Second Philosophy.

PENELOPE MADDY
University of California, Irvine, USA
Perhaps some of the movie-goers among you have had the experience
of sitting through a film with no discemable plot and no significant
action, only to be accused by your companions of having missed the
point. 'It's not supposed to be dramatic', they tell you, 'it's a Character
Study!' The conventions of this genre seem to require that it centre on
an otherwise inconspicuous person who undergoes some familiar life
passage or other with terribly subtle, if any, reactions or results. Well,
if a thesis is to a philosophy talk what a plot is to a movie, I'm afraid
I'm about to inflict the counterpart to a Character Study: I'll introduce
a distinctive inquirer and record her progress through a particularly
venerable philosophical neighbourhood. My apologies in advance for
what will be more a saunter than a journey.
I. DESCARTES
To explain what 'Second Philosophy' is supposed to be, I should begin
with Descartes and his Meditations on First Philosophy (1641). The
key to this work is Descartes' dramatic Method of Doubt.' It starts
modestly enough, noting that our senses sometimes deceive us about
objects that are very small or very distant, but quickly moves on to
perceptual reports that seem beyond question, like my current belief
that 'this is a hand', Still, the meditator wonders, might I not be mad,
or asleep?
Yet at the moment my eyes are certainly wide awake ... as I stretch
out and feel my hand I do so deliberately, and I know what I am
doing. All this would not happen with such distinctness to one asleep.
Indeed! As if I did not remember other occasions when I have been
14 PENELOPE MADDY
tricked by exactly similar thoughts while asleep! As I think about
this more carefully, I see plainly that there are never any sure signs
by means of which being awake can be distinguished from being
asleep. The result is that I begin to feel dazed ... Perhaps ... I do not
even have ... hands ... at all. (Descartes [1641]. p. 13)
In his dizziness. the meditator anxiously grasps for a fixed point
.., whether I am awake or asleep, two and three added together are
fivc, and a square has no more than four sides. In seems impossible
that such transparent truths should incur any suspicion of being false.
(op. cit., p. 14)
But the midnight fears cannot be stopped. What if God is a deceiver
or worse, what if there is no God, and I am as I am by mere chance'
Mightn't I then be wrong in absolutely all my beliefs'?
I have no answer to these arguments, but am finally compelled to
admit that there is not one of my former beliefs about which a doubt
may not properly be raised. (op. cit., pp. 14-15)
And he concludes that
in future I must withhold my assent from thesc fomlcr beliefs just
as carefully as I would from obvious falsehoods. if I want to dis-
cover any certainty. ... I will supposc therefore that ... some mali-
cious demon of the utmost power and cunning has employed all his
energies in order to deceive me. J shall think that the sky. the air, the
earth, colours, shapes, sounds and all cxtcmal things are merely the
delusions of drcams which he has devised to ensnare my judgement.
I shall consider myself as not having hands or eyes, or flesh. or
blood or senses, but as falsely believing that I have all these things
... this is an arduous undertaking ... (op. cit., p. 15)
Arduous, indeed, to deny that I have hands, that I'm now standing here
giving this talk, that you are all sitting in yollr chairs, listening 3.." I
rehearse the familiar Cartesian catechism. We might fairly ask, what i~
the ooint of this difficult exercise?
.S'econd Phi/osoph.v
The point is not that I am somehow unjustified in believing these
things. Despite the doubts that have just been raised, Descartes and his
meditator continue to regard my ordinary beliefs as
highly probable ... opinions, which, despite the fact that they are in
a sense doubtful... it is still much more reasonable to believe than
to deny. (op. cit., p. 15)
The very reasonableness of these beliefs is what makes it so difficult
to suspend them. For this purpose, some exaggeration2 is needed:
I think it will be a good plan to turn my will in completely the
opposite direction and deceive myself. by pretending for a time that
these former opinions are utterly false and imaginary. (op. cit.)
So. the Evil Demon Hypothesis is designed to help to unseat my oth-
erwise reasonable beliefs, though the doubt raised thereby is 'a very
slight. and, so to speak, metaphysical one' (op. cit.. p. 25).
But this just pushes the question back one step. We now wonder:
why should I wish to unseat my otherwise reasonable beliefs? The
meditator is explicit on this point. He is concerned about the status of
natural science, and he realizes that
It [is] necessary, once in the course of my life, to demolish every-
thing completely and start again right from the foundations if I [want]
to establish anything at all in the sciences that [is] stable and likely
to last. (op. cit., p. 12)
The Method of Doubt, the suspension of belief in anything in any way
doubtful, is just that. a method-designed to lead us to a firm founda-
tion for the sciences:
I must withhold my assent from these former beliefs just as carefully
as I would from obvious falsehoods, ~r 1 want to discover any cer-
tainty in the .\'ciences. (op. cit., p. 15, emphasis mine, underlined
phrase from the 1647 French edition)
The hope is that once we set aside all our ordinary beliefs, reasonable
or not, some absolutely indubitable foundational beliefs will then emerge,
on the ba."is of which science and common sense can then be given a
PENElOP MADDY
fiffi1 foundation. The Method of Doubt is the one-time expedient that
enables us to carry out this difficult task.
Janet Broughton, the scholar whose account of Descartes J've been
following here, describes the meditator's situation like this:
Of course, there is nothing about the strategy of this [Method of
Doubt] that guarantees it will do what we want it to do. Perhaps we
will find that all claims can be impugned by a reason for doubt.
Perhaps we will find some that cannot, but then discover that they
are very general or have few interesting implications. (Broughton
[2002], p. 53)
But. as we all know, this is not the fate of Descartes' meditator. Jn the
second Meditation, he quickly establishes that he must exist---as he
must exist even for the Evil Demon to be deceiving him!-and that he
is a thinking thing. From there, he moves to the existence of a benevo-
lent God, the dependability of 'clear and distinct ideas', and so on,
returning at last to the reasonable beliefs of science and common sense.
Alas, a sad philosophical history demonstrates that the path leading
from the Evil Demon Hypothesis to hyperbolic doubt has always been
considerably more compelling than the route taken by the meditator
back to ordinary belief. Still, the Cartesian hope of securing an unas-
sailable foundation for science has pe~isted, down the centuries. So,
for example, the good Bishop Berkeley (1710) suggested that our sense
impressions are incontrovertible evidence for the existence of physical
objects, because such objects simply are collections of impressions,
but the price he paid-subjective idealism-was one nearly all but
Berkeley have found entirely too high. More recently, Russell (1914)
and early Carnap (1928, on some readings, anyway) applied the full
scope and power of modem mathematical logic to the project of con-
struing physical objects as more robust logical constructions from sen-
sory experiences, but both efforts ultimately failed, even in the opin-
ions of their autho~. There is surely much in this historical record-
both in the detail of each attempt and in the simple fact of this string
of failures-to lead us to despair of founding science and common
sense on some more trustworthy emanations Qf First Philosophy. Thus,
Ouine speaks of a 'forlorn hope' and a 'lost cause' ([1969]. p. 74).
11
Second Philosophy
But perhaps the situation is not as tragic as it is sometimes drawn.
Let's consider, for contrast, another inquirer, one entirely different from
Descartes' meditator. This inquirer is bOrn native to our contemporary
scientific world view; she practices the modem descendants of the
methods found wanting by Descartes. She begins from common sense,
she trusts her perceptions, subject to correction, but her curiosity pushes
her beyond these to careful and precise observation, to deliberate
Av...~rimentation, to the formulation and stringent testing of hypotheses,
to devising ever more comprehensive theories, all in the interest of
learning more about what the world is like. She rejects authority and
tradition as evidence, she works to minimize prejudices and subjective
factors that might skew her invest4gations. Along the way, observing
the forms of her most successful theories, she develops higher level
principles-like the maxim that physical phenomena should be explained
in terms of forces acting on a line between two bodies, depending only
on the distance between them-and she puts these higher level principles
to the test, modifying them as need be, in light of further experience.
Likewise, she is always on the alert to improve her methods of
observation, of experimental design, of theory testing, and so on,
undertaking to improve her methods as she goes.
We philosophers, speaking of her in the third person, will say that
such an inquirer operates 'within science', that she uses 'the methods
of science', but she herself has no need of such talk. When asked why
she believes that water is H2O, she cites information about its behav-
iour under electrolysis and so on; she doesn't say, 'because science
says so and I believe what science says'. Likewise, when confronted
with the claims of astrology and such like, she doesn't say, 'these
studies ar~ ;unscientific'; she reacts in the spirit of this passage from
Feynman ~n astrology:
Maybe it's ... true, yes. On the other hand, there's an awful lot of
information that indicates that it isn't true. Because we have a lot of
knowledge about how things work, what people are, what the world
is, what those stars are, what the planets are that you are looking at,
what-makes them go around more or less ... iAnd furthermore, if you
look very carefully at the different astrologers they don't agree with
each other, so what are you going to do? Disbelieve it. There's no
PENELOPE MADDY
78
evidence at all for it. ... unless someone can demonstrate it to you
with a real experiment, with a real test ... then there's no point in
listening to them. (Feynman [1998], pp. 92-3)
My point is that o.ur inquirer needn't employ any general analysis of
what counts as 'scientific' to say this sort of thing, though we philoso-
phers use the tenn 'science' in its rough and ready sense when we set
out to describe how she behaves.
This, then, is the Character of our Character Study, a mundane and
unremarkable figure, as the genre dictates. Following convention, we
hope to tease out the hidden elements of her temperament by tracing
her reactions to a familiar philosophical test: the confrontation with
scepticism. So, how will she react to the challenge Descartes puts to his
meditator? Does she know that she has hands?
In response to this question, our inquirer will tell a story about the
workings of perception-about the structure of ordinary physical objects
like hands, about the nature of light and reflection, about the reactions
of retinas and neurons, the actions of human cognitive mechanisms,
and so on. This story will include cautionary chapters, abOut how this
nonnally reliable train of perceptual events can be undennined-by
unusual lighting, by unusual substances in the blood-stream of the
perceiver, and so on-and she will check as best she can to see that
such distorting forces are not present in her current situation. By such
careful steps she might well conclude that it is reasonable for her to
believe, on the basis of her perception, that there is a hand before her.
Given that it is reasonable for her to believe this, she does believe it.
and so she concludes that she knows there is a hand bef;>re her, that she
has hands.
But mightn't she be sleeping? Mightn't an Evil Demon be deceiving
her in all this? Our inquirer is no more impressed' by these empty
possibilities than Descartes' meditator; with him, she continues to think
it is far more reasonable than not for her to believe that she has hands,
that she isn't dreaming, that there is no Evil Demon. The question is
whether or not she will see the wisdom, as he does, in employing the
Method of Doubt. Will she see the need 'once in [her] life, to demolish
everything completely and start again' (Descartes [1641], p. l2)?
79
Second Philosophy
This question immediately raises another, which we haven't so far
considered, namely, what is it exactly that Descartes' meditator sees as
forcing him to this drastic course of action? The only answer in the
Meditations comes in the very first sentence:
Some years ago I was struck by the large number offalsehoods that
I had accepted as true in my childhood, and by the highly doubtful
nature of the whole edifice that I had subsequently based on them.
(Descartes [1641], p. 12)
Our inquirer will agree that many of her childhood beliefs were false,
and that the judgements of common sense often need tempering or
adjustment in light of further investigation, but she will hardly see
these as reasons to suspend her use of the very methods that allowed
her to uncover those errors and make the required corrections! It's hard
to see why the meditator feels differently.
The reason traces to Descartes' aim of replacing the reigning Scholas-
tic Aristotleanism with his own Mechanistic Corpuscularism. As he
was composing the Replies that were to be published with the first
edition of the Meditations, he wrote to Mersenne:
I may tell you, between ourselves, that these six Meditations contain
all the foundations of my physics. But please do not tell people, for
that might make it harder for supporters of Aristotle to approve
them. I hope that readers will gradually get used to my principles,
and recognize their truth, before they notice that they destroy the
principles of Aristotle. (Descartes [1641a], p. 173)
To get a sense of the conflict here, notice that on the view Descartes
comes to by the end of the Meditations, all properties of physical
objects are to be explained in terms of the geometry and motions of the
particles that make them up; the features we experience-like color,
weight, warmth, and so on-exist, strictly speaking, only in us. For the
Aristotelians, in contrast, physical objects themselves have a wide variety
of qualities, which brings Aristotelianism into close alliance with com-
mon sense.
This background is laid out beautifully by Daniel Garber, who thet:l
takes the final step:
80 PENELOPE MADDY
Descartes thought [that] the common sense worldview and the Scho-
lastic metaphysics it gives rise to is a consequence of one of the
universal afflictions of humankind: childhood. (Garber [1986], p.
88)
On Descartes' understanding of cognitive development, children are 'so
immersed in the body' (Descartes [1644], p. 208) that they fail to
distinguish mind and reason from matter and sensation, and
The domination of the mind by the corporeal faculties ... leads us to
the unfounded prejudice that those faculties represent to us the way
the world really is. (Garber [1986], p. 89)
So these are the 'childhood falsehoods' and Aristotelianism is the re-
sulting 'highly doubtful edifice' that the meditator despairs of in the
opening sentence of the Meditations.) As the errors of childhood are
extremely difficult to uproot in adulthood, only the Method of Doubt
will deliver a slate clean enough to allow Descartes' alternative to
emerge: the resulting principles of First Philosophy will be completely
indubitable, and as such, strong enough to undermine the authority of
common sense.4
Now our contemporary inquirer, unlike the meditator, has no such
Cartesian reasons to believe that her most reasonable beliefs are prob-
lematic,5 so she lacks his motivation for adopting the Method of Doubt.
Still, if application of the Method does lead to First Philosophical
Principles that are absolutely certain, principles that may cOnflict with
some of our inquirer's overwhelmingly reasonable, but ever-so-slightly
dubitable beliefs, then she should, by her own lights, follow this course.
Even if all her old beliefs re-emerge at the end, some of them might
inherit the certainty of First Philosophy.6 Though she quite reasonably
regards such outcomes as highly unlikely, she might well think it proper
procedure to read past the First Meditation, to see what comes next.
The unconvincing arguments that follow will quickly confirm her ex-
pectation that there is no gain to be found in this direction!
So our inquirer will continue her investigation of the world in her
familiar ways, despite her encounter with Descartes and his meditator.
She will ask traditional philosophical questions about what there is and
how we know it, just as they do, but she will take perception as a
Second Philosophy
81
mostly reliable guide to the existence of medium-sized physical
objects, she will consult her astronomical observations and theories to
weigh the existence of black holes, and she will b'eat questions of
knowledge as involving the relations between the world-as she un-
derstands it in her physics, chemistry, optics, geology, and so on-and
human beings-as she understands them in her physiology, cognitive
science, neuroscience, linguistics, and so on, While Descartes' medita-
tor begins by rejecting science and common sense in the hope of found-
ing them more firmly by philosophical means, our inquirer proceeds
scientifically, and attempts to answer even philosophical questions by
appeal to its resources, For Descartes' meditator, philosophy comes
first; for our inquirer, it comes second-hence 'Second Philosophy' as
opposed to 'First', Our Character now has a label: she is the Second
Philosopher ,8
D. STROUD'S DESCARTES
The Descartes we've been examining so far-perhaps he should be
called Broughton's Descartes-regards the sceptical hypotheses as a
mere tool in his search for a new foundation for science,' but contem-
porary epistemologists tend to entertain a more potent scepticism that
takes centre stage all on its own. To see how our Second Philosopher
fares in this context, let's turn our attention to this Descartes, of whom
Barry Stroud writes:
By the end of his First Meditation Descartes finds that he has no
good reason to believe anything about the world around him and
therefore that he can know nothing of the external world. (Stroud
[1984], p. 4)
This Descartes would seem to stand in clear conflict with common
sense, and with our Second Philosopher.
Stroud's analysis brings us back to the dream argument. The medi-
tator realizes that the senses sometimes mislead him, when the light is
bad, or he is tired, and so on, so he focuses on a best possible case: he
sits comfortably by the fire with a piece of paper in his hand. At first,
it seems to him impossible that he could be wrong about this-until
he's hit by the thought that for all be knows he might be dreaming.
82 PENELOPE MADDY
'With this thought,' Stroud writes, 'Descartes has lost the whole world'
(Stroud [1984], p. 12).
At this point, the Second Philosopher is tempted to answer in the
spirit displayed by Descartes himself at the end of th~ Meditations:
The exaggerated doubts ... should be dismissed as laughable ... es-
pecially .., my inability to distinguish between being asleep and
being awake .,. there is a vast difference between the two, in that
dreams are never linked by memory with all the other actions of life
as waking experiences are ... when I distinctly see where things
come from and where and when they come to me, and when I can
connect my perceptions of them with the whole of the rest of my life
without a break, then I am quite certain that when I encounter these
things 1 am not asleep but awake. (Descartes [1641], pp. 61-2)10
.
But the trouble, says Stroud's Descartes, is that 1 might be dreaming
that 1 distinctly see where the paper in my hand came from, I might be
dreaming that my current perception of my hand is connected with the
rest of my life without a break, and so on. If I think there is some test
I can apply to determine whether or not my current experience is or
isn't a dream, I might be dreaming that the test is satisfied-1 might
even be dreaming that this test is effective! J I
To this, the Second Philosopher might reply that she knows, by
ordinary means, that she is not dreaming, just as Descartes suggests:
her tests centre on doing things now that she can't do while dreaming;
her belief that she can't do them while dreaming is based on her past
dreaming experiences, and so on . Surely we do, in fact, operate in this
way. But even if such an adherence to everyday methods could be
maintained to rule out the possibility of dreaming, it would be of no
use against the Evil Demon hypothesis, for which there cannot in prin-
ciple be any ordinary tests: the Demon makes it seem to me exactly as
it would if there were no demon, so no aspect of my experience could
count against his existence. This suggests that the debate over ordinary
tests is beside the point, so let's leave this style of response aside, and
continue with Stroud's line of thought, assuming the dream hypothesis
to be as impregnable as that of the Evil Demon.12
83
Second Philosophy
So, Stroud's meditator reasons like this. First, if I'm now dreaming,
I can't know there's a hand before me, (and this is true, by the way,
even if there happens to be a hand there). Second, I can't tell whether
or not I'm dreaming-which the Second Philosopher must admit, given
the recently-adopted strong sense of the dream hypothesis. From which
it is to follow that I can't know that there's a hand before me. But I
chose the case of the hand to give perception the greatest possible
advantage; if I don't know in this case, I can't know anything at all on
the basis of perception. And so, as Stroud puts it, I have lost the world.
The remaining kink here is the unspoken assumption that I can't
know the hand is there if I can't rule out the possibility that I'm dream-
ing. This is hardly required in ordinary life, even under the most rig-
orous conditions, as Stroud appreciates:
If I testify on the witness stand that I spent the day with the defend-
ant, that I went to the museum and then had dinner with him, and
left him about midnight, my testimony under normal circumstances
would not be affected in any way by my inability to answer if the
prosecutor were then to ask 'How do you know you didn't dream the
whole thing?' The question is outrageous. ... Nor do we ever expect
to find a careful report of the procedures and results of an elaborate
experiment in chemistry followed by an account of how the experi-
menter determined that he was not simply dreaming that he was
conducting the experiment. (Stroud [1984], pp. 49-50)
The worry arises that Stroud's Descartes is simply imposing an artifi-
cially high standard on knowledge, a'standard we don't in fact consider
reasonable. If this is right, then I could be said, under the proper per-
ceptual conditions, to know that there is a hand before me, even if I
can't prove that I'm not dreaming.
Understandably, Stroud is keen to argue that his Descartes isn't sim-
ply changing the subject from knowledge to some kind of ultra-knowl-
edge. He begins his response by pointing out that its being inappropri-
ate to criticize the witness's or the chemist's knowledge claims in this
way doesn't by itself show that ruling out the dream hypothesis isn't
necessary for knowledge:
84 PENELOPE MADDY
The inappropriately-asserted objection to the knowledge-claim might
not be an outrageous violation of the conditions of knowledge, but
rather an outrageous violation of the conditions for the appropriate
assessment and acceptance of assertions of knowledge. (Stroud
[1984], p. 60)
The witness and the chemist make their claims to know!edge 'on just
about the most favorable grounds one can have for claiming to know
things' (op. cit., p. 61), so it isn't appropriate to criticize them for
failing to rule out, or even to consider, the possibility that they're
dreaming. But this doesn't show that they do in fact know what they
claim to know.
Having found this opening, Stroud's Descartes takes it: he thinks it's
appropriate for me to assert that I know when there is no reason to
think I might be dreaming, but that I still do not in fact know unless
I can rule out that possibility. The reason for this discrepancy between
conditions for knowledge assertions and conditions for knowledge lies
in the contrast between the practical and the theoretical:
It would be silly to stand for a long time in a quickly filling bus
trying to decide on the absolutely best place to sit. Since sitting
somewhere in the bus is better than standing, although admittedly
not as good as sitting in the best of all possible seats, the best thing
to do is to sit down quickly... there is no general answer to the
question of how certain we should be before we act, or what possi-
bilities of failure we should be sure to eliminate before doing some-
thing. It will vary from case to case, and in each case it will depend
on how serious it would be if the act failed, how important it is for
it to succeed by a certain time, how it fares in competition on these
and other grounds with alternative actions which might be performed
instead, and so on. This holds just as much for the action of saying
something, or saying that you know something, or ruling out certain
possibilities before saying dtat you know something, as for other
kinds of actions. (Stroud [1984]. pp. 65-6)13
The picture, then, is of a sliding scale of strictness on proper assertions
of knowledge:
Second Philosophy 8S
From the detached point of view-when only the question of whether
we know is at issue--our interests and assertions in everyday life are
seen as restricted in certain ways. Certain possibilities are not even
considered, let alone eliminated, certain assumptions are shared and
taken for granted and so not examined. (Stroud [1984], pp. 71-2)
In ordinary life, then, under good perceptual conditions, it's reasonable
for me to risk asserting that there is a hand before me, and so, to claim
to know that there is a hand before me. But the knowledge claim is just
a loose way of speaking, for practical purposes. In a theoretical con-
text--one without practical time pressures, with no limit on the amount
of 'effort and ingenuity' (op. cit., p. 66) we can bring to bear on the
question of the truth of our cla1ms-in such a context, free of practical
restrictions, we have no excuse for speaking loosely and we shouldn't
claim to know until we have ruled out every possibility that would
preclude our knowing-in particular, we must rule out the possibility
that we are dreaming. So Stroud's Descartes hasn't changed the sub-
ject; he's simply working with the usual notion of knowledge in an
unrestricted or theoretical context. I.
Now there is considerable appeal in this notion of a sliding scale of
stringency. The Second Philosopher imagines a shopkeeper concerned
about the coins he takes in: are they pure metal or fakes? 15 He instructs
his hired assistant to bite each coin to be sure, knowing that many
counterfeits are laced with harder metals. He also knows that more
sophisticated counterfeiters produce fake coins with softness compara-
ble to pure coins by a different. more difficult process, and that these
finer fakes can be detected by an optical device he keeps in the back
of his shop. But the fellows capable of this fine work are now in jail,
so he doesn't bother to include this extra twist in his instructions to his
assistant. Under these conditions, when the assistant says he knows a
particular coin is pure metal, the shopkeeper realizes that the fellow
doesn't really know, because he hasn't used the optical device in the
back room and doesn't know that the coin isn't one of the finer fakes,
but the knowledge claim is appropriate in the context, and the shop-
keeper would be out of line to correct him.
likewise, the chemist knows that there are impure metals that pass
both the biting test and the optical test, so he can see that the~
86 PENELOPB MADDY
shopkeeper's claim to know, after using his optical device, is also
restricted, despite being appropriate in the given circumstances. Even
the chemist's claim to know that the metal is pure will appear restricted
to the physicist who realizes that there are atomic variations undetectable
by chemical means. And even the physicist may have to admit that
there are possible variations he doesn't yet know how to test for, and
he will always realize that there may be possibilities he's unaware of
that will be uncovered by future scientists. So, even his claim to know
that the metal is pure will be subject to the proviso, 'at least as far as
current science can determine'.
All this gives the idea of a sliding scale of restrictiveness some
initial plausibility. But the Second Philosopher remains "troubled by the
conclusion that such a scale somehow presupposes an underlying entirely
unrestricted notion of the sort proposed by Stroud's Descartes. After
all, the scale as she understands it. no matter how stringent it gets, will
never require ruling out the hypothesis that the relevant inquirer is
dreaming, or deceived by an Evil Demon; these doubts still seem
artificial. Furthermore, no notion of a scale seems relevant in a simple
perceptual case like my seeing my hand before me; there I'm not
hampered by time pressure or ignorance or anything else, no further,
more strenuous investigation or special expertise seems relevant.
From the Sccond Philosopher's point of view, the situation looks
like this. She has various methods of finding out about the world,
beginning with observation, and as she builds and tests and modifies
her theories, she also studies, tests and refines thosc methods them-
selves. She has seen, in her day. implementations of various bad pro-
cedures for finding out about the world, like astrology and creationism,
and she can explain in detail what's wrong with these methods. Now
Stroud's Descartes present.. her with an alternative hypothesis: perhaps
everything she believes is false and she is dreaming, or an Evil Demon
has made it seem to her as if what she thinks she knows is true when
it is not. This alternative hypothesis is deliberately designed so that
none of her tried and true methods can be brought to bear on it. I
imagine she will reply along these lines: 'I admit I can't refute the
hypothesis, though it's hard for me to see the point of entertaining it. 1ft
Perhaps this shows that I can't be absolutely certain that I know what
Second Philosophy 87
I think I know, but that doesn't surprise me so much. 17 I constantly
work to remove as many "restrictions" as possible, to conduct my
inquiries in a detached and unhurried way, as unimpeded as possible
by practical limitations and lingering prejudices. This seems to me the
best way there is to find out what the world is like. The semantics of
the word "know" seem to me quite complex, and I don't pretend to
understand them completely,IS but it still seems to me reasonable to
think that we know, in a straightforward and unrestricted sense, that we
have hands (in which case scaling seems irrelevant), that water is ~O
(in which case we seem suitably high on the scale), and much, much
more.'
This will hardly satisfy Stroud's Descartes, but to avoid an unap-
pealing debate over the concept 'know',19 let me approach the issue
from another direction. Stroud calls the epistemological challenge a
'theoretical' or 'philosophical' one:
We aspire in philosophy to see ourselves as knowing all or most of
the things we think we know and to understand how all that knowl-
edge is possible. (Stroud [1994], p. 296)
The Second Philosopher thinks she has at least the beginnings of an
answer to this question, in her account of how and when perception is
a reliable guide, in her study of various methods of reasoning, and her
efforts to understand and improve them.
But this obviously isn't what Stroud has in mind:
In philosophy we want to understand how any knowledge of an
independent world is gained on any... occasions ... through sense-
perception. So, unlike ... everyday cases, when we understand the
particular case [like my hands] in the way we must understand it for
philosophical purposes, we cannot appeal to some piece of knowl-
edge we think we have already got about an independent world.
(Stroud [1996], p. 132)20
From the 'philosophical' or 'external' point of view:
A II of my knowledge of the external world is supposed to have been
brought into question in one fell swoop ... I am to focus on my
relation to the whole body of beliefs which I take to be knowledge
88
PENELOPE MADD~
whether
of the external world and ask, from 'outside' as it were
and how I know it ... (Stroud [1984], p. 118)
In other words, I'm to set aside all my hard-won methods, all my
carefully checked and double-checked beliefs, and then explain ... Well,
the Second Philosopher will hardly care what she's now asked to ex-
plain; the demand that she explain anything without using any of her
best methods seems barmy.
From Stroud's point of view, the problem with the Second Philoso-
pher's explanation can be illuminated by a comparison.21 Suppose a
pseudo-Cartesian inquirer gives the following account of his knowl-
edge of the world: 'I know because I have a clear and distinct idea, and
God makes sure that I only have clear and distinct ideas about things
that are true; furthermore, I came to believe this about God by means
of clear and distinct ideas, so I have good reason to believe I am right.'
This account is to run parallel to the Second Philosopher's: 'I know
because my belief is generated by such-and-such methods, and such-
and-such methods are reliable; furthermore, I came to believe that they
are reliable by means of such-and-such methods, so I have good reason
to believe that I'm right.' We may be inclined to think that the Second
Philosopher is right- that perception and her other methods of belief
formation are reliable-and that the pseudo-Cartesian is wrong-that
there is no such accommodating God-but the best either of these
inquirers can say is:
'If the theory I hold is true, I do know... that I know... it, and I do
understand how I know the things I do.' (Stroud [1994], p. 30 I)
Given that all knowledge is being called into question at once, neither
of them can detach the antecedent, so neither can give a philosophi-
cally satisfying account of their knowledge.
In fact. this just repeats the previous observation that the Second
Philosopher can't explain her knowledge without using her methods of
explanation, but the rhetorical force is heightened by the suggestion
that she's in no better position than this woeful pseudo-Cartesian. Of
course, she doesn't see it that way; to her, the pseudo-Cartesian is just
another in a long line of the benighted-like the astrologer and the
creationist-all of whom she can dispatch on straightforward grounds.
Second Philosophy 89
What Stroud's comparison invites her to attempt is an explanation of
the pseudo-Cartesian' s errors that uses none of her methods, a task that
seems to her no more reasonable than the original challenge to explain
her knowledge using none of her methods.
Perhaps the Second Philosopher's reaction can be clarified by com-
parison with a few close cousins. Moore, like the Second Philosopher,
tends to stick to the 'internal' or 'everyday' versions of the sceptic's
questions. Stroud writes:
It is precisely Moore's refusal or inability to take his own or anyone
else's words in [the] 'external' or 'philosophical' way that seems to
me to constitute the philosophical importance of his remarks. He
~teadfastly remains within the familiar, unproblematic understanding
of those general questions and assertions with which the philosopher
would attempt to bring all our knowledge of the world into question.
He resists, or more probably does not even feel, the pressure towards
the philosophical project as it is understood by the philosophers he
discussess22 ... But how could Moore show no signs of acknowledg-
ing that [those questions] are even intended to be taken in a special
'external' way derived from the Cartesian project of assessing all our
knowledge of the external world at f'T'I.'C? That is the question about
the mind of G.E. Moore that ( canll(It .~nswer. (Stroud [1984], pp.
119,125-6)
Here the Second Philosopher must sympathize with Stroud. Though
she, too, fails to feel the 'lure' of the philosophical project, she surely
realizes that those who do feel it intend the question of the external
world to be understood in a sense that explicitly marks 9ff everything
she has to offer as beside the point. For this reason, she, unlike Moore,
cannot honestly claim to have answered the sceptic's challenge.
Quine's naturalized epistemology is another obvious relative, but the
Quinean opus includes many themes, some of which seem to conflict
with his naturalism, and many statements and restatements, so the
assessment of agreement and disagreement here is an arduous one.23
Still, one note is salient for our purposes here. Quine poses the epis-
temolo.2ical challenge this way:
PENELOPE MADDY
90
We are studying how the human subject ... ... from his
data. (Quine [1969], p. 83)
The Second Philosopher, relying more on cognitive science than on
empiricism or behaviourism, is less inclined to speak of 'data' and
'positing' and more inclined to cite studies of how prelinguistic infants
come to perceive and represent physical objects}4
This brings us to one of Stroud's central concerns about Quine's
naturalism. He is unbothered by the idea of relying on science, which
he sees as an update of Moore's dogged trust in common sense:
What Moore says is perfectly legitimate and unassailable ... The
results of an independently-pursued scientific explanation of knowl-
edge would be in the same boat. (Stroud [1984], p. 230)
As we've seen, Stroud thinks 'there is wisdom is that strategy' ([1984],
p. 248), though it doesn't answer the sceptic's challenge as he under-
stands it. The trouble comes in Quine's distinctive conception of the
scientific undertaking, in his description of humans as positing objects
on the basis of data, where what
posits bodies
... can be said .,. in common-sense tenns about ordinary things are
... far in excess of any available data. (Quine [1960], p. 22)
For Quine, the naturalized epistemologist studies
the relation between the meager input and the torrential output ... in
order to see how evidence relates to theory, and in what ways one's
theory of nature transcends any available evidence. (Quine [1969],
p.83)
Though Quine doesn't begin from a sensory given, but from 'the lim-
ited impingements' of our sensory surfaces (Quine [1974], p. 3), he
persistc; in the language of 'evidence', 'information' and 'data'.
Stroud worries that this way of describing the scientific project pro-
vides a new foothold for the sceptic. If I regard my beliefs about the
external world as the result of my own positing, a positing that could
have gone any number of different ways without coming into conflict
with my sensory inputs, it's hard to see how I can use those beliefs to
explain how I come to know what the world is like:
~
Second Philosoph)' 91
Countless 'hypotheses' or 'theories' could be 'projected' from those
same slender' data', so if we happen to accept one such 'theory' over
others it cannot be because of any objective superiority it enjoys
over possible or actual competitors ... our continued adherence to
our present 'theory' could be explained only by appeal to some fea-
ture or other of the knowing subjects rather than of the world they
claim to know. And that is precisely what the traditional epistemolo-
gist has always seen as undermining our knowledge of the external
world. (Stroud [1984], p. 248)
Though Quine hopes to use ordinary science in his epistemological
project, the project itself is formed by 'the old epistemologist's prob-
lem of bridging a gap between sense data and bodies'; it is 'an enlight-
ened persistence ... in the original epistemological problem' (Quine
[1974], pp. 2-3). Stroud's point is that Quine's enlightenment has not
saved him; as soon as he allows
a completely general distinction between everything we get through
the senses, on the one hand, and what is or is not true of the external
world. on the other ...
he is 'cut ... off forever from knowledge of the world around us' (Stroud
[1984], p. 248).25 As Quine's line on positing and underdetermination
is supposed have resulted from scientific inquiry, Quine's science has
undermined itself from within.26
But the commitment to science and common sense doesn't force us
to conceive the problem in Quine's way. In fact, if we are interested in
explaining how a causal process beginning with light falling on and
reflecting off an object, continuing through stimulations of our sensory
surfaces, proceeding through various levels of cognitive processing,
often results in reliable belief about the external world, we find nothing
in the story about 'data' or 'theory', no grounds for identifying,one
episode in the causal chain-the 'irritation' of our 'physical receptors'-
as data or information or evidence that radically underdetermines the
rest}' Ironically, Quine himself, at other times, counsels us to drop
such talk of 'epistemological priority' (see Quine [1969], p. 85), but if
we do so
92 PENELOPE MADDY
We are left with questions about a series of physical events, and
perhaps with questions about how those events bring it about that we
believe what we do about the world around us. But in trying to
answer these questions we will not be pursuing in an 'enlightened'
scientific way a study of the relation between 'observation' and 'sci-
entific theory' or of the 'ways one's theory of nature transcends any
available evidence' ... (Stroud [1984], p. 252)
Any suggestion that we are addressing the sceptic's original challenge
now evaporates.
Of course, the Second Philosopher never embraced this Quinean
conception of the project in the first place: she isn't out to explain how
we project or infer objects from sensory data, but how we come to bc
able to detect external objects by sensory means. Stroud has no objec-
tion to this project of the Second Philosopher, or for that matter, to
Moore's persistence in the everyday or 'internal' reading of the ques-
tion; in fact, it's hard not to be struck by Stroud's admiration for Moore,
in particular. Speaking for himself rather than Descartes, Stroud won-
ders whether
the fully 'external' or 'philosophical' conception of our relation to
the world, when pressed, is really an illusion. (Stroud [1984], pp.
273-4)
His distrust extends not only to scepticism, but to all efforts to answer
the 'external' question:
It is what all such theories purport to be about, and what we expect
or demand that any such theory should say about the human condi-
tion that we should be examining, not just which one of them comes
in first in the traditional epistemological sweepstake. In that tough
competition, it still seems to me, skepticism will always win going
away. (Stroud [1994], p. 303)
Here Stroud and the Second Philosopher come into partial agreement,
in their reservations about the very problem of traditional epistemol-
ogy, the problem of justifying our knowledge of the world without
using any of our ordinary means of justification. The difference is that
Stroud suspects the problem is somehow incoherent-that there's some
93 Second Philosophy
obstacle in principle to posing the completely general question--while
the Second Philosopher thinks it's simply misguided.28
III. VAN FRAASSEN
In any case, scepticism of this traditional variety seldom troubles the
sleep of our level-headed philosophers of science. Still, they do worry
over some partial versions, most often scepticism about unobservables.
Van Fraassen's constructive empiricism is a conspicuous example. I'd
like to take my Second Philosopher on one last ramble, around this
comer of the philosophy of science. For future reference, let me begin
by summarizing the development of the Second Philosopher's reasons
for believing that there are atoms, despite her inability to see them,
with or without her eye-glaSses.
Beginning with Dalton in the early 1800s, the atomic hypothesis was
used in chemistry to explain various laws of proportion and combining
volumes, then Boyle's law and Charles' law, and was elaborated to
good effect with the notions of isomers, substitution and valence.29 For
a time, difficulties in the determination of atomic weights produced
severe worries-one prominent chemist proposed that the atom be
'erased' from science because 'it goes beyond experience, and never in
chemistry ought we to go beyond experience' (Dumas}-but by 1860,
all this was settled and, in the words of the historian, 'the atom [came]
into general acceptance as the fundamental unit of chemistry' (Ihde). In
the second hal f of the 19th century. atoms entered physics by way of
kinetic theory, again with dramatic success, including the determina-
tion of absolute atomic weights.
Despite all this, scientists concerned about atoms still asked 'who
has ever seen a gas molecule or an atom?' (Bertholet, 1877) and com-
plained about the appeal to 'forces, the existence of which we cannot
demonstrate. acting between atoms we cannot see' (Ostwald, 1895).
Even supporters recognized that it might well be thought 'more danger-
ous than useful to employ a hypothesis deemed incapable of verifica-
tion' and that scepticism here was 'legitimate' (Perrin). In this climate,
a leading textbook of 1904 warned that
94
PENELOPE MADDY
The atomic hypothesis has proved to be an exceedingly useful aid to
instruction and investigation ... One must not, however, be led astray
by this agreem~nt between picture and reality and combine the two.
(Ostwald, 1904)
In one of his remarkable series of papers of 1905, Einstein declared his
'major aim' was to 'find facts which would guarantee as much as pos-
sible the existence of atoms'.
A decade before, Gouy had argued that the phenomenon of Brownian
motion 'places under the eyes the realization of a" [the] hypotheses' of
the kinetic theorists. Einstein was unaware of this work and only dimly
aware of Brownian motion itself, but he concluded that:
According to the molecular-kinetic theory of heat, bodies of micro-
scopically visible size suspended in a liquid will perform move.
ments of such magnitude that they can be easily observed in micro-
scope.
Einstein took the presence or absence of this phenomenon in the exact
mathematical terms predicted by kinetic theory as a crucial test, but he
apparently thought actual experiments would require a level of preci-
sion beyond human reach.
But it was not beyond Jean Perrin, as we now know. He manufac-
tured tiny particles of exact and uniform size and weight and studied
how they suspended in a liquid-balancing the scattering forces of
Brownian motion against gravity. In this way, he used
the weight of the particle, which is measurable, as an intermediary
or connecting link between masses on our usual scaie of magnitude
and the masses of molecules ...
... and he obtained measurements of absolute atomic weights and
Avogadro's number that matched the predictions of kinetic theory. From
this beginning, he went on to verify the rest of Einstein's predictions.
In 1908, Ostwald, the same chemist who issued the textbook warn-
ing, described the work of Perrin and others as constituting
experimental proof for the discrete or particulate nature of matter-
proof which the atomic theory has vainly sought for a hundred years
Second Philosophy 95
Poincare, another former sceptic, writes:
we no longer have need of the infmitely subtle eye of Maxwell's
demon; our microscope suffices us .., atoms are no longer a useful
fiction... The atom of the chemist is now a reality.
The contemporary Second Philosopher agrees, on these grounds and
others that have accumulated since.
Now van Fraassen takes a different view, stunning in its sweep:
When the theory has implications about what is not observable, the
evidence does not warrant the conclusion that it is true. (van Fraassen
[1980], p. 71)
The sting of this denial is temporarily drawn by van Fraassen's admis-
sion that the evidence doesn't even warrant belief in a 'simple percep-
tual judgement', not because the evidence is sense-data-this way lies
traditional scepticism-but presumably because such a judgement in-
volves belief about matters I haven't actually observed (like the other
side of the moon, and so on). Indeed, van Fraassen holds that there are
no 'rationally compelling' reasons for extending ones belief beyond the
evidence precisely as far as he advocates and no further ([ 1980], pp.
72-3). He draws the line where he does following the lead of his
underlying Empiricism, which counsels him 'to withhold belief in
anything that goes beyond the actual, observable phenomena' ([1980],
p. 202).
Parallel to Stroud's Descartes on the subject of hands, van Fraassen
thinks that no evidence whatsoever could rationally compel belief in
atoms. Still, the Second Philosopher .is somewhat heartened, because
she finds here no general sceptical argument-corresponding to the
dream hypothesis-that's intended to establish this. Perhaps van Fraassen
is simply in the position of the chemists and physicists of 1900-
imagining, with some reason, that the existence of atoms 'goes beyond
experience' and is 'incapable of verification'. Perhaps he can be per-
suaded. as they were, by Perrin's experimental evidence and its like. If
he is not persuaded, she is eager to hear his objections to that evidence:
perhaps there is a misunderstanding she can clear up; perhaps there is
some weakness she hasn't noticed!
96 PENELOPE MADDY
To her surprise, van Fraassen's reactions don't seem to take this
form. Instead, he presents a range of arguments against people who
connect the truth of a scientific theory in one way or another to its
explanatory power, or who think the terms of mature scientific theories
typically refer, or whatever. Now she is disinclined to think a theory
true simply because it is the best explanation of the phenomena: the
atomic hypothesis gave an excellent account of a wide range of chemi-
cal and physical phenomena by 1900, but the existence of atoms still
hadn't been established; of course, their existence explains Brownian
motion, but this bare description of the situation leaves out the details
that made Perrin's experiments so compelling. Furthermore, the 'matu-
rity' of the theory in which an entity appears seems to her an unclear
and oddly-chosen indicator of the existence of its objects; what matters
is the particular experimental evidence available for the particular en-
tity in question! There is more, of course, but the nature and source of
the Second Philosopher's befuddlement should be clear.3O
Here van Fraassen makes the helpful suggestion that she has misun-
derstood the terms of the debate. The Second Philosopher is speaking
as one 'totally immersed in the scientific world-picture' (van Fraassen
[1980], p. 80). From this point of view,
the distinction between [atom]31 andflying horse is as clear as be-
tween racehorse and flying horse: the first corresponds to something
in the actual world, and the other does not. While immersed in the
theory, and addressing oneself solely to the problems in the domain
of the theory, this objectivity of [atom] is not and cannot be quali-
fied. (van Fraassen [1980], p. 82)
So, while the Second Philosopher is immersed in atomic theory, the
Perrin experiments do provide compelling reason to classify atoms as
real, as opposed, say, to phlogiston or whatever. But she has not yet,
according to van Fraassen, taken an 'epistemic' stance. He writes of the
working scientist:
If he describes his own epistemic commitment, he is stepping back
for a moment, and saying something like: the theory entails that
[atoms] exist, and not all theories do, and my epistemic attitude
towards this theory is X. (op. cit.)
I
Second Philosophy 97
To grasp what's at stake in van Fraassen' empiricism, the Second
Philosopher must step back and adopt an epistemic stance. And, at that
level, van Fraassen's empiricism counsels belief in the empirical ad-
equacy of the theory-that is, belief in what the theory tells us about
observable events and things-rather than belief in its truth.
The Second Philosopher imagines that she understands the distinc-
tion between immersion and an epistemic stance. It seems to be what
Ostwald recommended in his 1904 textbook: use atoms all you want
while you're doing your chemistry, treat them as real, just as you
would medium-sized physical objects, when you're explaining chemi-
cal phenomena, making chemical predictions, and so on; but, when you
step back, notice that the existence of atoms hasn't actually been estab-
lished and don't confuse t!te atomic picture with reality. The Second
Philosopher knows many examples of this 'epistemic' phenomenon: a
theory is used, taken as true during 'immersion', while the theorist
nevertheless retains doubts about certain aspects or entities involved.
But atomic theory has now passed beyond this, as a result of Perrin's
experiments: it was once regarded as empirically adequate; now it is
regarded as true.
Once again van Fraassen insists that the Second Philosopher has
misunderstood. The 'immersion' and 'stepping back' she describes is
all happening within what he calls 'the scientific world-picture':
Ostwald's reservations about atomic theory were 'immersed', part of
the internal scientific process of distinguishing between, say, race horses
and flying horses; from the epistemic stance, atomic theory should still
be and should always be regarded as empirically adequate only.
The bewildered Second Philosopher might be inclined to think we
should seek to understand why atomic theory is empirically adequate,
so as to understand the world better. Van Fraassen grants that
The search for explanation is valued in science because it consists
for the most part in the search for theories which are simpler, more
unified, and more likely to be empirically adequate ... because hav-
ing a good explanation consists for the most part in having a theory
with those other qualities. (van Fraassen [1980], pp. 93-4)
98
PENELOPE MADDY
So it's best that working scientists think as the Second Philosopher
does, always searching for explanations-it makes for progress-but
the interpretation of science, and the correct view of its methodol-
ogy, are two separate topics. (van Fraassen [1980], p. 93)
From the interpretive or epistemic point of view
that the observable phenomena ... fit the theory, is merely a brute
fact, and mayor may not have an explanation in tenns of unobservable
facts ... it really does not matter to the goodness of the theory, nor
to our understanding of the world. (van Fraassen [1980], p. 24)
All the Second Philosopher's impulses are methodological, just the
thing to generate good science. Much as Stroud's Descartes recom-
mends that we make knowledge claims in practical life that aren't
properly justified theoretically, van Fraassen finds it beneficial to speak
the language of current science 'like a native' (p. 82). But the correct
interpretation of science-the empiricist interpretation-is entirely in-
dependent of its methodology .32
So the Second Philosopher is once again silenced. Stroud's Descartes
left her no reply, because she was asked to justify her knowledge
without using any of her means of justification. Similarly, van Fraassen
has ruled all her evidence for the existence of atoms as ultimately
irrelevant: good, even admirable, for the purposes of science, to one
immersed; but not rationally compelling to the epistemologist. Her
trouble. is that she is so completely immersed: she doesn't speak the
language of science 'like a native'; she is a native. Van Fraassen intro-
duces her to his epistemic foreign language, where this baffling empiri-
cism reigns: her best theories are taken to be empirically adequate,
rather than true, and the desire for an explanation of why they are
empirically adequate is perhaps useful as a heuristic, but in truth
unmotivated.
In an effort to understand, she asks why we should adopt empiricism
in the foreign language. Van fraassen answ~rs, because 'it makes bet-
ter sense of science, and of scientific activity' ([ 1980], p. 73). Better
than what?, the Second Philosopher wonders. Better than those
accounts of scientific truth in terms of 'best explanations' and 'mature
r
99
Second Philosophy
theories' that van Fraassen explicitly engages and that so befuddled her
before. The salient difference now is that these opponents of van
Fraassen think science aims for truth; not just empirical adequacy, and
though she may flinch at generalities about 'science', the Second Phi-
losopher was dissatisfied before Einstein and Perrin, and perhaps it
isn't too great a distortion to say that she wanted to know if atomic
theory was empirically adequate because it was true or for some other
reason. Still, she concedes that it wasn't clear, before Perrin showed
what could be done, that this question could be answered, that the
existence or non-existence of atoms could be established, and if it
couldn't have been, she might have settled for the empirically adequate
theory. She figures the aim is to do the best we can in determining
what the world is like, but all this, from van Fraassen's point of view,
is just the thinking of one immersed: fine for scientific purposes; irrel-
evant epistemically.
Under the circumstances, the Second Philosopher seems unlikely to
get the hang of this new language-she can't see what style of argu-
ment is appropriate there, given that all hers are 'merely immersed'-
not to mention that she has little motivation for trying33-given her
watchful and considered confidence in her own methods. From her
perspective, the empiricist challenge is hardly more compelling than
the Cartesian. To Stroud's Descartes, she concedes that she cannot
justify her knowledge without using her means of justification; to van
Fraassen, she concedes that she cannot defend the existence of atoms
if all her best evidence is ruled irrelevant. But neither of these gives her
,"on to doubt her methods or to change her ways.
This final stroll has taken the Second Philosopher even further from
Quine's naturalist, as Quine's justification for the atomic hypothesis,
like those .of van Fraassen's proper opponents, depends on general
features of the theory rather than detailed experimental results. This
separates Second metaphysics, like Second epistemology, from meta-
physics and epistemology naturalized, which leads in turn to disagree-
ments in philosophy of mathematics and philosophy of logic. I'll talk
about some of these things on Monday. For now, I hope the Second
Philosopher's character has been brought into some degree of focus by
100 PENELOPE MADDY
this excursion into scepticism. And I hope the lack of philosophical
action hasn't left you wishing you'd gone to a different movie!34
NOTES
. This is the Lakatos A ward Lecture delivered at the London School of
Economics in May 2003.
I. The following account of Descartes goals and strategies comes ftom the
elegant and fascinating Broughton (2002).
2. In the 'Fourth Replies', Descartes refers to 'the exaggerated doubt which
I put forward in the First Meditation', and in dte 'Seventh Replies' he
reminds us that '1 was dealing merely with the kind of extreme doubt
which, as I ftequently stressed, is metaphysical and exaggerated and in no
way to be transferred to practical life' ([1642), pp. 159,308). See Broughton
(2002), p. 48.
3. As Broughton points out ([2002), p. 31), the meditator comes 'uncomfort-
ably equipped with Cartesian theories' at the outset of the Meditations,
though those theories aren't revealed to him until the end.
4. The need to undercut our most tenacious common sense beliefs explains
Descartes' interest in certainty: ifp and q conflict, and there is some slight
reason to doubt p, but q is certain, we take q to undermine p. See Broughton
[2002), p. 51.
5. She doesn't see the errors of childhood as based on a serious inability to
distinguish mind ftom body, so she thinks her ordinary methods of in-
quiry can correct them.
6. Not all of the new science will be indubitable, of course. See Garber
[1986), pp. 115-16, and the references cited there. Even perceptual be-
liefs are only trustworthy when properly examined by Reason. so some
room for error remains here as well (see the final two sentences of
Descartes [1641)) .
7 . Recall that our Second Philosopher has no grounds on which to denounce
First Philosophy as 'unscientific'. Open-minded at all times, she's willing
to entertain Descartes' claim that the Method of Doubt will uncover use-
ful knowledge. If, by her lights, it did generate reliable beliefs, she'd have
no scruple about using it. But if it did. by her lights-tbat is, by lights we
tend to describe as 'scientific'-then we'd also be inclined to describe the
Method of Doubt as 'scientific'.
8. The Second Philosopher is a development of the naturalist described in
Maddy [2001) and [to appear), building on [1997).1 adopt the new name
here to avoid larv;ely irrelevant debates about what 'naturalism' should be.
Second Philosophy
101
9. Both Broughton ([2002], pp. 13-15) and Garber ([1986], p. 82) would
allow that Descartes has some interest in replying to the sceptical argu-
ments current among his contemporaries, but they see this as something
of a side benefit to carrying out his real project of revising the founda-
tions of science.
10. On Broughton's reading, it seems Descartes' meditator could have said
this in the first Meditation.
11. For Stroud on ordinary tests, see [1984), pp. 2(}-23, 46-8.
12. I suspect that Stroud's Descartes gains some rhetorical advantage by stick-
ing to dreaming, a familiar phenomenon, rather than plumping for the
Evil Demon-this makes the challenge seem less like one based on an
objectionable requirement of logical certainty. But the familiar phenom-
enon might well be ruled out in familiar ways.
13. An odd note here. When I say I have hands, there is a risk that I'm wrong,
just as there's a risk that I might not get the best seat on the bus if I sit
down quickly, but it might still be best, in both cases, to take the action.
So I say I have hands without ruling out all possible defeaters. This might
incline me also to say that I know I have bands. But if Stroud's Descartes
is right, this second utterance" is different: it's not that there's a small risk
I might be wrong in saying that I know 1 have bands; there's no chance
at all that 1 might be right!
14. Williams describes this nicely as a sort of 'vector addition': 'The concept
of knowledge, left to itseif so to speak. demands that we consider every
logical possibility of error, no matter how far-fetched. However, the force
of this demand is ordinarily weakened or redirected by a second vector
embodying various practical or otherwise circumstantial limitations. The
effect of philosophical detachment is to eliminate this second vector,
leaving the concept of knowledge to operate unimpeded' ([ 1988], p. 428).
15. I use this example in place of Stroud's plane-spotter ([1984], pp. 67-75,
80-8 I) to bring out the role of scientific inquiry on the sliding scale.
16. Unlike Broughton's Descartes, Stroud's Descartes doesn't suggest that
sceptical hypotheses are themselves means to deeper knowledge.
17. Opinion seems divided on the role of certainty: e.g., Williams holds that
knowledge doesn't require certainty ('there is no obvious route from
fallibilism ... to scepticism' ([1988], p. 430), while Lewis takes the idea
of fallible knowledge to be prima facie 'madness' ([ 1996], p. 221). Stroud
sees the requirements of certainty and foundationalism (the epistemic
priority of experience), not as presuppositions of the scepticaJ argument,
but as 'natural consequence[s] of seeking ... a certain kind of understand-
ing of human knowledge in general' ([1989], p. 104). (I take up this
formulation of the challenge below). My own feeling (which I apparently
I
102 PENELOPE MADDY
share with Williams) is that the sceptical challenge isn't of much interest
(unless as a Method, as for Broughton's Descartes) if it rests on a require-
ment of certainty. (Fallibilism doesn't trouble me as it does Lewis.)
18. The Second Philosopher may well suspect that her linguistic inquiry into
the semantics of 'know' will not turn up anything determinate and unified
enough to play the role of the 'underlying concept' Stroud's argument
requires. Williams ([1988], p. 428) seems to make a similar suggestion,
though in a different argumentative setting. Lewis finds the concept com-
plex, but still more strictly codifiable than seems likely for a rough and
ready notion like knowledge.
19. For example, over the certainty requirement: it's hard to see how the
argument just rehearsed from the possibility that I'm dreaming to the
conclusion that I don't know I have hands can be pressed without requir-
ing that knowledge be certain; the sliding scale seems to be an attempt to
defend this requirement. But leaving aside the word or .:oncept (see pre-
vious footnote) of 'knowledge', what really matters is whether or not I
have good reason to believe I have hands.
20. From the philosophical perspective, no certainty requirement seems to be
presupposed: if 1 admit (as I must) that I might be dreaming, I have no
grounds on which to count this hypothesis as unlikely-because 1 can't
appeal to other knowledge of the world-and thus I have no good reason
to believe that I have hands. (Stroud doesn't put the case quite this way
in [1996], p. 132, but 1 think the spirit is the same.) As Stroud claims, the
requirement of certainty emerges from the sceptical reasoning, because,
in the absence of other information, any room for doubt leaves me with
no good reason to believe. Thus, it seems to me that two different argu-
ments for the sceptical conclusion are being offered, depending, on which
considerations support the key move from 'I could be dreaming 1 have
hands' to 'I don't know 1 have hands', the move, that is, that rules out the
response that the dream hypothesis is sufficiently unlikely to be dismissed:
in the first version, the sliding scale argument purportedly shows that
certainty is required for knowledge, so the dream hypothesis must be
conclusively defeated; in the second version, the philosophical perspec-
tive disallows the appeal to collateral information that would show the
dream hypothesis to be unlikely.
21. This is adapted from Stroud {1994], a reply to externalism. See also
Stroud [1989].
22. Stroud notes ([1984], p. 120) that 'even Homer nods'-there are places
where Moore leans farther than perhaps he should toward the 'external'
understanding.
Second Philosophy 103
23. See Fogelin [1997] for a discussion of one major fault line in the Quinean
opus, between naturalism on one side and the likes of holism and onto-
logical relativity on the other. I emphasize the tension between naturalism
and holism in [1997]. Stroud [1984], chapter VI, highlights the difficulty
of finding a single, consistent Quincan doctrine on scepticism. On epis-
temology itself, Quine sometimes says, 'why not settle for psychology?'
([1969], p. 75), while the Second Philosopher imagines a broader study,
including various other human studies, plus her accounts of the things
known. But see also Quine [1995], p. 16, where naturalized epistemology
is described as 'proceed[ing] in disregard of disciplinary boundaries but
with respect for the disciplines themselves and'appetite for their input'.
24. Fogelin traces Quine's approach to Camap's Aujbau: 'Quine's inspiration
comes from the library, not the laboratory' ([ 1997], p. 561).
25. I'm not sure Quine would disagree with this diagnosis. After all, his views
on proxy functions suggest that the world could be made of numbers
instead of physical objects, for all our evidence tells us. Stroud needn't
take a God's eye view and declare all these ontologies as equally good
(Quine [1981], p. 21); he need only point out that science itself has told
us that it.. evidence doesn't support its ontology over many rivals. Thus,
it's hard to see how Quine has 'defend[ed] science from within, against its
self-doubts' (Quine [1974], p. 3).
Quine replies that his 'only criticism of the sceptic is that he is over-
reacting' when he 'repudiates science' (Quine [198Ia], p. 475). I'm not
sure what this repudiation comes to, apart from denying that science is
knowledge. But Quine himself declares that there is no sense in which the
world can be 'said to deviate from ... a theory that is conformable to every
possible observation' (op. cit., p. 474). It sounds as if there is no fact of
the matter about ontology that we can be said to know or fail to know.
It's hard to resist Fogelin's conclusion (Fogelin [1997]) that Quine's
naturalism sits ill with his ontological relativity. Surely ordinary science
thinks there is a fact of the matter about whether the world is composed
of physical objects, as opposed to numbers.
26. See Stroud [1984], pp. 225-34.
27. See Quine ([1960], p. 22, [1969], pp. 82-3, [1974], pp. 2-3). Gibson
([1988], p. 66) suggests that Quine should be understood as linking the
'irritations'to 'hoiophrastically acquired observation sentences'. If so, then
we're back to the previous situation, with a gap between those observa-
tion sentences and 'theory' (and the latter includes observation statements
understood referentially).
28. I'm not sure whether Stroud's worry is over the coherence of the way the
sceptical challenge is raised, that is, by generalizing from a particular case
104 PENELOPE MADD~
(Descartes' inability to know by perception that he has hands) to the
whole of our purported knowledge of the world, or over the coherence of
the philosophical perspective itself, whether or not it's forced upon us by
the sceptical argument. If the former, he might still take the challenge to
be coherent and unanswerable (as the Second Philosopher does).
29. For details and references, see [1997], pp. 135-42. Achinstein [2002]
adopted the same general tone on the efficacy of the Perrin experiments.
30. Cf. van Fraassen [1985], p. 252: 'A person may believe that a certain
theory is true and explain that he does so, for instance, because it is the
best explanation he has of the facts or because it gives him the most
satisfying world picture. That does not make him irrational, but 1 take it
to be part of empiricism to disdain such reasons.' These aren't the Second
Philosopher's reasons.
31. van Fraassen uses 'electron', in this quotation and the next, but the same
would seem to go for atoms.
32. Like van Fraassen, Broughton's Descartes thought his contemporary sci-
entists were wrong, but he clearly didn't take this fact to be methodologi-
cally irrelevant-his aim was to change the way science was done.
33. Notice that van Fraassen, like Stroud's Descartes but unlike Broughton's
Descartes, is not offering an improvement in scientific methods.
34. My thanks to Sam Hillier, David Malament. and Kyle Stanford for help-
ful conversations and comments on earlier drafts, and to the audience at
the Lakatos Lecture at LSE for stimulating discussions.
REFERENCES
Achinstein, Peter [2002]. 'Is there a valid experimental argument for scientific
realism?', Journal of Philosophy 99, pp. 470-95.
Broughton, Janet [2002]. Descartes's Method of Doubt. (Princeton, NJ: Princeton
University Press).
DeRose. Keith. and Warfield, Ted, eds. [1999]. Skepticism. (New York: Oxford
University Press).
Descartes, Rene [1641]. Meditations on First Philosophy, in J. Cottingham et al.,
trans., The Philosophical Writings of Descartes, volume II, (Cambridge:
Cambridge University Press, 1984), pp. 3-62.
[164la]. 'Letter to Mersenne', in J. Cottingham et al., trans., The Philo-
sophical Writings of Descartes, volume III, (Cambridge: Cambridge
University Press, 1984), pp. 171-3.
[1642). Objection.f and Replies, in J. Cottingham et al., trans., The Philo-
sophical Writings of Descartes, volume II, (Cambridge: Cambridge Uni-
versify Press. 1984), pp. 66-383.
Second Philosophy
105
[1644]. Principles of Philosophy, in J. Cottingham et al., trans., 11re
Philosophical Writings !f Descartes. volume I, (Cambridge: Cambridge
University Press, 1985), pp. 179-291.
Feynman, Richard [1998]. 11re Meaning of It All: 11wughts of a Citizen-Scientist,
(Reading, MA: Perseus Books).
Fogelin, Robert [1997]. 'Quine's limited naturalism', Journal of Philosophy 94,
pp. 543-63.
Garber, Daniel [1986]. 'Semel in vita: the scientific background to Descartes'
Meditations', in A. Rorty, ed., Essays on Descartes Meditations, (Berkeley,
CA: University of California Press).
Gibson, Roger [1988]. Enlightened Empiricism: an Examination of w. V. Quine's
Theory of Knowledge, (Tampa: University of South Florida Press).
Lewis, David [1996]. 'Elusive knowledge', reprinted in Derose and Warfield. pp.
220-39.
Maddy, Penelope (1997). Naturalism in Mathematics, (Oxford: Oxford Univer-
sity Press).
[2001]. 'Naturalism: friends and foes', in J. Tomberlin, ed., Metaphysics
2001. Phi/oslJphical Perspectives 15, (Malden, MA: Blackwell), pp. 37-
67.
[Forthcoming]. 'Three forms of naturalism', to appear in S. Shapiro, ed.,
Oxford Handbook of Philosophy of Mathematics and Logic.
Quine, W.V.O. [1960]. Word and Object. (Cambridge, MA: MIT Press).
- [1969]. 'Epistemology naturalized', reprinted in his Ontological Relativ-
ity and Other Essays, (New York: Columbia University Press, 1969), pp.
69-90.
-- - [1974J. 11re ROlJts of Re.ference, (La Salle, IL: Open Court).
- [1981]. 'Things and their place in theories', in Theories and 11rings,
(Cambridge, MA: Harvard University Press), pp. 1-23.
[198Ia]. 'Reply to Stroud', Midwest Studies in Philosophy 6, pp. 473-4.
- -- [1995]. From Stimulus to Science, (Cambridge, MA: Harvard University
Press).
Stroud. Bany [1984]. The Significance of Philosophical Skeptici.\"m, (Oxford:
Oxford University Press).
[1989]. 'Understanding human knowledge in general', reprinted in his
[2000], pp. 99--121.
--~- [1994]. 'Scepticism, "externalism", and the goal of epistemology', re-
printed in his [2000], pp. 139-54, and in DeRose and Warfield, pp. 292-
304.
~- [1996]. 'Epistemological reflection on knowledge of the external world',
reprinted in his [2000J, pp. 122-38.
106
PENELOPE MADDY
[2000]. Understanding Human Knowledge, (Oxford: Oxford University
Press).
van Fmassen, Bas C. [1980]. The Scientific Image, (Oxford: Oxford University
Press).
-- [1985]. 'Empiricism in the philosophy of science', in P.M. Churchland
and C.A. Hooker, eds., Images of Science, (Chicago: University of Chi-
cago Press), pp. 245-308.
Williams, Michael [1988]. 'Epistemological realism and the basis for skepticism',
Now 91, pp. 415-39.