C 2016 The Philosophical Forum, Inc.

IS INTUITION CENTRAL IN PHILOSOPHY?*

TINGHAO WANG

Some experimental philosophers criticize standard philosophical methodology on

the basis of survey data reporting variation of intuition according to irrelevant factors
like culture and order. I will refer to them as “experimentalists” and their critique as
the “experimental critique.” Recently, a few philosophers (e.g., Williamson, Deutsch,
and Cappelen) have responded by noting that the experimental critique relies on the
“Centrality” assumption—the thesis that intuition plays a central evidential role in
philosophical inquiry.1 They then deny the Centrality thesis and claim that, therefore,
intuition variation has no significant implications for philosophical methodology. In
this paper, I defend Centrality in response to two recent objections: the “argument
from non-neutrality” and the “argument from reasoning.” According to the argument
from non-neutrality, we should not believe the truth of Centrality because it is ill-
motivated by a particular dialectical standard of evidence. According to the argument
from reasoning, philosophical practice relies on argumentation rather than intuition
as its central evidence. As will be seen, both objections have different implications
for different versions of Centrality. Though they constitute some prima facie strong
reasons to deny some particular versions of Centrality, I shall argue, neither of them
successfully undermines the version of Centrality that experimentalists need. Along
the way, I will draw some parallels between intuition and perception.

UNDERSTANDING CENTRALITY

Centrality, roughly put, is the statement that intuition serves as philosophers’

central evidence. In this section, I am going to elaborate this definition by

*Thanks to Jennifer Nado, Max Deutsch, Darrell Rowbottom, and Chris Atkinson for helpful
comments on an earlier draft.
1
I borrow the term “Centrality” from Cappelen. See Herman Cappelen, Philosophy without Intuitions
(Oxford: Oxford UP, 2012) 3.

281
 TINGHAO WANG

looking at recent views concerning the nature of evidence. The philosophical

literature on evidence tends to cluster around the following three interrelated
questions: What sort of things does evidence consist of? Under what circumstan-
ces does a subject possess x as her evidence? And, under what circumstances is x
evidence for y?2 Different answers to these questions—that is, different accounts
of the ontology of evidence, evidence possession, and evidential relations—will
lead to different understandings of Centrality.
To start with, one major controversy among epistemologists concerns whether
evidence always consists of propositions. Some philosophers hold that only
propositions count as evidence;3 others adopt a broader view, according to which
evidence can at least sometimes be non-propositional. Philosophers of the latter
kind can read Centrality as the thesis that intuitions qua mental states are used as
evidence. Even if some of them might not think of intuitions as ultimate
evidence, they typically will agree that intuitions can at least be derivative
evidence.4 And the assumption of Centrality does not require that the evidential
role of intuition is ultimate or foundational.5
However, for proponents of propositional views of evidence, intuitions
themselves cannot work as evidence because they are mental states rather than
propositions. They might understand Centrality as the view that the propositional
contents of intuitions are used as evidence in philosophy—though, as I will show
later, this version of Centrality is not one which critics intend to reject.
Alternatively, a supporter of the propositional view of evidence might understand
Centrality as the thesis that philosophers’ central evidence consists of proposi-
tions concerning intuitions, such as the proposition that this intuition exists,
occurs, or the like, instead of intuitions themselves. To the extent that speakers

2
For more about different possible ways to answer these three questions, see Thomas Kelly,
“Evidence: Fundamental Concepts and the Phenomenal Conception,” Philosophy Compass 3.5
(2008): 933–55.
3
See, for example, Timothy Williamson, Knowledge and its Limits (Oxford: Oxford UP, 2002).
4
Conee and Feldman appear to adopt such a view about intuitive evidence. They claim that intuitive
judgments about thought experiments can “gain evidence from awareness of conceptual relations.”
They seem to suggest that one’s conscious experience of conceptual relations is ultimate evidence,
but one’s intuitive judgments can work as derivative evidence. See Conee and Feldman,
“Evidentialism,” Epistemology: New Essays, ed. Q. Smith (Oxford: Oxford UP, 2008) 93.
5
By contrast, Cappelen thinks that Centrality often has this requirement, because many metaphiloso-
phers claim that intuitions “provide evidence for other claims without themselves requiring
evidence.” See Herman Cappelen, Philosophy without Intuitions (Oxford: Oxford UP, 2012) 6–7.
However, though “thick” theorists of intuition do sometimes make this sort of claim, “thin” theorists
of intuition can accept that its evidential status depends on more fundamental evidence (e.g., percep-
tual experience). For the distinction between thick and thin views of intuitions, see Weinberg and
Alexander, “The Challenge of Sticking with Intuitions through Thick and Thin,” Intuitions, eds. A.
Booth and D. Rowbottom (Oxford: Oxford UP, 2014) 187–212.

282
 IS INTUITION CENTRAL IN PHILOSOPHY?

talk of intuitions as evidence, propositional theorists of evidence might take that

to be merely an abbreviation. According to them, though it is expedient and com-
mon in ordinary language to talk of footprints, DNAs, and mental states as evi-
dence, evidence actually consists of propositions, such as the proposition that
this footprint exists. Dougherty, for example, contends that speaking of experien-
ces as evidence is a kind of loose talk, which is “innocent enough unless we take
it to reveal bedrock truth.”6
Indeed, some critics of Centrality have formulated the thesis in ways which
apparently assume some propositional theory of evidence. Cappelen, for exam-
ple, thinks of Centrality as suggesting that “it is A has the intuition that p that
serves as evidence.”7 Also, according to Williamson, supporters of Centrality
take it that our evidence “consists of the psychological facts to the effect that we
have intuitions with those contents.”8 By contrast, I will take Centrality as true if
either intuitions themselves or propositions about intuitions are philosophers’
central evidence. In either case, we might say that “philosophers use intuitions as
central evidence” but only in a loose sense. Or, to put it slightly more rigidly, we
can say that Centrality is true if and only if intuitions play a central evidential
role in philosophical inquiry.
Here is a rough definition: x plays an evidential role for a subject S in cases
where either (i) x is the evidence that S uses or (ii) whether x obtains is always
somehow intimately related to the evidence that S uses.9 I stress that this is just a
sketch of a definition, rather than a full definition of evidential-role. In particular,
there are multiple ways to spell out the notion of a “somehow intimate” relation
between two things. It might be defined as, for instance, some sort of truth-
making relation or causal relation. So far as I can tell, nothing in this paper hangs
on the exact definition of evidential-role; the provisional definition is sufficient
for my purposes. This definition is compatible with both the propositional and
the non-propositional theory of evidence. If the non-propositional theory is cor-
rect, then intuitions can play an evidential role by themselves being the evidence
6
Trent Dougherty, “In Defense of Propositionalism about Evidence,” Evidentialism and its Discon-
tents, ed. T. Dougherty (Oxford: Oxford UP, 2011) 230.
7
Herman Cappelen, Philosophy without Intuitions (Oxford: Oxford UP, 2012) 13. It is unclear
whether this is really how Cappelen interprets Centrality, for in another place (14) he also says that
he is neutral on whether only propositions can be evidence.
8
Timothy Williamson, Philosophy of Philosophy (Oxford: Blackwell, 2007) 235. Some might want to
distinguish between facts and propositions. Williamson, however, explicitly uses “fact” as a syno-
nym for “true proposition.” See ibid: 209.
9
Earlenbaugh and Molyneux offer the definition that intuition plays an evidential-role if and only if
intuition is treated as evidence. My definition of evidential-role is broader than theirs, for, on my def-
inition, intuition has an evidential role to play in cases where what is treated as evidence is a proposi-
tion describing an intuition instead of the intuition itself. See Joshua Earlenbaugh and Bernard
Molyneux, “Intuitions are Inclinations to Believe,” Philosophical Studies 145.1 (2009): 91–2.

283
 TINGHAO WANG

that philosophers use. But if the propositional view is correct, then intuitions can
play a key evidential role as long as philosophers use propositions describing
intuitions as their central evidence, since there is always an intimate relation
between intuitions and propositions describing those intuitions. The definition
also excludes the versions of Centrality that critics do not intend to reject, such
as the “content” reading: “many philosophical arguments treat the contents of
certain intuitions as evidence.”10 On this reading, the absence of an intuition is
not always relevant to philosophical evidence, since philosophers might use the
propositional content of an intuition as evidence due to reasons having nothing to
do with the intuition itself (e.g., theoretical arguments).
It is worth noting that Cappelen also considers an alternative way to understand
Centrality: “p is the evidence and the source of that evidence is that A has an intu-
ition that p.”11 This interpretation is added to include a view like Bealer’s,12
according to which intuition itself is not evidence but a source of evidence; it is
rather the propositional content of intuition that counts as evidence. I will not focus
on this interpretation, mainly because I find the notion of “source of evidence”
rather obscure. Note that, for supporters of Centrality, it is not enough for intuition
to be a merely causal source of philosophical evidence. If any distinction between
the context of discovery and the context of justification can be made in philosophi-
cal inquiry, then a causal source of evidence might not be granted any important
epistemological status.13 However, it is far from obvious how to interpret such a
notion of “source of evidence” as not merely the causal source, but neither Bealer
nor Cappelen analyses the notion in further detail.
Another related debate in epistemology concerns the nature of evidence pos-
session. For supporters of propositional theories of evidence, it is natural to think
that, in order for e to be one’s evidence, one has to believe the truth of e. But this
might not be sufficient for evidence possession; Williamson, for example, sets a
stricter requirement that e has to be part of one’s total knowledge. Williamson
develops the E 5 K theory of evidence, which “equates S’s evidence with S’s
knowledge, for every individual or community S in any possible situation.”14
10
Max Deutsch, The Myth of the Intuitive (Cambridge: MIT Press, 2015) 36.
11
Herman Cappelen, Philosophy without Intuitions (Oxford: Oxford UP, 2012) 13.
12
George Bealer, “Intuition and the Autonomy of Philosophy,” Rethinking Intuition: The Psychology
of Intuition and its Role in Philosophical Inquiry, eds. M. DePaul and W. Ramsey (Lanham, MD:
Rowman & Littlefield, 1998) 205.
13
Both Deutsch and Cappelen make a similar point. See Max Deutsch, “Intuitions, Counter-examples,
and Experimental Philosophy.” Review of Philosophy and Psychology 1.3 (2010): 447–60; Herman
Cappelen, Philosophy without Intuitions (Oxford: Oxford UP, 2012) 230. I stress that a causal
source of evidence might, and probably often does, play some important evidential role or other
non-evidential epistemological role in philosophical practice. The point made here is only that the
causal source is not always important when it comes to debates about methodology.
14
Timothy Williamson, Knowledge and its Limits (Oxford: Oxford UP, 2002) 185.

284
 IS INTUITION CENTRAL IN PHILOSOPHY?

Assuming that evidence is what justifies belief, the E 5 K theory entails that
“knowledge, and only knowledge, justifies belief.”15 Conversely, for non-
propositional theories, mental states like experiences and intuitions count as
one’s evidence only if one has those mental states; but again, merely having a
mental state might not be enough for possessing it as evidence, and therefore
additional conditions (e.g., being relatively easy to bring to awareness) might be
added. Centrality, as the claim that intuition serves as philosophers’ central evi-
dence, seems to presume that philosophers possess intuitions (or propositions
about intuitions) as genuine evidence, but it does not. Note that, if a subject pos-
sesses x as her evidence, then x has legitimate evidential status for her. Yet, a
subject can use x as evidence even if x does not have legitimate evidential status
for her. For instance, if Williamson’s E 5 K theory is correct, then one cannot
possess the propositional contents of any false beliefs as one’s evidence; they do
not have legitimate evidential status because the subject does not know them.
However, plausibly, one can still use them as evidence. Centrality only holds that
philosophers use intuitions as central evidence; whether they genuinely possess
intuitive evidence is a further question.
That being said, evidence possession might have a bearing on the question of
what it is for one to use x as evidence. Given the E 5 K theory, for example, it is
natural to think that to use x as evidence is just to treat x as if it were known. This
implies that one can only use the contents of one’s beliefs as evidence, for, plausi-
bly, if one does not believe x then one is not treating it as a piece of knowledge.16
By contrast, if one’s evidence is limited to a special kind of mental states, then one
might hold that only one’s mental states can be used as evidence and that to use a
mental state as evidence for one’s belief is just to “base” one’s belief on that men-
tal state.17 In any case, my definition of Centrality will be neutral on the issue of
what conditions one needs to satisfy to use something as evidence.18
Finally, Centrality implies that intuition plays strong evidential roles in the eval-
uation of philosophical theories, but differing views on evidential relation will dis-
agree as to what exactly this means. Some have adopted Bayesian approaches,
15
Ibid: 185.
16
Williamson thinks the reverse is also true: if one does not treat x as knowledge then one does not
believe x. He thereby adopts the view that “to believe p is to treat p as if one knew p.” See ibid: 46.
17
For a variety of approaches to characterize this “basing” relation, see Ram Neta, “The Basing
Relation,” The Routledge Companion to Epistemology, eds. S. Bernecker and D. Pritchard (New
York: Routledge, 2011) 109–18.
18
It is important to distinguish between using x as evidence and thinking of x as evidence. Though the
former notion is restricted by theories of evidential possession, the latter notion is not. For example,
even if the E 5 K theory is correct, one can still think of a mental state as evidence. There might be
a sense of “use” according to which Centrality is true as long as philosophers think of intuitions as
central evidence. But this will not be the relevant version of Centrality here; this paper considers
whether intuitions are used as evidence, no matter whether we are thinking of them as such.

285
 TINGHAO WANG

according to which evidential relations can be put in probabilistic terms; by con-

trast, some insist that the evidential relation is best understood as an abductive rela-
tion. I will stay neutral on this issue, but one point frequently made in the Bayesian
context is worth highlighting, namely the distinction between the balance and the
weight of evidence. There are two senses in which a piece of evidence x is “strong”
evidence for a philosophical theory t. One can say that the balance of the evidence
is strong, in cases where x makes t a highly probable theory. Alternatively, one can
say that x has strong evidential weight, in cases where it provides a substantial size
of evidential data.19 We might therefore ask whether Centrality is the claim that
intuitions (or propositions about intuitions) are given strong balance or weight. In
my view, Centrality is false if either intuition is treated as having rather weak bal-
ance or rather weak weight relative to other kinds of philosophical evidence. Intu-
ition’s importance in philosophical practice will be undermined, if either it only
very slightly raises a philosophical theory’s probability, or it is thought of as pro-
viding a rather small size of relevant information.20

THE ARGUMENT FROM NON-NEUTRALITY

Williamson argues against Centrality by claiming that it is ill-motivated. Accord-

ing to Williamson, the idea that philosophical evidence consists of intuitions is
driven by the principle of “Evidence Neutrality,” which he defines as follows:
Whether a proposition constitutes evidence is in principle uncontentiously decidable, in the sense
that a community of inquirers can always in principle achieve common knowledge as to whether
any given proposition constitutes evidence for the inquiry. . . in a debate over a hypothesis h,
proponents and opponents of h should be able to agree whether some claim p constitutes evidence
without first having to settle their differences over h itself.21

Note that this is not the claim that evidence is always known by participants of
the debate; it is rather the claim that evidence is always in principle known,
19
For more about the balance/weight distinction, see James M. Joyce, “How Probabilities Reflect
Evidence,” Philosophical Perspectives 19.1 (2005): 153–78.
20
The balance/weight distinction is seldom made in debates about intuitive evidence, probably
because a traditional non-Bayesian framework of epistemology is often assumed. An exception is
Weatherson, who defends Centrality by claiming that intuition provides strong but rather fragile
evidence. See Brian Weatherson, “Centrality and Marginalisation.” Philosophical Studies 171.3
(2014): 517–33. However, since Weatherson takes intuitive evidence to have rather weak evidential
weight, his position does not vindicate Centrality, as I use the term.
21
Timothy Williamson, Philosophy of Philosophy (Oxford: Blackwell, 2007) 210. He assumes in this
paragraph that some propositional theory of evidence is correct. In the current section, I will follow
Williamson in this assumption. However, the assumption is essential neither to Williamson’s argu-
ment nor to my points made in this section; similar points can be made if one endorses a non-
propositional theory of evidence.

286
 IS INTUITION CENTRAL IN PHILOSOPHY?

which means that inquirers can achieve common knowledge if they overcome all
the “accidental mistakes and confusions.”22 This suggests that evidence plays the
role of neutral arbiter between rival theories: one cannot argue from a piece of
evidence which presupposes the falsity of the opponent’s position. Take the
debate between supporters of the descriptivist theory of reference and supporters
of the causal theory as an example. According to Evidence Neutrality, one cannot
reject the descriptivist theory by putting forward evidence which presupposes the
truth of the causal theory, for in that case the descriptivist will never, even in
principle, accept that piece of evidence.
To illustrate why Williamson thinks Evidence Neutrality might lead to Central-
ity, consider the G€odel case, which is frequently cited as a thought experiment
against the descriptivist theory of reference. Suppose that Schmidt rather than
G€odel actually proved the incompleteness of arithmetic. Also suppose that we only
associate one description “the prover of incompleteness” with the name “G€odel.”
What evidence does this case provide? Consider the following two propositions:
(G1) “G€odel” does not refer to Schmidt.
(G0) It is intuitive that “G€
odel” does not refer to Schmidt.

According to Evidence Neutrality, one cannot use (G1) as evidence against

descriptivism. The descriptivist theory entails that “G€odel” refers to Schmidt,
and therefore the descriptivist will never, even in principle, accept that (G1) is
true. By contrast, (G0) seems to be a better candidate; even the descriptivist can
agree that (G1) is intuitive, but insists that this intuition does not falsify descripti-
vism. The above sort of consideration, according to Williamson, “tempts one to
retreat into identifying evidence with uncontentious propositions about psycho-
logical states.”23 To satisfy Evidence Neutrality, one needs to start with the
premises with which even one’s opponents might agree. And it seems that the
psychological premises describing people’s intuitions best satisfy this condition.
Williamson denies the principle of Evidence Neutrality by pointing out that
even psychological premises do not meet its requirement. One might face an
opponent who is committed to saying that we cannot even have the intuition that
“G€odel” does not refer to Schmidt. Some radical eliminativists, for instance, will
not accept that there are intuitions at all. On their view, attributions of folk psy-
chological mental states like intuition are false due to general theoretical prob-
lems with folk psychology. (G0) does not comply with Evidence Neutrality,
because these eliminativists will never, even in principle, accept that (G0) is true.
Williamson infers that Evidence Neutrality is false, for the notion of Evidence as

22
Ibid: 210.
23
Ibid: 211.

287
 TINGHAO WANG

a neutral arbiter between participants lays us “open to exploitation by ruthless

opponents.”24 We should not limit our evidence simply because the opponent
insists “an impoverished skeptical starting-point”; rather, we sometimes must
“abandon skeptics to their fate.”25 There is thus no need to retreat into psycho-
logical premises like (G0). Williamson concludes that “our evidence in philoso-
phy consists of facts, most of them non-psychological, to which we have
appropriate epistemic access.”26 I will refer to Williamson’s above argument
against Centrality as the “argument from non-neutrality.”
Even if Centrality is not well-motivated, however, this does not mean that it is
false. Both Alexander and Brown have responded to Williamson by providing rea-
sons to accept Centrality without appealing to Evidence Neutrality.27 Also, Wein-
berg argues that the experimentalists’ challenge does not need to assume Evidence
Neutrality in any event.28 Though I have some worries about their arguments, I am
inclined to endorse the general idea: the principle of Evidence Neutrality is not
essential to the experimental critique. My reason is this: the version of Centrality
that Evidence Neutrality initially seems to give rise to is indeed much stronger
than what experimentalists need. As Williamson puts it, Evidence Neutrality
“exerts general pressure to psychologize evidence”;29 since non-psychological
premises never pass the Evidence Neutrality test, supporters of Evidence Neutral-
ity will contend that only psychological premises play important evidential roles in
philosophy. Thus, they will conclude that only intuitions (plus maybe some other
psychological states) serve as philosophers’ central evidence. But supporters of
Centrality do not need to stick to such a strong conclusion. Instead, they can accept
that both psychological premises and non-psychological premises are granted cen-
tral evidential status. Thinking of intuitions as constituting central philosophical
evidence does not need to exclude other sorts of evidence as equally central.
This weak version of Centrality is enough for most experimentalists’ purpose.
Experimentalists usually aim to argue that the standard philosophical methodol-
ogy should be significantly revised, and they can achieve this purpose as long as
one of the central tools in philosophy is shown to be in trouble. They can accept
a skeptical attitude towards intuitions, but not towards other kinds of evidence in
philosophical practice. Admittedly, experimentalists occasionally sound as if
24
Ibid: 238.
25
Ibid: 238–39.
26
Ibid: 241.
27
See Joshua Alexander, “Is Experimental Philosophy Philosophically Significant?” Philosophical
Psychology 23.3 (2010): 377–89; Jessica Brown, “Thought Experiments, Intuitions and Philosophi-
cal Evidence,” Dialectica 65.4 (2011): 493–516.
28
See Jonathan Weinberg, “On Doing Better, Experimental-Style,” Philosophical Studies 145.3
(2009): 455–64.
29
Timothy Williamson, Philosophy of Philosophy (Oxford: Blackwell, 2007) 211.

288
 IS INTUITION CENTRAL IN PHILOSOPHY?

they think that, if intuition does not qualify as evidence, then the whole discipline
of philosophy is undermined. However, this sort of radical claim is not essential
to their critique, and surely their experimental data can have important methodo-
logical implications without being able to dismiss the discipline as a whole.
Indeed, Williamson’s two other related worries with Centrality are also only
effective against the strong version of Centrality. First, he argues that Centrality
can be “self-defeating.” Williamson seems to think that, ceteris paribus, support-
ers of Centrality prefer those explanations of intuitions on which the intuitions
are true to those explanations on which the intuitions are false.30 He then ques-
tions what justifies this preference. Even if it is justified by an intuition, it is
unclear why this particular intuition has such a special privilege that we adopt “a
methodology that assumes its truth.”31 But this problem is merely true of the
strong version of Centrality, according to which only intuitions work as philo-
sophical evidence. Supporters of the weak version of Centrality, by contrast, can
claim that the above preference is justified by non-psychological considerations
rather than intuitions. One might suggest that, for instance, intuitions are more
likely to be true than false as a result of evolutionary pressure. It is also worth
noting that Williamson’s self-defeating problem is nothing unique to philosophy;
empirical sciences face a similar problem of why, ceteris paribus, we prefer the
explanations on which the observational data come out true. This problem, how-
ever, obviously does not constitute any good reason against what we might refer
to as “P-Centrality”—the thesis that perception plays central evidential roles in
the empirical sciences. It is thus unclear why a similar objection in the case of
intuition should be considered as a serious problem with Centrality.
This leads to another point made by Williamson, which is precisely based on
an analogy between philosophy and the sciences. He puts it as follows:
If Evidence Neutrality psychologizes evidence in philosophy, it psychologizes it in the natural
sciences too. But it is fanciful to regard evidence in the natural sciences as consisting of
psychological facts rather than, for example, facts about the results of experiments and
measurements. . . The psychologization of evidence by Evidence Neutrality should be resisted in
the natural sciences; it should be resisted in philosophy too.32

Williamson might be right that the tendency of generally psychologizing evi-

dence should be resisted both in the natural sciences and in philosophy. However,
it is absurd to deny that the psychological occupies at least one of the central
kinds of evidential roles in scientific practice. It remains highly plausible that
30
Note that this argument is not intended against experimentalists, for they do not have such a prefer-
ence. Instead, the targets of this argument are those philosophers who support Centrality and also
think that intuition has legitimate evidential status in philosophy.
31
Ibid: 236.
32
Ibid: 212.

289
 TINGHAO WANG

perceptions or observations play key evidential roles, and experimental data

reporting extensive observational bias are definitely relevant to the methodology
of the sciences. If we take the analogy between philosophy and the sciences seri-
ously, then we have good reason to believe that intuitions are granted central
(though probably not the only central) evidential status in philosophy and that
the experimentalists’ results are relevant to philosophical methodology.
A final issue is worth addressing. Though I have been describing the argument
from non-neutrality as an argument against Centrality, it is not totally clear whether
this is Williamson’s intention. According to an alternative interpretation, when Wil-
liamson asserts that philosophical evidence does not consist of intuition, he does not
mean to deny Centrality; instead, he means that philosophers shouldn’t use intu-
itions as evidence.33 I maintain that Williamson’s target is Centrality, for he explic-
itly claims that his “rethinking of philosophical methodology” concerns “how
philosophy is actually done.”34 But there is still a crucial difference between Wil-
liamson and other critics of Centrality like Deutsch and Cappelen, both of whom are
inclined to adopt the extreme view that intuition never plays any evidential roles in
philosophy. Cappelen, for example, aims to argue that “it is not true that philoso-
phers rely extensively (or even a little bit) on intuitions as evidence.”35 Also,
Deutsch claims that “philosophical arguments never appeal to the intuitiveness of a
judgment about a case to justify belief in that judgment.”36 In contrast, Williamson’s
position seems to be more moderate. He admits that unawareness of failures of Evi-
dence Neutrality “does more than distort philosophers’ descriptions of philosophy”
and “alters their first-order philosophizing.”37 This suggests that intuition has some
non-central evidential role in philosophy, mainly because beliefs about Evidence
Neutrality have changed part of philosophical practice. In any event, the current sec-
tion discusses whether one could reasonably reject Centrality based on the argu-
ments Williamson offers, whatever Williamson’s intended target is. Similarly, this
section examines the question of whether the argument from non-neutrality under-
mines the experimental critique, no matter whether Williamson himself intends to
reject the experimentalists’ project on the basis of this argument.38

33
Cappelen, for instance, assumes such an interpretation of Williamson’s view. See Herman Cappe-
len, Philosophy without Intuitions (Oxford: Oxford UP, 2012) 204.
34
Timothy Williamson, Philosophy of Philosophy (Oxford: Blackwell, 2007) 6.
35
Herman Cappelen, Philosophy without Intuitions (Oxford: Oxford UP, 2012) 1.
36
Max Deutsch, The Myth of the Intuitive (Cambridge: MIT Press, 2015) 76–7.
37
Timothy Williamson, Philosophy of Philosophy (Oxford: Blackwell, 2007) 213.
38
Williamson does not explicitly use the argument from non-neutrality to criticize experimentalists.
Instead, he has two other worries about the experimental critique: (i) It leads to an unsustainable
form of “judgment skepticism” and (ii) it relies on data reporting variation of intuition among lay
people rather than trained philosophers. However, one might easily understand Williamson as
implicitly criticizing experimentalists from the denial of Evidence Centrality.

290
 IS INTUITION CENTRAL IN PHILOSOPHY?

To conclude, I share with Williamson the concern that experimentalists as

well as their opponents have sometimes exaggerated the importance of intuition
in philosophy. Philosophers, just as empirical scientists, can and have appealed
to a wide range of non-psychological evidence. However, I think Williamson
goes too far when he says that most of the evidence used in philosophy is non-
psychological in its nature. Though we do not need to always start from intuitive
evidence in philosophical theorizing, it remains plausible that we do start from
intuitive evidence in some central types of cases. But none of the arguments con-
sidered in this section succeeds in refuting that weaker claim.

THE ARGUMENT FROM REASONING

Another line of attack on Centrality involves the claim that philosophers use
reasoning rather than intuition in the evaluation of theories and hypotheses. Both
Deutsch and Cappelen have appealed to this approach in refuting Centrality.39 In
this section, I will mainly focus on Deutsch’s view. Take the G€odel case as an
example again. Remember that Williamson seems to think that evidence offered
in this case is the thought-experimental judgment (G1): “G€odel” does not refer to
Schmidt. Further, he seems to suggest that philosophers do not need to use any
further evidence to back up the premise (G1). Deutsch, by contrast, suggests that
philosophers need and have actually used further evidence—namely, argu-
ments—to support thought-experimental judgments like (G1).40
More specifically, he contends that Kripke presents the following arguments
for (G1) in the original presentation of the G€odel case. The first argument is
based on analogy with several real-life examples. “Einstein,” for instance, does
not refer to the inventor of the atomic bomb, despite the facts that some speakers
associate with “Einstein” only one description “the inventor of the atomic bomb”
and that Einstein did not invent the atomic bomb. The second is the “immunity to
error” argument. If descriptivism is correct, then it follows that people can never
make mistakes when they assert sentences like “G€odel is the prover of incom-
pleteness.” Since this consequence is false, there is good reason to believe that
descriptivism is false too; however, descriptivism is “the only reason to make the
opposing judgment [that G€odel refers to Schmidt].”41 The immunity to error
39
See Max Deutsch, “Experimental Philosophy and the Theory of Reference,” Mind & Language
24.4 (2009): 445–66; Max Deutsch, “Intuitions, Counter-Examples, and Experimental Philosophy.”
Review of Philosophy and Psychology 1.3 (2010): 447–60; Max Deutsch, The Myth of the Intuitive
(Cambridge: MIT Press, 2015); Herman Cappelen, Philosophy without Intuitions (Oxford: Oxford
UP, 2012).
40
Deutsch agrees with Williamson that (G1) is our evidence, but insists that the evidential status of
(G1) is derived from the arguments provided.
41
Max Deutsch, The Myth of the Intuitive (Cambridge: MIT Press, 2015) 110.

291
 TINGHAO WANG

argument thereby constitutes an indirect argument for (G1). For similar reasons,
Kripke’s objections to the descriptivist theory of meaning and positive arguments
for his own causal-historical theory (or “picture”) of reference also count as indi-
rect arguments for the judgment that G€odel does not refer to Schmidt.
According to Deutsch, none of the arguments above starts from premises stat-
ing people’s intuitions; the premises concern facts about reference rather than
intuitions about those facts. Both Deutsch and Cappelen have defended similar
conclusions in a series of important philosophical thought experiments, such as
Gettier cases, Lehrer’s Truetemp case, and the Trolley case. They both conclude
that Centrality is a misconception of philosophers’ practice and that intuition sel-
dom, if ever, plays significant evidential roles in philosophy. Call the above argu-
ment against Centrality the “argument from reasoning.”
Some philosophers have responded by claiming that, contra Deutsch and Cap-
pelen, arguments for the relevant thought-experimental judgments do depend on
intuitive evidence.42 Even supposing Deutsch is right that our intuition about
(G1) does not serve as evidence, it is sometimes suggested, some other intuitions
do and must play evidential roles at some stage of the argumentative chain for
(G1). Deutsch refers to this as the “relocation problem.”43 In what follows, I will
raise a different objection to Deutsch. In contrast to the relocation problem, my
objection does not assume the strong claim that intuition must serve as evidence
in philosophical arguments; rather, it is merely intended to show that Deutsch
does not provide any good argument against Centrality. I will focus on the fol-
lowing point, which is central to Deutsch’s argument:
(R) Philosophical arguments seldom start from premises stating people’s intuitions.

Here is my strategy. Instead of arguing against (R), I will claim that there are
indeed several at least prima facie strong arguments for it. But I will further con-
tend that we should not respond to those arguments by rejecting Centrality. The
reason is as follows. There are similar arguments for a parallel claim (P-R) in the
case of perception. Depending on what theory of evidence one has, one might
respond differently to these arguments. However, almost no philosopher in the
literature of perception responds by rejecting P-Centrality (the view that percep-
tion plays a central evidential role in empirical sciences). It thus seems unclear
why similar arguments for (R) should motivate us to discard Centrality.
There are at least three apparently strong arguments for (R). First, philosophers
usually do not mention “intuition” and its cognate terms in their writings. Both
42
See, e.g., Berit Brogaard, “Intuitions as Intellectual Seemings,” Analytic Philosophy 55.4 (2014):
382–93; Jonathan Ichikawa, “Review of Philosophy without Intuitions.” International Journal of
Philosophical Studies 21.1 (2013): 111–16.
43
Max Deutsch, The Myth of the Intuitive (Cambridge: MIT Press, 2015) 58.

292
 IS INTUITION CENTRAL IN PHILOSOPHY?

Cappelen and Deutsch have emphasized this point; they examine the original
texts surrounding famous thought experiments, and observe that philosophers
very occasionally use “intuition”-language. This is actually the main reason why
Deutsch thinks (R) is correct. Admittedly, sometimes one does not mention the
premises in one’s argument, such as in the case of hidden premises. However, if
philosophers often start their arguments from statements concerning intuitions,
then it is unlikely that “intuition”-talk would happen so rarely in philosophical
practice. Second, one might worry how one could infer from a premise describ-
ing intuition to a conclusion concerning non-psychological philosophical subject
matter. Williamson raises such a problem, claiming that there is a gap between
psychological premises like (G0) and non-psychological conclusions like (G1)
and the gap “is not easily bridged.”44 Third, philosophers often do not even con-
sider their own intuitions at the time of philosophical writing. When one writes
about the philosophy of reference, one considers what names refer to and what
the general nature of reference is; but one seldom considers one’s intuitions
about the nature of reference. This seems to suggest that they rarely hold beliefs
about their own intuitions. One might not believe one’s premises in some cases
of hidden premises, but in most cases one does hold beliefs about the premises
one uses. There is thus a prima facie good reason to think that philosophers do
not use premises about intuition in their arguments at all.
No matter how strong the above arguments are for (R), it is important to note
that one can make analogous arguments for the following judgment in the case of
perception:
(P-R) Arguments in the empirical sciences seldom start from premises stating people’s perceptual
experiences.

To start with, one can defend (P-R) by claiming that scientists seldom use
“perception”-language in their academic writings. For example, Williamson takes
it that “when scientists state their evidence in their publications, they state mainly
non-psychological facts.”45 One can thus argue that, if scientists usually start from
premises about experiences, then it is inexplicable that “perception”-talk happens
so rarely in scientific publications. Moreover, one can complain that there is a gap
between the psychological and the non-psychological in the case of perception too.
Brown points out that, on some accounts of perceptual evidence, there is a problem
of how one can infer from the proposition that “one is having an experience as of
p” to the conclusion that “p is the case.”46 She then responds to Williamson’s gap
44
Timothy Williamson, Philosophy of Philosophy (Oxford: Blackwell, 2007) 211.
45
Ibid: 212.
46
Jessica Brown, “Thought Experiments, Intuitions and Philosophical Evidence,” Dialectica 65.4
(2011): 506.

293
 TINGHAO WANG

objection by arguing that an externalist approach to bridging the gap in the case of
perception could be well applied to the case of intuition too. Finally, one can also
contend that (P-R) is true because empirical scientists plausibly often do not even
consider their own perceptual experiences. This suggests that scientists seldom
hold beliefs about perception, and thus their argumentation seldom begins with the
contents of those beliefs. These reasons support (P-R) in ways similar to how the
parallel reasons support (R) in the case of intuition.
What would philosophers of perception regard as the consequences of these
prima facie strong arguments for (P-R)? I take it that few philosophers will think
that, because those arguments support (P-R), they also constitute good reasons
for the view that perceptual experiences play no central evidential roles in scien-
tific practices. In what follows, I will present a dilemma: no matter whether a phi-
losopher accepts a propositional or a non-propositional theory of evidence, they
do not reject Centrality on the basis of the above reasons for (P-R). Take the last
argument for (P-R) as an example, which has been discussed frequently in recent
debates about the nature of perceptual evidence. For instance, Kelly writes that,
. . . some philosophers maintain that in typical cases of perception, one does not form beliefs about
how things appear to one, or about how one’s perceptual experience presents things as being:
rather, in response to one’s experiences, one simply forms beliefs about the external world itself.47

In a similar vein, Pollock and Cruz argue that “the beliefs we form are almost
invariably beliefs about the objective properties of physical objects—not about
how things appear to us.”48 According to them, we do not possess beliefs about
our own perceptual experiences in standard cases. One might thus allege that we
rarely argue from premises stating perceptual experiences either, for we typically
hold beliefs about our premises. This potential argument for (P-R) has exactly
the same structure as the last of the above arguments for (R).
The crucial point is that Neither Kelly nor Pollock and Cruz argue that,
because we seldom have beliefs about perception, perception has no evidential
status. Instead, they both infer that perception has an evidential role to play—it is
just that propositional theories of evidence fail to account for that role. They
claim that the propositional view is overly demanding and hyper-intellectual; for,
to characterize the evidential status of perception, the propositional theory
requires one to form a higher-order belief regarding one’s own perceptual experi-
ence. Indeed, (P-R) contradicts P-Centrality only if one assumes some proposi-
tional theory of evidence. But if the non-propositional view of evidence is
47
Thomas Kelly, “Evidence,” The Stanford Encyclopedia of Philosophy (Fall 2014 Edition), ed. E. N.
Zalta. The Metaphysics Research Lab, 21 Sept. 2014. Web. 1 June 2016. URL 5 <http://plato.stan-
ford.edu/archives/fall2014/entries/evidence/>.
48
John L. Pollock and Joseph Cruz, Contemporary Theories of Knowledge (Lanham, MD: Rowman &
Littlefield, 1999) 61.

294
 IS INTUITION CENTRAL IN PHILOSOPHY?

correct, then (P-R) is in fact compatible with P-Centrality. For, according to the
non-propositional view, a perceptual experience can be used as evidence for a
premise without itself being presented in the premise. Take Conee and Feldman’s
evidentialism, which is an influential non-propositional theory of evidence, as an
example. They argue that evidence includes “one’s private experiences” and that
“such evidence could not be put into an argument in any useful manner.”49 On
such a view, we can accept that (P-R) is true—arguments in sciences seldom start
with propositions describing perceptions—but insist that perception plays its evi-
dential role in some other way.
There seems to be nothing to prevent us from giving the same sort of reply in
the case of intuition. We can agree with Deutsch that (R) is true: philosophical
arguments rarely have their starting points as propositions presenting intuitions.
However, that is not the only way in which intuitions can play a central eviden-
tial role. If the non-propositional view of evidence is true, then intuitions are
used as evidence for the premises in the arguments, but they themselves do not
appear in those premises. Centrality thus remains cogent despite Deutsch’s argu-
ments for (R), assuming that evidence can be non-propositional in its nature.
Further, in my view, even if evidence must be propositional, the argument
from reasoning still does not give any good reasons for rejecting Centrality. Con-
sider the case of perception again. As I said above, some philosophers cast doubt
on the propositional theory of evidence by arguments for (P-R). But importantly,
even those who support the propositional theory do not reply by denying
P-Centrality; instead, they choose to reject (P-R) by undermining the apparently
strong arguments for it. For instance, Williamson, a proponent of the propositio-
nal theory, claims that we usually believe “the proposition that things appear to
be that way”50 and that propositions like this describe our perceptual experiences.
He admits the fact that in typical cases we do not consider such proposition; how-
ever, according to Williamson, we still often have beliefs and knowledge about
these propositions, because “one knows many propositions without considering
them.”51 One does not need consideration to believe or even know a proposition,
because “knowing is a state, not an activity.”52 In other words, while critics of
the propositional theory of evidence condemn it as being hyper-intellectual, Wil-
liamson maintains that beliefs about one’s own mental states are not as hard to
achieve as the critics suppose it to be. On his view, to achieve such beliefs, the
subject does not even need to grasp the notion of perception. One only needs to

49
Earl Conee and Richard Feldman, Evidentialism: Essays in Epistemology: Essays in Epistemology
(Oxford: Clarendon Press, 2004.) 2–3.
50
Timothy Williamson, Knowledge and its Limits (Oxford: Oxford UP, 2002) 198.
51
Ibid: 199.
52
Ibid: 199.

295
 TINGHAO WANG

grasp the notion of appearing by having “some inkling of the distinction between
appearance and reality.”53
Again, there seems to be nothing to stop us from making a similar reply in the
case of intuition. We can respond to Deutsch by stating that, despite the fact that
philosophers usually do not consider propositions about intuitions, they can still
possess beliefs about those propositions, because knowledge is a state but not an
activity. We might also add that, for Centrality to be true, philosophers do not
even need to grasp the notion of intuition. They only need to grasp the notion of
appearing, since propositions like “things appear to be that way” standardly
describe one’s intuitions in philosophical contexts. And if they believe such
propositions, then it remains plausible that such propositions are hidden premises
in philosophical argumentation.
The point is basically this. Though there are some prima facie strong argu-
ments for (P-R), virtually no one in the literature of perceptual evidence denies
P-Centrality on the basis of these arguments, no matter whether one supports the
propositional or the non-propositional theory of evidence. Quite the opposite, it
is a crucial aim for any plausible theory of evidence to capture perception’s evi-
dential role. As shown by Deutsch and Cappelen, there are similar prima facie
strong reasons for accepting (R). If one reacts in ways similar to how philoso-
phers react in the case of perception, then one should claim that the argument
from reasoning does not undermine Centrality, whatever account of evidence is
correct. It is rather the other way around: any plausible theory of evidence had
better be able to take account of intuition’s potential evidential role. Since critics
of Centrality do not give any reason why one should react differently in the case
of perception and the case of intuition, it remains convincing that intuition plays
a central evidential role in philosophical practice.

Lingnan University

53
Ibid: 199.

296