Professional Documents
Culture Documents
Akash Mehta
5/13/19
Word Count: 3219
Stop a random passerby on the street and ask: “what is philosophical skepticism?” Chances are,
you’ll be referred to the possibility that you’re dreaming or living in The Matrix. But in his landmark
essay “The Legacy of Skepticism”, Thompson Clarke accepts both these possibilities as genuine instances
of ‘plain skepticism’, while still claiming that philosophical skepticism is incoherent. On its face, this
may seem like evasion via redefinition—if Clarke concedes the two best-known examples (brain-in-a-vat
and dreaming possibilities) of philosophers trying to make skeptical arguments, isn’t he in effect
conceding philosophical skepticism? But the philosophical skeptic, Clarke argues, is trying to argue
against the possibility of certain knowledge as such—and it is only though a misunderstanding of her own
examples that she sees them as furthering that argument. The philosophical skeptical challenge to certain
knowledge as such, Clarke argues, is not even wrong; it is inconceivable, nonsensical, unintelligible.
In this essay, I first present what I take to be Clarke’s distinction between the plain and
philosophical skeptical possibilities and why he does not believe the latter to be genuine. I then raise
doubts about Clarke’s rather rejection of the philosophical skeptical possibility; he seems to rely on the
whereas I suggest that it may be possible to meaningfully speak of notions which our psychological
capacities are not being up to task of conceiving. Finally I consider two defenses of Clarke’s position
evoked by passages in Ludwig Wittgenstein’s On Certainty (which, I will suggest, contains the core of
1 of 8
Clarke’s argument). My treatment of the issues raised by these defenses is very far from comprehensive,
but again I raise doubts about their ability to disarm the philosophical skeptic.1
Clarke’s essay culminates in the discussion of PE, the “skeptic’s epistemic possibility”, and PNE,
the “nonepistemic possibility.” A characteristic expression of PE is that “all this now might turn out to be a
dream: I might wake up later… and discover I had just been dreaming” (Clarke 764); another expression
would be that I might be unplugged from the Matrix and discover myself to have been taken in by a
simulation. PE, then, is what the random passerby took to exemplify philosophical skepticism. But Clarke
points out that PE can only be a plain possibility, in that it takes for granted the coherence of some
perspective from which things are knowable (in our examples, my perspective after waking or being
unplugged). PE therefore does not challenge knowledge as such, as the philosophical skeptic aims to do,
but only points out that particular things I believe myself to know (e.g. that I am not dreaming) may turn
Realizing that PE cannot cast knowledge as such into doubt, the skeptic attempts to raise PNE, of
which Clarke supplies this characteristic expression: “it might be that we’re now asleep, dreaming….
There is no implication, pro or con, that we could (ever) find out.” (Clarke 766) But Clarke argues that in
order to imagine PNE, we must yet again take for granted some perspective from which things are
knowable. “I can imagine, it seems,” Clarke writes, “that I might be asleep now, dreaming, really in
surroundings very different from these. But the moment I am conscious that there will be real
surroundings, I realize I’m taking it for granted that these environs could be observed, known to be real,
by outsiders, if any, in appropriate positions” (Clarke 766). Descartes’ possibility that an Evil Demon
might be interfering with his arithmetic calculations, to take another expression of PNE, does indeed put
into raise the possibility that 2+2=5—but it only does so by taking for granted that, were that possibility
1 When one disagrees with a thinker of Clarke’s rank (let alone Wittgenstein’s), it is of course more likely
that one has misunderstood his argument than that one has refuted it; accepting this likelihood, I
nevertheless hope that the essay prompts readers to clarify for themselves what exactly my
characterization of his argument or my would-be rebuttal misses.
2 of 8
actual, the Evil Demon could know it to equal 5. Thus no less than PE, Clarke argues, PNE takes for
For Clarke, PE and PNE are genuine plain possibilities. It might be true that I really live in the
Matrix or that 2+2 really equals 5; “it seems almost beyond question that what plain PE and PNE suggest
could happen, could, indeed, just possibly” (Clarke 768). But what is not a genuine possibility is what
Clarke labels ‘PX’: that there is no ‘really’ to speak of, that whether or not I am in the Matrix is
unknowable in principle. Although this last claim is the crux of the entire argument for the
unintelligibility of philosophical skepticism, Clarke provides little in the way of justification for it. What
“We have no satisfactory techniques for handling a question like [whether PX is genuine]
objectively: we are forced winetasters of the conceivable. Acknowledging this, I feel confident,
nevertheless, that it is inconceivable that I could now be asleep, dreaming, if no outsider could
know my real environs because in the same boat, for the same reason, because he, too, could not
know he was not asleep, dreaming. Does Descartes's possibility even seem to make sense, if we
ask ourselves how the Evil Demon, or God, could know that he, too, wasn't dreaming—and
Clarke makes a move in this paragraph to which it bears paying attention closely. First he asserts
the inconceivability of PX; although he suggests that he cannot establish it “objectively”, he nevertheless
asserts his confidence that PX is indeed inconceivable. Then, as if continuing the same thought, he asks
whether “Descartes’s possibility even seem[s] to make sense” if we don’t take for granted that the Evil
2 As best as I can tell, Clarke refrains from saying PE and PNE are equivalent only because the former takes
for granted a perspective from which I may know the truth, whereas the latter takes for granted a
perspective from which an outsider may know the truth. Clarke does not take this distinction to be of
much relevance to the issue of philosophical skepticism; he writes that “the story of PNE is, in detail, the
tale told for PE.” (Clarke 767)
3 of 8
Demon is capable of knowledge. Clarke takes it for granted, it seems, that from the inconceivability of PX
But is this inference of nonsensicality from inconceivability justified? Consider the case of
infinity. When I try to conceive of an infinite universe, the image that comes before my mind is
indistinguishable from that of a large finite one. I look up into the sky, and imagine it going on and on and
on; but I do much the same when I try to picture the size of the (finite) observable universe. When I try to
conceive of an infinite number of points on a line segment, similarly, I find myself carrying out the same
mental operations involved in trying to imagine a trillion discrete points along that segment. In this sense,
then, infinity is inconceivable. Nevertheless I can meaningfully ask whether the universe is infinite in
extent and whether it will exist forever, and I can sensibly assert that there are an infinitude of points on a
line segment. In the case of infinity, then, the inconceivable still seems sensical.
How is it possible for me to meaningfully speak of infinity, if I can’t conceive of it? This is of
course an issue with a storied philosophical history (indeed, Descartes took it to give rise to the very proof
of God that alleviated his skepticism!) and deserves far more consideration than I can give it here. But
one plausible4 answer is that reason enables us to discuss concepts which we cannot imagine. By
reasoning, “no matter how large a number you give me, I can always add one,” I immediately grasp the
notion that there are infinity counting numbers and am able to utilize that notion (e.g. in various
mathematical proofs), without ever conceiving of that infinity. I raise our ability to make sense
of infinity not only as an analogy, but also to suggest that that very ability may enable us to similarly
3 “Palpable nonsensicality” might be more precise, as Clarke asks whether Descartes’s possibility “even
seems to make sense”, not whether it does make sense. Nevertheless, the rest of the essay demonstrates
that for Clarke, that it doesn’t make sense (i.e. is nonsense) follows from the fact that it doesn’t seem to
make sense (i.e. is inconceivable). This inference is what allows Clarke to rule philosophical skepticism
“illegitimate”, and is what I challenge in the remainder of the essay.
4 At the risk of cheekiness, I’ll note here that all my argument needs to work is for this answer to be
plausible or even barely possible. The same goes, in fact, for every step of my argument; as long as each
one is possibly correct, then the argument’s conclusions possibly follow—which, as it will turn out, is all
that the argument claims they do anyway!
4 of 8
make sense of PX without disputing Clarke’s assertion that we cannot conceive of it. As we saw earlier,
Clarke allow us the possibility that our apparent knowledge (let’s call it ‘K1’) that we are awake may be
revealed false by the knowledge (let’s call it ‘K2’) of our future awakened selves that they are awake. But
if we can imagine K2, surely we can imagine the possibility in which K2 is itself revealed to be false by
the knowledge (‘K3’) of subsequently awakened selves. Indeed, anyone who’s seen the movie “Inception”
or read Gogol’s “The Portrait” has imagined exactly such a scenario—and anyone who’s experienced a
false awakening has been in such a scenario! Similarly, we can conceive of the possibilities K4, K5, or
indeed any Kfinite; it simply takes imagining waking that many times.5
knowledge of wakefulness for granted in its final iteration. But the possibility of Kinfinity is a philosophical
skeptical possibility, an expression of PX; there is no claim of wakefulness that it does not cast into doubt.
Can Kinfinity be discussed meaningfully? It seems to me that in the same sense in which one can speak
meaningfully of infinity, it can. Neither notion is conceivable, but both are graspable by reason; just as I
can reason that any n is followed by n+1, so too I can reason that any Kn might be followed by a Kn+1.
(Similarly, I can reason that just as the Evil Demon caused me to miscompute 2+2, so too it is possible
that an Eviler Demon caused him to miscompute 2+2 as well, and so on.6)
As best as I can tell, Clarke does not provide a counter-argument to my defense of philosophical
skepticism. Ludwig Wittgenstein’s On Certainty, however, seems to suggest two such counter-
5Speaking precisely, the finitude of our computational capacities means that we cannot in fact imagine
numbers past a certain point; but if anything this only buttresses my argument, as we still can quite
meaningfully assert that numbers exist beyond that point.
6 Note that nothing in my argument denies that it is also possible that the first Evil Demon, for reasons
that I cannot comprehend, in fact is incapable of arithmetic errors. All I claim is that it is also possible that
he is indeed capable of them.
5 of 8
arguments.7 Before I consider them, let me justify rather abruptly bringing On Certainty into a discussion
of an article of Clarke’s by pointing to the two texts’ similarities.8 “If I imagine such a person [with non-
veridical experience] I also imagine a reality, a world that surrounds him,” Wittgenstein writes, “and I
imagine him as thinking (and speaking) in contradiction to this world.” (OC 595) Here I take Wittgenstein
to express Clarke’s argument that one can only imagine non-veridical experience (say, a case of
dreaming) if one takes for granted a perspective of knowledge from which it is revealed to be non-
veridical (the awakened self). Similarly, Wittgenstein says of Moore’s claims of knowledge (and would, I
think, also say of their skeptical opposites) that “for each one of these sentences I can imagine
circumstances that turn it into a move in one of our language-games, and by that it loses everything that is
philosophically interesting” (OC 622); I take this to anticipate Clarke’s celebration of a Moore concerned
only with the plain, and more broadly to anticipate Clarke’s distinction between the plain and the
philosophical. Finally, let me point out that like Clarke, both of the remarks in On Certainty I have thus
far quoted stress the importance of what we can imagine when considering skeptical challenges.
The first potential counter-argument I’ll consider is suggested by passages such as OC 383, in
which Wittgenstein holds that “the argument “I may be dreaming” is senseless for this reason: if I am
dreaming, this remark is being dreamed as well—and indeed it is also being dreamed that these words
7I cannot begin to do justice to On Certainty in the few pages I have allotted to its discussion here; my
project here is of course not to rebut On Certainty, but only to raise doubts about two arguments evoked
by a few of its passages. I say ‘evoked by’ rather than ‘made by’ because a full appreciation of these
passages’ arguments (let alone a rebuttal of them) would require their careful consideration in the context
of the whole book.
8 For reasons of space, I limit myself to quoting two passages that seem to clearly demonstrate that On
Certainty contains Clarke’s core argument. There are many others that I might have picked instead,
including OC 19, 115, 407, 467 and 554.
6 of 8
have any meaning.” (OC 383) Applied to PX,9 Wittgenstein’s thought could be perhaps be formulated as
follows: any challenge to knowability as such undercuts itself, because taken seriously it casts itself into
doubt, together with the words by which it is expressed. Thus we haven’t doubted knowability as such
any more than we’ve doubted the argument against it. But this doesn’t seem to me to disarm philosophical
skepticism, which as I see it aims not to establish the certainty of uncertainty (which would indeed
undercut itself), but the uncertainty of all things, including itself. Indeed one implication of PX is that
every argument may contain a mistake; that this implies that arguments for the genuineness of PX
themselves might contain mistakes hardly negates the possibility that, say, 2+2=4 might contain a
mistake; for after all, the arguments for PX might also not contain mistakes!10
But this might seem to deprive the skeptic of grounds for taking his own argument to be any more
compelling than his opponent’s.11 “If something happened (such as someone telling me something)
calculated to make me doubtful of my own name,” Wittgenstein writes in OC 516, “there would certainly
also be something that made the grounds of these doubts themselves seem doubtful, and I could therefore
decide to retain my old belief.” This leads to the second potential counter-argument, that philosophical
skepticism makes no “difference… in their lives” other than that skeptics just “talk rather more about
certain thing than the rest of us” (OC 338); that faced with the skeptic, “one might for example suppose
9 Note that Wittgenstein may have instead been thinking here primarily of what Clarke would call the
plain PE and PNE. If so, the passage should be read similarly to OC 54, which argues that a mistake in
identifying a hand is not merely improbable but inconceivable on the grounds that “if it were not so, it
would also be conceivable that we should be wrong in every statement about physical objects; that any we
ever make are mistaken.” What I read Wittgenstein as arguing here is that if one accepts a certain sort of
what Clarke would consider plain skeptical possibilities (e.g. that I may be dreaming, or that this may not
be my hand), inevitably one opens oneself up to the philosophical skeptical possibility. The ‘plain’
possibility generalizes, in other words, in a way that Clarke doesn’t acknowledge. On this point, far from
disagreeing with Wittgenstein, my essay aims to elaborate his argument!
10To take a simpler example, consider two propositions: a) it is possible that the toy broke; b) it is
possible that the toy is in fact indestructible and thus that proposition a is false. Is it not quite sensible to
assert “a and b”?
11Whether this is so depends on whether the notion of probability can survive philosophical skepticism.
Can the skeptic coherently say, “there is a high chance that PX is genuine”? Or are our very notions of
high and low chances intelligible only against the backdrop of ultimate probabilities 1 and 0?
7 of 8
that he has read skeptical philosophers, become convinced that one can know nothing, and that is why he
had adopted this way of speaking. Once we are used to it, it does not infect practice.” (OC 524) Drawing
on these passages, one might want to argue that philosophical skepticism is nonsensical insofar as the
sense of any sentence or belief is reflected in the way it makes a difference in our lives, the way in which
it shows up in our practice. In the very fact that one can only be a philosophical skeptic in one’s study, the
But philosophical skepticism can indeed, it seems to me, make a difference in our lives.
Confronted with the apparently unavoidable conclusion that nothing can be definitely known—that the
world may be unfathomably other than it seems and than it is even capable of seeming, that the most basic
of our assumptions and perceptions may not be tethered to reality, that the very words we use to think and
express our skepticism itself may not refer to anything—we feel our grip on the world loosening. Per
some ways of thought (most obviously within certain traditions of Hindu and Buddhist philosophy), this
loosened grip on the world enables the enlightened serenity of coming to peace with the enigma of all
things, and thus is precisely the aim of contemplation. Per others, the skeptical possibility subverts the
hubristic presumption that human reason can establish the coherence and meaningfulness of existence,
and thus teaches a lesson of the deepest import: the necessity of faith to provide that coherence and
meaning and tighten our grip on the world. Of course, skepticism can be charged with infecting practice
in less rosy ways, perhaps as a route to nihilism and suicide. Such arguments may provide reason to burn
all extant copies of the Meditations, but they do not negate the skeptical possibility considered therein;
they do not destroy the road to philosophical skepticism, but only attempt to blockade it.
8 of 8