistemically ought to believe given the evidence? In answer, I argue that beliefs can wrong.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Positivist realism

PDF

Belief as credal plan

PDF

Reasoning with degrees of belief

PDF

A perceptual model of evaluative knowledge

PDF

Units of agency in ethics

PDF

Rationality and the primacy of the occurrent

PDF

Rethinking reductive realism in ethics

PDF

The case for moral skepticism

PDF

Reasoning with uncertainty and epistemic modals

PDF

Process-oriented rationality

PDF

Contrastive reasons

PDF

Reasons, obligations, and the structure of good reasoning

PDF

A deontological explanation of accessibilism

PDF

Aggregating happiness: seeking and identifying a single plausible unifying theory

PDF

Aggregating complaints

PDF

The scope and limitations of young children’s belief understanding

PDF

Assessing the psychological correlates of belief strength: contributing factors and role in behavior

PDF

Public justification beyond legitimacy

PDF

Mistaken defense and normative conventions

PDF

Responding to harm

Asset Metadata

Creator Basu, Rima (author)

Core Title Beliefs that wrong

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Philosophy

Publication Date 07/25/2018

Defense Date 05/07/2018

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

Tag doxastic morality,doxastic voluntarism,ethics of belief,moral encroachment,normative conflicts,OAI-PMH Harvest,philosophy of race,pragmatic encroachment,wronging

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wedgwood, Ralph (committee chair), Bender, Aimee (committee member), Easwaran, Kenny (committee member), Finlay, Stephen (committee member), Schroeder, Mark (committee member)

Creator Email rima.basu88@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-31418

Unique identifier UC11672483

Identifier etd-BasuRima-6470.pdf (filename),usctheses-c89-31418 (legacy record id)

Legacy Identifier etd-BasuRima-6470.pdf

Dmrecord 31418

Document Type Dissertation

Format application/pdf (imt)

Rights Basu, Rima

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

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