ChapterPDF Available

Thesis - contents & intro

The Circumstances of
by Fleur Jongepier
Copyright 2017 © by Fleur Jongepier
All rights reserved
ISBN: 978-94-92380-46-3
Cover design: Fleur Jongepier & Marlaine Verhelst
Cover image: Our Little Secret II by Petyka
Inside design: ProefschriftOntwerp, Nijmegen
Printed by Ipskamp B.V.
The Circumstances of
ter verkrijging van de graad van doctor
aan de Radboud Universiteit Nijmegen
op gezag van de rector magnificus prof. dr. J.H.J.M. van Krieken,
volgens besluit van het college van decanen
in het openbaar te verdedigen op dinsdag 23 mei 2017
om 13.30 uur precies
Fleur Jongepier
geboren op 20 oktober 1986 te Eindhoven
Prof. dr. M.V.P. Slors
Prof. dr. J.A.M. Bransen
Prof. dr. Q. Cassam
Prof. dr. L.B.W. Geurts
Prof. dr. E. Schwitzgebel
Prof. dr. M. Düwell
Work on this thesis was funded by the Netherlands Organization for Scientific
Research (NWO). Research project no. 322-20-003.
Acknowledgements ix
1. Introduction 13
1. The Gist of the Thesis: Atomism versus Holism 13
2. Qualifications, Assumptions and Limitations 18
3. Plan for the Thesis 21
PART 1: Expressivism
2. Knowing Your Mind by Speaking Your Mind 27
1. Introduction 27
2. Expressivism: The Basics 28
3. Bar-On’s Neo-Expressivism 32
4. Expressivism as a Theory of Self-Knowledge 35
5. Avowals as Expressing Beliefs 38
6. Avowals as Expressing Justified Beliefs 40
7. Avowals as Expressing True Beliefs 42
8. Some Prima Facie Worries 46
9. Conclusion 50
3. The Limits of Expressivism: Failures of Self-Expression 53
1. Introduction 53
2. Ceteris Paribus I: Assuming I Am Avowing 55
3. Ceteris Paribus II: Assuming I Am Not Severely Intoxicated or Asleep 64
4. Self-Expression as Self-Interpretation 72
5. Diagnosis: Self-Expression and Communication 76
6. Avowals as Assertions: Theoretical Advantages 82
7. Conclusion 87
PART 2: Rationalism
4. Knowing Your Mind by Making Up Your Mind 91
1. Introduction 91
2. The Agential Account of Self-Knowledge 92
3. The Rationalist Account of Self-Knowledge 95
3.1 The Transparency Procedure 95
3.2 Alienation 99
4. Is the Transparency Procedure Necessary and Sufficient for
Self-Knowledge? 101
4.1 Common Objections to Rationalism 101
4.2 Possible Responses 104
5. A Transcendental Take on Transparency? 106
6. Is the Transparency Procedure an Inferential Procedure? 111
7. Conclusion 115
5. The Limits of Rationalism: Failures of Transparency 117
1. Introduction 117
2. Radical Rationalism and the Anscombean Constraint 120
2.1 The Anscombean Constraint 120
2.2 Radical Rationalism 122
3. Prima Facie Challenges and Problems for Radical Rationalism 124
4. Moderate Rationalism 127
5. The ‘Filtering Strategy’ as a Response to the ‘Garbage in,
Garbage out’ Objection 130
5.1 Why it Filters Out Too Much 131
5.2 Why ‘Considered Reasons’ are Neither Necessary Nor Sufficient
for Self-Knowledge 133
6. Holist Rationalism: The Self as Part of the World? 138
7. Conclusion 141
6. The Limits of Activism 145
1. Introduction 145
2. Boyle’s Epistemic View: Reflectivism 147
3. Boyle’s Metaphysical View: Activism 149
4. Radical Versus Moderate Dispositionalism 154
5. The Activist Argument Against Dispositionalism 160
5.1 Prima Facie Doubts Against Dispositionalism 160
5.2 The Principal Objection: Meeting ‘Boyle’s Constraint’ 162
6. Response to Boyle’s Constraint 163
7. The Apparent Gap Between Judgement and Belief 168
7.1 The Implicit Racist, the Unethical Professor and the Family Man 168
7.2 The ‘No Proper Judgement’ Response 171
8. The ‘What Gap?’ Response 173
9. Meta-Theoretical Questions about Metaphysics of Mind 177
10. Conclusion 182
Appendix: A Dispositionalist Account of Self-Knowledge and an Activist
Account of First-Person Authority 185
7. The Limits of Rationalism: Failures of Autonomy 189
1. Introduction 189
2. Rationalism as a Normative Project 192
3. Deliberation in Oppressive Circumstances: Three Cases 198
4. Having a Mind of One’s Own 203
5. Procedural Accounts of Autonomy 206
6. Relational Accounts of Autonomy 210
7. Implications for Self-Knowledge 215
7.1 The Simple View 215
7.2 A Nuanced View: Types of Self-Knowledge 219
8. The Metaphysical and Epistemic Sources of Atomism 221
9. Conclusion 229
8. Taking Stock and Looking Forward 233
Bibliography 239
Samenvatting 253
Curriculum Vitae 263
Introduction 13
1. The gist of the thesis: atomism versus holism
Do you have self-knowledge of your intention to be at work on time tomorrow if you’ve
had a few drinks too many? Do you know that you want a divorce if you express your desire
during a fit of anger? Do you know your own desires, hopes and beliefs if you’re depressed,
insecure or got out of bed on the wrong side? Do you have self-knowledge of your desire
to buy a healthy quinoa salad rather than fish and chips for lunch if it’s evident that you’ve
been ‘nudged’ into doing so? Do you have self-knowledge of your belief that having a baby
boy is better than having a baby girl if you’ve been manipulated by state propaganda?
These are fundamental and difficult questions about self-knowledge and about
what it means to know your own beliefs, desires, hopes, intentions and other attitudes.
The answers to these questions are by no means obvious. The aim of this thesis is not
to provide a concrete answer to such questions, but rather to ask what is required in
order to answer them. More specifically, the aim is to explore whether contemporary
philosophical theories of self-knowledge have the materials to handle such questions
in a satisfactory manner, i.e. in a way that respects their subtleties and intricacies.
The current philosophical debate on self-knowledge is mostly concerned
with the question of whether self-knowledge is a matter of looking into our minds
(introspectionism), interpreting our minds (interpretivism), speaking our minds
(expressivism) or making up our minds (rationalism). Hence, when working on
self-knowledge, one of the first questions one is confronted with is whether one
is an ‘introspectionist’, ‘interpretationist’, ‘expressivist’, ‘rationalist’ or perhaps
a ‘pluralist’ of some sort.1 In other words, much of the current self-knowledge
1 According to introspectionist or ‘hybrid’ introspectionist accounts, self-knowledge is a matter of inner looking
or ‘detecting’ one’s own mental states (e.g. Armstrong 1968; Lycan 1996; Goldman 1993; 2006; Nichols and Stich
2003; Rosenthal 2005; Peels forthcoming). Self-ascriptions or second-order beliefs about one’s first-order st ate
Chapter 1
debate focuses on the methods or procedures of self-knowledge – their differences,
similarities and (in)compatibilities.2
The focus of this thesis will instead be on what appears to be a widely shared
though implicit assumption, which I’ll refer to as the assumption of atomism
regarding self-knowledge. This is an assumption about what is required for
someone to acquire knowledge of her attitudes. The atomist assumption comes
down to this: following the theorist’s preferred procedure or method (such as
speaking or making up one’s mind) is sufficient for a subject to acquire knowledge
of her attitudes. The alternative ‘holist’ approach can be understood negatively as
the reverse of atomism: the mere following of any of these standard methods or
procedures by itself does not guarantee that they will yield self-knowledge.
The plan is to articulate and problematize the atomist approach to the question
of self-knowledge. More specifically, my aim is to address the preconditions of self-
knowledge, which I will address by asking under what circumstances following some
particular method is actually knowledge-conducive. In so doing I will concentrate,
specifically, on the expressivist and rationalist accounts.
The atomist view as I have just described it is a caricature, which, I take it, hardly
anyone will be happy to defend. In recognition of this fact, I distinguish between
two types of atomism. According to what I’ll call a radical atomist approach to self-
knowledge, the question of whether the ascription of a belief or other attitude to
oneself counts as self-knowledge is independent of the circumstances under which
one followed method X: there are no ‘bad’ circumstances of self-knowledge. So even
someone who is depressed, in a fit of anger, manipulated by state propaganda, under
the influence of mind-altering substances, hypnotized or tortured still has knowledge
of her attitudes, as long as she arrived at her self-ascription in the ‘right’ way.
are understood as ‘descriptions’ or ‘reports’ of underlying st ates. Interpretationists and/or inferentialists think
that one’s second-order belief is the result of an (unconscious or sub-personal) process of self-interpretation, a
process of ‘first-person mindreading’ or a process of inferences made on the basis of evidence (Bem 1972; Gopnik
1993; Carruthers 2009; 2011; Schwitzgebel 2001; 2010; 2012; 2013; 2016; Lawlor 2003; 2009; Cassam 2010; 2011;
2014; 2015). Expressivism, rationalism and interpretivism will be explained in detail in the chapters to come. I will
not discuss pluralist accounts in this thesis, but see, for example, Andrews (2015), Samoilova (2015) and Coliva
2 A brief note on terminology: the notion of a method or procedure can be understood either in a narrow way, as
referring to a way of doing something step by step or on the basis of a pre-established manual or recipe or, more
broadly, as ‘a way of doing something’ or an ‘act or a manner of proceeding in any action or process’. I shall
understand the notions of methods or procedures in the latter, broader, sense, i.e. as a way of doing something,
more specifically as a way of making or arriving at a self-ascription or coming to judge or coming to believe that
one φs that P.
Introduction 15
This view is a caricature, I think, because most philosophers working on self-
knowledge will allow, implicitly or explicitly, that there are at least certain
circumstances in which a self-ascription does not count as self-knowledge even if
the appropriate method is followed. If, say, a ‘mentalist’ like Derren Brown were to
hypnotize you into thinking you’re a chicken, and you go on to think you need to
brood your eggs, then, in some sense at least, you lacked knowledge of your beliefs,
desires or intentions, such as not wanting to be made a fool of, the belief that you are
not a chicken, the desire to be unhypnotized, and so on. The sort of self-knowledge
you have while hypnotized is only knowledge of your occurrent thoughts.3
Hence, I shall distinguish radical atomism from what I call moderate atomism,
which can be understood as a ‘hedged’ version of atomism. It is moderate atomism
that I shall be centrally concerned with. A moderate atomist claims that even though
there are certain extreme circumstances in which a self-ascription does not count
as self-knowledge, in ‘normal’, ‘standard’ or the ‘right’ conditions, the proposed
method is indeed knowledge-conducive, i.e. will allow the subject to know what
her attitudes are. The crux of the difference, then, between moderate atomism and
holism is that even the moderate atomist agrees with the holist that radical atomism
is mistaken; s/he thinks that the success of some or all of the above-mentioned
methods can be said to hold ceteris paribus, that is, all else being equal, right or
normal, or in the absence of countervailing forces or disturbing factors. Whereas
a moderate atomist thinks that we can talk meaningfully about ‘exceptions to the
rule’, I will try to make a case for the idea that in the end, the exceptions are the rule.
The moderate atomist thus proceeds to concern herself with ‘matters of method’
and questions regarding procedures by appealing to certain ceteris paribus clauses.
This, in turn, allows her to deal with, i.e. set aside, the sort of cases mentioned at the
beginning of this introduction. We can ignore angry spouses, tipsy colleagues and
depressed friends in our philosophical theories of self-knowledge because they’re
the exception to the rule, and they do not generalize in such a way that would require
abandoning moderate atomism and moving towards a holist approach. The main
challenge for the moderate atomist is to appeal to some non-arbitrary cut-off point
between normal and abnormal circumstances and to define the right circumstances
in a way that does not beg the question, i.e. does not define countervailing forces in
terms of self-ignorance or self-deception.
3 I say a bit more about the thought/attitude distinction below.
Chapter 1
It’s important to stress from the outset that I will not argue for the alternative holist
account of self-knowledge directly, only indirectly by arguing against moderate
atomism. I will argue that the moderate atomist’s appeal to normal circumstances
is hard to maintain, and that, in the end, the (implicit) ceteris paribus clauses end up
doing all of the work to determine the question of whether or not someone has self-
knowledge. The argumentative strategy pursued in this thesis, then, is to show that
moderate atomism is not a stable position and that it collapses into either radical
atomism or holism.
The notions of ‘atomism’ and ‘holism’ will probably be familiar to most readers,
given that these labels are often used in other areas of philosophy. The distinction
between the two has, for instance, been applied to issues surrounding theories
of meaning (Fodor and Lepore 1992; Hutto 2008), personal identity (Schechtman
1990; Slors 2001), personhood and individual agency (Hobbes 1651; Pettit 1996;
1996; Stoljar 2015), moral responsibility (Vargas 2013) and (moral) reasons for
action (Dancy 2004; 2007). As I will understand these terms, atomism and holism
are labels used to specify the role that circumstances play.4,5
Applied to self-knowledge, atomism can have either of two sources. One is to
be a ‘constructivist’ and think that some procedure or method M (e.g. making up
one’s mind) constitutes one’s attitudes and that by constituting one’s attitudes in this
way, one automatically comes to know about them. A radical atomist constructivist
thinks that by following M one always constitutes one’s attitudes one thereby knows
4 Dancy gives the following description of the distinction, as applied to (moral) reasons for action, which is in line
with the present project: “Atomism holds that any feature that is a reason in favour of action in one case will always
be a reason in favour of action wherever it occurs. The same feature always makes the same reason; or, a reason is
a general reason. This theory is false; something that is a reason in favour of action in one case may in another case
be no reason at all, or even a reason against action. It all depends on the circumst ances; reasons are sensitive to
context” (Dancy 2007, 80). Manuel Vargas draws the distinction, as applied to moral responsibility, along similar
lines: “Let atomism refer to the view that free will is a non-relational property of agents, that is, it is characterizable
in isolation from broader social and physical contexts. ... [atomistic] accounts specify some property—say, a real
self, an uncaused event, the presence of reasoning capacities, or what have you—that, at least in principle, one
could identify simply by looking inside the agent. If the relevant feature is there, then the agent has free will,
independent of context” (Vargas 2013, 204–5). He describes the alterative view as follows: “[The alternative]
picture suggests is that we cannot answer the question of whether an agent has free will simply by looking at the
agent. What we need to know are facts about both the agent and the circumstances. On this picture, free will turns
out to be a relational property, partly constituted by both agent and circumstance, and not the kind of thing that is
settled entirely by the presence of, say, a mechanism of practical reasoning or a general cross-situationally st able
capacity to recognize and respond to moral considerations.” (Vargas 2013, 206).
5 NB I certainly do not mean to suggest that if one accepts atomism or holism in one domain, one would be
committed to accept atomism in some other domain. One can certainly have a holist perspective on self-knowledge
without being a holist (or, in Dancy’s vocabulary, a so-called ‘particularist’) about (moral) reasons for action.
Introduction 17
about; a moderate atomist constructivist, on the other hand, thinks that following
M ‘under normal circumstances’ constitutes one’s attitudes one thereby knows
about. But not all self-knowledge theorists are constructivists. Alternatively,
one might be a ‘reliabilist’ about self-knowledge and think that the relevant
procedure is a highly reliable way of finding out or detecting certain facts of the
matter, and that one is entitled to assume that if one follows M, then one has
‘latched on to’ the relevant (mental) facts. A radical atomist reliabilist thinks
M is a reliable method in all circumstances; a moderate atomist reliabilist
thinks following M only has the desired result in the ‘right’ circumstances.
Despite these differences, what those with an atomist view of self-knowledge
in general share is the (implicit) idea that what makes self-ascription count as self-
knowledge does not require that the subject considers what went on before her self-
ascription or how she will be inclined to act in the future (the subject’s self-ascriptive
‘biography’) nor what’s going on, as it were, around or outside her self-ascription –
the self-ascriptive ‘context’, broadly conceived. Instead, what makes someone’s self-
ascription count as self-knowledge is a question the answer to which can be given
by focusing on what went on during, or immediately preceding, the moment of self-
ascription. This is so, in particular, for expressivist and rationalist views, since both
‘speaking one’s mind’ and ‘making up one’s mind’ are activities one does in the
present, here and now. The result of these activities likewise delivers (knowledge of)
what one thinks here and now. The question, therefore, is whether and why these
methods, which provide the subject with knowledge of her occurrent states, would
ipso facto tell her what her attitudes are: what she really believes, wants, hopes for,
expects, prefers, fears, and so on.
To use a metaphor, we might say that whereas an atomist ‘zooms in’ and
focuses on the methods and the moment of self-ascription, someone with a holist
perspective on self-knowledge instead ‘zooms out’ and considers the place that
a specific self-ascription occupies within the subject’s larger biography and self-
conception, as well as the concrete context in which her self-ascription was made,
including the overall psychological state or ‘mood’ of the self-ascriber. She also
takes into account the moral–political circumstances in which the subject finds
herself, such as whether or not she is subject to the will of another.6 Someone with
6 Another way of looking at it is to say that it’s the atomist who zooms out and looks at different people’s self-
ascriptions from a distance and concludes that some epistemic feature is stable across these cases, whereas a holist
zooms in and considers what is going on and what the good-making epistemic features are in each case.
Chapter 1
a holist view of self-knowledge thinks that the question of whether some subject S
has knowledge of her attitude A by following some procedure or method M can only
be answered by addressing ‘questions of circumstance’. Questions of circumstance
include questions like these: Who is S? Where is S? What will S do? What has S done?
How does S feel? What does S know? Who is S talking to? How is S related to A? How
is S treated by others? How does S think about or treat herself ?
On a holist view, it is not only the brainwashed, drugged and hypnotized among
us who might potentially lack knowledge of our attitudes, in spite of having spoken or
made up our minds but also people who are simply tired, hungry or insecure, who lost
their temper or got out of bed on the wrong side, or people who are distracted, confused,
heartbroken or in love. In other words, people like us. On a holist view, then, the question
of what is required in order for someone’s self-ascription to express self-knowledge can
only be answered by tailoring this question to the specific individual who ascribes an
attitude to herself, the life that she leads and the context in which she finds herself.
2. Qualifications, assumptions and limitations
Before proceeding, two important qualifications are in order. First, I do not mean to
deny in what follows that tipsy colleagues, angry spouses or people under oppressive
circumstances don’t have any self-knowledge, even though the process of my clarifying this
qualification will require us to have some patienc. This will be needed because in Chapter 7
I will suggest that we should not (just) distinguish between different procedures or routes to
self-knowledge, but that it will be helpful to talk about different types of self-knowledge and
allow for a sort of pluralism that at present is absent in the self-knowledge debate.
I also want to highlight an important assumption of this thesis, which is that
I will be exploring theories of self-knowledge, more specifically expressivism
and rationalism as theories of attitudinal self-knowledge, that is, as theories of how
a subject acquires knowledge of intentional mental states such as beliefs, desires,
hopes, intentions, and perhaps (some) preferences and emotions. The contrast is
the knowledge we have of ‘occurrent states’, such as one’s bodily sensations (the
experience of pain or nausea), one’s conscious thoughts or words “running willy-
nilly through her head”, as Harry Frankfurt (1988) would put it.7 Most theorists
7 As we’ll see, ‘judgements’ form a special case, at least on the rationalist view.
Introduction 19
of self-knowledge are explicit about the fact that their ambitions are to explain
attitudinal self-knowledge, not (merely) the epistemic relation we have with regard
to our own conscious thoughts, even though many of these theorists often go on
to equivocate between the two.8 An underlying assumption of this thesis is that
addressing the question of how someone knows her own attitudes is not necessarily
the same as addressing the question of how someone comes to know her occurrent
states, or indeed is not necessarily the same as addressing the question of how
someone knows her beliefs. It may well be that equivocating between propositional
attitudes and conscious thoughts is legitimate and that an account of how a subject
comes to know the latter ipso facto gives us an account of how she knows her own
desires, hopes and intentions, and (some of ) her preferences and emotions. My
plan, though, is not to assume that this is so from the outset. The hope is that the
thesis as a whole will go some way towards explaining why these two ‘domains’ of
the mental are importantly distinct and require different epistemic explanations.
I should also make explicit the limitations of this thesis. As said, I will for the
most part concentrate on expressivism and rationalism. This means I will not be
discussing a number of other interesting theories of and philosophers interested in
self-knowledge in the chapters to come. I will not, for instance, discuss any historic
or contemporary introspectionist accounts (e.g. Armstrong 1968; Nichols and Stich
2003), nor will I have much to say about constitutivist theories which have been fairly
dominant in the self-knowledge debate. One reason for this is that there’s only so
much I can do in the chapters to come, and something has got to give. A better reason,
perhaps, for leaving these theories aside is that they are relatively minimalist theories
of self-knowledge; self-knowledge only requires that the subject is conceptually
competent (constitutivism) and/or that she is equipped with the requisite faculty or
mechanism (introspectionism). It is not unlikely therefore that these theories are,
in current terminology, versions of radical atomism the latter of which I will not,
principally, be concerned with. Furthermore, traditional constitutivist views are
centrally concerned with so-called ‘cogito states’, which concern thoughts of the form
‘I am thinking that I am thinking that P’. In other words, I will for the most part ignore
8 Those who mention being interested in propositional attitudes generally, not beliefs specifically or indeed
occurrent thoughts, and who go on to equivocate between what attitudes and thoughts include are, for example,
Donald Davidson (1987, 2001), Tyler Burge (2013), Annalisa Coliva (2012), Brie Gertler (2011b) and, to some extent,
Peter Carruthers (2011). Exceptions include, for example, Quassim Cassam (2011, 2014), Eric Schwitzgebel (e.g.
2010), Johannes Roessler (2013, 2015b) and Christina Borgoni (2015).
Chapter 1
constitutivism because it is concerned mostly with occurrent thoughts rather than
propositional attitudes (by which I mean attitudes other than so-called conscious
or present beliefs). Whether or not atomist/holist approaches are (im)plausible
when it comes to these occurrent states is not an issue I will be addressing.
Another assumption I should mention is that I take it to be natural to think that
being in a certain mental state is one thing, but knowledge thereof is another. I can
want, feel or perhaps even believe something without knowing that I do; and I can
think that I want, feel or believe something without having the relevant attitudes.
I will be interested, therefore, in (versions of ) theories of self-knowledge that do
not claim that self-knowledge somehow ‘comes for free’ or is ‘self-intimating’,
and which at least allow for the possibility that there could be circumstances that are
epistemically undermining with respect to attitudinal self-knowledge. However, a
constitutivist or, according to my definition, a ‘radical’ version of rationalism will be
discussed in Chapter 6.
Finally, I should be clear that my claim in the chapters to come is not that all
existing theories of self-knowledge are obviously and/or necessarily committed
to (moderate) atomism. Rather, my aim is to examine the extent to which certain
contemporary views implicitly subscribe to a version of atomism. Or, in cases where
it’s not clear whether or not some view is ‘atomist’ or ‘holist’, the strategy will be to
read them along atomist lines for the sake of the argument, and to examine what
the implications and problems are and whether a holist rendering of the account is
possible and is capable of avoiding these problems.
Before moving to the summaries of the chapters to come, let me end this introduction
with a passage written by Wilfrid Sellars, who famously began his essay ‘Philosophy
and the Scientific Image of Man’ as follows:
The aim of philosophy, abstractly formulated, is to understand how things in
the broadest possible sense of the term hang together in the broadest possible
sense of the term. Under ‘things in the broadest possible sense’ I include such
radically different items as not only ‘cabbages and kings’, but numbers and
duties, possibilities and finger snaps, aesthetic experience and death. (Sellars
Introduction 21
The fundamental conclusion of my thesis is perhaps best described as a meta-thesis
about the self-knowledge debate as a whole, urging that before we dive into the
details of this or that procedure, we first of all need to question the preconditions
of self-knowledge: which questions are assumed to have been answered without
receiving explicit treatment? We must recognize that metaphysical questions such
as what we take beliefs, desires and intentions to be, and epistemological questions
about what we take ‘knowledge’ in ‘self-knowledge’ to be, are questions that
dovetail with the question of how someone acquires self-knowledge.
Echoing Sellars, we might say that the more specific aim of the philosophy of
self-knowledge, abstractly formulated, is, or at least should be, to understand how
things in the broadest possible sense of the term hang together in the broadest
possible sense of the term. Under ‘things in the broadest possible sense’ I include
such radically different things as not only selves and thoughts but friendships and
feelings, cultural practices and character traits, morality and propositional attitudes.
I have not even begun to chart, let alone give an account of, all of these and other
(inter)connections. My hope is that the present thesis will at least succeed in showing
that there are such interconnections and that at least some of them are crucial to a
proper understanding of the topic and question of self-knowledge.
3. Plan for the thesis
The thesis is divided into two parts. The first part deals with expressivism and the
second deals with (different versions of ) rationalism. I will spend significantly more
time on rationalism than on expressivism, because rationalism can and has been
understood in different ways, whereas the difference between various accounts of
expressivism appears to be less fundamental. I think it is therefore worthwhile to
consider these different versions of rationalism independently.
In terms of the arguments provided against expressivism and rationalism
generally, it will become clear, after a while, that a certain pattern begins to emerge.
The structure of this thesis is not linear. What I will try to show is that despite the
differences between expressivism and (different versions of) rationalism, these
views are confronted with similar problems and face similar challenges. This
pattern, in its turn, is an indication of the fact that moderate atomism in general is
Chapter 1
In Chapter 2, I describe Dorit Bar-On’s (2004) expressivist view. Expressivism is
usually read as merely offering a non-epistemic theory of first-person authority.
The main goal of this chapter is to describe expressivism taken as a theory of self-
knowledge; more specifically, it is to explain how we might understand that self-
ascriptions, or what expressivists refer to as ‘avowals’, can express self-knowledge.
In so doing, I explain how we might think of avowals as expressing (second-order)
beliefs about one’s mental condition, what would make such beliefs justified and
how an expressivist may approach the question regarding the truth conditions of
In Chapter 3, after having explained how expressivism can be understood as a
theory of self-knowledge, I turn to the question of how it must be evaluated in this
regard. I aim to show that on what seems to be a natural reading of the account,
expressivism betrays a form of moderate atomism. I argue that the atomist elements
in expressivism, and what stands in the way of a holist approach, can be traced
back to the expressivist’s reluctance to see avowals as assertions. I suggest that a
holist construal requires understanding self-ascriptions as assertions that serve
a communicative point but that understanding self-ascriptions in this way is
incompatible with one of expressivism’s central claims.
In the next block of chapters, 4–7, I turn to rationalism. In Chapter 4, I first of
all outline the rationalist view as set out by Richard Moran (2001), and then discuss
the central thesis that we can know our own attitudes by following the so-called
‘transparency procedure’. I discuss a number of common objections to rationalism
and consider ways of dealing with them. I end by discussing the relation between
judgements on the one hand and attitudes on the other, and point out, following
a number of recent authors, that for the transparency procedure to deliver self-
knowledge, it must be construed along inferentialist lines.
In Chapter 5, I distinguish between a ‘radical’ and a ‘moderate’ atomist version
of rationalism. I take issue with the latter view, according to which the transparency
procedure only has the right ‘output’ (self-knowledge) provided that the subject
made up her mind and/or followed the transparency procedure in the right or normal
circumstances of deliberation. I argue that the moderate rationalist is ultimately
unable to provide a satisfying account of what counts as the right circumstances,
at least without ruling out self-deception from the outset. This, then, leaves the
rationalist the choice of construing rationalism along holist lines or ‘going radical’
and ruling out the appeal to normal circumstances from the outset.
Introduction 23
In Chapter 6, I consider an alternative, so-called Activist’ take on rationalism, as
developed by Matthew Boyle (2009b, 2011a, 2011b). I suggest that Boyle’s view
appears to be committed to the radical claim that ‘making up our mind’ is a way of
acquiring self-knowledge whatever the circumstances or reasons involved in one’s
deliberation and so would qualify as a version of radical atomism. My main aim is
to show that the question of whether Boyle’s epistemic account of how one knows
one’s attitudes is plausible and thus whether radical atomism is problematic
depends on how one answers the metaphysical question of what attitudes are. To
this end, I contrast Boyle’s metaphysical view with an alternative ‘Dispositionalist’
view, taking my cue from Eric Schwitzgebel (e.g. 2001, 2010). My conclusion is that
radical atomism is implausible or absurd only given one’s (implicit) answer to the
metaphysical question, which in its turn depends on more fundamental issues such
as one’s views regarding the role (and relevance) of ‘intuitions’ about certain cases,
and one’s meta-philosophical views.
In Chapter 7, I present yet another reading of rationalism by taking the
rationalist account as a normative account of why the capacity to make up one’s
mind matters to autonomy. The connection between autonomy and self-knowledge
has received little attention. I suggest that a thoroughly normative reading of the
rationalist account may have a lot going for it, but that it has not been systematically
compared with other theories of autonomy, such as ‘proceduralist’ theories on
the one hand and ‘relational’ theories on the other. This, then, is what I set out to
do. I then abstract away from the details of rationalism and conclude that we need
to distinguish not just between different objects of self-knowledge or different
routes to them (as Moran proposes, ‘alienated’ versus ‘non-alienated’) but also
between different types of self-knowledge, such as introspective, non-alienated and
autonomous self-knowledge. I finish the chapter by taking a step back to reflect on
the deeper metaphysical and epistemic sources of atomism.
In the concluding chapter, I take stock and summarize some of the key points
in the thesis, reflect on the contribution of this thesis to our understanding of self-
knowledge and spell out possible directions for future research.
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.