OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
Introduction
Johan Gersel and Rasmus Thybo Jensen
Empirical thought is made true or false by how the world is. Through experience subjects
are brought into contact with the empirical world. Hence, the following premise
seems as solid a starting point for philosophical enquiry as any: Perceptual experience
provides reasons for empirical thought. Though not uncontested1 most philosophers
agree that this premise must be true, understood some way or another. Experiences
provide reasons for empirical thought by making our surroundings accessible to us in
some epistemically salient way. The compelling nature of this premise allows one to
ask questions on the form: what must experience/thought/reasons/the world be like
if perceptual experience is to provide reasons for empirical thought? If anything
unites the papers in this anthology it is their outset in a question of roughly this
form. However, the seeming unity in outset is shattered once we look closer at how the
individual authors understand the notions of ‘experience’, ‘reason’, ‘the world’, and
‘accessible’. With different understandings of the key notions, the shared agreement
that perceptual experience must provide reasons disintegrates to a plethora of different
philosophical theories and opinions. Our goal with this introduction is to provide
what can at best be a minimal roadmap, which traces the various agreements and
differences in views.
A fruitful starting point is McDowell’s interpretation of the shared premise, as many
of the papers in this anthology situate themselves as responses to, or elaborations of,
McDowell’s views. The first notion to look at is ‘perceptual experience’. To McDowell
experience is an essentially passive mental occurrence.2 It is thus not simply a species
of belief that is especially tied to perceptual phenomenology or held in some peculiar
way. Experience is a non-doxastic passive mental state attributable to the subject.3
Next in line is the notion of ‘provides’. One way of being the provider of something is
by being the very thing provided. Milk provides nutrition because milk is nutritious.
Another way of being a provider, which we may call mere providing, is by making that
1
2
Davidson (1986) famously rejects this.
McDowell (1996) p. 10.
On this point McDowell’s view is in contrast with the view expounded by Ginsborg in Paper 4 of this
anthology.
3
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
2 Johan Gersel and Rasmus Thybo Jensen
which does the actual providing available. The milkman is a mere provider of nutrition
by making the milk available; the milkman need not be nutritious himself. Equivalently,
experience may be the provider of reasons by figuring as a reason itself or merely by
making available other entities that are reasons or by granting these entities their status
as reasons. We take McDowell to endorse the former position where the experience
itself figures as the reason.4 There are two famous versions of the latter view. On the
first, it is the perceived worldly entities (particulars, states of affairs, facts) themselves
that are reasons. Experience merely makes these entities accessible to us.5 On the
second version it is our perceptual beliefs that figure as reasons and these acquire their
positive epistemic status or justificatory significance through a suitable connection to
experience.6 All of these three views are compatible with explanations of one’s reasons
for belief along the lines of: I believe that P because I saw that P. On what we take to be
McDowell’s understanding, it is my state of seeing, the experience itself, that constitutes
the reason. On the world-based view it is the mind-independent entity itself that is my
reason, only this reason is made accessible to me through my state of seeing. Finally, the
reference to my state of seeing that P can be seen as expressing my belief, or knowledge,
that I see that P, in which case my reason is my belief about my perceptual state, rather
than the experiential state itself.7 While all three versions are thus compatible with
speaking of experience as providing reasons, we still take McDowell to favour the first
view where experience figures as a reason itself.
With these terminological clarifications in hand, we are now in a position to present
McDowell’s more particular version of the shared premise, which he calls ‘Minimal
Empiricism’. Minimal Empiricism claims that passive experiential occurrences or
states must themselves figure as reasons for thought.8 The central motivation for
Minimal Empiricism is the intuition that our thinking must be rationally constrained
by how things are in the world if it is to be recognizable as thinking that aims at getting
right how things are in the world. And what else but our experiences could possibly
4
While this interpretation is controversial, see McDowell (1996) p. 162 where he talks of himself as
‘crediting experiences with rational relations to judgement and belief ’ and later (p. 168) where he criticizes Peacocke on the grounds that his view makes it difficult to see ‘how experiences . . . could constitute a
believer’s reasons’. Later (2009a, p. 268) he also states that judgements are ‘displayed as rational in light of
the experiences themselves, not just in light of beliefs about experiences’. All these quotes suggest a closer
connection between the experience and one’s reason than a role as mere provider.
5
The view that one’s experiential reasons are worldly facts is held by Dancy (2000). McDowell rejects
the attribution of this view to him in McDowell (2006, p. 134).
6
Gauker thinks that this latter solution is all we need to accept in order to have suitable empirical constraint on thought. Ginsborg’s version is close to this in that she thinks experiential reasons are a form of
belief. However, given her identification of experience with a form of belief it is a bit odd to say experience
grants the belief its status; rather, it’s the belief ’s status as an experiential belief normatively constrained by
reality which grants it a special epistemic status.
7
The view put forward by Millar in his contribution is not captured by these three interpretations of the
idea that experience provides reasons. Millar takes the fact that I see that P and not my belief that I see that
P to be my basic experiential reason and regards my seeing that P as belief-involving and as constituting
my knowledge that P.
8
McDowell (1996) p. xii.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
introduction
3
deliver such an external constraint on our thinking? Because what is at stake with
Minimal Empiricism is the very possibility of recognizing our thinking as having
empirical content, McDowell also sometimes refers to it as Transcendental Empiricism
(McDowell 2009b, p. 246). While Minimal Empiricism is indeed a version of our shared
premise its more specific formulation of the premise is in no way accepted by all
participants in this anthology.9
McDowell goes on to ask a question on the form: what must experience be like if
passive experiential states are to, themselves, figure as reasons for thought? He draws
two major conclusions. First of all, experiences must possess conceptual content.10 If
they didn’t they could not figure as reasons for thought at all. This conceptualism has
been widely contested, and in light of our current discussion we can already see one
easy point of contention. One may reject the part of Minimal Empiricism, which states
that the experiences themselves must figure as the reasons. Instead one could cast
experiences in the role of mere providers of reasons that aren’t reasons themselves. This
allows one to acknowledge that the very elements that figure as the reasons themselves,
for example a perceptual belief, are conceptual. One can, thus, acknowledge McDowell’s
claim that the space of reasons is the space of the conceptual11 and acknowledge our
initial premise that experience provides reasons for belief, all the while one denies that
experiences are conceptual. However, the cost of doing so is that one denies Minimal
Empiricism.12 Alternatively, one can accept Minimal Empiricism, but question that
there is any sound argument from the claim that experiences are reasons, to the claim
that they must be conceptual. McDowell claims that any theory which extends rational
relations outside the bounds of the conceptual is a version of the Myth of the Given;
however, one might question that there is anything mythical about such positions.13
McDowell’s second major conclusion is that experiences must be understood as
world-involving if they are to provide adequate reasons for empirical thought.14 If what
we experience falls short of the very reality about which we think, then according to
McDowell, experience cannot figure as an adequate reason for belief. McDowell’s line
of reasoning goes something along the following lines: Knowledge-yielding reasons
for belief cannot fall short of ensuring the truth of what one believes.15 Therefore,
9
Clear adherents are McDowell, Gersel, and probably also Logue. Clear opponents are Gauker,
Ginsborg, and Millar. Travis and Brewer are difficult to position. Their main concern is that what we are
given in experience isn’t conceptually structured. However, it is difficult to discern whether they think that
our experiencing what is given or merely that which is experienced constitutes our reason. Importantly,
Travis, Brewer, Gauker, Ginsborg, and Millar all accept something like our shared starting premise in its
unspecific form.
10
11
See especially lecture two of McDowell (1996).
McDowell (1996) p. 14.
12
This line of response is taken by Gauker. Ginsborg is a bit trickier insofar as she equates experience
with a form of perceptual belief. She accepts the conceptual nature of experiential reasons, but denies their
passive nature.
13
This is roughly the response adopted by Travis in this anthology and Brewer (2011). Gersel defends
and elaborates McDowell’s appeal to the Myth of the Given.
14
McDowell (1996) p. 26.
15
See McDowell (1998a, 1998b). In his contribution, Gauker voices worries about such a requirement.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
4 Johan Gersel and Rasmus Thybo Jensen
if experiences are to figure as knowledge-yielding reasons for belief about external
reality, then what we encounter in experience must be this external reality itself. Presented
this way McDowell’s arguments lead him from requiring truth-ensuring reasons to
the acknowledgement of epistemological disjunctivism. Once one requires that reasons
be truth-ensuring then one cannot avoid scepticism without attributing different
justificatory status to hallucinations and perceptions. Hallucinations do not ensure
the truth of the empirical beliefs based upon them. Hence, if perception is to provide
truth-ensuring reasons it must provide reasons with a justificatory force different from
that of subjectively indistinguishable hallucinations. Now, if our reasons must have this
truth-ensuring character and experiences themselves must figure as the reason, then it
is natural to think that we cannot be in the same metaphysical type of experiential state
when we perceive as when we hallucinate. Hence, beyond his obvious epistemological
disjunctivism, McDowell might be viewed as committed to metaphysical disjunctivism as well. This view claims that we cannot metaphysically account for hallucinations
and perceptions in terms of a single shared type of mental state, which simply differs in
its external relations to reality. Instead, perceptions are metaphysically considered as
some form of relational state in which the world figures as an inherent Relatum. However,
it isn’t clear that McDowell needs to go this far. He could restrict his characterizations
of experiences as relational, or world-involving, to explanations in which we are
concerned with the justificatory role of experience, and remain neutral as to how this
relates to experiences when they are considered purely metaphysically, that is, independently of their role in epistemic explanation.16 However, we do not think this is indeed
McDowell’s strategy and the feasibility of this approach has been questioned by both
Brewer (2011) and Travis (this anthology), who explicitly endorse both metaphysical and
epistemological disjunctivism.17 An argumentative strategy different from McDowell’s,
but with largely the same consequence, is to start one’s argument from the premise of
metaphysical disjunctivism. One could then take the possibility of truth-ensuring
reasons (and thus epistemological disjunctivism) to simply be a positive upshot of one’s
independently motivated metaphysical theory, rather than consider such truth-ensuring
a requirement on the adequacy of reasons.18
Both metaphysical and epistemological disjunctivism leave the notion of phenomenology in a somewhat awkward position. Traditionally, people have tried to
motivate the similarity in status, epistemological or metaphysical, between perception
and hallucination in terms of their indistinguishability to the involved subject. This has
often gone through some argument to the effect that shared phenomenology accounts
for both the indistinguishability and metaphysical or epistemological sameness.
16
Another way to avoid metaphysical disjunctivism, but not one congenial to McDowell, is once again
to reject that experiences themselves must figure as reasons for belief. This is the strategy that Millar
explores in this anthology.
17
See Haddock and Macpherson (2008) for a discussion of the relation between McDowell’s epistemological disjunctivism and metaphysical disjunctivism.
18
This seems more in line with Brewer’s (2011) later style of argument.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
introduction
5
Epistemological and metaphysical disjunctivism thus seem to sever the envisaged
connection between phenomenology and epistemic and/or metaphysical status. While
McDowell indeed acknowledges that his arguments take no basis in phenomenology,19
others have found this severance problematic. What has been called phenomenal disjunctivism (Haddock and Macpherson 2008) is thus a third branch of disjunctivism
which claims that the phenomenology can differ between subjectively indistinguishable
experiences. While McDowell takes no stand on this form of disjunctivism, others
have investigated whether his view has consequences in this regard.20
At this stage it may seem that whatever unity there is among the views found in this
anthology, it is at a level of abstraction with little philosophical bite. It seems like little
to no agreement is left once we elaborate in more detail on the concepts involved in the
shared premise that experience provides reasons for thought. This conclusion would,
however, be too quick. A dominant view within mainstream epistemology is some
form of externalism of justification where all that is required for X to justify S’s thought
is that S’s thought is based on X and that this makes S’s thought reliably true. If one
holds this view, very little of interest can be deduced from the shared premise that
experience must provide reasons for thought. All it would entail was that some reliable
connection, which involved experience, must exist between a series of a subject’s
thoughts and the world those thoughts concern. The reason such a fruitful debate
exists between the authors of this anthology is that they all acknowledge some form of
internalist restriction on rampant epistemological externalism. Somehow, in some
way, the subject must be able to access or appreciate the connection furnished between
experience and the world about which she thinks. Once again, how one spells out this
internalist requirement will differ between the authors. At a minimal level one might
simply require that the subject must be conscious of that which figures as her reasons.
A stronger requirement endorsed by McDowell is that the subject must be able to
appreciate her reasons as such, whatever that may require more precisely.21 Thus even
though the precise role experience plays in providing reason differs between authors,
and while the specific versions of disjunctivism endorsed, if any, also differ, there is
still a core agreement that the subject’s consciousness should play a crucial role in our
characterization of her epistemic position. This puts the current anthology somewhat
off-centre on the current epistemological scene. To us one of the most interesting
things about the papers included here is precisely the way they contribute to the quest
for the holy grail of epistemology (Pritchard 2009, p. 472): some form of non-sceptic
internalist theory of empirical knowledge.
There is of course much more agreement between individual authors in the book
than what may be stipulated within the group of authors as a whole. As the reader will
find, each of the individual authors in this book engages with at least one, but most
19
20
21
In discussion at the workshop that led to this anthology.
Logue’s paper is concerned with the relation between epistemological and phenomenal disjunctivism.
Gersel attempts to spell out McDowell’s and other competing notions of internalism in more detail.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
6 Johan Gersel and Rasmus Thybo Jensen
often several, of the other authors. This engagement is most often based on the significant
background of agreement that helps make philosophical discussion fruitful and constructive, rather than stubborn and destructive. We hope you will read these papers with
the same constructive spirit in mind. The following section will present the included
papers in more detail and try to highlight the individual debates rather than focusing
on the overall picture.
The Papers of the Anthology
While the papers have connections and ties that crisscross the whole anthology, some
of them are more directly engaged with the same issues. Our division into four separate
topics should therefore be seen more as a hint to their focus than as an expression of
a sharp segregation.
1. The Myth of the Given
The first four papers all explicitly engage with the Myth of the Given. This notion originates with Sellars (1956/1997), but has figured as one of the central driving premises in
McDowell’s thinking on experiential reasons. Again borrowing from Sellars, McDowell
(1996) employs the notion of the space of reasons. The space of reasons is the logical
space in which we account for the occurrence and propriety of certain mental states in
light of what reasons the subject has for being in those states. The Myth of the Given
consists in the extension of the space of reasons beyond its legitimate bounds. Mythical
theories are those which argue that certain types of entities can figure as a subject’s
reasons even though those entities are of such a kind that they are unsuitable to fulfil this
role. The core of McDowell’s thinking on experiential reasons starts from the claim that
any theory which extends a subject’s reasons beyond what is conceptually given to the
subject is a form of the Myth of the Given. In slogan form: the space of reasons is the
space of concepts.22 This idea in combination with Minimal Empiricism leads him to
the conclusion that experiences must possess conceptual content, as experiences must
be able to figure as reasons for a subject’s thinking.
Earlier Travis (2013a) has argued against the view that experiences have content. Yet
given his adherence to the idea that experience must provide the subject with reasons,
and perhaps even to Minimal Empiricism as a specific interpretation of this idea,
Travis has felt obliged to explain how his theory does not succumb to the Myth in spite
of McDowell’s arguments to the opposite conclusion (Travis 2013b, 2013c). Travis’s
arguments have typically had two prongs. On the one hand he has attempted to show
that if experiences were conceptual then they couldn’t provide adequate reasons for
thought. The second prong has been to elucidate the constraints implied by the danger
22
McDowell (1996) p. 14.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
introduction
7
of the Myth and argue that his theory doesn’t transgress against these. McDowell’s
present paper is a response to Travis (2013c). He argues that Travis is wrong to think
that once we consider experiences to be conceptual then we lose access to the reality
that makes our judgement true.
Travis’s paper in this anthology continues their debate. The first line of attack remains
constant insofar as he still considers conceptually laden experiences to be unable to
present adequate reasons for thought. However, his second line of attack has altered.
Instead of accepting the idea of the Myth of the Given, Travis now expresses that he
is unable to feel the pull of McDowell’s account of the Myth of the Given. He tries
to expose various false premises, which he envisages may have led McDowell to accept
the constraint on the space of reasons which is elaborated via the idea of the Myth
of the Given.
Gersel’s contribution can be seen as an attempt to supply the elucidation of the
Myth of the Given which Travis claims is lacking. The goal of that paper is to clarify
why the claim that reasons are connected to the possibility of self-conscious scrutiny
has the consequence that the space of reasons is limited to the space of the conceptual.
The paper thus attempts to sharpen the disagreement between Travis and McDowell
by locating an argument that could defend the actuality of the Myth, while eschewing
the false premises that Travis thinks drive the argument. McDowell’s and Gersel’s
papers can thus in unison be seen as an attempt at answering respectively the first
and the second prong of attack present in Travis’s contribution.
Ginsborg’s paper engages with the Myth of the Given at one step’s remove. For
McDowell the insight of Davidson’s dictum that only a belief can justify a belief is that
only items with conceptual content can play a justificatory role. Ginsborg agrees but
finds the further insight that only a committal, and thereby active mental state, can
serve as a reason. This, however, brings her into conflict with Minimal Empiricism as
envisaged by McDowell. The purpose of Minimal Empiricism was to ensure a rational
yet passive constraint on our thinking. Ginsborg’s view of the nature of reasons forces
her to reject that experiences understood as passive can play such a rational role.
However, Ginsborg argues, this does not rule out that experiences can still play a
rational role because we can and should regard experience as itself belief-involving.
This move brings her position dangerously close to Davidsonian coherentism, with
its total dismissal of Minimal Empiricism—a position she, with McDowell, takes to be
untenable. Her solution is to separate two elements that are put together in McDowell’s
understanding of Minimal Empiricism, namely the idea that our thinking must be
normatively constrained by the world and the idea that our thinking must be rationally
answerable to experience.
McDowell accounts for the normative relation between world and belief via the
rational, normative relation between experience and belief. Ginsborg also takes
experience to stand in a rational relation to beliefs, only she conceives of experience
as belief-like. What clearly distinguishes Ginsborg’s position from Davidson’s is her
account of experience in terms of a normative, but non-rational relation between the
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
8 Johan Gersel and Rasmus Thybo Jensen
object of experience and the concept-application involved in experience. Our seeing
that a cube is green intrinsically involves recognition of the appropriateness of this
very perceptual, yet committal response to the cube. However, the appropriateness in
question is conceived as more primitive than the rational appropriateness of a judgement in light of certain reasons. Here Ginsborg introduces a notion of judgement and
conceptual capacities that is less demanding than the one McDowell works with. The
object is experienced as calling for a certain concept application, but such sensitivity to
norms does not require that one has been initiated into the space of reasons through
language acquisition. Importantly, the idea of an object calling for a certain concept
application is not claimed to be intelligible independently of the child’s actual ability to
apply such non-linguistic concepts. For this reason, Ginsborg argues, her account does
not succumb to the Myth of the Given.
The first four papers can be seen as aligned in roughly the following way. McDowell,
Gersel, and Ginsborg agree that the space of reasons is exhausted by the space of the
conceptual. Ginsborg and Travis agree that given how experiences are conceived by
McDowell, as passive and conceptual, they cannot provide reasons for belief. However,
Travis has a problem with the conceptual nature attributed to experiential reasons,
whereas Ginsborg objects to their passivity. As Minimal Empiricism is the idea of a
passive constraint on thought, this means that Travis23 can share the view with McDowell
and Gersel that Minimal Empiricism must be understood as a requirement for experiential reasons. Ginsborg is forced to reject such a view, but argues that a normative, yet
non-rational, guidance of experience by the world suffices for possession of empirical
concepts. However, Travis will have to reject the conceptual nature of experience and
thus his position is challenged by McDowell’s claim that all such views are versions of
the Myth of the Given.
2. The Epistemology of Empirical Knowledge
The following two papers still explore the general themes discussed by the first four
papers, only they are more narrowly focused on epistemic issues. Gauker’s contribution aligns with Ginsborg in questioning the idea that experiences distinct from
doxastic states can serve as reasons for beliefs. However, rather than argue that the
activity involved in belief-formation is a virtue not a vice, he suggests that belief-like
states can possess exactly the kind of involuntary and passive nature that McDowell’s
Minimal Empiricism requires.
The more specific target of Gauker’s paper is an argument that attempts to show that
experiences can justify beliefs on the ground that sentences of the form ‘A looks F’ can be
used to justify claims of the form ‘A is F’. Arguably, if such an argument is to succeed
sentences of the form ‘A looks F’ must report the content of experience. Gauker presents
23
On the assumption that Travis thinks it is the experience of the worldly entity, which is a reason for
thought, rather than merely the worldly entity. As mentioned, it is unclear what Travis’s view on this issue is.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
introduction
9
what he takes to be an exhaustive argument to the effect that if ‘A looks F’ reports the
content of experience then, on any relevant conception of a justificatory link, such
sentences are unable to justify the claim ‘A is F’. As Gauker notes, his paper targets a
notion of experiential reasons that takes such reasons to be non-conclusive. This leaves
us with an interesting question as to whether similar arguments will be effective against
someone who argues that experiences can provide truth-ensuring warrant for judgements, which is precisely what McDowell claims.
Leddington’s starting point is the basic intuition of infallibilist epistemology:
Knowledge-yielding reasons must be truth-ensuring. His paper challenges the infallibilist to explain how their view of truth-ensuring experiential reasons is compatible
with the undeniable fact that we are at times misled in our pursuit of perceptual
knowledge. McDowell explains the possibility of error in terms of the fallible nature
of our perceptual capacities for knowledge. Thus, we may exercise our perceptual
capacity to the best of our ability and yet be in a position where it merely seems to us
that we perceive, that is, where it only seems to us that we are in a position to know on
the basis of experience. Leddington argues that such a view is untenable. In a first step,
he argues that allowing for both non-defective and defective exercises of a perceptual
capacity for knowledge commits us to the idea that, even under optimal conditions,
such a capacity may fail. In a second step, he argues that such a conception of our
fallibility itself gives us a concrete reason to rationally doubt, on any given occasion,
that we in fact perceive. As a consequence, we can never be in possession of the kind
of self-conscious, conclusive reason that both he and McDowell thinks is needed for
knowledge. This is not the end of infallibilism according to Leddington. In Millar’s
work he finds a different conception of fallibility that takes every exercise of a knowledgeyielding capacity to be successful. On such a view, our fallibility is explained in terms
of seeming cases of exercising such a perfect capacity rather than in terms of defective
exercises of an imperfect capacity.
Leddington takes his argument to leave open what exactly constitutes our perceptual
reasons, whether it is, for instance, our experiences as particulars, facts about experiences, or the perceived object. Furthermore, his conclusion stays neutral between two
different conceptions of where we should place our infallible capacities for perceptual
knowledge. On McDowell’s conception these capacities are perceptual capacities
distinct from judgemental capacities, whereas Millar places them at the level of judgement. Leddington favours Millar’s view but emphasizes that the understanding of
the fallibility of infallible capacities he presents is, on this point, compatible with
McDowell’s view.
3. The Nature of Experience
The following two papers discuss the nature of experience itself. Their primary concern is thus within philosophy of mind. However, both Millar and Logue regard it as a
constraint on any account of the nature of experience that it must allow us to see
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
10 Johan Gersel and Rasmus Thybo Jensen
how our reasons for perceptual judgements can be such as to ensure the truth of our
perceptual judgements.
Millar questions whether an acceptance of epistemological disjunctivism should
lead us to embrace metaphysical disjunctivism in the form of a relational view of
experiences. Millar’s general strategy is to show that we can preserve what he takes to
be the real insight of relational views within a sophisticated non-relational view. The
real insight of relationalism is that both an adequate account of perceptual reasons and
of perceptually based demonstrative thought must appeal to relational mental items.
Millar argues that the main arguments, to the effect that a non-relational view cannot account for perceptual, demonstrative thought, only succeed if the non-relational
view implies that the primary object of perception could not be a mind-independent,
physical object, an implication no sophisticated non-relationalist need accept. The
positive account of perceptual, demonstrative thought sketched by Millar is inspired
by Evans’ account in The Varieties of Reference. Millar proposes that we should understand
the perceptual awareness involved in perceptual discrimination as constitutively
dependent, not only on the sensory experience as such, but also on certain behavioural
dispositions that relate us to the mind-independent object itself.
In the last part of his paper, Millar discusses McDowell’s Minimal Empiricism. He
accepts that we need to regard our thinking as answerable to experience if we are to
regard it as possessing empirical content. McDowell thinks this demand can only be
fulfilled if it is experience understood as a passive, non-judgemental state or occurrence
that constitutes our reason. Millar urges that it is sufficient that our empirical beliefs
are answerable to what we know perceptually. If I possess the right recognitional
capacity I can perceptually know that a seen bird is a magpie. A distinct recognitional capacity also applied to the bird allows me to know that the bird is seen by me
(see also Millar 2011). It is the fact that I see that the bird is a magpie that constitutes
my reason for the belief that the bird is a magpie. Here Millar departs from Minimal
Empiricism on two points. First, it is the fact that I see that P which constitutes my
reason, not the experience as such. Second, my seeing that P is understood as constitutively involving my belief that P. On the second point Millar is in agreement with
Ginsborg; on the first point Millar and Ginsborg diverge, since Ginsborg takes it to be
the experience as such, understood as a doxastic state, that justifies.
Logue’s paper also concerns the consequences that epistemological disjunctivism has
for our theory of experience. She presents her version of metaphysical disjunctivism in
order to develop a new and improved version of epistemological disjunctivism.
A question facing the epistemic disjunctivist is to explain why experiences that are
subjectively indistinguishable can provide different reasons for a subject. Millar’s
account allows him to evade this question because he denies that experience as such
constitutes our reasons. A more common response is to argue that the epistemic role of
an experience is not determined by phenomenal character. There are two typical suggestions: either it is said that it is the content that differs between hallucinations and
perceptions; or it is said that the one state is a direct conscious relation to reality
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
introduction
11
whereas the other is not. Logue goes against this trend in wanting to defend disjunctivism,
all the while claiming that the rational contribution of an experience is constituted
by its phenomenal character. Her suggestion is to reject that phenomenal character is
what explains the subjective indistinguishability between perceptions, illusions, and
hallucinations. She argues that phenomenal character is constituted by a relation
between the perceiver and her environment. An immediate consequence of this view
is that hallucinations do not possess any phenomenal character. Cases of illusions are
characterized by Logue as cases where the subject perceives the real property of the
object but the object appears differently from how it is.
To explain the indistinguishability in question, Logue appeals to the idea that hallucinations and illusions present defective contexts for our capacity for self-knowledge.
Inspired by Evans she adopts an outward-looking model for self-knowledge according
to which I gain knowledge of the phenomenal character of my experience by attending to
the perceived object. Because there is no such object in the case of hallucinations, such
cases present a defective context for our capacity for self-knowledge. The defectiveness
of the context provides part of the explanation for how we can be fooled into believing that the hallucinatory experience does have a phenomenal character. In cases of
illusions, we are said to be blocked from attending to the real property of the object
and this again provides an unfriendly context for the exercise of our outward-looking
capacity for self-knowledge. The general strategy here is similar to McDowell’s. Just
because an exercise of a capacity for knowledge can, under non-favourable circumstances, mislead one into thinking one is in a position to know, we should not conclude
that such a capacity for knowledge cannot, under any circumstances, provide us with
knowledge. This strategy commits Logue to an explanation of our fallibility that goes
against the one favoured by Leddington and Millar.
On Logue’s view it is the phenomenal character of experience qua mental state that
grounds our perceptual judgements, though we only gain knowledge about such characters by attending to the world. Here we find an interesting parallel between Logue’s and
Millar’s views. They both take perceptual knowledge about the world to be explanatorily
prior to our access to the reasons that ground such knowledge; something they both
claim is compatible with a reasonable internalism concerning justification. On Millar’s
view, perceptual knowledge is even possible without the capacity to access perceptual
reasons, which makes his view come out as clearly at odds with what Cunningham in
his contribution calls ‘the Reasons Priority Thesis’.
4. The Object of Experience
The two previous papers considered the rational roles of different types of experiential
states and investigated how our obligation to account for these differences restricted a
feasible theory of experience. Another way of elucidating the nature of experiential
reasons is by considering what limits are placed on the objects of experience when they
must be such that perception of such objects can provide reasons for thought.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
12
Johan Gersel and Rasmus Thybo Jensen
Brewer defends the view that there exist natural continuants. These are mindindependent objects that are naturally unified over time in such a way that no two
natural continuants share precisely the same location at any time. He defends this
view by arguing that, if we are to understand our experiences as bringing us into contact with a mind-independent reality then we are committed to the view that we
encounter natural continuants in experience. According to Brewer it is only if we view
our experiences as encounters with natural continuants that we can make sense of our
experiential conditions as jointly determined by our spatio-temporal route and by
what is there anyway, present to be viewed independently of whatever location we may
be at. Brewer’s claim is that only if we can provide such a simple theory of our perceptual condition can we understand ourselves as encountering the mind-independent
reality in perception. Brewer’s defence of this position is tied to his endorsement of
naïve realism, according to which experiences do not possess content of any kind.
Rather, like Travis, his view is that the experiences involved in perception are a simple
conscious relation between a subject and the entities present in his surroundings.
In the final paper of the anthology, Cunningham argues against what he calls the
truth-maker theory of reasons, a view he tentatively ascribes to Brewer. The truth-maker
theory of reasons argues that the concrete entities, which make our empirical judgements true, can also count as our reasons for those judgements. It seems a reasonable
claim that we can perceive those entities. Hence, our perceptual reasons can be the
truth-makers of our judgements. Given their commitment to a relational conception
of experience and the idea that one’s experiential reasons are the perceived entities
missing in hallucinatory cases, truth-maker theorists are committed to both metaphysical and epistemological disjunctivism. Cunningham’s argument is built on the
idea that reasons must be capable of being identical to the explanantia of rationalising
explanations. In effect Cunningham is arguing that we should identify what Ginsborg
separates as reasons1 and reasons2. Reasons1 are reasons understood as considerations
that favour certain beliefs. Reasons2 are reasons ascribed to the subject from a third
person point of view in order to make her behaviour rationally intelligible. Cunningham
considers particulars (objects, properties, and events) and states of affairs as entities
that could be truth-makers for our beliefs. Each of these connects with a version of the
truth-maker theory of reasons. He then argues that neither particulars nor states of
affairs are capable of playing the required rationalizing explanatory role. The only
alternative candidate left for the role as both reason and rationalizing explanans,
Cunningham argues, is the category of true propositions.
Comparing Brewer’s and Cunningham’s arguments we may say the following:
Brewer’s argument starts from a presumption of naïve realism and argues that if we are
to understand the nature of experience in the light of a simple theory of perception,
then we must conceive of the objects we encounter in experience as natural continuants. Cunningham argues that, if all we are given in experience are entities such as
Brewer’s natural continuants, then what is given in experience cannot be identified
with our experiential reasons.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
introduction
13
A Brief Overview
The following questions highlight some of the important fault lines throughout
the papers of this anthology: Do experiences themselves figure as reasons or are
they mere providers of reasons? Are experiential reasons belief-independent? Do
our experiential reasons consist in facts, states of affairs, or particulars? Do they
consist in mind-dependent or mind-independent entities? Should we conceive of
experiences as having content? Should we conceive of the experiences involved in perceptions as relational? Should we conceive of experiential reasons as truth-ensuring?
Assuming that such reasons are truth-ensuring, how should we account for our
fallibility? Depending on which question we ask, different divisions amongst the
authors will show up.
Most, if not all, of the papers in this anthology favour the idea that experience
must be capable of providing the subject with reasons. However, Gauker, Millar, and
Ginsborg explicitly oppose the idea that experiences understood as belief-independent
states or occurrences can as such figure as reasons. Gauker suggests that beliefs caused
by experience may act as entry-level justifiers, whereas Ginsborg takes experiences to
be intrinsically belief-involving. Millar departs from Gauker and Ginsborg in taking
the relevant notion of reasons to be reasons understood as facts (‘I see that P’) and not
as mental states or occurrences. Cunningham argues in favour of Millar’s general conception of reasons. Millar, however, is in line with Gauker, Ginsborg, and Logue in
thinking that perceptual reasons are mind-dependent entities, whereas Travis and
Brewer seem to regard the mind-independent objects of experience as reasons.
Cunningham puts pressure on the idea that we can account for perceptual reasons
without ascribing content to experience, while Millar thinks there is a serious question
of whether we need ascribe any content to experience in order to give such an account.
Travis and Brewer give a negative answer to this question whereas McDowell and Gersel
maintain that we need to attribute conceptual content to experience. Travis, Brewer,
and Logue all share an explicit commitment to both epistemological and metaphysical
disjunctivism. Millar agrees that our perceptual reasons must be truth-ensuring but
argues that the crucial disjunctive move in our explanation of perceptual knowledge
should be located at the level of judgement not at the level of experience. This view is
also favoured by Leddington, who argues for Millar’s conception of our perceptual
fallibility opposing the views of McDowell and Logue.
On the grand philosophical scene most of the contributions to this anthology are in
wide agreement as to how we should approach an investigation of experiential reasons.
Some internalist notion of reasons and some kind of disjunctivist inclinations shape
or form a central part of most of the arguments. However, once the details are in focus
the appearance of unity is shattered. There may be many views on perception and
rationality that have had no say at all in this anthology. However, its explicit goal is to
present and further a debate within the general philosophical approach exemplified by
the contributions.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
14 Johan Gersel and Rasmus Thybo Jensen
We hope that you, the reader, will find the book as interesting and illuminating as we
have found the editing and writing of this book.24
References
Brewer, Bill. 2011. Perception and its Objects. Oxford: Oxford University Press.
Dancy, Jonathan. 2000. Practical Reality. Oxford: Oxford University Press.
Davidson, Donald. 1986. “A Coherence Theory of Truth and Knowledge”. In Truth and
Interpretation: Perspectives on the Philosophy of Donald Davidson, edited by Ernest LePore,
pp. 307–19. Oxford: Basil Blackwell.
Haddock, Alan and Fiona Macpherson. 2008. “Introduction: Varieties of Disjunctivism”. In
Disjunctivism: Perception, Action, Knowledge, edited by Alan Haddock and Fiona Macpherson,
pp. 1–24. Oxford: Oxford University Press.
McDowell, John. 1996. Mind and World. Cambridge, MA: Harvard University Press.
McDowell, John. 1998a. “Knowledge by Hearsay”. In Meaning, Knowledge, and Reality, pp. 413–43.
Cambridge, MA: Harvard University Press.
McDowell, John. 1998b. “Knowledge and the Internal”. In Meaning, Knowledge, and Reality,
pp. 395–413. Cambridge, MA: Harvard University Press.
McDowell, John. 2006. “Response to Dancy”. In McDowell and his Critics, edited by Cynthia
McDonald and Graham McDonald. Oxford: Blackwell Publishing.
McDowell, John. 2009a. “Avoiding the Myth of the Given”. In Having the World in View: Essays
on Kant, Hegel and Sellars, pp. 256–72. Cambridge, MA: Harvard University Press.
McDowell, John. 2009b. “Experiencing the World”. In The Engaged Intellect, pp. 243–56.
Cambridge, MA: Harvard University Press.
Millar, Alan. 2011. “Knowledge and Reasons for Belief ”. In Reasons for Belief, edited by Andrew
Reisner and Asbjørn Steglich-Petersen, pp. 223–43. Cambridge: Cambridge University Press.
Pritchard, Duncan. 2009. “Wright Contra McDowell on Perceptual Knowledge and Scepticism”.
Synthese 171: 467–79.
Sellars, Wilfred. 1956/1997. Empiricism and the Philosophy of Mind. Cambridge, MA: Harvard
University Press.
Travis, Charles. 2013a. “The Silence of the Senses”. In Perception: Essays after Frege, pp. 23–58.
Oxford: Oxford University Press.
Travis, Charles. 2013b. “Reason’s Reach”. In Perception: Essays after Frege, pp. 118–43. Oxford:
Oxford University Press.
Travis, Charles. 2013c. “Unlocking the Outer World”. In Perception: Essays after Frege, pp. 223–58.
Oxford: Oxford University Press.
24
Acknowledgement: We would like to thank Hagit Benbaji, Joe Cunningham, Christopher Gauker,
Alan Millar, and Hannah Ginsborg for their written feedback on prior versions of this introduction. Thanks
also to the two Oxford University Press referees for their critical comments and constructive advice.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
The Travis–McDowell Debate
Johan Gersel
The included papers by John McDowell and Charles Travis deserve a separate
introduction. They are the latest instalments in an ongoing debate that has stretched
across a series of exchanges (the main texts are Travis (2013a), Travis (2013b),
McDowell (2009), Travis (2013c)). The present paper is meant as an aid to the reader
unfamiliar with these earlier exchanges or to the reader merely in need of a reminder
of the discussion. The outset of the debate is McDowell’s view which, as mentioned,
holds that experience must be conceived both as possessing conceptual content and as
a passive element of cognition, separate from the active judgements one might form on
the basis of undergoing some experience. McDowell’s motivation for endorsing this
view is his conviction that only if the passive element in our cognition, ‘experience’ in
McDowell’s terminology, is conceptual can we make sense of how it provides a rational
constraint on our thinking. Travis’s debate with McDowell consists of two lines of argument. The first line is developed in ‘The Silences of the Senses’, where he argues against
McDowell and other representationalists that the passivity of experience is incompat
ible with its possession of representational content. The second line not only defends
the view that experiences can provide reasons for thought without having content, but
also argues, based on interpretations of Frege, that only if experience provides us with
acquaintance with the nonconceptual can it provide the proper constraint on thought.
This second line of argument is developed in his ‘Reason’s Reach’ and ‘Unlocking the
Outer World’. I want to briefly present both arguments and McDowell’s response.
In ‘The Silence of the Senses’ Travis argues that whichever way experience provides
reasons for thought it must be in terms of how things look to the subject of the experience
(2013a, p. 34). However, according to Travis, we can only sensibly talk of ‘looks’ in two
ways. On the one hand, there is how things look. This is an objective feature that vari
ous things have under various conditions. In this sense, a wax lemon looks like a lemon
under most conditions. A blue shirt in this lighting may look exactly like a white shirt
looks under certain different lighting conditions. Looks in this sense are objective and
independent of any activity of subjects. They are fully determined by features of the
visual appearances that things possess in certain circumstances (ibid. p. 35). Hence, if
experience is conceived as awareness of how things look in this sense, then experience
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
16
Johan Gersel
can indeed be conceived as a form of passive awareness of our surroundings. However,
this sense of ‘looks’ is unsuited to provide any determinate representational content for
the experience. Representational content presents the world as being a certain deter
minate way. For the representation to be true, correct, or veridical is just for the world
to be in that determinate way. However, on the sense of ‘looks’ under consideration
there is no determinate way the world should be for it to be as it looks. For given how
things look, one could equally well be aware of a lemon, a waxlemon, or a strangely
shaped orange under weird lighting conditions, or any number of different objects. An
encounter with either of these would be compatible with the world looking as it does.
A myriad of things has what are objectively similar looks under various conditions,
and a particular look gives no priority to one of the scenarios in which it obtains. Hence,
while the first sense of ‘looks’ may be something we are passively aware of, it is not
something which can suffice to determine any representational content (ibid. p. 37).
Things are opposite for the second sense of ‘looks’. ‘Looks’ in this sense are ways in
which people take things to be, or should or would take things to be if they went by
their visual appearance alone. It is in this sense that we can say: ‘it looks to me as if Pia
will sink the put’; meaning I think that she will sink it given how things look to me.
Likewise, we can say of someone else that it looks to him as if P, most often in order
to explain why he judged that P (he based his judgement on his experience) when we
know that P is in fact not the case (ibid. p. 40). This second sense of ‘looks’ certainly
suffices to settle a way the world must be in order for it to be as it looks. Pia must sink
the put, and P must be true. However, that something looks like this to someone isn’t
a passive element of cognition anymore. Rather, it is an active judgement in which
someone takes the world to be a certain way in light of his experiences, or at the very
least a claim that one ought to judge so if one went on how things looked alone. Travis
argues that these two notions of ‘looks’ are the only ones available, and concludes that
McDowell’s view of experience as both passive and representational is incoherent
(ibid. p. 47).
His rejection of conceptualism about experience leaves Travis a supposed victim of
the Myth of the Given as it is developed by McDowell. McDowell famously claims that
only if experiences present us with conceptually structured generalities, things being
some way, can experience provide reasons for the perceptual judgements in which
we judge things to be those very ways (McDowell 1996). To McDowell (2009, p. 264)
experience counts as conceptual as long as it presents the world as being in some
general way it could be judged to be. Travis (2013c, p. 237) has a more liberal notion
of the conceptual according to which we should consider something to fall within the
conceptual as long as it presents things as being a way at all. The key difference is that
McDowell’s definition leaves logical room for nonconceptual representational con
tent, whereas Travis’s does not.1 However, as long as Travis maintains the generality of
1
Importantly, McDowell doesn’t think we can suffice with attributing nonconceptual content to
experience, but he acknowledges a distinction between conceptual and nonconceptual content. On the
other hand, when Travis objects to attributing conceptual content to experience he means this to target any
representational view of experience.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
The Travis–M c Dowell Debate
17
all representation and the utter particularity of what is given in experience their
disagreement is genuine in spite of the slight difference in terminology. Both McDowell
and Travis agree that we must hold onto the idea that experiences figure as a reason
for thought, that is, to Minimal Empiricism.2 If McDowell (1996) is right, Minimal
Empiricism entails that experience must have conceptual content. Yet, according to
Travis, experience cannot be conceptual for there is no particular way in which experi
ence presents the world as being. Only subjects can take the world to be in some way
and that is not something passive, but rather an active commitment. Experience
merely acquaints us with things being as they are; it presents us with what Travis, in his
contribution to this anthology, calls ‘the unfolding of the historical’. Travis responds
to the challenge of explaining how the nonconceptual can provide reasons for thought
in ‘Reason’s Reach’. However, his argument has a peculiar structure. Rather than
showing us how an encounter with something nonconceptual can provide a reason for
thought, he argues that if anything that is given in experience can provide a reason
for thought at all, then it must be possible that something nonconceptual can provide
reasons. He follows Frege (Travis 2013b, p. 123) in drawing a line between particular
things, like the setting of the sun, and generalities, like that the sun has set. To Frege,
and Travis, the former, a particular, is a thing that can be perceived, whereas the latter,
a way that things can be, is not something that can be perceived at all, and thus not
something we can be given through perception. Hence, if anything that is given to us in
perception forms a reason for thought, then something nonconceptual must be able
to figure as a reason.3 For general ways that things can be are not perceivable at all.
Travis goes on to suggest that McDowell’s mistake is in restricting the sum of reason
giving relations to those of logical relations which merely hold between the general
ways that things can be (ibid. p. 141). One lacuna in Travis’s argument is that we are
given very little in the way of a positive story as to how these nonlogical reasongiving
relations function. We are hardly told anything as to how the particular unfolding of
reality we encounter in experience can provide reasons for our conceptual judgement
that the world is a certain general way. Travis simply informs us that part of what it is to
be a thinker is to have the expertise required to recognize that the surroundings one
experiences instance a particular generality, say that the sun is setting (ibid. p. 128).
In ‘Avoiding the Myth of the Given’, McDowell concedes some points to Travis. In
this paper McDowell attempts to insert a wedge between experiential content and
judgemental content, which to some degree accommodates Travis’s argument in ‘the
Silence of the Senses’. McDowell (2009, p. 267) claims that things are only presented as
so in judgements with propositional content. In contrast, experience presents us with
conceptualitystructured intuitional contents, from which we carve out the conceptual
2
As mentioned in the introduction, there is a worry as to whether Travis views the experience as the
reason or merely as the provider of the reason. I read Minimal Empiricism as requiring that the experience
figures as the reasons itself.
3
Notice that one may worry here whether Travis is guilty of misreading McDowell. Travis seems to
focus on what is given in experience as a reason. To McDowell, by contrast, it is the episode of having it
given, the experience itself, which plays the role of the reason.
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
18
Johan Gersel
contents which are then employed in the judgements that we choose to form upon its
basis. The crucial element in his response to Travis is that McDowell maintains that,
unlike propositional content, intuitional content brings the very mindindependent
objects of reality into view. Unlike judgements where we represent the world as being
such that these objects are in some specific way, in intuition, we are presented with
the ways things are (ibid. p. 268). One might question why intuitional content should
be said to be conceptual at all, given its significant difference from the content of
judgements. However, McDowell maintains this point by arguing that the very unity
present in intuitions is a function of the very same capacity that gives unity to our
thoughts: ‘The unity of intuitional content reflects an operation of the same unifying func
tion that is operative in the unity of judgement, in that case actively exercised’ (McDowell
2009, p. 264). McDowell’s second concession to Travis is the acknowledgement that
recognitional abilities may at times enable us to rationally and noninferentially form
a judgement whose content isn’t present in the experience upon which it is based
(ibid. p. 259). This seems to go some way towards acknowledging Travis’s (rather slim)
picture of how we recognize particular occurrences as falling under some general
way things can be. However, McDowell is adamant that we cannot rest content with
Travis’s picture of how “experience provides reasons for thought without falling
prey to the Myth of the Given (ibid. p. 269). While we may rationally form perceptual
judgements whose contents extend beyond the content of the experiences upon which
they are based, this extension cannot amount to a leap from experiencing something
entirely nonconceptual to the formation of a conceptual judgement. Thus, McDowell
maintains the charge that Travis’s theory is a form of the Myth of the Given.4
Travis’s ‘Unlocking the Outer World’ forms the last instalment of their debate prior
to this book. In this paper Travis challenges the idea that the function of the under
standing is to unify in any significant sense. According to Travis (2013c, p. 223), we
should follow Frege rather than Kant and question that unification is required for
experience and judgement. According to Frege judgements are the basic elements of
thought rather than something unified out of concepts. Instead, any talk of concepts is
by way of abstracting from or decomposing the inherently unified structure of thought
(ibid. 252). Likewise, no unification is required at the level of experience. If unification
was required for the objects of experience to come into view, then Travis questions that
those experienced objects could be mindindependent (ibid. p. 230). Hence, according
to Travis, when McDowell claims that the same unifying function is at play in both
judgement and experience, his argument in favour of the conceptual nature of intu
itions fails at two stages. He is wrong in assuming that unification is required for, or
even compatible with, acquaintance with mindindependent objects and, secondly,
wrong in assuming that unification forms a central element in judgement. McDowell’s
contribution to this anthology is a response to this critique. Travis’s contribution tries
4
For an extended discussion of this issue see Gersel, Jensen, and Thaning (2017).
OUP CORRECTED PROOF – FINAL, 05/10/2018, SPi
The Travis–M c Dowell Debate
19
once and for all to provide a principled argument as to why the requirements McDowell
imposes on experiential reasons cannot be an acceptable demand.
References
Gersel, J., Jensen, R.T., & Thaning, M.S. 2017. “McDowell’s New Conceptualism and the
Difference Between Chickens, Colours and Cardinals”. Philosophical Explorations 20 (1):
88–105.
McDowell, J. 1996. Mind and World. Cambridge, MA: Harvard University Press.
McDowell, J. 2009. “Avoiding the Myth of the Given”. In Having the World in View—Essays on
Kant, Hegel and Sellars, pp. 256–72. Cambridge, MA: Harvard University Press.
Travis, C. 2013a. “The Silence of the Senses”. In Perception—Essays after Frege, pp. 23–58.
Oxford: Oxford University Press.
Travis, C. 2013b. “Reason’s Reach”. In Perception—Essays after Frege, pp. 118–43. Oxford: Oxford
University Press.
Travis, C. 2013c. “Unlocking the Outer World”. In Perception—Essays after Frege, pp. 223–58.
Oxford: Oxford University Press.