Faith, Flourishing, and Agnosticism

Faith, Flourishing,
and Agnosticism
T. RYA N B Y E R LY
 Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© T. Ryan Byerly 2023
The moral rights of the author have been asserted
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2023941735
ISBN 978–0–19–286571–7
DOI: 10.1093/oso/9780192865717.001.0001
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
 Acknowledgments

There are many individuals and organizations who deserve my gratitude for
their role in supporting this book’s coming to fruition.
In the summer of 2021, I was a fellow with the New Visions in Theological
Anthropology project at St. Andrews University, supported by the John
Templeton Foundation. The project also awarded me follow-­on funding
that supported two of the empirical studies described in Chapters 6 and 7
and provided me with a semester of research leave. I am grateful to the proj-
ect and its leaders and to attendees of the 2022 capstone conference where
I gave a short paper reporting aspects of this work. John Perry, Joanna
Leidenhag, Tasia Scrutton, Simon Hewitt, Jason Stigall, Kate Finley, Dan
Sartor, and Sarah Lane Ritchie, among others, offered helpful feedback. I am
also grateful to the Philosophy Department at the University of Sheffield
and Head of Department Chris Bennett for arranging and supporting my
period of study leave in autumn 2022.
Some of the ideas in this book are ones I have mulled over for a long
time. This is true especially of the ideas in Chapter 5. I first discussed them
informally at The Paradise Project conference, which Eric Silverman and
I co-­organized in 2015. I thank Jerry Walls, Hud Hudson, and Eric for
­discussion. I also presented ideas that formed the basis of this chapter at
a meeting of the Centre for Philosophy of Religion at Leeds, at the
Southampton Workshop on God and Morality, and to an audience at the
University of Birmingham. I am grateful to Mark Wynn, Robin Le Poidevin,
Sophie Grace Chappell, St.John Lambert, and Scott Sturgeon among others
for their conversation. Material from Chapters 5 and 3 was first published in
my chapter “Being Good and Loving God” (2022), and I am grateful to
Oxford University Press for permitting its re-­use here. I am grateful to the
anonymous reviewers and editors of that volume for their comments on
that work as well.
Some of the material in Chapter 6 was first published in my paper “The
Transformative Power of Accepting God’s Love” (2022b), and I am grateful
to Religious Studies and Cambridge University Press for permitting its
re-­use here. I am grateful to Tasia Scrutton, Peter Hill, and Hans Van
vi Acknowledgments

Eyghen for their comments on a draft of that paper. I am also grateful for
the feedback provided by two reviewers for Religious Studies.
Chapter 7 borrows with permission from several paragraphs of my 2021
Psyche article, “How Awesome Natural Beauty Drops the Jaw but Lifts the
Spirit”; I am grateful to the publisher.
I presented some of the arguments from Chapter 3 at the 2022 meeting of
the British Society for the Philosophy of Religion. I am grateful to Maria
Rosa Antognazza, Jason Stigall, and Robin Le Poidevin, among others, for
discussion.
In the summer of 2022 I received a fellowship from the Science Engaged
Theology—­Foundations project led by Meghan Page. I spent a lovely week
with the fellowship cohort in May 2022, learning about the Philosophy of
Science and receiving helpful feedback on some of my planned work for this
book. I am grateful to Kyle Stanford, Meghan Page, Andrei Buckareff, Liz
Jackson, and Eric Yang, among others, for their comments, and I am also
grateful to the funders for this opportunity.
The following individuals also volunteered to read and comment on
aspects of this work: Hans Van Eyghen, Logan Gage, Kevin McCain, Chris
Tweedt, and Sylwia Wilczewska. I am very grateful for their feedback.
I am grateful to OUP editors Tom Perridge and Aimee Wright for their
support, enthusiasm, and guidance.
My immediate family—­Meghan, Tommy, and Samuel—­have been char-
acteristically supportive and patient as I worked on this project. Meghan
was particularly heroic in the summer of 2022 when she took the boys to
Texas for five weeks without me. I am ever grateful to all of you.
Finally, and in the spirit of the argumentation of the book, I am grateful
to God.
 Contents

List of Figure and Tables  ix

Introduction1

PA RT O N E F O U N DAT IO N S F O R PAT H WAYS

T O F L O U R I SH I N G

1. Minimal Theism 11
2. Ambiguous Evidence for God 24
3. Faith and Its Justification 61
4. Virtue and Flourishing 93

PA RT T WO PAT H WAYS F R OM FA I T H
P R AC T IC E S T O F L O U R I SH I N G

5. Giving God the Benefit of the Doubt 117

6. Accepting God’s Love 138
7. Spiritual Excellence 159
Conclusion180

References 185
Index 199
 List of Figure and Tables

Figure
7.1 Moderation of change in connectedness 177

Tables
6.1 Mean scores for mental health for secure theists and different groups
of agnostics150
6.2 Mean scores for mental health for agnostics with different God attachment 153
7.1 Key bivariate correlations 175
 Introduction

How would you assess your evidence for the existence of God? Many of us
find it hard to say. We find that our evidence is ambiguous—­neither strongly
supporting God’s existence nor strongly supporting God’s nonexistence.
If you find yourself in this camp, what should you do? Is it still possible to
respond in faith toward God? Could doing so be good for you in the here
and now? Specifically, could it help you to grow in virtue and thereby to
lead a more flourishing life, even if it turns out there is no God?
This book is primarily concerned with that last question. Using both con­
ceptual and empirical methods, the book argues that many of us indeed do
have ambiguous evidence for God, and that taking up simple practices of
faith toward God can help us in these circumstances to grow in virtue and
attain greater flourishing. The practices on which I will focus are thanking
God, praising God, apologizing to God, accepting God’s love, and cultivating
a sense that some of the awesome things in life reflect God’s grand at­tri­
butes, such as wisdom and love. In short, by faithfully taking up some of the
practices that have often featured in the everyday religious lives of believers,
the many of us who may lack evidence sufficient for belief can move toward
virtue and flourishing.
Given the above description, it will strike many readers that this book’s
focus overlaps with that of other important historical and contemporary
projects in the Philosophy and Psychology of Religion. In Philosophy of
Religion, for instance, readers may wonder about its relationship to Pascal’s
Wager. And in Psychology of Religion, readers may wonder how it relates to
recent research concerned with the value of religious or quasi-­religious rit­
ual practices. Here I explain how the book overlaps with but is distinct from
these projects, thereby helping to contextualize its contributions and high­
light some of its distinctive features, particularly its focus on what I call
“practices of faith.” I also highlight one of the most valuable features of the
book from my vantage point—­that it helps to initiate a new empirical
research area concerned with the value of agnostic faith practices.

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0001
2 Faith, Flourishing, and Agnosticism

Start with the book’s relationship to Pascal’s Wager. As is well known,

“Pascal’s Wager” is a bit of a misnomer. According to many interpreters,
Pascal offers us not one wager, but several. The version of Pascal’s Wager that
is nearest to the argumentative arc of this book runs roughly as follows.1
The likelihood that God exists is about as good as the likelihood that God
doesn’t exist. But believing in God is much better than not believing if God
does exist, and also a bit better if God doesn’t exist. Since believing is better
than not believing in both cases, and both cases are serious pos­si­bil­ities,
you’re better off trying to believe.
This version of Pascal’s Wager is near to the argumentative arc of this
book for two reasons.2 First, this version aims to appeal to people whose
evidence regarding God’s existence qualifies as ambiguous in the sense I will
develop in Chapter 2. I will work with a distinctive and wider notion of
ambiguity than that which seems to be in view in this version of the Wager,
but the focus is nonetheless fairly similar. The notion of ambiguity with
which I will ultimately work is one according to which either a person’s evi­
dence neither strongly supports God’s existence nor strongly supports God’s
nonexistence, or their evidence strongly supports that it neither strongly
supports God’s existence nor strongly supports God’s nonexistence.
Second, this version of the Wager points to both postmortem benefits of
wagering on God, which obtain only if God exists, as well as to ante-­mortem
benefits of doing so, which obtain independently of God’s existence. The
inclusion of a focus on ante-­mortem benefits that do not depend on God’s
existence is probably the most unique feature of this version of Pascal’s
Wager in comparison to the other versions Pascal gives us. In a similar way,
this book will be occupied with ante-­mortem benefits of engaging in prac­
tices of faith toward God that are not dependent upon God’s existence. The
book aims to show that even if God does not exist, a person whose evidence
is ambiguous regarding God’s existence can grow in virtue and flourishing
in the here and now by engaging in practices of faith toward God.
Despite this overlap, the focus of this book is distinct from Pascal’s Wager
in several ways. First, the focus of the book is narrower than Pascal’s Wager.
Pascal’s Wager aims to make an argument in favor of wagering on God
based on the assessment of both the comparative value of wagering given

1 Cf. the fourth version of the Wager discussed in (Rota 2017).

2 The two similarities noted are also similarities between the argumentative arc of this book
and the thought of William James (cf., e.g., James 2010). Chapter 2 discusses some ideas derived
from James in further detail.
 Introduction 3

that God exists and the comparative value of wagering given that God does
not exist. This book, however, does not address in any detail benefits that
may obtain exclusively given God’s existence, but only benefits that may
obtain independently of God’s existence. Moreover, it focuses on only one
such group of benefits—­benefits pertaining to growth in virtue. These bene­
fits, of course, are weighty in value if they obtain. But they are not the only
values one would want to consider when estimating the total value of en­
gaging in practices of faith toward God. Moreover, I do not intend to be
understood as suggesting that individuals with ambiguous evidence for
God should engage in practices of faith toward God solely in an instrumen­
tal fashion in order to attain these goods. Rather, these goods, attainable
through sincere practices of faith that involve attempting to respond to gen­
uine values, are just one consideration alongside others that might contrib­
ute to individuals’ decision-­making in this area. In this way, the book offers
only a partial assessment of the value of engaging in practices of faith
toward God, and not the sort of full assessment of wagering for God that
Pascal requires.
In other ways, the book’s focus is even more different from the Wager. As
it is often interpreted, the Wager treats belief that God exists as an end goal.
“Wagering” itself as Pascal understands it may not require belief. But it tends
to be conceptualized as taking steps that may help put one in a better pos­
ition to believe, and it is regarded as valuable because it can help one believe.
The Wager argument is then only relevant for people who do not yet believe.
This is quite different from the argument given in this book. Here the
focus is on the value of engaging in practices of faith. As explained more
fully in Chapter 3, these are practices that involve positive cognitive com­
mitments to God’s existing and being related in certain ways to oneself and
others, such as loving oneself and others. But, crucially, these cognitive
commitments need not be understood as requiring belief. Following a
growing consensus in Philosophy of Religion, I propose that nondoxastic
assumptions can play a similar role to that played by belief, and that these
may undergird the practices of faith that are the focus in this book. For
instance, a person may belieflessly assume that God exists and has benefit­
ted them, and on the basis of this assumption express gratitude to God. The
book argues that engaging in such practices of faith—­whether they include
belief or not—­can help a person with ambiguous evidence for God to grow
toward virtue.
Because the book focuses on practices of faith understood in this way, it
differs from Pascal’s Wager in at least two important respects. First, it does
4 Faith, Flourishing, and Agnosticism

not treat belief in God as a goal to be pursued. In fact, it is compatible with

the arguments I will advance that belief in God’s existence in the face of
ambiguous evidence is not overall preferable to lack of belief. Second, while
the arguments of this book are likely to be perceived as most relevant for
people who lack belief that God exists (particularly, people on the fence
about whether God exists), they are also relevant for people who already
believe that God exists. Just because someone believes that God exists, this
does not imply that they habitually engage in the kinds of practices of faith
that are the focus of this book. The arguments of this book may help such a
person better evaluate the value of going beyond mere belief and engaging
in these practices of faith.
The book’s focus on practices of faith so understood can also help to dis­
tinguish it from some more recent research in Psychology of Religion con­
cerned with religious or quasi-­religious ritual practice. I have in mind
research on such topics as synchronous movement or communal celebra­
tions of certain transitions in life, such as the transition to adulthood. These
kinds of ritual practices are likely to come to mind for some readers when
they read the term “practices of faith.” And it may very well be that there is
value to engaging in such practices, including conduciveness to virtue
development.3 Yet it is not clear that sincere engagement in these kinds of
practices requires any sort of cognitive commitment to God or ultimate
spiritual realities. Indeed, some authors concerned with the value of these
practices emphasize this very fact, arguing that human beings ought to take
seriously the project of developing practices that mimic these religious
practices without any associated cognitive religious commitments (DeSteno
2021). While I find research on these kinds of faith practices important and
interesting, it is simply not the focus of the argument in this book. Here I
am concerned with whether there may be religious practices that do require
cognitive commitments to God’s existence and activities that may promote
virtue development and flourishing.
Where the project of this book does overlap with research on these other
kinds of faith practices, however, is in its reliance upon empirical research
findings. In total, I will offer four arguments for thinking that engaging in
practices of faith toward God (in my sense) can help a person with ambigu­
ous evidence for God to grow in virtue. Two of these are primarily concep­
tual and two of them are primarily empirical, though all incorporate both

3 For one interesting defense of the latter conclusion that engages with Confucian ritual
practice in particular, see (Stalnaker 2016).
 Introduction 5

elements to some extent. The arguments break new ground conceptually by

addressing how virtue may involve giving God a kind of benefit of the doubt
(Chapter 5) and by addressing the virtuousness of spirituality, which has
only recently begun to attract more attention from philosophers (Chapter 7).
The empirical ground broken is even more novel, and indeed points
toward the value of developing a broader empirical research program
focused on the topic. In Chapters 6 and 7, I will be presenting recent em­pir­
ic­al evidence I have gathered concerned with the values associated with faith
practices among agnostics—­individuals who neither believe that God exists
nor believe that God does not exist. Strikingly little empirical research has
focused on the spiritual lives of these individuals in general. As one group of
authors point out, Psychology of Religion has studied the link between reli­
giosity and well-­being for over 100 years, but it has only turned in the last
decade to consider the link between nonreligiosity (which is construed to
include atheism and agnosticism) and well-­being (Farias and Coleman
2021). One finding that has received support in this emerging literature is
that there is a U-­shaped curve that suggests that agnostics (at the bottom of
the U) may be at a well-­being disadvantage in comparison with both more
confident and committed theists and more confident and committed athe­
ists (Uzarevic and Coleman 2021). This finding suggests that there may be
special reason to attend to the sorts of spiritual practices that might mitigate
against this well-­being risk for agnostics. This book breaks ground on this
topic by addressing the potential value of some agnostic faith practices.
The idea of this U-­shaped curve can also help to illustrate one final general
point I should make about the arguments of this book before turning to a
summary of the chapters. While this book does address the topic of religion
and morality, it does not contend that atheists or agnostics cannot be virtu­
ous, or that theists in general are morally superior to atheists or agnostics.
I have no desire to feed into harmful stereotypes about this topic. Rather, the
book aims to show that for people with ambiguous evidence for God’s
existence (and likewise, though less focally, people who have evidence that
favors God’s existence), engaging in practices of faith can be a source of
moral growth. The aim is to highlight the potential value of engaging in prac­
tices of faith for such people, but I do not in general contend that en­gaging in
these practices is necessary for being virtuous, nor that there are not alterna­
tive strategies for pursuing virtue development that may be of comparable
value. The arguments of the book are offered in a spirit of service for people
who find themselves with ambiguous evidence for God’s existence, who may
be inclined toward agnosticism and thus perhaps at greater risk than others
6 Faith, Flourishing, and Agnosticism

for well-­being deficits. The book explores whether such people may partially
mitigate these well-­being risks through faith practices that can help them
grow toward virtue and experience greater flourishing.
The chapters in Part One lay the groundwork for the four main arguments
developed in the book. They examine key ideas and arguments that are rele­
vant for all four of the pathways to virtue development discussed in Part Two.
Chapter 1 explains how God is understood in the book and why God is
understood this way. I develop an account of what I call “minimal theism,”
which is intended as a view about God that provides just enough content
about God that, if true, most of us would conclude that a God of the sort
envisioned in the Abrahamic religions exists. I illustrate the flexibility of
minimal theism by showing how advocates of a variety of more specific the­
ories of God’s fundamental nature can embrace the view, though min­imal
theism does not require the truth of any of these more specific theories.
Chapter 2 makes a case that there is a sizable population of individuals
with ambiguous evidence for minimal theism. I explain what I mean by
“ambiguous” evidence and contribute to defenses of agnosticism that other
authors have developed by making contributions to the assessment of the­
ism’s intrinsic probability and to the assessment of a sample of theistic and
atheistic arguments. This includes providing an up-­to-­date discussion of my
own previous strategy of responding to the gap problem facing certain
­theistic arguments. I also discuss the relevance of higher-­order evidence for
agnosticism and identify a novel version of agnosticism—­ higher-­
order
agnosticism. By offering this support for a limited form of agnosticism,
I explain why there is likely to be a sizable population of individuals with
ambiguous evidence for God, whether their evidence includes all of the
public evidence for God or only various subsets of it.
Chapter 3 defends a distinctive perspective on the cognitive component
required by the practices of faith that the book focuses on. Following other
authors, I argue that this cognitive component can be satisfied by either
beliefs or assumptions. But uniquely, I argue that ten other candidates for
this cognitive component ultimately fail. I then address the objection that
adopting the beliefs or assumptions required by practices of faith when one
has ambiguous evidence for God is epistemically unjustified. I argue that
these commitments are either not epistemically unjustified, are epi­stem­ic­
al­ly excused, or their epistemic disvalue is outweighed by their moral value.
In developing this argument, I provide the most detailed discussion to date
of the difference between the epistemic norms governing belief and those
governing assumption.
 Introduction 7

Chapter 4 develops and defends simple, attractive accounts of virtues and

flourishing and argues that the character traits that are the focus in Part
Two all qualify as virtues—­including, most controversially, the trait of spiritual
excellence. The chapter explains what is meant by “growth in virtue” in the
book, and why growth in virtue is likely to lead to growth in flourishing as
well. I address the relationship between flourishing and well-­being and
argue that even if flourishing is not the same as well-­being, individuals
should be motivated to maintain or enhance their level of flourishing.
Chapter 5, the first of Part Two, identifies a first pathway whereby indi­
viduals with ambiguous evidence for God can grow in virtue by engaging in
faith practices. The faith practices in view are all ones that involve giving
God the benefit of the doubt, either by praising, thanking, or displaying
contrition toward God. I argue that by giving God the benefit of the doubt
in these ways, individuals with ambiguous evidence for God cultivate more
general tendencies to give other people the benefit of the doubt in similar
ways—­and that these general tendencies are virtuous.
Chapter 6 identifies a second pathway. It argues that individuals with
ambiguous evidence for God can accept God’s love for them and that this
provides an indirect pathway to virtue development because it can boost
their mental health and thereby remove impediments to virtuous behaviors.
As part of the defense of this conclusion, the chapter describes two em­pir­
ic­al studies of my own that support the claim that agnostics who are more
accepting of God’s love for them experience better mental health.
Chapter 7 identifies two further pathways to virtue development that
both focus on a virtue I call “spiritual excellence.” I develop a novel account
of this trait and argue that it is virtuous in itself, and that acting in ac­cord­
ance with it—­whether it is virtuous or not—­promotes growth in other more
standard virtues. I then present an original empirical study that suggests
that agnostics who adopt a more faithful orientation toward God can act in
accordance with spiritual excellence, making use of the worldview of min­
imal theism in order to cultivate transformative experiences of awe and
connectedness.
The Conclusion addresses the significance of the arguments developed
in the book and points forward to avenues for future research that can build
upon them. I address the potential added value of engaging in theistic faith
practices within a community of like-­minded others; the extent to which
the practices I have surveyed have unique value in comparison to other
possible practices; and how these theistic faith practices might be combined
with nontheistic faith practices in interesting and fruitful ways.
 PART ONE
F OUNDAT ION S F OR PAT H WAYS
TO F LOU R ISHING
 1
Minimal Theism

This book identifies four pathways whereby engaging in simple practices of

faith toward God can help a person with ambiguous evidence for God’s
existence to grow in virtue, thereby enabling them to experience greater
flourishing. The first four chapters provide a foundation for articulating
these pathways by clarifying some of the key concepts and arguments
involved in all of them. This chapter begins that work by clarifying how God
is understood in this book, why God is understood in this way, and how
this way of understanding God is very flexible with respect to competing
accounts of God’s fundamental nature. The way of conceptualizing God
identified should be attractive to theorists who advocate a variety of differ-
ent conceptions of God’s fundamental nature, and useful for purposes such
as adjudicating the quality of evidence there is for God’s existence.

1 Why Minimal Theism?

For the purposes of this book, God is understood to be the one who is the
ultimate source of contingent reality, who loves each human person as
much as anyone does, and who has benevolently bestowed each good in
each human person’s life to them. The claim that there is a God of this sort
I call “minimal theism” because it provides a quite minimal characterization
of what is required for God to exist, in contrast to many other conceptions
of God that can be found in the literature in Philosophy of Religion. Before
unpacking these claims about God in more detail in Section 2, I start here
by explaining why I am employing the characterization of God offered by
minimal theism in this book.
I am employing minimal theism in this book because it has three key
features that are valuable for my argumentative purposes. First, minimal
theism provides enough content about God to support the arguments I will
develop in Part Two. If a person adopts positive cognitive attitudes toward
the claims about God made by minimal theism, this will be enough to

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0002
12 Faith, Flourishing, and Agnosticism

underpin the practices of faith that will be my focus in those arguments. For
instance, if a person believes or assumes that what minimal theism claims
about God is true, then they will be in a cognitive position to engage in
practices such as thanking God for having benevolently bestowed the goods
in their life, accepting God’s love for them, and employing the ideas of
minimal theism in order to experience some awe-­ inducing stimuli as
reflecting divine attributes of intelligence and love. I more fully explain the
connection between taking a positive cognitive attitude toward minimal
theism and these faith practices in the chapters concerned with them, but
here I simply note that one valuable feature of characterizing God in accord-
ance with minimal theism is that it provides the content regarding God
needed to underpin the faith practices that are the focus of my arguments.
A second valuable feature of minimal theism is that it secures a concep-
tion of God that if satisfied is just robust enough to serve as the primary
object of religious devotion depicted in the Abrahamic faiths of Judaism,
Christianity, and Islam. This is not an accident but an intentional aim. I am
concerned in this book with whether engaging in practices of faith directed
toward the sort of God envisioned in the Abrahamic faiths can help some-
one with ambiguous evidence for this God to cultivate or maintain virtues.
My claim is not, of course, that minimal theism is all that the Abrahamic
faiths claim to be true about God—­quite the contrary. Nor am I aiming to
take a stand on the question whether practitioners of these faiths “worship
the same God”—a subject of some recent debate (e.g., Bogardus and Urban
2017). Rather, my claim is twofold. First, followers of Abrahamic faiths do
make the claims of minimal theism about God. Second, if it turned out that
of all the claims that followers of Abrahamic faiths make about God, only
minimal theism was true, it would still be reasonable to conclude that a God
of the sort envisioned as the primary object of religious devotion in these
faiths exists.1 Adherents of these faiths will just have gotten certain other
details about the God that exists incorrect.
Jonathan Kvanvig’s (2021) recent work in metatheology is helpful for
supporting this point. Kvanvig’s aim is rather different from mine here. He
aims to take initial steps toward identifying a foundational concept of an
Abrahamic God that can provide an adequate starting point for constructing

1 I suggest that this claim is probably best understood as a conditional of deliberation rather
than as a metaphysically or conceptually necessary truth. Roughly, if we rationally updated our
information with the claim that minimal theism is true, we should conclude that an Abrahamic
God exists. See (Edgington 2020: sect.3) on this approach to the indicative conditional.
 Minimal Theism 13

a full theology. This is a different task from identifying a concept of God

that if satisfied is just enough for it to be reasonable to conclude that an
Abrahamic God exists. Still, the concept of God that Kvanvig arrives upon is
instructive for us here because, while stronger than that required by min­
imal theism, it points in the same two key directions that feature in minimal
theism—(i) that God is the ultimate source and (ii) that God is the proper
object of certain faith practices.
Kvanvig’s preferred conception is that God is the worship-­ worthy
source of all else. This combines what he identifies as two different starting
points for metatheology—­a sourcehood approach and a worship-­worthiness
approach. Kvanvig’s conception of God is a logically stronger conception
of God than minimal theism in that it entails minimal theism but not
vice versa. It is stronger in two respects. First, on Kvanvig’s account, God
is the source of all else and not just the source of contingent reality.
Second, on Kvanvig’s account, God is worthy of the specific faith practice
of worship. Worship, for Kvanvig, involves absolute surrender, submission,
and subordination.
My suggestion here is that while incorporating these two features of into
one’s concept of God may be helpful for metatheological purposes, in that it
may help to identify a more fecund concept of God for constructing a full
theology, it would also remain reasonable to believe that an Abrahamic God
existed if we found that only minimal theism was true and that nothing sat-
isfied Kvanvig’s more lofty conception of God. Kvanvig’s work points us in
the direction of the two key features needed for a minimal concept of an
Abrahamic God—­being the source and being the proper object of faith
practices. But because his own purpose is metatheological, he provides a
conception of God with more than these two features, which is more than
the minimum needed for an Abrahamic God. Minimal theism identifies a
concept of God that likewise has these two features but that remains more
minimal. By leaving out commitments to God being the source of noncon-
tingent realities other than God and to God being worthy of the sort of wor-
ship that Kvanvig identifies, minimal theism avoids commitments about
two potential features of God about which there is significant controversy.2
The fact that this controversy exists suggests that these features are not
required for a minimal concept of an Abrahamic God and so supports

2 For a recent introduction to God’s relationship to other necessities, see (Bøhn 2019); for
discussion of the meaning and appropriateness of worship of God, see (Bayne and Nagasawa
2006) and (Burling forthcoming).
14 Faith, Flourishing, and Agnosticism

thinking that minimal theism identifies a conception of God that if satisfied

is just enough for an Abrahamic God to exist.
If minimal theism does have this feature, then it provides something that
I would suggest is valuable for certain purposes in Philosophy of Religion
that overlap with practical life purposes of the sort with which this book is
concerned. Chiefly, such a conception is valuable for purposes of consider-
ing whether the available evidence supports the existence of God. When we
debate whether an Abrahamic God exists, we get off on the wrong foot if we
begin with a lofty conception of God that requires far more than is min­
imal­ly necessary for an Abrahamic God to exist. I would suggest that this
mistake sometimes occurs and is not always adequately addressed in litera-
ture concerned with theistic and atheistic arguments. Beginning with a con-
ception of God along the lines of minimal theism puts us in a much better
position for this kind of purpose.
The foregoing observations lead us naturally to the final valuable fea-
ture that minimal theism has for my argumentative purposes. My main
goal is to argue that for people with ambiguous evidence for God’s exist-
ence, en­gaging in faith practices can promote their flourishing. However,
I also aim to make the case in Chapter 2 that the number of people who
have ambiguous evidence for God is sizable, and larger than might ini-
tially be thought. Specifically, I want to suggest that some people who may
have thought their evidence for God clearly and strongly supported God’s
nonexistence may in fact have ambiguous evidence for God’s existence.
Thus, I aim to suggest that at least some people who may antecedently be
inclined toward an atheistic position—­believing that God does not exist—­
may in fact be more reasonable or at least excused in adopting an agnostic
position—­suspending belief about God’s existence. Adopting minimal
theism makes this task somewhat easier because less is required for minimal
theism to be true than for more lofty conceptions of God to be satisfied.
This is true even if, as I suggest in some places in Chapter 2, some of the
best arguments in favor of minimal theism provide support for the exist-
ence of a God that satisfies a more lofty conception.
This book employs a minimal theistic conception of God. Because
­min­imal theism is just enough for an Abrahamic God to exist, it makes it
more plausible that a larger number of people (particularly those who may
lean toward atheism) have ambiguous evidence for God’s existence, and it
provides enough content about God to underpin the practices of faith that
are the book’s focus.
 Minimal Theism 15

2 Minimal Theism Explained

While I surveyed above three benefits of employing minimal theism in this

book and perhaps more broadly for purposes of debating God’s existence,
I have not yet unpacked the meaning of the claims that minimal theism
makes about God. That is the purpose of this section, which explains min­
imal theism in more detail.
Minimal theism claims that God is “one”—the one who is the ultimate
explanation for contingent reality. Minimal theism is thus committed to
monotheism. It claims that there is one and only one who plays the roles that
it specifies—­providing the ultimate explanation for contingent reality and
loving each human person and benevolently bestowing each good in each
human person’s life. The notion of oneness involved is not intended to rule
out from the outset conceptions of the oneness of God that claim that God is
a Trinity of persons or something similar, as long as these can maintain as
they purport to that there is only one God. The key idea is simply that there
is only one God of whom all the claims of minimal theism are true.
Minimal theism specifies a cosmic explanatory role for God. God is the
source of contingent reality. Contingent reality is that part of reality that
exists but could have failed to exist. This is understood to include at least all
spatiotemporal entities, from the smallest particle to the largest supernova.3
All of these exist because of God’s sourcing of them. Minimal theism allows
that some contingent things exist at least in part because they are sourced in
other contingent things. But at some point, this sourcing in other contin-
gent things ends with God as the ultimate source. Every contingent thing is
either sourced directly in God or sourced indirectly in God.
Minimal theism is intended to be compatible with multiple approaches
to understanding God’s provision of sourcehood. God may provide source-
hood for contingent reality through a historic creation event in such a way
that contingent reality has a finite history. Alternatively, God may provide
the ultimate explanation for a contingent reality that is infinitely old. God
may be a causal source or a noncausal explanatory source, such as a ground
or an ultimate end, provided that these can sustain the idea that God’s
sourcing explains the existence of contingent entities. To put these ideas
together, minimal theism claims that God provides the ultimate explanation

3 This formulation allows for a substantivalist view of space or time themselves, if desired.
16 Faith, Flourishing, and Agnosticism

for the existence of all the contingent entities there are, without committing to
precisely when or by what metaphysical means God provides this explanation.
Minimal theism also makes claims about God’s love. God’s love for each
human person is as great or greater than the love that any human person
has for any other human person. There is no one who loves any human per-
son more than God does. The conception of love with which I am working
is one derived from Thomas Aquinas and developed in contemporary work
by Eleonore Stump (2006). On this account, loving someone includes two
components—­willing the good for them and willing fitting union with
them. Thus, minimal theism can be understood as claiming that God wills
the good for each human person at least as well as anyone does and that
God wills at least as much as anyone does to have a union with each human
person that is fitting given their relationship. God’s love for human persons
is expressed in part through God’s benevolently securing each good in each
human person’s life—­the third claim of minimal theism—­as well as in God’s
willing interpersonal union with each human person. This latter aspect of
God’s love involves God’s openness to and pursuit of close personal rela-
tionship with each human person.
The third claim of minimal theism is that God benevolently bestows each
good in each human person’s life. Since each good in each human person’s
life will be a contingent thing, God will be its ultimate source. But this third
claim requires more than God’s being the ultimate source of these goods.
God must source these with benevolent intentions for those whose goods
they are. God sources these goods as an expression of love for these human
persons. In this way, the third claim of minimal theism is closely inter­
related to the first and second.
As alluded to in the previous section, by endorsing these claims minimal
theism does not aim to provide anything like a foundational theory of God.
As a candidate for such a theory, it would suffer several defects. Not only
would it be less fecund than other theories, as noted previously, but it may
well be unnecessarily semantically complex. As a theory, it is clunky. This is
because we may suspect that at least some aspects of it can be derived from
other aspects of it. This is especially tempting with regard to the third claim
about benevolent bestowals being derivable from previous claims about
divine sourcehood and love. More generally, we may suspect that there is a
simpler account of the nature of the being of whom minimal theism is true,
which explains well why the claims of minimal theism are true of this being.
Indeed, the project of metatheology can be understood as in part a project
 Minimal Theism 17

of assessing candidates for such a theory. This possibility leads us naturally

to our final topic for this chapter.

3 The Flexibility of Minimal Theism

This section will further illustrate the flexibility of minimal theism by

exploring its relationship to several other influential ways of conceptualiz-
ing God that have received attention in recent Philosophy of Religion—­
many of which aim to provide a theory of God’s fundamental nature. I will
argue that minimal theism is compatible with but does not demand most of
these views. This illustrates how there are many ways that minimal theism
could be true, which is both illuminating in its own right and relevant for
the assessment of its intrinsic probability, which we will discuss in Chapter 2.
First, consider what is probably the most common approach to conceptual-
izing God in contemporary Philosophy of Religion—­perfect being theism.
According to perfect being theism, God is an absolutely perfect being, pos-
sessing every perfection.4 There is some controversy over exactly how to
understand “perfections,” though the rough idea is that for any property that
makes its possessor better or greater, a perfection is the highest possible
degree of that property, if it has one. An example would be maximum or com-
plete power. Power is plausibly a good- or great-­making property, and plaus­
ibly admits of a maximum—­an extent of power beyond which no more power
is possible. If so, maximum power will be a perfection, and God will possess
it. More generally, it is very common for advocates of perfect being theism to
claim that God is an omniGod. In particular, God possesses omnipotence
(maximum or complete power), omniscience (maximum or complete know­
ledge), and omnibenevolence (maximum or complete goodness).
Minimal theism is compatible with perfect being theism but does not
demand it. Supposing that it is possible for an absolutely perfect being to
exist, it is also possible that the claims of minimal theism are true of this
being. It could be, in other words, that the ultimate source of contingent
reality, who loves and benevolently bestows goods to each human being, is
also an absolutely perfect being. Yet, minimal theism does not demand per-
fect being theism. Even if it were not possible for there to be an absolutely

4 For sample contemporary discussions of perfect being theism, see (Morris 1987), (Rogers
2000), and (Speaks 2018).
18 Faith, Flourishing, and Agnosticism

perfect being, it could still be possible for minimal theism to be true. The
being of whom the claims of minimal theism are true needn’t be absolutely
perfect for these claims to be true of it. In particular, minimal theism does
not demand that God possesses the attributes of omnipotence, omniscience,
and omnibenevolence.
This difference between minimal theism and perfect being theism is
worth stressing in the context of this book because there are many objections
to the coherence of the omni- attributes both individually and jointly.
These objections are objections to the existence of God if God is defined
in accordance with perfect being theism, but not if God is defined in
terms of minimal theism.
There is an approach to conceptualizing God that is a variant of perfect
being theism but weaker than that characterized above, which has received
growing attention in recent research. It is most systematically developed by
Yujin Nagasawa in his Maximal God proposal (2017). According to this
view, God is not to be defined as an absolutely perfect being but as the
greatest possible being. In particular, God possesses the greatest possible
combination of power, knowledge, and goodness. This combination may
not include the omni- attributes, and so objections to the coherence of these
attributes need not threaten the existence of a Maximal God.
Again, minimal theism is compatible with the Maximal God proposal
but does not demand it. If it is possible for there to be a Maximal God, then
it may also be possible that the being of whom minimal theism is true is a
Maximal God. The one who is the loving, benevolent source of contingent
reality may also be the greatest possible being and may have the maximal
consistent set of power, knowledge, and goodness. But this is not demanded
by minimal theism. Minimal theism could be true even if there weren’t a
Maximal God.
While the Maximal God proposal does not have the same disadvan-
tages as the form of perfect being theism with which we began, it does
seem to face a different problem for my purposes. It is not clear that it
identifies a concept of God that if satisfied is enough for an Abrahamic
God to exist. This is because of a concern, raised by Jeff Speaks (2018),
that the greatest possible being might not be very impressive. In particu-
lar, the greatest possible being might not be the source of all else and
might not love and ben­evo­lent­ly bestow goods to each human person. If
not, then the concept of the Maximal God is not adequate for the argu-
mentative purposes of this book, which require the existence of a being
that satisfies minimal theism.
 Minimal Theism 19

A second family of approaches to conceptualizing God are deistic

approaches. Deistic approaches tend to emphasize God’s ultimate sourcehood
and intelligence but attempt to distinguish themselves from theistic approaches
by resisting claims about divine intervention into the natural world. The
most significant source of potential tension between deistic views and min­
imal theism derives from minimal theism’s claim that God loves each
human person. As we saw in the previous section, this is understood to
include God’s being open to and pursuing close personal relationship with
each human person. Minimal theism can be squared with deism only if this
aspect of divine love can be understood in a way that does not require inter-
vention of a sort incompatible with deism.
One recently discussed deistic view—­or at least a view that seems to
be regarded as deistic by its primary exponent—­is Paul Draper’s (2022)
“aesthetic deism.” According to this view, “an eternal, non-­physical, omnipo-
tent, and omniscient being created the physical world.” However, this being
“prioritize[s] aesthetic goods over moral ones” and is not morally perfect. It
is not entirely clear why (or whether) Draper regards this as a version of
deism, though the thought may be that the fundamental difference between
deistic and theistic approaches is that the latter prioritize God’s moral good-
ness, and it is this fundamental difference that may make a difference for
what the views say about divine interventions if there is a difference on that
score. At any rate, if it is a version of deism, it may be a version that is com­
pat­ible with minimal theism. This is so as long as God’s prioritizing of aes-
thetic goods over moral ones does not imply that God fails to love each
human person as much as anyone does, or that God does not benevolently
bestow the goods of each human person’s life. It seems to me there may be
ways of developing aesthetic deism that do conflict with these requirements
of minimal theism as well as ways of developing it that do not. So, I would
suggest that some versions of aesthetic deism are compatible with minimal
theism. Yet, again, minimal theism does not require aesthetic deism. As
we’ve seen, it is compatible with omniGod theism, which is inconsistent
with aesthetic deism.
A third family of proposals for conceptualizing God are pantheist pro­
posals. On these proposals, God and the cosmos are not distinct but are in
some sense the same. They may not be numerically identical, but they
may be the same in some other way. Pantheistic conceptions of God tend
to treat God and the cosmos as more intimately related than proposals
according to which God is a separate entity from the cosmos that causes
the latter to exist.
20 Faith, Flourishing, and Agnosticism

I suggest that minimal theism is compatible with a subset of pantheistic

views of God. For a pantheistic view of God to be compatible with minimal
theism, there are two key challenges that must be met. First, the view must
allow that God provides the ultimate explanation for contingent reality.
Second, the view must allow that God is able to exhibit seemingly personal
properties, such as loving humans and benevolently bestowing benefits
upon them. I am optimistic that pantheism can be developed in this way.
For instance, one might take the relationship between God and the cosmos
to be analogous to the relationship between Socrates and seated Socrates.
The cosmos is the way that God is. God is this way contingently, not neces-
sarily, just as Socrates is contingently seated, not necessarily seated.
Moreover, there is a sense in which the cosmos is God, much as there is a
sense in which seated Socrates is Socrates. If it can be maintained that the
cosmos can be a way that a minded thing is as some authors have in­de­pend­
ent­ly argued,5 then this pantheistic view can also overcome the second chal-
lenge and thereby accommodate minimal theism. Of course, minimal
theism does not demand being developed in this way, as it is also com­pat­
ible with views on which God’s relationship with the cosmos is precisely of
the sort that pantheists aim to avoid.
There is a variant of this pantheistic proposal that has recently been
defended by Kenny Pearce (Pearce and Oppy 2022) and which I would sug-
gest is a panentheistic view of God that is compatible with minimal theism,
although Pearce does not himself categorize it this way. On Pearce’s view,
contingent reality is noncausally grounded in a necessary God’s contingent
choices. God’s choices provide the metaphysical grounding of aspects of
contingent reality, perhaps in the way that Socrates provides the meta­phys­
ic­al grounding for seated Socrates—­indeed, some of Pearce’s analogies for
grounding are much like the Socrates case. On such a view, it seems that
contingent reality exists in God, though God is more than contingent
reality—­thus securing a distinctively panentheistic perspective about God.
It seems that on this view there isn’t a robust sense in which contingent real-
ity or its aspects are God, but there is a robust sense in which contingent
reality or its aspects are God’s choices. Minimal theism is compatible with
this panentheistic view, since this view will allow that God is the ultimate
source of contingent reality and that God loves and benevolently benefits

5 See, e.g., (Buckareff 2022).

 Minimal Theism 21

human beings. Of course, minimal theism does not demand development

in these panentheistic terms.
Another perspective on the divine that has received a good deal of atten-
tion in recent years is John Schellenberg’s ultimism. Ultimism is not
intended by Schellenberg to provide an account of the nature of an
Abrahamic God, though it can be developed in this way if it is insisted that
the Ultimate is a Personal Ultimate. For anything to be an ultimate reality
for Schellenberg, it must be triply ultimate—­metaphysically ultimate, axio-
logically ultimate, and soteriologically ultimate. To be metaphysically ul­tim­
ate is to be metaphysically most fundamental; to be axiologically ultimate is
to be of the greatest possible value; and to be soteriologically ultimate is to
be that in relation to which human beings can find the greatest possible ful-
fillment (Schellenberg 2005: 17–38). Schellenberg argues that the publicly
available evidence supports atheism about the existence of a Personal
Ultimate but suspension of judgment regarding the broader truth of
ultimism. He defends the value of having faith in ultimism in much the
sense in which I will defend the value of engaging in practices of faith
toward the God of minimal theism.
Minimal theism is compatible with ultimism. If there is a Personal
Ultimate, then it plausibly follows that minimal theism is true. Being meta­
phys­ic­ally ultimate, the Personal Ultimate will be the ultimate source of
contingent reality. Being an axiologically and soteriologically ultimate per-
son, it will plausibly love and benefit human persons in the ways required
by minimal theism. Yet it is debatable whether minimal theism requires
ultimism. Minimal theism allows that there are necessary realities that are
not ultimately sourced in God, and so God may not be metaphysically most
fundamental but may be equally fundamental to these other necessities. It
takes an argument to show that if there is a God that satisfies minimal the-
ism, this God is axiologically ultimate, unsurpassable in its value; this is not
a straightforward commitment of minimal theism. I do think this argument
can be made persuasively. It requires arguing that if there is a God that satis-
fies minimal theism, this God must be a necessary being, and that given the
additional features it must have to satisfy minimal theism, there could be no
other being that exceeds it in value. Insofar as there are ways such an argu-
ment can be resisted, it can be resisted whether minimal theism implies that
God is axiologically ultimate. Finally, the soteriological ultimacy of the God
of minimal theism is likewise not a straightforward implication of minimal
theism but must be defended by an argument some readers will resist. Much
rides on how the notion of “fulfillment” is understood. And, even if there
22 Faith, Flourishing, and Agnosticism

are some aspects of fulfillment that human beings cannot experience with-
out being rightly related to the God of minimal theism if this God exists,
minimal theism does not straightforwardly imply that these goods do more
than add incrementally to human fulfillment. If that is all they do, this may
run counter to the spirit of soteriological ultimacy, if not its letter. Minimal
theism is consistent with personal ultimism but may not require it.
Finally, consider classical theism. Classical theism is a model of God with
a venerable history in the Abrahamic faiths, being associated with such
leading lights as Augustine, Boethius, and Aquinas.6 What is distinctive of
classical theism is its affirmation that God is timeless, immutable, im­pass­
ible, and simple. It is controversial exactly how to understand each of these
attributes. But for my purposes, it is not essential to delve into the details of
this controversy here. I will, however, note two facts about recent discus-
sions of classical theism that are relevant for my purposes.
First, many contemporary philosophical advocates of theism have been
critical of classical theism because they are concerned that it conflicts with
the concept of God found in lived Abrahamic religious practice and
reflected in the scriptures of the Abrahamic faiths. In particular, the concep-
tion of God as timeless, immutable, impassible, and simple is thought to
conflict with the idea of a personal, interacting, relational God.7 These con-
cerns can be extended to raise suspicions about the compatibility of classical
theism and minimal theism, since minimal theism takes its cue from lived
Abrahamic religious practice and articulates a concept of God as a personal,
interacting, relational agent. While the concerns raised for classical theism
are substantial, there are capable authors who have argued in response that
the God of classical theism can be reconciled with the personal, interacting
God of the scriptures.8 While I won’t evaluate these defenses of classical the-
ism here, I will suggest that if they are successful, then minimal theism is
compatible with classical theism. Or, more weakly, if they are successful,
then there are versions of classical theism with which minimal theism is
compatible. Of course, minimal theism does not require these versions of
classical theism. If God is neither timeless, immutable, impassible, nor sim-
ple but is the source of contingent reality, loves each human person as much
as anyone does, and benevolently benefits each human person with the
goods of their life, then minimal theism can still be true.

6 For a recent explanation and critical discussion of the approach, see (Mullins 2021).
7 See (Bishop and Perszyk 2017) for discussion and references to the literature.
8 One example is (Stump 2016).
 Minimal Theism 23

Second, I note that some advocates of classical theism will be less concerned
to argue that classical theism can be reconciled with minimal theism.9
They will claim that God is not a person and may argue that to describe
God in the way that minimal theism does risks sacrificing God’s transcend-
ence. There may be room to accommodate such descriptions of God as part
of religious practice where it is understood that what is being affirmed does
not provide a deep understanding of God’s nature. Such ascriptions may
have a metaphorical truth to them or be made true by something other than
God’s nature.
I think it is more questionable whether these versions of classical theism
are reconcilable with minimal theism. Minimal theism cannot allow for an
“anything goes” attitude about what makes the claims of minimal theism
true or appropriately affirmed in religious practice. For the arguments of
this book to be cogent, people with ambiguous evidence for God must be
able to believe or assume the claims of minimal theism. If certain ways of
developing classical theism undercut this, then they conflict with minimal
theism as presented here. It should not be surprising that minimal theism
would conflict with such versions of classical theism, since the articulation
of the former is driven by the very practices that have seemed to some
authors to conflict with such versions of classical theism.
This section has illustrated the flexibility of minimal theism. Minimal
theism is compatible with versions of perfect being theism, deism, panthe-
ism, ultimism, and classical theism. Yet it does not require being developed
in any of these ways. If there is someone who is the ultimate source of con-
tingent reality, who loves each human person as much as anyone does, and
who has benevolently bestowed the goods of each human person’s life, then
minimal theism claims there is an Abrahamic God, even if this God is not
perfect, deistic, pantheistic, ultimate, or classical.

9 Here I am thinking of approaches such as those exemplified in (Bishop and Perszyk 2017)
and (Davies 2016).
 2
Ambiguous Evidence for God

The previous chapter explained the concept of God used in this book. This
chapter argues that there is a sizable population of individuals who have
ambiguous evidence for the existence of such a God. This task is important
for my purposes because it establishes that there is a sizable target audience
for whom the book’s arguments are relevant. After clarifying what I mean
by ambiguous evidence, I then defend the existence of this population by
defending a limited form of agnosticism, offering support to a style of argu-
ment for agnosticism that others have developed previously. I contribute to
this style of argument for agnosticism by making contributions to the
understanding of the intrinsic probability of theism, to arguments for the-
ism and atheism, and to how higher-­order evidence about the first-­order
evidence about God is relevant to the ambiguity of evidence for God. My
aim is to build on existing work by offering some unique contributions that
will help to explain why a sizable population of individuals may find them-
selves with ambiguous evidence for God.

1 Ambiguous Evidence

Let me begin by clarifying the claim I wish to defend regarding this sizable
population. First, the claim I wish to defend is concerned with their evi-
dence. More specifically, it is concerned with their total available evidence.
My claim is that there is a sizable group of individuals whose total available
evidence is ambiguous regarding God’s existence.
I won’t make further commitments regarding the exact nature of evi-
dence here. That is, I won’t plump for some specific theory of evidence. But
it is important for my arguments that a person who has available to them
arguments for or against God’s existence, or information about the flexibil-
ity of minimal theism, or who has had private experiences as of God’s pres-
ence or absence or has received testimony from others about their having
had such experiences, can thereby gain evidence concerning God’s existence.

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0003
 Ambiguous Evidence for God 25

This view is one that various accounts of the nature of evidence can
accommodate.
I talk of evidence that is “available” to someone (or “possessed” by
them—­I don’t intend to make a distinction between these here). There is no
consensus among philosophers about what it is for evidence to be available
to someone, and indeed some authors have suggested that it is vague when
evidence is available or possessed.1 Most philosophers will think that there
are few people, only those with significant expertise, whose total available
evidence for God includes all of the public, shareable evidence for God. I’m
inclined toward this view too, and the view will prove helpful for defending
the argument I will develop. But it is not strictly required for this argument
to be successful. By developing an argument for limited agnosticism, I will
aim to offer support for the claim that both the total public evidence for
God and, even more so, various subsets of it are ambiguous. This makes it
possible to defend the view that there is a sizable group of individuals whose
total available evidence for God is ambiguous even if everyone has available
to them all of the total public evidence for God, and much more so if that is
not the case and instead many individuals have available to them only dif-
ferent combinations of limited subsets of this evidence.
This brings us to the notion of ambiguity. Here I do wish to be more pre-
cise about my use of terminology.2 By “ambiguous evidence,” I mean evi-
dence that neither strongly supports God’s existence nor strongly supports
God’s nonexistence. I will discuss a second kind of ambiguity, which I call
“higher-­order ambiguity,” in Section 3.2.5 and suggest that the arguments of
this book may work for individuals whose evidence for God is ambiguous
in that sense as well. But for now, it is easier if we stick with the simpler
account of ambiguity just provided.
Given this definition, “ambiguous” evidence does not have to lack clarity
as the term might suggest, though this is one salient way for evidence to be
ambiguous. A person’s evidence for God’s existence could still qualify as
ambiguous, on my usage, if it was clearly exactly counterbalanced—­
supporting God’s existence exactly as much as supporting God’s non­ex­ist­
ence. It could also qualify as ambiguous if it clearly supported God’s

1 On this point, see the discussion in (Feldman and Conee 2018).

2 Other authors have described the evidence for religious claims as “ambiguous.” This
includes most famously John Hick (1989), as well as Robert McKim (2008), who has developed
a similar approach in more detail. My use of “ambiguous” has much in common with theirs but
differs subtly as well.
26 Faith, Flourishing, and Agnosticism

existence or God’s nonexistence but did not strongly support either. And, of
course, it could qualify as ambiguous if it was vague whether it supported
God’s existence or nonexistence, or if it vaguely supported God’s existence
or vaguely supported God’s nonexistence while not strongly supporting
either. My claim, then, is that there is a sizable population of individuals
whose total available evidence neither strongly supports God’s existence nor
strongly supports God’s nonexistence.
To further clarify the idea of ambiguous evidence, it may be helpful to
explain what I mean by evidence “strongly” supporting a claim, as opposed
to supporting it but not strongly. It is difficult to be precise about what it
takes for evidence to strongly support a claim without begging important
epistemological questions. But the basic idea can at least be illustrated by
appeal to a fairly common way of thinking about epistemic justification.3
According to this approach, for a person to be epistemically justified in
believing a claim, their evidence must sufficiently strongly support that
claim. Not just any level of support for the claim will suffice for belief to be
epistemically justified; a threshold of strength must be crossed for this ep­i­
ste­mic justification to obtain. It is this sort of level of support that I mean to
identify when I talk about evidence “strongly” supporting God’s existence
or not. We might put it this way: for a person’s evidence to strongly support
God’s (non)existence is for it to satisfy the evidential support requirement
for it to be epistemically justified for the person to (dis)believe God exists
according to views of the kind just sketched. Thus, for a person’s total avail-
able evidence regarding God’s existence to be ambiguous, it must be that
their total available evidence does not support God’s existence or God’s
nonexistence sufficiently strongly for belief or disbelief in God to be ep­i­ste­
mically justified according to such views. I say “according to such views”
because I am not aiming to commit myself to these views being correct
here. I am only using them for illustrative purposes to clarify the notion of
strong support.
The foregoing remarks should help to clarify the sort of individuals I have
in mind—­individuals whose total available evidence for the God of mini-
mal theism is ambiguous. My claim is that there is a sizable population of
such individuals. “Sizable” is a very slippery term, but not without content.
We often say, for instance, that there are “quite a few” or a “good deal of ”
such-­and-­such, and this is different from saying there is “virtually no”

3 For a recent discussion of this general view, see (Magalotti forthcoming). For the contrary
view, see (Feldman and Conee 2018).
 Ambiguous Evidence for God 27

such-­and-­such. I am making at least this strong a claim about the population

of individuals whose evidence for God is ambiguous.
Perhaps a good hermeneutic for understanding my use of “sizable” is to
consider the population of agnostics—­individuals who neither believe that
God exists nor believe that God does not exist but suspend belief regarding
God’s existence. I’d say this population is sizable. Indeed, I’ll stipulate that
this is how I’m using “sizable.” It is hard to estimate exactly how big the
agnostic population is on the basis of existing publicly available evidence,
for many reasons. Still, the evidence we do have suggests that agnosticism is
a noteworthy minority stance on the question whether there is a God. It is
noteworthy, in part, because it is, percentage-­wise and in terms of absolute
numbers, large enough that empirical researchers can conduct studies com-
paring this population to theist and atheist populations and can make dis-
coveries, such as the U-­shaped curve discussed in the Introduction, that
suggest this population may be at risk of a well-­being deficit. The population
of agnostics is large enough for the group to appropriately garner public
concern.
In much these same ways, I want to suggest that the population of indi-
viduals with ambiguous evidence for God is sizable. I don’t suggest there is
an exact overlap between the population of agnostics and the population of
individuals whose evidence for God’s existence is ambiguous. Not everyone
who is agnostic about God’s existence is such that their evidence fails to
strongly support God’s existence or nonexistence; likewise, not everyone
who believes God exists or disbelieves God exists has evidence that strongly
supports God’s existence or nonexistence. Indeed, one of my more minor
goals in this chapter will be to suggest that some individuals who disbelieve
God exists have ambiguous evidence for God’s existence. Yet, while there is
not an exact overlap between agnostics and individuals with ambiguous
evidence for God, if I could show that the latter population is roughly as
large as the former or larger, that would demonstrate that it is sizable—­and
this helps us to get some grip on that term.
Of course, the foregoing considerations suggest one simple way to
argue that the population I’m interested in is sizable. Namely, we argue
that the population of agnostics is sizable; that most agnostics are reason-
able in holding their agnosticism and that some theists and atheists
would also be reasonable in being agnostics; and that if these individuals
are all reasonable in being agnostics, then there is a sizable population of
individuals whose total available evidence is ambiguous regarding God’s
existence.
28 Faith, Flourishing, and Agnosticism

I think this simple argument isn’t bad. At least, I think it supports its
c­ onclusion. But it’s not very informative about the sorts of pathways that
can put a person in a position where their total available evidence for God is
ambiguous. The argument suggests that this happens for a sizable group of
people but doesn’t explain how it happens. In the remainder of this chapter,
I want to try to do more to explain how. To do so, I’ll offer support to a
strategy previously used by others to defend agnosticism.

2 Limited Agnosticism

While there is not a standard philosophical usage of “agnosticism” in the

way that there is for “agnostic,” the term is commonly used to refer to a view
about the epistemic status of belief and disbelief in God. Paul Draper (2022),
for instance, describes several different versions of agnosticism of this sort.
What they all have in common is the denial that either theistic belief or
atheistic belief has some positive epistemic status, such as justification or
knowledge, at least for a relevant group of individuals.
Draper describes the relevant group as individuals who are “intellectually
sophisticated,” though he does not explain what this means. I doubt that he
means to imply by the term something laudable about the individuals’ ep­i­
ste­mic character; rather, he is probably aiming to identify those individuals
who have as part of their own available evidence most or all of the public
evidence concerning God’s existence. Draper’s versions of agnosticism,
then, will claim that for people whose total available evidence includes most
or all of the public evidence concerning God, neither theistic nor atheistic
belief has some positive epistemic status. For instance, one version would
say that no intellectually sophisticated person knows whether there is a
God. As we will see, Draper and others have discussed arguments for agnos-
ticism of these sorts.
I will contribute to an argument for a limited version of such agnosti-
cism. The version I will offer support for is limited in two ways. First, I won’t
aim to show that everyone whose total available evidence includes the total
public evidence for God’s existence is such that their total available evidence
for God is ambiguous. I won’t aim to show this because I wish to allow—­
though not insist—­that people can have private, incommunicable evidence
for or against God’s existence that, when combined with the total publicly
available evidence for God’s existence, strongly supports God’s existence or
nonexistence. I’m especially willing to grant this in the case of individuals
 Ambiguous Evidence for God 29

who persistently engage in the very sorts of practices with which this book
is concerned (cf. Lougheed 2018). Perhaps by so engaging they acquire
incommunicable private evidence (e.g., religious experiences) that tips the
scales of support to “strong” for theism. So, the version of agnosticism I will
defend is limited in that it restricts the group of individuals with whom it is
concerned more than Draper’s versions of agnosticism do. My limited ver-
sion of agnosticism is limited not just to the intellectually sophisticated but
to those among them who do not have compelling private evidence for or
against God’s existence.
The second way in which my agnosticism is limited is that it is focused
specifically on evidential support rather than some other epistemic status.
Thus, the version of agnosticism I will offer support for claims that for any
individual whose total available evidence for God includes only the total
public evidence for God, their evidence neither strongly supports God’s
existence nor strongly supports God’s nonexistence. Or, much more simply
put, the total public evidence for God is ambiguous. While I present my
arguments as defending this version of limited agnosticism in the first
instance, I consider in Section 3.2.5 whether some of them might support a
different form of limited agnosticism, higher-­order agnosticism, which
could also work for my later arguments in this book.
I have said I will offer support for arguments for limited agnosticism
because doing so will help to show how it could be that there is a sizable
population of people with ambiguous evidence for God’s existence. Why
would a defense of the claim that the total public evidence for God is ambig-
uous help us see how there could be a sizable population with ambiguous
evidence for God? In several ways, I suggest.
First, and most straightforwardly, if limited agnosticism is true, then any
individuals whose only evidence for God is the total public evidence for
God will be in the target population, having ambiguous evidence for God.
Second, if limited agnosticism is true, any individuals who have all of the
publicly available evidence for God together with only flimsy private evi-
dence will also be in the population. By “flimsy” private evidence I just
mean private evidence that, when combined with the total available public
evidence, does not strongly support God’s existence or nonexistence. While
I wish to allow, as noted above, that some individuals may have nonflimsy
private evidence for God, I would suggest that this is more the exception
than the norm, particularly among individuals not already engrossed in
religious practice. Most individuals whose total available evidence for God
includes the total public evidence but who are not already engrossed in
30 Faith, Flourishing, and Agnosticism

religious practices of the sort discussed in this book will therefore be in the
target population if limited agnosticism is true.
Third, I suggest that there is a kind of trickle-­down effect from those
whose total available public evidence for God includes all the total public
evidence for God to those whose total available public evidence for God is
more limited. The fact, if it is a fact, that the former group’s total available
public evidence is ambiguous will make it more likely that members of the
latter group also have ambiguous evidence. This is because of the fact that
the ambiguous evidence available to the sophisticated will produce evidence
of this evidence for the nonsophisticated, often in the form of testimony.
The presence of sources of ambiguity in the evidence available to the sophis-
ticated makes more available to the nonsophisticated sources of ambiguity.
A similar sort of trickle-­down mechanism helps to explain why, for instance,
human-­caused climate change denial in the general public is often unrea-
sonable. It is unreasonable because of the way the evidence available to
experts supports human-­caused climate change, and this evidence leaves
evidence of itself, making more available to the public evidence in favor of
there being human-­caused climate change.4
Fourth, if limited agnosticism is true, this suggests that as the public evi-
dence for God becomes more widely available, the population of individuals
with ambiguous evidence for God will increase, unless what that evidence
supports changes significantly. On any understanding of availability on
which not everyone has available all of the publicly available evidence for
God, it is plausible that enhancements in technology and communication
are continuing to make available more of the total available public evidence
to more individuals. This suggests that the population of individuals with
ambiguous evidence for God is only likely to grow. In this case, something
close to the paradoxical idea that “the more you learn the less you know”
is true.
Finally, and ultimately of most importance for the strategic purposes of
this chapter, the arguments I will give below in support of existing strategies
for defending agnosticism can support the main contention of the chapter
by identifying various subsets of the total public evidence for God that are
ambiguous. For many individuals, it is likely that their total available evi-
dence for God includes just these subsets of the total public evidence for
God rather than the entirety of that evidence. And, as these subsets are

4 A similar trickle-­down story is defended by (Wykstra 2002) to explain how lay religious
believers can acquire justification for their beliefs from the experts in their community.
 Ambiguous Evidence for God 31

ambiguous, their total available evidence is then likely to be ambiguous,

either because this public evidence is all it includes or because they have
only “flimsy” private evidence in addition to it. Importantly, this pathway
whereby the argumentation below can support the main contention of this
chapter, unlike the four outlined above, is one that can work even if limited
agnosticism is not true.
The support I will offer to defenses of limited agnosticism can therefore
be quite helpful in showing how it could be that a sizable population of indi-
viduals has ambiguous evidence for God. It can be helpful by supporting the
claim that limited agnosticism is true and, therefore, in some or all of the
ways suggested above people may come to have ambiguous evidence for
God. Or, and here it may do its most potent work, it may be helpful by
pointing to considerations that suggest that individuals with access only to
certain subsets of the public evidence for God thereby have ambiguous evi-
dence for God, even if limited agnosticism itself is not true.

3 Pathways to Limited Agnosticism

In this section, I identify several different pathways to limited agnosticism.

They are different ways in which limited agnosticism could be true, or dis-
tinct arguments for limited agnosticism. They fall into two broad camps.
First, there are pathways that appeal only to first-­order evidence concerning
God’s existence. Second, there are pathways that appeal to higher-­order evi-
dence about this first-­order evidence. It is controversial exactly how to char-
acterize the distinction between first-­order and higher-­order evidence. But
the rough idea is that first-­order evidence regarding God’s existence bears
directly on the question whether God exists, while higher-­order evidence
regarding God’s existence is evidence about the nature of the first-­order evi-
dence regarding God’s existence that has bearing on what it supports.

3.1 Pathways from First-­Order Evidence

There are several different pathways whereby facts about the first-­order evi-
dence concerning God’s existence alone could support limited agnosticism.
If the first-­order total publicly available evidence for God were exactly
counterbalanced, this would support limited agnosticism. If it clearly sup-
ported theism or atheism to some precise extent but did not do so strongly,
32 Faith, Flourishing, and Agnosticism

this would support limited agnosticism. If it supported theism or atheism

vaguely but not strongly, this would support limited agnosticism. And if it
was vague whether it supported theism or atheism, this would support lim-
ited agnosticism.
Why think that the total first-­order publicly available evidence for God
may be like this—­neither strongly supporting God’s existence nor strongly
supporting God’s nonexistence? Paul Draper (2002) and Robin Le Poidevin
(2010) have each defended forms of agnosticism on this basis. They argue
that the intrinsic probability of theism is not much lower than the intrinsic
probability of atheism, and that the arguments for theism and atheism are
not individually very strong and collectively cancel each other out so that
the total available public evidence neither strongly supports theism nor
strongly supports atheism. I will add to this style of argument for agnosti-
cism in this section by making some observations about the intrinsic proba-
bility of minimal theism and testimonial evidence for it, and about some of
the arguments for theism and atheism.

3.1.1 The Intrinsic Probability of Minimal Theism and Testimonial

Evidence for Minimal Theism
I will begin with some thoughts about intrinsic probability. In his argument
for agnosticism, Robin Le Poidevin (2010) begins by arguing that “there is
no firm basis” on which to judge that theism or atheism is intrinsically more
probable than the other; they “begin the race on exactly the same line” (76).
The version of theism he is assessing is one which claims that there is a
being that is the ultimate and intentional cause of the universe and the ulti-
mate source of love and moral knowledge. Le Poidevin argues that the
intrinsic probability of this version of theism is not lower than the intrinsic
probability of its denial because the only feature that is relevant for assessing
the intrinsic probability of a claim is its specificity (the more specific it is,
the fewer ways it can be true and so the less probable it is), but it is impossi-
ble to show that theism is more specific than atheism.
My first observation is that the advantage that Le Poidevin’s theism has
because of its lack of specificity is even greater for minimal theism. Le
Poidevin’s theism is not very specific. It does not require, for example, per-
fect being theism or panentheistic theism. Still, it does require that God is
the “cause” of the universe, and that God is the “ultimate source” of both
love and moral knowledge. In the first and third of these respects, it seems
more specific than minimal theism. Minimal theism, as we saw in Chapter 1,
claims only that God is the ultimate source of the universe, allowing that
 Ambiguous Evidence for God 33

this sourcehood may not be causal. And while minimal theism specifies
a loving and benevolent role for God, it does not straightforwardly imply
that God plays an essential role in humans’ acquisition of moral knowledge.
Thus, minimal theism appears to be even less specific than Le Poidevin’s
theism.
The lack of specificity in both minimal theism and Le Poidevin’s theism is
related to another feature that is relevant for their intrinsic probability.
Some versions of theism, such as omniGod theism, face concerns about
their intrinsic probability because of worries that the claims they make
about God’s properties are impossible. This is an assessment of their intrin-
sic probability. Neither Le Poidevin’s theism nor minimal theism face these
challenges to their intrinsic probability.
My second observation has to do with the simplicity of minimal theism,
where this is distinguished from its specificity. The simplicity of minimal
theism includes at least its ontological simplicity—­how many entities and
how many kinds of entities it is committed to. Some authors will contend,
contrary to Le Poidevin, that ontological simplicity is a feature that is rele-
vant for the assessment of a claim’s intrinsic probability. The more entities
and more kinds of entities it claims that there is, the lower its probability.5
Theism does seem to be less simple than atheism, in that it claims there is an
entity, and perhaps even an entity of a different kind, that atheism does not
claim there is. If ontological simplicity is a criterion of intrinsic probability,
then theism may have a lower intrinsic probability than atheism.
I want to grant, at least for the sake of argument here, that ontological
simplicity is a criterion of intrinsic probability and to attempt to neutralize
its effect on the likelihood of theism in two ways. First, I suggest that its
effect can be neutralized, at least to some extent, when we consider the testi-
monial evidence for theism. The criterion of ontological simplicity would
tell us that the intrinsic likelihood that there are apple trees is lower than the
intrinsic likelihood that there aren’t. Yet it seems clear that if lots of people
report that there are apple trees, this intrinsic unlikelihood can be neutral-
ized, and indeed overcome, even if we ourselves have not directly observed
an apple tree. Likewise, I suggest that while the greater ontological com-
plexity of minimal theism may render it intrinsically less likely than athe-
ism, widespread testimony to the existence of a God of whom minimal
theism is true can help to neutralize and perhaps overcome this intrinsic

5 For a defense of the relevance of ontological simplicity for assessments of intrinsic proba-
bility, see (Swinburne 2001: ch.4).
34 Faith, Flourishing, and Agnosticism

unlikelihood. My simple point here is just that if we grant that it is intrinsi-

cally less likely that there are some xs than that there are not because of the
criterion of simplicity, we should also allow testimony to the existence of xs
based on experiences of xs to constitute a significant counterweight to this
when it is present.
Other recent defenders of agnosticism have similarly stressed the eviden-
tial importance of testimony to experiences of God. A good example is
Bryan Frances (2021: esp. ch.9). Frances’s discussion suggests that, from his
perspective, the evidence for theism provided by testimony of experiences
of God is comparable to the evidence against more lofty forms of theism
provided by the evidential argument from evil. Much as it can seem implau-
sible to think that of all the many, many cases where it seems that there is an
evil that God would not have allowed, God indeed would have allowed that
evil in every last case; it can likewise seem implausible to think that of the
many, many cases where some person sincerely reports that they have had
an experience of God, not a single one of these reports was due to the person
in fact experiencing God.
Of course, none of this is to suggest that there are not atheistic approaches
to trying to account for purported religious experiences—­explanations as to
why it would be that so many people would report having God-­like experi-
ences despite there being no God. There are such explanations,6 and investi-
gation of them is important for determining the evidential significance of
this testimony. Frances goes so far as to suggest that future research on this
topic could really help to settle the question of God’s existence once and for
all. But he also acknowledges that research on this topic is in its infancy. At
present, it does not serve to undermine the basic thought that it is at least
somewhat more likely that these reports would exist if God did exist than if
God did not exist, even if it does suggest that the evidence from this testi-
mony is not as strong as in the apple tree case. And it is this basic thought
that I am suggesting can help to neutralize the intrinsic evidence against
theism from its ontological complexity.
A second way of attempting to neutralize to some extent the evidence
against theism from its complexity involves considering the complexity of
what may be the most likely atheistic alternatives to theism. This issue is
intertwined to some extent with arguments for theism, particularly the
argument from contingency as discussed below. Let’s suppose that views

6 For a recent, critical discussion, see (Van Eyghen 2020).

 Ambiguous Evidence for God 35

that offer an explanation of the existence of contingency by appealing to a

necessary being that is its source are more plausible than views that deny
this, as that argument suggests. Some atheists agree with this view and argue
that the necessary being that is the source of contingent reality is not God,
and indeed not anything supernatural.7
Now, I want to suggest that if it is this sort of atheism that we are compar-
ing with minimal theism, then minimal theism does not suffer much of a
disadvantage in terms of its complexity. Both views claim that there is a
necessary source of contingency; theism claims it is a personal being char-
acterized by love and benevolence, and atheism says it is not that but some-
thing natural, such as an initial state of the universe together with its laws.
Defenders of this form of atheism have argued that their view of the neces-
sary source of contingency is simpler than the theist’s view because their
view doesn’t postulate anything supernatural (Oppy 2013). I am not sure
that the terms “natural” and “supernatural” are very helpful for this purpose.8
It is true that the theist ascribes to the necessary source attributes the atheist
doesn’t—­personal attributes. But, then again, the atheist’s view requires
­postulating kinds of things the theist view doesn’t, too—­necessarily existent
initial states and necessary laws of nature.
I am not aiming to argue here that minimal theism is simpler than this
form of atheism. Rather, I am aiming to argue that whatever advantage
there may be for this form of atheism over minimal theism due to its sim-
plicity is not enough to strongly support it. We have two very different views
about the nature of the necessary source of contingency here. Each is sim-
pler than the other in some respects and more complex in other respects.
Deciding them on the basis of simplicity alone seems out of the question.
We need to investigate whether there are arguments that favor one over
the other.
In this section, I have offered some comments concerning the intrinsic
probability of minimal theism aimed at helping to make agnosticism more
plausible. First, minimal theism is not rendered improbable because of its
specificity, since it is very flexible. Second, any intrinsic unlikelihood of
minimal theism due to its complexity may be counterbalanced by testimo-
nial evidence. And third, the unlikelihood of minimal theism due to its

7 A leading defender of this perspective is Graham Oppy. See, e.g., (Oppy 2013).
8 On their vagueness and various ways of precisifying them, see (Papineau 2020). See (Ellis
2014) for an example of a philosopher who aims to reconcile naturalism and theism.
36 Faith, Flourishing, and Agnosticism

complexity in comparison to what some will consider leading atheistic

rivals is quite minimal.
These aspects of the overall publicly available evidence for God are not
the only aspects, of course. However, finding that they do not render either
theism or atheism strongly supported is relevant in two ways for the overall
aim of this chapter, of arguing that there is a sizeable population whose evi-
dence for God is ambiguous. First, it is more likely that the total public evi-
dence will strongly support theism or atheism if the evidence covered in
this section does, and so less likely that it will if this evidence, as argued,
does not. Indeed, if the evidence discussed here is ambiguous and the
remaining public evidence is also ambiguous—­as it may well be—­then
there is a very good chance that the total public evidence is ambiguous.
Second, there may be some individuals who are such that the only public
evidence that is available to them is roughly the evidence discussed in this
section concerning simplicity and testimony. It’s a fairly widely available
fact that the claim that there is a God commits one to there being things
that the denial of this claim does not. It’s also a fairly widely available fact
that there are lots of reports from people claiming to have had experiences
of God. If these sources of first-­order evidence do not together strongly sup-
port theism or atheism, then individuals with access only to these aspects of
the publicly available evidence for God are likely to have ambiguous evi-
dence for God.

3.1.2 Theistic Arguments

In addition to public evidence bearing on the intrinsic probability of theism
and atheism, and testimonial evidence regarding God’s existence, there is
also a well-­developed literature discussing various arguments in favor of
theism. Those who defend agnosticism on the basis of an engagement with
the first-­order evidence for God tend to argue, as noted above, that the
arguments for theism are not very strong and tend to be canceled out by
arguments for atheism, leaving neither theism not atheism strongly sup-
ported by the total public evidence. Now, I can’t hope to address the entire
literature on theistic arguments here, or even to comment on every family
of arguments for theism within it. I will limit myself to offering some com-
ments about theistic arguments that will contribute to making this sort of
broader case for agnosticism. My comments will focus on cosmological and
fine-­tuning arguments. I will suggest that these arguments, considered in
isolation from atheistic arguments, do offer support for minimal theism,
but not strong support for it.
 Ambiguous Evidence for God 37

Start with fine-­tuning arguments. These arguments contend that theism

gains an explanatory advantage over atheism when it comes to explaining
why it is that several of the free parameters of our best current fundamental
physical theories are finely tuned for the universe to support intelligent life.9
These free parameters are numbers that our current physical theories do not
themselves specify but that scientists have to determine by examining
nature. As they are not specified by the theories themselves, there is no basis
in the theories for thinking they could not have had values other than the
ones we observe them to have. Moreover, research with cosmological mod-
els indicates that for many of these parameters, a life-­sustaining universe
would not have been possible if the value for the parameter did not fall
within a narrow range of possible values it could have taken. The question,
then, is why the parameters take values that fall within the narrow range
required for a life-­supporting universe?
Theism offers one potential explanation. If there is a loving, benevolent
personal being capable of sourcing the universe, this being may have wanted
to source a universe capable of sustaining intelligent creatures who could
enjoy its benefits and engage in loving relationship with it. This would raise
the likelihood that the values for the free parameters would fall within the
range needed, potentially many times over what this likelihood is given
background knowledge alone. Unless there is an equally good atheistic
explanation for the phenomenon, its existence would seem to support the-
ism, perhaps even strongly so.
I will discuss two responses to this argument aimed at showing that the
support it offers for theism is not strong—­one, a multiverse response, which
is very commonly discussed, and the second, an anti-­realist response, which
is much less commonly discussed. The second response is more closely
related to the first than is commonly appreciated and is, I suggest, a more
powerful response. After discussing both responses, I suggest a way of
thinking about fine-­tuning arguments according to which they confer a
modest advantage to theism over atheism even granting the success of the
anti-­realist response.
Multiverse responses to fine-­tuning are pitched as offering an alternative,
nontheistic explanation of the fine-­tuning phenomenon. These approaches
maintain that there are many universes, not just one, and that the process
that gave rise to these universes allowed for many chances for there being a

9 For recent sympathetic discussions of these arguments, see (Barnes 2019) and (Waller
2021).
38 Faith, Flourishing, and Agnosticism

universe with parameters such as ours. The fact that there is such a universe
wasn’t so unlikely because there were lots of chances for it.10
Something that is not always appreciated about multiverse responses is
that they are best understood as contending that our current best funda-
mental physical theories are incorrect or incomplete and will ultimately be
replaced.11 More physics is needed for a multiverse than for a universe.
Perhaps this point is not always appreciated, in part because it is rare for
advocates of multiverse theories to try to provide such a replacement phys-
ics. Moreover, in order for multiverse responses to work, not only is a
replacement physics needed, but it must be a replacement physics that does
not itself have finely tuned free parameters. If it did contain finely tuned free
parameters, then the problem would just arise again for the replacement
physics. So, multiverse responses are committed to arguing that the ulti-
mately correct physics is not our current physics but is one that contains no
finely tuned free parameters, and moreover one that gives rise to the ex­ist­
ence of many universes in which the free parameters of our current best
physical theories take different values.
These commitments of multiverse responses weaken them, in my view.
I would suggest that they point us toward a similar response that does not
have all the same commitments and that has independent support. The lat-
ter response is to simply embrace a kind of anti-­realism about the relevant
fundamental physical theories. According to the kind of anti-­realism I have
in mind, our current best fundamental physical theories are not a good
guide to what the ultimately correct physics is. In particular, the fact that
our current best physical theories contain finely tuned free parameters does
not strongly support the claim that the ultimately correct physics will have
such. It is not clear to me that this kind of anti-­realist response has been
given its due in discussions of fine-­tuning. Waller (2021), for instance, dis-
cusses anti-­realist responses to fine-­tuning but focuses on versions of anti-­
realism that concern the proper interpretation of the unobservable entities
postulated in the relevant scientific theories—­a different form of anti-­
realism than I am suggesting.
The anti-­realist response in view here shares in common with multiverse
responses a certain kind of skepticism about our current best physical theo-
ries. The skepticism of multiverse responses is stronger, though—­it requires
rejecting these theories outright and replacing them, whereas the anti-­realist

10 For a recent discussion of multiverse approaches, see (Isaacs et al. forthcoming).

11 Barnes (2019) stresses this point.
 Ambiguous Evidence for God 39

response cautions suspension of judgment about them. But the anti-­realist

response is much better off than the multiverse responses in terms of its
simplicity. It is not committed to there being a multiverse. And it gets about
the same payoff in terms of offering a response to the fine-­tuning argument
without this major commitment.
Moreover, while advocates of the multiverse approach struggle to iden-
tify independent support for the existence of a multiverse, there is already a
tradition of argumentation defending this kind of anti-­realism about scien-
tific theories—­especially fundamental physical theories singled out and
contrasted with other kinds of scientific theories. One prominent approach
here, defended by Kyle Stanford (2006), is to argue that the track record of
research on fundamental physical theories shows us that we are not good at
conceiving of alternatives (especially radically different alternatives) to our
current best theories that fit our data as well or better than these theories.
The history of revolutions in fundamental physical theorizing suggests that
for long periods of time we fail to even identify these possibilities, while
having theories that are quite good in terms of the practical guidance they
offer us. Moreover, features of the way science is practiced today may only
exacerbate this concern, as there are mechanisms that work to keep scien-
tific theorizing conservative rather than revolutionary. On these bases, we
might conclude that our evidence supports only that these theories are
practically useful and not that they are true. Evidence for fine-­tuning won’t
strongly support theism, then, because it doesn’t strongly support the claim
that there are finely tuned parameters in the correct physics.12 No appeal to
multiverses is needed to show this.
While I think the anti-­realist response to the fine-­tuning argument offers
an important objection to that argument, it does not show that consider-
ations pertaining to fine-­tuning offer no support for theism. I think it can
be shown that fine-­tuning considerations do favor theism, but only weakly.
This can be done, in fact, through an a priori argument. One thing that dis-
cussion of the fine-­tuning argument highlights is that theism coheres better
with there being finely tuned free parameters in the ultimately correct phys-
ics. There is, of course, some possibility that there will be such fine-­tuning,
whether we think that the fact that our current best physics is like this

12 An alternative reading of this kind of argument would be that it shows not that our cur-
rent evidence doesn’t strongly support this fine-­tuning but that our current evidence supports
that it does not strongly support this fine-­tuning. This move is analogous to the move I suggest
below as to how the higher-­order evidence for God may support higher-­order agnosticism.
40 Faith, Flourishing, and Agnosticism

makes it likely that this possibility is actual or not. Thus, theism is better
able to tolerate one way that things may be—­the correct physics being finely
tuned—­than atheism is. It isn’t clear why theism or atheism would fare any
better on the other possibility—­of there not being finely tuned free parame-
ters in the ultimately correct physics. So, theism gains a modest advantage
over atheism on this a priori basis. Unfortunately, given scientific anti-­
realism about fundamental physical theories, we are unable to assess how
great this advantage is and so cannot conclude that fine-­tuning consider-
ations strongly support theism.
Move now to arguments for theism from contingency. These arguments
contend that theism gains an explanatory advantage over atheism because
theism better explains why there are contingent beings—­being that didn’t
have to exist—­than atheism does. Arguments from contingency needn’t be
developed in ways that require appeal to necessary principles like the prin-
ciple of sufficient reason. They can be developed more modestly on the basis
of the idea that if theism can provide a better explanation of some phenom-
enon than atheism, this counts in its favor, and that this is true in the case of
explaining the existence of all the contingent beings.13
Why are there the contingent beings that there are? An explanation that
appeals to a source outside these beings themselves will be better than one
that doesn’t. The best sort of explanation that appeals only to the contingent
beings themselves is one that posits an infinite sequence of such beings with
the existence of the beings at any one time in the sequence being explained
by contingent beings that existed at previous times. But while such an expla-
nation may get us some answer to why there are these beings, it does not
entirely remove our puzzlement. Why is there this series of contingent
beings, this history of contingent beings explaining other contingent
beings? Why are there any contingent beings at all? The history, too, is con-
tingent, as is the fact that there are any contingent beings at all. An answer
that appeals to something other than the contingent beings can do more to
remove our puzzlement. It provides a particular sort of explanation of the
contingent beings—­an explanation external to them—­that we commonly
seek. And in doing so, it gains an explanatory advantage over answers that
don’t do this.
But an answer to why there are contingent beings that appeals to some-
thing other than the contingent beings will be appealing to a necessary

13 For recent articulations of such arguments, see (Rasmussen 2021) and (Pearce and Oppy
2022). My presentation in the next paragraph is very similar to Rasmussen’s approach.
 Ambiguous Evidence for God 41

being—­something that couldn’t have failed to exist. The answer will be pro-
posing that the things that exist but didn’t have to exist do because they are
sourced in a being that had to exist. Ultimately, the best answer to why there
are these beings that didn’t have to exist is that there is some being that did
have to exist that sourced these. It’s not that they just happened to exist
(which isn’t an explanation), nor that they sourced themselves (which is not
as good an explanation).
A major difficulty facing theistic arguments from contingency is the gap
problem. This is the problem of getting from the conclusion that there is a
necessary source of contingency to the claim that theism is true. Other the-
istic arguments face gap problems too. For instance, the fine-­tuning argu-
ment may be thought to support the existence of an intelligent and
life-­desiring source of the universe, but this is not all that theism requires.
Admittedly, though, the gap for arguments from contingency is greater than
this. At least with the fine-­tuning argument we have personal qualities in
addition to sourcehood. But the argument from contingency seems only to
yield necessary existence and sourcehood but not anything personal.
Indeed, some leading defenders of atheism, such as Graham Oppy, are
happy to grant that there is a necessary source of contingency, as noted pre-
viously. They simply claim that this necessary source of contingency does
not have the other features necessary for it to qualify as an Abrahamic God.
If arguments from contingency are to support theism, then they need a way
of overcoming this gap problem.
Various approaches to solving the gap problem have been proposed.14
Some proceed one divine attribute at a time, arguing that if there is a neces-
sary, ultimate source of contingency, then this source will also have each
other attribute needed for it to be God. Another approach, one I have con-
tributed to developing, aims to derive the other divine attributes more
quickly as a whole by arguing that the supposition that this source is God is
the best available explanation for why it is a necessary being (or necessary
source of contingency), or why we have found it to be such. I’ll briefly com-
ment on examples of these approaches here, suggesting that they provide
weak support for theism.
Start with an approach of the first kind recently defended by Jonathan
Kvanvig (2021: ch.7). Kvanvig’s starting point is not quite the same as the
intermediate conclusion of the argument from contingency above. He starts
with the claim that there is a source of all else, not just a source of

14 See (Baker-­Hytch forthcoming) for a survey of some approaches.

42 Faith, Flourishing, and Agnosticism

contingency. This source, he argues, must be necessary. And he then proceeds

to argue that there is good reason to think it is agential, having a mind and
will, and that it is omnipotent, omniscient, and omnibenevolent—­or near
enough to be God.
Kvanvig gives two arguments for thinking the source must be agential.
First, characterizing the source as having both a mind and will makes for a
ready explanation for how it can source both contingent truths and neces-
sary truths. Contingent truths are sourced in the source’s will, whereas nec-
essary truths are sourced in its mind. Kvanvig calls this the “beautiful view”
and suggests that nonagential approaches to the source do not have compa-
rable resources for sourcing both necessary and contingent truths in it.
The second argument appeals to the idea that we need to be able to
explain how it is that the source gives rise to the particular character of its
effect—­and here it seems primarily that it is the contingent effect that
Kvanvig has in mind. Kvanvig suggests that an agential approach has a ready
explanation: the precise effect is intended by the agential source—­
understood and willed by it, giving rise to it. He argues that attempts to
explain how a nonagential source gives rise to this particular effect face seri-
ous difficulties.
A natural approach that Kvanvig considers involves linking a nonagential
source to its precise effect via laws, and the best candidate for the kind of
laws for this purpose is natural laws. This approach does cohere well with
the sort of nontheistic view described earlier that Oppy has defended—­an
approach on which the necessary initial state will give rise to its effect via
necessary laws.
Kvanvig’s argument against this approach is a bit difficult to follow. He
seems to think that we should allow that the source could have sourced
something other than it did, in which case there will need to be laws, just as
necessary as the ones that link the source to its actual effect, that link it to
these other possible effects. But then the problem is that by appealing to the
source and the necessary laws, we don’t get any explanation for why we get
the effect there actually is rather than some other effect.
There are details of this argument that could be challenged, but my incli-
nation is to think that those who defend a nonagential view of the source,
like Oppy, will grant that they cannot offer an explanation for why the
source gave rise to its actual effect rather than another effect.15 They may

15 Indeed, Oppy (Pearce and Oppy 2022) denies that there are contrastive explanations in
cases of nondeterministic causation.
 Ambiguous Evidence for God 43

point out, additionally, that something similar is true for the agential view.
For the agential view will claim that God’s intending of the actual effect is
sourced in God, but they will struggle to be able to offer a contrastive expla-
nation for why God sourced this intention rather than another. It seems that
they have only necessary facts about God to appeal to in order to explain
God’s intentions, and these necessary facts won’t generate the sort of con-
trastive explanation needed.
Perhaps there is still reason to prefer the theist’s explanatory gap to the
atheist’s. We may have independent reason to think that in cases of free
choices of a certain kind, there will not be contrastive explanations, and that
God’s creative decision would be a paradigmatic instance of this kind of free
choice. By contrast, the independent reason we have for thinking that con-
trastive explanations will be lacking for other kinds of contingent facts may
not be as strong. To put the point somewhat differently, we may view it as
less of a cost to maintain that the decisions of God would be free than to
maintain that there is genuine indeterminacy in nature of a sort that
requires the absence of contrastive explanations. If so, the agential approach
may gain an edge.
Once the agency of the ultimate source is established, Kvanvig’s argu-
ments for its other attributes proceed more straightforwardly. Omnipotence
(or something near enough) is secured because of the nature of the explana-
tory link between the agential source and its effect. What we are supposing
is that the agential source gives rise to its intended effect of metaphysical
necessity—­that, necessarily, if it wills some effect, that effect arises. This
already secures unlimited conditional power for the agential source. It is
controversial whether more than this is required for omnipotence, and for
purposes of evaluating minimal theism it is unnecessary to consider it.
Unlimited power comes with unlimited power to know. Anything the agen-
tial source wills to know it will know. Since knowledge is valuable and since
with unlimited power it will come at little cost to acquire such knowledge,
we would need some good reason to think that the source wouldn’t have the
knowledge. Without such, there is good reason to think it would be
omniscient.
Finally, and more importantly for our purpose, once the foregoing attri-
butes are established, there is reason to think the source will also be good—­
even perfectly good. Kvanvig’s preferred approach to defending this
conclusion appeals to the idea that God is the source of both contingent and
necessary truths, and that for God to act badly would require an impossible
incoherence in either God’s mind or God’s will. But he also points with mild
44 Faith, Flourishing, and Agnosticism

approval to other approaches to defending the goodness of an omnipotent

and omniscient source that are perhaps more relevant for us. These
approaches argue that there are connections between moral knowledge,
power, and good behavior. The better someone understands what is good
and what is bad, the better they understand that doing the good is good for
them too, and the easier it is for them to do the good, the more likely it is
they will do the good.
Callum Miller (2021) has shown how to develop this sort of argument in
order to show that God, if there is one, is probably good, thereby defusing
the so-­called evil God challenge to theistic arguments. This challenge,
derived from Stephen Law (2010), is a limited form of the gap problem,
which demands that the theist explain why it is that the being that is sup-
ported by arguments such as the fine-­tuning argument or the argument
from contingency is good rather than bad. If the ideas just canvassed are
along the right track, this aspect of the gap problem is less difficult to handle
than other aspects of it. Once we have gotten an agential source of contin-
gency, particularly if it also has unlimited power and extensive knowledge,
it becomes quite likely that it will also be good, even if not perfectly so. This
will offer significant support for minimal theism—­that not only is God the
source of contingency, but God has the properties of love and benevolence
needed for this view.
This argument will offer significant support, that is, only if the initial hur-
dle of agency can be overcome. It’s here that I would urge caution. Kvanvig’s
first argument for the agency of the source is based on the idea that the
source sources not only contingent truths but necessary ones. This is not a
claim established by arguments from contingency. Indeed, Kvanvig is not
arguing that we can get to the agency of the source via arguments from con-
tingency. Still, one might argue that the hypothesis that the source of con-
tingency is God would allow for this “beautiful” view and thereby gain some
explanatory advantage over rival views that are unlikely to allow an expla-
nation for all necessary truths. This explanatory advantage must be bal-
anced, of course, by concerns about the simplicity or specificity of the
hypothesis.
Kvanvig’s second argument is more directly useful for defenders of the
argument from contingency. But, as we saw, it seems that it ultimately
depends on a trade-­off of costs regarding where we allow for noncontrastive
explanations in our theories. As these topics are controversial, it seems this
argument is unlikely to secure strong support for an agential view of the
source of contingency. So, I would suggest these arguments may provide
 Ambiguous Evidence for God 45

weak support for an agential source of contingency. They provide stronger

support for thinking that an agential source of contingency would be good,
particularly if it is also very powerful and knowledgeable. This is important
because what we want to be able to defend is the plausibility of agnosticism
about minimal theism, not agnosticism about there being a “bad God.” We
need to be able to defend the conclusion that a sizable population of indi-
viduals has ambiguous evidence for a minimal theistic God while having
strong evidence against there being a bad God. What is needed is reason for
thinking that if there is an agential source of contingency, it is many times
more probable that it is a minimal theistic God than a bad God. I suggest
that the kinds of arguments given above for the goodness of the agential
source are very useful to this end.
I have discussed Kvanvig’s arguments as an example of arguments that
can be used to try to address the gap problem for arguments from contin-
gency in a step-­by-­step fashion, arguing that the source of contingency must
have other divine attributes. As an example of the other kind of approach to
addressing the gap problem, I will offer an updated discussion of an
approach I have defended in the past (Byerly 2019a), that others have also
advocated more recently.16
This approach begins with a pair of questions: Why is that which is the
source of contingency necessary? Or why have we found through our inves-
tigations that this being has necessary existence? The answer I initially pro-
posed was that the being is necessary, or we have found that it is necessary,
because it is a perfect being—­but no comparably good answer to these
questions is available for the atheist.
While I initially developed this approach as defending a single argument,
I have come to think that in fact there were two different arguments for the
conclusion that I had not sufficiently distinguished in my thinking or writ-
ing at the time (cf. Anderson 2022). I want to suggest here that both argu-
ments have some purchase, though they fall short of providing strong
support for solving the gap problem.
One of the arguments is more metaphysical in nature. It is concerned
with the question: Why is the source a necessary being? In asking this ques-
tion, we aren’t asking for an argument for thinking that the source is a nec-
essary being. We already have that with the argument from contingency.
Rather, we are asking a question about the nature of the being itself. Why is

16 In addition to Pearce as discussed in the text, (Rasmussen 2021) develops a similar

approach.
46 Faith, Flourishing, and Agnosticism

it that the sort of being that this being is should be a necessary being? What
makes it so special compared to all the contingent beings? As with the argu-
ment from contingency as developed above, we don’t presume through
some sort of necessary principle that there must be an answer to this ques-
tion; we only assume that if theism could provide a good answer to this
question while atheism cannot, this would yield some explanatory advan-
tage to theism.
I suggest that theism can offer a better answer to the question than athe-
ism. One way of putting the answer, as I did in the original paper, is to say
that what it is about the nature of the being that is the source of contingency
that explains why it is necessary is that it is a perfect being. As a perfect
being, it possesses all perfections. One of these is necessary existence. This
yields an explanation of why the being that is the source of contingency is
the sort of being to possess necessary existence.
Kenny Pearce’s (Pearce and Oppy 2022) recent work has helped me to
understand better how this more specific strategy can be generalized. Pearce
is also interested in appealing to the nature of God in order to explain God’s
necessary existence. His approach appeals to the idea of God’s “real
­definition”—a definition that would state what God really is, as opposed to
just what is meant by the term “God,” which would provide a nominal
definition. Pearce argues that God’s real definition—­whatever it is—­will
explain why God is a necessarily existent being. Perhaps the real definition
of God is that God is a perfect being, and so the real definition of God
explains why God necessarily exists in exactly the way just outlined. But we
can allow for more flexibility. We may not be sure exactly what the real
definition of God is, but we may contend that God is the sort of being whose
correct real definition will explain why God necessarily exists.
Tien-­Chun Lo (2020) has responded to my argument understood in
accordance with this metaphysical reading and has suggested an atheistic
alternative. Lo’s proposal appeals to a view of modality according to which
whatever being is explanatorily fundamental necessarily exists. Since the
ultimate source of contingency is explanatorily fundamental, we get an
explanation of why it is necessary. Graham Oppy (Pearce and Oppy 2022:
282), interacting with Kenny Pearce, likewise suggests that the atheist can
contend just as well as the theist that the source of contingency, on their
view, has a real definition that will explain why it is a necessary being.
I find this suggestion from Oppy doubtful, and the reasons that make it
seem doubtful to me also make me find Lo’s proposal even more doubtful.
Real definitions are supposed to identify what a thing really is, to uncover
 Ambiguous Evidence for God 47

its essence. It would seem that this project is not going to be carried out well
by identifying extrinsic features of a thing—­its relationships to other things.
But this is just what Lo’s suggestion does. It suggests that what it is about the
source of contingency that explains why it is necessary is that it is explana-
torily fundamental—­a relational property. This isn’t telling us what it is in
itself. It isn’t providing a real definition of the thing.
Indeed, when we consider the sort of thing that, on Oppy’s view, is the
source of contingency, it does seem a mistake to think that its real definition
includes its being explanatorily fundamental. The source of contingency for
Oppy is a state of the universe and its laws. That description of it gets us
much closer to a real definition of the thing. But it also gets us much further
away from an explanation of why the thing is necessary. It’s very tempting to
think that there is nothing about the intrinsic nature of that state itself that
would explain why it rather than some other state is necessary. If it is neces-
sary, this is just because it happened to be the initial state. “Happened to be”
not in the sense that it is metaphysically possible that something else was
the initial state, but in that it is conceptually possible, or not ruled out by the
real definition of the thing, that this was so. There is nothing about the
intrinsic nature of the state itself that would explain why it must have been
the initial state.
This is how things seem to me, but I must confess that I’m not very confi-
dent about this. It requires heavy work to be done by the idea of real defini-
tions, which are controversial,17 and it also requires that the atheist can’t
identify a plausible real definition of the state of the universe that is initial
that would explain why it is initial. Little effort has been given to this latter
task, and I caution that we need to be patient and hear what defenders of the
view have to say. It would be a mistake, in my view, to judge this argument
to strongly support the theist’s ability to solve the gap problem, even if it
does support it.
What of the other, more epistemologically oriented way of developing
my argument? Here the suggestion is to build an argument for thinking that
the source has all perfections because we have found that it has the perfec-
tion of necessary existence. The basic idea is that universal generalizations
are confirmed by observations of their instances. Just as the fact that I have
observed a black raven confirms the claim that all ravens are black, so the
fact that we have come to find that the source has the perfection of

17 For general discussion of real definitions, see (Gupta 2021).

48 Faith, Flourishing, and Agnosticism

necessary existence confirms that the source has all perfections. An appealing
story about why confirmation works this way is that the generalizations
provide the best available explanation for the observations of the instances
in each case. It is because all ravens are black that I have found that this one
is black, and it is because the source has all perfections that we have found
that it has necessary existence.
This version of the argument can be strengthened if there is independent
reason to think it is plausible that the being in question has any other per-
fections, such as perfect or maximal power. I think there is independent
reason to think that the being in question has this attribute, and that the
arguments of Lo and Oppy support this contention. On the atheistic views
they develop, anything that is possible to bring about is something that the
source can bring about. If this is not omnipotence, it is very close to it. It is a
salient candidate for being the perfect or maximal power attribute, like nec-
essary existence is a salient candidate for being the perfect or maximal
ex­ist­ence attribute. Finding support for thinking that the source has both
the perfect existence attribute and the perfect power attribute strengthens
our inference to the claim that it has all perfect attributes, just as finding
additional black ravens strengthens our inference to the claim that all ravens
are black.
While I think this argument does offer some support for bridging the gap
between the necessary being of arguments from contingency and a perfect
being, I also think it faces some problems that render the support it offers
for minimal theism weak. First, the sample size of observations is small,
leading to the inference being weak, even if supportive.18 Second, this ver-
sion of the argument relies more heavily than does the metaphysical version
on the coherence of the notion of “perfections.” But this is a contested
notion. Are there really such things as perfections? Can we be confident
that it will turn out that a being that possesses all of the features that qualify
as perfections on the most attractive theory of perfections is one that will
make minimal theism true? A strong defense of positive answers is required
for this argument to provide strong support for minimal theism.19

18 Anderson (2022) makes this point. She also argues against the general strategy by sug-
gesting that it is analogous to a situation in which I find that a large object is partly blue and
conclude on that basis that is also partly all the other colors. There is, however, an important
disanalogy between the perfections case and the partly blue case. When I observe that some-
thing has a blue part, I also at the same time gain evidence that this part of it is not some other
color. There is nothing analogous in the perfections case.
19 Kvanvig (2021: ch.8) defends a pessimistic outlook on these prospects.
 Ambiguous Evidence for God 49

In this section, I have attempted to offer further support for the kind of
case for agnosticism developed by Draper and Le Poidevin by arguing that
while arguments from fine-­tuning and contingency may support minimal
theism, they do not support it strongly. The anti-­realist response to fine-­
tuning poses a greater challenge to this argument than is sometimes appre-
ciated, and it is difficult for theists to bridge the gap between the existence of
a necessary source of contingency and God, even if they can offer some
support for this conclusion. If these theistic arguments do not strongly sup-
port God’s existence, that makes it more likely that the overall public evi-
dence does not strongly support God’s existence. And if they support God’s
existence weakly, that makes it more likely that the overall publicly available
evidence does not strongly support God’s nonexistence.
In addition to these more muted results, we also found a more positive
result when it came to comparing minimal theism with the hypothesis of a
bad God. In that case we found that it is plausible to think that the probabil-
ity of a minimal theistic God is much greater than that of a bad God. This
allows for the publicly available evidence to be ambiguous about whether
there is a minimal theistic God while strongly supporting the nonexistence
of a bad God, which is a good result for the arguments of this book. What
we desire a defense of is a sizable population whose evidence is ambiguous
regarding the existence of a God who loves and has benefitted them benev-
olently, not a God who hates them.

3.1.3 Atheistic Arguments

I will be more brief and less detailed in my treatment of sample atheistic
arguments and their bearing on minimal theism. My contribution to the
case for agnosticism here will be to suggest that these arguments offer little
direct evidence against minimal theism, and at best serve to undercut some
of the positive support for minimal theism provided by theistic arguments.
My focus will be on the two most widely discussed kinds of atheistic argu-
ments: arguments from evil and arguments from divine hiddenness.
Contemporary defenses of arguments from evil against theism tend to
focus on the threat to theism posed by apparent cases of unjustified evils.20
Unjustified evils are intrinsically bad events that are not required for God to
secure a greater good or prevent a worse evil. Even if God might allow
intrinsically bad events that are required to secure a greater good or prevent

20 An excellent recent defense is (Ekstrom 2021).

50 Faith, Flourishing, and Agnosticism

a worse evil, it is thought to be implausible that God would allow such when
they aren’t required for this. This is because of God’s moral and power prop-
erties. Being perfectly good, God would want to prevent any such events
God could, and being perfectly powerful, God could prevent them all. Yet it
appears that they do happen. There are many cases where it seems to us that
some evil is not required for God to secure a greater good or prevent a
worse evil. This provides significant evidence against God’s existence.
There are, of course, many responses to such arguments on behalf of the-
ism. Some have argued that the presence of cases of apparently unjustified
evil is to be expected even if God exists, and so provides little evidence
against God’s existence (Dougherty and Pruss 2014). This is especially so if
apparently unjustified evils are not the norm of our experience, and if we
are able to see how many evils that may have seemed unjustified at first
glance may in fact be justified on reflection. Others have suggested that we
should expect there to be some examples of unjustified evils (and not just
apparently unjustified ones) if God exists—­that God’s existence does not in
fact preclude such. At least two lines of defense can be given for this. First,
defenders of this version of the argument from evil allow that God may
need to allow some evils in order to accomplish God’s purposes. Yet, it may
be that there is no minimum amount of evil that will do the trick, just as in
more mundane cases there is often no minimum penalty just severe enough
to serve our purposes. If this is true, then God may have to allow some
excess, unjustified evil in order to accomplish God’s purposes.21 Second,
God may allow unjustified evils so that creatures’ lives can make a bigger
difference. If God’s existence implied that every evil is required for securing
a greater good or preventing a worse evil, then creatures will not be able to
make decisions between bringing about unjustified evils or not, or between
preventing them or not. Yet, other things being equal, if creatures can make
such decisions, then their decisions make a bigger difference than if they
can’t. God may be motivated to allow creatures’ decisions to be meaningful
in this way, even if the value of meaningful decisions does not outweigh the
disvalue of the evils incurred by allowing for these.22
A more fundamental problem with arguments from evil for our purposes
is that they do not target the God of minimal theism. They instead target a
loftier conception of God, such as that of perfect being theism or of omni-
God theism. And these loftier conceptions of God play an important role in

21 See (Van Inwagen 1988) and (Tucker 2016).

22 See (Hasker 1992) and (Swenson 2022).
 Ambiguous Evidence for God 51

the way the arguments are defended. For it is by appealing to God’s perfect
goodness that it is maintained that God would never fail to want to prevent
an unnecessary evil; and it is by appeal to God’s perfect power that it is
maintained that God would always be able to prevent any unnecessary evil.
Without these divine attributes, defense of these key claims in the argument
becomes more questionable. Even if we agree that it is likely that a God who
loves each human person as much as anyone does and benevolently benefits
each with all the goods of their life will also be such as to always want to
prevent any unjustified evil and always be able to do so, this will be much
less likely than it is on a more lofty conception of God, such as perfect being
theism. Accordingly, arguments from evil, whatever evidence they provide
against these more lofty Gods, will provide significantly less against the God
of minimal theism.
At least, such arguments provide much less direct evidence against the
God of minimal theism. I would go as far as to claim that when we combine
all of the above pathways of response, arguments from evil do not provide
strong direct evidence against this God. However, what they may do, by
providing a rebutting defeater for them, is cancel out the support for the
existence of this God provided by some theistic arguments. This is because,
as we saw in the previous section, some theistic arguments work to support
the existence of a God of minimal theism by supporting the existence of a
more lofty God, such as the God of perfect being theism. This is true, for
instance, of the second version of my own proposed solution to the gap
problem facing arguments from contingency, and to a lesser extent of the
arguments derived from Jonathan Kvanvig for bridging this gap. I should
also note that it will be true of other theistic arguments not discussed here,
such as ontological arguments. What this suggests is that, insofar as argu-
ments from evil are successful against more lofty versions of theism, this
weakens the support that minimal theism can receive from these kinds of
theistic arguments.
Turn now to arguments from divine hiddenness.23 These arguments aim
to show that God does not exist on the basis of the fact that it seems God
has not done all God could do to ensure that each human person is in a
position to engage in positive relationship with God. God, according to
these arguments, would have a pro-­relationship motivation. If there are any
humans who are not in a position to engage in a relationship with God—­for

23 For a recent defense, see (Schellenberg 2015).

52 Faith, Flourishing, and Agnosticism

example, because they do not believe in God or do not have evidence for
God’s existence sufficient for belief—­then this could only be through their
own fault. If it were not their own fault, then God would have been able to
do something about it. In particular, God could have ensured that they
believe in God or had enough evidence to do so. But if God could have done
this, God would have. So, if there are individuals who through no fault of
their own are not in a position to engage in relationship with God, God
does not exist.
Theists, of course, have their responses to these arguments too. One
approach parallels the above discussion of arguments from evil. We must
consider both how expectable are the apparent cases of no-­fault nonbelief
(or whatever) that there are given both God’s existence and God’s non­ex­ist­
ence. It may be that it is not much less expectable given God’s existence than
given God’s nonexistence that we would have these apparent cases, as sug-
gested above regarding evils.24
Another common response to arguments from hiddenness has been to
stress the possibility of engaging in relationship with God without belief,
and without strong evidence for God’s existence (e.g., Weidner 2019). The
arguments of this book are of a piece with this kind of response. I will be
maintaining that people who have ambiguous evidence for God’s existence
and who may therefore not believe or not be capable of believing in God
may nonetheless exhibit faith toward God and benefit from doing so. In
fact, I take my arguments to contribute to the viability of this response to
some extent by showing that the hypothesis that individuals can do this is
no mere abstract philosophical speculation but a measurable, difference-­
making empirical reality for some individuals. Thus, the broader arguments
of this book may contribute somewhat to weakening the evidence from
divine hiddenness against theism.
But again, as in the case of arguments from evil, there is also a funda-
mental problem with arguments from hiddenness as arguments against
minimal theism. For these arguments are typically pitched as arguments
against the existence of a more lofty conception of God, such as perfect
being theism. And this loftier conception of God does play a role in the
defense of the arguments. In particular, a lofty conception of God’s power
plays a role when considering what God can do to put people in a position
where they can engage in a relationship with God (say, by providing the

24 See (Anderson 2021) for an approach along these lines.

 Ambiguous Evidence for God 53

right sort of evidence for this). Other things being equal, it is less likely
given minimal theism than given a loftier view, such as perfect being
­theism, that God will always be able to do whatever needed to put people in
such a position if they have not resisted it. So, arguments from hiddenness
will be less persuasive as arguments against minimal theism than they are as
arguments against more lofty conceptions of God.
The same point made earlier about the rebutting effect of arguments
from evil applies here. While arguments from hiddenness pose less of a
direct threat to minimal theism than they do to loftier conceptions of God,
they may weaken the overall support for minimal theism by rebutting some
theistic arguments for minimal theism.

3.1.4 Conclusion about First-­Order Evidence

In this section, I have attempted to offer additional support to existing argu-
ments for agnosticism based on an evaluation of first-­order evidence for
God. These arguments contend that the evidence against theism based on
its intrinsic improbability is not strong, and that theistic and atheistic argu-
ments tend to cancel each other out so that the total public first-­order evi-
dence for God is ambiguous. I have attempted to offer support for this
strategy here by arguing (i) that the evidence against minimal theism based
on its intrinsic improbability is weaker than that against other forms of the-
ism, particularly when considered against its most compelling atheistic
rivals and when counterbalanced by evidence from testimony for theism;
(ii) that fine-­tuning arguments and arguments from contingency offer only
weak support for minimal theism; and (iii) that the most widely discussed
arguments for atheism do not offer strong support for atheism, though they
may work to defeat the support minimal theism receives from some theistic
arguments.
These conclusions are helpful in two ways for defending the claim that
there is a sizable population of individuals with ambiguous evidence for
God. First, they make it more likely that the total public evidence for God is
ambiguous. This is because they preclude some of the avenues whereby this
evidence could fail to be ambiguous. If the total public evidence does
strongly support theism or atheism, it must be at least in significant part
because of evidence not surveyed here. If that additional evidence is no less
ambiguous than the evidence surveyed here, as defenders of agnosticism
such as Draper and Le Poidevin would contend, then limited agnosticism is
very likely true. And if it is true, this makes it likely that individuals who
possess the total public evidence for God will have ambiguous evidence for
54 Faith, Flourishing, and Agnosticism

God. This, in turn, makes it likely that other individuals who do not possess
all of the public evidence for God will also have ambiguous evidence for
God, in the ways noted previously.
I don’t want to put too much emphasis, however, on this first way of
defending the conclusion that there is a sizable population of individuals
with ambiguous evidence for God. I am willing to allow that perhaps there
is a convincing first-­order cumulative case argument for theism or atheism
once all of the publicly available first-­order evidence is taken into account.
What is more central for my broader argumentative purposes is seeing that
there are various combinations of significant portions of public first-­order
evidence for God that are ambiguous, including many combinations of such
portions that may exhaust the evidence available to some individuals. Any
combination of the elements of the evidence discussed above that does not
strongly support theism or atheism will do. Likewise, any combination that
adds to these elements additional evidence not surveyed here that does not
render theism or atheism strongly supported will also do. As there are many
such combinations that are likely to be ambiguous and likely to exhaust
what is available to many individuals, there will be many individuals with
ambiguous evidence for God because of the first-­order evidence for God
they possess.

3.2 Pathways from Higher-­Order Evidence

The previous sections discussed aspects of the first-­order evidence for God.
One purpose of this discussion was to support an argument others have
made that the first-­order public evidence for God is ambiguous, and so the
total public evidence for God is ambiguous. But there was also a second
purpose to my doing so. Discussing these aspects of the first-­order evidence
for God also will enable me to illustrate in this section some points about
the higher-­order evidence for God—­that is, the evidence we have about the
first-­order evidence for God. Here I will be discussing several aspects of the
higher-­order evidence for God that may also be taken to support agnosti-
cism about the total public evidence for God. My arguments here resemble
similar arguments for agnostic views developed by Robert McKim (2008)
and John Schellenberg (2007), though for the latter author these consider-
ations are applied to ultimism rather than theism.25

25 (King 2016) also discusses how these kinds of higher-­order considerations might be used
to motivate religious skepticism in the primary way envisioned here.
 Ambiguous Evidence for God 55

3.2.1 Vagueness
One observation about the first-­order evidence for God is that it often
appears vague. What I mean is that it often appears to us unclear, both for
individual aspects of our first-­order evidence as well as for the evidence
taken as a whole, whether and to what extent it supports theism or atheism.
I illustrate in some places above how this is true for me, by hedging in vari-
ous ways and being vague in my assessments of how strongly the evidence
discussed supports theism or atheism. And I am hardly alone in this
approach. It seems to be a very common experience of people who engage
with theistic and atheistic arguments, and for that matter philosophical
arguments more generally. Very often, it appears unclear to us to what
extent, if at all, these arguments support their intended conclusions, and it
may be even less clear to us what is indicated by the overall evidence
for a view.
There are two salient explanations for why our first-­order evidence often
appears this way that may each support agnosticism, though in different
ways. First, it may be that what the evidence supports is in fact vague. On
this approach, the vagueness is out there in the world. There’s no fact of the
matter about what the first-­order evidence supports, or how strongly it sup-
ports what it supports.26 If this view is true, it helps to support agnosticism.
If it is true of the total public first-­order evidence that there is no fact of the
matter about whether it supports theism or atheism, or how strongly it sup-
ports whichever it does support, then it will neither be the case that it
strongly supports theism nor that it strongly supports atheism. But in that
case agnosticism is true. If instead we limit the claim to aspects of our first-­
order evidence, this too will increase the likelihood that the total available
public first-­order evidence is ambiguous. For if there is no fact of the matter
about what some aspects of this evidence support or how strongly they sup-
port it, this makes it more likely that there will likewise not be a fact of the
matter about what the total evidence supports or how strongly it supports it.
So, if the vagueness is in the world as on this first proposal, this supports
agnosticism.
A second salient option is that the apparent unclarity is due to us and
does not reflect vagueness in the world. There is some fact of the matter
about what the first-­order evidence supports and exactly how much it sup-
ports it. But this fact is not clear to us. We can’t tell what the fact is. This

26 For a recent defense of the view that what evidence supports can be vague, see (Feldman
and Conee 2018). I am not aware of any previous application of this view to the evidence
for God.
56 Faith, Flourishing, and Agnosticism

approach, too, may support agnosticism. On this approach, people will have
first-­order evidence for God that may or may not support God’s existence.
They then also have higher-­order evidence that they cannot tell whether
their first-­order evidence supports God’s existence or not, or to what extent
it does. According to a fairly common view of higher-­order evidence, this
combination of first-­order and higher-­order evidence will result in a total
body of evidence that does not strongly support theism or atheism. Indeed,
this will be true even if the first-­order evidence does by itself strongly sup-
port one or the other. The reason why is that the person’s higher-­order evi-
dence that they can’t tell what their first-­order evidence supports somehow
defeats the support offered by their first-­order evidence, regardless of what
it supports.27 Thus, if the apparent unclarity of the first-­order evidence for
God is due to our inability to identify what this evidence supports and how
much, this too may support agnosticism.

3.2.2 Complexity
A second item of the higher-­order evidence about the first-­order evidence
for God is that the latter is very complex. What I mean is that not only is
there a lot of evidence to consider, but the evidence points in different
directions, and some elements interact with others. Above, we considered
only two theistic arguments and two atheistic arguments. But even in con-
sidering only these, it should be clear that the evidence items point in differ-
ent directions and interact with each other. Any one argument may be
defended by several lines of evidence in support of its premises, and there
are objections and replies to consider to its premises too, making for a very
complicated picture, even for individual arguments. This complexity only
multiplies when we consider—­as we did not above—­all of the arguments on
the topic and the way in which support and objections to their premises
interact.
The complexity of the first-­order evidence provides another source of
caution regarding our abilities to properly assess this evidence. This com-
plexity makes it quite likely that any person evaluating this evidence will
have made a mistake somewhere along the way in their evaluation. But
then, anyone who has considered all the available evidence will have reason
to think that they will have made a mistake in their evaluation. And this

27 See (Whiting forthcoming) and (Tiozzo forthcoming) for discussion of views that allow
higher-­order defeat. I discuss an alternative reading of the significance of higher-­order
­evidence in Section 3.2.5.
 Ambiguous Evidence for God 57

reason they will have to think they have made a mistake, when combined
with their first-­order evidence, may no longer yield strong support for
either theism or atheism, even if the first-­order evidence alone did. We have
another potential source of higher-­order defeat. Whatever the first-­order
evidence may support taken on its own, its support may be at least partially
defeated through the observation that the person evaluating the evidence is
likely to have made mistakes in evaluating it.

3.2.3 Instability
Another observation about the first-­order evidence for God that is illus-
trated by the discussion in the previous section is that this evidence is not
stagnant. Across time, we have gained further first-­order evidence that
bears on the question. The fine-­tuning data is one example. Some of the
argumentation surveyed concerning the gap problem is another. Evidence
from the cognitive science of religion that may help to explain why God-­
like experiences are common even if there isn’t a God is another. The
changes in this evidence push in different directions; there’s not a clear
direction in which it all pushes. We have reason to think further changes
will come, and we cannot be confident in which direction they will push.
If the total first-­order public evidence for God is unstable in this way, this
provides us with reason to doubt that its current constitution is representa-
tive of the total public first-­order evidence for God. Yet, if the higher-­order
evidence concerning the first-­order evidence contains this reason to doubt
the latter’s representativeness, then it is plausible that the combination of
the two will not yield strong support for theism or atheism, even if the first-­
order evidence alone does. Evidence to doubt that the total available public
first-­order evidence is representative provides another potential source of
defeat for that evidence, again pushing us toward the conclusion that the
total available public evidence does not offer strong support for theism or
atheism.

3.2.4 Disagreement
A final fact about the first-­order public evidence is that there is persistent
disagreement about it, including among those who understand it the best.
Fellow academics who understand this evidence as well or better than I do
and who read my treatment of the part of it assessed earlier are very unlikely
to completely agree with that treatment. Disagreement will come from both
sides, with some scholars urging that I have underestimated the force of the
pro-­theistic considerations and others arguing that I have underestimated
58 Faith, Flourishing, and Agnosticism

the force of the pro-­atheistic considerations. On all sides of the disagree-

ment are significant populations of sincere individuals with good access to
and understanding of the relevant evidence and the best skills for evaluating
it that humans have.
This kind of disagreement can again provide a further source of defeat for
the first-­order publicly available evidence for God. The fact of persistent
disagreement may signal that even at our best human beings are not reliable
in assessing the total first-­order public evidence for God. When this fact is
combined with the first-­order evidence for God, the conjunction may no
longer strongly support theism or atheism even if the first-­order public evi-
dence alone does. This provides a source of support for thinking that the
total available public evidence for God taken together neither strongly sup-
ports theism nor strongly supports atheism.

3.2.5 Higher-­Order Agnosticism

Throughout this section, I have spoken as if the higher-­order evidence sur-
veyed provides sources of defeat for the first-­order evidence for God such
that in combination the evidence does not strongly support either theism or
atheism, even if the first-­order evidence did strongly support one of these in
isolation. It is a commonly held view among epistemologists that higher-­
order evidence can function in this way. And other authors in the
Philosophy of Religion who have developed arguments for agnosticism on
the basis of such higher-­order features, including Schellenberg and McKim,
take it to function in this way.
However, I note here that while the possibility of higher-­order defeat is
commonly affirmed, it is also a hotly debated topic how exactly such defeat
works, and some authors doubt that it can in fact happen. These authors
tend to locate the skeptical significance of higher-­order evidence elsewhere.
A common alternative approach of this kind suggests that in a situation in
which a person’s first-­order evidence alone strongly supports p, higher-­
order evidence can function to support the claim that their evidence does
not support p, though it does not function to defeat the first-­order evidence
for p itself.28 This yields a situation in which the person’s total evidence sup-
ports p while their total evidence supports that their total available evidence
does not support p. If we link evidential support to justification, it yields a
situation in which a person can be justified in believing p and justified in
believing that they are not justified in believing p.

28 See (Whiting forthcoming) for discussion of views of this kind.

 Ambiguous Evidence for God 59

I want to suggest here that this alternative way of thinking about the
function of higher-­order evidence should make little difference to the argu-
ments I will develop in this book. The kind of situation described is still one
in which a certain kind of ambiguity obtains. For a person in this situation,
from their subjective point of view, it will be unclear what they are to
believe. Indeed, such cases are often described as epistemic dilemmas. This
is a different kind of ambiguity than that which has been my focus in the
preceding subsections. But my suggestion is that it is a kind of ambiguity
that can still function well for my argumentative purposes in this book.
If the features of the higher-­order evidence about the public first-­order
evidence for God are as suggested here, then someone with access to all of
the public evidence for God is in the following sort of situation. Even if their
first-­order evidence does in fact strongly support God’s existence or God’s
nonexistence, what that evidence supports is unclear to them; they have
reason to think that they likely have made errors in assessing what it sup-
ports; they have reason to doubt that it is representative of the total evidence
for God; and they know that there are other competent judges of the evi-
dence who evaluate it differently from them. This is enough to leave them in
a position where they may be bewildered about what to believe. In particu-
lar, they may be justified in believing that their evidence does not strongly
support either theism or atheism. I will suggest that the arguments I will
give in favor of the value of engaging in practices of faith toward God if one
has ambiguous evidence for God apply for this kind of ambiguity as well as
the kind that has been my primary focus. A suitable name for this kind of
view is higher-­order agnosticism. On this view, regardless of whether the
total public evidence for God strongly supports God’s existence or non­ex­ist­
ence, it strongly supports that it doesn’t strongly support either. Even if a
person’s evidence doesn’t exhibit first-­ order ambiguity, it may exhibit
higher-­order ambiguity in which it strongly supports that it exhibits first-­
order ambiguity.

3.2.6 Conclusion about Higher-­Order Evidence

This section has sought to explain how various facts about the higher-­order
public evidence for God may support limited agnosticism about the total
public evidence for God. The facts that the first-­order evidence seems
unclear in what it supports, is complex, is unstable, and is the subject of
disagreement may defeat the first-­order evidence for God such that the total
public evidence does not strongly support either theism or atheism. Or
these facts may instead show that the total public evidence strongly
60 Faith, Flourishing, and Agnosticism

supports that it does not strongly support theism or atheism, even if it does
in fact strongly support one of them, thereby supporting a kind of higher-­
order agnosticism.
Examining this support for agnosticism is helpful in two ways for
defending the claim that there is a sizable population of individuals whose
evidence for God is ambiguous. First, if this support for agnosticism succeeds,
then it shows that individuals whose total evidence for God is the total pub-
lic evidence for God have ambiguous evidence for God, or strong evidence
for thinking that their evidence for God is ambiguous. The ambiguity of
these individuals’ evidence, in turn, can promote ambiguity in the evidence
possessed by other individuals who do not possess all of the public evidence
for God, as explained previously. Second, even if this support for agnosti-
cism does not succeed—­for example, because the higher-­order evidence is
not strong enough to defeat the full body of first-­order evidence—­it may be
that individuals whose total public evidence for God consists of a subset
containing some or all of this higher-­order evidence alongside some more
limited portion of the first-­order evidence will possess ambiguous evidence
for God. If any of the sources of higher-­order ambiguity indeed can defeat
some of the first-­order evidence for God or can strongly support that it, in
combination with this first-­order evidence, does not strongly support the-
ism or atheism, then individuals who possess this combination of higher-­
order evidence and first-­order evidence will have ambiguous evidence for
God. For example, if the higher-­order evidence regarding complexity and
disagreement were enough to defeat the first-­order evidence pertaining to
theism’s intrinsic probability and testimonial evidence of theism, then
someone possessing this combination of evidence would have ambiguous
evidence for God. This is an important observation for defending the popu-
lation claim because these facts about higher-­order evidence—­especially
regarding vagueness, complexity, and disagreement—­do seem to be quite
widely available, even more so than the details of the first-­order evidence.
Thus, these pathways toward limited agnosticism or limited higher-­order
agnosticism may have much to contribute to the main argument of this
chapter that there is a sizable population of individuals with ambiguous evi-
dence for God. Ultimately, whether it is by possessing all of the first-­order
and higher-­order evidence for God or only some subset of this, there are
many avenues whereby an individual may come to possess ambiguous evi-
dence for the God of minimal theism.
 3
Faith and Its Justification

The previous chapters explained how God is understood in this book and
argued that there is a sizable population of individuals whose evidence for
God is ambiguous. The rest of this book is ultimately concerned with the
value for such individuals of engaging in practices of faith toward God.
This chapter begins addressing this topic by explaining what is meant by
“practices of faith” and by responding to an important concern about the
potential epistemic disvalue of engaging in such practices when having
ambiguous evidence for God. I focus primarily in this chapter on the cogni-
tive components of these practices because it is these components of the
practices for which this concern is relevant. I argue, first, that these compo-
nents may be supplied by beliefs or assumptions about God, while arguing
that it is more doubtful that they can be supplied by other cognitive states.
I then argue that the cognitive attitudes required for these practices need
not be ep­i­ste­mically unjustified and that even if they are unjustified, taking
them up may be epistemically excused or the disvalue of doing so may be
outweighed. The upshot is that individuals with ambiguous evidence for
God need not be very concerned by the potential epistemic disvalue
involved in adopting the cognitive attitudes necessary to engage in practices
of faith toward God if engaging in these practices has the value for virtue
development that I argue in later chapters it does.

1 Practices of Faith

The kinds of faith practices on which this book focuses are very simple
practices, such as thanking God for the goods of one’s life, apologizing to
God for wronging those God loves, accepting God’s love for oneself, and
cultivating perceptions of awe-­inducing phenomena as reflecting divine
intelligence or love. These practices are first and foremost actions: they are
things people do, and do intentionally. Moreover, when engaged in as

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0004
62 Faith, Flourishing, and Agnosticism

practices, they are done repeatedly over some interval of time. They come to
form a kind of habit of faith.
Each of these or similar practices plausibly involves a range of compo-
nents, including cognitive components, evaluative components, and cona-
tive components. Indeed, according to a model of faith defended both
independently and jointly by Dan Howard-­Snyder and Dan McKaughan1,
this is true of any sort of faith, whether it is propositional, relational, or
characterological. My thinking about faith practices is heavily indebted to
their work, though we will see in this chapter that I depart from them subtly
with regard to which cognitive states can support faith and with regard to
the epistemic norms governing these states. I also flag here that my aims in
this section are not (like theirs) to analyze faith or faith practices but to
explain how I am using the term “faith practices” in the book. I do aim to
say something about what comprises the kinds of practices described in the
previous paragraph, regardless of whether they are best regarded as faith
practices. The thesis of the book is that engaging in the kinds of practices
described briefly in the previous paragraph and more fully in later chapters
can promote virtue and flourishing for people with ambiguous evidence for
God, not that engaging in faith practices—­however those are ultimately best
analyzed—­can do so.
I will remark only briefly here on the evaluative and conative elements of
faith practices because I will comment on them more thoroughly in later
chapters. When a person sincerely thanks God for goods in their life, this
plausibly requires that they evaluate God’s benevolence toward them posi-
tively. Even if they may feel they do not deserve this benevolence, they view
the benevolence itself as valuable. Likewise, when accepting God’s love, the
person who does this evaluates God’s love, and their acceptance of it, as
valuable. Even the person who apologizes to God for wronging others does
so with a positive evaluation of God’s love for those others and God’s dispo-
sition toward them as a source of forgiveness. Indeed, in each case, these
positive evaluative elements help to explain why the person is engaging in
the practice in the first place.
Moreover, as Howard-­Snyder (2013a) has argued, once these positive
­evaluative elements are in place, it is also plausible that there will be corre-
sponding positive conative components of some kind. A person who evalu-
ates God’s benevolence positively will also desire, at least to some extent, to

1 See, inter alia, (McKaughan 2016), (Howard-­

Snyder 2013a), and (McKaughan and
Howard-­Snyder forthcoming).
 Faith and Its Justification 63

display gratitude to God. A person who evaluates God’s love for them posi-
tively will also, at least to some extent, want to accept this love. Or, if they
find that they cannot desire to give thanks or to accept God’s love, they may
at least desire to have such desires.
My main focus here, however, will be on the first kind of element involved
in these practices—­ their cognitive elements. Both Howard-­ Snyder and
McKaughan propose that the kind of cognitive attitude required for faith
somehow involves taking a stand. For McKaughan in particular, anytime
one acts on faith that p, they must have a cognitive attitude toward p that
constitutes taking a cognitive stand on p.2 Indeed, it does seem that in order
to engage in the kinds of practices of faith that are my focus, such a cognitive
attitude is required. In giving thanks to God for the goods of your life, you
take a cognitive stand on the claim that God has benevolently benefitted
you with those goods. In accepting God’s love for you, you take a cognitive
stand on that love being there for you to accept. And so on. I will return to
this topic and discuss it more fully in later chapters, further defending the
idea that these practices require adopting cognitive attitudes that involve
taking a stand on the claims of minimal theism or their consequences for
oneself. But for now, let us take this as a working assumption and allow it to
guide our thinking about which cognitive states could fulfill the needed
role. Which cognitive states can play this stand-­taking role, needed to
underpin practices of faith?
The most obvious candidate is belief.3 If a person believes that God has
benevolently benefitted them with the goods of their life, this is a way for
them to exhibit a cognitive attitude that constitutes taking a stand on God’s
having done this. By believing, they take a stand on God’s having benefitted
them in at least the sense that, if God has not done this, their attitude is in
error. They have a false belief. Indeed, for this result to obtain, it should not
matter which theory of belief is correct. Rather, yielding the result would
be a plausible constraint on a theory of belief. Part of the very nature of a
belief that p is taking a cognitive stand on p. We might even be tempted

2 See (McKaughan 2016) and (Howard-­ Snyder 2017). Howard-­ Snyder (2013a) and
McKaughan and Howard-­Snyder (forthcoming) say instead that propositional faith that p
requires a disposition to take a stand on p. Mark Wynn offers a similar view of spiritual practices
as involving practical actions in which cognitive commitments are “embedded in practical
commitments” (2020: 192). Cf. also (Cottingham 2003: ch.3).
3 Some, such as John Schellenberg (2005), will object to belief playing this role because they
claim that faith is incompatible with belief. See Howard-­Snyder (2013b) for a critical discus-
sion of this view.
64 Faith, Flourishing, and Agnosticism

to think this is the distinctive province of belief—­that no other cognitive

state involves taking a cognitive stand like belief does.
If that view is correct, however, it will pose a significant challenge to the
argumentative arc of this book. The problem is that belief is widely regarded
to be involuntary. At least, it is not directly voluntary.4 Nobody can believe a
claim just through an act of will. And, moreover, belief is sensitive to evi-
dence, at least in the sense that if a person’s evidence does not strongly sup-
port a claim and they believe that their evidence does not strongly support a
claim, this either prevents belief or poses a very significant obstacle to it.
But if these things are true of belief, then people with ambiguous evidence
for God are unlikely to be able to hold the beliefs about God required to
sustain faith practices. Such people either have evidence that does not
strongly support God’s existence or they have evidence that strongly sup-
ports that their evidence does not strongly support God’s existence. Yet the
sort of argument I wish to make is that people whose evidence for God is
ambiguous can decide to engage in practices of faith, and that doing so can
be a source of virtue development and flourishing for them. To facilitate this
argument, it will prove very helpful if there is some other cognitive state
that also constitutes taking a stand in much the way that belief does but that
does not have these same features of involuntariness and sensitivity to evi-
dence that belief appears to have.
Thankfully, thought experiments suggest that there is such a state. Or, at
least, they suggest that there is a state that involves taking a cognitive stand
on p and that does not require the above features of involuntariness or sen-
sitivity to evidence—­whether it is best regarded as a special type or case of
belief or as another kind of state.5 Following Howard-­Snyder (2013a), I will
call this state an assumption.6
Howard-­Snyder (2016) offers a case of a defensive football captain and an
army general, tweaked from earlier cases developed by William Alston
(2007), to make the point. In both cases, the protagonists find that they do

4 For discussion of this general topic, see (Vitz n.d.).

5 I thus intend to remain open about whether there is only one state, belief, which can serve
as the cognitive state underpinning faith. For recent defenses of this view, see (Mugg forthcom-
ing) and (Rettler 2018). I differ from these authors, however, in suggesting that for belief to be
the only cognitive state that underpins faith, it must be capable of being under direct voluntary
control. For a defense of the view that belief can be under voluntary control, see (Rinard 2019a).
6 As Howard-­Snyder notes, there are states we sometimes call “assumption” that do not fit
the profile envisioned here—­e.g., where someone “assumes” a claim merely for the sake of an
argumentative reductio. It is only a particular use of the term “assumption” that is applicable here.
 Faith and Its Justification 65

not have enough information to go on in order to form a belief about some

topic needed to guide a decision they must make. For the football captain,
the topic is what play the opposing quarterback will call, and for the army
general, the topic is how the enemy general has positioned their troops.
While they do not have enough information to form belief, they each find
that of the credible options, there is one that seems more likely than the
others. Each protagonist assumes that this option is how things are and uses
this assumption to guide their action, positioning their players or troops
accordingly.
In these cases, it seems that the protagonists take a cognitive stand on the
topic at hand—­which play the opposing quarterback will call or how the
enemy has positioned their troops. By making assumptions about these top-
ics, they take a stand at least in the sense noted above that they expose
themselves to cognitive error. If things are not as they assume them to be,
then their assumptions are false. Moreover, making these assumptions is a
voluntary matter. It is something they choose to do under the circumstances
in order to guide their activity. They decide to act on these assumptions.
Finally, while the assumptions they make are guided by the evidence and in
that sense are sensitive to it, they are not sensitive to the evidence in the
stronger sense noted above. For the protagonists to make the assumptions
they do, it does not have to be that their evidence strongly supports what
they assume, or that their evidence strongly supports that their evidence
strongly supports what they assume. They can assume what they do even if
their evidence is ambiguous or it strongly supports that it is ambiguous.
Thus, it seems that there is indeed a state, which I am calling assumption,
which has the features needed to support the argumentative arc of this
book. Individuals with ambiguous evidence for minimal theism can engage
in faith practices that require cognitive commitments to minimal theism
and its consequences for themselves that constitute taking a cognitive stand
on these topics. If they cannot do this by believing minimal theism, they
can do it by assuming minimal theism. By assuming minimal theism, they
can voluntarily take a cognitive stand on it despite having ambiguous evi-
dence for that position.
What I wish to investigate in the remainder of this section is whether
there is any other state besides belief and assumption that can fulfill the role
of the cognitive state required by practices of faith. I will consider ten other
candidates for fulfilling this role. I argue that in none of these cases is it clear
that there is a cognitive state distinct from assumption and belief that con-
stitutes taking a stand on its topic, and so it is not clear that there is any state
66 Faith, Flourishing, and Agnosticism

other than belief or assumption that can play the role of the cognitive state
needed to underpin practices of faith. In defending this view, I part com-
pany from Howard-­Snyder and McKaughan to some extent because they
have either explicitly advocated that some of these alternative states can play
the role in question or have briefly suggested that they might be able to.7 In
other cases, I am considering proposals that they do not identify but that
readers may wonder about. Thus, I offer a critical and more thorough explo-
ration of the candidates for fulfilling this role than has been provided in
previous scholarship.
One suggestion is that the cognitive attitude needed for faith practices
may be the sort of attitude that is expressed by exclamations such as “Thank
God!” Notably, individuals who disavow belief in God can still make such
exclamations when something good happens. (And, indeed, they can say
more negative things like “Goddamn it!” when something bad happens.)
Perhaps there is some cognitive attitude toward God, distinct from belief
and assumption, that underpins these expressions and that could play the
role of the cognitive attitude needed for faith practices.
Initially, this proposal sounds like a total nonstarter. The expressions
referenced are simply cultured (or brutish, depending on your point of
view) ways of expressing approval or disapproval. If there is a cognitive
content underpinning them, it needn’t have anything to do with God. It
may simply be an evaluative assessment of a state of affairs, such as this is a
good thing. Or, if we insist that it must have some content about God in it, it
may be a hypothetical assessment, such as this is something God would
approve of, if there were a God. Whatever cognitive content it has, if it has
any, is not a content that allows the state to constitute taking a stand on
there being a God.
Nonetheless, while this initial suggestion is a nonstarter, it may point us
in the direction of a more interesting and less obviously problematic sug-
gestion. Drawing on research in the cognitive science of religion, it might
be proposed that while nonbelievers explicitly disavow belief in God, much
to their chagrin they may in fact have implicit beliefs in God. Indeed, some
researchers have argued for precisely this view, claiming that there are uni-
versal cognitive mechanisms that give rise to belief in God so naturally that,
despite peoples’ best efforts to undo their work, implicit belief in God is

7 The cases of explicit advocacy are noted later in this section. See (McKaughan and
Howard-­Snyder forthcoming) for an example of their briefly suggesting other candidates.
Jackson (2022) also briefly suggests several of the candidates outlined in this chapter.
 Faith and Its Justification 67

inescapable (Barrett 2004; Bering 2010). Even if it can be escaped depending

on one’s cultural experiences, many self-­avowed atheists do not escape it
because of their exposure to religious practices (Hitzman and Wastell
2017). Thus, it might be proposed that in addition to (explicit) belief and
assumption, another state that could underpin practices of faith could be
implicit belief.
However, several difficulties face this proposal. First, these implicit beliefs
are not voluntary. Indeed, in the case of atheists and, perhaps, some agnos-
tics, their possession will be contrary to the individual’s wishes. Because of
this, they are not good candidates for underpinning practices of faith, where
those practices are conceived of as voluntary actions.
Second, it is questionable whether these states involve taking a cognitive
stand for what they are about. One way to try to show this is by pointing
out their compatibility with contradictory explicit beliefs. By hypothesis of
those who claim there are such states, an implicit belief that there is a God is
straightforwardly, nonproblematically compatible with an explicit belief
that there is no God. But does a person who implicitly believes there is a
God and explicitly believes there isn’t take the cognitive stand that there is a
God? Perhaps McKaughan and Howard-­Snyder could rule out this possibil-
ity by adjusting their account so that it requires an overall cognitive stance
that constitutes taking a stand for p. It would be questionable whether
implicit belief would be adequate to secure such because of its straightfor-
ward compatibility with cognitive states that involve the contrary stance.
Notably, it is far more questionable whether explicit belief that p is straight-
forwardly compatible with explicit belief that not-­p.8 And the same seems
to ring equally true of (explicit) assumption. Perhaps, then, the lesson to be
learned from considering these states is that what is needed for practices of
faith is a sort of cognitive state that involves taking a stand on p where the
very nature of the state rules out or works to undermine any cognitive states
involving a contradictory stand. If so, then implicit beliefs will not be the
right kind of state, while beliefs and assumptions will be.
Finally, as some researchers have argued (see Coleman et al. 2019), it is
questionable that implicit beliefs are existential beliefs at all. What is meas­
ured in studies of implicit theism are (i) the accessibility of associations
between ideas about God and other ideas, and (ii) whether being primed
with ideas about God leads people to behave more prosocially, despite their

8 For a recent defense of the view that it is metaphysically impossible for a person to explic-
itly believe p and explicitly believe not-­p, see (Marcus 2021).
68 Faith, Flourishing, and Agnosticism

explicit nontheism. Yet, believing something exists is not just a matter of

being able to access ideas about it quickly. Someone who studies literary
works about the fictional character Sherlock Holmes will have all kinds of
ideas about Sherlock quickly available to them without this implying that
they believe Sherlock Holmes exists. Likewise, the fact that being primed
with religious ideas leads some people who don’t believe in God to behave
more prosocially doesn’t show that they implicitly believe that God exists
and that the priming activated this belief that contributed to their behavior.
Another salient possibility is that ideas about God and ideas about moral
ideals are closely associated in their thinking despite the absence of explicit
or implicit belief in God. Priming the former primed the latter, and the lat-
ter is what made a difference for their behavior. Thus, for multiple reasons,
I suggest that implicit beliefs, if they exist at all, are not a good candidate for
playing the needed cognitive role in underpinning faith practices.
Consider, then, a third proposal—­ one that McKaughan (2016) has
endorsed. This proposal is that the state of hope can play the needed role.
The defensive captain can pick a play to call because they hope that it will
thwart the quarterback’s plans. Likewise, someone with ambiguous evi-
dence for God can thank God for the goods in their lives based on their
hope that God has provided these goods.
Evaluating this proposal is complicated by the fact that, like faith, hope is a
multifaceted state that incorporates cognitive and conative elements, if not
also evaluative ones (see Bloeser and Stahl 2022: sect.3). In order to hope
that a political candidate will lose the election, it’s not enough for me to have
a cognitive attitude about this. I must also evaluate it positively and in some
sense want it to happen. Properly understood, the proposal that hope
supplies an alternative to assumption and belief that can play the needed role
is that the cognitive component distinctive of hope can play this role.
Now, there are various cognitive states that are plausibly compatible with
hope. Classically, all that is required for hope is endorsing the epistemic
possibility of what is hoped for (Milona 2019). But epistemic possibility is
very broad, ranging from near zero to one. On some views, hope is only
possible within a restricted part of this range—­for example, between 0 and
0.5, noninclusive.9 Once we have understood this about the cognitive atti-
tude of hope, I suggest we can see that it does not provide an alternative to
belief and assumption that can do the needed work.

9 For a view that highlights how hope may have such restrictions, though not quite as stated
in the text, see (Benton 2021).
 Faith and Its Justification 69

It is true that sometimes we describe people using the language of hope

in such a way as to imply that they do have a cognitive attitude that involves
taking a stand. We can do this in the examples of the football captain and in
the God case. Best expressed, what we should say is that the football cap-
tain, or the individual with ambiguous evidence for God, behaves as they do
“in the hope” that what they take a cognitive stand for is correct. My sugges-
tion here is that in these kinds of cases, the kind of hope we are referencing
incorporates either assumption or belief. What we have is hopeful believing
or assuming. The individual does have a cognitive attitude toward what they
hope for that involves taking a stand on it, but the attitude whereby they
take this stand needn’t be understood as anything other than belief or
assumption.
Alternatively, there are cases where we hope without taking a cognitive
stand. In these cases, what the cognitive attitude distinctive of hope requires
is just that we judge what we hope for to be an epistemic possibility—­
perhaps one that is less likely to obtain than not. But taking this kind of
cognitive attitude does not involve taking a cognitive stand for what is
hoped for. Indeed, to echo the earlier argument about implicit beliefs, this
kind of cognitive attitude is straightforwardly, unproblematically compati-
ble with taking the contradictory cognitive stand. That is, judging it to be
epistemically possible but unlikely that p is straightforwardly compatible
with believing not-­p. So, the cognitive attitude distinctive of hope is not a
good candidate for supplying an alternative to belief and assumption that
can fulfill the needed role in practices of faith. Either hope includes a cogni-
tive attitude that involves taking a stand but is belief or assumption, or hope
includes no such attitude.
These remarks also help us to see why another proposal, defended by
Howard-­Snyder (2013a), will not work. Howard-­Snyder suggests that in
addition to belief that p and assumption that p, various beliefs about the
likelihood of p can play the role of the cognitive attitude needed for faith.
For instance, believing p is likely, or believing p is twice as likely as not, or
believing p is the most likely of the plausible candidates, can play the needed
role. I doubt this.
Here’s how I suggest we think about the issue. Beliefs about the likeli-
hood of p are not the sorts of things that involve taking a stand on p; rather,
they are the sorts of things that, in some cases, can help us make up our
minds about whether to take a stand on p or not. Sometimes when we
inquire into a claim p, we just look to the evidence for or against p, and if we
take a stand, we do it directly on that basis. I do this when I look out the
70 Faith, Flourishing, and Agnosticism

window to check whether the dog is in the back garden, for instance. Other
times, especially when it comes to more complicated matters, what we do is
try to assess what the evidence supports. This is a different, metacognitive
activity—­one that is a step too many in the dog case. Here we try to figure
out to what extent our evidence makes p likely, and if we take a cognitive
stand for or against p, we use this information to guide the stand we take.10
The God case, and perhaps the football case, can be like this. Here, we may
attempt to discern to what extent our evidence makes it likely that God
exists or that the quarterback will call a certain play, and depending on what
we find we might subsequently take a cognitive stand.
Now, if this way of thinking about the function of our judgments about
likelihoods is correct, it implies that these judgments cannot play the role
needed to underpin practices of faith. This is because these judgments don’t
involve taking a cognitive stand. They are the sorts of things that, in certain
cases, come explanatorily prior to a person’s taking a stand for p, but they do
not by themselves constitute a cognitive stand for p.
There are two additional ways to try to demonstrate this result. First, the
way we construed the idea of taking a cognitive stand when discussing
belief and assumption was in terms of exposing oneself to error. Someone
who believes p or assumes p errs when p is false because they have a false
belief or assumption. But someone who makes some judgment regarding
the likelihood of p doesn’t thereby expose themselves to error in the case of
p’s falsity. If p is false, nothing follows about whether their estimation of its
likelihood was in error. And the same is true as well of judgments of possi-
bility, such as those characteristic of hope.
Second, we can trot out our good friend—­the contradictory attitudes
argument. Making the likelihood judgments that Howard-­Snyder has in
mind regarding some claim p is compatible with taking a contradictory
cognitive stand on not-­p. Specifically, it is compatible with making a contra-
dictory assumption. A person can judge, for instance, that p is the most
likely of the plausible alternatives but assume not-­p. More generally, cases in
which one estimates that the likelihood of p is such that neither p nor not-­p
is strongly supported are cases where it is plausible one can assume not-­p.
But, if so, then again estimations of likelihood are not the right sort of state
to play the needed role in underpinning practices of faith.
This discussion suggests that another potential candidate for a state other
than belief and assumption also cannot supply the needed role—­namely,

10 For a similar description of these two different ways of inquiring, see (Staffel 2019).
 Faith and Its Justification 71

the state of credence. Views of the nature of credences (or, as they are
­sometimes called, “degrees of belief ”) tend to understand them either as
subjective probabilities or as degrees of confidence (see Jackson 2020). Yet if
we think of credences as subjective probabilities, then it is tempting to
understand them as the same kinds of states just discussed—­namely, as
agents’ beliefs about epistemic likelihoods (see Moon and Jackson 2020).
If this is what credences are, then the proposal that credences supply a
­candidate for a state other than belief and assumption that can underpin
faith practices faces the same objections as the previous proposal.11
On the other hand, the proposal seems to fare no better if we understand
credences as degrees of confidence (see Moon 2019), distinguishing these
from subjective probabilities. The most convincing reason for thinking
there are degrees of confidence in this sense, and one that is widely cited in
the literature, is the observation that we hold some of our beliefs with higher
degrees of confidence than others. Thus, degrees of confidence are not the
same thing as beliefs. Fair enough. But notice that given this way of moti-
vating the existence of degrees of confidence, it seems they will always be
properties of beliefs—­or, perhaps, assumptions. We hold the beliefs and
assumptions we do with some level of confidence, but we don’t have free-­
floating degrees of confidence that aren’t degrees of confidence with which
we hold one of these other attitudes. So, if credences are degrees of confi-
dence, they are not a cognitive attitude independent of belief and assump-
tion that can play the needed role in underpinning faith practices. Anytime
they are present, they will be properties of beliefs or assumptions, and it will
be the latter in virtue of which a person has taken a cognitive stand.
Imagining that p is another state that does not involve taking a stand for
p. At least when we use the language of “imagination” in a way that is dis-
tinctive to it as opposed to being used as a synonym for belief or assumption,
the kind of state we identify does not involve taking a stand on p. On its dis-
tinctive usage, imagining p involves representing p, but not as how things
are. When one imagines that p and it is not the case that p, one has not erred.
Imagining p is straightforwardly compatible with believing not-­p. So, imagi-
nation is not the right sort of cognitive state to underpin practices of faith.
This is not to say that imagination cannot play a role in initiating or sus-
taining states that are the right sort of state for underpinning practices of

11 For a similar argument that credences do not involve “settling” or taking a stand on a
claim in the way that beliefs do but that makes use of a different notion of “settling” or taking a
stand than the one employed here, see (Friedman 2019).
72 Faith, Flourishing, and Agnosticism

faith. On the contrary, I suggest that it can indeed play this kind of role. For
instance, one can make skillful use of multisensory imagination in order to
cultivate or reinforce acceptance of God’s love, or experiences in which one
takes awe-­inducing stimuli to reflect divine attributes of love or intelligence.
When one accepts God’s love or takes awe-­inducing stimuli to reflect divine
attributes, one isn’t merely imagining this, however; one is taking a cogni-
tive stand on God’s love or God’s intelligence being manifested in the world.
Using the imagination can causally support such an attitude, but it cannot
constitute it.
Another possibility is that one might take a cognitive stand on p by act-
ing as if p. There are at least two serious difficulties facing this proposal.
First, it is not clear that there is any such thing as acting as if p. Nobody ever
acts as if p. Rather, talk of “acting as if p” is a shorthand way of referring to
acting as if one takes a stand on p—­say, by believing or assuming p. People
do act as if they believe or assume things, but they do not act as if those
things.12
Second, there seem to be two possibilities as to what is happening when
someone acts as if they take a stand on p. Either they act this way and they
do take a stand on p, or they act this way and they don’t take a stand on p.
The latter happens, for instance, when people lie, mislead, or pretend. But, if
these are the options we have for understanding what it is to “act as if p,”
then it should be clear that this proposal does not get us any distance to
identifying a candidate other than belief or assumption that can play the
role of being the cognitive state that underpins faith practices. If one acts as
if p but doesn’t take a stand on p, then one hasn’t taken a stand on p, and so
this way of acting as if p cannot underpin faith practices. If one acts as if p
and does take a stand on p, this could underpin faith practices, but we have
been given no reason to think that the cognitive attitude in virtue of which
one takes a stand on p is not a belief or assumption.
Another commonly discussed attitude, which is sometimes thought dis-
tinct from belief and assumption and which could underpin faith practices,
is acceptance. In fact, in the literature on acceptance, the state is conceptual-
ized in two quite different ways, each of which is worth considering.
One way of conceptualizing acceptance takes it to be very similar to
assumption. This is an approach to conceptualizing acceptance developed
by William Alston (2007) and followed by Howard-­Snyder (2013a). On this

12 This idea was suggested to me in conversation by Dan Howard-­Snyder.

 Faith and Its Justification 73

approach, we can individuate assumption and acceptance on the basis of

their dispositional profiles.
These profiles are extremely similar, but Howard-­Snyder proposes that
they differ on one particular point. Both assuming and accepting p include
dispositions to assume or accept known consequences of p, to use p as a
premise in practical and theoretical reasoning, and to act in appropriate
ways given one’s goals. Both differ from belief in that neither requires dispo-
sitions to be surprised upon learning that not-­p nor to feel it to be the case
that p when considering p. Howard-­Snyder argues, however, that ac­cept­
ance and assumption differ in that the former requires a tendency to assert
p when asked whether p, while the latter does not (2013a: 363).
However, I would suggest that these are slim and ultimately unconvinc-
ing grounds on which to argue that acceptance is a distinct cognitive state
from assumption.13 In distinguishing assumption from acceptance,
Howard-­Snyder says that someone who assumes p “will lack a tendency to
assert that p when asked whether p, unless it is clear to her that she will not
be misunderstood for expressing a more positive cognitive stance” (2013a:
366). But the problem is that exactly this is also true of acceptance. As
Howard-­Snyder notes, the affirmations of p by someone who accepts p but
doesn’t believe it will be “less confident and more hesitating” (362) than
those of someone who believes p. Presumably, the reasons for this hesitation
will include not wanting to be misunderstood as endorsing a more positive
cognitive stance such as belief. So, I suggest that we do not have a genuine
difference between acceptance and assumption here.
The other way of conceptualizing acceptance is quite different and is
inspired by anti-­realist approaches to scientific theories.14 On the relevant
sort of anti-­realism, one denies that leading scientific theories are true. But
one still has a certain kind of positive cognitive stance toward them. How
should we characterize it? It seems that its primary element involves taking
these theories to be practically useful in the sense that they issue in accurate
predictions regarding observable reality despite their falsehood. To take this
sort of attitude toward minimal theism, then, would be to treat it as false but
issuing in true predictions.
But what predictions does minimal theism make? Well, it “predicts” (or
implies) that each good in some particular human’s life is a benevolent
divine benefit and that God loves this person and any persons they have

13 For another view that treats acceptance as a kind of assumption, see (Buckareff 2005).
14 I am thinking, in particular, of van Fraassen’s (2002) approach.
74 Faith, Flourishing, and Agnosticism

wronged. Is the idea, then, that one can refrain from believing in minimal
theism but believe these predictions it makes and thereby accept it? This
seems a rather unmotivated idea. What would make more sense and be
more analogous to the scientific case would be to accept some more specific
fundamental theory about God that predicts minimal theism, such as per-
fect being theism or a version of pantheistic theism. One might regard that
theory, like one regards scientific theories, as false but as generating true
predictions—­in particular the true prediction of minimal theism.15
More fundamentally, the problem with this proposal is that it does not
provide an alternative to belief or assumption, which can serve as the
­positive cognitive state that underpins practices of faith. When one thanks
God for the goods of one’s life, on this view, one isn’t accepting that God
has provided these benefits—­one is believing God has done so. What one
accepts is something else—­the best candidate being some theory about
God’s fundamental nature. So, this proposal just ends up suggesting that
the cognitive attitudes that underpin practices of faith are beliefs, which is
nothing new.
Finally, consider the proposal that a different type of attitude that can
underpin practices of faith is a conditional attitude. Even an atheist can sin-
cerely say things such as “God, if you’re there, thank you for the benefits of
my life” or “God, if you’re there, I’m sorry for the wrongs I’ve committed.”
Sometimes when people make these kinds of statements, the “if ” signals an
assumption—­they are assuming that God has benefitted them and are
expressing gratitude given this assumption, for instance. However, it is also
possible that the “if ” does not signal an assumption, as would be the case
for a coherent atheist who says the above things. In their case, they believe
God isn’t there, but they express conditional gratitude and conditional apol-
ogies to God.16 What they give expression to are their conditional attitudes.
Perhaps these can underpin practices of faith.
The problem with this proposal for our purposes is that, like so many
others, it does not involve taking a cognitive stand.17 If I give you a

15 Much this approach is suggested by some of van Fraassen’s own comments about
religion—­e.g., (van Fraassen 2002: 1, 29, 255).
16 For a recent discussion of conditional apologies as genuine apologies, see (Baumann
2021). I know of no comparable discussion of conditional gratitude.
17 For an account of conditional attitudes in general that makes this clear, see (Barnett
2006). See (Engel 2021) for a similar take on suppositions. Note that even authors, such as
Baumann (2021), who argue that conditional apologies are genuine apologies agree with
this point. The perspective that conditional apologies are genuine apologies has been a
minority view.
 Faith and Its Justification 75

conditional apology, saying “I’m sorry if I’ve wronged you,” and it turns
out that I did not do anything wrong, then I did not commit any cognitive
error. My conditional attitude is like a contingency plan: I plan for apology
if I wronged, but not otherwise. But this planning isn’t incompatible with it
turning out that I didn’t wrong you. Moreover, my attitude of being
remorseful if I wronged you is perfectly compatible with also not being
remorseful if I didn’t wrong you. And, indeed, it is compatible with my
believing that I didn’t wrong you. Conditional attitudes are, then, compati-
ble with contradictory cognitive stands. So, conditional attitudes are not
adequate to underpin practices of faith.
We have found in this section that it is quite difficult to identify any other
cognitive state besides belief and assumption that can underpin practices of
faith. This is because the cognitive states we have examined are either not
clearly distinct from belief and assumption or they fail to have the feature of
taking a positive cognitive stand that is needed of cognitive attitudes that
can underpin faith practices. Thus, I will assume henceforth that the cogni-
tive attitudes exhibited in faith practices are either beliefs or assumptions.
I want to close this section, however, with a brief word of caution. Just
because the other attitudes I have surveyed cannot underpin practices of
faith, it does not follow that adopting these attitudes cannot be valuable for
individuals with ambiguous evidence for God. I will be arguing that engag-
ing in practices of faith that include adopting either beliefs or assumptions
about minimal theism or its consequences for oneself can help individuals
with ambiguous evidence for God to grow in virtue. But I am not arguing
that adopting these other cognitive attitudes could not also contribute to
this goal. That is a separate subject deserving of attention in its own right.

2 The Epistemic Status of Faith

Now that we have a better understanding of the nature of the cognitive

commitments involved in practices of faith, I wish in this section to con-
sider the epistemic status of these commitments. In particular, I wish to
address a potential concern about their epistemic status that maintains that
such commitments would have to be epistemically unjustified because they
would not be adequately supported by a person’s evidence, given that their
evidence is ambiguous. While the concern itself can be explained straight-
forwardly in this way, the task of addressing it is somewhat complicated
because we must consider both the different ways in which a person’s
76 Faith, Flourishing, and Agnosticism

evidence can be ambiguous and the different kinds of cognitive states that
can underpin practices of faith. My aim here will be to argue that depending
on which cognitive states are in view and in what way a person’s evidence
for God is ambiguous, their cognitively committing to God will either not
be epistemically unjustified, or they will be epistemically excused in holding
this epistemically unjustified attitude, or the epistemic disvalue of their
holding this attitude may be outweighed by nonepistemic values to be
attained through maintaining the attitude. In any case, the concern that by
engaging in practices of faith a person with ambiguous evidence for God
will thereby be adopting epistemically unjustified cognitive commitments is
blunted. While the concern is blunted in all these cases, it is perhaps least
worrisome in the case where their cognitive commitment to God is an
assumption rather than a belief, as I will offer an original argument for
thinking that the epistemic norms governing assumption are weaker than
those governing belief.

2.1 First-­Order Evidential Ambiguity

Consider first a case in which a person’s evidence for God exhibits first-­
order ambiguity in the sense that it neither strongly supports God’s ex­ist­
ence nor strongly supports God’s nonexistence. Suppose that in these
circumstances they engage in practices of faith, adopting the requisite
cognitive commitments—­
­ either beliefs or assumptions regarding God’s
ex­ist­ence, love for them and others, and so on. How should we assess the
epistemic status of these commitments?
Let’s start with the case of belief. Supposing that their commitments are
beliefs, the worry is that such beliefs would be epistemically unjustified
because any belief that p is epistemically justified for a person S only if S’s
total evidence strongly supports p, and otherwise it is unjustified. But the
individuals we are considering are such that their evidence does not strongly
support God’s existence and love for them, and so on; so, in believing these
claims, they are adopting epistemically unjustified beliefs.
There are many ways to reply to this concern. Some of them depend on
the exact nature of the ambiguity of the evidence.
Suppose that the person’s evidence supports God’s existence, just not
strongly. In that case, some epistemological views—­such as the version
of evidentialism recently defended by Richard Feldman and Earl Conee
(2018)—will imply that the person’s beliefs are epistemically justified, not
 Faith and Its Justification 77

unjustified. Defenders of these views deny the premise of the argument

just sketched—­that epistemic justification requires strong support, not just
­support. On Feldman and Conee’s view, any proposition supported even
slightly by one’s evidence is epistemically justified. Such a view may seem
more plausible if we allow that the beliefs in question are held with a low
degree of confidence.
But suppose such views are incorrect. There are still at least three addi-
tional ways one might attempt to defend the claim that the person’s beliefs
needn’t be epistemically unjustified, drawing on recent work in mainstream
epistemology.18
First, consider permissivist views.19 According to permissivism, it is
false that for any body of evidence and any proposition p, there is just one
cognitive attitude toward p that is epistemically justified and all others
­ep­i­ste­mically unjustified. In other words, permissivism denies the rational
uniqueness thesis that evidence makes one unique attitude rational. Instead,
permissivists maintain that some bodies of evidence permit multiple dis-
tinct attitudes toward a proposition p to be epistemically justified.
Now, there are different versions of permissivism. A chief difference
between the versions has to do with the range of attitudes they will allow to
be epistemically justified by the same body of evidence. Some permissivists
have suggested that permissivism is more plausible when it comes to more
fine-­grained attitudes, such as credences, than when it comes to more
course-­grained attitudes, like belief, suspension of judgment, and disbelief
(cf. Kelly 2014). Perhaps, for example, given the same body of evidence for
some claim p, both believing p with one level of confidence and believing it
with a slightly higher level of confidence can be epistemically justified. It
may be more questionable whether, given the same body of evidence for p,
both believing p and disbelieving p are epistemically justified.
Permissivism is often defended by appealing to the Jamesian idea that
when we inquire, we must somehow balance two competing goals—­
obtaining true beliefs and avoiding false ones (James 2010). The goals are
not equivalent, and some people may weigh one of the goals more heavily
than the other, while others assign the opposite weighting. Some people, in
other words, may approach belief-­forming in a more risk-­averse way, and
others may approach belief-­forming in a more risk-­tolerant way. If we
­cannot provide convincing epistemic grounds for preferring one of these

18 The following discussion borrows with permission from (Byerly 2022a).

19 For an overview of permissivism in epistemology, see (Kopec and Titelbaum 2016).
78 Faith, Flourishing, and Agnosticism

weightings to the other, then it seems we must follow the permissivist and
allow that individuals who have the same evidence but weigh it differently
may be equally justified in adopting different attitudes in response to this
evidence. Given their different goals, what is rational for each is different
(cf. Riggs 2008).
A different way of making the same point is to appeal to the subjectivity
of what qualifies as “adequate” support for a claim (cf. Schoenfield 2014).
The permissivist may agree that for a person to be justified in believing a
claim, their evidence must adequately support the claim. But they may
argue that whether a person’s evidence adequately supports a claim is deter-
mined in part by their own tendencies in terms of risk-­aversion or risk-­
tolerance. What do they themselves require for evidence to support a claim
enough to believe it for purposes of accomplishing their weighted ep­i­ste­
mic goals?
Permissivist views of this sort offer some aid for responding to the pres-
ent worry, depending on how permissive they are. Suppose, for instance,
that a person’s evidence very nearly strongly supports minimal theism. Even
the less permissive permissivist views may allow that in this case, believing
minimal theism may be epistemically justified for someone who weighs
believing the truth about minimal theism more heavily than avoiding error
about it. It will take a more permissive permissivist view to allow that belief
is epistemically justified for any degree of support that is positive but less
than strong, though this sort of view may seem more plausible if it is granted
that the beliefs in such cases will be held with low confidence. Moreover,
permissivism may interact in an interesting way with a view along the lines
of that of Feldman and Conee described above, according to which weak
evidential support is adequate for justified belief. If that view is true and
permissivism is true, it may be that individuals with evidence that slightly
supports p or slightly supports not-­p may be equally justified in believing or
disbelieving p with low confidence, depending on their risk tolerances.
Generally, however, permissivist views that would allow for belief in God to
be justified despite evidence that supports God’s nonexistence are regarded
as more extreme and less plausible than the less permissivist views with
which we began this paragraph.
A second approach to defending the epistemic justification of belief
despite ambiguous evidence appeals to pragmatic—­and more specifically,
moral—­encroachment (see Kim and McGrath 2018 for an overview). On
this view, moral reasons for or against adopting a cognitive attitude can
affect the level of evidential support needed in order for that attitude to be
 Faith and Its Justification 79

epistemically justified. Not just any version of moral encroachment will

allow us to maintain that the cognitive commitments required for faith are
epistemically justified given ambiguous evidence. What is needed here is a
view according to which moral reasons for holding an attitude can lower
the level of support needed for that attitude to be epistemically justified.
Michael Pace (2011) has defended a version of moral encroachment of
this sort. On Pace’s view, the fact that greater moral value is attainable via
belief than via its absence can lower the evidential standards necessary for
belief to be epistemically justified. As he puts it, “When there are significant
positive benefits to be gained by having a true belief and relatively little
practical cost of error, the evidential standards sufficient for justification dip
below what they would be in contexts in which nothing much is at stake”
(257). Pace stipulates, however, that in order for belief that p to be justified,
one’s evidence must make p at least more likely than not. It is only in cases
where one’s evidence for p already makes p more likely than not that moral
considerations can kick in to make a difference for how strongly this evi-
dence must support p in order for believing p to be epistemically justified.
Pace claims that there are principled reasons for imposing this “more
likely than not” requirement. The first is the worry that if one’s evidence
doesn’t make p more likely than not and one realizes this, then it is not psy-
chologically possible for one to believe p. But we can put this concern to the
side here, since we are granting for the sake of argument that the cognitive
commitments of faith are beliefs and are investigating their epistemic status
given this supposition. In any case, as we have noted, there is controversy
about whether believing despite unsupportive evidence is possible. Pace’s
second worry, which is worth addressing here, is that even if not psycholog-
ically impossible, believing on the basis of such reasoning “may require
deceiving oneself about the quality of one’s evidence” (2011: 252). I suggest
that on Pace’s own approach this concern is unmotivated.
On Pace’s approach, the best way to understand how moral encroach-
ment works is that moral factors affect one’s “evidential standards” (254) for
the extent to which evidence must support a claim p in order for belief that
p to be epistemically justified. Pragmatic factors matter, in other words, for
whether the extent to which a claim is supported by evidence is adequate for
belief. Yet there doesn’t seem to be anything self-­deceptive involved in a
person recognizing that their evidence only supports a claim p just as much
as its denial, or even less than its denial, and yet reflectively taking this
degree of support as adequate for believing p given the moral gains to be
had if the belief is true and the relative costlessness if the belief is false.
80 Faith, Flourishing, and Agnosticism

A moral encroachment view like Pace’s but without its “more likely than
not” requirement therefore seems about as well-­motivated as Pace’s own
stated view.
If such a moral encroachment view is correct, then it will support the
judgment that the beliefs involved in faith practices may be epistemically
justified rather than epistemically unjustified. For on the one hand, the
value of acting on these beliefs if they are true is significant—­it will, for
example, result in one successfully expressing gratitude to God and accept-
ing God’s love. On the other hand, this book can be read largely as a defense
of the contention that the disvalue of having these beliefs if they are false is
not high. I am arguing that even if the commitments of faith are false, there
is significant value to be gained through having them in terms of one’s
growth in virtue. If this is correct, then the standards for the degree of evi-
dential support necessary for these commitments to be adequately sup-
ported may be lowered—­even lowered to the point that the evidence needn’t
make God’s existence more likely than not. A person might self-­reflectively
assess their evidence regarding God’s existence as being counterbalanced or
as weakly supporting God’s nonexistence and yet take this level of evidential
support to be adequate for cognitively committing to minimal theism. On
the present moral encroachment view, these commitments would be ep­i­ste­
mically justified.
This discussion may give the impression that permissivist and pragmatic
encroachment views are quite similar. They can be, but it is worth pointing
out that they do offer somewhat unique resources in defense of the views
needed here. Permissivist views can but needn’t be tied to moral encroach-
ment. A permissivist view can be motivated simply on the basis of the
thought that it is not always (or ever) possible to argue that one way of
weighing the competing goals of gaining truth and avoiding error is ep­i­ste­
mically superior to another. Moral encroachment views, for their part,
needn’t be developed in a way that supports permissivism. One can be a
moral encroacher while defending the uniqueness thesis, claiming that
whatever is adequately supported by one’s evidence—­where adequacy is
partly determined by moral considerations—­is the unique epistemically
justified attitude to have.
Turn finally to a third and more markedly different approach recently
developed by Susanna Rinard (2017, 2019b). Rinard defends a view she
calls “Equal Treatment” for belief. On this view, there is no special guidance-­
giving normativity that applies only to beliefs; questions about what one
should believe or is permitted to believe are to be answered in the same way
 Faith and Its Justification 81

as questions about what one should do or is permitted to do more generally.

As a consequence, if having a belief is all-­things-­considered justified, it can-
not be unjustified according to some standard of justification that applies
only to beliefs. The implication for faith commitments is straightforward:
if believing minimal theism when in possession of ambiguous evidence
for it is all-­ things-­considered justified, then it cannot be epistemically
­unjustified—­it cannot be unjustified according to some standard of justifi-
cation that applies only to beliefs.
It is instructive to note the kind of case that plays a central role in Rinard’s
defense of Equal Treatment. She asks us to imagine a case in which taking
a pill will cause you to have a certain belief you would not otherwise
have without taking the pill, and you know this. Moreover, as the case is
described, it is supposed to be one in which you all-­things-­considered
should take the pill, but your all-­things-­considered reasons favor this only
slightly. Denying Equal Treatment and maintaining that there is a special
guidance-­giving normativity that applies only to beliefs allows it to be the
case that you should not hold the belief that is caused by taking the pill,
despite it being the case that you should take the pill. For there could be
distinctively epistemic reasons against holding the belief that do not apply
to taking the pill and are such that they are sufficient to shift the balance of
all-­things-­considered reasons pertaining to holding the belief but not to
taking the pill. Rinard finds this consequence implausible because it violates
the principle of agglomeration, that if you should x and you should y then
you should (x and y). Denying Equal Treatment violates this principle
because it allows that you should take the pill and should not hold the belief,
whereas surely it is not the case that you should (take the pill and not hold
the belief), as the latter is not an option for you.
The case of faith commitments significantly parallels this example of the
belief-­inducing pill. Let us suppose that the moral reasons in favor of engag-
ing in faith practices for those with ambiguous evidence for God render
engaging in these practices all-­things-­considered justified. If we suppose
that there is a guidance-­governing normativity distinctive to belief, then it
could be that while engaging in practices of faith is justified, believing the
claims of minimal theism is not all-­things-­considered justified. This possi-
bility runs afoul of agglomeration, if we suppose—­as we are doing here—­
that it is not possible to engage in the faith practices without belief in
minimal theism. Thus, if we, with Rinard, are unwilling to give up agglom-
eration, we have a way of resisting the conclusion that the beliefs necessary
for faith commitments are epistemically unjustified for individuals with
82 Faith, Flourishing, and Agnosticism

ambiguous evidence for God. Notably, this strategy differs from the two
previous strategies in that it does not offer support for the conclusion that
these beliefs will have the positive property of being epistemically justified;
instead, on this strategy there is no such thing as distinctively epistemic
­justification for these beliefs to have or lack.
So far in this section, we have explored epistemological views that offer
support for the view that if faith practices require beliefs, these beliefs
needn’t be epistemically unjustified despite the evidence for them being
ambiguous. I now want to consider what can be said about the value of the
beliefs if all of these views fail and the beliefs are epistemically unjustified.
Supposing that the beliefs are epistemically unjustified, could there still be
an epistemic excuse for having the beliefs? Or, whether or not they are ep­i­
ste­mically excused, could the epistemic disvalue associated with their being
unjustified be outweighed by the moral value of adopting them? I will point
to resources that can be used to support positive answers to both questions.
First, regarding epistemic excuses, I suggest that we might regard a per-
son as epistemically excused in adopting an unjustified belief if their evi-
dence strongly supported that the belief was not unjustified.20 What I have
in mind is a scenario in which a person’s evidence does not in fact ­adequately
support a claim p for their believing it to be justified, yet their evidence also
does adequately support the claim that believing p is not epistemically
unjustified. We might think that the support available for the epistemologi-
cal views discussed above could provide just this sort of excuse for believing
minimal theism for those whose evidence for it is ambiguous. We are sup-
posing that these views are in fact false and that individuals with ambiguous
evidence for minimal theism are unjustified in believing minimal theism.
Yet such individuals can have misleading but strongly supportive evidence
for these epistemological views. If they do, then they may have an epistemic
excuse—­ an excuse derived from epistemic considerations alone—­ for
adopting the beliefs necessary for faith. Despite these beliefs being ep­i­ste­
mically unjustified, such individuals will have an epistemic excuse for
adopting them and will be epistemically blameless.
My view is that this tactic may be of some use in showing how some indi-
viduals with ambiguous evidence for God may be excused in adopting
unjustified beliefs in God in order to pursue practices of faith. But its use
is probably quite limited. For I doubt that most individuals who have

20 For a similar discussion of epistemic excuses, see (Littlejohn forthcoming).

 Faith and Its Justification 83

ambiguous evidence for God also have access to evidence that strongly
­supports the epistemological views discussed above. You, reader, may have
it, because you’ve just been reading about those views. But that’s the trouble—­it
seems that access to this evidence is limited to individuals like you who take
in readings in general epistemology.
Let’s suppose, then, that the beliefs of faith practices are usually not ep­i­
ste­mically excused and that they are epistemically unjustified. Could the
badness derived from their being epistemically unjustified be outweighed
by the goodness to be obtained through them in terms of their facilitation
of virtue development and (if true) relationship with God? Supposing that
this moral goodness to be obtained through engaging in faith practices is
not itself counterbalanced by some other moral badness that comes from
them, I suggest that the answer is yes. The total expected value of adopting
practices of faith may be positive, and significantly so, despite it being the
case that the beliefs required to engage in these practices are epistemically
unjustified and unexcused. The moral values to be gained through these
commitments, both in the case where they are true and in the case where
they are false, are just too weighty.
Similar arguments have been defended by others in the literature on faith
and on epistemic partiality in friendship. Preston-­Roedder appeals to this
sort of strategy in his defense of the all-­things-­considered value of behaving
in accordance with faith in humanity. He is willing to grant the possibility of
some conflict between such faith and epistemic rationality, writing that in
some cases “someone who has faith can, without any failure of [this] virtue,
form beliefs about people. . . that are to some degree irrational, given the evi-
dence” (2013: 685–86). Yet, he continues, “unless we assume that the moral
importance of epistemic rationality is implausibly great, or the importance
of [the moral aims of faith] implausibly slight, we should conclude that a
virtuous person may sacrifice some degree of epistemic rationality, in cer-
tain respects and in certain cases, in her pursuit of these other aims” (687).
Preston-­Roedder concludes that having faith is constitutive of a “practical
ideal, concerned with the sort of life one should live” (686). Epistemically
disvaluable features of faith can only prevent faith from contributing to this
all-­things-­considered ideal, according to Preston-­Roedder, if they are of
comparable moral importance to the moral values toward which such faith
is conducive. Similarly, in the literature on epistemic partiality in friendship,
it has been maintained that if there is conflict between the norms of friend-
ship and the norms of rationality, it is the norms of rationality that must
give way, as the epistemic values in play are not of comparable moral
84 Faith, Flourishing, and Agnosticism

importance to friendship (cf. Stroud 2006). Here, too, we may maintain that
if the arguments of the second part of this book are correct and so the
beliefs necessary for faith practices really are conducive toward significant
goods of virtue development or (in the case where they are true) relation-
ship with God, their epistemic disvalue may be outweighed by these
moral goods.
We have, then, a defense of the claim that for individuals with ambiguous
first-­order evidence for God, adopting beliefs in minimal theism necessary
for engaging in faith practices is either not epistemically unjustified, is ep­i­
ste­mically excused, or is such that its epistemic disvalue is outweighed by its
moral value. But what of the other possibility—­that beliefs are not neces-
sary, and that faith practices can be engaged in on the basis of assumptions
rather than beliefs?
It seems generally to be the case that epistemologists who grant that there
are subdoxastic states, such as assumptions, that are distinct from belief in
terms of their voluntariness and sensitivity to evidence tend to think that
these states are either subject to less stringent epistemic norms than belief
or are not subject to such norms at all. Palmqvist (2022: 503) characterizes
the tendency this way: “The assumption seems to be that since ep­i­ste­mic
rationality is primarily about belief-­regulation; it is not a very relevant
notion for non-­doxasticism.” Indeed, several authors who have explicitly
addressed the topic have maintained the view that nondoxastic attitudes
such as assumption do not require as strong a support as belief in order to
be epistemically justified. Liz Jackson, for example, claims that while belief
that p is epistemically justified only when one has “a good amount of
­evidence for p,” the sub-­doxastic cognitive attitudes of faith require only
“a moderate amount of evidence” to be epistemically justified (2022: 213).
Howard-­Snyder, writing about epistemic reasonableness, claims that “even
if it is not reasonable for you to believe or accept [p], it might yet be reason-
able for you to belieflessly assume it” (2017: 163).
The unfortunate thing is that neither of these authors offers any argu-
ment for this claim about the epistemic norms governing assumption. Dan
McKaughan (2016) makes a similar claim about the epistemic status of the
nondoxastic cognitive commitments required of faith and gives a very brief
argument in its defense. After maintaining that acting on faith that p
requires a cognitive commitment to p that involves taking a cognitive stand
on p (though not belief that p), he claims that “Epistemic rationality . . . is
solely a matter of finding yourself with a credence level or subjective proba-
bility judgment that fits the evidence.” As a result, someone who acts on
 Faith and Its Justification 85

faith that p when having evidence that does not strongly support p is
“entirely free” to abide by the norms of epistemic rationality, and “there is
nothing inherent in action-­centred faith that leads her into epistemic irra-
tionality” (84–5).
While we at least have an argument for the view here, it is an unsatisfying
one, for the claim that epistemic rationality is solely concerned with cre-
dences or probability judgments implies that even garden-­variety beliefs are
not subject to standards of epistemic evaluation. And if we modified the
claim so that it implied that beliefs, too, but not assumptions are subject to
epistemic evaluation, then the argument will just beg the question in favor
of the desired view.
While Howard-­ Snyder doesn’t defend his view about the epistemic
norms (not) governing assumption, he does criticize the view that the ep­i­
ste­mic norms governing acceptance are weaker than those governing belief.
On the account of acceptance he is discussing, acceptances differ from
beliefs in that the former can be acquired at will and the latter cannot, and
the latter involve a disposition to feel that the claim in question is true when
reflecting on it, while the former do not. But, Howard-­Snyder claims, ­“neither
difference seems relevant to any epistemic status related to evidence, reasons
or grounds” (2017: 155). For each difference, Howard-­Snyder asks whether
it is relevant, says “[i]t seems wholly irrelevant,” and concludes that it isn’t
relevant.
I contend that these differences—­which are also differences between
belief and assumption on Howard-­Snyder’s own view and mine—­are rele-
vant, and in so doing I will provide a rare non-­question-­begging argument
in defense of the commonly endorsed view of assumption’s less demanding
epistemic norms. The argument builds on a recent approach to thinking
about the nature of belief defended by Eric Marcus (2021). On Marcus’s
view, beliefs are a kind of state that is by its very nature rational in a certain
way. One way this is illustrated is that when a person believes a claim p and
has this belief clearly in mind, it is metaphysically impossible for them to
believe not-­p while retaining the former belief. Marcus calls this necessity a
“rational necessity” in that this metaphysically necessary fact follows from
the nature of beliefs being rational states of a certain kind.
The phenomenon of not being able to hold contradictory beliefs when
these are held clearly in mind is a consequence of a broader feature of belief.
Every time we believe something, Marcus claims, we also take what we
believe as “to-­be-­believed” or, as he puts it in some places, as what “should
be represented as true” (2). It is because belief is like this that we cannot
86 Faith, Flourishing, and Agnosticism

believe p and not-­p when holding these clearly before the mind, because we
recognize that (p and not-­p) is not to-­be-­believed.
Now, I suggest that this way of thinking about belief is consonant with
and perhaps a more perspicuous way of characterizing the two differences
between belief and assumption noted a few paragraphs ago. Representing
the content of one’s belief as to-­be-­believed is a different way of characteriz-
ing the feeling that the belief is correct which accompanies belief. And the
metaphysical necessity of believing p when p is represented as to-­ be-­
believed and not doing so when it is not captures how belief is involuntary.
On both scores, assumption differs. When the football captain assumes the
opposing quarterback will call a certain play, they don’t represent the quar-
terback’s calling that play as to-­be-­believed or as something that should be
represented as true. I would suggest that they needn’t even represent the
quarterback’s calling that play as to-­be-­assumed. There is more optionality
than that involved in assumption. This also helps us see how assumption is
more voluntary than belief. Although it is metaphysically impossible to
believe p while not representing p as to-­be-­believed, no such metaphysical
necessity attaches to assumption.
So, what does this have to do with the norms governing belief and
assumption? That is a hard question to answer, but I will offer a proposal.
My suggestion is that the epistemic norms governing belief and assumption
derive from their distinct natures. The fact that we cannot believe p without
representing p as to-­be-­believed explains why it is that one is epistemically
justified in believing p only if p is to-­be-­believed, or what one should
believe. There are different theories of what it is for a claim to be to-­be-­
believed or what one should believe, one of which is that the claim must be
strongly supported by one’s evidence. If that theory is correct, then in com-
bination with what we have learned of belief ’s normativity from its nature, it
follows that one is epistemically justified in believing p only if p is strongly
supported by one’s evidence. No such argument will apply to assumption,
however, precisely because its nature differs from that of belief. We will not
be able to take the first step of arguing that assuming p is epistemically justi-
fied only if p is to-­be-­believed or even to-­be-­assumed.
The most difficult question facing this line of argument is why it should
be that the fact that believing p requires representing p as to-­be-­believed
should imply that believing p is epistemically justified only if p is to-­be-­
believed. I suggest a kind of ought-­implies-­can-­like defense. If we are won-
dering whether the reasons someone has justify their taking a cognitive
attitude, we can consider whether they could take the attitude if they had a
 Faith and Its Justification 87

full understanding of those reasons. If they couldn’t, then the reasons don’t
justify their taking the attitude. But in that case, only reasons that make p
to-­be-­believed will be able to justify believing p, for a full understanding of
the reasons will include an understanding of whether they make p to-­be-­
believed or not. If they fail to make p to-­be-­believed and one understands
this, then one cannot believe p as a matter of rational necessity; but if they
do make p to-­be-­believed and one understands this, then one can believe
p and, on Marcus’s view, will do so as a matter of rational necessity. Of
course, the same line of argument cannot be employed to defend the con-
clusion that assuming p is epistemically justified only if p is to-­be-­believed
or even to-­be-­assumed. And this is precisely because of the differences
between belief and assumption—­that belief and not assumption includes a
representation of what is believed as to-­be-­believed and is metaphysically
impossible without such.
So, contrary to what Howard-­Snyder says and consonant with the seem-
ing consensus view of scholars working on the topic, the differences between
the nature of beliefs and assumptions do make a difference for the epistemic
norms governing these. Assumptions are not subject to as stringent ep­i­ste­
mic norms as belief. While I won’t try to fully spell out what epistemic
norms do govern assumptions,21 our argument suggests that the only sort of
evidence that would render an assumption epistemically unjustified is evi-
dence that, if fully understood, would make assumption impossible. The
cases used to motivate the existence of assumptions as distinct states from
belief also then motivate the claim that assumptions can be epistemically
justified in the cases that matter to us—­cases of evidential ambiguity. It
needn’t be epistemically unjustified for a person with first-­order ambiguity
in their evidence for God to assume the claims of minimal theism.
Not everyone will be convinced by this defense of the common view of
assumption’s epistemic norms, nor of the common view, for that matter.
I am, after all, more or less starting from scratch here, and I indeed would
strongly advocate that additional attention be given to the issue. So, let us
briefly consider what can be said about the value of the assumption compo-
nents of faith practices if my defense of the common view does not succeed.
If assumptions are subject to the same epistemic norms as beliefs, then it
seems that much the same things can be said about the assumption compo-
nents of faith practices as were said about the belief components in detail

21 I come back to the topic in the next section, suggesting a norm for when assumptions are
not epistemically justified.
88 Faith, Flourishing, and Agnosticism

above. We could argue that the epistemological views that allow beliefs to be
epistemically justified (or at least not epistemically unjustified) despite an
absence of strong evidence can do the same for assumptions. Or we could
argue that, despite being epistemically unjustified when strong evidence is
not present, assumptions can be epistemically excused, or their epistemic
disvalue can be outweighed by their moral value. Thus, as with beliefs, if the
cognitive components of faith practices are assumptions, there seems to be
a viable pathway of arguing that these are either not epistemically unjusti-
fied, are epistemically excused, or their epistemic disvalue is outweighed for
individuals with ambiguous evidence for God. Indeed, the case that they are
not epistemically unjustified is somewhat stronger than the comparable
case for beliefs, depending on the success of my defense of the common
view of the epistemic norms of assumption.

2.2 Higher-­Order Evidential Ambiguity

We have been considering at length the question of the epistemic status of

the cognitive commitments required for faith practices in circumstances
where a person’s evidence is ambiguous in the sense that it neither strongly
supports minimal theism nor strongly supports its negation. Yet, as high-
lighted in Chapter 2, this may not be the only way that a person’s evidence
for God can exhibit a kind of ambiguity. Instead, it may exhibit ambiguity if,
despite strongly supporting God’s existence or nonexistence, it also strongly
supports that it does not strongly support either. What can be said about the
epistemic status of a person’s beliefs or assumptions about God in that
circumstance?
Let’s begin by considering beliefs in minimal theism under these circum-
stances. We are granting here that whether these beliefs are epistemically
justified is determined by whether their evidence strongly supports mini-
mal theism. If it does, then their belief is epistemically justified; if it does
not—­which is the more interesting case—­then their belief is ep­i­ste­mically
unjustified.
What difference might their higher-­order evidential ambiguity make?
We saw in the previous section that higher-­order evidence may function to
provide epistemic excuses. For example, a person whose evidence strongly
supports p but whose evidence also strongly supports that it doesn’t strongly
support p has an epistemic excuse for not believing p. Perhaps something
similar applies in this case. Perhaps, although a person’s evidence strongly
 Faith and Its Justification 89

supports that minimal theism is false, the fact that their evidence strongly
supports that it neither strongly supports minimal theism nor strongly sup-
ports its denial furnishes them with an epistemic excuse.
Which attitude would be excused by the excuse? Perhaps most straight-
forwardly, they are excused for not believing that minimal theism is false.
Although their evidence justifies this belief, they are excused in not adopting
it because of their higher-­order evidence. But, more tentatively, we may be
able to go even further.
Think about it from the person’s own perspective—­as we often do when
thinking about what is excusable. From their own perspective, it may seem
that their evidence neither strongly supports minimal theism nor its negation.
Indeed, their evidence strongly supports this claim about itself. They are, in
fact, justified in believing that their evidence neither strongly supports
­minimal theism nor its negation. But if they are justified in believing this,
then they may be justified in believing that they are in the sort of position
where believing minimal theism is either not epistemically unjustified or is
epistemically excused. Their being justified in believing these things in turn
supplies them with an epistemic excuse for believing minimal theism.
Let me unpack that a bit. Given that a person’s evidence strongly supports
that their evidence does not strongly support minimal theism or its nega-
tion, that person is epistemically justified in believing that their evidence
neither strongly supports minimal theism nor its negation. However, in the
previous section, we reviewed several epistemological views that provide
ways of arguing that if a person’s evidence strongly supports neither mini-
mal theism nor its negation, then their believing minimal theism can either
avoid being epistemically unjustified or it can be epistemically excused.
Someone aware of the evidence for these theories, then, who justifiably
believes that their evidence for minimal theism strongly supports neither it
nor its negation, may justifiably infer that their believing minimal theism
either avoids being epistemically unjustified or is epistemically excused. Yet
these justified beliefs themselves provide epistemic excuses for their believ-
ing minimal theism. Because the person in question justifiably believes that
their believing minimal theism would not be epistemically unjustified (or
may even be epistemically justified) or is epistemically excused, they have
epistemic excuses for believing minimal theism.
Higher-­order evidential ambiguity regarding God’s existence can there-
fore provide a person with an epistemic excuse for believing minimal
­theism, despite believing minimal theism being epistemically unjustified
for them because their evidence strongly supports its negation. Notably,
90 Faith, Flourishing, and Agnosticism

however, it seems that it can do this only for someone who has access to the
evidence concerning the epistemological theories discussed in the previous
section. Or, at least, it can do this only for someone who is justified in
believing that their believing in minimal theism when their evidence does
not strongly support it is not epistemically unjustified or is excused. So, this
approach to showing that a person whose evidence for minimal theism has
only higher-­order ambiguity has some limited purchase in blunting con-
cerns about the epistemic disvalue of believing minimal theism on ambigu-
ous evidence.
What about the question of whether any epistemic disvalue involved in
believing in minimal theism can be outweighed when one’s evidence exhib-
its only higher-­order ambiguity? Here it seems that the situation is nearly
the same as it is in the case where one’s evidence exhibits first-­order ambi-
guity. In that case, what we considered is whether the moral value to be
gained through adopting the relevant beliefs could outweigh the epistemic
disvalue of those beliefs, where we assume that they are both epistemically
unjustified and epistemically unexcused. Here, it seems, we must consider
exactly the same situation—­with one difference.
Consider someone whose evidence strongly supports that minimal the-
ism is false but whose higher-­order evidence strongly supports that it nei-
ther strongly supports minimal theism nor its negation. If they believe
minimal theism and their belief is both epistemically unjustified and ep­i­ste­
mically unexcused, it is also plausible that they have missed out on having a
belief that was epistemically justified—­namely, believing minimal theism is
false. This is the difference between the present case and the case considered
in the previous section. There is somewhat greater epistemic disvalue
involved in believing minimal theism in this case, because in doing so one
foregoes an epistemic good that is not attainable in the other case.
Notably, however, as we saw above, there is a relatively uncontroversial
epistemic excuse for foregoing this epistemic good in this case. The individ-
ual’s evidence strongly supports that it doesn’t strongly support minimal
theism’s falsity; so, their not believing minimal theism to be false is ep­i­ste­
mically excused. This suggests that while there is a difference in the disvalue
associated with believing minimal theism in cases where one’s evidence
exhibits only higher-­order ambiguity versus first-­order ambiguity, one has
an epistemic excuse for not responding to this difference. If one’s evidence
for minimal theism exhibits only higher-­order ambiguity, one has an ep­i­ste­
mic excuse for acting exactly as if one’s evidence exhibits first-­order ambi-
guity for minimal theism.
 Faith and Its Justification 91

What about the case where one has only higher-­order ambiguity in one’s
evidence for God and one assumes rather than believes in minimal theism?
The answer will depend in part on what the epistemic norms for assump-
tion are. In particular, it will depend on whether assuming p when one’s
evidence strongly supports not-­p is epistemically unjustified. If my argu-
ment about the epistemic norms governing assumption in the previous
­section are sound, then this will in turn depend on whether it is possible for
someone who fully understands that one’s evidence strongly supports not-­p
to assume p. I suggest that this is not possible, given what was said in the
previous section about belief. For, recall that we said there that it is impossi-
ble for someone who fully understands that their evidence strongly sup-
ports not-­p not to believe not-­p. This is because, in fully understanding that
their evidence supports not-­p, they represent not-­p as to-­be-­believed, which
metaphysically necessitates that they believe it. Believing not-­p, however, is
not compatible with assuming p, since both involve taking a cognitive stand
on p. So, I suggest that a person whose evidence strongly supports not-­p
would be epistemically unjustified in assuming p.22
In this respect, assuming minimal theism when one has only ambiguous
higher-­order evidence is much like believing minimal theism when one has
only ambiguous higher-­order evidence. Both states are epistemically unjus-
tified. Both also can be epistemically excused by the higher-­order evidence.
In the case of assumption, the excuse is easier to come by, since it is reason-
able to believe that the epistemic standards governing assumption are
weaker than those governing belief. As such, someone whose evidence
strongly supports that their evidence neither strongly supports minimal
theism nor its negation is more likely to have an epistemic excuse for
assuming minimal theism than for believing it.
Finally, suppose that assuming minimal theism when having only higher-­
order ambiguity in one’s evidence for minimal theism is both epistemically
unjustified and unexcused. The story about how moral considerations may
outweigh this epistemic disvalue of assuming minimal theism will be just
like the story given above regarding believing minimal theism in these
circumstances.
In this section, I have explained how individuals with only higher-­order
ambiguity in their evidence for God can be epistemically excused in

22 This makes my view different from more lenient views that have been defended recently,
such as Mark Wynn’s (2020: 189–91) view that it is possible to adopt a faith commitment to a
worldview that one regards as very unlikely (cf. McKaughan 2016).
92 Faith, Flourishing, and Agnosticism

believing or assuming minimal theism because of this ambiguity, and how

the epistemic disvalue of these beliefs or assumptions can be outweighed by
their moral value. This helps to complete the main argument of the chapter
that the cognitive commitments of faith practices can be supplied either by
beliefs or assumptions and that in either case these commitments are not
epistemically unjustified, or they are epistemically excused, or their ep­i­ste­
mic disvalue is outweighed by their moral value—­if indeed adopting them
helps one to grow in virtue, as this book maintains. This leads us, then, to
the question whether adopting these commitments as part of one’s faith
practices really does contribute to virtue and flourishing. Our next step in
addressing that question is to consider the nature of virtue and flourishing.
 4
Virtue and Flourishing

This book argues that individuals with ambiguous evidence for God can
grow in virtue and experience greater flourishing by engaging in practices
of faith. The previous chapters have expanded upon the idea of God used in
the book, addressed why a sizable group of individuals likely have ambigu-
ous evidence for God, explained what the cognitive commitments of faith
practices involve, and addressed a concern about the epistemic disvalue of
those commitments. This chapter turns to the subject of virtues and flour-
ishing. Specifically, I will discuss the nature of character virtues in general
and their relationship to flourishing, with the aim of defending views on
these topics necessary for the broader argument of this book to succeed. In
Section 1, I develop a basic account of what makes a character trait a virtue,
showing how this account can be embraced by advocates of many different
theories of virtue and arguing that on this account the character traits that
are my focus will count as virtues. In Section 2, I develop a basic account of
flourishing that many theorists of flourishing can accept, arguing that
growth in virtue is likely to promote growth in flourishing and that this fact
is and should be motivating regardless of whether growth in flourishing
also constitutes or promotes growth in well-­being. The chapter thus defends
the claims that the traits that I will argue are promoted by faith practices are
virtues, that growth in these virtues also promotes greater flourishing, and
that these facts should provide some motivation for individuals with ambig-
uous evidence for God to engage in the faith practices.

1 Character Virtues

My focus is on how engaging in practices of faith can help a person grow

toward or in their possession of virtuous character traits. Such character
virtues may not be the only kinds of virtues humans can have. For example,
they may also be able to have faculty virtues that involve such things as
good eyesight or sharp memory (cf. Battaly 2015: ch.1). I won’t be arguing

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0005
94 Faith, Flourishing, and Agnosticism

that practices of faith support those virtues or others if there are yet other
virtues humans can have. My focus is just on virtuous character traits.
Chapters 4–6 will identify four pathways whereby engaging in three dif-
ferent faith practices can have this effect. One pathway involves promoting
character traits that involve giving others the benefit of the doubt; one
involves accepting God’s love; one involves promoting the virtue of spiritual
excellence; and one involves growing in other virtues through practicing
spiritual excellence. For the first, second, and fourth of these pathways, I do
not think any defense of a view about the general nature of character virtues
is needed for it to be persuasive that the traits these pathways promote are
character virtues. But for the third pathway, it is helpful to have in hand a
widely appealing view about the nature of character virtues. I will offer such
a view here, arguing that it implies that all of the traits that are the focus of
my later arguments—­including, most contentiously, spiritual excellence—­
are indeed character virtues.
So, what makes a character trait a virtue? It is helpful to start by saying
what character traits are. Here I intend to follow a widely held view among
virtue theorists (cf. Miller 2014). Character traits are tendencies to perceive,
think, feel, and act in characteristic ways under characteristic triggering cir-
cumstances, out of characteristic motives or values. As tendencies, charac-
ter traits are dispositions; a person who possesses a character trait is inclined
by that trait to behave in certain ways. The “behaviors” they are inclined to
exhibit are a diverse bunch, making any character trait a multitrack disposi-
tion. A character trait makes a difference for what a person notices and how
they perceive the world, for the judgments they make (especially their eval-
uative judgments), for the emotions they experience, and for the overt
actions they take or try to take. A character trait inclines one toward such
behaviors under characteristic circumstances—­not equally all the time in
whatever circumstance one finds oneself in. And when a character trait
inclines one to perceive, think, feel, and act in some way, it inclines one to
do so on account of values one has (i.e., what one cares about) as a possessor
of that trait.
To illustrate the idea, consider honesty (for a recent account, see
Christian Miller 2021). Someone who possesses the character trait of hon-
esty values honest behaviors and disvalues dishonest behaviors. Because of
these values they hold as someone with the character trait of honesty, they
are inclined to behave in characteristically honest ways in characteristic
triggering circumstances. They are inclined to notice and pay attention to
facts that bear on whether a behavior is or would be honest; to evaluate
 Virtue and Flourishing 95

honest behaviors positively and dishonest behaviors negatively; to feel upset

or put off by dishonest behaviors and elevated or irenic toward honest
behaviors; and to act or try to act honestly, for example, by telling the truth
as they see it or refraining from cheating. Not just any circumstance (e.g.,
when playing pretend) will trigger such behaviors, but only ones where
honest behavior is relevant.
Our question, then, is what would make a character trait such as honesty
a virtuous character trait. I think we can make a good start toward an
answer by appealing to the widely endorsed principle that it is valuable to
value what is valuable (Hurka 2001). Character traits by their very nature
involve ways of valuing. The behavioral tendencies to which they give rise
are generated because of the foundational values the character trait involves.
Thus, according to the widely endorsed principle, a valuable or good char-
acter trait will be one that involves valuing what is in fact valuable. I suggest
that this provides us with an illuminating and foundational necessary con-
dition for character virtues: for a character trait to be a virtuous character
trait, it must be that what the character trait involves valuing is in fact valu-
able and that what it involves disvaluing is in fact disvaluable. Honesty, for
instance, will be a character virtue only if honest behaviors are valuable and
dishonest behaviors are disvaluable.
Can we get from this simple necessary condition for a character virtue to
an answer to the general question that will allow for a defense of the view
that spiritual excellence is a virtue? I think we can, and rather easily. What is
needed, first, is a second principle that is just as intuitive as the first. It says
that not only is it valuable to value what is valuable but that the value of
valuing the valuable increases in accordance with how well the valuable is
valued. It is, for instance, better to value the valuable well than to value it
poorly. According to this principle, character traits that involve valuing
what is valuable well will be better—­even more valuable—­than character
traits that involve valuing what is valuable less well.
The second thing we need is the idea that virtues are a “satis” concept
(Russell 2009). Using the language we’ve been working with here, the idea is
that one has a virtue when one has a character trait that involves valuing
something valuable well enough. There’s some point along the measurement
axis of how well one values something valuable after which one has a charac-
ter virtue of valuing that value. Honesty, for instance, involves valuing well
enough honest behaviors and disvaluing well enough dishonest behaviors.
This basic view of character virtue gets us a long way toward being able to
defend the view that spiritual excellence and the other virtues that are my
96 Faith, Flourishing, and Agnosticism

focus are virtues. To defend the claim that they are virtues, we need to argue
that they are ways of valuing well enough something valuable. It will assist
us in defending this claim, however, if we can say a bit more about some of
what is involved in valuing a value well. I will offer two uncontroversial
­suggestions, then develop an initial argument that spiritual excellence is a
virtue, and then respond to some objections. In doing so, I will suggest that
the view of character virtue given is one that is both adequate for the argu-
mentative purposes of this book and one that can be accepted—­perhaps
with some additional modifications in some cases—­by authors who defend
a variety of different fundamental theories of virtue.
My first suggestion about what is involved in valuing something valuable
well is that this involves adopting fitting attitudes to it. For instance, to
shriek in terror at something that is not very fearful would involve a mis-
match of fit between one’s attitude and its object. In this way, someone who
otherwise was inclined to overcome their fears for the sake of promoting
good ends could fail to be courageous because they do not tend to fear
appropriately—­they are too fearful toward what is not frightening. Likewise,
to strongly desire something that is only minimally desirable would be a
failure of fit, but to strongly desire what is highly desirable would be to
adopt a much more fitting attitude toward it. In order for a character trait to
involve valuing something valuable well, it must involve tending to adopt
fitting attitudes toward the value in question.
My second suggestion is that to value well involves valuing with skill. It is
an uncontroversial idea that there is some sort of analogy between character
virtues and skills (cf. Annas 2011), and indeed that character virtues include
components of skill. Generosity, for instance, is thought to involve not just
caring about others’ welfare and trying to benefit them with one’s resources,
but also employing skill in identifying what sorts of potential benefits really
would benefit them and how one can give these benefits in an excellent way
(Roberts and Wood 2007). For a character trait to involve valuing some-
thing valuable well, then, it must incorporate skill in valuing that thing.
The basic view of character virtues that we are working with thus far says
that a character trait is a virtue if it involves valuing something valuable well
enough. And we have sought to partially illuminate what is involved in
valuing a value well by noting that this includes tending to adopt fitting atti-
tudes toward that value and employing skill in one’s valuing of it. I suggest
that these views are adequate for getting an argument off the ground in
defense of the claim that spiritual excellence and the rest of the character
traits I’m concerned with are virtues. The argument may be straightforwardly
 Virtue and Flourishing 97

convincing for some, and for others we may need to first modify our basic
view by adding further claims about what is involved in valuing a value well,
but we can then offer an argument they will find convincing as well.
To state the argument, we need a basic understanding of spiritual excel-
lence. I will expound on the idea in more detail in Chapter 6, but what is
needed here is just the general idea. The way I understand it, spiritual excel-
lence is a tendency to make skillful use of a worldview for which one has
ambiguous evidence or better in order to experience morally transformative
awe of the awesome. So understood, spiritual excellence is a character trait
that is available to both practitioners of various religions and to the nonreli-
gious. It is concerned with appropriately valuing and pursuing moral trans-
formation, and with appropriately valuing and pursuing experiences of awe
in order to facilitate this transformation. The experiences of awe it aims to
induce are experiences where one experiences awesome things as awesome.
And one of the particular qualities of the transformative experiences of awe
it is concerned with is an experience of connectedness to large wholes such
as the earth or the universe or humanity or everything. The person with the
character trait of spiritual excellence makes use of a way of understanding
the world and their place in it that helps them to cultivate experiences in
which they have awe for what is awesome; this awe includes experiences of
connectedness to large wholes and leads to moral growth.
It should not be difficult to see that this character trait is a candidate for a
virtuous character trait given the basic account of character virtues we have
developed. First, spiritual excellence involves valuing some things that are
in fact valuable. Specifically, it involves valuing experiencing awe for the
awesome, connectedness to larger wholes, and moral transformation.
Experiencing awe is pleasant in its own right, and when what is experienced
as awesome is in fact awesome, awe is also a fitting attitude toward it and
valuable for that reason. Likewise, the experience of connectedness to larger
wholes is pleasant in its own right. And given that we are all in fact con-
nected in one way or another to larger wholes, to experience ourselves as in
some way so connected is to have a veridical experience, which is valuable.
Finally, it will be uncontroversial among virtue theorists that moral trans-
formation in the sense of growth toward or in virtue is valuable. So, in vari-
ous ways, spiritual excellence involves valuing things that are in fact
valuable.
Moreover, spiritual excellence involves valuing these valuable things well
in the sense that it involves adopting fitting attitudes toward them. Those
who possess spiritual excellence appropriately value moral transformation,
98 Faith, Flourishing, and Agnosticism

awe, experiences of connectedness, and the like. For example, they tend to
desire experiences of awe and connectedness on account of both their
pleasantness and on account of the contribution they can make to moral
transformation, which they also desire. They also tend to adopt attitudes of
awe toward what is awesome. The character of their desires for these experi-
ences, and the experiences themselves, fit their objects: awe experiences,
experiences of connectedness, and moral transformation are desirable in
the ways the spiritually excellent desire them; and the attitude of awe fits
what is awesome.
Spiritual excellence also involves using skills. I will discuss this in more
detail in Chapter 6, but suffice it to say here that making use of a worldview
for which one may have ambiguous evidence in order to induce transforma-
tive experiences of awe for the awesome does not come naturally to every-
one. It is learned through practice and involves making use of certain
human technologies. Spiritual excellence therefore involves valuing some
valuable things in a way that is not only fitting but skillful. Thus, spiritual
excellence appears a good candidate for a virtue given our account of virtue.
This argument may be enough to convince some readers that spiritual
excellence is a virtue, or at least a good candidate for one. Several leading
authors have, in fact, defended theories of the nature of character virtues
that imply the truth of the basic approach to character virtue developed
above—­that a character trait is a virtue if it involves valuing something
valuable well enough.
This is true, for instance, of Jason Baehr’s (2011) personal worth concep-
tion of virtue. For Baehr, a character trait is a virtue if possessing it makes
its possessor better as a person. Notably, Baehr defends a view according to
which the basis of personal worth, or that in virtue of which character
traits make their possessors better as people, involves valuing something
valuable. As he puts it, “A subject S is good or better qua person to the
extent that S is positively oriented toward or ‘loves’ what is good and is
negatively oriented toward or ‘hates’ what is bad” (97). Moreover, Baehr
argues that not just any positive orientation toward some good or negative
orientation toward some bad will suffice. The orientation cannot, for
instance, be a fanatical orientation, and Baehr argues that where the good
in question is intrinsically valuable, virtues require valuing that good for its
own sake. These seem to be ways of specifying how well one must be posi-
tively oriented toward some good in order for the orientation to count as a
virtue. Thus, it seems that Baehr’s personal worth conception of virtue
implies our basic conception.
 Virtue and Flourishing 99

Something similar is true of Robert Adams’s (2006) approach. For

Adams, “capital V virtue,” or overall good character, is virtuous because it
involves “persisting excellence in being for the good” (32). The individual
moral virtues contribute to good overall character by themselves being
excellent ways of being for some good.1 There are many ways to be for a
good, including “loving it, liking it, respecting it,” and so on. To be an excel-
lent way of being for the good is to be a way that is “worth prizing for its
own sake” (24). This too seems to imply our basic approach, that a character
trait that involves valuing something valuable well enough will count as a
virtue. For Adams, to value something valuable well enough to constitute a
virtue is to value it in a way that is worth prizing for its own sake.
A third and final example is provided by Christine Swanton (2021),
whose approach to conceptualizing virtue is perhaps closest to the one
developed here. On her view, “A virtue itself is understood as a disposition
of good or excellent responsiveness to evaluatively significant features of the
world, within its field” (221). Each character trait focuses upon some field of
concern and is concerned with evaluatively significant features within that
field. When a character trait “latches onto these features in a characteristi-
cally good or correct way,” that is when it is a virtuous trait. Here again, it
appears that the view implies that if a character trait involves valuing well
enough what is valuable, it will be a virtue. What makes Swanton’s view a bit
more similar to the basic view offered here is that she seems to take more of
a piecemeal approach to identifying what the “targets” of virtue are in virtue
of which virtues count as responding in characteristically good ways, rather
than—­as Baehr’s and Adams’s views do—­attempting to provide a simple
theory about why these characteristically good ways of responding are
good. Her aim in articulating the above theory, like my own, is to provide a
“minimalist definition of virtue” that “is compatible with a great variety (if
not all) rival conceptions of virtue” (206).
On all these views, then, a character trait that involves valuing well
enough something valuable will count as a virtue. I have argued that spir­it­
ual excellence, and indeed the other virtues that will be my concern in later
chapters, involve valuing well enough something valuable. In the case of
spiritual excellence, I have argued that it involves valuing certain specific

1 Adams offers a different view of what makes structural virtues virtuous. I note here that I
intend to allow for a pluralistic approach to what makes character virtues virtuous. The view I
offer in the text is just offered as one way whereby a character trait can count as virtuous; there
may be others. Cf. (Baehr 2011: 89–90).
100 Faith, Flourishing, and Agnosticism

values and that there is reason to think it involves valuing these well because
it involves valuing them in accordance with certain key features involved in
valuing goods well. This gives us reason to think it involves valuing these
goods well enough. One might attempt to challenge the claim that spiritual
excellence is a virtue based on one of the above theories by appealing to the
particular way of conceptualizing what it is to value well that is involved in
that theory. For instance, one might try to argue that while spiritual excel-
lence involves valuing certain values well in some sense, it doesn’t make its
possessors better as people, or it isn’t an admirable way of valuing. But I
doubt that this pathway to resisting the virtuousness of spiritual excellence
will be very compelling. We can find a more potent pathway of resistance
elsewhere, and addressing it may help to alleviate this source of concern
as well.
While defenders of the above theories of character virtue or closely simi-
lar theories may be persuaded that spiritual excellence is a virtue by the
argument developed so far, advocates of other theories of virtue may be
more hesitant. I am thinking here especially of views in the neo-­Aristotelian
family, which may give us pause about whether the basic view of virtue
I have outlined is sufficient for identifying human virtues. I will discuss two
concerns that defenders of these theories of virtue may have with the pres-
ent argument, showing how one might make additional modifications to
the basic view of virtue in order to accommodate these concerns while still
supporting the conclusion that spiritual excellence is a virtue.
The first concern is based on the idea that the basic view of virtue does
not adequately reflect the humanness of virtue. It offers an account of a way
that a character trait may count as a virtue, perhaps, but not an account of
how a trait may count as a human virtue. Advocates of naturalistic
approaches to ethics will find this suspicious because, in their view, ethics
must be founded upon a conception of human nature, and the basic view
I have sketched is not so based. This is important because there may be ways
of valuing well things that are valuable that are alien to human beings. To be
a human virtue, a trait must not only involve valuing something valuable well;
it must involve doing so in a way that is accessible to human beings and
reflective of their nature as human beings. It must not involve valuing some-
thing valuable in a beyond-­human way. As Martha Nussbaum (1990) puts
it, “there are some very general conditions of human existence that are also
necessary conditions for the values that we know, love, and appropriately
pursue” (79).
 Virtue and Flourishing 101

The concern may be clear enough couched in these general terms, but we
can also develop a more specific version of it by referring to a leading exam-
ple of this kind of neo-­Aristotelian theory, such as Rosalind Hursthouse’s
(1999). On Hursthouse’s view, the broad structure of ethical evaluation for
human beings is similar to that involved in the evaluation of plants and ani-
mals. We evaluate any of these things as a good specimen of its kind insofar
as its evaluative aspects tend to foster the ends characteristic of its species.
Human beings are by nature social animals whose ends are individual sur-
vival, continuance of the species, characteristic enjoyments and freedom
from pain, and the good functioning of their social group. Thus, “human
beings are ethically good in so far as their ethically relevant aspects foster
the four ends appropriate to a social animal, in the way characteristic of the
species. And the structure—­the appeal to just those four ends—­really does
constrain, substantially, what I can reasonably maintain is a virtue in human
beings” (224). To count as a virtue, a trait must not just involve valuing
something valuable well, but it must involve doing so in such a way as to
adequately promote the four ends of social animals and not be inimical
to them.
I think a promising response can be made to this concern, both in its
generic form and its specific articulation appealing to Hursthouse’s view, by
modifying our basic view of character virtue and arguing that spiritual
excellence satisfies the modified view. In response to the generic objection,
for example, we could modify our basic view so that it says that a character
trait is a human virtue if it involves valuing well something valuable in a
way that fits well with human nature or is characteristic of human beings.
Exactly what it is for some way of valuing to “fit well with human nature” or
to be “characteristic of human beings” is a fraught notion. But however we
understand the notion, there is a great deal of plausibility to the claim that
spiritual excellence as described ought to count as a way of valuing some-
thing valuable that fits with human nature and is characteristic of human
beings. As David McPherson (2017) argues, the recorded history of human-
ity to the present indicates that human beings are “homo religious—­i.e.,
­naturally drawn to spirituality” (74). Practicing spiritual excellence is
something human beings can do; most human beings do it to some extent,
and some do it very well. Spiritual excellence is a very human way of valu-
ing valuables.
A parallel approach can be made in response to the concern as expressed
through Hursthouse’s view. Though we could perhaps attempt to resist her
102 Faith, Flourishing, and Agnosticism

claim that human beings have just the four ends she lists,2 I don’t think
doing so is necessary for answering the concern. We can again modify our
basic view so that it now claims, for example, that a character trait is a
human virtue if it is a way of valuing valuables well that when exercised by
human beings tends to promote their four ends and not be inimical to them.
We then need to argue that spiritual excellence, when practiced by human
beings, tends to promote the four ends and not be inimical to them. And
again, I suggest a plausible case can be made for thinking this is true.
First, if practicing spiritual excellence is indeed conducive to developing
or maintaining other more standard virtues, and it is granted—­as it is by
Hursthouse—­that these other virtues tend to promote the four ends and are
not inimical to them, then spiritual excellence will be indirectly conducive
toward promoting these ends. And this makes good sense, since as concep-
tualized spiritual excellence is supposed to be an example of a virtue of self-­
improvement (cf. Swanton 2015). Second, I suggest that spiritual excellence
is directly conducive to the third end of human beings without being inimi-
cal to the other ends. It is directly conducive to human beings experiencing
characteristic enjoyments in the form of experiencing awe of the awesome
and connectedness with large wholes. These experiences, we have noted, are
pleasant, and they are characteristically available to human beings. As John
Cottingham puts it, they are part of our “ordinary human birthright”
(2014: 63).
Talk of these experiences being characteristically available to human
beings brings us to a second concern that is focused on one way of spelling
out exactly what the characteristic human way is. As Hursthouse explains
the view, “Our characteristic way of going on, which distinguishes us from
all the other species of animals, is a rational way. A ‘rational way’ is any way
that we can rightly see as good, as something we have reason to do” (1999:
222). One might worry that even if practicing spiritual excellence is charac-
teristic of human beings in some other sense of fitting with their nature or
being generally available to them as a way of valuing that promotes their four
ends, it could still fail to be a virtue on account of involving ir­ra­tion­al­ity—
­as not being something humans have reason to do. Indeed, Hursthouse
offers some remarks about theistic spirituality, or piety, in particular that
seem to suggest this view.

2 McPherson (2017) seems to prefer this approach. It may be necessary for defending the
virtuousness of piety as he understands it, but it is not necessary for defending the virtuousness
of spiritual excellence as I conceptualize it in Chapter 6.
 Virtue and Flourishing 103

Hursthouse asks us to imagine an atheist evaluating a theist’s practice of

theistic piety. The sort of piety in view is a sort that “prompts them to pray,
to refrain from blasphemy, to go to church, to spend time thinking about
God and trying to get closer to an understanding of Him” (233) and the
like. An open-­minded atheist making such an evaluation may well grant
that such piety in the pious “is inseparably intertwined with, and positively
reinforces, their other virtues” (232) and so in this way indirectly “fosters
the four ends” (233) of human beings. They may grant, moreover, that “piety
undoubtedly brings great joy and serenity to its possessors” (233). In these
respects, it would otherwise be a good candidate for a virtue. The problem,
however, is that “from the atheist's standpoint [practicing piety] is based on
a complete illusion; reason cannot endorse it.” Indeed, “the right reasons
[the pious] think they have . . . for doing these things, are no reasons at all”
(233). Thus, “the atheist cannot judge piety to be a virtue without abandon-
ing her atheism” (234).
Now, this argument from Hursthouse is not exactly an easy one to inter-
pret. But charitably understood, it seems that she must be assuming that
whatever reasons the theist in question thinks they have for practicing piety,
the atheist has access to exactly those same reasons. The atheist knows what
the reasons are that the theist thinks they have for practicing piety and
judges these to be no reasons at all. Since these reasons the atheist has access
to are the only reasons the theist has for engaging in theistic piety, and the
atheist does not regard these “reasons” as reasons, she cannot judge the the-
ist’s practice of piety to be something the theist has reason to do. If she did,
she would have to give up her own atheism because she would have to judge
that she too has reasons to engage in theistic piety, since she has access to
the same reasons for this that the theist does.
Expressed in this way, I think Hursthouse’s argument makes sense and is
unobjectionable. But it also will not present a challenge to the rationality of
practicing spirituality—­or even theistic spirituality—­of a sort that should
concern us in the context of this book’s larger project. Understood as indi-
cated above, Hursthouse is just making an illustrative point using a particu-
lar example where an atheist and a theist have access to all the same evidence
for theism, and she is only considering what evaluation the atheist must
make of the theist’s piety in that circumstance. But this does not imply—­and
if it did, the implication would be highly questionable—­that every atheist
and every theist share exactly the same evidence for theism. If an atheist
thinks that a theist may have rather different evidence for theism from her-
self, then she will not be in a position to judge as Hursthouse’s atheist does
104 Faith, Flourishing, and Agnosticism

that the reasons this theist thinks they have for practicing theistic piety are
no reasons at all. More broadly, in the context of the project we are con-
cerned with in this book, it’s not the perspective of the atheist that is rele-
vant for assessing the rationality of practicing spiritual excellence. What
should matter for us is whether engaging in spiritual excellence—­and prac-
ticing theistic spirituality in order to do so—­is rational when one has
ambiguous evidence for the worldview one employs in doing so. But the
question whether it is rational to engage in faith practices when one has
ambiguous evidence for God is just the question that we considered already
at length in Chapter 3. And there we saw that it is highly plausible that if a
faith practice such as practicing theistic spirituality has the kinds of values
Hursthouse’s atheist is willing to grant that it does, then for a person with
ambiguous evidence for God it is indeed rational—­at least practically, and
perhaps epistemically as well.
An atheist who thinks that no individual has ambiguous evidence or bet-
ter for God may not be able to judge that practicing theistic spirituality as a
means to cultivating spiritual excellence is rational. But such an atheist will
need to reckon with the arguments of Chapter 2 that there is a sizable popu-
lation of individuals who do have ambiguous evidence for God. If those
arguments are on the right track, and those of Chapter 3 are as well, then
there is a sizable population of individuals for whom practicing theistic
spirituality as a means to developing spiritual excellence will be rational.
There is also a more general point that needs to be made in response to
the concern that spiritual excellence in general, and not just theistic spir­it­
uality, is irrational. Hursthouse’s argument is concerned only with practic-
ing theistic piety, not with piety or spirituality more generally.3 Indeed,
I would suggest that, as characterized above and in Chapter 6, an atheist can
endorse the rationality of practicing spiritual excellence. They can do this
even if they take the extreme view that no individual has ambiguous evi-
dence or better for God and so practicing theistic spirituality is irrational.
The point is simply that there are nontheistic ways to practice spiritual
excellence. I do not think that Hursthouse’s argument suggests that she
thinks otherwise, and in the Conclusion of this book I will discuss some
examples of nontheistic ways of practicing spiritual excellence. To contend
that spiritual excellence in general must be irrational would seem to require

3 McPherson (2017) seems to miss this point, claiming that Hursthouse “narrowly con-
strues piety as theistic piety” (75). But, in context, she is just discussing theistic piety for illus-
trative purposes and doesn’t comment on a more generic sort of piety.
 Virtue and Flourishing 105

arguing that no way of acting when one has ambiguous evidence is ever
rational, but the arguments of the previous chapter strongly indicate that
this is a wrongheaded way of thinking about human rationality. Indeed, if
anything is characteristic of human beings, it is that we often have to act on
the reasons we have when we have ambiguous evidence.

2 Flourishing

The previous section argued that the character traits that I will argue can be
fostered via faith practices are indeed virtues. If they are, this provides indi-
viduals with ambiguous evidence for God some reason to engage in these
faith practices, insofar as growing in these virtues is valuable for its own
sake. In this section, I explore the question whether growth in these virtues
might be valuable for another reason—­namely, that it is likely to enhance a
person’s level of flourishing. I develop a basic account of flourishing and
argue that growth in virtue is likely to promote greater flourishing. I then
respond to concerns about this argument that threaten to restrict the popu-
lation for whom its conclusion holds true or to undermine its significance.
I argue that its conclusion holds true for most individuals with ambiguous
evidence for God and that the conclusion is significant and should be moti-
vating for many such people.
Before addressing the topic of flourishing, it will be helpful to have more
clearly in mind the idea of “growth in virtue” that is relevant here. A person
grows in a virtue through engaging in a faith practice if they move closer
toward possessing a virtue, or come to possess a virtue more fully, or sustain
their current level of or closeness to possessing a virtue that would be
diminished without their engaging in the practice. The topic of the present
section is whether such growth would facilitate growth in the person’s flour-
ishing in much the same sense. That is, by growing in character virtues,
would a person also get closer to experiencing flourishing, or experience
even greater flourishing than they are already experiencing, or retain their
level of or closeness to flourishing that would be diminished in the absence
of this character growth? I will argue that this is indeed likely.
Start, then, with a basic approach to conceptualizing flourishing—­one
which I will suggest can be affirmed by advocates of several different views
of flourishing that differ in their details. By flourishing, I mean living well.
More specifically, I mean living well as a human. And since evaluations of
flourishing are most properly concerned with life as a whole, to flourish is
106 Faith, Flourishing, and Agnosticism

to live well as a human over the course of a complete human life. A human
who leads a flourishing life has lived a good human life.
Thought of in this way, flourishing stands to human lives much as virtue
stands to human character traits, as outlined in the previous section.
Flourishing human lives are good human lives, and the degree to which a
human life is a flourishing life is a measure of the degree of the goodness of
that life, just as virtues are good human character traits, and the degree of
virtuousness of a character trait is a measure of how good that character
trait is. What makes a human character trait good and virtuous, we saw, is
that it involves valuing well what is valuable in a way that fits with human
nature. I suggest that a parallel story is true of what makes a human life
good and flourishing. A human life is good and flourishing if it is a com-
plete human life that involves persistently acting well in a way that fits with
human nature.
Given this basic view of human flourishing, I argue that growth in char-
acter virtue is likely to promote growth in flourishing. The reason for this is
that to grow in virtue is to gain, or at least avoid loss, in one’s capacities to
act well in a way that fits with human nature. By growing in virtue, one
either becomes better at valuing valuables in a human way, or one avoids
getting worse at valuing valuables in a human way. But such growth either
constitutes strengthening or retaining one’s capacities to act well in a human
way. In turn, when one’s capacities for acting well are strengthened or
retained, one is thereby more likely to use these capacities to act better than
if they had not been strengthened or retained. Acting better contributes to
how well one is living—­to one’s flourishing—­given the above basic view of
flourishing. So, if a person grows in virtue, this makes it likely they will
experience greater flourishing; growing in virtue makes it likely that a per-
son will live a better life.
I will assess this argument by considering two topics. First, I will argue
that the basic view of flourishing employed in the argument is one that can
be embraced by defenders of several different views of flourishing that differ
in their details. Thus, the argument should have somewhat broad appeal.
Second, I address some concerns about the argument that threaten to
restrict the population for whom its conclusion holds true or to undermine
its significance.
Begin, then, with the flexibility of the basic view of flourishing appealed
to in the argument—­the view that a human life is a flourishing life to the
extent that it is a complete human life that involves persistently acting well
in a way that fits with human nature. This view will, of course, sound very
 Virtue and Flourishing 107

familiar because it is very much in the vein of Aristotle’s own view of human
flourishing or eudaemonia. As Kristjánsson (2019) explains Aristotle’s view,
eudaemonia “involves virtuous, reason-­infused activity, suitable and pecu-
liar to human beings, achieved over a complete life” (9). “Virtuous” here is a
synonym for “reason-­infused,” and it is a more detailed way of describing
activity that is performed well (cf. Russell 2012: ch.3). “Suitable and peculiar
to human beings” is another way of describing a way of life that fits with
human nature. So, Aristotle’s view does seem to be a more detailed way of
expressing the basic view that flourishing involves acting well in a way that
fits human nature over a complete life.
But Aristotle is not the only philosopher who can embrace the basic view,
and further consideration of the particularities of Aristotle’s view can reveal
some choice points where philosophers may differ with Aristotle over
details while retaining the basic view offered above. For instance, one of the
well-­known features of Aristotle’s view is his stance on the sufficiency of
virtuous activity for eudaemonia. Aristotle defended the view that virtuous
activity within whatever circumstances one finds oneself is not sufficient for
flourishing, whereas some other ancient philosophers—­notably the Stoics—­
argued that acting virtuously within one’s circumstances was sufficient for
flourishing (see Annas 2011). For Aristotle, fortunate circumstances, such
as having adequate wealth, friends, and physical attractiveness, were needed
for performing sufficiently valuable activities in at least two ways (Curzer
2012: 422–23). In some cases, valuable activities cannot be performed with-
out these things because they are a necessary means to performing those
activities, whereas in other cases, if these necessities of life are not fulfilled,
they distract one away from performing valuable activities that are needed
for flourishing.
What I want to point out here is that both Aristotle’s view and the Stoic
view are compatible with the basic view of flourishing developed above.
They are both refinements or further specifications of it, and the argument
advanced above for thinking that growth in virtue is likely to promote
growth in flourishing can be accepted by defenders of each account.
Roughly, Aristotle’s view is that the activities needed for flourishing must be
of sufficient value that to perform enough of them over the course of a com-
plete human life, one will need the external necessities. The Stoic view does
not make this requirement for flourishing. The kinds of valuable activities
one can perform with limited resources and in unfortunate circumstances
are adequate for flourishing. One can even lead a complete human life—­one
that “come[s] to some sort of fruition” (Curzer 2012: 414)—despite one’s life
108 Faith, Flourishing, and Agnosticism

being brought to an abrupt end due to bad luck, if one handles this bad luck
with sufficient virtue.
Another choice point on which others will differ with Aristotle concerns
the necessity of virtue for flourishing. Most commentators interpret
Aristotle as claiming that while virtuous activity is not sufficient for flour-
ishing, it is necessary. To flourish, a person must possess the virtues and
exercise them. Some contemporary neo-­Aristotelians differ from Aristotle
on this point. They suggest instead that acquiring the virtues is “the only
reliable bet” (Hursthouse 1999: 172) for achieving flourishing. As
Hursthouse explains the view, “To claim that the virtues, for the most part,
benefit their possessor, enabling her to flourish, is not to claim that virtue is
necessary for happiness. It is to claim that no ‘regimen’ will serve one
better—­no other candidate ‘regimen’ is remotely plausible” (173). We might
put the point the following way. Living well involves acting in accordance
with virtue. But one can act in accordance with virtue without acting from
virtue. Possessing virtue makes it more likely that one will act in accordance
with virtue and thereby flourish, but it is strictly possible to act in ac­cord­
ance with virtue without possessing virtues.
Again, both views are compatible with the basic view of flourishing, and
adherents of each view can accept the argument given above for thinking
that growth in virtue is likely to facilitate greater flourishing. One way of
specifying what it takes to act well enough to flourish is that this takes pos-
sessing and acting from virtue; another is that it merely involves acting in
accordance with virtue. On both views, one is likely to experience greater
flourishing if one grows in virtue, since on both views this makes it more
likely that one will act better.
Finally, some contemporary authors have defended views of flourishing
that seem even to reject the claim that developing virtues is the most reli-
able way to achieve flourishing. This seems to be implied by Kristján
Kristjánsson’s (2019) view. Kristjánsson defines flourishing as “the (rela-
tively) unencumbered, freely chosen and developmentally progressive
activity of a meaningful . . . life that actualises satisfactorily an individual
human being’s natural capacities in areas of species-­specific existential tasks
at which human beings . . . can most successfully excel” (1). There are two
features of Kristjánsson’s view that are worth highlighting here.
First, on Kristjánsson’s view, flourishing is fundamentally about the actu-
alization of human capacities. This can include capacities for engaging in
virtuous activity, but Kristjánsson wants to make room for the idea that it
need not include this. For example, he wants to make room for the idea that
 Virtue and Flourishing 109

a person who develops outstanding talents in sports or medicine or academia

or art can thereby flourish even if they have not also developed stand­ard
moral or intellectual or civic virtues.
Second, Kristjánsson emphasizes the contribution to flourishing that can
be made by states of continence as opposed to virtue. Continent individu-
als, roughly, are those who tend to act in accordance with virtue but
only after struggling against contrary temptations. On Kristjánsson’s view,
­“continence [is] a somewhat morally under-­estimated character trait” (21)
that can be exercised “with grace and ingenuity” (22). Noting that even on
Aristotle’s view most people only reach continence as opposed to full virtue,
he expresses his view as follows: “my general claim is that continence suf-
fices for flourishing” (22).
It is the second aspect of Kristjánsson’s view that suggests he may deny
that developing virtues is the “only reliable bet” for achieving flourishing.
Another reliable bet, one that is perhaps even more reliable, is to develop
states of continence. This is more reliable in the sense that people are more
likely to achieve flourishing by developing states of continence than by
developing virtues, since they are more likely to achieve states of continence
that enable them to flourish than they are to achieve virtues that enable
them to flourish.
Although Kristjánsson’s view differs from the others we have surveyed
in these respects, it is still, like these others, a way of developing the basic
view of flourishing in more detail. Kristjánsson can agree that what it is
for a human life to be a flourishing life is for it to be a complete human life
of acting well in a way that fits human nature. He just wants to stress that
acting well can be achieved through the actualization of capacities that are
not typically regarded as involving virtue as well as through acting
continently.
Kristjánsson can also accept the argument given above for thinking that
growth in virtue is likely to promote greater flourishing. This is true even
though on his view there is a sense in which developing virtues is not the
most reliable means of achieving flourishing. For we must recall what was
meant by growth in virtue. This could consist either in growing closer to
virtue if one isn’t virtuous already, in becoming more fully virtuous if one is
already virtuous, or in maintaining one’s level of virtuousness or closeness
to virtue rather than deteriorating. Kristjánsson can accept, for example,
that a person who moves from being vicious to being continent is thereby
more likely to flourish to a greater extent, or that the same is true of some-
one who is continent and avoids becoming incontinent. Indeed, even a
110 Faith, Flourishing, and Agnosticism

person who moves from being continent to being virtuous lives better, since
as Kristjánsson puts it, continence is only a “second best tack” (22).
I will conclude this section by addressing some concerns about the
­argument I have given that growth in virtue makes growth in flourishing
more likely. One kind of concern focuses on the scope of individuals to
whom the conclusion of the argument applies. It aims to show that there are
some groups of individuals for whom growth in virtue is not likely to
­promote growth in flourishing, and thereby to identify a limitation of the
argument. I grant that the argument does have such limitations, but I will
suggest that they are not very limiting.
One kind of individual to whom the conclusion of the argument might not
apply is the sort of “great achievers” that Kristjánsson focuses on, such as
sportspeople or artists. Perhaps their level of “focus on their specific talents”
really does require “such concentration of motivation and effort that displays
of moral virtue . . . will get squeezed out” (21). Any growth in virtue would
come as an opportunity cost in that it would interfere with the maintenance
of their specific talents, and so would be injurious to their overall flourishing.
I do not think that even Kristjánsson would buy this argument. As he
clarifies, on his view, great achievers will not qualify as flourishing if they
are vicious, but only if they are continent. Indeed, their continence must be
preserved for them to qualify as flourishing. But acting in ways that pre-
serve one’s character as continent qualifies as a kind of “growth in virtue” in
the sense of that term here. Thus, if engaging in faith practices could help
such a person sustain their character as continent, it could contribute to
their flourishing. And I suggest it indeed could play that role. In fact, it may
be an especially attractive way of doing this, given that several of the faith
practices on which we will focus, such as giving thanks and praise to God,
are fairly undemanding on one’s time and concentration.
There is another point that should be made about the role of character in
the lives of great achievers as well. For many great achievers, the part of the
human lifespan in which these achievements can take place is limited. We
sometimes call the time when they can make these achievements their
“prime.” But a complete human life for such people extends beyond their
prime. Oftentimes, the great achievers we most admire are those who, after
their prime has come and gone, turn their energies to invest in younger
generations, offering inspiration and wisdom and supporting good causes.
Here it seems that growth in virtue may be particularly relevant. Even if
growth in virtue will not contribute now to their level of flourishing, it may
contribute to this later.
 Virtue and Flourishing 111

There is another group of individuals for whom the tension between

growth in virtue and maintenance of their “specific talents” may be even
more extreme than in the case of great achievers. I am thinking of people
whose legitimate and valuable roles require them to display role-­specific
character traits that seem to conflict with their possession of standard
­virtues. For instance, there is research that suggests that first responders
perform better when they tend not to experience compassionate feelings
for people suffering. These feelings only get in the way; what they need is
aplomb. But these same first responders are known to struggle in their per-
sonal lives because they seem to react with indifference to the suffering of
their close others (Carrico 2012; Regehr et al. 2002; Rudd and D’Andrea
2015). Here it may seem that growth in a standard virtue such as compas-
sion may not just distract, as in the case of great achievers, but may pull in
exactly the opposite direction needed for sustaining their valuable activities.
I think these kinds of cases are very important and should be cause for
public concern and receive more attention from philosophers. Some philos-
ophers have addressed cases like this, often arguing in one way or another
that role-­differentiated virtues cannot, despite appearances, conflict with
basic virtues. The best and most recent work I know of on this topic is
Christine Swanton’s (2021: ch.7). But here I don’t want to insist on an easy
or abstract solution for these cases. I want to acknowledge the tension and
treat it as prima facie a genuine conflict between growing in (basic) virtue
and achieving greater flourishing.
Even granting this genuine conflict, it does little to challenge the scope
of applicability of the above argument, for the fact that growth in one
virtue would not make it more likely that an individual would attain
greater flourishing does not imply that their growth in other virtues
would not make it more likely that they will attain greater flourishing.
Perhaps, for instance, a firefighter in the above circumstances could grow
in their ability to own their numbness toward their close others’ suffer-
ing, acknowledging this and its effect on these people they care for, and
experiencing remorse for it while recognizing it as a regrettable conse-
quence of excellent functioning in their work. Growth in that sort of vir-
tue may help them to live better and may not undermine their ability to
display the role-­ differentiated virtues demanded in their vocation.
Indeed, more generally, whatever the best available approach is for the
first responders—­whether this involves sacrificing the one virtue for the
other, the other for the one, or somehow threading a needle to possess
both—­is likely to itself demand a display of virtue. Growth in whatever
112 Faith, Flourishing, and Agnosticism

that virtue is—­and here I mean to be agnostic, as I have no tidy solution

for their predicament—­would contribute to their flourishing.
In addition to the above individuals for whom growth in virtue may seem
to be in tension with their continued performance of valuable activities cen-
tral to their flourishing, we might also wonder about individuals at the more
extreme ends of the spectrum in terms of virtue and vice. At the extreme
end of virtue, for instance, it would seem obvious that the perfectly virtuous
person won’t attain greater flourishing through growing in virtue, since they
can’t grow in virtue. And at the other end, someone who is downright
vicious may not be capable of growth in virtue; their character may have
become immovably stuck in vice. Regarding the perfectly virtuous, we
should note again that one kind of “growth” in virtue is sustaining rather
than deteriorating in one’s possession of virtue. And it may indeed seem
plausible that even for the perfectly virtuous, their virtue could atrophy
through lack of use (cf. Byerly 2017). If, when faced with a situation in
which virtuous activity is appropriate and they don’t display that activity,
this could weaken their possession of virtue. And, indeed, some of the faith
practices I will be describing are practices that involve behaving in ac­cord­
ance with virtue under some set of circumstances. So, at least when it comes
to these faith practices, it remains plausible that by engaging in them, a per-
fectly virtuous person could retain their level of perfect virtue rather than
having it deteriorate and in that sense could “grow” in virtue, making it
more likely that they will grow in flourishing. But, of course, the main reply
to emphasize about the perfectly virtuous is that this group is almost, if not
entirely, unpopulated. The fact—­if it is a fact—­that the above argument
won’t apply to the perfectly virtuous is not a serious limitation on the scope
of that argument.
A more serious concern might derive from the other end of the spec-
trum, focusing on the stubbornly vicious, since this region is sadly more
densely populous. In this vein, Dan Russell (2012) asks us to consider the
character Crumb, who is just incapable of genuinely loving others. Russell
argues that this is a pitiful fact, but that if it is true, it should regulate how
we go about caring for Crumb and promoting his flourishing. Given that he
really can’t become more loving, we shouldn’t “force the issue” with him.
“Crumb is just Crumb”; the best we can do is “make him comfortable, and
leave him alone” (55).
Now, one comment I want to make about such Crumb-­like individuals is
that we can sometimes be too quick to give up on them. In particular,
­following Kristjánsson (2019), I want to suggest that even if gradual growth
 Virtue and Flourishing 113

in virtue through the typical channels of practicing ordinary virtuous acts

will not work for such a person, there may be a more dramatic alternative yet
available. This is for them to experience a kind of “epiphanic moral conversion”
(114) involving a radical readjustment of their moral outlook. Epiphanes
like this may be promoted particularly through one of the faith practices
I will discuss—­the one that focuses on spiritual excellence. As Kristjánsson
suggests, the emotion of awe may “hold the key” (128) to epiphanic experi-
ences. If so, then a practice that focuses precisely on cultivating awe of the
awesome may be just what Crumb needs.
But, of course, a transformative epiphany won’t work for every vicious
person. Given the original setup of the Crumb case, we really were sup-
posed to take it that Crumb just cannot via any means grow in virtue. And if
that’s true, then growing in virtue won’t help Crumb experience greater
flourishing. Crumb is stuck where he is. Leave him alone.
I’ve been addressing here a concern focused on ways that the conclusion
of my argument might be limited in scope, in that it will not be applicable to
all people with ambiguous evidence for God. I’ve been arguing that while
there are some such limitations, they are not very limiting. A different con-
cern with the argument, however, challenges whether its conclusion is or
should be motivating even for those for whom it is true. One way to develop
the concern is to ask whether living a better life, in the sense in view in the
argument, is “better for” the person whose life it is. If not, it might be argued
that the fact that growing in virtue will help them live a better life does not
provide them with reason to pursue such growth. All that does and should
matter to them is living a life that is better for them, not living a better life.
Perhaps the most common approach to answering this kind of concern is
to argue that a flourishing life in fact is a life that is good for the person who
lives it (cf. Hursthouse 1999: ch.8; Russell 2012: ch.2). Indeed, this is just
what it is to live a life that is good for oneself—­it is to live well in the sense of
flourishing. That’s the very thing that the concept of flourishing was always
supposed to have been about. Theories of flourishing are just theories of
what it is to live a life that is good for oneself.
I do not think that theorists of flourishing need to circumscribe their
project in this way. Theories of flourishing can stand on their own inde-
pendently of whether they are also theories of what it is to live a life that is
good for oneself. Living a good human life and living a better life than one is
currently living are both desirable independently from whether these also
involve living a life that is good for oneself or better for oneself. If growing
in virtue helps you to live better, that’s a reason to pursue such growth
114 Faith, Flourishing, and Agnosticism

even if it doesn’t also help you live a life that is better for you (cf. Curzer
2012: 422).
What does it mean, then, to live a life that is better “for you”? It’s tempting
to think that what is meant is a life that you will enjoy or that you will be
satisfied with, or a life filled with many of the sorts of goods that humans
appropriately desire. Indeed, there are hedonic and desire-­satisfaction and
objective list theories of what well-­being consists in, where the notion of
well-­being is equated with living a life that is good for the one living it (see
Crisp 2021 for an overview). If living a life that is good for one is living an
enjoyable life, or a life in which one gets most of what one wants out of life,
or a life that ticks the right objective list boxes, then it’s an open question
whether growth in virtue will help someone live a life that is better for them.
Theorists of flourishing do often emphasize that there is a certain charac-
teristic pleasure involved in exercising the virtues (cf. Kristjánsson 2019: 7).
And they point out that acting in accordance with virtue helps a person’s
relationships go better, which helps them get more of what they appropri-
ately want out of life and experience less relational stress and better inter-
personal pleasures (Battaly 2015: ch.6). These are good points that highlight
a potential relationship between flourishing and well-­being understood in
terms of these other theories. But they may only support the conclusion that
large gains in virtue are likely to be conducive to well-­being. It’s less clear
that just any growth in virtue will make it more likely that one will experi-
ence greater pleasure or be more satisfied with one’s life or tick more of the
objective list theorist’s boxes. The latter seems to me an open question.
Living well in the sense of flourishing is something we do or should care
about. The fact that engaging in faith practices when one has ambiguous
evidence for God helps one to grow in virtue, and that this in turn makes it
likely that one will live better, gives one reason to engage in those practices.
It does this whether well-­living also contributes to “well-­being” or not, and
regardless of whether it contributes to living an enjoyable life, a life one is
satisfied with, or a life that ticks certain objective list boxes. Living well
might also promote some of these other things, and I will explore this possi-
bility to some extent in some of the chapters that follow. If it does, that may
provide one with even further reason to engage in the relevant practices.
But even if it doesn’t, there is already significant value in engaging in prac-
tices that help one grow in virtue and live a better life.
 PART T WO
PATHWAYS F ROM FA I T H
PR AC T IC E S TO F LOU R I SH I NG
 5
Giving God the Benefit of the Doubt

This is the first chapter to address a specific faith practice—­or, rather, set of
practices—­and to argue that engaging in these faith practices can enable
individuals with ambiguous evidence for God to grow in virtue and flour-
ishing. The focus here is on a set of faith practices that together involve giving
God the benefit of the doubt in different ways, and the traits of character to
which I will argue they are conducive are likewise traits that involve giving
other people more generally the benefit of the doubt in similar ways. In
Section 1, I explain the nature of the character traits that I will argue are
promoted by engaging in the faith practices. I offer accounts of each of the
traits, discussing relevant philosophical literature that can illuminate them,
illustrating their operation, and explaining how manifesting them can
incorporate the kinds of cognitive attitudes characteristic of faith discussed
in Chapter 3. Section 2 then argues that these character traits are virtuous
by showing how they are related to other commonly accepted virtues and
providing a deep story about their moral value. Section 3 describes the
practices of faith that involve giving God the benefit of the doubt. And
Section 4 explains how engaging in these practices of faith can promote the
virtuous traits of character described in the earlier sections, and how in
doing so these practices may also promote other aspects of a practitioner’s
well-­being.

1 Praisefulness, Thankfulness, and Contrition

There are various ways that the character traits which are my focus here can
be conceptualized as following a unified pattern or contributing to a unified
ideal.1 All of them can be thought of as tendencies to err in one way rather
than another—more specifically, to give other people the benefit of the
doubt in certain ways. And, as I will discuss in more detail in Section 2, the

1 This section and the next adapt with permission content from (Byerly 2022a).

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0006
118 Faith, Flourishing, and Agnosticism

way they involve giving others the benefit of the doubt tends to be conducive
toward cultivating valuable interpersonal relationships. So, we might think
of them as traits that involve giving others the benefit of the doubt for the
sake of relationship. Moreover, while the examples of these pro-relationship
character traits I will focus on in this section may at first strike some readers
as rather narrow features of character, I will explain in Section 2 how they
are related to broader features of character equally concerned with promot-
ing valuable interpersonal relationships. As I will argue there, these narrow
features of character have a fitting place within a broad, virtuously other-­
oriented character.
While there are many candidates for these kinds of traits, I will focus my
attention here on three that I call, respectively, “praisefulness,” “thankfulness,”
and “contrition”. I use these labels stipulatively rather than in an effort to
analyze some pre-­theoretic phenomena already generally recognized using
these terms. In speaking of “thankfulness,” for example, I should not be
understood as offering a rival conception of the trait of gratitude, which has
received extensive attention from philosophers and psychologists, though
thankfulness as I will conceive of it is closely related to gratitude as it is
commonly conceived, as I spell out further below.
Praisefulness is stipulatively defined as a tendency to err on the side of
giving credit to others for their accomplishments rather than refraining
from giving such credit. The praiseful person would rather give credit when
credit isn’t deserved than refrain from giving credit when credit is deserved.
They are more tolerant of erring by offering credit when it isn’t due than
they are of erring by failing to offer credit when it is due. The credit the
praiseful person tends toward giving, they tend toward giving sincerely.
Theirs isn’t a tendency to feign giving others’ credit for their accomplish-
ments but a tendency to sincerely give credit. Nor is theirs a tendency to
give others more praise than their accomplishments would merit but a ten-
dency to err on the side of giving others the praise their accomplishments
would merit—­if indeed they are accomplishments. The praiseful person
therefore tends to err on the side of sincerely giving others credit commen-
surate with their accomplishments rather than the side of refraining from
giving others credit commensurate with their accomplishments.
In the contemporary philosophical literature, the trait that is most closely
related to praisefulness so conceived is appreciation. The kind of apprecia-
tion most commonly discussed by philosophers is aesthetic appreciation—­
the appreciation of beauty (e.g., Budd 2002). But aesthetic appreciation is
just one kind of appreciation. Tony Manela (2016), for example, notes that
there are also cognitive, ethical, and prudential kinds of appreciation. While
 Giving God the Benefit of the Doubt 119

noting that “there is no consensus philosophical account of what apprecia-

tion is,” Manela suggests that what is common to all these forms of appreci-
ation is that they are each “a mode of valuing, that is, a certain kind of
response to something good” (289). Accordingly, Manela proposes that
appreciation includes both cognitive elements and affective elements.
Focusing on the case of prudential appreciation—­appreciation for the good
things in one’s life—­he proposes that “when I appreciate [such things], I do
more than just get right certain facts about the value of those things; in
addition I enjoy them as well” (289). Similarly, I propose that the praiseful
person errs on the side of adopting a stance toward others’ achievements
that includes both positive cognitive and positive affective elements. The
stance includes both a positive cognitive stance toward the achievements as
achievements and an appropriate positive valuing of those achievements as
such. Without adopting such a stance, their praise of others’ achievements
would not be sincere, as it would not express the attitude of appreciation
that sincere praise expresses. The praiseful person thus errs on the side of
adopting a positive cognitive and affective orientation toward others’
achievements rather than refraining from adopting such an orientation.
They err on the side of offering others sincere praise for their achievements
as an expression of appreciation of those achievements.
Thankfulness is structurally very similar to praisefulness. It is a tendency
to err on the side of giving thanks to others for the valuable things others
have done for one rather than refraining from giving such thanks. The
thankful person would rather give thanks when thanks isn’t deserved than
refrain from giving thanks when thanks is deserved. They are more tolerant
of erring by offering thanks when it isn’t due than they are of erring by fail-
ing to offer thanks when it is due. The thanks the thankful person tends
toward giving they tend toward giving sincerely. Theirs isn’t a tendency to
feign giving others’ thanks for their help but a tendency to sincerely give
thanks. Nor is theirs a tendency to give others more thanks than their help
would merit but a tendency to err on the side of giving others the thanks
their help would merit—­if indeed they have given such help. The thankful
person therefore tends to err on the side of sincerely giving others thanks
commensurate with the benefits they have given rather than the side of
refraining from sincerely giving others thanks commensurate with the ben-
efits they have given.
While thankfulness is structurally similar to praisefulness, it is not the
same trait, nor is thankfulness a subordinate species of praisefulness. To
recognize and value the excellence of someone’s performance in the way
characteristic of praisefulness is a different thing from thanking them for
120 Faith, Flourishing, and Agnosticism

the contribution this performance made to one’s own well-­being. Giving

thanks is not just what someone who has benefitted from another’s excellent
performance does when they recognize the excellence of that performance;
it is its own distinctive sort of activity. Giving thanks involves adopting an
orientation toward another as one’s benefactor and not just an orientation
toward an achievement of another.
There is a large philosophical literature on gratitude that is directly rele-
vant for informing our conception of thankfulness. Where the attitude
expressed by sincere praise for others’ accomplishments is appreciation of
those accomplishments, the attitude expressed by sincere thanks for the
benefits others have given to one is gratitude. According to a widespread
view, such gratitude is understood to have a “to–­for” structure (see Manela
2015: sect.1). The grateful person is grateful to their benefactor for the bene-
fit they have received from them. As with appreciation, philosophers gener-
ally agree that gratefulness includes both positive cognitive (e.g., Walker
1980) and positive affective (e.g., Bruton 2003) elements. Likewise, I pro-
pose that the thankful person tends to err on the side of adopting a stance
toward benefactors that includes positive cognitive and positive affective
elements—­chiefly, the positive cognitive recognition of benefits these bene-
factors have given them and of the benevolent intentions of the benefactors,
and the positive valuing of their benefactors as sources of these benefits.
The thankful person would rather offer sincere thanks that expresses such a
stance when such a stance is not merited than fail to offer sincere thanks
expressing such a stance when such a stance is merited.
Whereas praisefulness and thankfulness so conceived govern how a per-
son approaches certain positive behaviors of others, contrition2 governs
how a person approaches their own negative behaviors. The contrite person
errs on the side of apologizing for wrongs done to others and seeking their
forgiveness. They would rather apologize and seek forgiveness when an
apology and forgiveness are not warranted than fail to apologize and seek
forgiveness when apology and forgiveness are warranted. They are more
tolerant of erring by offering an apology when it is not warranted than they
are of erring by failing to offer an apology when it is warranted. The apolo-
gies they tend toward giving are sincere apologies, and apologies commen-
surate with the wrongs done. Thus, theirs is a tendency to err on the side of
giving sincere apologies commensurate with the wrongs they’ve done and

2 Perhaps a better term would be “contritefulness.” I’ve kept “contrition” here mainly
because I use this term in (Byerly 2022a) and didn’t want to introduce confusion with a different
term here.
 Giving God the Benefit of the Doubt 121

seeking forgiveness for these wrongs rather than the side of refraining from
giving sincere apologies commensurate with the wrongs they’ve done and
seeking forgiveness for these.
The philosophical literature most relevant to contrition so conceived is
the limited literature on apology and even more limited literature on
contrition.3 As with gratitude, apology has a “to–­for” structure; a person
apologizes to someone they have wronged for the wrong they’ve done.
Sincere apologies are typically regarded as including both cognitive and
affective elements. Radzik and Murphy write that “a well-­formed apology
requires at least acknowledgement of both the fact of wrongdoing and
responsibility by the wrongdoer, as well as an expression of regret or
remorse” (2015: sect.3.1). Here, the cognitive element is positive while the
affective element is negative. Yet, as Roberts (2007: 104–6) emphasizes in
his work on contrition, contrition is not the same as mere guilt, because in
displaying contrition one is hopeful about the possibility of forgiveness,
which involves positive affective elements as well and can even shade into
joy. Accordingly, I propose that the contrite person errs on the side of
adopting a stance toward their wrongdoing that involves a positive cogni-
tive recognition of this wrongdoing as such and a negative affective evalua-
tion of this wrongdoing as such, together with a hopeful attitude about the
possibility of being forgiven for their wrongdoing. The contrite person
would rather offer a sincere apology expressing such a stance when none is
called for than fail to offer such an apology when it is called for.
Before illustrating how praisefulness, thankfulness, and contrition oper-
ate in mundane circumstances, it is worth pausing to consider more pre-
cisely the cognitive commitments that may be involved in each trait. As we
have seen, each trait tends to manifest in ways that include adopting posi-
tive cognitive commitments—­commitments to someone’s having achieved
something valuable (praisefulness), to someone’s having benevolently bene-
fitted oneself (thankfulness), or to one’s having wronged someone else
­(contrition). It should come as a natural suggestion here, given the discussion
of doxastic and subdoxastic cognitive commitments offered in Chapter 3,
that these commitments might be supplied either by beliefs or by nondoxas-
tic assumptions. The kind of cognitive commitment needed in each of these
cases is the kind that we observed in Chapter 3 involves taking a stand
for the truth of the relevant claims. And this stand requires adopting a

3 I am aware of only two article-­length pieces in contemporary philosophical research

focused on contrition: (Flood 2021) and (Roberts 2007).
122 Faith, Flourishing, and Agnosticism

cognitive commitment such that if what one has committed to is false, the
commitment was erroneous. As we saw, it appears that only beliefs and
nondoxastic assumptions can fill this role. So, praisefulness, thankfulness,
and contrition should be understood to involve tendencies to err on the
side of believing or assuming that others have achieved valuable feats, have
benevolently benefitted oneself, or have been wronged by oneself.
Now that I have explained the basic nature of praisefulness, thankfulness,
and contrition, I wish to illustrate their operation by considering how they
may make a difference for a person’s behaviors in some mundane cases.
Imagine, for instance, that you are watching the final seconds of a basketball
game with a tie score. An offensive player gets the inbounds pass and drib-
bles down the middle of the lane. They’re swarmed by the defense, so much
so that you can hardly tell what’s happening. The crowd stands to its feet,
further obscuring your view. What you are able to see clearly, though, is the
ball popping up out of the crowded lane toward the basket, bouncing about,
and falling in.
Details about the case could be fleshed out in various ways. Let’s suppose,
though, that the offensive players had spread the floor, and so there was no
other offensive player in the lane. And let’s stipulate that your ability to see
the events was affected in just such a way that your evidence is ambiguous
regarding whether the player who drove into the lane deserved credit for
having made a winning shot. Your evidence neither strongly supports that
they deserve credit for making it nor strongly supports that they don’t,
though your evidence does strongly support that no one else deserves credit
for making the shot.
Our question is whether characterological features of the sort in view in
this section might make a difference for how you behave in this case. Here,
it is the first character trait of praisefulness that is relevant. Suppose that you
have a tendency to err on the side of giving others credit for their accom-
plishments. You’d prefer to give credit when it isn’t deserved than fail to give
credit when it is deserved. The ambiguity of your evidence in this case may
leave you otherwise on the fence about whether or not credit is deserved.
Yet the trait of praisefulness could indeed make a determinative difference
for what you do. If you are praiseful, you will tend to offer sincere praise to
the player for their having made the winning shot.
It is of course important to recall from Chapter 2 that there is a good deal
of variability in ways that a person’s evidence can be ambiguous. This vari-
ability in ways that evidence can be ambiguous can make a difference for
how strong a tendency of praisefulness is needed to make a difference in
 Giving God the Benefit of the Doubt 123

this kind of case. For instance, suppose that your evidence in the case is
ambiguous because it either weakly supports the player’s deserving credit
for making the shot or it is exactly counterbalanced regarding whether the
player deserves this credit. In that case, a weak tendency to err on the side
of giving praise can make a determinative difference for whether you give
them credit. Alternatively, if your evidence is ambiguous because it weakly
supports that the player does not deserve credit, or because it strongly sup-
ports that it supports that the player does not deserve credit, then a stronger
tendency to err on the side of praise will be needed to lead to your praising.
This then raises an important question about how strong a tendency to err
on the side of praising is characteristic of praisefulness.
My suggested answer is a relatively simple one. For any potentially praise-
worthy accomplishment, the strength of the tendency to err on the side of
offering praise that is characteristic of the trait of praisefulness will be pro-
portional to the value of the accomplishment. The praiseful person will tend
to err more strongly on the side of giving praise for more valuable accom-
plishments and will tend to err less strongly on the side of giving praise for
less valuable accomplishments. For instance, a praiseful person might not
be inclined by their praisefulness to offer praise if the above basket was
made in a low-­stakes situation in the second quarter and their evidence
weakly supported that the player did not deserve credit. But given the same
evidence, their praisefulness may indeed incline them toward offering
praise if the shot was made to break a tie at the final buzzer.
Structurally similar cases illustrate how thankfulness and contrition can
make a parallel difference for a person’s thanking and apologizing behav-
iors. A thankful person whose evidence is ambiguous regarding whether
someone else has benefitted them will tend toward sincerely thanking them
for the benefit and will tend more strongly toward doing so when the bene-
fit and benevolent intentions would be of greater value. Perhaps the sup-
posed benefactor was aiming to give the benefit undetected but couldn’t
avoid leaving just enough evidence for the beneficiary’s evidence to be
ambiguous regarding whether they had given the benefit. Here, the thankful
person would rather err on the side of offering sincere thanks to the sup-
posed benefactor than on the side of refraining from giving such thanks.
Similarly, the contrite person whose evidence is ambiguous regarding
whether they have wronged another person will tend to sincerely apologize
to this other, preferring to sincerely apologize when no apology is necessary
than to fail to apologize when an apology is necessary. Here, we might
imagine that you’ve been hashing over the details of whether you have
124 Faith, Flourishing, and Agnosticism

wronged your partner for some time and have reached the point that your
evidence that you’ve wronged them is ambiguous in one way or another.
Acting in accordance with contrition will push you toward offering a sin-
cere apology, and it will push you more strongly toward doing so depending
on the severity of the transgression.
The aim of the foregoing discussion has been to illustrate how the charac-
ter traits of praisefulness, thankfulness, and contrition can make a determi-
native difference for one’s behaviors in mundane cases. But it is important
to note that these character traits needn’t make the sorts of differences cited
in every such case. In particular, sometimes there will simply be something
more important to do than offer the relevant praise, thanks, or apology. For
example, we might imagine in the basketball case that at the moment the
shot goes through, you receive a phone call notifying you that your child
has just been injured and is at the hospital. Given the relative significance of
this turn of events and the time-­sensitive nature of it, you may quickly exit
the stadium without so much as a sincere cheer in order to quickly make
your way to the hospital—­and this remains the case even if you are a praise-
ful person. Your praisefulness leads you to err on the side of offering sincere
praise rather than refraining from doing so, but only other things being
equal, and here other things are not equal. The same kind of ceteris paribus
clause applies to the tendencies of thankfulness and contrition. Sometimes,
for example, we might have evidence that if the praise we offer is not mer-
ited this may lead the recipient to feel shame, or that if the apology we offer
is not merited this could supply misleading evidence to the recipient of the
apology; in these cases, other things are not equal.
Let the foregoing suffice for an explanation of the nature of the traits that
I will argue are promoted by the faith practices of giving God the benefit of
the doubt that are my focus in this chapter. The next step is to argue that
these traits are virtuous—­that they are either themselves virtues or that by
promoting them one promotes closely related virtues.

2 The Virtuousness of Praisefulness, Thankfulness,

and Contrition

It is worth starting with the observation that, for some readers, little argu-
ment may be needed for the conclusion that the traits of praisefulness,
thankfulness, and contrition are virtuous. Some readers may simply, upon
understanding the nature of the traits, be inclined to think that these traits
 Giving God the Benefit of the Doubt 125

are human excellences. They might be inclined to judge that these are traits
they wished their colleagues had, or that their children will come to possess
one day. Such intuitions regarding the virtuousness of these traits may be as
persuasive as any philosophical argument could be for leading readers to
the conclusion that these traits are virtuous. Thinking in that way makes
perfect sense, in my opinion.4
Another approach to defending the value of these traits, which will help
us to uncover a deep account of their value, involves attending to other
traits that have been regarded as virtues and arguing that if these latter traits
are virtues, then so are praisefulness, thankfulness, and contrition. Or, if the
latter are not themselves virtues, they are at least related to other virtues in
such a way that by promoting them one promotes virtues to which they are
closely related. In either case, we could sensibly regard praisefulness, thank-
fulness, and contrition as “virtuous” tendencies.
Several philosophers have been attracted to the idea that it is morally
excellent to give others a certain kind of benefit of the doubt—­to err, we
might say, on the side of viewing others more positively or charitably. Susan
Wolf, for example, in her classic essay on moral saints, proposes that a moral
saint “should try to look for the best in people” and “give them the benefit of
the doubt as long as possible” (1982: 422). Similarly, Ryan Preston-­Roedder
defends the value of a virtue he calls “faith in humanity” at length, where
this virtue involves both a cognitive element and a volitional element. Of
the cognitive element, he writes that “when someone who has faith in
humanity morally evaluates other people’s actions, motives, or characters,
she tends to give them the benefit of the doubt.” Moreover, she tends to
“believe in people, trust in them, make presumptions in their favor, or see
them in a favorable light, morally speaking” (2013: 666). Michael Pace like-
wise writes that “thinking charitably of others, may in fact be a prima facie
moral obligation regarding evidential standards that one has to every-
one. . . . Other things being equal, adjusting one’s standards to give people
the benefit of the doubt seems to be a moral good that flows from the good
of treating others with respect” (2011: 258–59).

4 Nor am I the only one to hold such an opinion. At least, I am not the only one to embrace
the basic methodological point relied upon here, that a person’s intuitions regarding which
traits they would want their children to have can provide them with as much reason to believe
a trait is a virtue as any philosophical argument could. Cf. (Hursthouse 1999: chs. 8 and 9). To
embrace the application of this point in the present case we needn’t embrace the broader and
more dubious view that all intuitions are equally epistemically probative.
126 Faith, Flourishing, and Agnosticism

In each of these cases, we find a proposal that a tendency to err on the

side of viewing others favorably is a central component (or perhaps the
whole) of some virtue or other—­whether that virtue is the virtue of giving
the benefit of the doubt, the virtue of faith, or the virtue of respect. What is
important to note for our context is that if any of these traits is indeed a
virtue and centrally includes the disposition to err on the side of viewing
others favorably, then this provides reason for thinking that the traits
of praisefulness and thankfulness are virtuous. For these traits, too, are
­centrally constituted by dispositions to err on the side of viewing others
favorably. It is just that these other-­favoring dispositions are restricted to
particular domains—­namely, domains in which it is others’ achievements
or acts of beneficence that are up for consideration—­and they are accompa-
nied by fitting favorable evaluative stances, where not all of the virtues
­identified above require such evaluative stances in addition to favorable
cognitive stances. We might say, then, following Daniel Russell (2009: ch.7),
that praisefulness and thankfulness as defined here are “unique specifications”
of more cardinal virtues such as faith or giving the benefit of the doubt or
respect, in much the way that magnificence is a unique specification of gen-
erosity. They are what giving the benefit of the doubt or showing respect or
demonstrating faith is like when specialized to circumstances focused on
evaluating others’ accomplishments or thankworthiness.
Moreover, we can locate a plausible deep story about the moral value of
not only praisefulness and thankfulness but contrition as well by attending
to why it is that these traits involving erring on the side of viewing others
favorably are valuable. A key idea appealed to by several authors who have
written in favor of giving others the benefit of the doubt in one way or
another has been the following. When we err on the side of giving someone
the benefit of the doubt, we thereby enhance the expected value of personal
relationship goods we will enjoy with this other over what we would enjoy if
we instead erred on the side of refraining from giving them the benefit of
the doubt.
Preston-­Roedder implicitly relies upon this kind of point in his defense
of the value of having faith in humanity. He writes that faith in humanity
“partly constitutes a certain morally important relation, namely, a kind of
harmony or solidarity, between the virtuous person and other members of
the moral community” (2013: 676). “Having faith in people’s decency,” he
continues, “despite reasons for doubt, is a way of standing by them.” By put-
ting her faith in others, the faithful person “ties her own flourishing, in
­certain respects, to the quality of these people’s characters and actions” (683).
 Giving God the Benefit of the Doubt 127

This tying of one’s flourishing to others is a significant personal relationship

good, and it is one that is promoted in greater measure when one errs on
the side of adopting the positive cognitive and volitional stances toward
others that Preston-­Roedder has in view than when one errs on the side of
not adopting such stances. Having faith in humanity, Preston-­Roedder
emphasizes, is one way in which “a morally virtuous person escapes her sol-
itude and enters into [a valuable] form of community” (684).
A similar instance of this pattern of argument can be found in the grow-
ing body of literature on epistemic partiality in friendship (see, e.g., Keller
2004; Stroud 2006).5 The driving thought behind the central problem in this
literature is the thought that the value of a maintained friendship is better
advanced by erring on the side of viewing one’s friends favorably than by
not doing so. Writing about cases in which one friend, S, tells another, A,
that p, thereby inviting A to trust S, Goldberg says that “A risks jeopardizing
the friendship in any case in which it is true both that S is worthy of A’s trust
and that A fails to trust S.” By contrast, “the case in which S is not worthy of
A’s trust but A trusts anyway is not one in which A damages the friendship”
(2019: 8). Other things being equal, it will follow from this observation that
erring on the side of trusting rather than not trusting one’s friends better
promotes a maintained friendship with them, and it is this fact that has
seemed to some authors to provide moral reason for friends to err on the
side of trusting one another—­even if this involves epistemic irrationality.
I propose here a similar deep defense of the virtuousness of praisefulness,
thankfulness, and contrition. By erring on the side of giving others credit
commensurate with their achievements, thanks commensurate with the
benefits they have given us, and apologies commensurate with wrongs we
have done to them, we enhance the expected value of personal relationship
goods with these others over what we would achieve if we erred on the side
of refraining from praising, thanking, or apologizing to them. For, just as in
trusting a trustworthy friend one cultivates this friendship but in trusting
an untrustworthy friend one does not comparably harm the friendship,
likewise in praising a praiseworthy person and thanking a thankworthy
person and apologizing to one’s victims one cultivates significant personal
relationship goods, but in praising an unpraiseworthy person, thanking an
unthankworthy person, or apologizing to someone one has not wronged,

5 Wynn (1997) offers an earlier discussion of a similar idea regarding values obtainable in
“trust relationships,” defending its significance in supporting the moral value of theistic belief
in a way that parallels some of the argumentation of this chapter.
128 Faith, Flourishing, and Agnosticism

one does not typically comparably damage these personal relationship

goods. Similarly, just as one does greater harm to a friendship by failing to
trust a trustworthy friend than by trusting an untrustworthy friend, like-
wise one does more harm to a relationship by failing to praise, thank, or
apologize to one who deserves it than by offering praise, thanks, or apology
to one who doesn’t deserve it. To harken back to the language we used in
describing virtues in the previous chapter, praisefulness, thankfulness, and
contrition are valuable ways of valuing something valuable—­personal rela-
tionship goods. As such, there is available a powerful defense of their
virtuousness.
At this point it will be helpful to stress the significant value of the per-
sonal relationship goods promoted by praisefulness, thankfulness, and con-
trition. The importance of personal relationship goods for human well-­being
is stressed in growing bodies of both philosophical and psychological litera-
ture. In philosophy, several growing strands of research in recent decades
have focused on personal relationship goods, including research in care
ethics and research on associative duties. Where relationship goods in gen-
eral are conceived as “those goods of constitutive (as well as, often, instru-
mental) value that accrue to individuals in virtue of them being in
relationships with other people,” personal relationship goods are such goods
that “accrue to individuals in virtue of them being in relationships that
involve some kind of direct, personalized interaction” (Gheaus 2018: sect.1).
Examples include “companionship, affection, intimacy, attachment, love,
friendship, empathy, social respect, solidarity, trust. . . attention, sympathy,
encouragement, [and] acceptance” (sect.1) among others. Philosophers
have been in broad agreement that such goods “generate weighty reasons for
action” (sect.1) and “represent a significant and non-­substitutable compo-
nent of individuals’ well-­being, are a significant kind of personal resource as
well as a major determinant of individuals’ opportunities” (Gheaus 2018:
Introduction).
It is commonly affirmed that these personal relationship goods are non-
instrumentally valuable (e.g., Seglow 2013) and are constitutive of good
lives (e.g., Lynch et al. 2009). They are, in addition, indispensable to subjec-
tive life satisfaction (Vaillant 2012) and instrumentally valuable as sources
of self-­confidence, self-­respect, and self-­esteem (Honneth 1995), as well as
even mental, emotional, and physical health (Brownlee 2016). The value of
having at least some minimal level of such goods is dramatically illustrated
by the devastation that ensues when people are bereft of such goods for
extended periods. Brownlee argues that there is a right against social
 Giving God the Benefit of the Doubt 129

deprivation on the basis that chronic lack of adequate social contact ­“generates
the same threat response as pain, thirst, hunger, or fear by setting off a chain
of anxiety-­inducing physiological reactions known as the ‘fight or flight’
response” (2013: 211). She writes, further, that “when we are deprived of
adequate social connections. . . we tend to break down mentally, emotion-
ally, and physically” (2016: 55). The value of personal relationship goods is
such that justice in the distribution of these resources has recently become a
major topic of philosophical debate, with several philosophers defending
the existence of various rights to personal relationship goods such as
­adequate social contact or even love (Liao 2015), and others defending the
ex­ist­ence of duties to cultivate personal relationship goods such as friendship
(Collins 2013).
Recent psychological literature, some of which is relied upon in the phil-
osophical research just surveyed, also provides confirmation of both the
value of personal relationship goods and the way in which a concern for
these goods unifies the dispositions of gratitude, appreciation, and apology.
Even nearly 30 years ago, at a time when it was relatively unfashionable for
psychologists to argue in favor of the existence of basic psychological needs,
Baumeister and Leary (1995) nonetheless found the evidence in favor of a
basic need to belong so widespread and powerful that they published a sem-
inal article on the topic. Now the need to belong, or for belongingness, is
commonly recognized in psychological research. The need to belong is a
“pervasive drive to form and maintain at least a minimum quantity of last-
ing, positive, and significant interpersonal relationships” in which “frequent,
affectively pleasant interactions” take place in a context of a “temporally sta-
ble and enduring framework of affective concern for each other’s welfare”
(497). Given the existence of such a need, it is hypothesized and confirmed
that “real, potential, or imagined changes in one’s belongingness status will
produce emotional responses, with positive affect linked to increases in
belongingness and negative affect linked to decreases in it” (497). Moreover,
as reflected above in recent philosophical literature, the absence of adequate
relationships will be detrimental toward mental, emotional, and physical
health, while their presence will predict better health as well as life satisfac-
tion. While not every person will be equally motivated to cultivate a positive
personal relationship with every other person, “rejecting social attachment
goes against some deeply rooted aspect of human nature” (520), and when
one experiences such rejection, “as in unrequited love, the result is typically
distress and disappointment” (505). Personal relationship goods are here
seen to serve an indispensable role in fulfilling a basic psychological need.
130 Faith, Flourishing, and Agnosticism

And just in case you might be tempted to think I am overstating the case
for the value of personal relationship goods, it is important to observe that
there is research that reveals that human beings tend to systematically
underestimate the value of personal relationship goods to their own detri-
ment. For instance, we misjudge the impact that small acts of kindness will
have on both others and ourselves (Kumar and Epley 2022). We likewise think
that speaking kindly to a stranger won’t matter to us—­but it does (Epley and
Schroeder 2014). We mistakenly think that spending some unearned money
on ourselves will make us happier than spending it on others, when the reverse
is true (Dunn et al. 2008). And the reason it’s true is precisely that in spending
money on others, or benefitting others rather than ourselves more broadly,
we do more to satisfy our needs for relatedness by promoting personal
relationship goods (Titova and Sheldon 2022). Indeed, in some of my own
empirical research (Byerly et al. 2022), my colleagues and I have found that
people who tend to prioritize others’ interests over their own because they
value personal relationship goods tend to be more satisfied with their lives, to
experience greater meaning in life, and to endure less stress and cope better
with it. Tendencies to value personal relationship goods well are very im­por­
tant for human well-­being and are a sorely needed corrective to our tendencies
to devalue these goods.
Let me conclude this section by briefly restating the basic case here
offered in favor of the virtuousness of praisefulness, thankfulness, and con-
trition. These traits all involve a particular sensitivity to personal relation-
ship goods that are better promoted through erring on the side of praise,
thanks, and apology than through erring on the side of avoiding these. The
sensitivity to personal relationship goods they involve seems to be a good
way of valuing these goods given the significant role that these goods and
caring for these goods plays in human life. There are different ways we could
understand the virtuousness of these traits. We might regard them as con-
stituting distinctive virtues in their own right. Or we might regard them as
unique specifications of more cardinal virtues to give the benefit of the
doubt, to show faith in humanity, or to show others respect. Still another
possibility is that we might simply regard them as a kind of upper limit on
virtuous appreciation, gratitude, and contrition. They are tendencies dis-
played by people who are ideally appreciative, grateful, and contrite.
Whichever of these accounts of their place in the life of the overall virtuous
person we find most attractive, it will remain true that by fostering these
traits in oneself, one will grow in virtue.
 Giving God the Benefit of the Doubt 131

3 Giving God the Benefit of the Doubt

Now that we have a clear idea of what praisefulness, thankfulness, and con-
trition are and why they should be regarded as virtuous traits of character, it
is time to describe the practices of faith that a person with ambiguous evi-
dence can undertake that can promote growth in these tendencies. These
practices can, in fact, be described fairly simply—­they just involve offering
praise, thanks, and contrition to God. A person can give thanks to God for
the many goods of their life, which find their ultimate benevolent source in
God. They can praise God for God’s many valuable achievements, which
include God’s foundational explanatory work in generating spatiotemporal
reality and God’s benevolently benefitting all other people with the goods of
their lives. And a person can apologize to God for wronging those God
loves, if not for more directly wronging God by having previously failed to
display adequate gratitude toward God or having previously treated God’s
love for them coldly.
If minimal theism is true, these simple practices of praising, thanking,
and demonstrating contrition to God would be perfectly fitting. Minimal
theism would imply that God is the ultimate personal source of the uni-
verse, that God has benevolently secured each good in each person’s life and
loves each person as much as anyone does. But these simple implications of
minimal theism are adequate to make these acts of gratitude, praise, and
apology fitting. A person who has exercised their agency in generating the
universe and benevolently securing benefits for all people is praiseworthy for
this. A person who has benevolently bestowed the goods of one’s life is
thankworthy. And a person who loves those one has wronged as much as
anyone does deserves apology for these wrongs. They also may deserve apol-
ogy if their love and beneficence toward oneself has not been adequately
well-­regarded in the past and if their good gifts have been misused for ill.
By engaging in these simple acts of gratitude, praise, and contrition, a
person with ambiguous evidence for God would be giving God the benefit
of the doubt in much the way described above. Given that someone has
ambiguous evidence for the God of minimal theism, they will thereby have
ambiguous evidence that God is praiseworthy, thankworthy, and apology-
worthy in the ways just described. When a person with ambiguous evidence
then gives God fitting thanks or praise or apology for these things for which
God would be praiseworthy, thankworthy, or apologyworthy if God exists, they
give God the benefit of the doubt in much the same way that we saw at work
132 Faith, Flourishing, and Agnosticism

in the mundane illustrative examples from Section 1. They err on the side of
giving God fitting praise, thanks, and apology.
The case of God is, of course, a somewhat unique one. In this case, a person’s
evidence for thinking that another person has done something worthy of
praise or thanks or has been wronged by them is just as strong as their
­evidence for thinking this person exists. Insofar as they have evidence for
thinking this person, God, exists at all, they also have evidence regarding
God’s praiseworthiness, thankworthiness, and apologyworthiness. This is
not typical of our evidential situation with fellow human beings. Typically,
with our fellow human beings, our evidence that they exist is much stronger
than our evidence that they are praiseworthy, thankworthy, or apologywor-
thy. Yet this difference should not make a difference for whether the acts of
praising, thanking, and apologizing in this case would involve erring on the
side of offering fitting praise, thanks, and apology. What matters for that is
only that one is offering praise, thanks, or apology when one has ambiguous
evidence of its fittingness.
Moreover, while the God case differs from typical cases involving our
fellow human beings, it is worth noting that there are atypical cases involv-
ing our fellow human beings that are closer to the God case. For example,
imagine a child attending to a parent on their deathbed. In some such cases,
depending upon the child’s beliefs about an afterlife, the medical facts about
the parent, and the parent’s treatment of the child during their life, the
child’s evidence for thinking their parent exists may be about as good as
their evidence for thinking this parent is praiseworthy, thankworthy, or
apologyworthy. Insofar as they have evidence for thinking their parent is
still with them, they likewise have evidence for thinking that someone is
with them who is praiseworthy, thankworthy, or apologyworthy. If we
imagine their evidence for thinking their parent is still with them is ambig-
uous, then this is a sort of case, a bit closer to the God case, where offering
praise, thanks, or apology would involve erring on the side of so doing.
Indeed, a tendency to err on the side of offering praise, thanks, or apology
may help to explain the extraordinary commonality of continued commu-
nication with the deceased relatives, even among individuals who do not
explicitly espouse belief in an afterlife (Steffen and Klass 2018).6

6 We can imagine additional cases even closer to the God case. For example, imagine that a
late adolescent who has spent most of their life in the foster system has recently discovered
ambiguous evidence of the existence of someone who showed them great love and care when
they were very young. Their evidence for thinking this person exists may be about as good as
their evidence for thinking this person is thankworthy.
 Giving God the Benefit of the Doubt 133

Giving God the benefit of the doubt in these ways can involve manifesta-
tions of faith as understood in Chapter 3. As seen above in the discussion of
praisefulness, thankfulness, and contrition, someone who sincerely praises,
thanks, and apologizes must adopt certain evaluative, conative, and cogni-
tive attitudes. They evaluate God’s achievements and benevolent benefices
as good and are motivated to acknowledge their value, while they evaluate
their own wrongdoing as bad and are motivated to apologize for it, while
also evaluating God’s love for them as indicative of a possibility of forgive-
ness, which they are motivated to seek. They also adopt some kind of posi-
tive cognitive stance toward God’s being praiseworthy, thankworthy, and
apologyworthy. This might involve beliefs or assumptions about these top-
ics. In either case, it will include a cognitive commitment characteristic
of faith.
While thanking, praising, and apologizing to God are acts, intentionally
engaging in these acts regularly constitutes a practice. What counts as
“regularly” engaging in these acts may be vague, but an example might be
engaging in these acts monthly, weekly, daily, or multiple times per day.
Someone who engaged in these acts this often would seem to qualify as per-
forming them enough as to count as regularly engaging in them so as to be
engaged in practices of faith toward God. It is noteworthy that engaging in
these acts of praise, thanks, and apology needn’t be particularly time con-
suming. Many people attempt to engage in these practices in about as wide
of a variety of settings as tend to characterize human life. At just about any
time and in just about any circumstances, one can engage in these practices,
and attempting to engage in them in this kind of a way seems about as good
a means of trying to achieve their communicative aims as any. Also, in con-
trast to the human case is the fact that, given God’s loving nature, there are
fewer grounds for anxiety or doubt about one’s attempts being well-­received
if God indeed is there to receive them. Even the kinds of worries that too
often wrongly lead us to miss out on erring on the side of offering praise,
thanks, or apology in the human case are less of a concern with God. In
summary, the opportunity costs of these practices of faith are low.
At the same time, the absolute value of what one gives praise, thanks, and
apology for in this case is quite high. Every good in one’s life is one for
which God is properly praised and thanked, and similarly every good in
others’ lives is one for which God is properly praised. So, the praise and
thanks fitting for God if God exists is constantly growing insofar as one’s
own life and the lives of others are good. Since there is no one else who
stands in relation to the goods of one’s own life and the lives of others in this
134 Faith, Flourishing, and Agnosticism

same way, it is plausible that God is supremely deserving of praise and

thanks. And if God is supremely deserving of one’s praise and thanks but
one has failed to give thanks and praise to God, this is a significant wrong.
Moreover, if it is true that whenever one wrongs anyone, one also wrongs
those who love that person, then the debt of apology one owes to God is
constantly growing insofar as one commits any wrongs. Furthermore,
whenever one wrongs anyone, one will have used for ill goods of one’s own
life benevolently bestowed by God, which is wrong to do. As there is no one
else who stands to all of one’s wrongdoing in this same way, it is plausible
here, too, that God will be supremely deserving of apology.
The practices of faith that are our focus here, then, are practices of thank-
ing, praising, or displaying contrition toward God despite having ambigu-
ous evidence for God’s being thankworthy, praiseworthy, or apologyworthy.
The practices incorporate the characteristic evaluative, conative, and cogni-
tive attitudes involved in faith practices. And engaging in these acts of faith
toward God with some regularity constitutes engaging in these acts as prac-
tices of faith. Even if performed regularly, these practices come with low
opportunity cost and involve responding to particularly important goods if
God exists.

4 How Giving God the Benefit of the Doubt Promotes

Giving Others the Benefit of the Doubt

There is one final step left to defend the main argument of this chapter. Thus
far, we have identified a suite of character traits that involve giving other
people generally the benefit of the doubt, and we have argued that these
character traits are virtuous. We have also identified faith practices that
involve giving God the benefit of the doubt. The final step is to explain why,
for individuals with ambiguous evidence for God, engaging in the faith
practices of giving God the benefit of the doubt is likely to promote growth
in the character traits that involve giving other people generally the benefit
of the doubt.
The first and most obvious thing to say about this involves appealing to
the simple and well-­trodden Aristotelian idea that we grow in the virtues by
practicing the acts that are characteristic of them (see Nicomachean Ethics
II.1). Just as one grows toward being a kind person by practicing perform-
ing kind actions, one grows toward being the kind of person who errs on
the side of giving others fitting praise, thanks, or apology by practicing
 Giving God the Benefit of the Doubt 135

erring on the side of giving others fitting praise, thanks, or apology. One
way one can practice this, if one’s evidence for God is ambiguous, is by err-
ing on the side of giving God fitting praise, thanks, or apology. Indeed, err-
ing on the side of giving God praise, thanks, or apology is what would be
characteristic of general tendencies to give others the benefit of the doubt if
one has ambiguous evidence for God. Failing to give God the benefit of the
doubt in these ways when one has ambiguous evidence for God would be
uncharacteristic of general tendencies to give others the benefit of the doubt
and would work to undermine one’s possession of such tendencies. As
Russell expresses Aristotle’s view about virtue cultivation, “character is what
one has as a result of building up of customary and familiar ways of acting”
(2015: 24). As such, the Aristotelian view should lead us to expect that
someone who builds up as customary ways of acting that involve giving
God the benefit of the doubt should be moved thereby toward developing a
character that involves giving others the benefit of the doubt more generally.
We might supplement this general Aristotelian point by observing that
the case of God may provide an especially useful one for practicing virtues
of giving others the benefit of the doubt. This is because, as observed earlier,
this is a case where giving the other the benefit of the doubt is plausibly less
difficult than many other cases while also being more important. It is less
difficult because the kinds of activities through which we seem to be able to
give God the benefit of the doubt are ones that can be practiced in a wide
variety of circumstances and with little fear that the acts will be poorly
received. It is particularly important because of the absolute value of the
goods for which we give praise, thanks, or apology in this case. Yet, for
many kinds of improvement through practice, it is precisely these kinds of
cases that are particularly useful in developing skill through practice, espe-
cially at earlier stages of development. We start with cases where it is easier
to display the skill and move to harder cases. We start with cases that if we
fail to get right are more costly and move to cases that if we fail to get right
are less costly. The case of giving God the benefit of the doubt, then, is prob-
ably a case where those who are not very far along in mastering the virtues
of giving others the benefit of the doubt stand to gain considerably, while
for those who are closer to mastering the virtue already, it would be particu-
larly uncharacteristic of them to fail to act in accordance with it in this case
and so may pose a particular threat to their virtue if they fail to do so.
Some readers may wonder, however, whether practicing giving God
the benefit of the doubt might not promote giving others the benefit of
the doubt more generally but may only promote a tendency to give God the
136 Faith, Flourishing, and Agnosticism

benefit of the doubt. One important source of this kind of concern is

research appealed to by philosophical situationists who argue that people
generally do not have cross-­ situationally consistent traits of the kind
required by Aristotelian virtues but have only locally consistent traits, such
as “rock-­climbing courage” (Waggoner et al. 2022: 14) rather than courage
simpliciter. Yet, several philosophers (e.g., Adams 2006; Annas 2011; Snow
2009) have contended that growth in more local character traits can pro-
mote growth in more global character traits. As Nancy Snow puts it in an
early defense of this view, although “our virtues might start out being local,
they need not remain so” (2009: 27). More recently, Snow (2016) has aimed
to show how psychological research supports this idea. Basically (and
removing some technical jargon), the thought is that by developing patterns
characteristic of virtue in some local set of circumstances, a person builds
mental resources that make acting similarly in similar circumstances more
likely. The process whereby these resources are accessed to guide conduct in
novel though similar situations may be spurred on by reflection on the
similarities in the circumstances (Kamtekar 2016: 192).
In the case in view here, it may seem that this process is particularly apt
to occur, because we are talking about situations that are more similar to
one another than what is found for other cases where cross-­situational con-
sistency is more challenging to develop. What is being claimed here is just
that by practicing giving God the benefit of the doubt through praising,
thanking, or apologizing to God when one has ambiguous evidence of
God’s being praiseworthy, thankworthy, or apologyworthy, one makes it
more likely that one will develop tendencies to give other people more gen-
erally the benefit of the doubt by praising, thanking, or apologizing to them
when one has ambiguous evidence of their praiseworthiness, thankworthi-
ness, or apologyworthiness. This is a rather different case from, say, expect-
ing someone who displays honesty through not cheating on a test to also
display honesty through not lying or not stealing. It is more analogous to
expecting that someone who displays honesty by not cheating on one type
of test is likely to display honesty by not cheating on another type of test. Yet
seminal research on students’ honesty, which is still regarded by some
authors as “the best evidence” (Kamtekar 2016: 190) for the situationist con-
cern in view here, shows that the latter sort of cross-­situational consistency
is much more expectable than the former (Hartshorne and May 1928). So,
we might expect that a tendency to give God the benefit of the doubt in the
ways highlighted here is not so unlikely to promote giving others the benefit
of the doubt in similar ways.
 Giving God the Benefit of the Doubt 137

There is, in fact, some more specific empirical support for this idea.
Researchers have observed that religiosity is related both to greater generic
gratitude (Emmons 2005) and greater tendencies to seek forgiveness from
other human persons when wronging them (Toussaint and Williams 2008).
In the case of gratitude, researchers have found strong support for the
­possibility of religious individuals’ tendencies to display gratitude to God
specifically accounting for why they are more grateful in general than their
nonreligious counterparts and have found strong support for this link.
Being more grateful to God specifically appears to be causally related to
being more grateful to people in general, and it thereby influences other
variables, such as individuals’ satisfaction with their lives (Aghababaei et al.
2018; Rosmarin et al. 2011). A similar pattern is detectable in the case of
seeking forgiveness from other people and experiencing God’s forgiveness.
The best evidence I am aware of in this case comes from studies in which
religious participants are induced to experience God’s gracious forgiveness
for their wrongdoing, where this has been found to promote their experi-
ence of guilt for their wrongdoing and intentions to make amends to the
human persons they have wronged (Bassett et al. 2020). Of course, neither
of these bodies of research focuses specifically on being thankful to God or
showing contrition to God in a way that involves giving God the benefit of
the doubt and whether this prompts isomorphic behaviors toward human
beings. But the fact that there do appear to be causal pathways from show-
ing gratitude or contrition to God to showing gratitude or contrition toward
other humans is surely suggestive that giving God the benefit of the doubt
through thanksgiving or contrition would be causally linked to giving
­others the benefit of the doubt in similar ways.
In summary, there is good reason to think that by practicing giving God
the benefit of the doubt a person is likely to promote more general tenden-
cies to give other people the benefit of the doubt in similar ways. This is con-
firmed by the foundational Aristotelian idea that virtues are acquired
through practicing the acts characteristic of them, as well as by research that
has addressed the mechanisms whereby more local virtuous features may be
expanded into more global virtuous features and research concerned specifi-
cally with the relationships between displaying gratitude and contrition
toward God and displaying the same toward human persons. By giving God
the benefit of the doubt through praising, thanking, or apologizing to God
when having ambiguous evidence of God’s praiseworthiness, thankworthi-
ness, or apologyworthiness, a person is likely to grow in virtuous tendencies
to give other people more generally the benefit of the doubt in similar ways.
 6
Accepting God’s Love

This chapter explores a second potential route whereby a practice of faith

may enable those with ambiguous evidence for God to grow in virtue.1
Whereas the previous chapter focused on a route whereby certain specific
virtues are cultivated directly, this chapter focuses on a route whereby a
wider array of virtues can be developed indirectly. Specifically, the conten-
tion is that by adopting a richly accepting orientation toward God’s love,
those with ambiguous evidence for God can experience better mental
health and thereby neutralize obstacles to their development or retention of
virtues.
In Section 1, I identify some of the central features involved on the
human side in adopting a richly accepting orientation toward God’s love.
Section 2 then builds an initial conceptual and empirical argument for the
conclusion that accepting God’s love can enhance a person’s mental health
and can indirectly enable a person to cultivate or maintain moral virtues,
focusing on research on attachment to God among theists. Section 3 extends
this argument by presenting some of my own empirical research on God
attachment and acceptance of God’s love among agnostics. Section 4
responds to three objections to the arguments developed, further highlight-
ing the empirical support for the unique contribution that acceptance of
God’s love has for agnostics and pointing forward to fruitful avenues for
future research on the topic.

1 Accepting God’s Love

The God of minimal theism is understood to be at least fairly loving. This

God, if they exist, has intentionally brought about each good in each human
person’s life with benevolent intent. Moreover, they love each human person
at least as much as any human person loves any other human person,

1 The chapter borrows with permission from (Byerly 2022b).

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0007
 Accepting God ’ s Love 139

willing both the good for each human person and willing relational union
with that person fitting for their relationship to them. Of course, stronger
claims are sometimes made about the love of God in specific theistic tradi-
tions. And minimal theism is compatible with stronger claims being true of
God’s love. The God of minimal theism loves each person at least in the
ways described but may love each person even more than this. Perhaps this
God even is love in some sense, as some more specific theories of God’s
nature have suggested. Yet even the fairly minimal claims about God’s love
that are definitive of minimal theism may be adequate to undergird an
argument for the indirect value of accepting God’s love for growing in virtue.
My focus here is on the human activity of accepting God’s love.2 It is one
thing for God to love a person in the ways specified and another thing for
that person to accept God’s love. Instead of accepting God’s love, a person
could be unaware of God’s love; ignore or reject it; or misconstrue, resist, or
doubt it. Here I wish to offer an explanation of some of the chief elements
involved in adopting a certain richly accepting orientation toward accepting
God’s love. In subsequent sections, I will argue that adopting a richly
accepting orientation toward God’s love is indirectly conducive toward vir-
tue development for individuals with ambiguous evidence for God.
At first glance, it may be tempting to think of accepting God’s love as
requiring that God’s love exists. The language of “acceptance” sounds fac-
tive: you can’t accept something if it’s not there to be accepted. While this
observation may be correct about how the language of “acceptance” is typi-
cally used, that’s not how I will be using the language here. Instead, I will be
using it to refer just to what happens on the human side in accepting God’s
love. Accepting God’s love in this sense involves a pattern of attitudes and
behaviors directed toward God as an intentional object. These attitudes and
behaviors can be displayed whether or not there is, in fact, a God for them
to be directed toward and whether or not God loves one in the ways one
takes God to.
While there doesn’t have to be a God in order for a human person to
accept God’s love in the sense I am concerned with here, it does seem plau-
sible that this person must at least have some sort of cognitive commitment
to there being a God in order to accept God’s love. The person must some-
how assume, or assent, or take it to be the case that God loves them. In this
way, accepting God’s love constitutes a faith practice of the kind described

2 Two recent similar discussions focusing on Christian sanctification are (Porter and
Rickabaugh 2018) and (Stump 2018).
140 Faith, Flourishing, and Agnosticism

in Chapter 3. Those who accept God’s love tend to believe or assume that
God exists and loves them. They tend to adopt a positive cognitive stance
toward God’s loving them consistently with minimal theism. As explained
in Chapter 3, adopting such cognitive attitudes is possible not just for those
who believe that God exists but also for those who are agnostic regarding
God’s existence—­a group of individuals that have been the focus of the
empirical work I will describe later in this chapter.
While cognitive commitments to God’s love are required for adopting a
richly accepting orientation toward that love, I would suggest that the cog-
nitive commitments are not all that is involved. After all, even the person
who believes in God but rejects God’s love may be cognitively committed to
God’s having allowed each of the many good things in their life as an
expression of love for them, and so on. It is just that they repudiate this love
from God, wish it didn’t exist, long to escape from it, or oppose it. There
may be a sense in which they “accept” God’s love—­the sense satisfied merely
by their being cognitively committed to it—­but there also seems to be a
richer sense in which they do not “accept” God’s love. It’s this richer sense of
accepting or embracing God’s love that is my focus here.
One of the main differences between the person who repudiates God’s
love and the person who embraces God’s love concerns their affections.
Those who reject God’s love are negatively affectively oriented toward God’s
love for them rather than positively affectively oriented toward it. My sug-
gestion, then, is that a second ingredient for accepting God’s love is that a
person be positively affectively oriented toward this love. They must tend to
experience positive emotions directed toward what they are committed to
taking to be God’s love for them. They will tend, for example, to be joyful
about and thankful for God’s bringing into their lives the many good things
God does out of love for them. They will appreciate God’s attentiveness to
them and be glad to be cared for by God.
This isn’t to say that a person can’t embrace God’s love while also
­experiencing some negative emotions related to God’s love. They might
appreciate God’s love for them yet feel all the more regretful of their own
wrongdoing—­perhaps in ways consistent with the tendency of contrition
described in the previous chapter. Or they might feel intimidated by the
ways in which God’s love might challenge them to change. Yet these nega-
tive affections seem not so much directed toward God’s love itself as toward
negative features of oneself or the prospects of the effects that God’s love
may bring about. Someone who embraces God’s love in the rich way I have
in mind will tend to view God’s love itself as positive, and it would be a
 Accepting God ’ s Love 141

deficiency in their orientation toward that love if their affections toward

God’s love did not align with this positive evaluative stance.
I suggest that this positive affective orientation will also be comple-
mented by positive desiderative and volitional orientations. The person who
accepts God’s love does not only experience positive emotions directed
toward God’s love for them; they also want God to love them and want to
experience God’s love for them. They are motivated to enjoy and acknowl-
edge God’s love for them. They try to express thanks for God’s love and to
show suitable affection in return toward God in response to the love they
take God to have shown them.
What I have briefly described in this section is a well-­integrated, richly
accepting orientation toward God’s love. Notably, this sort of orientation
toward accepting God’s love comes in degrees: a person can be more
strongly or more weakly disposed toward accepting God’s love in this way.
The orientation can also issue in particular acts of accepting particular ele-
ments of God’s love. Each act of acceptance of divine love, in turn, may itself
be more or less thoroughly accepting of that element of divine love, depend-
ing upon whether the relevant cognitive, affective, desiderative, and voli-
tional elements are present with respect to that particular element of divine
love. Such an orientation is typically developed through consistent practice,
and it is for this reason that we can think of accepting God’s love as a faith
practice. Continued practices of accepting God’s love lead individuals to
develop accepting orientations toward God’s love.
For my purposes here, this should suffice as a sketch of what is involved
in accepting God’s love. I do not intend this sketch to be exhaustive; there
may be more involved in accepting God’s love than I have identified here.
But I do intend the sketch to have highlighted several of the chief aspects
involved in accepting God’s love. To practice accepting God’s love is, at least,
to regularly adopt a positive cognitive, affective, desiderative, and volitional
orientation toward the varied aspects of divine love here highlighted. My
next question concerns the value of accepting God’s love in this sense.

2 Evidence for the Transformative Power of

Accepting God’s Love among Theists

There are many ways that accepting God’s love may be valuable. For exam-
ple, if a person accepts God’s love and God does love them in the ways
they accept, then they respond in an appropriate, fitting way to God’s love.
142 Faith, Flourishing, and Agnosticism

The cognitive commitments they adopt are accurate; the affective responses
they have are fitting; the desires and volitions they have track attain-
able values.
Moreover, if a person accepts God’s love and God does love them in the
ways they accept, this may lead to further additional goods. It may secure a
valuable form of relationship with God. This relationship may have implica-
tions for the person’s long-­term future. According to some approaches to
thinking about experiencing a heavenly afterlife, for example, forming such
a relationship with God is necessary for experiencing heaven and remains
eternally a significant component of the experience of heaven. Relating to
God in this way is thought of as the greatest good there could be for a per-
son (Stump 2018).
My focus here, however, will be on a value that accepting God’s love may
have whether or not God exists. Specifically, I will argue that accepting
God’s love has the particular value of being conducive to developing or
retaining moral virtues. Moreover, accepting God’s love can have this value
both for individuals who believe that God exists and loves them as well as
for individuals who lack belief that God exists and loves them but who
assume that God exists and loves them. More broadly, it can have this value
for individuals who have ambiguous evidence for God but who accept God’s
love in the way articulated in the previous section.
The main way I have in mind whereby accepting God’s love can be con-
ducive to moral virtue is indirectly, as opposed to directly. A direct approach
to developing or maintaining a virtue is to practice the characteristic activi-
ties of that virtue—­the characteristic behaviors, feelings, thoughts, and so
on associated with that virtue (cf. Porter and Baehr 2020). A direct approach
to developing or maintaining generosity, for instance, is to practice giving
things one values to benefit others with appropriate joy and thoughtfulness.
We saw this approach at work in the arguments developed in Chapter 5 for
thinking that giving God the benefit of the doubt is conducive to the devel-
opment of praisefulness, thankfulness, and contrition.
The indirect approach to virtue development I have in mind, which can
complement the direct approach, instead focuses on removing certain kinds
of obstacles to a person’s acting in accordance with virtue (cf. Porter and
Baehr 2020). There are many temptations and challenges that lead us away
from acting in accordance with virtue. If these obstacles can somehow be
neutralized, their power over us reduced through the “scaffolding” of our
personalities (Snow 2013), then this could free us to act in accordance with
virtue and thereby aid us in developing virtue. For example, in the case of
 Accepting God ’ s Love 143

generosity, we might be inclined to fear the loss of things we value or to

worry about embarrassing ourselves when we attempt to aid others with
our gifts, leading us not to act generously. Likewise, more generally, if we
suffer from depressive ideation or dissatisfaction with our lives, or lack self-­
esteem, we may struggle to act in accordance with virtue, doubting our own
abilities, becoming preoccupied with self-­rumination, and being distracted
from other-­oriented virtuous conduct. Shaping our personalities so that
these obstacles to virtuous behavior are neutralized can indirectly promote
our virtue development.
What I want to suggest here is that accepting God’s love can help to neu-
tralize these kinds of obstacles to our acting in accordance with virtue, and
it can thereby free us to develop or retain virtue. To construct an initial case
as to why it is reasonable to think that accepting God’s love can play this
role, it will be helpful to look at a substantial body of research on attach-
ment, including attachment to God. This research provides strong reason to
think that secure attachments to other people can play this indirect role in
virtue development and that secure attachment to God can play this role for
believers. In the next two sections, I will extend this argument to focus on
agnostics, presenting findings from two recent studies of my own concerned
with God attachment and acceptance of God’s love among agnostics. The
focus on agnostics should add to the plausibility of the focal conclusion of
this chapter that individuals with ambiguous evidence for God can grow in
virtue through accepting God’s love, since a charitable view of agnostics and
believers in God would suggest that agnostics probably have more ambigu-
ous evidence of God’s existence on average. Indeed, in some of my own
recent empirical work, reported in Chapter 7, I have found that there is a
significant difference in how agnostics and theists themselves rate their evi-
dence for God, with agnostics on average evaluating their evidence as
36.62% supporting God’s existence and theists on average evaluating their
evidence as 87.76% supporting God’s existence (Byerly 2023b). If we assume
charitably that most agnostics have ambiguous evidence for God and some
believers do as well, then the evidence surveyed in the remainder of this
chapter will support the conclusion that individuals with ambiguous evi-
dence for God can grow in virtue through accepting God’s love for them.
I begin with some brief background on attachment theory. Attachment
theory, as originally developed in psychology, was focused primarily on the
child–­caregiver relationship (Bowlby 1969). According to the theory, there
were three different types of attachment orientation a child might develop
toward a caregiver. They might be avoidant, trying to do as much as they
144 Faith, Flourishing, and Agnosticism

can on their own without relying on their caregiver, rejecting the affection
of their caregiver, or being cold toward them. They might be anxious, con-
stantly seeking their caregiver’s presence, feeling distraught about their
absence, being unable to engage their environment without their caregiver,
and worrying that their caregiver might abandon them or might prefer others
to them. Or they might be securely attached, a kind of happy medium in
which they are confident that their caregiver will be available to them and
supportive of them when needed, warm toward their caregiver, and unafraid
to engage their environment on their own and to return to their caregiver
when necessary. Anxious and avoidant attachment are both referred to as
insecure forms of attachment, in contrast to secure attachment.
Researchers soon realized that these patterns of attachment could apply
to a much wider range of relationships (Ainsworth 1985; Bowlby 1973),
including adult romantic relationships, relationships with friends, relation-
ships with inanimate objects, and relationships with deities (Kirkpatrick
and Shaver 1992). In these relationships, much as in the child–­caregiver
relationship, a person can be avoidant toward the other party in the rela-
tionship, anxious toward them, or securely attached to them. Moreover,
these orientations may come in degrees.
From a theoretical perspective, it should be expected that secure attach-
ment could indirectly support virtue development. One of the main func-
tions of securely attached relationships is to enable a person to regulate
affect (Bowlby 1988). The child explores their environment, experiences a
stressor, returns to their caregiver for support, and is better able to manage
the stressful trigger and resume exploring their environment. Similarly for
the adult romantic partner or friend. Securely attached relationships are a
source of mental well-­being and stability that enable us to confidently
engage our world. The security they provide can reduce the influence of the
kinds of obstacles to virtue development highlighted above. “Attachment
security,” Dwiwardani et al. put it, “provides a foundation for the practice of
relational virtues” (2014: 84).
This theoretical perspective is now supported by a wealth of empirical
evidence. Attachment security is very important for personal development.
Secure attachment is associated with higher needs for achievement, greater
likelihood of adopting mastery goals, and weaker fear of failure (Elliot and
Reis 2003). Secure attachment is related to greater curiosity (Mikulincer
1997), greater openness to new ideas (Bourne et al. 2014), and less biased
information seeking (Mikulincer 1997). Secure attachment is related to
greater self-­control (Tangney et al. 2004), greater attentiveness to one’s
 Accepting God ’ s Love 145

projects (Webster et al. 2009), and better planning and organization

(Learner and Kruger 1997). All of these features are important for develop-
ing and maintaining virtues—­they are precisely the sort of “personality
scaffolding” we are looking for. Studies have also confirmed more directly
the link between secure attachment and virtue. For example, securely
attached individuals exhibit greater empathic concern, compassion, and
altruism (Granqvist et al. 2010), and they are more forgiving, grateful, and
humble than their insecure counterparts (Dwiwardani et al. 2014).
Since researchers first posited that God may function as an attachment
figure, evidence has mounted that secure attachment to God can function
in much the same way as secure attachment to caregivers or romantic part-
ners when it comes to features such as mental health and virtue. Secure
attachment to God is associated with experiencing less negative pressure
regarding body image and self-­esteem (Ellison et al. 2011), and being less
susceptible to problematic internet use (Knabb and Pelletier 2013) and alco-
hol and drug abuse (Horton et al. 2010). Those with secure attachment to
God experience greater satisfaction with life and less loneliness and depres-
sion (Kirkpatrick and Shaver 1992; Reiner et al. 2010). The relationship
between secure God attachment and virtues has also been studied more
directly, with secure attachment positively linked to humility (Jankowski
and Sandage 2014; Sandage et al. 2015), forgiveness (Davis et al. 2008), and
gratitude (Byerly 2023b).
It is important that these benefits of attachment to God appear to go
beyond benefits attained from other secure attachment relationships. That
is, even controlling for other secure attachments, researchers have found
that secure attachment to God still makes a significant contribution to these
kinds of variables (Keefer and Brown 2018; Njus and Scharmer 2020). Thus,
it appears that secure attachment to God can play an important and unique
role in an indirect approach to virtue development.
If secure God attachment can play this role, it is plausible that accepting
God’s love can as well. Accepting God’s love, as described in Section 1,
largely overlaps with what researchers are measuring when they measure
attachment to God. A person with a richly accepting orientation toward
God’s love is much more likely than their counterpart to have a secure
attachment to God.
There are two widely used scales for measuring God attachment in the
literature. One, a 28-­ item measure developed by Beck and McDonald
(2004), is more emotionally oriented. Avoidant attachment is measured
using items such as “I just don’t feel a deep need to be close to God” and
146 Faith, Flourishing, and Agnosticism

“My experiences with God are very intimate and emotional” (reverse
scored). Anxious attachment is measured using items such as “I worry a lot
about my relationship with God” and “I fear God does not accept me when
I do wrong.” And secure attachment is operationalized as low avoidant and
low anxious attachment. The other measure is a nine-­item measure devel-
oped by Rowatt and Kirkpatrick (2002), which leans more in a cognitive
direction. Avoidant attachment is measured using items such as “God seems
to have little or no interest in my personal affairs” and “I have a warm rela-
tionship with God” (reverse scored), while anxious attachment is measured
using items such as “God’s reactions to me seem to be inconsistent.” Secure
attachment, again, is operationalized as low anxious and low avoidant
attachment.
It should be clear enough that someone who adopts the richly accepting
orientation toward God’s love identified in Section 1 would tend to respond
to these items in the way a person with secure God attachment would. For
example, given their tendencies to adopt positive cognitive attitudes toward
God’s having shown love to them in various ways, they will tend to disagree
with the idea that God seems to have little or no interest in their personal
affairs. Their tendencies to respond to what they take to be God’s love for
them with positive affect will lead them to regard their relationship with
God as more warm, intimate, and emotional. And adopting a well-­
integrated, accepting orientation toward God’s loving them with a love that
at least matches any human person’s love for anyone else will tend to work
against perceptions that God is inconsistent toward them or fears that God
does not accept them. Accordingly, this research on God attachment sup-
ports the claim that accepting God’s love can play a significant role in indi-
rect virtue development.
To put it more carefully, this research supports, primarily, the claim that
accepting God’s love can play this role for believers. The research confirms
that, at least in the case of those who believe in God, it is important for their
virtue development that they accept God’s love—­failing to accept it by either
being avoidant or anxious toward God negatively influences the believer’s
ability to develop or maintain moral virtues.
I say that the research primarily supports these conclusions about believ-
ers because, with few exceptions, this research has focused on the potential
significance of God attachment for those who believe in God, not for those
who lack belief in God. In most cases, samples collected contain few if any
nonbelievers. In some cases, while data were collected on God attachment
for nonbelievers, these data were purposefully excluded from the analysis
 Accepting God ’ s Love 147

by researchers. Leman et al. (2018), for example, wrote of their procedures,

“Given our interest in how people view God or their relationship with God,
we limited the sample to participants who had high certainty about their
belief in God” (165).
Yet not all researchers would agree with the idea that nonbelieving par-
ticipants should be excluded from research on attachment to God. In their
paper on God attachment and eating disorders, Strenger et al. (2016) make
precisely the opposite contention. They write, “Although it may seem
counter-­intuitive to assess attachment to God in people who do not claim
belief in a deity, previous research has demonstrated that people who do
not believe in God still hold mental representations of God that affect their
behaviours, emotions, and cognitions” (25). Their own analysis included
both believing and nonbelieving participants, and they found that for the
whole sample, anxious attachment to God was positively related to eating
disorder symptoms. Moreover, they found that the way in which anxious
attachment moderated the link between sociocultural pressure and eating
disorder symptoms did not differ between believing and nonbelieving par-
ticipants. They therefore endorse the idea that “future research is needed to
understand if/how attachment to God affects non-­believers” (33).
There is also complementary research on attachment to God with Jewish
populations (Pirutinsky et al. 2019) that is at least suggestive of the potential
significance of attachment to God for those without strong cognitive com-
mitments to God. It is not that Jews are agnostics. Rather, as emphasized by
the researchers who have conducted these studies, Judaism tends in empiri-
cally verifiable ways to downplay the importance of the cognitive dimen-
sions of religion and play up the importance of practice. Because Judaism
downplays the importance of the cognitive in this way, researchers expected
that attachment to God would not be significant for Jewish participants. But
they found exactly the opposite. Attachment to God was significant for anx-
iety and depression in their participants. In the authors’ summary of their
results, they write that these results appear to indicate that “attachment to
God—­as opposed to belief, faith, or even conviction—­may be a unique
internal variable [linking] religiosity and mental health” (167).
These observations are what prompted my own research interest in
assessing how accepting God’s love may relate to mental health and virtue
for agnostics. As we have seen, there is a strong case to be made for thinking
that accepting God’s love can facilitate better mental health and can thereby
indirectly promote virtue development for believers in God. I have sought
to explore more thoroughly than previous researchers whether the same
148 Faith, Flourishing, and Agnosticism

kind of relationship holds in the case of agnostics. If so, this would

strengthen the case for thinking that individuals with ambiguous evidence
for God can indirectly grow in virtue through engaging in the faith practice
of accepting God’s love.

3 Evidence of the Transformative Power of Accepting

God’s Love among Agnostics

Thus far, I have completed two studies of accepting God’s love among
agnostics. One is a secondary data analysis that assesses the relationship
between God attachment and depression and self-­ esteem among self-­
identified agnostics (Byerly 2022b). The second is a more ambitious original
study with a larger group of self-­identified agnostics (Byerly 2023b). It more
rigorously assesses the relationship that God attachment and newly devel-
oped measures of accepting and resisting God’s love have with several well-­
being indicators, using methods that enable us to ascertain to what extent
accepting God’s love is uniquely significant for these well-­being indicators.
Both studies support the claim that accepting God’s love is significantly
related to better mental health for agnostics, with the second study offering
stronger support for thinking that accepting God’s love makes a unique
contribution to agnostics’ mental health.
The first study reanalyzes an existing data set in which data about God
attachment were collected from self-­identified agnostics but not analyzed
(Njus and Scharmer 2020: study 2). This study included 790 participants, of
whom 120 identified as agnostic. The central finding of the study was that
there are significant differences in depression and self-­esteem when groups
of theists with secure attachment to God are compared with agnostics. The
mean for securely attached theists’ depression was 12.67 compared to a
mean of 20.76 for agnostics, while the mean for securely attached theists’
self-­esteem was 28.57 compared to a mean of 22.57 for agnostics. Depression
was measured using the Center for Epidemiological Studies Depression
Scale Short Form (Cole et al. 2004), while self-­esteem was measured with
the Rosenberg self-­esteem scale (Rosenberg 1965). God attachment was
measured using the instrument created by Beck and McDonald (2004).
In my study, I looked at bivariate relationships between agnostics’ anx-
ious and avoidant God attachment and self-­esteem and depression, and
I conducted difference-­in-­means tests to determine whether subgroups of
agnostics with secure, anxious, and avoidant God attachment differed
 Accepting God ’ s Love 149

significantly in self-­esteem and depression. It is important to understand

that these tests are different and that both are needed to fully understand
the significance of secure attachment to God because of the way secure God
attachment is operationalized. It is operationalized as scoring low for both
anxious and avoidant attachment. Thus, to study secure God attachment,
the researcher has to create subgroups of individuals who score highly in
anxious attachment (an anxiously attached group), highly in avoidant
attachment (an avoidantly attached group), and low in both anxious and
avoidant attachment (a securely attached group).
The bivariate relationship between anxious God attachment and depres-
sion was significant in my study (r = .33, p < .001), while the bivariate rela-
tionship between anxious God attachment and self-­esteem trended toward
significance (r = −.16, p < .1). Relationships between avoidant God attach-
ment and self-­esteem and depression were not significant. Taken alone, this
might suggest that agnostics who experience anxious attachment to God
may experience worse mental health, while agnostics who avoid God may
experience no adverse effects. To put this into the idiom of accepting God’s
love, we might take the lesson to be that it is important for agnostics not to
worry and fret over their relationship with God and God’s love for them,
but if they just pay this relationship no mind, ignoring God’s love for them,
that may be okay. This finding alone, then, does not tell us whether agnos-
tics who adopt a richly accepting orientation toward God’s love are better
off than agnostics who ignore or resist God’s love for them.
The difference-­in-­means tests help to remedy this deficiency. These tests
revealed that the difference between securely and insecurely attached
groups of agnostics are significant for both depression and self-­esteem
(p = .02 in both cases). More specifically, differences between anxiously and
securely attached agnostics were significant for depression (p = .01); the
differences between anxiously and securely attached and between avoid-
antly and securely attached agnostics both approached significance for self-­
esteem (p = .06); and the differences between anxiously and avoidantly
attached agnostics were insignificant for both depression and self-­esteem.
This pattern of results suggests that when it comes to depression and self-­
esteem among agnostics, having a secure attachment to God is better than
both having an anxious attachment to God and having an avoidant attach-
ment toward God. Developing an attachment relationship toward God
more characteristic of what would be expected of someone who accepts
God’s love for them seems to be better for agnostics’ depression and
self-­esteem.
150 Faith, Flourishing, and Agnosticism

Table 6.1 Mean scores for mental health for secure theists and
different groups of agnostics

Self-­Esteem Depression

Securely attached theists 28.57 12.67

Securely attached agnostics 26.86 14.86
Agnostics 22.57 20.76
Insecurely attached agnostics 21.81 23.03
Avoidantly attached agnostics 21.60 22.40
Anxiously attached agnostics 21.55 25.82

Table 6.1 provides group means for depression and self-­esteem for four
groups: securely attached theists, agnostics as a whole, securely attached
agnostics, insecurely attached agnostics, avoidantly attached agnostics, and
anxiously attached agnostics. As the reader can see, secure attachment
roughly makes up the difference in scores for depression and self-­esteem
observed in Njus and Scharmer’s original study between securely attached
theists and agnostics. In other words, these results suggest that being
securely attached to God erases the observed differences in depression and
self-­esteem between agnostics and securely attached theists.
While this secondary data analysis suggests that agnostics’ God attach-
ment is significant for their mental health and may even make up for
observed differences between securely attached theists’ and agnostics’ men-
tal health, further study of the topic is highly desirable for several reasons.
First, the sample of agnostics used in this study is fairly small, which makes
it less likely for effects of God attachment to be observed. A study with a
larger number of agnostics may be better able to detect effects of agnostics’
God attachment. In a larger sample, for instance, relationships that
approached significance in this study may be found to cross that conven-
tional threshold. Second, it is desirable in general to attempt to replicate
these findings. If they can be replicated, then this, together with their coher-
ence with the body of research described in the previous section, provides
more confidence in their conclusions. Third, this study alone does very little
to reveal the potential unique significance of accepting God’s love for agnos-
tics’ mental health. While it indicates that agnostics’ God attachment is
related to their self-­esteem and depression, it leaves open the possibility that
when additional variables are controlled for, agnostics’ attachment to God
may no longer be significant for these variables. Finally, this study used only
 Accepting God ’ s Love 151

agnostics’ God attachment as a way of trying to capture whether agnostics

accept God’s love. As explained in Section 2, it would seem that securely
attached agnostics probably exhibit greater acceptance of God’s love.
However, there is also reason to think that measures of God attachment
may not fully capture the accepting orientation toward God’s love described
in Section 1. For instance, only two items in the Beck and McDonald meas­
ure reference God’s love: “Sometimes I feel that God loves others more than
me” and “I crave reassurance from God that God loves me.” Both of these
assess participants’ lack of acceptance of God’s love due to anxious attach-
ment rather than assessing their positive acceptance of God’s love. A meas­
ure that more directly targets whether agnostics assume that God loves
them, sincerely act as if God loves them, and have faith that God loves them
may do a better job of capturing whether agnostics adopt an accepting
­orientation toward God’s love.
All of these limitations informed my work in running a second, more
ambitious study of acceptance of God’s love among agnostics (Byerly 2023b).
For this study, I recruited a larger sample of 360 self-­identified agnostics
who answered the question “Which of the following best characterizes your
views about God?” with “Agnostic: I neither believe God does exist nor
believe God doesn’t exist” rather than “Theist: I believe God does exist” or
“Atheist: I believe God does not exist.” The study was designed to provide an
opportunity to replicate the previous findings concerning bivariate relation-
ships between God attachment and depression and self-­esteem and con-
cerning difference-­in-­means for depression and self-­esteem among securely
attached and insecurely attached groups of agnostics. The larger sample
raised the likelihood of observing significant effects for these tests.
The study was also designed to take research on agnostics’ acceptance of
God’s love and mental health further. It did this primarily in two ways. First,
I constructed a novel measure designed to more directly assess agnostics’
acceptance of God’s love, guided by the philosophical literature on nondox-
astic faith discussed in Chapter 3. This takes the average of their responses
to the items “I assume that God really loves me,” “I hope that God really
loves me,” “I act as if God really loves me,” “I accept that God really loves
me,” and “I have faith that God really loves me.” The study allowed me to
examine bivariate relationships between this new measure and agnostics’
self-­esteem, satisfaction with life, depression, and gratitude. But it also
allowed me to compare this new measure with both the Beck and McDonald
and Rowatt and Kirkpatrick measures of attachment to God, which takes
me to the second way in which this study advances research on God
152 Faith, Flourishing, and Agnosticism

attachment among agnostics. In this study, unlike the first, I conducted

hierarchical regressions in order to assess the unique significance of God
attachment and the new measure of acceptance of God’s love for these out-
come variables when controlling for other variables. At a first step, I entered
participants’ age, sex, Big Five personality traits (neuroticism, agreeable-
ness, extroversion, conscientiousness, and openness), importance of reli-
gion, importance of spirituality, image of God (whether they imagine God
as being more cruel, kind, or distant), past experiences with God (past
anger with God, past belief that God was angry with them, past doubts
about God, and past positive emotions toward God), and how they rate
their evidence for God. The second step added the measures of God attach-
ment to determine whether these predict additional variance in the out-
come variables beyond the variance predicted by those variables included at
step 1. As a third step, I added the new measure of acceptance of God’s love
to determine whether it predicts additional variance beyond the variables
included at both steps 1 and 2.
Bivariate correlations in this study replicated the previous findings and
extended them. Anxious God attachment, particularly as measured using
the Beck and McDonald instrument, was significantly related to higher
depression (r = .43, p < .001), lower satisfaction with life (r = −.48, p < .001),
lower self-­esteem (r = −.46, p < .001), and lower trait gratitude (r = −.15,
p < .01). Similarly, difference-­in-­means tests replicated and extended the
findings of the previous study, with participants in the securely attached
group scoring significantly higher than their anxiously and avoidantly
attached counterparts for self-­esteem and satisfaction with life but signifi-
cantly lower for depression. Mean values for these variables for each group
are reported in Table 6.2.
The most salient findings from the hierarchical regressions concerned the
additional predictive value of anxious God attachment as measured by the
Beck and McDonald instrument and the newly developed acceptance of
God’s love measure. Anxious God attachment predicted additional variance
beyond the step 1 variables in agnostics’ life satisfaction (B = −.25, SE = .08,
p < .001), self-­esteem (B = −.17, SE = .07, p < .01), and depression (B = .17,
SE = .10, p < .05). Accepting God’s love predicted additional variance beyond
the step 1 and step 2 variables in agnostics’ gratitude (B = .17, SE = .06, p < .01),
life satisfaction (B = .20, SE = .07, p < .01), and self-­esteem (B = .20, SE = .07,
p < .01). These findings suggest that anxious attachment to God and
acceptance of God’s love are uniquely predictive of these outcome variables
when controlling for the other variables in the study. Together, God
 Accepting God ’ s Love 153

Table 6.2 Mean scores for mental health for agnostics with different God
attachment

Self-­Esteem Depression Life Satisfaction

Securely attached agnostics 5.21 2.07 5.68

Avoidantly attached agnostics 4.34 3.20 4.64
Anxiously attached agnostics 3.90 3.58 4.27

attachment and acceptance of God’s love accounted for an additional 9.3%

variance in life satisfaction, 8.5% variance in self-­esteem, 6.4% variance in
depression, and 2.6% variance in gratitude.
These two studies, when considered in light of the much larger literature
on God attachment among theists and the philosophical perspective on
nondoxastic faith discussed and developed in Chapter 3, provide a coherent
and powerful narrative that supports the idea that accepting God’s love can
play an indirect role in virtue development for individuals with ambiguous
evidence for God. Charitably considered and supported to some extent by
the arguments of Chapter 2, many agnostics probably have ambiguous evi-
dence for God. Some theists, too, probably have ambiguous evidence for
God. So, the fact that psychological theory and research support the claim
that theists and agnostics who accept God’s love experience better mental
health that can serve as an indirect support for virtue development also
confirms that this is true for individuals with ambiguous evidence for God.

4 Responses to Objections

I will close this chapter by responding to three objections that might be

raised against the argument developed in previous sections. This will pro-
vide an occasion for supplying some further support for the argument thus
far developed as well as for indicating how future research could further test
the main hypothesis of the chapter.
First, it might be objected that even if the argument succeeds in showing
that individuals with ambiguous evidence for God who accept God’s love
for them are likely to experience better mental health and thereby to grow
in some virtues, this will only come at the expense of other virtues. An
objection along these lines might be gleaned from criticisms of theistic reli-
gion that maintain theistic religious practice involves treating God as a
“crutch.” The criticism that religious belief acts as a kind of psychological
154 Faith, Flourishing, and Agnosticism

crutch for believers has a venerable history and is commonly associated

with Freud and other advocates of wish-­fulfillment explanations of religion
(for a review, see Guthrie 2006).
In one of very few articles in Philosophy of Religion written on this topic,
Angelo Juffras (1972) points a way forward for how the details of this sort of
criticism might be developed. Juffras, who frames the debate as focused on
using “theology” as a crutch, writes, “Presumably, the opponents of theology
wish to train man to a higher state of virtue” (256). Either using theology as
a crutch tends toward the development of vice or it tends away from the
development of virtue. Clarifying which virtues might be involved, Juffras
writes, “When theology is disparaged as a crutch, this also suggests what is
approved, viz. self-­sufficiency” (256). Thus, one might take the objection
to be that, if an individual with ambiguous evidence for God accepts
God’s love in part in order to boost their mental health and thereby be in
a position to become a better person, they are failing to be self-­sufficient;
they are succumbing to too strong a dependence on God in their pursuit
of virtue.
I would suggest, however, that this criticism may be met, or at least
largely blunted. Recent research on the virtue of autonomy has strongly sug-
gested that, when properly understood, autonomy should not be thought to
involve a strong reluctance toward relying on others. Rather, it involves a
reflective and attentive reliance upon others. This is especially clear in work
on the virtue of intellectual autonomy in particular. For instance, Roberts
and Wood (2007), in their discussion of the virtue of intellectual autonomy,
argue that virtuous autonomy “involves a wise dependence” on others (258).
Indeed, they go so far as to suggest that the autonomous person “sees his
indebtedness [to others] as a good and fitting thing, not at all second-­rate or
to be regretted” (258). Similarly, we might propose here that the virtue of
proper autonomy does not demand that individuals with ambiguous evi-
dence for God shy away from depending on God to fulfill their needs for
relational support; rather, it is compatible with their accepting God’s love
reflectively and wisely given their understanding of themselves and the
potential values that accepting God’s love may hold for them.
A second objection targets the uniqueness of accepting God’s love for
one’s virtue development rather than accepting the love of some other
potential lovers. The most powerful version of this objection, in my view,
appeals to accepting the love of one’s mother- and father-­like figures and
one’s romantic partner.3 Is accepting love from God really needed beyond

3 Another version focuses on the love of one’s ancestors; see (Byerly 2022b).
 Accepting God ’ s Love 155

accepting love from these figures? Could it be that, instead, whether a

person accepts God’s love is indicative of how they tend to relate to other
potential lovers, and it is their orientation toward the love of other
human lovers that really explains the relationship between accepting
God’s love and participants’ mental health? Put in terms of the attach-
ment idiom, the question is whether God attachment predicts unique
variance in mental health when controlling for attachments to significant
human others.
We have already seen that previous studies have suggested that there is in
fact a unique relationship between God attachment and mental health for
theists. Theistic participants’ attachment to God is predictive of their mental
health even when controlling for their attachments to human others (e.g.,
Keefer and Brown 2018; Njus and Scharmer 2020). So, when it comes to
individuals who believe that God exists, it seems that the objection does not
have much force. However, I do not know of any studies with agnostics that
have addressed this question. We might wonder whether in the case of indi-
viduals who lack belief that God exists, the lack of belief may put them in a
position where their acceptance of God’s love will make little difference for
their mental well-­being beyond what is contributed by their orientations
toward other humans’ love for them.
We can make a start at answering this question using the data set col-
lected for the second study with agnostics summarized in the previous sec-
tion. For that study, I had participants’ complete measures of not only their
attachment to God but their attachment to mother- and father-­like figures
and romantic partners, if they had them, using the experience in close
relationships—­relationship structures questionnaire (Fraley et al. 2011). I
have not previously reported analysis of these data; the report that follows is
original to this chapter.
Two tests we can run that could be useful for answering our question
about whether accepting God’s love predicts agnostics’ mental health
beyond their human attachment relationships involve hierarchical regres-
sions. First, we could enter participants’ human attachments at step 1 and
their God attachment at step 2, seeking to discern whether God attach-
ment predicts additional variance in mental health variables beyond par-
ticipants’ human attachments. Second, we could repeat this same process,
adding the direct measure of participants’ acceptance of God’s love as a
third step. This would allow us to discern whether participants’ acceptance
of God’s love is related to their mental health only because it covaries with
their other attachment relationships or whether it is uniquely related to
their mental health.
156 Faith, Flourishing, and Agnosticism

When we complete these procedures, we get results that complement

those that have been found with theistic participants. When participants’
human attachments are controlled for at step 1, the addition of God
attachment—­specifically, anxious God attachment—­remains a significant
predictor of participants’ depression (B = .29, SE = .11, p < .001), self-­esteem
(B = −.34, SE = .09, p < .001), and life satisfaction (B = −.35, SE = .08, p < .001)
at step 2. Similarly, when the direct measure of participants’ ac­cept­ance of
God’s love is added at step 3, it is a significant predictor of participants’
self-­esteem (B = .21, SE = .08, p < .01), life satisfaction (B = .17, SE = .08,
p < .05), and gratitude (B = .19, SE = .06, p < .05). Together, after controlling
for participants’ human attachments, their God attachment and acceptance
of God’s love accounted for an additional change in variance of 14.7% for
life satisfaction, 11.3% for self-­esteem, 5.9% for depression, and 8.2% for
gratitude.
These results suggest that for agnostic participants, as for theistic partici-
pants, accepting God’s love may have a significance for participants’ mental
health that is not simply reducible to their acceptance of other humans’
love. There is variance in agnostics’ mental health that is uniquely predicted
by agnostics’ attachment to God or acceptance of God’s love. Although
additional confirming evidence would be useful, the second objection does
not seem to be very compelling given this evidence. Individuals with ambig-
uous evidence for God, whether they believe God exists or not, seem to be
in a position where accepting God’s love may be uniquely related to their
experiencing better mental health.
A third and final objection targets the claim that accepting God’s love
may causally bring about better mental health for individuals with ambigu-
ous evidence for God and in this way promote an indirect pathway for their
virtue development. The objection can be sharpened by focusing on the
studies with agnostic participants summarized in Section 3. Both of these
studies are cross-­sectional in nature, measuring participants’ God attach-
ment or acceptance of God’s love alongside other variables at a single time.
As such, while the findings can demonstrate covariance between these vari-
ables, they cannot speak directly to the causal direction of this relationship.
They may establish that accepting God’s love is related to experiencing
­better mental health for agnostics; but they don’t by themselves indicate that
it is the acceptance of love that causes the better mental health, or vice versa,
or whether the two have some other common cause.
Of the three objections, I think this one is the closest to being on track. It
must be admitted that the studies summarized in Section 3 do not directly
 Accepting God ’ s Love 157

speak to the question of the direction of causation because of the nature of

the study design. Future research utilizing different study designs, such as
longitudinal designs that measure changes in agnostics’ acceptance of God’s
love and mental health over time, would be highly desirable for the pur-
poses of more directly addressing this question about causation. I hope to
be involved in such work myself in the near future.
While I do not think there is evidence available yet that directly addresses
the causal relationship between agnostics’ acceptance of God’s love and
their mental health, there is evidence that may indirectly favor the causal
hypothesis at the heart of this chapter. First, the hierarchical regressions
help to eliminate potential candidates for common causes of both ac­cept­
ance of God’s love and better mental health. They show that accepting God’s
love and mental health variables remain related when controlling for several
other variables, so at least it cannot be those variables that are common
causes of both accepting God’s love and better mental health. Or, if they are
common causes, they are only common partial causes.
Second, research with theistic participants using longitudinal designs
does indicate that God attachment is causally related to mental well-­being.
Bradshaw and Kent (2018) found that secure God attachment was prospec-
tively predictive of increases in optimism and self-­esteem over time. Ellison
et al. (2011) showed that secure God attachment at baseline was predictive
of decreases in stress over time and moderated the deleterious effects of
stressful life events. And Calvert (2010) found that insecure God attach-
ment was predictive of worse mental health over time even after controlling
for interpersonal human attachments. There is also research that suggests
that better mental health may also foster more secure attachment to God
(e.g., Thauvoye et al. 2018). Taken together, these results may suggest that
there is a reciprocal relationship between God attachment and mental
well-­being, with each variable sometimes causally influencing the other
(Cherniak et al. 2021).
This research only directly confirms the causal link between God
attachment and mental health for believers. But as we have seen above,
the pattern indicated in existing research seems to be that what goes for
believers with respect to their God attachment or acceptance of God’s
love tends to go for agnostics as well. Until further evidence using longi-
tudinal designs with agnostics is available, it would seem that this pattern
favors the view that there is likely to be a reciprocal relationship between
acceptance of God’s love and mental health for agnostics, too, with
acceptance of God’s love sometimes causally influencing agnostics’
­
158 Faith, Flourishing, and Agnosticism

mental health. To be sure, however, this is an area where further research

is desirable.
None of the objections surveyed in this section is clearly successful in
overturning the main argument of this chapter. Engaging in the faith practice
of accepting God’s love may help individuals with ambiguous evidence for
God to experience better mental health, which can in turn enable them to
grow in virtue. This seems to be true for both those whose acceptance of
God’s love involves belief and for those whose acceptance of God’s love
involves a cognitive commitment short of belief. While further research
with nonbelieving participants is desirable, the hypothesis that even for
these individuals accepting God’s love can serve as a pathway toward better
mental health and virtue development is very promising.
 7
Spiritual Excellence

This chapter identifies two final pathways whereby individuals with

ambiguous evidence for God may grow in virtue by engaging in practices of
faith toward God. Both pathways focus on a trait I call “spiritual excellence.”
The first pathway proposes that spiritual excellence itself is a virtue that can
be recognized as such by both the religious and nonreligious alike. Given
that it is a virtue, individuals who engage in faith practices that promote
spir­it­ual excellence thereby grow in this particular virtue. The second path-
way is focused on the transformative effects of spiritual excellence and its
exercise, independently of whether spiritual excellence is itself a virtue.
It proposes that individuals with ambiguous evidence for God may grow in
other more standard virtues, such as generosity or kindness, by acting in
accordance with spiritual excellence. The chapter begins in Section 1 by
offering an account of spiritual excellence and defending its conduciveness
toward growth in other virtues. It then appeals in Section 2 to the results of
an original empirical study in order to argue that individuals with ambiguous
evidence for God can cultivate spiritual excellence by engaging in contem-
plative theistic faith practices.

1 An Account of Spiritual Excellence

Spirituality is highly valued within many religious traditions and is often

associated with character development. But spirituality is increasingly rec-
ognized to have a life of its own, independent from religion, in which its
value is also maintained. Indeed, the idea that there could be a human vir-
tue or cluster of virtues concerned with spirituality, recognizable as such
from both religious and nonreligious perspectives, has become increasingly
popular among both psychologists and philosophers. I will review a few
examples here before presenting my own account of spiritual excellence,
which shares much in common with these views.

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0008
160 Faith, Flourishing, and Agnosticism

One example of this kind of treatment of spirituality derives from

Christopher Peterson and Martin Seligman’s landmark work Character
Strengths and Virtues (2004), which initiated the field of positive psychol-
ogy. Peterson and Seligman identified spirituality as one of the four
­character strengths of the virtue of transcendence. Indeed, they describe
spirituality as the “prototype” of the whole cluster, which also includes
appreciation of beauty, hope, gratitude, and humor. What unites the entire
cluster is that the various traits enable “individuals to forge ­connections
to the larger universe and thereby provide meaning to their lives” (519).
People with the character strength of spirituality are described as
“hav[ing] a theory about the ultimate meaning of life” that is “linked to
an interest in moral values and the pursuit of goodness” (533), and it
is claimed that “spiritual people are likely to experience frequent and
powerful awe” (539). As with other character strengths, spirituality is
conceptualized as a positive trait-­like feature and a candidate to be a
­virtue in the Aristotelian sense. Peterson and Seligman are explicit that
spirituality can be recognized as virtuous from multiple religious perspec-
tives and none.
John Cottingham is another author who has written widely on spir­it­u­al­
ity, religion, and ethics. He claims that the “two main components of
spir­it­ual­ity” are “spiritual praxis” and “spiritual experience” (2017: 14). The
former comprises “spiritual techniques,” such as prayer, meditation, or fast-
ing, that may or may not be derived from or at home within a religious tra-
dition. Spiritual experiences, according to Cottingham, contain both a
“human dimension,” which concerns our deep human responses and aspi-
rations, and a “cosmic dimension” which “draws us forward and beyond
ourselves . . . and enables us somehow to be part of, or one with, something
mysterious” (18). Both spiritual praxis and spiritual experience contain
ineliminable moral aspects. Cottingham writes that “the overriding aim of
authentic spiritual praxis is to facilitate the emergence of a better self ” (24),
while paradigmatic spiritual experiences are “infused with awe and charred
with moral significance, where the individual feels him or herself to be
checked, to be scrutinized, and to be called on to respond and to change”
(25). On Cottingham’s view, a close interrogation of these features of spir­it­
ual experiences reveals that their occurrence makes best sense only if
­theism is true and these experiences in fact involve a confrontation with a
personal source of meaning and value, whether or not they are always
understood by the experient as such.
David McPherson, another recent author on spirituality, has vigorously
defended its importance in the good life against contemporary versions of
 Spiritual Excellence 161

Aristotelianism that threaten to exclude it. McPherson offers a definition of

spirituality as “a practical life-­orientation that is shaped by what is taken to
be a self-­transcending source of meaning, which involves strong normative
demands, including demands of the sacred” (2017: 64). As with Cottingham,
spirituality is understood to be expressed through practices such as “self-­
examination, repentance, mindfulness, study, contemplation,” and many
others. To practice spirituality, one must adopt “a way of seeing and direct-
ing one’s life as a whole” in which one seeks “to orient one’s self better
toward the good” (65). Moreover, the spiritual life is not just concerned
with any old goods but is particularly concerned with qualitatively superior
and superlative goods of a sacred, holy, or reverence-­worthy nature, which
are taken to generate strong normative demands on the practitioner’s life.
Spirituality itself is not a virtue and can be practiced both well and poorly.
Piety is “the virtue concerned with a proper relationship in feeling and in
action to the sacred or the reverence-­worthy” (74).
Finally, Pierre Hadot is well known for his characterization of all of the
ancient schools of Hellenistic philosophy as spiritual traditions. On Hadot’s
account, these schools each combine a certain practical ideal as to how to live
one’s life with a philosophical worldview or discourse that supports practi-
tioners in pursuing that ideal. Epicureanism, for instance, is “a philosophy
which seeks, above all, to procure peace of mind” (1995: 222). The atomistic
world­view of Epicureanism articulated in its philosophical discourse aims
to “liberate mankind from everything that is a cause of anguish for the soul:
the belief that the gods are concerned with mankind; the fear of postmortem
punishment; the worries and pain brought about by unsatisfied desires”
(1995: 222).
Mark Wynn outlines Hadot’s view in more detail as follows:

The ‘philosopher’ begins with a conception of the good human

life . . . . Granted some such conception, the sage then seeks to adopt the
requisite means for realizing the good life so understood, and to this end
engages in various spiritual disciplines, which are designed to help them
internalize the world view specified in the relevant philosophical dis-
course. If these exercises are efficacious, then the adept will become habit-
uated to thinking of themselves in terms of that world view, and will
thereby achieve a way of life in which their favoured psychological and
moral condition can be enduringly realized. (2020: 12–13)

Importantly, Hadot does not require that the worldviews of philosophical

discourse be endorsed by the adept as true in all their details for them to
162 Faith, Flourishing, and Agnosticism

play this role. Rather, as argued by Wynn, the adept need only take the
world­view to have “some prospect of providing at least an approximation to
the truth” (161) or to be “a serious contender for truth, considered as a gen-
eral guide to the nature of things” (15). They cannot, however, “treat their
world view as simply a matter of make-­believe” (161). The worldview itself
may be distilled into “a small number of principles, tightly linked together,
which derived great persuasive force and mnemonic effectiveness” (Hadot
1995: 267).
My own approach here will be to articulate an account of spiritual excel-
lence as a virtue that brings together several of the key elements found in
the above views. The account I offer does not overlap entirely with any of
these accounts, as certain details from each are excluded or new features are
added. I also stress that my aim is only to identify one virtue that seems to
be central to living the spiritual life well. It may be that there are others, too,
and that some of the features I leave out from the above accounts are
important components of these other traits. In calling the trait “spiritual
excellence” rather than “piety,” I am using terminology closer to what is
found in contemporary psychology than philosophy and that has a looser
connection to institutional religion, although I’m not one to fuss over labels.
On my account, spiritual excellence is a tendency to make skillful use of a
worldview for which one has ambiguous evidence or better in order to
experience morally transformative awe of the awesome. The person who
possesses this virtue, as introduced in Chapter 3, appropriately values expe-
riencing awe of the awesome, and they also appropriately value moral trans-
formation. They make skilled use of a worldview in order to pursue
transformative experiences of awe. This worldview articulates their place in
the larger whole of reality, though it needn’t be very fine-­grained. Minimal
theism, for instance, will count as a worldview because it implies that each
human being has a place in reality as a whole that includes them being an
object of love of the ultimate source of reality, God, alongside other human
persons who are also objects of God’s love. In making skillful use of a world­
view, the adept engages in spiritual practices. These practices bring about
transformative experiences of awe, which as I will unpack in more detail
subsequently include an experience of feeling connected to a wider whole,
such as one’s group, humanity as a whole, the earth, the universe, or reality.
There are very clear areas of overlap between this proposal and the con-
ceptions of spirituality previously described. With Peterson and Seligman,
I affirm that the person of spiritual excellence “has a theory” about the ulti-
mate, or a “world view,” as Wynn puts it, or a “way of seeing” one’s life as a
 Spiritual Excellence 163

whole, following McPherson. I won’t try to enter into debates about exactly
what worldviews are, as I am using the term stipulatively, at least in part.
What is key for me is the idea articulated above that worldviews must con-
tain some account of where the person fits within reality as a whole. This, of
course, allows worldviews, and also individuals who practice spiritual excel-
lence, to be both religious and nonreligious, since there are both religious
and nonreligious accounts of how people fit within reality as a whole.
Allowing that both the religious and nonreligious can exemplify spiritual
excellence is a point of agreement between my approach and most of those
surveyed above.
Indeed, my view goes further than this, allowing that individuals who are
adherents of different religions can equally possess the virtue of spiritual
excellence. It is not required for practicing spiritual excellence that one’s
worldview be true. Nor is it required that one have strong justification for
believing this worldview. In these respects, my account is especially reso-
nant with Hadot’s description of the spiritual life. What is required of the
epistemic status of one’s worldview is only that one has at least ambiguous
evidence for it. This will allow, as argued in Chapter 3, for one to either
believe or assume the contents of this worldview to be true in a way that can
guide one’s actions. This is part of what I mean by “making use” of a world­
view. The worldview needs to be internalized in much the way described by
Hadot and Wynn, and my own view is that a requirement of this internal-
ization is that the claims of the worldview are regularly assumed or believed.
Internalizing and living in accordance with a worldview also requires
spiritual practices. This is why I refer to the “skillful” use of a worldview.
I agree with Cottingham, McPherson and Hadot/Wynn about the centrality
of practice to spirituality. I will return in Section 2 to describe an example of
a theistic spiritual practice that can enable a person to internalize a world­
view, but many of the examples identified by these authors are good candi-
dates as well. I will also comment in a general way later in this section on
some of the features that may be a part of many practices that can be used to
internalize worldviews.
The proper internalization of a worldview, when performed in ac­cord­
ance with spiritual excellence, alters the practitioner’s experiences. In par-
ticular, it promotes their experience of transformative awe, as suggested by
Peterson and Seligman, Cottingham, and McPherson. Empirical research
provides strong support for the link between spirituality and religion and
awe. Keltner and Haidt (2003), whose work reignited interest in awe among
contemporary psychologists, describe awe as both a “spiritual emotion” and
164 Faith, Flourishing, and Agnosticism

the “prototypical emotion of religious conversion.” Both religiosity and spir­

it­ual­ity are predictive of dispositional and state awe, although research has
found that spirituality is an even stronger predictor of dispositional awe
than religiosity (Kearns and Tyler forthcoming). When people are asked to
recall a spiritual experience, this activity leads them to experience awe
(Preston and Shin 2017). Experiencing awe is likewise related to spirituality.
Experiencing awe has been shown to lead people to adopt more spiritually
oriented goals (Van Cappellen et al. 2013) and to be more inclined to adopt
supernatural beliefs (Valdesolo and Graham 2014). If this sounds like a
reinforcing cycle, it is. Researchers have developed the upward spiral model
precisely to explain the ways in which religious engagement and positive
emotions, including awe, tend to be mutually reinforcing (Van Cappellen
et al. 2021).
Awe itself is a powerful emotion, typically experienced in the presence of
something great or vast (Keltner and Haidt 2003).1 Researchers refer to it as
an “epistemic” emotion because the kinds of objects that lead people to
experience it are rich in information (Valdesolo and Graham 2014).
Experiences of awe are typically positive, and people who experience awe
want their experience to continue. Awe is associated with a dropping of the
jaw and an open, gaping mouth (Shiota et al. 2003). In experiencing awe, we
are stunned by something, enraptured by it. Among the most common elic-
itors of awe is natural scenery, especially panoramic views, which has led
some authors to defend a “nature-­first” theory of awe’s evolutionary origins.
In a hunter-­gatherer context, it would be extremely beneficial for small
groups to find and maintain a safe haven from which they could easily see
approaching enemies. Thus, tending to have a positive and enjoyable expe-
rience of awe directed toward sweeping views, especially one that also
prompted behaviors that knit the group even more tightly together and kept
them in an elevated location, could have been highly adaptive (Chirico and
Yaden 2018).
This leads us directly to the transformative benefits of awe, especially the
aspect of awe called “connectedness.” People with stronger tendencies to
experience awe are rated as more humble by their friends (Stellar et al.
2018), display more generous behavior (Piff et al. 2015), and have stronger
prosocial dispositions (Guan et al. 2019). When experiences of awe are
experimentally induced, they lead people to present a more balanced

1 The next few paragraphs borrow with permission from (Byerly 2021).
 Spiritual Excellence 165

account of their own strengths and weaknesses (Stellar et al. 2018), to dis-
play more helping behavior (Piff et al. 2015), to display less aggressive
behavior toward others (Yang et al. 2016), and to be more willing to endure
unpleasant experiences to obtain a desired goal (Jiang et al. 2018).
The leading explanation of how experiencing awe prompts these social
effects has to do with the “small self ” (Perlin and Li 2020). When we experi-
ence awe, we experience ourselves as smaller and the world beyond us as
larger. This can be quite literal: people who experience awe judge their own
bodies to be smaller in size (van Elk et al. 2016). Yet it is also figurative: the
self and its concerns are less salient, and the world beyond the self more
salient, for those who experience awe. Crucially, in awe, we also experience
our smaller self as more connected with the larger world. People who expe-
rience awe report that they feel themselves to be “part of a greater entity”
(Piff et al. 2015). When people experience awe, they report a greater sense
of connection to groups they belong to, to their nation, and to their species
(Shiota et al. 2007). They experience greater oneness with others and friends
(Van Cappellen and Saroglou 2012) and feel more integrated into their
communities (Bai et al. 2017). Astronauts experiencing awe in spaceflight
report a greater sense of connection both to other people and to the earth in
general (Yaden et al. 2016). In this sense, awe is referred to as a “self-­
transcendent” emotion (Yaden et al. 2016). When we experience it, we tran-
scend ourselves by experiencing our small selves as connected to larger
wholes. Several studies have confirmed that the transformative social effects
of experiencing awe are mediated by this experience of connectedness (Bai
et al. 2017; Luo et al. 2022; Piff et al. 2015). Intense experiences of awe cause
or include an experience of feeling connected to a large whole, which in
turn promotes more prosocial, virtuous behavior.
Individuals characterized by the virtue of spiritual excellence, then, make
skilled use of a worldview for which they have ambiguous evidence or bet-
ter in order to cultivate these kinds of transformative experiences of awe
and connectedness, which help them to live more virtuously in general.
Thus, in agreement with all of the accounts surveyed above, spiritual excel-
lence is conspicuously concerned with moral improvement. And with
Cottingham, and to a lesser extent McPherson, this transformation is
achieved particularly via experiences of awe and accompanying feelings of
oneness with larger wholes.
The idea that developing this sort of virtue of spiritual excellence would
be a possibility for human beings is not especially surprising. It is well
known that there are individual differences in how prone people are to
166 Faith, Flourishing, and Agnosticism

experience awe, with some more inclined to experience more frequent and
powerful awe and others less so. It is only because of such individual varia-
tion that the results noted above concerning dispositional awe are possible.
Moreover, it is also known that there are skill- or ability-­like features,
regarding which people differ, that are predictive of how awe-­prone individ-
uals are. Probably the best-­ known example is the trait of absorption
(Tellegen and Atkinson 1974), which has to do with individuals’ tendencies
to become deeply immersed and engaged with the objects of their attention,
often in a multisensory or cross-­modal way. People who are high in trait
absorption are very imaginative and creative, and they apply these powers
to their engagement with both mental and external objects in a way that
leads them to experience more frequent and powerful awe (Maij and van
Elk 2018; Maij et al. 2018).
There is also reason to think that individuals can improve with respect to
abilities of these kinds that facilitate their experience of awe. In fact, this is
basically what happens for many individuals who participate in religious
communities for extended periods of time. They acquire skills in making
use of the worldviews associated with these traditions, which enable them
to have more immersive, imaginative, and transformative experiences of
the world.
Anthropologist Tanya Luhrmann is probably the single scholar who has
done the most work, in collaboration with research partners, to advance
understanding of both absorption and training and how they influence
spiritual experiences, including awe. Luhrmann emphasizes that viewing
and experiencing the world in accordance with worldviews that incorporate
invisible others, especially caring and powerful invisible others such as the
God of minimal theism, is not an easy task. Those who wish to adopt a
“faith frame”—a “mode of thinking in which gods and spirits really matter”
(2020: 21)—must make a “decision to enter into another mode of thinking
about reality that calls on the resources of the imagination to reorganize
what is fundamentally real and that exists in tension with the ordinary
expectations of everyday reality” (23). They must “think with the faith frame
as much as they can, despite how easy it can be to get distracted or discour-
aged, despite the competition from and contradictions of the everyday” (22).
While some people have more natural talent for engaging in this kind of
activity—­they are more naturally high in trait absorption—­training also
matters, even for those already high in absorption. For example, Luhrmann
describes how the cultivation of inner senses is often used to promote more
vivid and transformative spiritual experiences. Inner sense cultivation
 Spiritual Excellence 167

frequently uses what Luhrmann calls “interaction” and “sensory enhancement”

(2020: 73–4). Interaction involves back-­and-­forth engagement with what
the practitioner is imagining, while sensory enhancement involves using
multiple inner senses, such as imagining the smell, color, and sounds
associated with the object of imagination. Luhrmann has found that when
individuals in a religious tradition are instructed to employ all of their
senses to engage with a scriptural passage, this leads to more unusual spir­it­
ual experiences than when individuals instead listen to academic lectures
about that passage. In a similar way, other researchers have found that even
giving participants the simple instruction to “focus on the details of their
surroundings” when sitting in an arboretum can lead them to experience
more powerful experiences of awe (Ballew and Omoto 2018).
What all of this suggests is that when it comes to cultivating transforma-
tive spiritual experiences, some people have more natural talent than others,
and there are techniques that can be used to become better at it. This is not so
different from other areas of human activity that admit of degrees of exper-
tise or excellence. My own focus here is simply on one particular dimension
of spiritual experience cultivation—­namely, the cultivation of transforma-
tive experiences of awe through making use of worldviews and techniques
derived from religious or spiritual traditions. It should be expected that
some people will be better at this kind of activity than others and that there
are ways to practice becoming better at it. Individuals who are more charac-
terized by spiritual excellence will tend to be more adept comfortable m­ aking
use of worldviews for which they have ambiguous evidence or better to
cultivate experiences of awe; they will also find this easier to do and will be
able to focus on engaging in this practice while resisting distractions. We
might also expect that, as with individuals high in absorption, they are able
to recruit the resources of imagination and the use of inner senses in order
to prompt these experiences. Ultimately, they are able to prompt experiences
of themselves and their relation to the world that are shaped by their faith
frame and that lead them to feel connected to larger wholes.
To qualify as spiritual excellence, the experiences produced by using
these techniques must promote awe of what is indeed awesome. As a guide
to thinking about the extension of what is awesome, we might imagine
human beings perfecting their tendencies to experience awe over time in a
way that is analogous to how philosophers of science sometimes talk about
the idealization of science. Whatever such idealized awe would find to
be awesome is indeed awesome. On moral exemplarist views such as
Zagzebski’s (2017), the ultimate test of what is admirable is admiration;
168 Faith, Flourishing, and Agnosticism

likewise, we may allow that the ultimate test of what is awesome is what
people over time will reflectively find awesome. This, I suggest, leaves much
room for a diversity of awe-­elicitors, including those that are among the
most common awe-­elicitors today: beautiful natural scenery, heroic and
virtuous acts, childbirth, views of earth from space, and so on.2 Spiritual
excellence will lead individuals to feel awe for paradigmatic awe-­elicitors
such as these.
Spiritual excellence must also promote genuine moral transformation. If
a person employs a worldview and uses techniques, such as inner sense cul-
tivation, that lead them to experience feelings of connectedness but these
feelings fail to promote their moral growth or even corrupt them morally,
then their practice will not qualify as spiritual excellence. I think this possi-
bility must be acknowledged as a genuine danger. Anytime it is claimed that
religion or spirituality somehow qualifies as a virtue, objections will be
raised about the ills of religion. David McPherson (2017) discusses this con-
cern under the banner of what he calls the “social peace objection,” citing
the work of David Hume in particular. The objection alleges that spirituality
cannot be a virtue because of all of the ills that spirituality, particularly as
practiced in religious traditions, has led to over millennia. McPherson
responds that spirituality can be practiced well or poorly, and that the virtue
of piety must “avoid certain problematic directions the religious impulse
can take” and must instead promote “achieving ethical and spiritual fulness”
(2017: 83).
I agree with this basic response to the concern. In my own case, focusing
as I am on experiences of connectedness, I am inclined to suggest that the
more problematic experiences of connectedness that might be prompted
through practices similar to those characteristic of spiritual excellence may
be experiences of connectedness that knit certain human groups together,
but only at the expense of opposition toward outgroups. It is not difficult to
imagine worldviews being made use of that prompt individuals within cer-
tain groups to feel connected to each other and opposed to other individu-
als or groups. Indeed, there is a body of research linking religion and
religiosity with prejudice (Rowatt et al. 2014). To guard against this kind of
danger, my suggestion is that the virtue of spiritual excellence focuses on
forging experiences of connectedness to wider wholes, such as humanity,
the earth, or reality.

2 For more on what is known regarding the nature of awe-­elicitors, see (Byerly 2019b).
 Spiritual Excellence 169

Much as spiritual excellence must lead to experiences of awe for what is

indeed awesome, it must likewise prompt experiences of connectedness to
what the experient is indeed connected to. I do not think this is a particu-
larly difficult requirement to meet. We are all, in some way or other, con-
nected to other human beings, to the earth, and to reality as a whole. The
experience of connectedness can be a vague one that feels roughly the same
regardless of how one conceptualizes the precise way one is connected to
these larger wholes. It is this vague experience of connectedness that I am
primarily concerned with.
There may be sharper, more differentiated and specific experiences of
connectedness that can be facilitated through internalizing certain world­
views. A paradigmatic example here is certain kinds of mystical experiences
that seem to involve a loss of ego, in which one experiences a kind of con-
nectedness (if that is even the right term) with wider wholes that eliminates
one’s own distinct being (Hood 2017). Whether spiritual excellence can lead
to these kinds of experiences of connectedness or other specific and differ-
entiated experiences of connectedness, as opposed to the vague experience
of connectedness that is my focus, depends on whether people really are
connected to the larger wholes they experience such connectedness to in
these specific ways. I won’t argue here that people aren’t connected to larger
wholes in these specific ways but will only point out that it is more likely a
priori that they are connected in some way or other than that they are con-
nected in these precise ways. So, it is more likely that spiritual excellence
can be practiced in a way that gives rise to vague experiences of connected-
ness than that it can be practiced in a way that gives rise to these more spe-
cific kinds of experiences of connectedness. My own focus, then, is on the
skillful use of worldviews to promote transformative experiences of a vague
sort of connectedness with wide wholes, such as all of humanity, the earth,
the universe, or reality.
The main aim of this section has been to describe in some detail the char-
acter trait that I am calling the virtue of spiritual excellence. I explained in
Chapter 4 why I regard this trait as a virtue. This is because it involves valu-
ing well some things that are indeed valuable—­chiefly, experiences of awe
and connectedness as well as moral transformation—­in a way that is char-
acteristic of human beings. A subsidiary aim in this section has been to
defend the idea that acting in accordance with the trait of spiritual excel-
lence promotes growth in other more standard virtues. This amounts to
claiming that cultivating experiences of connectedness with large wholes by
making skilled use of a worldview for which one has ambiguous evidence or
170 Faith, Flourishing, and Agnosticism

better tends to promote virtue development. This claim has been defended
here by referring to empirical studies that document the role that such
experiences of connectedness play in explaining the link between experi-
ences of awe and prosocial tendencies and behaviors reflective of standard
virtues, such as generosity and kindness. I have therefore offered reason
to think both that spiritual excellence is a virtue and that acting in
­accordance with spiritual excellence tends to promote the development
of other virtues.
My overall aim in this chapter, however, is not focused on spiritual excel-
lence in the abstract but, more specifically, on how individuals with ambig-
uous evidence for the God of minimal theism can develop spiritual
excellence through engaging in practices of faith toward God. To demon-
strate this, I need to address how particular faith practices can be used by
people who have ambiguous evidence for God in such a way as to enable
them to make skilled use of the worldview of minimal theism in order to
prompt the kinds of experiences of connectedness characteristic of trans-
formative awe. The descriptions above probably give some indication of
how the worldview of minimal theism may be used to cultivate spiritual
excellence. But the question remains as to whether it is a genuine possibility
for individuals with ambiguous evidence for God to make salutary use of
minimal theism in this way. Can individuals with ambiguous evidence for
God really make use of minimal theism in practices of faith that promote
their experience of transformative connectedness? Showing that this is
indeed a genuine possibility is my aim in the next section.

2 Theistic Faith Practices for Cultivating

Spiritual Excellence

I maintain that individuals with ambiguous evidence for the God of mini-
mal theism can cultivate the character trait of spiritual excellence by making
skilled use of the worldview of minimal theism. They can do this, I suggest,
in the standard Aristotelian fashion described in Chapter 5, by practicing
acts characteristic of spiritual excellence. That is, they can make skilled use
of minimal theism to promote experiences of the sort of connectedness
with large wholes characteristic of morally transformative awe. This helps
them develop or maintain spiritual excellence, which in turn helps them
develop or maintain other more standard virtues, such as kindness or
generosity.
 Spiritual Excellence 171

Why think individuals with ambiguous evidence for God can make
skilled use of minimal theism to promote experiences of connectedness?
Here again, my argument is empirical, and in a way that parallels my
approach in the previous chapter. There, I argued that individuals with
ambiguous evidence for God can accept God’s love for them in a way that
promotes their mental well-­being by referring to research on agnostics’
acceptance of God’s love and its relationship to their mental well-­being.
I will argue here that individuals with ambiguous evidence for God can
cultivate experiences of connectedness by making use of minimal theism,
again via referring to experimental research on agnostics’ experiences of
connectedness using a theistic faith practice.
To set the stage for reporting this research, it is important to note its
exploratory and original nature. While there is a robust body of research,
surveyed briefly in the previous section, that indicates the transformative
benefits of awe, there has been much less research on how to intervene to
prompt individuals to experience more frequent or powerful awe. In a
recent review article, Chen and Mongrain (2021) raise the question of what
can be done to prompt more frequent and powerful experiences of awe in
daily life, and they contend that “here is where the research on awe falls
short” (3). They speculate that “the application of brief interventions with
credible rationales for daily practice” could be a valuable source of promot-
ing experiences of awe, specifically suggesting that, among other possibili-
ties, “spiritual quests or practices emphasizing one’s place in a greater
universal order could further reinforce the experience of something greater
than oneself ” (6). The research reported below takes this suggestion seri-
ously, investigating a brief, two-­and-­a-­half-­minute intervention for promot-
ing experiences of connectedness that draws on theistic ideas.
One source of limitations for research on interventions to promote awe is
the lack of tools for measuring awe experiences. Often, researchers have
either measured dispositional awe—­the general tendency to experience fre-
quent and powerful awe—­or single-­item questions about how much awe
participants were feeling, or they have made reasonable assumptions about
which experimental conditions likely prompt greater awe in participants.
But only recently has a thorough measure of state awe been developed that
can enable researchers to compare individuals’ levels of awe and its various
subdimensions at different times (Yaden et al. 2019). This measure, devel-
oped rigorously and based on themes of awe experiences identified in the
experimental literature, includes six factors: slowing or loss of time; perceiv-
ing oneself to be smaller or less significant; feeling connected; experiencing
172 Faith, Flourishing, and Agnosticism

something vast; physiological sensations such as goosebumps; and difficulty

comprehending what one is encountering. The research reported below
employed a shortened version of this scale, focusing on the factor of con-
nectedness. The full subscale for this factor includes five items: “I had a
sense of being connected to everything”; “I felt a sense of communion
with all living things”; “I experienced a sense of oneness with all things”;
“I felt closely connected to humanity”; and “I had a sense of complete
connectedness.”
To test the effectiveness of an intervention for promoting experiences of
awe and connectedness measured with this instrument, I recruited 460
­participants—230 theists and 230 agnostics. The theists claimed to believe that
God exists, while the agnostics claimed to neither believe that God exists
nor that God does not exist, where God was understood to be a supernatu-
ral creator of the universe who cares lovingly for all creatures within it.
I wanted to investigate theists alongside agnostics for two reasons. First,
because little research on awe interventions has been done in general,
I wanted to test whether this simple intervention would at least be effective
for theists, even if it wasn’t effective for agnostics. Second, studying theists
alongside agnostics provides a useful point of comparison when consider-
ing how agnostics responded to the intervention, as elaborated below.
The sample was 63% female, with a mean age between 35 and 44 years
old. All participants were US residents; 77% identified as White, 12% identi-
fied as Black, and fewer than 3% identified as any other racial category.
There were 121 participants whose highest level of educational attainment
was a high school diploma or equivalent; 191 had attained a college degree;
9 had not attained a high school diploma; and 132 had attended some col-
lege but had not attained a degree. The full data set is available online
(Byerly 2023a).
The group of theists and the group of agnostics were randomly divided
into two groups—­an intervention group and a control group. The control
groups viewed a series of five images, each for thirty seconds. The images
were taken from OASIS—­the Open Affective Standard Image Set (Kurdi
et al. 2017)—and selected because they were images of known awe-­elicitors.
The images were of a flower, a galaxy, a lightning bolt, the moon reflecting
off water, and a sleeping baby. Before viewing the images, participants were
given the following instructions: “You will now be asked to complete a short
CONTEMPLATIVE EXERCISE. You will be shown a series of images. Each
image will be accompanied by a short phrase. USE THE PHRASE TO HELP
YOU CONTEMPLATE EACH IMAGE FOR 30 SECONDS.” For each
 Spiritual Excellence 173

image, the phrase was “Focus on the details of the image.” Participants in
the intervention groups viewed the same series of images and had the same
instructions, but the phrases they were given were different. The phrases
given to the intervention groups were: “The flower bursts open with God-­
given abundant life”; “The swirling galaxy reveals God’s extraordinary wis-
dom and design”; “The lightning’s strike displays God’s magnificent power”;
“God calls to us and lights our path”; and “Every creature is treasured and
loved intimately by God.”
Before completing the contemplative exercise, both groups completed
abbreviated versions of the awe experiences scale described earlier (α = .86),
responding to the three top-­loading items from each of its subscales (for the
connectedness subscale, these were the first three items listed earlier). They
also completed the measure of accepting God’s love described in the previ-
ous chapter (α = .95) and answered demographic questions and a question
about how they evaluate their evidence for God’s existence using a 100-­
point sliding bar. After completing the exercise, both groups completed the
abbreviated measure of state awe, focused on how they were feeling during
the exercise, and answered single-­item questions about how comfortable
they were during the exercise, how easy they found it to do the exercise, and
how engaged they were with it. They also completed the Sensory Delights
Scale (Luhrmann 2020), designed to measure absorption (α = .94). They
were given a (fake) opportunity to sign up to receive information about vol-
unteer opportunities in their local area securely via email. They could select
up to four types of volunteering opportunities to receive information about,
or they could select “Please do not contact me about volunteer opportunities.”
It was my expectation going into the experiment that both theists and
agnostics in the control group would experience greater feelings of awe and
connectedness following the contemplative exercise. This expectation was
based on two facts. First, the groups in this condition were engaging with
known awe-­elicitors. Second, the phrases they were given, instructing them
to “focus on the details” of the images, have been successfully used in previ-
ous research to promote participants’ feelings of awe when sitting in an
arboretum (Ballew and Omoto 2018). I was, however, wrong in this expec-
tation. Paired sample t-­tests revealed that there was no significant difference
between agnostics’ pre- and post-­awe or pre- and post-­connectedness in the
control condition. Theists experienced significantly lower awe in the control
condition after the contemplative exercise than before it (t(97) = −2.87;
p < .01), with a Cohen’s d of −.23. Theists’ connectedness was not significantly
different following the exercise than preceding it.
174 Faith, Flourishing, and Agnosticism

My expectation for the intervention groups was that theists in the inter-
vention condition would experience greater awe and greater connectedness
following the intervention, while agnostics as a whole would experience less
awe and connectedness following the intervention. The expectation for
agnostics was driven by the expectation that, as a whole, agnostics would be
less comfortable, at ease, and engaged with the exercise in the intervention
condition than theists would be. I expected, however, that agnostics who
were more comfortable, more at ease, and more engaged during the exercise
would be more likely to experience greater awe following the intervention.
Crucially for purposes of this chapter, I also expected that agnostics’ ac­cept­
ance of God’s love for them would moderate whether they experienced
greater awe and connectedness following the intervention than before it.
The idea is that agnostics who were more willing to adopt a cognitive stance
reflecting faith toward God would be in a better position to experience
enhanced feelings of awe and connectedness as a result of contemplating
the images using the theistic phrases.
These hypotheses were largely confirmed. Theists in the intervention
condition experienced significantly greater connectedness following the
intervention (t(126) = 4.02; p < .001), with a Cohen’s d of .32 and a mean
difference of .48—the largest effect size in the study. Theists did not experi-
ence significantly greater awe, however. Agnostics as a whole experienced
significantly lower awe (t(107) = −2.49; p < .05) and connectedness (t(107)
= −3.19; p < .01) following the intervention. Table 7.1 lists several variables
that were related to change in awe or change in connectedness in the full
intervention sample, as well as whether being a theist or agnostic was related
to these variables in the treatment condition. For change in connectedness,
participants’ comfort with the exercise, engagement with it, trait absorption,
state of connectedness prior to the exercise, and acceptance of God’s love
were all significant. For change in awe, only participants’ engagement with
the exercise and state awe prior to the exercise were significant. Theists in
the treatment condition were significantly more comfortable, at ease, and
engaged with the exercise; they also scored higher in trait absorption, awe
and connectedness prior to the exercise, and acceptance of God’s love than
did agnostics in the treatment condition.
To evaluate whether agnostics with a more faithful orientation toward
God tended to experience an increase in awe or connectedness as a result of
the intervention, I conducted multiple regressions for awe and connected-
ness using several independent variables and an interaction between the
 Spiritual Excellence 175

acceptance of God’s love measure and participants’ condition (whether

­control or treatment) for the full sample of agnostics. This is a complex
­procedure, but vital for understanding the significance of the study for this
chapter. What we are considering here is whether agnostics who tended to
adopt a more accepting orientation toward God’s love also tended to experi-
ence a greater upswing in awe or connectedness as a result of engaging in
the contemplative exercise in the intervention condition in comparison to
the control condition, when controlling for several other variables. The
other variables controlled for were participants’ sex and age, comfort, ease,
and engagement with the exercise, trait absorption, and awe prior to the
exercise (in the case of regressing for change in awe) or connectedness prior
to the exercise (in the case of regressing for change in connectedness). These
variables are controlled for primarily because of their potential significance
for change in awe or connectedness, as revealed in Table 7.1.
In the case of awe, there was not a significant interaction effect between
agnostics’ acceptance of God’s love and their being in the intervention con-
dition. In the case of connectedness, however, there was. When controlling
for agnostics’ sex and age; comfort, ease, and engagement with the exercise;
absorption; and level of connectedness prior to the exercise, there was a sig-
nificant interaction effect between their acceptance of God’s love and their

Table 7.1 Key bivariate correlations

Change in Change in Theism in

Connectedness—­ Awe—­ Intervention
Intervention Condition Intervention Condition
Condition

Comfort .15* .09 .32***

Ease .12 .06 .18**
Engagement .27*** .18** .38***
Absorption .23*** .12 .28***
Awe Prior to .12 −.17** .43***
Exercise
Connectedness −.27*** −.12 .25***
Prior to Exercise
Accepting God’s .33*** .07 .76***
Love

Note. * = p < .05; ** = p < .01; *** = p < .001

176 Faith, Flourishing, and Agnosticism

condition as predictors of their change in connectedness.3 Agnostics who

were in the intervention condition, using the theistic prompts described
above to engage in contemplation of the images, tended to experience an
increase in connectedness if they were more accepting of God’s love, but a
decrease in connectedness if they were less accepting of God’s love (B = .51,
SE = .12, p < .01).
The effect can be represented graphically (see Figure 7.1). The left-­hand
side of the figure represents agnostics’ change in connectedness in the con-
trol condition, while the right-­hand side of the figure represents agnostics’
change in connectedness in the intervention condition. The solid line shows
the trajectory of change in connectedness for agnostics who score at two
standard deviations above the mean for acceptance of God’s love. The dot-
ted line represents agnostics who score at two standard deviations below the
mean for acceptance of God’s love. And the dashed line represents agnostics
who score at the mean for accepting God’s love. The use of two standard
deviations above the mean for agnostics’ acceptance of God’s love (6.3) is
useful as a point of comparison with theists, because theists’ mean for
acceptance of God’s love ( X = 6.21) is roughly equivalent to this value. In
other words, agnostics who are two standard deviations above the mean for
agnostics in acceptance of God’s love score at about the mean for acceptance
of God’s love for theists. About 5% of agnostics in this sample scored this
high. Figure 6 shows that these agnostics tend to respond to the interven-
tion exercise in much the way theists do on average. They experience a sig-
nificantly greater increase in feelings of connectedness in the intervention
condition in comparison with the control condition.
We can be a bit more exact. We can use the regression equation depicted
in Figure 7.1 to predict the average increase in connectedness that an agnos-
tic who is highly accepting of God’s love would experience as a result of the
intervention. To do this, we must specify values for the other variables in
the equation. The simplest way to do this is to set them all at their means.
When we do this, the equation predicts that, on average, an agnostic who
scores two standard deviations above the mean for agnostics on accepting
God’s love will have a decrease in connectedness in the control condition of
.28, while they will have an increase in connectedness in the intervention
condition of .21. This may be compared to the average gain in connectedness

3 The interaction effect is also present when these other variables are not controlled for. We
simply get a more informative regression equation by including them, and the unique signifi-
cance of the effect is confirmed by the fact that it remains when they are included.
 Spiritual Excellence 177

3 Moderator: Acceptance
6.3
2 3.8
1.3
Connection Change

–1

–2

–3

Figure 7.1 Moderation of change in connectedness

in the control condition experienced by theists (.09) and the average gain in
connectedness in the intervention condition experienced by theists (.48),
which is a statistically significant difference (p < .05). In other words, we
find that just as theists tend to experience significantly greater increases in
connectedness when undergoing the intervention compared to the control
exercise, agnostics who score higher on accepting God’s love do so as well.
Increasing the values chosen for other variables, such as agnostics’ engage-
ment with the exercise or absorption, yields even higher predicted growth
in connectedness. For example, the predicted value for change in connect-
edness for agnostics scoring at one standard deviation above the mean for
comfort, ease, engagement, and absorption as well as two standard devia-
tions above the mean for accepting God’s love is .63 in the control condition
and 1.13 in the intervention condition.
These results are supportive of the main contention of this section that
individuals with ambiguous evidence for God can make use of the world­
view of minimal theism in order to cultivate experiences of connectedness
characteristic of transformative awe. As with the argument in the previous
chapter, it helps here to make the charitable assumption, supported by the
arguments given in Chapter 2, that most agnostics have ambiguous evi-
dence for God. Agnostics in this study, most of whom probably have ambig-
uous evidence for God, tended to be able to cultivate greater experiences of
connectedness through making use of the worldview of minimal theism
when contemplating awe-­ inspiring pictures than when instead simply
focusing on the details of these images—­if they had more of an orientation
of faithful acceptance of God’s love. If they were more inclined to say that
they accept or assume or act as if or have faith that God loves them, taking
on board positive cognitive attitudes toward God’s existence and love for
178 Faith, Flourishing, and Agnosticism

them that may fall short of belief, then they tended, like theists in the study,
to experience greater connectedness in the intervention condition.
We should, of course, exercise caution in interpreting the results of the
study. The prediction that an agnostic who has a more faithful attitude
toward God’s love for them will experience a gain in connectedness through
the intervention is a prediction about what such agnostics will experience
on average. Yet this gain itself is small, and the regression equation also
exhibits margin for error (S = 1.10). This allows that some agnostics with a
more faithful orientation toward God may experience less of a feeling of
connectedness following the intervention than prior to it, or that they will
experience little or no gain in feelings of connectedness. But it also allows
that most such agnostics will experience positive gains. Additionally, of
course, this is only one study with one relatively small sample, and more
data would be very welcome. Still, these findings are enough to support the
modest contention of this section that individuals with ambiguous evidence
for God can cultivate experiences of connectedness by making skilled use of
the worldview of minimal theism. This is especially so when we consider
how altering additional variables, such as absorption or engagement
with the exercise, tends to increase agnostics’ predicted change in
connectedness.
It should be kept in mind how short and simple this intervention was.
Participants were asked to view a series of images and contemplate these,
guided by theistic prompts, for two and a half minutes. They did not receive
prior training in how to do this. It was a one-­off exercise. They were not
engaging with the natural world itself, but with images of it. They were not
told to use multiple inner senses in their contemplative efforts. They did not
elect to undertake this exercise of their own accord. In all these ways, we
might expect that practices like the one these participants undertook could
have a stronger effect for individuals with ambiguous evidence for God if
modestly modified. If participants were provided with training for how to
engage in tasks of this sort, if they engaged with more powerful awe-­
elicitors, if they used multiple inner senses, if they did this regularly for a
longer stretch of time, we might find that their efforts lead to even greater
changes in their experiences of connectedness. Indeed, given the data, we
should expect it. By engaging in these kinds of practices, individuals with
ambiguous evidence for God may, in standard Aristotelian fashion, cultivate a
general tendency to make use of a worldview—­minimal theism—­for which
they have ambiguous evidence in order to cultivate morally transformative
experiences of awe.
 Spiritual Excellence 179

The results of this study did indicate that there was a relationship between
experiencing greater feelings of connectedness and behaving more prosocially.
Participants who experienced stronger feelings of connectedness after their
contemplative exercise were more likely to request to receive information
about volunteer opportunities in their local area (r = .19; p < .001). This
finding coheres well with the findings surveyed in the previous section
indicating that experiences of connectedness tend to be predictive of prosocial
behaviors characteristic of virtues, mediating the relationship between awe
and such behaviors. It reinforces the idea that if individuals with ambiguous
evidence for God are able to act in accordance with spiritual excellence, this
can enable them to grow in other more standard virtues even if spiritual
excellence itself is not itself a virtue.
The results of this study, then, help to complete the main argument of
this chapter. As argued in the previous section, spiritual excellence may
itself be considered a virtue, and its exercise (whether virtuous or not) can
lead to growth in other more standard virtues. Yet in this section we have
seen that there is reason to think that individuals with ambiguous evidence
for God can cultivate spiritual excellence by engaging in practices of faith
toward God. By engaging and becoming absorbed with awe-­elicitors in a
way that is guided by theistic ideas accepted by faith, individuals with
ambiguous evidence for God can cultivate experiences of connectedness
they may not have been able to cultivate as well without making use of this
worldview. Individuals with ambiguous evidence for God can engage in acts
characteristic of spiritual excellence by making skilled use of the worldview
of minimal theism to cultivate transformative experiences of connected-
ness. By doing so, they can cultivate spiritual excellence and its downstream
effects on character in standard Aristotelian fashion.
 Conclusion

This book has identified four pathways whereby individuals with ambiguous
evidence for God may grow in virtue and flourishing by engaging in practices
of faith toward God. They may cultivate general virtuous tendencies to give
others the benefit of the doubt by giving God the benefit of the doubt; they
may grow indirectly in virtue by experiencing better mental health as a
result of accepting God’s love; they may cultivate the virtue of spiritual
excellence by making skilled use of the worldview of minimal theism to
experience transformative awe; and their acting in accordance with spiritual
excellence can further facilitate their growth in other more stand­ard virtues.
These results are in many ways only a beginning when it comes to explor-
ing the value of faith practices for agnostics and individuals with ambiguous
evidence for God. We began the book by noting that current research has
suggested that such individuals may be at a well-­being deficit in comparison
to confident theists and confident atheists. By identifying the four pathways
to developing virtue and flourishing described above, this book has contrib-
uted to exploring how faith practices might mitigate these deficiencies. But
there are many ways for future research to build upon these findings to
address this important topic more thoroughly.
First, there is a great deal of room for further empirical research address-
ing the faith practices that have been my focus, as well as other very similar
practices. I do not know of any empirical research that has focused specifi-
cally on giving others, including God, the benefit of the doubt in the way
that figures into the first pathway. Researchers could develop measures of
these tendencies and investigate their significance. We saw in Chapter 6 that
there has not been longitudinal research on agnostics’ acceptance of God’s
love and its relationship to their mental health and virtue possession. The
same is true of agnostics’ use of practices that make use of minimal theism
in order to cultivate transformative awe experiences. Additionally, it would
be valuable for researchers to explore the effectiveness of interventions
designed to increase uptake of these faith practices among agnostics, and
the effects of such uptake on their virtue and flourishing.

Faith, Flourishing, and Agnosticism. T. Ryan Byerly, Oxford University Press. © T. Ryan Byerly 2023.
DOI: 10.1093/oso/9780192865717.003.0009
 Conclusion 181

Second, it would be valuable to investigate whether engaging in practices

of faith within a community of like-­minded others, and not just individually,
is conducive to virtue development and flourishing. I would envision the
investigation of this topic taking interdisciplinary shape in which contribu-
tions from both philosophy and psychology could be valuable. A growing
number of philosophers have suggested that groups, not just individuals,
can possess and exercise virtues (Lahroodi 2019). One might wonder, then,
whether by engaging in practices of faith with like-­minded others one might
cultivate social forms of virtue with their own value, which goes beyond the
value displayed in one’s own individual virtues and contributes to the flour-
ishing of the community one participates in, not just to one’s own individual
flourishing. In some of my own research with members of Christian con-
gregations, I have found that how congregants perceive the virtuousness of
their congregations is uniquely predictive of their satisfaction with that
community as well as their own spiritual well-­being, satisfaction with life,
and experience of their lives as meaningful (Byerly et al. 2022). Future
research might investigate whether agnostics’ participation in faith prac-
tices within a community of like-­minded others could have similar value for
the community and its members.
Thinking somewhat in the other direction, researchers might consider
the various ways in which engaging in faith practices within a like-­minded
community might enhance the practitioner’s ability to engage in and benefit
from those practices themselves. In Chapter 7, we considered the idea that
there may be spiritual technologies that consist in specific techniques for
making use of worldviews to cultivate experiences of transformative awe. It
is a live hypothesis, one we saw to be associated with the work of Hadot and
Wynn, that spiritual traditions, including religious traditions, retain such
technologies, and that one’s access to them could be enhanced through
engagement with such communities which preserve these traditions. It is also
known that communities can exert influence over their members’ conduct
through more and less explicit means, for ill or good (Vessiére et al. 2019).
Are there communities available that provide sources of encouragement
and influence to spur on individuals with ambiguous evidence for God to
engage in practices of faith that are successful in helping them grow in virtue
and flourishing? Can participating in faith practices within a community of
fellow practitioners help one to get the most out of those practices? These
are important questions worthy of future investigation.
Finally, it would be valuable for future research to explore the value of
nontheistic faith practices alongside the kinds of theistic faith practices that
182 Faith, Flourishing, and Agnosticism

have been my focus. I have not sought to argue in this book that the faith
practices that are my focus are entirely unique in the sense that there are no
alternative nontheistic faith practices that can make comparable contribu-
tions to individuals’ virtue and flourishing. I did suggest in Chapter 5 that
individuals with ambiguous evidence for God who fail to give God the ben-
efit of the doubt through expressing praise, thanks, or contrition to God
would be failing to act in accordance with more general virtuous tendencies
to give others the benefit of the doubt. I also suggested that accepting God’s
love may make a unique contribution to indirect virtue development beyond
what is made by accepting certain others’ love for oneself. Yet these conten-
tions are compatible with thinking that there is comparable value to be
attained through other nontheistic faith practices. I have left the pathways
to virtue development that run through the virtue of spiritual excellence
even more wide open to making use of nontheistic faith practices. To the
extent that there are nontheistic worldviews for which an individual has
ambiguous evidence or better that they can make use of in order to cultivate
transformative experiences of awe, they can engage in nontheistic faith
practices to cultivate virtue and flourishing via the routes examined in
Chapter 7 by making use of this worldview.
Are there other worldviews that are good candidates for practicing spir­it­
ual excellence? I think there may be for some individuals. For instance, one
example is described by Philip Ivanhoe (2018) in his treatment of Neo-­
Confucian oneness. Ivanhoe highlights how Neo-­Confucian authors such as
Chen Hao (1032–1085) and Wang Yangming (1472–1529) affirmed a meta-
physical view according to which each person forms one body with all other
persons, creatures, and things. According to this way of thinking, most of us
wrongly think of our selves as too small. Rather than being contained within
our skin or our brains, our selves extend outward beyond the boundaries of
our individual bodies to include all people, creatures, and things in the uni-
verse. Our “faculty of pure knowing” retains awareness of this unity with all
else, and this is why we experience sympathy with other creatures’ distress.
But often our understanding of this unity is obscured. Coming to appreciate
it better can lead to our moral transformation—­to manifesting benevolence
(ren) in which we care for all others as for ourselves.
This Neo-­Confucian idea of the extended self may provide a “worldview”
in my terminology, one that individuals could make use of in order to culti-
vate transformative feelings of connectedness. The idea of the extended self
provides an account of how individuals fit within reality as a whole, which
indicates that individuals are indeed robustly connected to all other people,
 Conclusion 183

creatures, and things. As such, it seems ripe for being used as part of practices
aimed at cultivating experiences of connectedness. At least, it is ripe for this
among individuals who have ambiguous evidence or better for this view of
the self.
Another example is yielded by the foundational Buddhist principles of
impermanence and dependent co-­origination. According to impermanence,
everything is impermanent; nothing lasts or endures or has a fixed, enduring
essence. According to dependent co-­origination, everything depends for
its existence on everything else, and all is in constant flux. These ideas of
impermanence and dependent co-­origination apply to selves as much as to
anything else. Thus, the complementary doctrine of no self teaches “not
only the denial of a substantial, fixed entity we call the self but also a recog-
nition of the self and reality as processes in immanent relationship with one
another in their dynamic unfolding. The ‘great chain of being’ is dynami-
cally linked in a stream of creative processes in which nothing persists or
endures” (Davis 2014: 308). Understanding and learning to experience the
world in terms of these foundational ideas through practices such as mind-
fulness is thought to help individuals to overcome their wrongful attach-
ments to themselves and to other things in the world so that they can
eliminate their own and others’ suffering (see especially Bodhi 2011). We
might surmise that engaging in faith practices that make use of these ideas
could prompt individuals to experience greater feelings of connectedness.
Indeed, research does suggest that engaging in forms of mindfulness and
loving-­kindness meditation inspired by Buddhist thought can enhance
individuals’ feelings of connectedness with other people and the natural
world (Aspy and Proeve 2017).
I wish to briefly suggest that these kinds of nontheistic faith practices
needn’t be pursued exclusively from theistic faith practices but may in fact
be fruitfully combined with them. An individual with ambiguous evidence
for God needn’t choose only one route or the other, because the worldview
of minimal theism is not clearly incompatible with these nontheistic world-
views. A person might have ambiguous evidence both for minimal theism
and for either the view that they form one body with all other humans, crea-
tures, and things, or the view that all (except God, anyway; cf. McNabb and
Baldwin 2022: ch.1) is impermanent and depends for its existence on every-
thing else. If a person has ambiguous evidence or better for both minimal
theism and one of these other worldviews, then they could make use of both
in efforts to cultivate experiences of connectedness, thereby growing in
spiritual excellence. In fact, it may be that this sort of intermingling of
184 Faith, Flourishing, and Agnosticism

worldviews may be expected of those who exercise spiritual excellence. It

may be part of the skill they employ in cultivating experiences of connect-
edness that they are open to and able to creatively intermix elements of
worldviews that are often disassociated from one another.
I don’t mean, of course, to minimize genuine differences between fully
developed theistic worldviews and the detailed religious traditions of
Confucianism or Buddhism, which certainly do contain sources of tension
or outright conflict with each other.1 Rather, my suggestion is that the
particular, narrower ideas that one finds within these traditions that qualify
as worldviews in the sense developed in Chapter 7 might be fruitfully
employed in the pursuit of spiritual excellence by someone who has
ambiguous evidence for them. So, I suggest that future research should attend
to how agnostics and individuals with ambiguous evidence for other religious
or nonreligious worldviews can grow in virtue through engaging in non-
theistic practices of faith—­potentially even combining these with theistic
faith practices.
There is a great deal of room for future research to build upon the results
presented in this book. As it does, we may hope that it will become increas-
ingly clear how agnostics and others with ambiguous evidence for God or
other ultimate spiritual realities can grow in virtue and flourishing through
faith practices and other means.

1 On the relationship between Confucianism and theism, see (Sim 2017); on the relation-
ship between Buddhism and theism, see (McNabb and Baldwin 2022: ch.2).
 References

Adams, Robert. (2006). A Theory of Virtue. Oxford University Press.

Aghababaei, Naser, Blachnio, Agata, and Aminikhoo, Masoume. (2018). “The
Relations of Gratitude to Religiosity, Well-Being, and Personality.” Mental Health,
Religion & Culture 21, 4: 408–17.
Ainsworth, M. D. S. (1985). “Attachments across the Life Span.” Bulletin of the New
York Academy of Medicine 61, 792–812.
Alston, William. (2007). “Audi on Nondoxastic Faith.” In Rationality and the Good:
Critical Essays on the Ethics and Epistemology of Robert Audi, ed. Mark Timmons,
John Greco, and Alfred Mele, 123–41. Oxford University Press.
Anderson, Charity. (2021). “Divine Hiddenness: An Evidential Argument.”
Philosophical Perspectives 35, 1: 5–22.
Anderson, Tina. (2022). “From a Necessary Being to a Perfect Being: A Reply to
Byerly.” European Journal for the Philosophy of Religion 14, 1: 257–68.
Annas, Julia. (2011). Intelligent Virtue. Oxford University Press.
Aspy, Denholm and Proeve, Michael. (2017). “Mindfulness and Loving-Kindness
Meditation: Effects on Connectedness to Humanity and to the Natural World.”
Psychological Reports 120, 1: 102–17.
Baehr, Jason. (2011). The Inquiring Mind: On Intellectual Virtues and Virtue
Epistemology. Oxford University Press.
Bai, Y., Maruskin, L. A., Chen, S., Gordon, A. M., Stellar, J. E., McNeil, G. D., and
Keltner, D. (2017). “Awe, the Diminished Self, and Collective Engagement:
Universals and Cultural Variations in the Small Self.” Journal of Personality and
Social Psychology 113, 2: 185–209.
Baker-Hytch, Max. (forthcoming). “Natural Theology and Religious Belief.” In The
Cambridge Handbook of Religious Epistemology, ed. Jonathan Fuqua, John Greco,
and Tyler McNabb. Cambridge University Press.
Ballew, M. T. and Omoto, A. M. (2018). “Absorption: How Nature Experiences
Promote Awe and Other Positive Emotions.” Ecopsychology 10, 1: 26–35.
Barnes, Luke. (2019). “A Reasonable Little Question: A Formulation of the Fine-
Tuning Argument.” Ergo 6, 42: 1220–57.
Barnett, David. (2006). “Zif Is If.” Mind 115, 459: 519–66.
Barrett, Justin. (2004). Why Would Anyone Believe in God? Altamira.
Bassett, Rodney, Scott, Sarah, Bodgett, Renee, Barnhart, Mika, Batterson, Noelle,
Hall, Ashley, Pursel, McKennzie, and Emerson, Melissa. (2020). “May Grace
Abound: Making God’s Grace Cognitively Salient May Increase Reparative Action.”
Journal of Psychology and Theology 48, 3: 218–28.
Battaly, Heather. (2015). Virtue. Polity Press.
Baumann, Peter. (2021). “Sorry If! On Conditional Apologies.” Ethical Theory and
Moral Practice 24, 5: 1079–90.
186 References

Baumeister, Roy and Leary, Mark. (1995). “The Need to Belong: Desire for
Interpersonal Attachments as a Fundamental Human Motivation.” Psychological
Bulletin 117, 3: 497–529.
Bayne, Tim and Nagasawa, Yujin. (2006). “The Grounds of Worship.” Religious
Studies 42, 3: 299–313.
Beck, R. and McDonald, A. (2004). “Attachment to God: The Attachment to God
Inventory, Tests of Working Model Correspondence, and an Exploration of Faith
Group Differences.” Journal of Psychology and Theology 32: 92–103.
Benton, Matthew. (2021). “Hope, Knowledge, and Fallibilism.” Synthese 198:
1673–89.
Bering, Jesse. (2010). The God Instinct: The Psychology of Souls, Destiny and the
Meaning of Life. Nicholas Brealey Publishing.
Bishop, John and Perszyk, Ken. (2017). “The Divine Attributes and Non-personal
Conceptions of God.” Topoi 36: 609–21.
Bloeser, Claudia and Stahl, Titus. (2022). “Hope.” In Stanford Encyclopedia of
Philosophy, ed. Edward Zalta. The Metaphysics Research Lab, Stanford University.
Available at https://plato.stanford.edu/entries/hope/
Bodhi, Bhikkhu. (2011). “What Does Mindfulness Really Mean? A Canonical
Perspective.” Contemporary Buddhism 12, 1: 19–39.
Bogardus, Thomas, and Urban, Mallorie. (2017). “How to Tell Whether Christians
and Muslims Worship the Same God.” Faith and Philosophy 34, 2: 176–200.
Bøhn, Einar. (2019). God and Abstract Objects. Cambridge University Press.
Bourne, K., Berry, K., and Jones, I. (2014). “The Relationships between Psychological
Mindedness, Parental Bonding and Adult Attachment.” Psychology and
Psychotherapy: Theory, Research and Practice 87, 167–77.
Bowlby, J. (1969). Attachment and Loss, vol. 1: Attachment. Basic Books.
Bowlby, J. (1973). Attachment and Loss, vol. 2: Separation. Basic Books.
Bowlby, J. (1988). A Secure Base: Parent-Child Attachment and Healthy Human
Development. Basic Books.
Bradshaw, Matt and Kent, Blake. (2018). “Prayer, Attachment to God, and Changes
in Psychological Well-Being in Later Life.” Journal of Aging and Health 30,
5: 667–91.
Brownlee, Kimberly. (2013). “A Human Right Against Social Deprivation.” The
Philosophical Quarterly 63, 251: 119–222.
Brownlee, Kimberly. (2016). “Ethical Dilemmas of Sociability.” Utilitas 28, 1: 54–72.
Bruton, Samuel. (2003). “Duties of Gratitude.” Philosophy in the Contemporary World
10, 1: 1–5.
Buckareff, Andrei. (2005). “Can Faith be a Doxastic Venture?” Religious Studies 41,
4: 435–45.
Buckareff, Andrei. (2022). Pantheism. Cambridge University Press.
Budd, Malcolm. (2002). The Aesthetic Appreciation of Nature. Oxford University
Press.
Burling, Hugh. (forthcoming). “Do We Owe God Worship?” Religious Studies.
Byerly, T. Ryan. (2017). “Virtues of Repair in Paradise.” In Paradise Understood: New
Philosophical Essays about Heaven, ed. T. Ryan Byerly and Eric Silverman, 136–60.
Oxford University Press.
 References 187

Byerly, T. Ryan. (2019a). “From a Necessary Being to a Perfect Being.” Analysis 79,
1: 10–17.
Byerly, T. Ryan. (2019b). “The Awe-Some Argument for Pantheism.” European Journal
for Philosophy of Religion 11, 2: 1–21.
Byerly, T. Ryan. (2021). “How Awesome Natural Beauty Drops the Jaw but Lifts the
Spirit.” Psyche. Available at https://psyche.co/ideas/why-awesome-natural-beauty-
drops-the-jaw-and-lifts-the-spirit
Byerly, T. Ryan. (2022a). “Being Good and Loving God.” In Oxford Studies in
Philosophy of Religion, vol. 10, ed. Lara Buchak and Dean Zimmerman. Oxford
University Press.
Byerly, T. Ryan. (2022b). “The Transformative Power of Accepting God’s Love.”
Religious Studies 58, 4: 831–45.
Byerly, T. Ryan. (2023a). “An Intervention for Connectedness with Theists and
Agnostics.” Figshare. DOI: 10.6084/m9.figshare.21896781
Byerly, T. Ryan. (2023b). “Agnostics Who Accept God’s Love Experience Greater
Well-Being.”
Byerly, T. Ryan, Hill, Peter, and Edwards, Keith. (2022). “Others-Centeredness:
A Uniquely Positive Tendency to Put Others First.” Personality and Individual
Differences 186, A: 1111364.
Calvert, Sarah. (2010). “Attachment to God as a Source of Struggle and Strength:
Exploring the Association between Christians’ Relationship with God and Their
Emotional Wellbeing.” Dissertation. Massey University. Available at http://hdl.
handle.net/10179/1699
Carrico, C. P. (2012). A Look inside Firefighter Families: A Qualitative Study.
University of Northern Colorado: ProQuest Dissertations Publishing.
Chen, Susan and Mongrain, Myriam. (2021). “Awe and the Interconnected Self.”
Journal of Positive Psychology 16, 6: 770–8.
Cherniak, Aaron, Mikulincer, Mario, Shaver, Phillip, and Granqvist, Pehr. (2021).
“Attachment Theory and Religion.” Current Opinion in Psychology 40: 126–30.
Chirico, Alice and Yaden, David. (2018). “Awe: A Self-Transcendent and Sometimes
Transformative Emotion.” In The Function of Emotions: Why and When Emotions
Help Us, ed. Heather Lench, 221–33. Springer.
Cole, J. C., Rabin, A. S., Smith, T. L., and Kaufman, A. S. (2004). “Development and
Validation of a Rasch-Derived CES-D Short Form.” Psychological Assessment 16,
4: 360–72.
Coleman, Thomas, Sevinç, Kenan, Hood, Ralph, and Jong, Jonathan. (2019). “An
Atheist Perspective on Self-Esteem and Meaning Making while under Death
Awareness.” Secular Studies 1, 2: 1–25.
Collins, Stephanie. (2013). “Duties to Make Friends.” Ethical Theory and Moral
Practice 16, 5: 907–21.
Cottingham, John. (2003). On the Meaning of Life. Routledge.
Cottingham, John. (2014). Philosophy of Religion: Toward a More Humane Approach.
Cambridge University Press.
Cottingham, John. (2017). “Philosophy, Religion, and Spirituality.” In Spirituality and
the Good Life: Philosophical Approaches, ed. David McPherson, 11–28. Cambridge
University Press.
188 References

Crisp, Roger. (2021). “Well-Being.” In Stanford Encyclopedia of Philosophy, ed.

Edward Zalta. The Metaphysics Research Lab, Stanford University. Available at
https://plato.stanford.edu/entries/well-being/
Curzer, Howard. (2012). Aristotle and the Virtues. Oxford University Press.
Davies, Brian. (2016). “Aquinas on What God Is Not.” Royal Institute of Philosophy
Supplement 78: 55–71.
Davis, D., Hook, J., and Worthington, E. (2008). “Relational Spirituality and
Forgiveness: The Roles of Attachment to God, Religious Coping, and Viewing
the Transgression as a Desecration.” Journal of Psychology and Christianity 27, 4,
293–301.
Davis, Leesa. (2014). “Mindfulness, Non-attachment, and Other Buddhist Virtues.”
In The Handbook of Virtue Ethics, ed. Stan van Hooft. Acumen.
DeSteno, David. (2021). How God Works: The Science Behind the Benefits of Religion.
Simon & Schuster.
Dougherty, Trent and Alexander Pruss. (2014). “Evil and the Problem of Anomaly.”
In Oxford Studies in Philosophy of Religion, volume 5, ed. Jonathan Kvanvig,
49–87. Oxford University Press.
Draper, Paul. (2002). “Seeking but Not Believing: Confessions of a Practicing
Agnostic.” In Divine Hiddenness: New Essays, ed. Daniel Howard-Snyder and Paul
Moser, 197–213. Cambridge University Press.
Draper, Paul. (2022). “Atheism and Agnosticism.” In Stanford Encyclopedia of
Philosophy, ed. Edward Zalta. The Metaphysics Research Lab, Stanford University.
Available at https://plato.stanford.edu/entries/atheism-agnosticism/
Dunn, E. W., Aknin, L. B., and Norton, M. I. (2008). “Spending Money on Others
Promotes Happiness.” Science 319: 1687–88.
Dwiwardani, C., Hill, P., and Bollinger, R. (2014). “Virtues Develop from a Secure
Base: Attachment and Resilience as Predictors of Humility, Gratitude, and
Forgiveness.” Journal of Psychology and Theology 42, 1, 83–90.
Edgington, Dorothy. (2020). “Indicative Conditionals.” In Stanford Encyclopedia of
Philosophy, ed. Edward Zalta. The Metaphysics Research Lab, Stanford University.
Available at https://plato.stanford.edu/entries/conditionals/
Ekstrom, Laura. (2021). God, Suffering, and the Value of Free Will. Oxford
University Press.
Elliot, A. J. and Reis, H. T. (2003). “Attachment and Exploration in Adulthood.”
Journal of Personality and Social Psychology 85: 317–31.
Ellis, Fiona. (2014). God, Value, and Nature. Oxford University Press.
Ellison, C. G., Bradshaw, M., Kuyel, N., and Marcum, J. P. (2011). “Attachment to God,
Stressful Life Events, and Changes in Psychological Distress.” Religious Research
Association 53, 4: 493–511.
Emmons, Robert. (2005). “Emotion and Religion.” In The Handbook of the Psychology
of Religion, ed. Ray Laloutzian and Crystal Park, 235–52. Guilford.
Engel, Pascal. (2021). “Dr. Livingstone, I Presume?” Episteme 18, 3: 477–91.
Epley, N., and Schroeder, J. (2014). “Mistakenly Seeking Solitude.” Journal of
Experimental Psychology: General 143: 1980–99.
Farias, Miguel and Coleman, Thomas. (2021). “Non-religion, Atheism, and Mental
Health.” In Spirituality and Mental Health Across Cultures, ed. Alexander Moreira-
Almeida, Bruno Mosqueiro, and Dinesh Bhugra, 259–76. Oxford University Press.
 References 189

Feldman, Richard and Conee, Earl. (2018). “Between Belief and Disbelief.” In
Believing in Accordance with the Evidence, ed. Kevin McCain, 71–89. Springer.
Flood, Anthony. (2021). “Aquinas on Contrition and the Love of God.” American
Catholic Philosophical Quarterly 95, 2: 235–48.
Fraley, R., Heffernan, M., Vicary, A., and Brumbaugh, C. (2011). “The Experiences
in Close Relationships—­Relationship Structures Questionnaire: A Method for
Assessing Attachment Orientations Across Relationships.” Psychological Assessment
23, 3: 615–25.
Frances, Bryan. (2021). An Agnostic Defends God: How Science and Philosophy
Support Agnosticism. Palgrave Macmillan.
Friedman, Jane. (2019). “Inquiry and Belief.” Nous 53, 2: 296–315.
Gheaus, Anca. (2018). “Personal Relationship Goods.” In Stanford Encyclopedia of
Philosophy, ed. Edward Zalta. The Metaphysics Research Lab, Stanford University.
Available at https://plato.stanford.edu/entries/personal-relationship-goods/
Goldberg, Sanford. (2019). “Against Epistemic Partiality in Friendship: Value-
Reflecting Reasons.” Philosophical Studies 176, 8: 2221–42.
Granqvist, P., Mikulincer, M., and Shaver, P. (2010). “Religion as Attachment:
Normative Processes and Individual Differences.” Personality and Social Psychology
Review 14, 1, 49–59.
Graves, S. (2014). “God and Moral Perfection.” In Oxford Studies in Philosophy of
Religion, vol. 5, ed. Jonathan Kvanvig, 122–46. Oxford University Press.
Guan, Fang, Chen, Jun, Chen, Outong, Liu, Lihong, and Aha, Yuzha. (2019). “Awe
and Prosocial Tendency.” Current Psychology 38: 1033–41.
Gupta, Anil. (2021). “Definitions.” In Stanford Encyclopedia of Philosophy, ed.
Edward Zalta. The Metaphysics Research Lab, Stanford University. Available at
https://plato.stanford.edu/entries/definitions/
Guthrie, S. (2006). “Anthropological Theories of Religion.” In The Cambridge
Companion to Atheism, ed. Michael Martin, 283–99. Cambridge University
Press.
Hadot, Pierre. (1995). Philosophy as a Way of Life: Spiritual Exercises from Socrates to
Foucault, trans. M. Chase. Blackwell.
Hartshorne, H. and May, M. (1928). Studies in the Nature of Character, vol. 1: The
Nature of Deceit. Macmillan.
Hasker, William. (1992). “The Necessity of Gratuitous Evil.” Faith and Philosophy 9,
1: 23–44.
Helm, B. (2017). “Love.” Stanford Encyclopedia of Philosophy, ed. Edward Zalta. The
Metaphysics Research Lab, Stanford University. Available at https://plato.stanford.
edu/entries/love/
Hick, John. (1989). An Interpretation of Religion: Human Responses to the
Transcendent. Yale University Press.
Hitzman, Cortney and Wastell, Colin. (2017). “Are Atheists Implicit Theists?”
Journal of Cognition and Culture 17, 1–2: 27–50.
Honneth, Axel. (1995). The Struggle for Recognition: The Moral Grammar of Social
Conflicts, trans. Joel Anderson. Polity Press.
Hood, Ralph. (2017). “Mysticism and Hypo-egoicism.” In The Oxford Handbook
of Hypo-egoic Phenomena, ed. Kirk Brown and Mark Leary, 285–96. Oxford
University Press.
190 References

Horton, K. J., Ellison, C. G., Loukas, A., Downey, D. L., and Barrett, J. B. (2010).
“Examining Attachment to God and Health Risk-Taking Behaviours in College
Students.” Journal of Religion and Health 51, 2: 552–66.
Howard-Snyder, Daniel. (2013a). “Propositional Faith: What It Is and What It Is
Not.” American Philosophical Quarterly 50, 4: 357–72.
Howard-Snyder, Daniel. (2013b). “Schellenberg on Propositional Faith.” Religious
Studies 49, 2: 181–94.
Howard-Snyder, Daniel. (2016). “Does Faith Entail Belief?” Faith and Philosophy 33,
2: 142–62.
Howard-Snyder, Daniel. (2017). “The Skeptical Christian.” In Oxford Studies in
Philosophy of Religion, vol. 8, ed. Jonathan Kvanvig, 142–67. Oxford University Press.
Hurka, Thomas. (2001). Virtue, Vice, and Value. Oxford University Press.
Hursthouse, Rosalind. (1999). On Virtue Ethics. Oxford University Press.
Isaacs, Yoav, Hawthorne, John, and Russell, Jeffrey. (forthcoming). “Multiple
Universes and Self-Locating Evidence.” Philosophical Review.
Ivanhoe, Philip. (2018). Oneness: East Asian Conceptions of Virtue, Happiness, and
How We Are All Connected. Oxford University Press.
Jackson, Elizabeth. (2020). “The Relationship between Belief and Credence.”
Philosophy Compass 15: e12668.
Jackson, Elizabeth. (2022). “Faith, Hope, and Justification.” In Propositional and
Doxastic Justification, ed. Luis Oliveira and Paul Silva, 201–16. Routledge.
James, William. (2010). The Will to Believe and Other Essays in Popular Philosophy.
Floating Press.
Jankowski, P., and Sandage, S. (2014). “Attachment to God and Humility: Indirect
Effect and Conditional Effects Models.” Journal of Psychology and Theology 42,
1: 70–82.
Jiang, L., Yin, J., Dongmei, M., Zhu, H., and Zhou, X. (2018). “Awe Weakens the
Desire for Money.” Journal of Pacific Rim Psychology 12, 4: 1–10.
Jordan, J. (2018). “Pragmatic Arguments and Belief in God.” Stanford Encyclopedia
of Philosophy, ed. Edward Zalta. The Metaphysics Research Lab, Stanford
University. Available at https://plato.stanford.edu/entries/pragmatic-belief-god/
Juffras, A. (1972). “Is Theology a Psychological Crutch?” International Journal for
Philosophy of Religion 3, 4: 251–56.
Kamtekar, Rachana. (2016). “Becoming Good: Narrow Dispositions and the Stability
of Virtue.” In Developing the Virtues: Integrating Perspectives, ed. Julia Annas,
Darcia Narvaez, and Nancy Snow, 184–203. Oxford University Press.
Kearns, Peter and Tyler, James. (forthcoming). “Examining the Relationship between
Awe, Spirituality, and Religiosity.” Psychology of Religion and Spirituality.
Keefer, L. and Brown, F. (2018). “Attachment to God Uniquely Predicts Variation in
Well-Being Outcomes.” Archive for the Psychology of Religion 40, 2: 225–57.
Keller, Simon. (2004). “Friendship and Belief.” Philosophical Papers 33, 3: 329–51.
Kelly, Thomas. (2014). “Evidence Can Be Permissive.” In Contemporary Debates
in Epistemology, ed. Matthias Steup, John Turri, and Ernest Sosa, 298–311.
Wiley-Blackwell.
Keltner, Dacher and Haidt, Jonathan. (2003). “Approaching Awe, a Moral, Spiritual,
and Aesthetic Emotion.” Cognition and Emotion 17, 2: 297–314.
 References 191

Kim, Brian and McGrath, Matthew, eds. (2018). Pragmatic Encroachment in

Epistemology. Routledge.
King, Nathan. (2016). “Religious Skepticism and Higher-Order Evidence.” In Oxford
Studies in Philosophy of Religion, vol. 7, ed. Jonathan Kvanvig, 126–56. Oxford
University Press.
Kirkpatrick, L. A. and Shaver, P. R. (1992). “An Attachment-Theoretical Approach to
Romantic Love and Religious Belief.” Personality and Social Psychology Bulletin
18, 3: 266–75.
Knabb, J. J., and Pelletier, J. (2013). “The Relationship between Problematic Internet
Use, God Attachment, and Psychological Functioning among Adults at a
Christian University.” Mental Health, Religion & Culture 17, 3: 239–51.
Kopec, Matthew and Titelbaum, Michael. (2016). “The Uniqueness Thesis.”
Philosophy Compass 11, 4: 189–200.
Kristjánsson, Kristján. (2019). Flourishing as the Aim of Education: A Neo-Aristotelian
View. Routledge.
Kumar, Amit and Epley, Nicholas. (2022). “A Little Good Goes an Unexpectedly
Long Way: Underestimating the Positive Impact of Kindness on Recipients.”
Journal of Experimental Psychology: General 152, 1: 236–52. DOI: 10.1037/
xge0001271
Kurdi, B., Lozano, S., and Banaji, M. R. (2017). Introducing the Open Affective
Standardized Image Set (OASIS). Behavior Research Methods, 49, 2: 457–70.
Kvanvig, Jonathan. (2021). Depicting Deity: A Metatheological Approach. Oxford
University Press.
Lahroodi, Reza. (2019). “Virtue Epistemology and Collective Epistemology.” In
The Routledge Handbook of Virtue Epistemology, ed. Heather Battaly, 407–19.
Routledge.
Law, Stephen. (2010). “The Evil-God Challenge.” Religious Studies 46: 353–73.
Le Poidevin, Robin. (2010). Agnosticism: A Very Short Introduction. Oxford
University Press.
Learner, D. G. and Kruger, L. J. (1997). “Attachment, Self-Concept, and Academic
Motivation in High-School Students.” American Journal of Orthopsychiatry
67: 485–92.
Leman, Joseph, Hunter, Will, Fergus, Thomas, and Rowatt, Wade. (2018). “Secure
Attachment to God Uniquely Linked to Psychological Health in a National,
Random Sample of American Adults.” International Journal for the Psychology of
Religion 28, 3: 162–73.
Liao, Matthew. (2015). The Right to Be Loved. Oxford University Press.
Littlejohn, Clayton. (forthcoming). “A Plea for Epistemic Excuses.” In The New Evil
Demon Problem, ed. Fabian Dorsch and Julien Dutant. Oxford University Press.
Lo, Tien-Chun. (2020). “The Gap Problem Made Easy?” Analysis 80, 3: 486–92.
Lougheed, Kirk. (2018). “Is Religious Experience a Solution to the Problem of
Religious Disagreement?” Logos and Episteme 9, 2: 173–97.
Luhrmann, Tanya. (2020). When God Becomes Real: Kindling the Presence of Invisible
Others. Princeton University Press.
Luo, Li, Zou, Rong, Yang, Dong, and Yuan, Jiajin. (2022). “Awe Experience Triggered
by Fighting against COVID-19 Promotes Prosociality through Increased Feelings
192 References

of Connectedness and Empathy.” The Journal of Positive Psychology 1–16. DOI:

10.1080/17439760.2022.2131607
Lynch, Kathleen, Baker, John, Lyons, Maureen, Cantillon, Sarah, Walsh, Judy, Feeley,
Maggie, Hanlon, Niall, and O’Brien, Maeve. (2009). Affective Equality: Love, Care
and Injustice. Palgrave.
Magalotti, Tricia. (forthcoming). “Value Promotion and the Explanation of
Evidential Standards.” Erkenntnis.
Maij, David and van Elk, Michiel. (2018). “Getting Absorbed in Experimentally
Induced Extraordinary Experiences: Effects of Placebo Brain Stimulation on
Agency Detection.” Consciousness and Cognition 66: 1–16.
Maij, David, van Elk, Michiel, and Schjoedt, Uffe. (2018). “The Role of Alcohol in
Expectancy-Driven Mystical Experiences: A Pre-registered Field Study Using
Placebo Brain Stimulation.” Religion, Brain and Behavior 9, 2: 108–25.
Manela, Tony. (2015). “Gratitude.” In Stanford Encyclopedia of Philosophy, ed.
Edward Zalta. The Metaphysics Research Lab, Stanford University. Available at
https://plato.stanford.edu/entries/gratitude/
Manela, Tony. (2016). “Gratitude and Appreciation.” American Philosophical Quarterly
53, 3: 281–94.
Marcus, Eric. (2021). Belief, Inference, and the Self-Conscious Mind. Oxford
University Press.
McKaughan, Daniel. (2016). “Action-Centered Faith, Doubt, and Rationality.”
Journal of Philosophical Research 41: 71–90.
McKaughan, Daniel and Howard-Snyder, Daniel. (forthcoming). “Faith and
Faithfulness.” Faith and Philosophy.
McKim, Robert. (2008). “On Religious Ambiguity.” Religious Studies 44, 4: 373–92.
McNabb, Tyler and Baldwin, Erik. (2022). Classical Theism and Buddhism:
Connecting Metaphysical and Ethical Systems. Bloomsbury.
McPherson, David. (2017). “Homo Religiosus: Does Spirituality Have a Place in Neo-
Aristotelian Virtue Ethics?” In Spirituality and the Good Life: Philosophical
Approaches, ed. David McPherson, 64–83. Cambridge University Press.
Mikulincer, M. (1997). “Adult Attachment Style and Information Processing:
Individual Differences in Curiosity and Cognitive Closure.” Journal of Personality
and Social Psychology 72: 1217–30.
Miller, Callum. (2021). “Why God Is Probably Good: A Response to the Evil-God
Challenge.” Religious Studies 57: 448–65.
Miller, Christian. (2014). Moral Character: An Empirical Theory. Oxford University
Press.
Miller, Christian. (2021). Honesty: The Philosophy and Psychology of a Neglected
Virtue. Oxford University Press.
Milona, Michael. (2019). “Finding Hope.” Canadian Journal of Philosophy 49, 5: 710–29.
Moon, Andrew. (2019). “A New Puzzle about Belief and Credence.” Canadian
Journal of Philosophy 49, 2: 272–91.
Moon, Andrew and Jackson, Elizabeth. (2020). “Credence: A Belief-First Approach.”
Canadian Journal of Philosophy 50, 5: 652–69.
Morris, Thomas. (1987). “Perfect Being Theology.” Nous 21, 1: 19–30.
 References 193

Mugg, Joshua. (forthcoming). “Faith Entails Belief: Three Avenues of Defense

against the Argument from Doubt.” Pacific Philosophical Quarterly.
Mullins, Ryan. (2021). “Classical Theism.” In T&T Clark Companion to Analytic
Theology, ed. James Arcadi and James Turner, 85–100. T&T Clark.
Nagasawa, Yujin. (2017). Maximal God: A New Defence of Perfect Being Theism.
Oxford University Press.
Njus, D. M. and Scharmer, A. (2020). “Evidence that God Attachment Makes a
Unique Contribution to Psychological Well-Being.” The International Journal for
the Psychology of Religion 30, 3: 178–201.
Nussbaum, Martha. (1990). Love’s Knowledge: Essays on Philosophy and Literature.
Oxford University Press.
Oppy, Graham. (2013). “Ultimate Naturalistic Causal Explanations.” In Why Is There
Something Rather Than Nothing? ed. Ty Goldschmidt, 46–63. Routledge.
Pace, Michael. (2011). “The Epistemic Value of Moral Considerations: Justification,
Moral Encroachment, and James’ ‘Will to Believe’. ” Nous 45, 2: 239–68.
Palmqvist, Carl-Johan. (2022). “Desiderata for Rational, Non-doxastic Faith.” Sophia
61, 3: 499–519.
Papineau, David. (2020). “Naturalism.” In Stanford Encyclopedia of Philosophy, ed.
Edward Zalta. The Metaphysics Research Lab, Stanford University. Available at
https://plato.stanford.edu/entries/naturalism/
Pearce, Kenny and Oppy, Graham. (2022). Is There a God? A Debate. Routledge.
Perlin, Joshua, and Li, Leon. (2020). “Why Does Awe Have Prosocial Effects? New
Perspectives on Awe and the Small Self.” Perspectives on Psychological Science 15,
2: 291–308.
Peterson, Christopher and Seligman, Martin. (2004). Character Strengths and Virtues:
A Handbook and Classification. Oxford University Press.
Piff, P. K., Dietz, P., Feinberg, M., Stancato, D. M., and Keltner, D. (2015). “Awe, the
Small Self, and Prosocial Behavior.” Journal of Personality and Social Psychology
108, 6: 883–99.
Pirutinsky, S. D., Rosmarin, H., and Kirkpatrick, L. A. (2019). “Is Attachment to
God a Unique Predictor of Mental Health? Test in a Jewish Sample.” The
International Journal for the Psychology of Religion 29: 161–71.
Porter, S. and Baehr, J. (2020). “Becoming Honest: Why We Lie and What Can Be
Done about It.” In Integrity, Honesty, and Truth-Seeking, ed. Christian Miller and
Ryan West, 182–206. Oxford University Press.
Porter, S. and Rickabaugh, B. (2018). “The Sanctifying Work of the Holy Spirit:
Revisiting Alston’s Interpersonal Model.” Journal of Analytic Theology 6: 112–30.
Preston, Jesse and Shin, Faith. (2017). “Spiritual Experiences Evoke Awe through
the Small Self in Both Religious and Non-religious Individuals.” Journal of
Experimental Social Psychology 70: 212–21.
Preston-Roedder, Ryan. (2013). “Faith in Humanity.” Philosophy and Phenomenological
Research 87, 3: 664–87.
Radzik, Linda and Murphy, Colleen. (2015). “Reconciliation.” In Stanford
Encyclopedia of Philosophy, ed. Edward Zalta. The Metaphysics Research Lab,
Stanford University. Available at https://plato.stanford.edu/entries/reconciliation/
194 References

Rasmussen, Joshua. (2021). “The Argument from Contingency.” In Contemporary

Arguments in Natural Theology: God and Rational Belief, ed. Colin Ruloff and
Peter Horban, 17–34. Bloomsbury.
Regehr, C., Goldberg, G., and Hughes, J. (2002). “Exposure to Human Tragedy,
Empathy, and Trauma in Ambulance Paramedics.” American Journal of
Orthopsychiatry 72: 505–13.
Reiner, S. R., Anderson, T. L., Hall, M. E. L., and Hall, T. W. (2010). “Adult
Attachment, God Attachment and Gender in Relation to Perceived Stress.”
Journal of Psychology and Theology 38, 3: 175–85.
Rettler, Lindsay. (2018). “Faith, Belief, and Control.” American Philosophical
Quarterly 55, 1: 95–109.
Riggs, Wayne. (2008). “Epistemic Risk and Relativism.” Acta Analytica 23, 1: 1–8.
Rinard, Susanna. (2017). “No Exception for Belief.” Philosophy and Phenomenological
Research 94, 1: 121–43.
Rinard, Susanna. (2019a). “Believing for Practical Reasons.” Nous 53, 4: 763–84.
Rinard, Susanna. (2019b). “Equal Treatment for Belief.” Philosophical Studies 176, 7:
1923–50.
Roberts, Robert C. (2007). Spiritual Emotions: A Psychology of Christian Virtues.
Eerdmans.
Roberts, Robert C. and Wood, W. Jay. (2007). Intellectual Virtues: An Essay in
Regulative Epistemology. Oxford University Press.
Rogers, Katherin. (2000). Perfect Being Theology. Edinburgh University Press.
Rosenberg, M. (1965). Society and the Adolescent Self-Image. Princeton University
Press.
Rosmarin, David, Pirutinsky, Steven, Cohen, Adam, Galler, Yardana, and Krumrei,
Elizabeth. (2011). “Grateful to God or Just Plain Grateful? A Comparison of
Religious and General Gratitude.” Journal of Positive Psychology 6, 5: 389–96.
Rota, Michael. (2017). “Pascal’s Wager.” Philosophy Compass 12, 4: e12404.
Rowatt, W. C. and Kirkpatrick, L. A. (2002). “Two Dimensions of Attachment to
God and Their Relation to Affect, Religiosity, and Personality Constructs.” Journal
for the Scientific Study of Religion 41: 637–51.
Rowatt, Wade, Carpenter, Tom, and Haggard, Megan. (2014). “Religion, Prejudice,
and Intergroup Relations.” In Religion, Personality, and Social Behavior, ed.
Vassilis Saroglou, 170–92. Psychology Press.
Rudd, R. A. and D’Andrea, L. M. (2015). “Compassionate Detachment: Managing
Professional Stress While Providing Quality Care to Bereaved Parents.” Journal of
Workplace Behavioral Health 30, 3: 287–305.
Russell, Daniel. (2009). Practical Intelligence and the Virtues. Oxford University Press.
Russell, Daniel. (2012). Happiness for Humans. Oxford University Press.
Russell, Daniel. (2015). “Aristotle on Cultivating Virtue.” In Cultivating Virtue:
Perspectives from Philosophy, Theology, and Psychology, ed. Nancy Snow, 17–48.
Oxford University Press.
Sandage, S., Paine, D., and Hill, P. (2015). “Spiritual Barriers to Humility: A
Multidimensional Study.” Mental Health, Religion & Culture 18, 3: 207–17.
Schellenberg, John. (2005). Prolegomena to a Philosophy of Religion. Cornell
University Press.
 References 195

Schellenberg, John. (2007). The Wisdom to Doubt: A Justification of Religious

Skepticism. Cornell University Press.
Schellenberg, John. (2015). The Hiddenness Argument: Philosophy’s New Challenge to
Belief in God. Oxford University Press.
Schellenberg, John. (2017). “Divine Hiddenness: Part 1 (Recent Work on the
Hiddenness Argument).” Philosophy Compass 12, 4: e12355.
Schoenfield, Miriam. (2014). “Why Permissivism Is True and What It Tells Us about
Irrelevant Influences on Belief.” Nous 47, 1: 193–218.
Seglow, Jonathan. (2013). Defending Associative Duties. New York: Routledge.
Shiota, Michelle N., Campos, Belinda, and Keltner, Dacher. (2003). “The Faces of
Positive Emotion: Prototype Displays of Awe, Amusement, and Pride.” Annals of
the New York Academy of Sciences 1000: 296–99.
Shiota, Michelle N., Keltner, Dacher, and Mossman, Amanda. (2007). “The Nature
of Awe: Elicitors, Appraisals, and Effects on Self-Concept.” Cognition and Emotion
21, 5: 944–63.
Sim, May. (2017). “Identifying with the Confucian Heaven: Immanent and
Transcendent Dao.” In Spirituality and the Good Life, ed. David McPherson.
Cambridge University Press.
Snow, Nancy. (2009). Virtue as Social Intelligence: An Empirically Grounded Theory.
Routledge.
Snow, Nancy. (2013). “Notes Toward an Empirical Psychology of Virtue: Exploring
the Personality Scaffolding of Virtue.” In Aristotelian Ethics in Contemporary
Perspective, ed. Julia Peters, 130–44. New York: Routledge.
Snow, Nancy. (2016). “How Habits Make Us Virtuous.” In Developing the Virtues:
Integrating Perspectives, ed. Julia Annas, Darcia Narvaez, and Nancy Snow,
135–57. Oxford University Press.
Speaks, Jeff. (2018). The Greatest Possible Being. Oxford University Press.
Staffel, Julia. (2019). “Credences and Suspended Judgments as Transitional Attitudes.”
Philosophical Issues 29, 1: 281–94.
Stalnaker, Aaron. (2016). “In Defense of Ritual Propriety.” European Journal for
Philosophy of Religion 8, 1: 117–141.
Stanford, Kyle. (2006). Exceeding Our Grasp: Science, History, and the Problem of
Unconceived Alternatives. Oxford University Press.
Steffen, Edith and Klass, Dennis eds. (2018). Continuing Bonds in Bereavement: New
Directions for Research and Practice. Routledge.
Stellar, J. E., Gordon, A., Anderson, C. L., Piff, P. K., McNeil, G. D., and Keltner,
D. (2018). “Awe and Humility.” Journal of Personality and Social Psychology 114,
2: 258–69.
Strenger, A., Schnitker, S., and Felke, T. (2016). “Attachment to God Moderates the
Relation between Sociocultural Pressure and Eating Disorder Symptoms
as Mediated by Emotional Eating.” Mental Health, Religion & Culture 19, 1: 23–36.
Stroud, Sarah. (2006). “Epistemic Partiality in Friendship.” Ethics 116, 3: 498–524.
Stump, Eleonore. (2006). “Love, by All Accounts.” Proceedings of the Aristotelian
Society 80, 2: 25–43.
Stump, Eleonore. (2016). The God of the Bible and the God of the Philosophers.
Marquette University Press.
196 References

Stump, Eleonore. (2018). Atonement. Oxford University Press.

Swanton, Christine. (2015). “Cultivating Virtue: Two Problems for Virtue Ethics.” In
Cultivating Virtue: Perspectives from Philosophy, Theology, and Psychology, ed.
Nancy Snow, 111–34. Oxford University Press.
Swanton, Christine. (2021). Target-Centred Virtue Ethics. Oxford University Press.
Swenson, Philip. (2022). “Moral Luck, Free Will Theodicies, and Theological
Determinism.” In Theological Determinism: New Perspectives, ed. Peter Furlong
and Leigh Vicens, 184–94. Cambridge University Press.
Swinburne, Richard. (2001). Epistemic Justification. Oxford University Press.
Tangney, J. P., Baumeister, R. F., and Boone, A. L. (2004). “High Self-Control Predicts
Good Adjustment, Less Pathology, Better Grades, and Interpersonal Success.”
Journal of Personality 72: 271–322.
Tellegen, Auke and Atkinson, Gilbert. (1974). “Openness to Absorbing and Self-
Altering Experiences (‘Absorption’), a Trait Related to Hypnotic Susceptibility.”
Journal of Abnormal Psychology 83, 3: 268–77.
Thauvoye, E., Granqvist, P., Golovchanova, N., and Dezutter, J. (2018). “Attachment
to God, Depression and Loss in Late Life: A Longitudinal Study.” Mental Health,
Religion & Culture 21: 825–37.
Tiozzo, Marco. (forthcoming). “Explaining Higher-Order Defeat.” Acta Analytica.
Titova, Luidmila and Sheldon, Kennon. (2022). “Happiness Comes from Trying to
Make Others Feel Good, Rather Than Oneself.” Journal of Positive Psychology 17,
3: 341–55.
Toussaint, Loren and Williams, David. (2008). “National Survey Results for
Protestant, Catholic, and Nonreligious Experiences of Seeking Forgiveness and of
Forgiveness of Self, of Others, and by God.” Journal of Psychology and Christianity
27, 2: 120–30.
Tucker, Chris. (2016). “Satisficing and Motivated Submaximization (in the Philosophy
of Religion).” Philosophy and Phenomenological Research 93, 1: 127–43.
Uzarevic, Filip and Coleman, Thomas. (2021). “The Psychology of Nonbelievers.”
Current Opinion in Psychology 40: 131–38.
Vaillant, George. (2012). Triumphs of Experience: The Men of the Harvard Grant
Study. Belknap Press.
Valdesolo, Piercarlo, and Graham, Jesse. (2014). “Awe, Uncertainty, and Agency
Detection.” Psychological Science 25, 1: 170–78.
Van Cappellen, P. and Saroglou, V. (2012). “Awe Activates Religious and Spiritual
Feelings and Behavioral Intentions.” Psychology of Religion and Spirituality 4,
3: 223–36.
Van Cappellen, Patty, Edwards, Megan, and Barbara Fredrickson. (2021). “Upward
Spirals of Positive Emotions and Religious Behaviors.” Current Opinion in
Psychology 40: 92–98.
Van Cappellen, Patty, Saroglou, Vassilis, Iweins, Caroline, Piovesana, Maria, and
Fredrickson, Barbara. (2013). “Self-Transcendent Positive Emotions Increase
Spirituality Through Basic World Assumptions.” Cognition and Emotion 27, 8:
1378–94.
 References 197

van Elk, M., Karinen, A., Specker, E., Stamkou, E., and Baas, M. (2016). “ ‘Standing
in Awe’: The Effects of Awe on Body Perception and the Relation with Absorption.”
Collabra: Psychology 2, 1: 4.
Van Eyghen, Hans. (2020). Arguing from Cognitive Science of Religion: Is Religious
Belief Debunked? Bloomsbury.
van Fraassen, Bas. (2002). The Empirical Stance. Yale University Press.
Van Inwagen, Peter. (1988). “The Magnitude, Duration, and Distribution of Evil:
A Theodicy.” Philosophical Topics 16, 2: 161–87.
Vitz, Rico. (n.d.) “Doxastic Voluntarism.” Internet Encyclopedia of Philosophy.
Available at https://iep.utm.edu/doxastic-voluntarism/
Vessiére, Samuel, Constant, Alex, Ramstead, Maxwell, Friston, Karl, and Kirmayer,
Laurence. (2019). “Thinking through Other Minds: A Variational Approach to
Cognition and Culture.” Behavioral and Brain Sciences 43, e90: 1–75.
Waggoner, Maria, Doris, John, and Vargas, Manuel. (2022). “Situationism, Moral
Improvement, and Moral Responsibility.” In The Oxford Handbook of Moral
Psychology, ed. Manuel Vargas and John Doris, 629–60. Oxford University Press.
Walker, A. D. M. (1980). “Gratefulness and Gratitude.” Proceedings of the Aristotelian
Society 81: 39–55.
Waller, Jason. (2021). Cosmological Fine-Tuning Arguments: What (if Anything)
Should We Infer from the Fine-Tuning of Our Universe for Life? Routledge.
Webster, L., Hackett, R. K., and Joubert, D. (2009). “The Association of Unresolved
Attachment Status and Cognitive Processes in Maltreated Teens. Child Abuse
Review 18: 6–13.
Weidner, Veronika. (2019). Examining Schellenberg’s Hiddenness Argument. Palgrave
Macmillan.
Whiting, Daniel. (forthcoming). “Recent Work on Higher-Order Evidence.” Analysis.
Wolf, Susan. (1982). “Moral Saints.” Journal of Philosophy 79, 8: 419–39.
Wykstra, Stephen. (2002). “Not Done in a Corner: How to be a Sensible Evidentialist
about Jesus.” Philosophical Books 43: 81–135.
Wynn, Mark. (1997). “Trust Relationships and the Moral Case for Religious Belief.”
International Philosophical Quarterly 37, 2: 179–88.
Wynn, Mark. (2020). Spiritual Traditions and the Virtues: Living between Heaven
and Earth. Oxford University Press.
Yaden, D. B., Iwry, J., Slack, K. J., Eichstaedt, J. C., Zhao, Y., Vaillant, G. E., and
Newberg, A. B. (2016). “The Overview Effect: Awe and Self-Transcendent
Experience in Space Flight.” Psychology of Consciousness 3, 1: 1–11.
Yaden, David, Kaufman, Scott, Hyde, Elizabeth, Chirico, Alice, Gaggioli, Andrea,
Zhang, Jia, and Keltner, Dacher. (2019). “The Development of the Awe
Experiences Scale (AWE-S): A Multifactorial Measure for a Complex Emotion.”
The Journal of Positive Psychology 14, 4: 474–88.
Yang, Y., Yang, Z., Bao, T., Liu, Y., and Passmore, H. A. (2016). “Elicited Awe
Decreases Aggression.” Journal of Pacific Rim Psychology 10: e11.
Zagzebski, Linda. (2017). Exemplarist Moral Theory. Oxford: Oxford University Press.
 Index

For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on
occasion, appear on only one of those pages.

Abrahamic faiths, God of the 6, 12–14, compassion 111, 144–5

18–23, 41 complexity (of evidence for God) 56–7
absorption (personality trait) 165–7, 173–8 conditional attitudes 74–5
acceptance (the cognitive attitude) 72–4, 85 Conee, Earl 25n.1, 26n.3, 55n.26, 76–8
acting as if 72, 150–2 confidence 70–1, 76–8
Adams, Robert 99, 135–6 congregation see community
afterlife 2, 132, 142 connectedness 97–8, 102, 164–79, 182–4
Alston, William 64–5, 72–3 contemplation 160–1, 172–9
Annas, Julia 96, 107, 135–6 continence (of character) 109–10
anti-realism contingency, argument from 34–5, 40–2, 44–9
scientific 37–40, 73 contrition 120–37, 140–2
theological 73–4 Cottingham, John 63n.2, 102, 160–1, 163–5
apology 61–2, 74–5, 120–37 creation, divine 15–16, 19
appreciation 118–20, 129–30, 140–1, 160 credences 70–1, 77, 84–5
Aristotle 106–9, 134–7, 160–1, 170, 178–9 credit 118, 122–3, 127–8
assumption (the cognitive attitude) 3, 64–76, crutch, religion as 153–4
83–92, 121–2, 133, 139–40, 163 Curzer, Howard 107–8, 113–14
attachment 143–5
to God 145–57 defeat, evidential 51, 53, 55–60
autonomy 154 deism 19
awe 7, 97–8, 112–13, 160–79 depression 142–3, 145, 147–53, 156
disagreement (about evidence for God) 57–8
Baehr, Jason 98–9, 142–3 Draper, Paul 19, 28–9, 32
Battaly, Heather 93–4, 114
Beck, Richard 145–53 ego dissolution 169
belief (the cognitive attitude) 3–4, 26, 63–92, epiphany 112–13
121–2, 133 epistemic dilemma 59
belongingness 129–30 epistemic excuse 82–4, 88–92
Bowlby, John 143–4 epistemic partiality 83–4, 127
Brownlee, Kimberly 128–9 epistemic possibility 68–9
Buddhism 183–4 equal treatment view of belief 80–2
eudaemonia 106–7
care ethics 128 evidence
character virtues 93–105 nature of evidence 24–5
characteristic enjoyments (of humans) availability of evidence 25
101–2, 114 ambiguity of evidence 25–6
classical theism 22–3 evidential support 26
cognitive science of religion 57, 66–7 private evidence 24–5, 28–31
Coleman, Thomas 5, 67–8 higher-order evidence 54–60
community (religious or spiritual) 30, 181 vague evidence 25–6, 55
200 Index

evidentialism 76–7 limited agnosticism 25, 29–32, 53–4

evil God challenge 44–5 Lo, Tien-Chun 46, 48
evil, evidential argument from 34, 49–53 local character see global character
exemplarism 167–8 Luhrmann, Tanya 166–7, 173

Feldman, Richard 25n.1, 26n.3, 55n.26, 76–8 Manela, Tony 118–20

fittingness (of attitudes) 96–8, 102, Marcus, Eric 67n.8, 85–6
131–5, 141–2 maximal God 18
flourishing 105–14 McDonald, Angie 145–6, 148, 150–3
forgiveness 120–1, 133, 137, 145 McKaughan, Daniel 62–3, 65–6, 68,
Frances, Bryan 34 84–5, 91n.22
free choice 43 McKim, Robert 25n.2, 54
friendship 83–4, 127–9 McPherson, David 101–2, 104n.3, 160–1,
163, 168
gap problem 41–9 meaning (in life or of life) 130, 160
generosity 96, 142–3, 164–5 metatheology 12–13, 16–17
global character 135–7 Miller, Callum 44
Goldberg, Sanford 127 Miller, Christian 94–5
gratitude 120–1, 129–32, 137, 145, 151–3, mindfulness 183
156, 160 Mongrain, Myriam 171
great achievers 110–11 monotheism 15
grounding (metaphysical) 15–16, 20–1 moral encroachment 78–80
moral saints 125
Hadot, Pierre 161–3 multiverse 37–9
hiddenness, divine 51–3 mystical experience 169
higher-order agnosticism 58–9
honesty 94–5, 136 Nagasawa, Yujin 13n.2, 18
hope 68–9, 121, 151–2, 160 natural laws 35, 42
Howard-Snyder, Daniel 62–7, 69–70, need to belong see belongingness
72–3, 84–7 Neo-Aristotelian virtue ethics 100–1, 108
human nature 100–1, 106–7 Neo-Confucianism 182
Hurka, Thomas 95 Njus, David 145, 148, 150, 155
Hursthouse, Rosalind 101–5, 108, 113, Nussbaum, Martha 100
125n.4
omniGod theism 17, 19, 33, 50–1
imagination 71–2, 165–7 omnipotence 17, 43, 48, 50–1
indirect virtue development 138–9, 142–8 Oppy, Graham 34–5, 41–3, 46–8
inner senses 166–8
instability (of evidence for God) 57 Pace, Michael 79–80, 125
intervention, divine 19 panentheism 20–1, 32–3
intrinsic probability 32–6 pantheism 19–21, 73–4
intuition 124–5 Pascal’s wager 1–4
Ivanhoe, Philip 182 Pearce, Kenneth 20–1, 40n.13, 46
perfect being theism 17–18, 32–3, 45–8,
Jackson, Elizabeth 66n.7, 70–1, 84 50–3, 73–4
James, William 2n.2, 77–8 permissivism, epistemic 77–8, 80
Juffras, Angelo 154 personal relationship goods 126–30
personal worth conception of virtue 98
Kristjánsson, Kristján 108–10, 112–14 Peterson, Christopher 160, 162–4
Kvanvig, Jonathan 12–14, 41–5 piety 102–5, 160–2, 168
 Index 201

pragmatic encroachment; see moral small self 165

encroachment Snow, Nancy 135–6, 142–3
prejudice 168 Speaks, Jeff 17n.4, 18
Preston-Roedder, Ryan 83–4, 125–7 Stoics 107–8
stress (psychological) 114, 130, 157
Rasmussen, Joshua 40n.13, 45n.16 Stump, Eleonore 16, 22n.8, 139n.2, 142
rationality (human) 83–5, 102–5 subordination (among character
real definition 46–7 traits) 119–20
respect (virtue of) 125–6 Swanton, Christine 99, 102
Rinard, Susanna 64n.5, 80–2 Swinburne, Richard 33n.5
ritual 1, 4
Roberts, Robert 121, 154 testimony (of God) 24–5, 33–4
role-differentiated virtue 111 transcendence (virtue of) 160
Russell, Daniel 95, 106–7, 112–13, 126, 134–5 transcendence, divine 23
Trinity 15
sacredness 160–1
satis concept 95 ultimism 21–2
satisfaction with life 128–9, 137, 142–3, 145, union, interpersonal 16, 138–9
151–3, 156, 181
Scharmer, Alexandra 145, 148, 150, 155 vice (of character) 112, 154
Schellenberg, John 21, 51n.23, 55–6,
58, 63n.3 Waller, Jason, 37n.9, 38
self-deception 79–80 well-being, theories of 114
self-esteem 128–9, 142–3, 145, 148–53, 156–7 Wolf, Susan 125
self-improvement 102, 165 worldview 97–8, 161–4
Seligman, Martin 160, 162–4 worship 13–14
simplicity (ontological) 33–6, 38–9, 44 Wynn, Mark 63n.2, 91n.22, 127n.5,
sincerity 2–4, 62, 118–24 161–3, 181
situationism 135–6
skill (of virtue) 96–8, 162–6, 183–4 Yaden, David 164–5, 171–2