Philos Stud (2013) 163:271–290

DOI 10.1007/s11098-011-9815-1

Rational intuition and understanding

Peter J. Markie

Published online: 21 September 2011

 Springer Science+Business Media B.V. 2011

Abstract Rational intuitions involve a particular form of understanding that gives

them a special epistemic status. This form of understanding and its epistemic effi-
cacy are not explained by several current theories of rational intuition, including
Phenomenal Conservatism (Huemer, Skepticism and the veil of perception, 2001;
Ethical intuitionism, 2005; Philos Phenomenol Res 74:30–55, 2007), Proper
Functionalism (Plantinga, Warrant and proper function, 1993), the Competency
Theory (Bealer Pac Philos Q 81:1–30, 2000; Sosa, A virtue epistemology, 2007) and
the Direct Awareness View (Conee, Philos Phenomenol Res 4:847–857, 1998;
Bonjour, In defense of pure reason, 1998). Some overlook it; others try to account
for it but fail. We can account for the role of understanding in rational intuition by
returning to the view of some of the early Rationalists, e.g. Descartes and Leibniz.
While that view carries a prohibitive cost, it does contain an insight that may help us
solve the problem of giving understanding its due.

Keywords Rational intuition  Understanding  Phenomenal conservatism 

Proper function  Direct awareness  Knowledge  Justification

We sometimes just intellectually ‘‘see’’ a proposition’s truth. This is rational

intuition, and, in the right circumstances, it is a source of justified belief and
knowledge. What does it involve and how does it support such normative epistemic
states? I shall argue that rational intuition involves a particular form of
understanding that is not explained by several current theories of its nature and

An early ancestor of this paper was preented to the Central States Philosophical Association.

P. J. Markie (&)
University of Missouri, Columbia, MO, USA
e-mail: markiep@missouri.edu

123
272 P. J. Markie

epistemic efficacy, including Phenomenal Conservatism (Huemer 2001, 2005,

2007), Proper Functionalism (Plantinga 1993), the Competency Theory (Bealer
2000; Sosa 2007) and the Direct Awareness View (Conee 1998; Bonjour 1998). I’ll
end by considering how the early Rationalists, e.g. Descartes and Leibniz,
accounted for the role of understanding in rational intuition. While their approach
carries a significant cost, I think it contains an insight that just might enable us to
give understanding its due.1

1 Understanding and epistemic superiority

Rational intuition requires understanding—we can’t rationally intuit what we don’t

understand—but mere understanding isn’t enough. My friend Elias has a fuller
understanding of fractions than I and can rationally intuit a truth I cannot:
(P) 237/148 is greater than 425/266.2
Nonetheless, we both understand (P), as we both believe it. He believes (P) on the
basis of his rational intuition. I believe it on the basis of his testimony. (P) just
seems true, indeed necessarily true, to him, and he feels a strong inclination to
believe it. (P) neither seems true nor false to me, and my understanding of it gives
me no inclination to believe it. We might then say that to intuit a proposition
rationally is to understand it so that it seems necessarily true and we have a felt
inclination to believe it. We might say this, but we would be mistaken. The
understanding involved in rational intuition is not properly characterized in terms of
its psychological effects—its seeming to us as if the claim is necessarily true; our
feeling inclined to believe it—even if those effects are reliably associated with true
beliefs.
Suppose that in an attempt to intuit (P), I start to study fractions, but I am unable to
attain Elias’s level of comprehension, and he decides to help me. He hooks me up to
the mathematics machine while I sleep. He types in (P) and the machine makes an
adjustment in my brain. From then on, when I consider (P) and understand it well
enough to believe it, it seems to me as if it is necessarily true and I feel a strong
inclination to believe it. Elias uses the machine to make similar adjustments for other
mathematical truths I understand well enough to believe but do not rationally intuit.
When I wake, I know nothing of what he has done. Some days later, when I again
consider (P), its truth just seems obvious to me. I proclaim, ‘‘Now I see it!’’ and stand
taller for my sense of accomplishment. Elias never tells me about the machine.

1
Although I shall sometimes write simply of intuition, my concern here is with rational intuition, as
opposed to intuition more broadly construed to include such things as hunches and ‘‘physical intuitions’’
(Bealer 2000). I do not claim that the four theories considered here are the only plausible contemporary
accounts of rational intuition. Limited space requires a limited examination. A consideration of these
theories shows how the problem at hand arises in a number of quite different contemporary views of
rational intuition.
2
All the theories I shall examine allow for the possibility of someone’s intuiting (P). In the last section of
the paper, I shall consider the possibility that (P) is beyond rational intuition.

123
Rational intuition and understanding 273

I now understand (P) in such a way that it seems necessarily true to me, and I feel
an inclination to believe it. My seeming state and inclination are truth-reliable.
Nonetheless, I do not rationally intuit (P). My understanding of (P) and of fractions
after the machine treatment is the same limited understanding I had before. I don’t
have the fuller understanding of fractions that Elias has, and I certainly do not
understand fractions, in general, or (P), in particular, in such a way as to ‘‘rationally
grasp’’ the truth of (P). Rational intuition involves a form of understanding that is
not correctly characterized in terms of its psychological effects, even if the effects
are a reliable source of true beliefs.
The basic point is not a new one. Descartes writes in The Rules for the Direction
of the Mind that intuition is ‘‘the indubitable conception of a clear and attentive
mind which proceeds solely from the light of reason’’ (p. 14; my emphasis).
Consider too Leibniz’s comment in The Discourse on Metaphysics, 24:
But distinct knowledge has degrees, because ordinarily the conceptions which
enter into the definitions will themselves have need of definition, and are only
known confusedly. When at length everything which enters into a definition or
into distinct knowledge is known distinctly, even back to the primitive
conception, I call that knowledge adequate. When my mind understands at
once and distinctly all the primitive ingredients of a conception, then we have
intuitive knowledge (pp. 41–42).
We need not endorse Leibniz’s view of intuitive knowledge as requiring a distinct
understanding of all the concepts in a proposition down to the primitive elements
(a view to which I’ll return at the end), but we should endorse the underlying notion
that rational intuition involves a distinctive form of understanding beyond what
mere belief requires.
As Leibniz indicates, rational intuitions provide the epistemic support they do
because of the distinctive understanding they involve. After my mathematics
machine treatment, I have a ‘‘machine-based intuition’’ of (P). It seems to me as if
(P) is necessarily true, and I feel inclined to believe (P) upon considering it. Elias,
however, has a rational intuition of (P), and his belief in (P) is epistemically superior
to mine. The point seems obvious, when put rhetorically: If Elias’ rational intuition
of (P) is not epistemically superior to my machine-based one, why is studying
epistemically preferable to plugging into the mathematics machine as a way to learn
(P)? The epistemic superiority of rational intuition can also be established through
two arguments by analogy.
The first analogy is between two forms of basic knowledge, our rational intuition
of necessary truths, on the one hand, and our introspective knowledge of
experiential contents, on the other, and focuses on the familiar case of the speckled
hen.3 When Edna and I each have the experience of a three-speckled hen, we are
each justified in the belief that we are appeared to three-speckledly, because we each
recognize our experience as an experience of three speckles. When I experience a
twenty-eight speckled hen, I am directly aware of my experience and its twenty-
eight speckles content, but I don’t recognize my experience as an experience of

3
For discussion of the speckled hen case, see Sosa (2003) and Fumerton (2005).

123
274 P. J. Markie

twenty-eight speckles, rather than twenty-seven or twenty-six, and I am not justified

in believing that I am so appeared to. Suppose, however, that Edna is an Alpha
Centaurian with more advanced cognitive abilities. She recognizes the content of
her experience as an experience of twenty-eight speckles, distinguishing it from
patterns with other numbers of speckles. She is justified in her belief that she is
appeared to twenty-eight speckledly,
The epistemic difference between Edna and me in this case remains, even if Elias
uses a machine to ‘‘rewire’’ my neurological connections, so that whenever I am
appeared to twenty-eight speckledly, it also seems to me as if I am so appeared to
and I feel inclined to believe that I am. I still don’t recognize my current experience
as an experience of twenty-eight speckles.4 As a result, my belief remains
epistemically inferior to Edna’s.5
Just as my machine-based, twenty-eight-speckles belief is epistemically inferior
to Edna’s recognition-based, twenty-eight-speckles belief, so too my machine-based
intuition of (P) is epistemically inferior to Elias’ rational intuition of (P). In the
speckled hen case, the machine treatment doesn’t give me the recognition of my
experiential content that Edna has of hers, and this makes a difference in the
epistemic status of our beliefs. In the intuition case, the mathematics machine
treatment never gives me the rational grasp of (P) that Elias has, and this again
makes for an epistemic difference in our beliefs.
It is helpful to consider a second analogy.6 An experienced birdwatcher makes
immediate and reliable bird identifications on sight. He gained his identification
ability over time. He first discovered a correlation between certain visual
experiences and the presence of a certain type of bird and used those justified
correlation beliefs to infer the bird’s presence. He then learned to make the
identifications non-inferentially, and his identifications are now basic beliefs for
him. Another birdwatcher makes the same immediate identifications on the basis of
the same experiences. He has been plugged into a bird identification machine and
had his brain modified so that, whenever he has these sorts of visual experiences, it
seems to him as if the birds are present and he feels a strong inclination to believe
that they are. A third birdwatcher is just like the second in having an unlearned
4
Some may object: If, after the machine treatment, whenever I have a twenty-eight-speckled hen
experience, it seems to me that I’m appeared to twenty-eight speckledly and I feel inclined to believe that
I’m so appeared to, then that is sufficient for me to recognize my experience as an experience of twenty-
eight speckles. Yet, suppose the machine treatment is a bit different. My neural connections are rewired
so that, upon having a twenty-eight-speckled hen experience, it seems to me as if Zeus exists and I’m
inclined to believe accordingly. That doesn’t mean that I now recognize my twenty-eight-speckled hen
experience as an experience of Zeus. It just means that when I have the experience, it seems to me as if
Zeus exists and I feel inclined to believe accordingly. The fact that, whenever I have an experience, things
seem a certain way to me and I feel inclined to believe that they are that way is not sufficient for me to
recognize the experience as an experience of things being that way.
5
After the machine modification, it seems to me as if I’m appeared to twenty-eight speckledly. For all
I argue here, it may, all other things being equal, be more epistemically appropriate for me to believe that
I am appeared to twenty-eight speckledly, if it seems to me as if I am so appeared to. My claim here is
that even such a seeming state does not raise the epistemic status of my belief to the level of Edna’s belief
that she is so appeared to, based on her recognition of the character of her experience.
6
The following case is roughly patterned after one described by Bergmann (2006, pp. 118–121). For
related discussion, see his (2008).

123
Rational intuition and understanding 275

identification ability, except that he gains his ability by a brain tumor. Due to the
brain tumor, whenever he has the relevant sort of visual experience, it just seems to
him as if a bird of the relevant sort is present and he feels a strong attraction to
believe accordingly. One day, all three see the same bird, have the same experience
and on its basis form the same identification belief. Is the belief of the first
birdwatcher epistemically superior to those of the others? The first birdwatcher’s
belief is epistemically superior to that of the third. Beliefs based on learned
identification skills are epistemically superior to those that result from tumor
induced seemings. The belief of the second birdwatcher is not, however,
epistemically superior to that of the third. There’s no relevant difference between
being plugged into the bird identification machine, on the one hand, and having a
brain tumor produce the same result, on the other. The first birdwatcher’s learned
identification belief is, therefore, epistemically superior to the second birdwatcher’s
machine-based one.7
Just as the learned identification belief of the first birdwatcher is epistemically
superior to the machine-based one of the second, Elias’ rational intuition of (P) is
epistemically superior to my machine-based intuition of it. An intuition of (P) that
results from a rich understanding of fractions is epistemically superior to one that
results from a machine modification of the psychological effects of a meager
understanding, just as an immediate identification belief that results from the
discovery of certain correlations between our experiences and the world is
epistemically superior to an immediate identification belief that results from a
machine modification of the psychological states associated with our experiences.
Let me forestall a misunderstanding and an objection.8 Suppose Elliot can see
that one tile has a rougher surface than another, but I can’t visually tell the
difference. I then have my vision enhanced by eyeglasses, and I see the difference
between the two surfaces. My belief in the texture difference is on an epistemic par
with Elliot’s, even though it was gained by the use of glasses. Analogously, my
machine-based intuition that (P) is on an epistemic par with Elias’ rational intuition
that (P). My reliance on the mathematics machine is just like my reliance on glasses.
This objection misses an important point. My glasses change the phenomenological
content of my visual experience so that the two surfaces look different to me where
they previously looked the same. The mathematics machine makes no adjustment in
the content of my understanding of fractions or of (P). It simply establishes certain
associations between my understandings, on the one hand, and how things seem to
me and my felt inclinations to believe, on the other.
A rational intuition of (P) is epistemically superior to a machine-based one by
virtue of the understanding it involves. What then is the nature of this
understanding? What sort of epistemic superiority does it support and how does it
support it? In response to the first question, we might say that Elias’ rational
intuition of (P) involves a ‘‘fuller’’ or ‘‘deeper’’ understanding of (P) than my

7
The second and third birdwatchers may gain some epistemic support for their beliefs from the fact that
it seems to them as if the bird is of a certain type. I do not claim that their beliefs are without epistemic
merit, only that their beliefs lack the epistemic merit of the first birdwatcher’s belief.
8
The following objection was suggested by Ernest Sosa.

123
276 P. J. Markie

machine-based intuition of it (Recall Leibniz’s reference to an understanding of all

the ‘‘primitive ingredients’’ of a conception.). Yet, the exact nature of this fuller or
deeper understanding still needs explanation. In response to the second question,
since epistemic value generally lies in forms of knowledge and justified belief, we
might say that beliefs supported by rational intuition have a status as knowledge or
as being justified that those supported by machine-based intuition lack (Recall
Leibniz’s reference to ‘‘degrees’’ of distinct knowledge.). The exact nature of this
knowledge or justification then needs explanation, as does the manner in which
rational intuition supports it. Alternatively, we might forego an appeal to knowledge
or justification and say that the epistemic superiority of rational intuition just lies in
the distinctive value of the understanding it involves. We then need an account, not
only of that understanding, but also of its distinctive value.
It is appropriate to expect accounts of rational intuition to identify the form of
understanding that it involves and to explain the epistemic status that that understanding
supports. Put simply, if a theory of rational intuition does not explain how rational
intuitions differ from, and are epistemically superior to, machine-based intuitions, it
hasn’t captured the role of understanding in rational intuition. The challenge is
daunting. Major theories—Phenomenal Conservatism, Proper Functionalism, the
Epistemic Competency Theory and the Direct Awareness View—fail to meet it.

2 Phenomenal conservatism

Huemer (2005) explains intuition as follows

An initial, intellectual appearance is an ‘intuition.’ That is, an intuition that p
is a state of its seeming true to one that p that is not dependent on inference
from other beliefs and that results from thinking about p, as opposed to
perceiving, remembering, or introspecting (p. 102).
He also adopts the Principle of Phenomenal Conservatism and explicitly argues
(2007, pp. 36–38) that it explains the epistemic force of rational intuitions in
particular.
PC: If it seems to S as if p, then S is prima facie justified in believing that p
(2001, p. 99).9
According to PC, if it seems to us as if a proposition is true, we are justified in
believing it in the absence of defeaters (2005, p. 105).
At first pass, this treatment of intuition contains no provisions by which to
distinguish between Elias’ rational intuition of (P) and my machine-based intuition
of it; let alone explain how one is epistemically superior to the other. From the
limited perspective afforded by Phenomenal Conservatism, Elias and I intuit (P) in
the same way: it seems to each of us as if (P) is true, that seeming results from our
thinking about (P), and it doesn’t depend on perception, memory or introspection.
We are both prima facie justified in believing (P). End of story.
9
See Huemer (2001) for an additional and quite helpful discussion of Phenomenal Conservatism in
general and as it applies to perception in particular.

123
Rational intuition and understanding 277

We can, of course, modify Phenomenal Conservatism to distinguish rational

intuitions from other intellectual seemings on the basis of their source. When the
intellectual seeming results from a ‘‘suitably full understanding,’’ let’s say, it’s a
case of rational intuition. When it results from a less than ‘‘suitably full’’
understanding, as by a modification through the mathematics machine, it is not a
case of rational intuition. Adjusted in this way, the theory distinguishes Elias’
rational intuition from my machine-based one, though we still need to explain when
someone’s understanding of a particular proposition is ‘‘suitably full.’’
The theory’s prospects are more limited when it comes to explaining the
epistemic superiority of rational intuition, even if rational intuitions are distin-
guished from machine-based ones in terms of the sort of understanding behind the
intellectual seemings. The theory awards prima facie justification to both rational
and machine-based intuitions on the same basis. Elias and I are both prima facie
justified in believing (P) because it seems to us as if it’s true. To explain the
epistemological superiority of Elias’ intuition, the theory must appeal to a quality
difference between prima facie justifications. It is hard to see what that could be.
There’s no difference in the rebutting defeaters that confront us with regard to (P),
either now or in the future. Neither of us has any evidence against (P) at this point,
and while that could change, I’m at no greater risk than Elias is. My belief in
(P) might be challenged, for example, by the misleading news that most experts
reject it. Then again, so might his. There’s no difference in the undercutting
defeaters that we have at this point. Like Elias, I have no reason to think that my
inclination to believe (P) stems from anything other than my grasp of it, and I’ve no
reason to think that my grasp of it is anything other than reliable.
Perhaps, my prima facie justification for believing (P) is less stable than Elias’, as
mine is open to some future undercutting defeaters that don’t threaten his. I might
learn that my inclination to believe (P) derives from the changes induced in my brain
by the mathematics machine, changes that could have been made relative to a false
proposition rather than a true one. I might then have an undercutting defeater for my
justification. Elias is not, it might be suggested, subject to a comparable threat from
potential undercutting defeaters. Yet, he can be. Suppose that the teacher who taught
Elias his fractions was highly incompetent, though he happened to teach Elias
correctly. Elias might then gain a defeater for his belief by learning that his teacher
was incompetent. He might later gain a defeater for that defeater by learning that the
teacher’s incompetence didn’t infect his abilities at rational intuition. Similarly, I
might gain a defeater by learning that my intuitions are machine-made but then gain a
defeater for that defeater by learning that the machine is highly reliable. We are on a
par in terms of potential threats to our prima facie justification. Yet, Elias is still
somehow in an epistemically superior position with regard to his belief in (P).
It might be suggested that the difference between us is simply this: We both
would believe upon consideration that (P) seems true to us because we have a
suitably full understanding of fractions, but in Elias’ case that belief is true and in
my case it is false.10 Perhaps, but this difference between us doesn’t explain how

10
If a ‘‘suitably full understanding’’ of fractions is simply one that leads P to seem true to us, then Elias
and I both have such an understanding of fractions. The proposal assumes that we have, already within

123
278 P. J. Markie

Elias’ rational intuition is epistemically superior to mine. The fact that he

potentially has a true belief about his understanding of (P) and I potentially have a
false one doesn’t, within Phenomenal Conservatism, imply that his current
epistemic position relative to (P) is in any way superior to my own.11
It is clear why Phenomenal Conservatism does not explain the epistemic
superiority of rational intuition. Phenomenal Conservatism takes the epistemic value
of rational intuition to rest in the seeming-state it involves. It ignores the crucial
component of the nature of the understanding that gives rise to that seeming state.12

3 Proper functionalism

According to Plantinga, to intuit a proposition is to directly see that it is true (1993,

pp. 106–107). He explains this sort of intellectual apprehension as follows.
So what is it then to see that a proposition p is true? All I can say is this: it is (1)
to form the belief that p is true and indeed necessarily true (when it is
necessarily true, of course), (2) to form this belief immediately, rather than as a
conclusion from other beliefs, (3) to form it not merely on the basis of memory
or testimony (although what someone tells you can certainly get you to see the
truth of the belief in question), and (4) to form this belief with that peculiar sort
of phenomenology with which we are well acquainted, but which I can’t
describe in any way other than as the phenomenology that goes with seeing that
such a proposition is true. We must add one further qualification…. One sees
that p only if the relevant cognitive module is functioning properly (p. 106).
Plantinga’s account of intuition differs notably from Huemer’s. He takes intuition
to be a special form of belief, rather than a special kind of seeming state.13 He
assumes that intuited propositions are believed to be necessarily true and that when
we form this belief, our relevant ‘‘cognitive module’’ is functioning properly.14

Footnote 10 continued
Phenomenal Conservatism, given an adequate analysis of the form of understanding integral to a rational
intuition and absent from a merely machine-based one.
11
Note that some people, e.g. children, who gain knowledge by rational intuition never form, or even
have the concepts to form, the meta-belief that their understanding of the proposition is ‘‘suitably full.’’
12
Critics have raised a similar objection to Phenomenal Conservatism’s treatment of perceptual beliefs.
It overlooks the fact that the cause of the perceptual seeming state can make a difference in the epistemic
appropriateness of the resulting belief. See (Markie 2005, 2006; Bergmann 2008; Huemer 2006, 2007;
Tucker 2010).
13
For a discussion of the difference between beliefs and seeming states, see (Huemer 2007; Conee 1998;
Bergmann 2008).
14
Where Huemer allows for our intuiting a false proposition, Plantinga does not allow for our seeing a
false proposition to be true. He does, however, introduce the concept of a priori belief, which he defines as
follows: ‘‘So what is it to believe p a priori? Take the conditions severally necessary and jointly sufficient
for seeing that p is true; to believe p a priori is to meet the set of those conditions minus the truth
conditions—that is, the condition that p be true in the case of seeing directly that p is true) and the
condition that p follows from q (in the case of seeing indirectly that p is true.)’’ (p. 106). I shall
concentrate on Plantinga’s account of (directly) seeing a proposition to be true. My points are equally
applicable to his account of a priori belief.

123
Rational intuition and understanding 279

Plantinga’s account enables us to distinguish Elias’ rational intuition of (P) from

my machine-based one. Elias meets all the conditions for seeing that (P) is true, and
the proper function condition, in particular. The cognitive module involved in the
formation of his belief is presumably some component of his understanding, and it
is functioning properly. Elias, like the rest of us, is designed so that, once he
understands (P) as well as he does, it just seems necessarily true to him and he
believes it to be so. My machine-based intuition of (P) isn’t a case of seeing it to be
true. I meet Plantinga’s first four conditions (I non-inferentially believe that (P) is
necessarily true, don’t rely on memory or testimony and have the same
phenomenology as Elias).15 I don’t meet the proper function condition. The
mathematics machine modifies my understanding so that I believe (P) without the
understanding of fractions that is required by my design plan.16
This difference between Elias’ rational intuition and my machine-based intuition
is sufficient, within Plantinga’s epistemology, to explain the epistemic superiority of
the former. Knowledge requires warranted belief, and beliefs are warranted only if
they are formed by properly functioning faculties successfully aimed at the truth.
Elias’ rational intuition supports a warranted belief in (P). My machine-based
intuition does not.17
Plantinga is mistaken, however. The epistemic superiority of Elias’ rational
intuition is not a matter of its being sanctioned by our design plan. It is superior
because it involves a form of understanding that my machine-based intuition lacks.
That form of understanding may be sanctioned by our design plan, assuming we
have one, but the point is irrelevant. The epistemic value of studying, rather than
plugging into the mathematics machine, isn’t based on the fact that using the
machine is ‘‘unnatural’’ for us. It is based on the fact that the result of using the
machine falls short of the fuller understanding study provides, a kind of
understanding that is required for rational intuition and supports an epistemically
superior belief.
Suppose that Alpha Centaurians are designed so that, up until the age of thirteen,
they are in the same position relative to (P) as I am, prior to being plugged into the

15
Plantinga may take the peculiar phenomenology of rational intuition to include more than a
proposition’s seeming to be (necessarily) true to us and our feeling inclined to believe it, but whatever the
nature of this peculiar phenomenology, there’s no reason to think that the mathematics machine cannot
produce it in us.
16
Consider Plantinga’s discussion of a similar case: ‘‘Suppose my cognitive faculties are redesigned by
an Alpha Centaurian superscientist in an experimental mood; he modifies them in such a way that when
I consider any proposition of the sort n is prime (where n is any of the first 10,000 natural numbers), it has
for me the very appearance of necessity enjoyed by even the most elementary of elementary truths of
arithmetic. I form the belief that n is prime, for some fairly large number n less than 10,000; chances are
I form a false belief; but even if it happens to be true, I don’t know that it is. Here the problem is that this
belief, though necessarily true, is not formed in me by virtue of faculties functioning properly and
successfully aimed at truth’’ (p. 108).
17
Consider too Michael Bergmann’s (2006) account of empirical doxastic justification: S’s belief B is
justified if and only if (i) S does not take B to be defeated and (ii) the cognitive faculties producing B are
(a) functioning properly, (b) truth-aimed and (c) reliable in the environment for which they were
‘designed’. If we extend Bergmann’s account to cases of rational intuition, it offers a Plantinga-like
account of why Elias’ rational intuition of (P) supports a justified belief and my machine-based intuition
does not.

123
280 P. J. Markie

mathematics machine. In these early years, they can understand (P) and believe it on
the basis of testimony, a hunch or a whim. They can’t rationally intuit it, however,
and it never just seems true to them on the basis of their understanding of it. When
they turn thirteen, things change under their design plan. Now when they consider
(P), they non-inferentially believe that it is necessarily true with no reliance on
memory or testimony, and their belief has the phenomenological character we
associate with rational intuition. This happens without any change in their
understanding of (P), the concepts it contains, or of fractions in general. According
to Plantinga’s account, they now see that (P) is true; their epistemic position relative
to (P) is superior to mine and on a par with Elias’.18 This is simply implausible.
Their design plan has only done for them what the mathematics machine did for me.
Their belief in (P) has changed to being immediate, non-testimonial and non-
memorial, and the phenomenological character of that belief has altered, but their
understanding of (P), in particular, and fractions, in general, has not changed a bit.
Such psychological changes, unaccompanied by any learning or increase in
understanding, are insufficient to produce an epistemic change, even when
mandated by a designer’s plan.
Plantinga’s position gains some plausibility from an apparent analogy between
perceptions and intuitions. Consider an example Bergmann (2006) presents in
defending a proper function requirement for empirical justification. Certain tactile
sensations lead us to believe noninferentially that there is a hard round object in our
hand; certain flowery smells lead us to believe that flowers are nearby. These
connections are different in an Alpha Centaurian. The ‘‘flowery’’ (to us) smells are
how she, equally reliably and naturally, identifies the presence of a hard round
object in her hand. We have no reason to epistemically favor our cognitive design in
this case. Our belief in a hard round object based on our tactile sensations is not
epistemically superior to her belief in the same proposition based on olfactory
experiences. Moreover, a plausible account of the common factor that accounts for
our jointly warranted beliefs/knowledge is proper function: We each form our belief
as provided for by our truth-aimed cognitive design plan. By analogy, the same may
seem to go for the intuitions of Elias and a teenage Alpha Centaurian. Here again,
we may be tempted to think, there is no basis to epistemically favor Elias’ cognitive
design over the Alpha Centaurian’s. Each rationally intuits and is warranted in
believing/knows (P). The difference in their degree of understanding of (P) or
fractions does not matter for their knowledge or warranted belief. What matters is
that they are both forming their belief in accord with their truth-aimed cognitive
design plan.
The problem here is clear enough. It is plausible that our phenomenal experiences
evidence our perceptual beliefs in such a way that which beliefs are evidenced by an
experience for a subject is a contingent property of the experience. The ‘‘flowery’’
smell that evidences our belief that there are flowers nearby evidences the Alpha
Centaurian’s belief that there is something hard and round in her hand. In contrast,
18
It seems that Plantinga would endorse this result. He tells us that, ‘‘a self-evident proposition is such
that a properly functioning (mature) human being can’t grasp it without believing it. This makes self-
evidence a species-relative notion; there may be angels or Alpha Centaurians for whom quite different
propositions are self-evident in this sense’’ (p. 109).

123
Rational intuition and understanding 281

the epistemic support relation is not contingent in rational intuition. What must be
understood for a rational intuition of (P) does not vary across possible subjects. A
rational intuition of (P) requires an appropriate understanding of fractions and that
can’t be replaced by a more meager understanding, any more than it can be replaced
by an understanding of the main characteristics of French Neo-Classical painting, no
matter what the subject’s design plan.
Rejecting Plantinga’s position does not preclude recognizing that variations in
cognitive design can play an important role in determining variations in the limits of
rational intuition. Perhaps angels or Alpha Centaurians are such that if they
understand (P) well enough to believe it (as even I do), then they also have the
understanding required to rationally intuit it (as Elias does). Their position relative
to (P) is like our position relative to the proposition that 1 = 1. There is no gap for
them, as there is for us, between having an understanding of fractions sufficient to
believe (P) and having an understanding sufficient to rationally intuit it. If they have
the one, they have the other. We can recognize possible variations in the range of
rational intuition and relate them to differences in cognitive design, even as we
reject Plantinga’s account of what it is to see intellectually a proposition’s truth.

4 The competency theory

Sosa (2007) takes rational intuitions to be intuitive intellectual seemings, which are
a kind of attraction to assent.
An intellectual seeming is intuitive when it is an attraction to assent triggered
simply by considering a proposition consciously with understanding (p. 60).
Rational intuitions are intuitive intellectual seemings that are explained by a certain
sort of epistemic competence.
S rationally intuits that p if and only if S’s intuitive attraction to assent
to hpi is explained by a competence (an epistemic ability or virtue) on the part
of S to discriminate, among contents that he understands well enough, the true
from the false, in some subfield of the modally strong (the necessarily true or
necessarily false), with no reliance on introspection, perception, memory,
testimony, or inference (no further reliance, anyhow, than any required for so
much as understanding the given proposition.) (p. 61)
We rationally intuit that p just when the following conditions hold: (1) it
intellectually seems to us as if p and this attraction to assent is triggered by our
considering and understanding p; (2) the explanation for the triggering is that it is an
instance of a general epistemic competence we have for distinguishing, through
such intellectual seemings, between true and false members of a set of modally
strong propositions that we understand well enough; and (3) our competence does
not involve any significant reliance on introspection, perception, memory, testimony
or inference.
An important point needs clarification. Sosa understands rational intuition in
terms of an attraction to assent that is explained by a competency at distinguishing

123
282 P. J. Markie

between true and false modally strong propositions that we understand well enough.
There are three components here: the quality of the understanding, the strength of
the attraction to assent and the reliability of the competency. Each comes in degrees.
How strong must the attraction to assent be? How well must the propositions be
understood? How reliable must the competency be? One option is to answer two of
these questions in terms of the third: The proposition intuited must be understood
well enough to trigger a strong enough attraction to assent, and an attraction to
assent is strong enough when it supports a suitably reliable ability to distinguish
between true and false propositions. More simply, an attraction to assent is strong
enough and a degree of understanding is full enough when they support a belief-
forming competency that is reliable enough. The only question then is how reliable
the competency has to be. Another option is to answer the questions separately. The
required degrees of attraction and understanding are not determined simply by what
will make for a suitably reliable faculty. What instead determines each is an open,
and difficult, question.
The choice between these options is important. If we develop the Competency
Theory along the lines of the first option, it won’t distinguish a rational intuition of
(P) from a machine-based intuition of it. Suppose that Elias is able by rational
intuition to distinguish between the true and false members of a set of mathematical
propositions, including (P), with 100% reliability. When he considers a true claim in
the set, he understands it well enough to have a felt attraction to believe it and the
attraction is strong enough that he actually believes it. When he considers a false
claim, he understands it well enough to have a felt attraction to believe its negation
and the attraction is strong enough that he actually does so. According to the
Competency Theory, he rationally intuits each proposition he believes, including
(P). I initially perform just as Elias does, except in the case of (P), where my
understanding of fractions doesn’t give rise to an attraction to assent to either it or
its negation. Then, unbeknownst to me, I am hooked into the mathematics machine,
and my current level of understanding becomes sufficient to support an attraction to
assent to (P). The Competency Theory implies that I rationally intuit (P), just as
Elias does. Any difference in our understandings of (P) and fractions is irrelevant, as
our understandings both support a completely reliable ability to distinguish the true
propositions in the set from the false ones.
The theory does better, if we take the second option.19 Rational intuition requires
that the attraction to assent be part of an epistemic competency that is reliable
enough and in the exercise of which the propositions believed are understood well
enough. The standard for a good enough understanding—whatever it may be—is
not specified simply in terms of a strong enough attraction to assent and a reliable
enough competency. Within the Competency Theory, we can now distinguish my
machine-based intuition of (P) from Elias’ rational intuition of it. I don’t understand
(P) well enough to have a rational intuition, even if I do understand it well enough to
have a reliable attraction to assent to it. So far so good.

19
Sosa has indicated in correspondence that this is his preference and speculated that contextual factors
may play a role in determining the required degrees of understanding, attraction and reliability.

123
Rational intuition and understanding 283

The theory still contains a serious problem, however. It fails to account for the
epistemic superiority of Elias’ rational intuition over my machine-based one.
Suppose again that Elias is able by rational intuition to distinguish with 100%
reliability between the true and false members of some set of mathematical
propositions including (P). I can do the same by machine-based intuition. Each of us
exercises a competency to distinguish the true from the false through an intuitive
attraction to assent that is not based in introspection, perception, memory, testimony
or inference. Each of us distinguishes the true from the false by intuitive intellectual
seemings, as Sosa defines them. The only difference is that Elias’s competency
involves a greater degree of understanding of the intuited propositions, one great
enough to count as rational intuition. How then is Elias’s position epistemically
superior to mine?
It is not a matter of justified beliefs on Sosa’s account. Sosa distinguishes
between basis-dependent justification and virtue justification.
Basis-dependent foundational justification is foundational justification that
derives essentially from the justified belief’s being based on a given state, a
psychological state of the subject’s, one that lies beyond justification and
unjustification (p. 50).
Virtue foundational justification is foundational justification that derives
essentially from the justified propositional attitude’s manifesting an epistemic
competence (p. 51).20
He takes rational intuition to be a source of virtue foundational justification. Yet, if
it is, then so is machine-based intuition. I am exercising an epistemic competence
just as much as Elias is, and my competence is just as reliable.
The difference between rational and machine-based intuitions also is not a matter
of knowledge on Sosa’s account. One form of knowledge within his theory is apt
belief or ‘‘animal knowledge.’’ A belief is apt when it is accurate, adroit—in the
sense of manifesting an epistemic competence—and accurate because adroit (p. 22).
Elias and I both have apt beliefs in (P). Our beliefs manifest different competencies
(different forms of intuitive intellectual seeming involving different levels of
understanding), but the competencies are equally accurate and each belief is
accurate because it is an exercise of the relevant competency. Sosa’s second form of
knowledge is reflective knowledge, which is apt belief aptly noted. Here too, Elias
and I are on a par. Insofar as Elias has the ability to aptly believe that he aptly
believes (P), so have I. The only difference between us lies in the degrees of
understanding involved in the different epistemic competencies by which we gain
animal knowledge of (P). This difference is irrelevant to our ability to gain apt
knowledge that our belief in (P) is accurate because formed by an epistemic
competency.

20
Sosa speculates (p. 51) that all basis-dependent foundational justification may be virtue foundational
justification, even as some cases of virtue foundational justification are not cases of basis-dependent
foundational justification.

123
284 P. J. Markie

Bealer (2000) seems to offer a way to explain the special epistemic status of
rational intuition within the Competency Theory. Like Sosa, he takes rational
intuitions to be intellectual seemings:
(W)hen you first consider one of de Morgan’s laws, often it neither seems to
be true nor seems to be false, but after a moment’s reflection, however,
something new happens: suddenly it just seems true. Of course this kind of
seeming is intellectual, not sensory or introspective (or imaginative) (p. 3).
He also, like Sosa, thinks that rational intuition requires a certain level of
understanding, the full or determinate possession of the relevant concepts.
A subject possess a concept in the full sense iff (i) the subject at least
nominally possesses the concept and (ii) the subject does not do this with
misunderstanding or incomplete understanding or just by virtue of satisfying
our attribution practices or in any other such manner (p. 11).
Bealer then adds a distinctive account of how rational intuition is a source of basic
justification. According to Bealer, a basic epistemic source must meet the demands,
not just of reliablism, but of modal reliablism: ‘‘something counts as a basic source
of evidence iff there is an appropriate kind of strong modal tie between its
deliverances and the truth’’ (p. 9). Rational intuition, because it involves the
determinate possession of the relevant concepts, can meet the demands of modal
reliabilism. My machine-based intuition presumably does not. We can now explain
how Elias’s rational intuition that (P) is superior to my machine based one. Because
it is modally reliable, rational intuition is a basic epistemic source. Elias’s belief in
(P) is justified. My machine-based intuition is not modally reliable and so not a
basic epistemic source. My belief that (P) is not justified.21
How though are we to understand the requirement of modal reliability? If we
specify it too weakly, we won’t distinguish Elias’ rational intuition from my
machine-based one. Some, such as Sosa (1996), have properly questioned whether
Bealer’s modal reliabilism actually does imply that sources such as my machine-
based intuition do not provide basic epistemic justification. Given the identity of
their contents, Elias’ rational intuition and my machine-based intuition are equally
modally reliable.22 If we specify modal reliability too strongly, the requirement will
have a counter-intuitive implication for other sources of belief. Perception is clearly
a source of basic justified beliefs about the external world. We see the tree before us,
and we are justified in believing that it’s there, without inferring that belief from any
others. Bealer’s modal reliablism implies otherwise. Perception is not a source of
basic justified beliefs as it is only contingently reliable. To explain the special
epistemic status of rational intuition, we will set too high a standard for a basic
epistemic source.

21
See Bealer (2000, pp. 8–9) for a discussion of a case similar to my mathematics machine example in
which space aliens use telepathy to give a someone a contingently reliable faculty of guessing what is
necessarily true.
22
See Bealer (1996) for a reply.

123
Rational intuition and understanding 285

5 The direct awareness theory

Conee (1998) presents a direct awareness theory of rational intuition.23 He considers

three propositions we just see to be true:
(A) Everything is self-identical.
(B) Two is more than one.
(C) Any golden trumpet is a trumpet.
He thinks that our seeing each proposition to be true is a form of direct
awareness.
The seeing is a direct awareness. The objects of this awareness are
intrinsically obvious propositions, their conceptual ingredients and the
relations among the latter (p. 852).
This direct awareness is a source of epistemic support.
There is something positively confirming that goes on in seeing the truth of
e.g. (C). We notice that (C) is only about things that are both golden and
trumpets and that (C) only attributes one of those very features to any such
things. This is not a matter of merely understanding the proposition and failing
to see how it could not be so. We see the relevant inclusion relation asserted,
and see that it is truth-making, even if we are unable to describe the bearing on
truth of what is seen in some illuminating way (p. 854).
Bonjour (1998) offers a similar direct awareness account of his rational intuition
that nothing can be red all over and green all over at the same time.
First, I understand the proposition in question. This means that I comprehend
or grasp the property indicated by the word ‘red’ and also that indicted by the
word ‘green,’ that I have adequate conceptions of redness and greenness
(which is not, of course, to say that I know everything about even their
intrinsic natures, let alone their relational properties.) Similarly, I understand
the relation of incompatibility or exclusion that is conveyed by the rest of the
words in the verbal formulation of the proposition, together with the way in
which this relation is predicated of the two properties by the syntax of the
sentence. Second, given this understanding of the ingredients of the
proposition, I am able to see or grasp or apprehend in a seemingly direct
and unmediated way that the claim in question cannot fail to be true….(p. 101)
As I understand them, Conee and Bonjour share the following view. When we
rationally intuit a proposition, we are directly aware of the proposition, its
conceptual elements and the relations among those elements; and we are directly
aware of the fact that the relations among the conceptual elements are (necessarily)

23
See Fales (1996) for another version of the Direct Awareness View. See Bergmann (2006) for a fine
discussion of some of the problems confronting Fales’ position and, in particular, it central notion of
transparency.

123
286 P. J. Markie

truth-making. This complex direct awareness gives our belief in the proposition a
positive epistemic status.24
We seem to have an explanation of how Elias’ rational intuition of (P) differs
from my machine-based intuition of it. Elias and I both understand (P). It seems true
to both of us, and we both feel inclined to believe it. We may both be unable to
conceive of it as being false. Yet, Elias is directly aware of (P), its conceptual
elements and the relations among them, and he is directly aware of the fact that the
relations among the elements of (P) are truth-making. This complex direct
awareness gives his belief that (P) some positive epistemic status. My machine-
based intuition does not involve such a direct awareness and, presumably, does not
give my belief that (P) the same epistemic status.
This is a start, but we need to know much more. What is direct awareness and
why does Elias have it with regard to (P) when I don’t? It is plausible to assume that
whenever we believe a proposition, we are directly aware of it and its conceptual
elements. When, for example, I believe (P), I am directly aware of (P), the concept
of 237/148, the concept of 455/266, and the relation of being greater than. It is also
plausible to assume that facts are just true propositions. In particular, the fact that
the relations among the elements of (P) are truth-making is just the true proposition
that the relations among those elements are truth-making. Suppose then that the
mathematics machine treatment causes me to believe, not just (P), but the
proposition that the relations among the elements of (P) are truth-making. I
understand and believe both of these propositions without ever gaining the
knowledge of fractions that supports Elias’ rational intuition of (P). I am now
directly aware of all that Conee and Bonjour require for seeing that (P) is true: (P),
the conceptual elements of (P), the relations among the elements and the fact that
the relations are truth-making. Yet, I still have not improved my understanding of
fractions, I still don’t have Elias’ rational intuition, and my belief that (P) still lacks
the epistemic status of his belief. Conee and Bonjour might, of course, reject the
assumption that believing (P) is sufficient for a direct awareness of (P), its
conceptual elements and the relations among them. They then owe us an alternative
account of what direct awareness is and what it requires.25
Moreover, assuming that Elias has this direct awareness and I do not, why does
that give his belief that (P) a positive epistemic quality that my belief lacks? One
possibility is that the direct awareness involved in Elias’s rational intuition supports
justification, while the seeming state involved in my machine-based intuition does
not. Yet, to allow for false beliefs to be a priori justified, Bonjour properly takes
apparent, rather than real, rational insight to be the source of a priori justification
(p. 113), and my machine-based intuition provides me with an apparent rational
24
Conee’s and Bonjour’s position is of a piece with Richard Fumerton’s general characterization of the
internalist approach to basic belief and knowledge:
‘‘(T)he internalist wants to ground all justification on a ‘direct confrontation’ with reality. In the
case of a non-inferentially justified belief, the internalist wants the fact that makes true the belief
‘there before consciousness’’’ (1995, p. 83).
25
Note the similarity between this problem in Conee’s and Bonjour’s accounts of rational intuition and
the problem of the speckled hen, as it plays out for direct awareness theories of epistemically basic
empirical beliefs. See, in particular, Fumerton (2005).

123
Rational intuition and understanding 287

insight. It seems to me as if, given the concepts involved, (P) must be true. A second
possibility is that Elias and I both have justified beliefs that (P), but only Elias has
knowledge that (P). Bonjour seems inclined to take this approach at one point.
(A)n apparent insight might fail to be genuine in the sense specified, i.e., might
fail to be a genuine perception of the necessary character of reality, and yet
might have as its content a proposition that is in fact true and necessary. Such
a situation could result from mere chance or from some more complicated
explanation—e.g., influence from someone who does have a genuine insight
(Such a case would be something like an a priori version of a Gettier case.)
(p. 113).
Elias’s rational intuition gives him knowledge that (P), while my machine-based
intuition does not. My situation is something like a Gettier case. Yet, how is that?
Unlike the subjects in the original Gettier cases, I have not reasoned through a false
premise. Unlike the subjects in subsequent, non-inferential Gettier-style cases, I
have not accidentially formed a true belief (That’s a barn) by a faculty (perception)
that is otherwise unreliable in the sort of situation I’m in (a valley salted with fake
barns). My machine-based intuitions are quite reliable, as are my intellectual
seeming-states more generally, and they are truth-sensitive: if I had considered a
false proposition, rather than P, it would not have seemed true to me. My machine-
based intuitions are appropriately connected to the truth.

6 A traditional approach and a suggestion

It’s time to take stock. Rational intuition involves a form of understanding that
several current theories either overlook or fail to explain. Phenomenal Conservatism
focuses on seeming states and ignores the understanding behind them. Proper
Functionalism mistakenly takes the distinctive form of understanding involved in
rational intuition to be a matter of proper function. The Competency Theory
acknowledges that rational intuition requires a certain form of understanding but
does not explain how that understanding makes rational intuition epistemically
superior to forms of intuition that lack it. The Direct Awareness View claims that
the understanding required for rational intuition is a direct awareness of the
proposition, its conceptual elements, the relations among them and the fact that the
relations are truth-making, but it fails to explain how this direct awareness is lacking
in machine-based intuitions or how this direct awareness is a source of special
epistemic support.
Consider then a different approach drawn from the past. Leibniz tells us that
when the ‘‘mind understands at once and distinctly all the primitive ingredients of a
conception, then we have intuitive knowledge.’’ Descartes presents a similar view in
the Rules for the Direction of the Mind. In the first instance, our intuitions are of
‘‘self-evident propositions’’, but we can extend the category of rational intuitions to
include propositions we deduce from what is self-evident, provided the deductions
proceed by ‘‘a continuous and uninterrupted movement of thought in which each
individual proposition is clearly intuited’’ (p. 15). Let us say then that some of the

123
288 P. J. Markie

claims we rationally intuit are ‘‘first principles;’’ these propositions are so simple it
is impossible for an inquirer to understand them without rationally intuiting them.
All our other rational intuitions are of claims we deduce, with no reliance on
memory, from currently intuited first-principles. If we proceed in this general way,
we still have the task of explaining just what rational intuition is and how it supports
epistemic justification and knowledge, but there’s no longer any problem in
distinguishing my machine-based intuition of (P) from Elias’s rational intuition of
it. Elias has no rational intuition of (P), as I have described his case. (P) is not a first
principle, since I understand it without rationally intuiting it, and Elias is not
deducing it from rationally intuited first principles, since, by assumption, he is not
deducing it at all. If Elias ever does rationally intuit (P) by deducing it from intuited
first principles, we can easily distinguish his epistemic state from my machine-based
intuition of (P). Elias’ deduction provides him with knowledge of more information,
and, in that information, a stronger basis for believing (P) than my machine-based
intuition of (P) provides me.26
This return to the past has significant drawbacks. It severely restricts the scope of
rational intuition to first principles, which must be intuited if understood, and what
can be deduced from them with no reliance on memory. As noted, our initial
assumption that Elias can non-inferentially intuit that (P) has to be set aside. This
approach also implies that no other species of inquirer could possibly have a non-
inferential rational intuition of a proposition that we can understand but can’t intuit.
It makes rational intuition infallible, since any proposition it is impossible to
understand without rationally intuiting is, presumably, true, and any deductions
involved in rational intuition are, presumably, logically valid. These implications
are generally regarded by current theorists as mistakes of the past.27
There is though a nugget of insight here that indicates a more promising approach
to our problem. Rationally intuiting a proposition involves both understanding it so
as to believe it, but also understanding how it is true. For Descartes and Leibniz,
first-principles are so simple that to understand them is to understand how they are
true, and, for more complex propositions, such as (P), understanding how they are
true requires deducing them from intuited premises that explain their truth. This is
too restrictive, but the basic insight is correct. Elias’ rational intuition that
(P) involves a deeper understanding of (P) than my machine-based intuition, and
that deeper understanding is an understanding of how it is that (P) is true. This

26
See Chisholm (1989) for an example of this general approach developed for a priori knowledge rather
than rational intuition.
27
Plantinga is representative in his rejection of these implications and his association of them with a
‘‘traditional account’’ of rational intuition.
(T)he tradition also held that self-evident propositions… are such that we can’t even grasp or
understand them without seeing that they are true. But is the idea that it is logically impossible (in
the broadly logical sense) that I understand such a proposition and fail to see that it is true? There
seem to be some such propositions, but surely there aren’t many—not nearly as many as,
according to the tradition, there are self-evident truths. A better position, I think, is that a self-
evident proposition is such that a properly functioning (mature) human being can’t grasp it without
believing it. This makes self-evidence a species-relative notion (pp. 108–109).
The tradition also ‘‘displayed an unhappy penchant for the view that intuition is infallible’’ (p. 109).

123
Rational intuition and understanding 289

deeper understanding does not involve Elias’ deducing (P) from other rationally
intuited propositions, all the way back to first-principles. Yet, it is plausible to take
Elias’ understanding of (P) to involve two other components: first, he has some
justified mathematical beliefs the contents of which provide an explanation of how
it is that (P) is true, and, second, the fact that he has those beliefs is why it seems to
him that (P) is true. His belief that (P) is, as initially assumed, non-inferential and
based on its seeming to him that (P), but it only seems to him that (P) because he has
these other beliefs, which together provide an account of how it is that (P) is true
and are justified for him. I, on the other hand, have the state of its seeming to me that
(P) because of the mathematics machine modifications, which do not involve my
having any justified beliefs that explain how it is that (P) is true.
Elias’ rational intuition that (P) is then epistemically superior to my machine
based intuition in at least two ways. First, in terms of the value of epistemic
justification, Elias has justified beliefs that are evidence for him for (P), so that his
belief in (P) is potentially justifiable, if not actually justified, for him in a way in
which it is not justifiable for me. Second, there is epistemic value just in the deeper
understanding that is causally responsible for his belief. Elias’ belief that (P) is
epistemically superior to mine just by virtue of the fact that it is based on a seeming
state that results from his having beliefs that explain how it is that (P) is true.
Believing that (P) based on a seeming state caused by (justified) beliefs that provide
an explanation of the truth of (P) is, in itself, epistemically superior to believing
(P) as a result of a seeming state caused by a mathematics machine modification. It
is not that Elias is justified in believing (P) and I am not, or that Elias knows (P) and
I do not. It is that the seeming state that supports whatever justified belief or
knowledge that (P) Elias has is the result of his having justified beliefs that explain
how it is that (P) is true. The seeming state that supports whatever justified belief or
knowledge that (P) I may have is the result of a machine modification that involves
no explanation of how it is that (P) is true.28
This approach clearly needs to be developed. Obvious points need attention.
Consider our rational intuition of what Descartes and Leibniz would have taken as a
first-principle, something we cannot understand without rationally intuiting it, e.g.,
that 1 = 1. It just seems to us as if this proposition is true. Does that seeming result
from our having other beliefs that explain how it is that 1 = 1 or is the fact that
1 = 1 a limiting case with no explanation beyond itself? Consider too a case in
which Elias forgets the particular mathematical beliefs that once were his available
explanation of why (P) is true, yet it still seems to him that (P) is true. His seeming
state results from his once having had those beliefs, but now that they are forgotten,
(P) is no longer justifiable for him on their basis and his beliefs no longer include an
28
Consider a similar difference, that between believing that (P) on the basis of expert testimony and
believing that (P) on the basis of having deduced it from first-principles of mathematics. The latter is
clearly epistemically superior to the former, but how? Each may be a case of justified belief or even
knowledge. The latter involves more information, more justified beliefs or knowledge, but it is not just
that. The additional information in the latter case provides an explanation of how it is that (P) is true. In
proving (P) for oneself, one gains an explanation of how it is that (P) is true. There is epistemic value in
having that explanation. Elias, in my example, directly intuits (P) without deducing it, but he is in a
similar position. It only seems to him that (P) because he has other (justified) mathematical beliefs that
provide an explanation of how (P) is true.

123
290 P. J. Markie

explanation of how it is that (P) is true. Does he still have a rational intuition that
(P); if so, what makes his intuition that (P) superior to my machine based one?
There is work to do here. Nonetheless, I suggest that this approach, drawn from
the past, is a promising way to give understanding its due in rational intuition and to
move beyond the difficulties in the four contemporary theories I‘ve examined.

Acknowledgement I am grateful to participants in the discussion for their helpful comments. I have
benefitted from discussions of this material with Matt McGrath and from the very helpful criticisms of a
referee for this journal.

References

Bealer, G. (1996). ‘A priori knowledge:’ Replies to William Lycan and Ernest Sosa. Philosophical
Studies, 81, 163–174.
Bealer, G. (2000). A theory of the a priori. Pacific Philosophical Quarterly, 81, 1–30.
Bergmann, M. (2006). Justification without awareness. Oxford: Oxford University Press.
Bergmann, M. (2008). Externalist justification and the role of appearances. Unpublished.
Bonjour, L. (1998). In defense of pure reason. Cambridge: Cambridge University Press.
Chisholm, R. (1989). Theory of knowledge. Englewood Cliffs, NJ: Prentice-Hall.
Conee, E. (1998). Seeing the truth. Philosophy and Phenomenological Research, 4, 847–857.
Descartes, R. (1628). Rules for the direction of the mind, (J. Cottingham, R. Stoothoff, & D. Murdoch,
Trans.). The philosophical writings of descartes. Cambridge: Cambridge (1985).
Fales, E. (1996). A defense of the given. Lanham, MD: Rowman and Littlefield.
Fumerton, R. (1995). Metaepistemology and skepticism. Lanham, MD: Rowman and Littlefield.
Fumerton, R. (2005). Speckled hens and objects of acquaintance. Philosophical Perspectives, 19,
121–139.
Huemer, M. (2001). Skepticism and the veil of perception. Lanham, Md: Rowman and Littlefield.
Huemer, M. (2005). Ethical intuitionism. Houndmills, UK: Palgrave Macmillan.
Huemer, M. (2006). Phenomenal conservatism and the internalist intuition. American Philosophical
Quarterly, 43, 147–158.
Huemer, M. (2007). Compassionate phenomenal conservatism. Philosophy and Phenomenological
Research, 74, 30–55.
Leibniz, G. W. (1973). 1686 Discourse on metaphysics (G. Montgomery, Trans.) Leibniz. La Salle, IL:
Open Court Publishing.
Markie, P. (2005). The mystery of direct perceptual justification. Philosophical Studies, 126, 347–373.
Markie, P. (2006). Epistemically appropriate perceptual belief. Nous, 40, 118–142.
Plantinga, A. (1993). Warrant and proper function. Oxford: Oxford University Press.
Sosa, E. (1996). Rational intuition: Bealer on its nature and epistemic status. Philosophical Studies, 81,
151–162.
Sosa, E. (2003). Epistemic justification: Internalism vs. externalism (with Laurence Bonjour). Oxford:
Blackwell.
Sosa, E. (2007). A virtue epistemology. Oxford: Oxford University Press.
Tucker, C. (2010). Why open minded people should endorse dogmatism. Philosophical Perspectives,
24(1), 529–545.

123