Intuition and Conscious Reasoning

Ole Koksvik∗
(Forthcoming in The Philosophical Quarterly.)

Abstract

This paper argues that, contrary to common opinion, intuition can

result from conscious reasoning. It also discusses why this matters.

Introduction
Can having an intuition justify belief? That depends, in all likelihood, on
what intuitions are. If intuitions are like emotions the answer is probably
‘no’, since emotions do not in general justify belief in their content. If intu-
itions are relevantly similar to perceptual experiences, on the other hand,
the answer is more likely ‘yes’, since perceptual experiences generally do.
Those who care about the epistemology of intuitions should therefore also
care about their nature.
Before counting a mental state as an intuition, many thinkers place re-
strictions on the state’s causal history. There are two broad types of such
restrictions: the state must either have a particular history, or it must lack
one. A common positive etiological restriction is that the state must derive
∗ Many thanks to David Chalmers, Daniel Korman, Leon Leontyev, Daniel Nolan, Joel
Pust and Weng Hong Tang for helpful discussion.

1
“from one’s understanding of one’s concepts”.1 A very widely accepted
negative etiological restriction is that the state must fail to result from con-
scious reasoning.2
I argue that this second restriction should be rejected: contrary to pop-
ular opinion, intuition can result from conscious reasoning. I then show
why this matters.

Aims and Approach

It may be worth noting that I do not aim to partake in—or start—a quarrel
about how we should use the word ‘intuition’. I take this to be a merely
verbal dispute3 which we would do well to avoid.
The interesting questions are whether there is a class of mental states
which constitutes a good candidate for a psychological kind—a kind
which cuts the mind at its natural joints—and which deserves the label
‘intuition’ well enough. If such a class exists it is furthermore interesting
to ask what its characteristics are. My aim here is to help answer questions
such as these, and not to argue over semantics.
1 George Bealer, “The Origins of Modal Error”, Dialectica 58 (2004), pp. 11–42, at p. 13;
George Bealer, “Intuition and Modal Error”, Epistemology: New Essays, ed. by Quentin
Smith, Oxford University Press, 2008, at p. 191; see also Paul A. Boghossian, “Virtuous
Intuitions: Comments on Lecture 3 of Ernest Sosa’s A Virtue Epistemology”, Philosophical
Studies 144 (2009), pp. 111–119, p. 119; and Laurence BonJour, In Defence of Pure Reason,
Cambridge University Press, 1998, at p. 101.
2 See, for example, Paul A. Boghossian, “Inference and Insight”, Philosophy and Phe-

nomenological Research 63 (2001), pp. 633–640, at p. 636; Laurence Jonathan Cohen, The Di-
alogue of Reason, Clarendon Press, 1986, at pp. 75–6; Alison Gopnik and Eric Schwitzgebel,
“Whose Concepts Are They, Anyway? The Role of Philosophical Intuition in Empirical
Psychology”, Rethinking Intuition, ed. by Michael DePaul and William Ramsey, Rowman
& Littlefield, 1998, at p. 77; Michael Lynch, “Trusting Intuition”, Truth and Realism, ed. by
Patrick Greenough and Michael P. Lynch, Oxford University Press, 2006; Alvin Plantinga,
Warrant and Proper Function, Oxford University Press, 1993, at p. 106; Joel Pust, Intuition
as Evidence, Garland Publishing, 2000, at pp. 44–5; but compare Michael Huemer, Ethical
Intuitionism, Palgrave Macmillian, 2005, at p. 101. See also Herman Cappelen, Philosophy
Without Intuitions, Oxford University Press, 2012, at pp. 33, 46.
3 David Chalmers, “Verbal Disputes”, The Philosophical Review 120 (2011), pp. 515–566.

2
 How might one support the claim that there is a psychological kind
which deserves to be called ‘intuition’ and which can result from conscious
reasoning? Well, either by arguing that we have reason to believe that
there is a nearby psychological kind which isn’t subject to this etiological
requirement, or by arguing that lack reason to believe that there is a nearby
psychological kind which is.
The former task might be carried out by providing a positive charac-
terisation of a class of mental states without making reference to etiology.
If the class deserves the label ‘intuition’ well enough, and if the members
share important and interesting features—perhaps they can play a certain
epistemological role, for example—the claim has been defended. Else-
where I have attempted to provide just such a characterisation, but it is,
I confess, quite long.4 Here I will therefore go the other route, and attempt
to support the claim that we lack reason to think that there is a nearby class
of mental states which cannot result from conscious reasoning and which
constitutes a psychological kind.

Intuitions Can Result from Conscious Reasoning

To fix ideas, suppose you adopted the view that intuitions are experiences
of a certain kind, such that when one has the intuition that p it seems to
one that p, and that simply having an intuition provides some justification
to believe what it represents. (This is my view.) Should you think that
intuitions in this sense can result from conscious reasoning?
It is true, of course, that one can come to believe a proposition p by
reasoning one’s way to p, and that this needn’t involve its coming to seem
to one that p: one might believe that certain premises are true and that p
follows from those premises, for example.5 But why should we think that
4 Ole Koksvik, “Intuition”, PhD thesis, The Australian National University, 2011.

Available from www.koksvik.net.

5 See for example George Bealer, “The Incoherence of Empiricism”, Proceedings of the

3
a process of conscious inference with p as its conclusion cannot result in it
seeming to the agent that p? If the task of interest is to delineate a good
candidate for a psychological kind there seems to be no reason to require
from the outset that conscious argument be absent.
In fact, such a requirement seems ad hoc. Even those who restrict their
attention to what they call ‘rational’ or ‘philosophical’ intuition do not ban
antecedent conscious deliberation about the concepts involved in p imme-
diately before the intuition arises. For example, no one thinks there should
be a ban on thinking about the logical connectives before having the in-
tuition that one of de Morgan’s laws holds. And it would be absurd to
hold that one cannot think about the law itself just before having the in-
tuition. As Bealer points out, it is most often precisely when you consider
these things that the law suddenly seems true to you (1992, p. 101). It is
very hard to see what reason one might have to allow the deliberation that
goes on beforehand to take any form whatever, except only the particular
form of an argument. Contrary to common opinion, then, the clear pre-
sumption should be that the negative etiological restriction does not apply
to intuition. We need an argument to deviate from this stance.
Joel Pust has recently presented just such an argument:6

[U]nless intuitions are non-inferential they cannot serve . . . as

the ultimate premises in philosophical argumentation and
analysis. Philosophical practice treats intuitions as basic, as not
admitting of further inferential support, and this provides us
with a reason for requiring of any genuine intuition that it not
be the result of conscious inference. (p. 45)

Similarly, L. Jonathan Cohen, to whom Pust attributes this argument, ar-

Aristotelian Society, Supplementary Volume 66 (1992), pp. 99–138, at p. 102. It is a separate
and interesting question whether one must have an intuition corresponding to each tran-
sition in a proof or argument, as Locke arguably thought; see An Essay Concerning Human
Understanding §§4.2.1–4.2.7.
6 Pust no longer defends this argument.

4
gues that “[i]f intuition is to provide the ultimate premises of philosoph-
ical argument, those premises should not themselves be the conclusions
of further reasoning” (p. 76). And it is reasonable to assume that the very
widespread acceptance of the negative etiological requirement is largely
due to sympathy with similar reasoning.
I want to make two comments about this argument. First, Cohen and
Pust are concerned in particular with philosophical intuition. Given that
project, perhaps it makes sense to limit the candidates according to the
role intuition is thought to play in theory construction. If, however, the
interest is in delineating a good candidate for a psychological kind this is
getting things the wrong way around: we first need to delineate a kind,
and then see what work it can do in philosophy.
But second, let us distinguish two senses of being ‘non-inferential’. In
one sense, S’s intuition that p is non-inferential if it is not the result of—
in the sense of being caused by—conscious deliberation. In another, the
intuition is non-inferential just in case S’s justification to believe that p af-
ter having the intuition does not wholly rest on the support p receives in
virtue of being the conclusion of an argument.
To provide foundational justification, it is clear that intuition must
be non-inferential in the second sense. But why think it must be non-
inferential in the first sense? I can think of no other reason than the belief
that the two do not come apart.
But in fact, clearly they do. To see this, imagine that I do not grasp de
Morgan’s laws, and that you set out to explain them to me:

Assume that it is not the case that p-and-q, which is to say that
p-and-q is false. One way for that to happen is if p is false. In
that case, p and q are obviously not both true (we just said that
p is false). And if p and q are not both true, p-and-q is false. So
one way for p-and-q to be false is for p to be false.
Naturally, another way to get the same result is for q to be false

5
 instead: the reasoning is just the same. And a third way is if
p and q are both false. But if p-and-q is false, one of these three
things has to to be the case: either p is false, or q is false, or both
p and q are false. There is no other way.
Now, not-p-or-not-q is true in exactly those three situations;
when either one of p and q is false, or both p and q are false.
So, you see, if it’s not the case that p-and-q, then it is the case
that not-p-or-not-q.

This may not be the snappiest of arguments. But it is an argument. It is

valid, and it yields one direction of one of de Morgan’s laws as its conclu-
sion. In a similar fashion, you could have explained the other direction to
me. But it is surely at least possible that at the end of such explanations it
comes to seem to me that the transformation in question is valid. After all,
that seems to be the point of the entire affair!
Had I been a little quicker I might have arrived at the point where
I could ‘just see’ that the transformation holds simply by staring at
¬(p & q) ↔ (¬p ∨ ¬q) for a while. But my being able to ‘just see’ that the
transformation holds can just as much be the result of your patient ex-
planation (along with that of the other direction), a result of your arguing
that it does, and my following along in a conscious reasoning process. In
such a circumstance, why should the value of my being able to just see this
depend on what took place just before?
If I really do just see it, my justification for believing the transformation
does not rest wholly on the support this receives in virtue of being the
conclusion of an argument. It rests in part on the fact that I just see it to be
so, that I have the intuition that it is so.
So: while a proposition which is the conclusion of an argument does
not have foundational justification if it is justified only because it follows
from premises that are also justified, it is no bar to its having foundational

6
justification in virtue of being the content of an intuition that what allows,
prompts, or causes the intuition to arise is a conscious reasoning process.
On the view of intuition mentioned above, for example, the question is
simply whether the right kind of experience obtains. If it does, what
brought this about is irrelevant. The same is true for other views. (Per-
haps knowledge about what brought about the experience in some cases
defeats the justification, as, for example, Peter Singer7 thinks is the case for
moral intuitions, but that is a separate matter.)
Some might try to save the negative etiological requirement by claim-
ing that two processes are at work here, one conscious reasoning process,
on the one hand, and another separate, non-conscious process on the other,
and that I count as having an intuition because my ‘just seeing’ that the
transformation holds is in part caused by the non-conscious process. (I am
grateful to an anonymous referee for raising this objection.)
We should acknowledge that there are such cases. But there are surely
also cases where the sole cause of my being able to ‘just see’ that the trans-
formation holds is the conscious reasoning process. Stipulating that the
case discussed above is one of these, the argument loses none of its force:
it is just as plausible as before that, because it really does seem to me that
the transformation holds, its so seeming provides me with foundational
justification for that belief.
To sum up: a restriction that intuition may not be the result of con-
scious reasoning is, in the absence of argument, ad hoc. The argument that
without the restriction, intuition cannot provide foundational justification,
fails. Absent further argument we therefore lack reason to think that there
is a class of mental states which deserves the label ‘intuition’, which con-
stitutes a psychological kind, but which cannot result from conscious rea-
soning.
7 Peter Singer, “Ethichs and Intuitions”, The Journal of Ethics 9 (2005), pp. 331–352.

7
Why It Matters
Why does any of this matter? For at least four reasons.
Most obviously, realising that intuition can result from conscious rea-
soning teaches us something about its nature, and provides a constraint on
theory. Someone who wishes to account for the nature of intuition must
also consider the intuitions which result from conscious reasoning, and
must ensure that her theory can capture these cases.
Second, the reasoning which shows that intuitions can result from rea-
soning also teaches us something about the structure of justification. A
conclusion, one might have thought, cannot provide foundational justifi-
cation. Not so, if what I have argued is correct: it can, albeit not in virtue
of being a conclusion.
Third, when a proposition c is the conclusion of an argument with
premises p1 , . . . , pn , it is usually thought that c cannot support further
conclusions d, e, . . . , independently of p1 , . . . , pn . That is, if the argument
is deductive and no other arguments support the conclusion, and bracket-
ing the possibility of a high prior credence in c, ones rational credence in
the most plausible of p1 , . . . , pn sets the upper bound for rational credence
both in c, and in d, e, . . . . But if what I have argued is correct, this is not
in general true. In each case we have to ask whether c is supported solely
because it follows from p1 , . . . , pn , or whether it is justified also because it
seems true to the agent. In the former case, one’s justification for c, d, and e
cannot exceed the justification for the most plausible of p1 , . . . , pn (holding
the assumptions fixed). But in the latter case it can. Suppose, for example,
that you have determined that a deductive argument from p1 , . . . , pn to
c supports a credence in the latter of .7, and that no other argument you
know of supports c. If you also have the intuition that c—prompted, per-
haps, precisely by the conscious reasoning process of going through the
argument—it may now be rational for you to have credence .85 in c, since
the intuition provides additional justification.

8
 Finally, the realisation that intuition can result from conscious reason-
ing suggests a possible reply to the experimental philosophy challenge to
the use of intuition in philosophy. Broadly speaking, the challenge is that
intuitions vary in ways that indicate, it is claimed, that they cannot be re-
liable guides to truth.8
If the above is correct, some intuitions result from conscious reasoning
and some do not. If, as seems likely, philosophers to a significant extent
use intuitions of the former type, and if, as I think is the case, the exper-
imental philosophy movement has almost exclusively probed those that
do not, the movement misses its target to that extent, because (in a slogan)
what they test is not what we use. It then remains possible that intuitions
that result from conscious reasoning are reliable guides to truth, even if
some of those that do not are not.

8 See, for example, Jonathan M. Weinberg, Shaun Nichols, and Stephen Stich, “Norma-

tivity and Epistemic Intuitions”, Philosophical Topics 29 (2001), pp. 429–460.