Skepticism
Skepticism
Skepticism
Duncan Pritchard
Module 1
What is Skepticism?
Introducing Skepticism
In a nutshell, skepticism is about doubt, in that to be skeptical about something is to have doubts
about it. Skepticism can apply to lots of different things: one can be skeptical about a person (e.g., a
used car salesman), a topic (e.g., horoscope predictions), and even objects (e.g., one might be
skeptical that one’s clapped-out air-conditioning is going to last the summer). What these different
forms of skepticism have in common is that one is concerned about whether one can rely on the
thing in question, whether it is the advice of the used car salesman, the predictions made by the
horoscope, or the effectiveness of the air-conditioning. In short, skepticism undermines belief (belief
that what the car salesman is telling you is correct, belief that the air-conditioning will last the
summer, and so on), and so this is what we will be focussing upon here.
A certain degree of skepticism is often a good thing. Indeed, we talk about the importance of having
a ‘healthy skepticism’, where this means not simply accepting whatever one is told. Skepticism in this
sense is the antidote to gullibility, and surely no-one wants to be gullible. Some things warrant
skepticism after all. Take the example I just mentioned of horoscopes. These predictions are
notoriously problematic, in that they are either very specific, in which case they are often wrong, or
more often they are so general that they would fit any eventuality, in which case there is no clear
sense in which they are ever right. Moreover, we also know that there is no scientific basis for
astrology—scientists long ago abandoned astrology in favour of astronomy, which unlike astrology is
an accepted scientific discipline. So there are excellent grounds for being skeptical about the
reliability of horoscopes.
A healthy skepticism can also prevent us from being taken in by those who want to deceive us. For
example, since we know that used car salesmen have a motive for wanting us to buy a car at the
highest price possible, hence we know that we should take what they say with a pinch of salt and not
simply accept it completely at face-value. More generally, if someone we don’t know well tells us
something that seems at first blush to be rather incredible⎯for example, that the Queen of England
has just been arrested for shoplifting⎯then our skeptical instinct should kick-in to ensure that we
don’t simply take this testimony on trust. That’s not to say that we should never accept incredible
testimony of this kind, but only that we should always demand extra grounds in such cases (for
example, we might want to switch on the news to verify this incredible tale).
3
But this type of quite specific skepticism can also easily drift into something more extreme and
that will be of particular
generalised. It is skepticism of this kind⎯what is known as radical skepticism⎯
interest to us. For example, what are we to make of the contemporary trend in some aspects of
public life to be skeptical about science itself, such as concerning the scientific consensus about
manmade climate change? Notice how this kind of skepticism is very different to skepticism about
horoscopes. In the latter case, the skepticism is grounded in a belief in the reliability of science. And
that seems legitimate precisely because we regard science as being a paradigmatic way of getting to
the truth about the world around us. But when one is skeptical about science itself, then of course
that kind of skepticism cannot be grounded in science. If we reject the authority of science to tell us
about the world around us, however, then what will be our basis for belief in this regard? The worry
is that a healthy skepticism is here morphing into a generalised skepticism that is an entirely different
beast.
Here’s another way of putting the point. There can be all kinds of good reasons why it might be right
to be skeptical about particular claims, such as regarding what horoscopes or used car salesmen tell
us. Such localised skeptical doubt is, however, grounded in what we know, such as our knowledge of
the unscientific nature of horoscopes and the motivations of used car salesmen. But once we shift
from a localised skeptical doubt to one that is much more general—as when we become skeptical of
scientific claims en masse⎯ then it is hard to make sense of how our skepticism is grounded in what we
know at all. For whatever we might cite as reasons for this doubt, wouldn’t they be equally in doubt
too? The worry is that rather than being grounded in something that we can rely upon, as a localised
skeptical doubt might be, a generalised or wholesale skeptical doubt becomes instead completely
free-floating.
When skepticism of this more radical kind takes root in our public life, then it has all kinds of
practical ramifications, many of them not particularly appealing. One consequence is a lack of
concern for accuracy and the truth. Think, for example, of such contemporary phenomena as ‘false
facts’ and ‘post-truth politics’, whereby people in public life utter manifest falsehoods seemingly with
impunity. A radical skepticism effectively licenses such phenomena, since once everything is open to
doubt, then there is nothing that is accepted as true, and hence what’s true starts to drop out of the
equation altogether.
In particular, once a concern for truth and accuracy is lost, then a radical skeptical doubt gives way to
a prevailing relativism about truth. This is the view that the truth is just simply whatever someone says
it is. For example, one camp endorses the scientific consensus that climate change is man-made,
whereas another camp argues that these scientists are all part of a global conspiracy to deceive the
public. According to relativism, both camps can be right, since truth is relative to one’s subjective
opinion.
It might initially seem liberating to shake off a concern for truth and accuracy in this way, and to
allow that opposing groups can both be right, but this is an illusion. For notice that to say that both
camps in a dispute are right is as empty as saying that they are both wrong. Once you abandon the
4
idea of accuracy, then it doesn’t matter any more what’s true and what isn’t. But getting things right
should matter to us.
In order to see this, just consider something that you take to be extremely important. For example,
imagine that you are charged with a serious crime that you didn’t commit. Wouldn’t it matter a great
deal to you that the truth was brought to light, and hence that you were acquitted? And wouldn’t it
appal you if, even after being acquitted, those who accused you continued to maintain, in a relativistic
spirit, that, contrary to the evidence produced in favour of your innocence at your trial, it was just as
much ‘true’ that you committed the crime in question? You would surely regard this as an injustice,
and you would be right to do so. But that just goes to show that the truth counts, and that means the
truth in the sense of what actually is the case. But of course there can’t be multiple inconsistent
accounts of what is actually the case. Either you did the crime or you didn’t. And if you didn’t, then
anyone who says otherwise is saying something false, not saying something true that is relative to
their subjective opinion.
In the last section we encountered relativism about truth. This is the idea that what is true is relative
to one’s subjective opinion, so that if your opinion is in conflict with someone else’s opinion
then—no problem!—you can both be right. We noted that relativism about truth is not all it’s cracked
up to be, not least because if everyone is right, then everyone is just as much wrong too. The point is
that if the idea of truth capturing whether a viewpoint is accurate or not is lost, then it really doesn’t
matter any more whether that viewpoint is true or false. And yet we clearly do care about getting
things right, particularly when this concerns something of personal importance, such as whether one
is wrongly found guilty of a serious crime. Or, to take another example, if a doctor is about to
operate on you, I think you would care a great deal whether the surgery she was about to perform
was based on accepted medical practice rather than just being the doctor’s subjective opinion.
5
But if relativism about truth is such a bad idea, then why might some people be attracted to it? I
think there are several reasons. One reason might be that some things are simply a matter of
subjective opinion. What foods you like, what you find funny, your favourite film, and so on. These
are all just matters of subjective opinion. In particular, there’s nothing odd about the fact that I might
hate sweet foods (this is true in fact, I simply cannot stand to eat anything sweet), but most people
really like sweet foods. But that some things—matters of taste, essentially—are just a matter of
subjective opinion doesn’t mean that everything is.
Suppose, for example, that a scientist claims to have discovered a new planet. Either what she says is
true, in which case the planet she describes exists, or it’s false, in which case the planet she describes
doesn’t exist. Opinions don’t come into the matter at all. And that’s precisely because when we talk
about whether something is true, we want to know whether it is actually the case, and not merely
what someone thinks might be the case.
(By the way, there is still truth even when we are talking about subjective matters of taste, such as
what foods one prefers. Take the claim that I can’t stand the taste of sweet foods. It is purely a
matter of subjective opinion that I don’t like sweet foods, unlike most other people. But it is not a
matter of subjective opinion that it’s true that I don’t like sweet foods, any more than it’s a matter of
subjective opinion that it’s true that other people tend to like them. As in the case just described with
the planet, either I do like sweet foods or I don’t. If I don’t like sweet foods, then the claim that I
don’t is true; otherwise it is false. And there’s nothing relative or subjective about that).
Another reason why some people might be attracted to relativism is that it can superficially seem to
be liberating, and even a way of respecting the views of those we don’t agree with. Rather than
having to take sides in a debate, one can now just say that both parties are right. Isn’t that a way of
respecting the views of everyone concerned? But notice that this isn’t to respect the other person’s
views at all. After all, they weren’t saying that this was their subjective opinion, but that it was
true—i.e., such that they were right and the other party was wrong. In saying that they have merely
claimed a relative truth, however, we are effectively saying that what they have claimed is not true at
all, at least not as they meant it. After all, they weren’t saying that this was just their subjective
opinion, but that it was actually (i.e., objectively) true. In particular, they were claiming that they were
right and the other person’s views were wrong, not that they were both saying something true (albeit
in a relativistic way).
What applies to other people’s opinions applies with even more force to one’s own. Think about
your most cherished beliefs, such as your ethical, political or religious convictions. Now contrast
these cherished beliefs with claims that are merely a matter of your own subjective opinion. You
clearly think the former are true, which is why you have such conviction in them. But what comfort
would you get in discovering that ‘true’ here just means your own subjective opinion and nothing
more? After all, you don’t think that your deepest convictions are just matters of taste, such as
whether you like (or dislike) sweet foods.
6
Suppose, for example, that you come across someone who has completely different political views to
you. Perhaps you are in favour of western liberal democracy, while they favour a totalitarian state
ruled by a strong leader. If this disagreement is just about a matter of taste, then there would be
nothing to disagree about, any more than there is any point in me ‘disagreeing’ with someone over
whether sweet foods taste nice (they simply don’t for me, but do for others). And yet there clearly is
something substantive here worth disagreeing about, as the political organisation of society has all
sorts of practical implications for the lives that we lead. This simply isn’t the sort of issue where one
can casually shrug one’s shoulders, as you might if I told you that I didn’t like sweet foods!
Yet another reason why I think some people are initially attracted to relativism is because of the
inherent fallibility of our judgements. What I mean by this is that even our best judgements can
sometimes be mistaken. Put another way, we are not infallible creatures. This is true even of science.
The best, most grounded, scientific theories of the day could give way, over time, to better scientific
theories that completely replace the earlier proposals. (The idea that the sun orbited the earth, rather
than vice versa, was once a widely held claim, after all). But if it is always possible that we might be
wrong, then how confident can we be that we are ever right? In particular, how confident can we be
that what we believe is really true, as opposed to merely being a subjective opinion that could turn
out to be false? Accordingly, why not jettison the idea of truth as objective and instead treat it as
merely relative?
There are several points that we need to tease out here. The first thing to notice is that this line of
reasoning is primarily motivating skepticism rather than relativism. This is thus a good juncture to
remind ourselves of the difference. Recall that skepticism concerns doubt, primarily doubt about
what is true. So construed, the skeptic is not proposing that truth is just subjective opinion in the way
that the relativist is. Indeed, what is motivating skepticism is rather the worry that our beliefs might
not be true in an objective sense. As we noted earlier, however, skepticism can slide into relativism if
the former becomes extensive enough. If we are inclined to doubt everything, then we might be
tempted to think that there is no such thing as an objective truth, and hence that everything is just a
matter of subjective opinion instead.
The general worry about our fallibility is primarily motivating skepticism rather than relativism
because it is giving us a reason to doubt even those things that we are most confident of. After all,
people have been very confident of things in the past (such as that the sun orbited the earth) and
have turned out to be mistaken. Accordingly, we seem to have a reason to doubt everything we
believe. The connection to relativism comes in only indirectly, via the point we just noted about how
once one’s skepticism becomes extensive enough, then it seems to invite relativism.
Crucially, however, while there are some good arguments for skepticism⎯indeed, we are going to be
considering one influential argument in this regard soon⎯the appeal to our fallibility is not a good
basis to be generally skeptical about the truth of our beliefs. Sure, the possibility that you are
mistaken is always a live one. But that in itself is not a good reason to be skeptical about everything
you believe (or even most of what you believe). What you need is rather some specific reason to
think that you are mistaken in this particular case.
7
In order to see this, consider some everyday scenarios. Take, for example, the question of whether a
particular tree is an oak. If I am the person making the judgement, then there are some grounds for
skepticism about this judgement, as I’m not very reliable when it comes to identifying types of tree.
But compare this case with someone making the very same judgement who is a qualified arborist
(i.e., a tree surgeon) with a long history of working with trees of different kinds. Sure, it’s still
possible that they might be mistaken, but unlike my judgement that the tree is an oak, this isn’t a
serious possibility at all, as it would frankly be incredible that someone with this level of expertise
should misidentify something so common as an oak tree. The point is that mere fallibility alone is not
a good basis for doubt; rather one needs some specific reason to think that one might be mistaken in
this particular case.
We can also see this point in action when we look at scientific practice. Scientists openly recognise
that their enterprise is fallible. Indeed, this fallibility is built into the structure of science. The fact that
even the best confirmed scientific theories could nonetheless be false is what leads scientists to be
constantly testing their theories. Experimental results are repeated by other scientists to check for
anomalies, experimental trials are conducted ‘blind’ in order to screen-out any possible bias in the
interpretation of the results, predictions are extracted from scientific theories and then tested, and so
on. The point of all of this is to expose scientific theorising to as much examination as is feasible in
order to ensure that it is as accurate as possible. Significantly, however, when a theory has been
road-tested in this fashion, then while it is still possible that it might be wrong, there isn’t any longer
any particular reason to be skeptical about it. Again, then, we find that mere fallibility alone is not a
good basis for doubt.
What is Knowledge?
Up until now we have been focussing on our beliefs about the truth. But we don’t just want to
believe the truth, we want to know it. Relatedly, if the skeptic is able to get us to doubt what we
believe, then they thereby ensure that we don’t have knowledge, since if we don’t even believe it,
then we can’t know it. And that really is the ultimate goal of the skeptic: to convince us that we don’t
know much, if anything, of what we take ourselves to know, and thereby in the process get us to
doubt everything that we hitherto took to be true.
The reason why knowing involves more than just believing is that one can form one’s beliefs in all
kinds of inappropriate ways. For example, a lucky guess is not knowledge. Or, to take another kind
of case, imagine someone who is completely gullible, in that they believe everything that they are
told, no matter how ridiculous. Most of what they are told is false, let us say, but every now and again
they are told something true. Since they believe everything they are told, amongst the many
falsehoods that they believe are also the odd truths. But clearly these true beliefs do not amount to
knowledge. One cannot gain knowledge by believing whatever one is told, even if one happens to
believe something true as a result. Gullibility is not a route to knowledge, even if it chances on the
truth.
8
So knowledge demands more than just true belief. Before we ask ourselves what else is required for
knowledge over and above true belief, let us pause for a moment to consider what it means for
knowledge to demand at least true belief.
The first thing to notice is that we are here talking about a particular kind of knowledge, what is
known as propositional knowledge. As the name suggests, this is knowledge of a proposition, where a
proposition is a statement that describes things as being a certain way. For example, that Paris is the
capital of France, or that the square of 2 is 4. When we believe that something is the case (e.g., that
Paris is the capital of France), what we are believing is a proposition. Thus, when we know what we
believe, what we know is a true proposition.
Not all forms of knowledge are propositional. For example, consider ability knowledge, or know-how. I
know how to ride a bicycle, how to swim, how to wire a plug, and many more things besides. Ability
knowledge is clearly very different from propositional knowledge. While I know how to ride a
bicycle, for instance, I couldn’t tell you what it is, exactly, that I’m doing when I’m riding my bike.
And a lot of ability knowledge is like that, if you think about it. Knowing how to do something is
often very different to knowing a bunch of propositions related to that activity. In any case, when we
say that there’s more to knowledge than merely true belief, we are clearly talking about a kind of
knowledge that is propositional, just as belief is propositional.
Next, consider the claim that knowledge requires true belief. Can there not be false knowledge (i.e.,
knowledge of false propositions)? It is certainly the case that someone can quite reasonably suppose
they have knowledge and yet what they think they know is false. This was the situation that many
people were in centuries ago when they quite reasonably thought that the sun orbited the earth rather
than vice versa. But thinking that you know, even reasonably thinking that you know, is not the same
as actually knowing. Since it’s not the case that the sun orbits the earth, then none of those people
centuries ago who believed this actually knew what they believed; they merely thought that they did.
Notice that even though knowledge does demand truth (i.e., true belief), it doesn’t thereby demand
infallibility or even certainty. For example, a lot of our knowledge of the world around us is gained
via our senses, such as by seeing, hearing, touching and so forth—this is known as perceptual knowledge.
Our senses are clearly fallible, in that sometimes they deceive us. Optical illusions might trick us into
seeing something that isn’t there (e.g., as when a straight stick looks bent when placed underwater).
Similarly, when under the influence of drugs one might have hallucinations. The fallibility of our
senses does not prevent us from having perceptual knowledge, however—that is, knowledge that is
acquired via our senses. For while you can’t come to perceptually know that there is an oasis before
you by having an hallucination to this effect, this doesn’t mean that in ordinary conditions when your
(fallible) perceptual faculties are working just fine that they can’t deliver knowledge. As we pointed
out above when we discussed fallibility in the context of relativism, that we sometimes make
mistakes is not a good basis to doubt everything we believe. For the same reason, it is also not a
good basis to doubt whether we know anything either. That our knowledge is fallible does not mean
that it is not bona fide knowledge nonetheless.
9
The same goes for certainty. I may be completely certain of the things that I know, but this isn’t
required, and often it isn’t the case anyway. Right now, for example, I believe that my car is parked
outside of my house. That’s where I parked it earlier (something that I clearly remember), and I live
in a safe area where car theft is very unusual. Plus, only I have the keys to drive it (and no-one else in
our household drives). Thus the likelihood that the car is not presently parked outside on my drive is
very low. If my belief is true, and my car is presently parked outside my house, then I think we would
grant that this is something that I know. But am I completely certain of this claim? Well, while I am
fairly confident of it, I don’t think I would say that I am completely certain. After all, it has been
some hours since I last saw my car on the drive, and although unlikely there are various scenarios
that are not completely far-fetched whereby something has happened to it in the interim (that it has
been stolen, that the hand brake has failed and it has quietly rolled down the road and into the
bushes at the bottom of the street, and so on). That one’s belief isn’t completely certain does not
entail that it fails to amount to knowledge.
So while knowledge demands truth, it doesn’t demand infallibility or certainty. This means that the
skeptic who seeks to deprive us of knowledge needs to do much more than show that our beliefs are
acquired in fallible ways, or that we often are not completely certain of what we believe. So how do
they motivate their skeptical doubt?
We noted a moment ago that knowledge demands more than just mere true belief. This is a good
juncture to revisit that claim, because it is in the gap between knowledge and mere true belief that
skepticism is able to gain some leverage. We noted above that mere true belief can be gained in all
kinds of haphazard and inappropriate ways—for example, through mere gullibility. In such cases it
wouldn’t amount to knowledge. What would it take to turn a true belief into knowledge? There are
various competing proposals on this score offered by contemporary epistemologists, but the general
idea is that knowledge at the very least requires a true belief that is grounded in good reasons.
In particular, knowledge requires good reasons for thinking that the belief in question is true. The
reason for this caveat is that there can be good reasons for believing a proposition which aren’t
thereby good reasons for thinking that what you believe is true. Imagine that someone has put a gun
to your head and tells you that they will shoot you unless you believe in the existence of alien life. In
such a scenario you clearly have a good reason to believe in the existence of alien life, since it’s the
only way to avoid being shot and killed. But that someone will shoot you if you don’t believe that
there is alien life is not a good reason to believe that it’s true that alien life exists. It is rather a merely
prudential reason to believe what you do—i.e., something which it would be prudent for you to believe,
given that you wish to stay alive.
In contrast, if scientists discovered that there is alien life and reported their discovery, then one
would now have a good reason for believing that it’s true that there is alien life. The scientists, after
all, know what they are talking about in this regard, will have made appropriate observations,
collected and verified supporting evidence for this claim, and so on. Reasons for thinking that a
belief is true are known as epistemic reasons. (This is because the area of philosophy that is concerned
10
with truth, knowledge and such like is called epistemology, with its practitioners known as epistemologists).
When we talk about knowledge being true belief that is grounded in good reasons, it is epistemic
reasons that we specifically have in mind⎯i.e., reasons for thinking that the belief is question is true.
It is good reasons of this kind that are precisely missing when it comes to someone forming their
beliefs through guesswork or mere gullibility. If one’s true belief is just down to guesswork, then one
has no good reasons at all for believing what one does, epistemic or otherwise. The gullible person
may think that she has good reasons for believing what she does. This is what someone told her, after
all, and she thinks that one should believe everything one is told, no matter how fantastical. But of
course her reasons for believing what she does are not good epistemic reasons, since being willing to
believe whatever one is told is not a good way to form true beliefs.
We can make this point crystal clear by comparing this way of forming a belief with someone who
isn’t at all gullible, and who hence forms their belief via testimony in an epistemically appropriate
manner—for example, by being careful about whose testimony they accept, reflecting on the
plausibility of what they have been told, and so on. Imagine, for example, someone who believes that
there is alien life because they have listened to the testimony from the scientists, which was widely
reported in all the main reputable news outlets. Such a person, if asked why they believe what they
do, will be able to offer all kinds of good epistemic reasons in support of their belief. For instance,
they could say that the person who told them is someone they know to be reliable about the relevant
subject matter (which is true, as the scientists who work in this field know that their reputations
depend on asserting only what the scientific evidence supports). They could also point out that the
story is running on reputable news outlets, outlets that are renowned for checking their sources. And
so on.
So while gullibility is not a route to knowledge, even when it results in a true belief, one can gain
knowledge from someone else’s testimony so long as one acquires one’s true belief in a careful
fashion. In particular, it is being careful about how one forms one’s belief that will generate the good
epistemic reasons that sets this belief apart from a groundless belief that is formed via gullibility. And
so long as the true belief is appropriately grounded in epistemic reasons, then it can amount to
knowledge.
We can now see how one might go about motivating skeptical doubt. The idea is to show that one
does not have good epistemic reasons in support of one’s beliefs, and hence that one lacks
knowledge. If that’s right, then one’s beliefs, even if true, are no better than someone who forms her
beliefs in a completely gullible fashion (or, for example, via mere guesswork).
A crucial point to notice here is that when skeptical doubt is motivated in this way it is not being
claimed that one’s beliefs are false. Indeed, for all the skeptic has claimed, one’s beliefs might be
entirely true. The point is rather that those beliefs do not enjoy good epistemic grounds, and hence
don’t amount to knowledge. This point is important since it illustrates that skepticism need not be
targeted at the truth of our beliefs at all. Relatedly, going back to our earlier discussion about
relativism, offering a relativistic account of truth, such that the truth is relative to one’s subjective
11
opinion, wouldn’t have any bearing at all on the skeptical problem under this construal (even if we
could find a way to make relativism coherent). The radical skeptic is saying that we lack knowledge of
an objective truth, which is compatible with one having objectively true beliefs that don’t amount to
knowledge, and for that matter also with having subjectively true beliefs that don’t amount to
knowledge. Relativism would thus be entirely irrelevant to skeptical doubt that is targeted specifically
at knowledge of objective truths in this way.
Recall too some of our other points from earlier. In particular, remember we noted that skeptical
doubt wouldn’t be very interesting if it was motivated via the claim that we need to be infallible or
completely certain, as knowledge doesn’t demand either. The same applies to skepticism that is
focused specifically at our epistemic reasons. One can have good epistemic reasons for believing
something true, and thereby have knowledge, even if those reasons are fallible, and even if those
reasons do not suffice to make one completely certain. Indeed, in the case we just gave of someone
who properly forms her belief that alien life exists by listening to the testimony of scientists, the
epistemic reasons are hardly infallible, and our agent is unlikely to be completely certain of what she
believes on this basis. Even so, she can come to have knowledge via forming her true belief in this
way.
Remember too that we noted above that it is not enough to motivate skeptical doubt to argue that
there is a mere possibility that one is mistaken. Just as knowledge can be acquired in a fallible way, so
one can have genuine knowledge even when there is a remote possibility of error. What undermines
knowledge are serious possibilities of error rather than mere possibilities of error. Forming one’s beliefs
via gullibility involves a serious possibility of error, as if one believes everything one is told, no matter
how ridiculous, then one is bound to end up believing falsehoods. But forming a scientific belief by
listening to the testimony of scientific experts, while inevitably fallible, does not involve a serious risk
of error, and hence can be a route to knowledge.
So we are going to be considering a type of skeptical doubt that is directed at the epistemic grounds
we have for our beliefs. Skepticism of this kind claims that we don’t have adequate grounds for our
beliefs, and hence we lack knowledge. This claim is entirely compatible with our beliefs being true,
and so the skeptic isn’t obliged to argue that they aren’t true. But the skeptic does need to show that
their skepticism doesn’t presuppose that knowledge is infallible or that we are completely certain of
what we believe. Relatedly, they need to demonstrate that they are appealing to a serious possibility
of error, and not merely noting that there is a possibility, however remote, that we might be
mistaken. As we will see, even given these constraints, there is a way of motivating skeptical doubt
that satisfies all these conditions.
Before we turn to considering this form of skeptical doubt, however, we will close this module by
considering why it might matter to us that we do have the widespread knowledge that we take
ourselves to have. Could we coherently accept the radical skeptical contention that we don’t know
12
very much with equanimity? I don’t think so. This is a topic that we are going to explore more fully
in module 4, so for now I will just make some preliminary remarks in this regard.
Imagine that the skeptic is right and you don’t know very much. Put this way, the claim might seem
very abstract, but it’s relatively straightforward to bring out the practical ramifications of what the
skeptic is contending. Think about it. You don’t know that your parents are your parents. You don’t
know that your friends are your friends. You don’t know anything about your past, about those
cherished events that you seem to recall, or those achievements that you are so proud of. And so on.
Indeed, as we will see in the next module, the radical skeptic isn’t just claiming that you lack
knowledge, but more specifically that you lack any good epistemic reason for believing what you do.
So it’s not as if one can counter the radical skeptic by maintaining that while one lacks knowledge
one does have good (epistemic) reasons for believing what one does, since if the skeptic is right then
one doesn’t even have the latter. That is, it’s not just that you don’t know that your parents are your
parents, but that you have no good (epistemic) reason for thinking that they are your parents—i.e., no
reason at all for thinking that it’s true that the people you believe are your parents actually are your
parents.
Once we understand what skepticism means in practice, it should be easy to see why conceding the
skeptical conclusion is so problematic. Indeed, wouldn’t living a life on this basis make one’s
existence absurd? How could one’s existence make any sense if one lacks knowledge of even the most
basic facts about one’s life? We will return to this point later on, but for now I just want to register
the fact that radical skepticism appears to have devastating consequences for the meaningfulness of
our own lives. It is in this sense an existential problem. If that’s right, then even if we set aside the
pernicious social consequences of radical skepticism that we noted earlier, it would remain a difficulty
that needs to be taken seriously due to how it threatens to make our existences absurd.
To learn more about the topics covered in this module, see chapter 1 of the book that
accompanies this MOOC, A Very Short Introduction to Skepticism, by Duncan Pritchard
(Oxford: Oxford University Press, 2019).
13
For an introduction to the theory of knowledge that doesn’t presuppose any prior knowledge of the
subject, see Pritchard (2018)—chapters 1 and 18-20 are particularly relevant to issues covered in this
module. See also Nagel (2014) and Pritchard (2016). For a very readable introduction to the problem
of skepticism, including its history, see Gascoigne (2002). For a helpful introduction to some of the
issues regarding truth, including the perils of adopting a relativist account of truth, see Blackburn
(2007). For an excellent and highly engaging account of the use and abuses of science, and thereby
why the scientific method is so important, see Goldacre (2008). For an accessible treatment of how
one might apply epistemology to practical questions, such as how to spot a conspiracy theory, or
how to determine which ‘experts’ one should believe, see Coady (2012). See also and Pritchard (2018,
part four) for further discussion of a range of topics in ‘applied’ epistemology. For a classic
discussion of the absurd, see Nagel (1971).
Module 2
Is Knowledge Impossible?
Introduction
In this module we are going to be looking at an influential argument that purports to show that we
do not know much of what we take ourselves to know. If this argument works, then it licenses a
radical skeptical doubt. Recall that in the last module we noted the difference between a localised
skeptical doubt, which is concerned with only doubting certain specific claims, with a more
generalised, or radical, skeptical doubt which calls much of our knowledge into doubt en masse. As we
saw, there can be all kinds of good reasons why we should be skeptical about specific matters, and
hence localised skepticism can be perfectly healthy from an intellectual point of view. It is rather the
radical form of skeptical doubt that seems to be intellectually problematic. Indeed, we noted that
once it takes root it has the tendency to generate all kinds of ills for public life, whereby people don’t
care any more what is true and what isn’t. In addition, we observed that radical skepticism has the
potential to make one’s own life absurd.
We also saw in the last module, however, that motivating radical skeptical doubt of this kind is not
straightforward. For example, it is not enough to merely appeal to the fact that we are sometimes
mistaken, or more generally to simply note that we are fallible creatures. As we made clear,
knowledge doesn’t demand infallibility, and so that we have bona fide knowledge is entirely compatible
with the fact that we are sometimes mistaken. Relatedly, it is not enough to note that we are often
not completely certain of the things that we believe, since knowledge doesn’t demand complete
certainty either.
If radical skeptical doubt cannot be motivated in these ways, then how might we motivate it? This is
where the argument that we are going to consider comes in. The point of this kind of skeptical
argument is to show that we don’t have any good reasons for thinking that our beliefs are true.
Remember that we called reasons of this kind epistemic reasons, and contrasted them with merely
prudential reasons. We have a prudential reason for believing something when it would be useful for us
to believe it, regardless of whether it is true. Epistemic reasons, in contrast, are specifically reasons
for thinking that the belief in question is true. It is epistemic reasons in support of our true beliefs
that we need if we are to know what we believe. Accordingly, the radical skeptical argument that we
will be examining will be claiming that we lack reasons of this specific kind in support of our beliefs.
15
Remember too that a skeptical argument that is focussed on the epistemic reasons for our beliefs is
not concerned with whether our beliefs are true. That is, skepticism of this kind is not claiming that
our beliefs are generally false, but rather that they don’t generally amount to knowledge (because they
are not supported by good epistemic reasons). But our beliefs could be generally true and yet fail to
amount to knowledge (just as someone who is completely gullible, and who hence doesn’t know very
much, might nonetheless have lots of true beliefs). Whether our beliefs are true, and whether, even if
true, they amount to knowledge, are distinct issues.
Cartesian Skepticism
With all this in mind, let’s start constructing our argument for radical skeptical doubt. This a
contemporary radical skeptical argument that is rooted in the work of René Descartes (1596-1650),
especially in his Meditations on First Philosophy. As such, this form of skepticism is standardly described
as Cartesian, although it departs from the particular skeptical argument that Descartes offered in some
crucial respects. (It is also often called external world skepticism, for reasons that will become apparent).
Note that Descartes wasn’t himself a radical skeptic, and he certainly wasn’t trying to convince
anyone to embrace radical skeptical conclusions. His interest in radical skepticism was rather
methodological. In particular, he wanted to find the secure foundations for knowledge by employing a
‘method of doubt’—i.e., doubting as many of his beliefs as he possibly could. The point of his use of
radical skepticism is thus as a kind of extreme ‘stress test’ applied to our system of belief. Descartes
thought that in this way he could find the indubitable, and thereby certain, fixed-points in this system
onto which our knowledge can be founded. In particular, if we can trace our knowledge back to such
epistemically secure foundations, then we can be assured that it is thereby immune to skeptical
challenges. This sort of epistemological project is thus known as foundationalism, and as should be
clear it is meant to be a decidedly anti-skeptical project, even though it employs skeptical arguments
along the way.
You may well be familiar with the indubitable fixed-point that Descartes claimed that he discovered
using this skeptical methodology, since this is where his famous remark ‘cogito ergo sum’ comes in
(known as the cogito, for short). This is usually translated, albeit not without some controversy, as ‘I
think, therefore I am’. In a nutshell, Descartes’s idea was that while one can doubt most of what one
believes—we will consider just why he claimed this in a moment—one cannot doubt one’s own
existence. After all, in doubting it, one is thereby thinking, and that means that one must be in
existence in order to be doing the thinking in the first place. That one exists is thus meant to be an
indubitable, and hence epistemically secure, foundation for our knowledge. More specifically, what
Descartes was seeking as a foundation for knowledge was certainty, and that the cogito is indubitable
is meant to show that it is something the truth of which we can be certain of.
Interestingly, while Descartes’s brand of foundationalism has not proved to be popular, the radical
skepticism that he set out in order to motivate this style of foundationalism continues to grip the
philosophical imagination. Part of the reason for this is that there are some fairly serious difficulties
16
that afflict the particular foundationalist account that Descartes goes on to develop, not least in terms
of how it essentially appeals to God’s existence (and also to God’s epistemically benevolent nature).
But another part of the reason for the enduring appeal of Descartes’s formulation of radical
skepticism concerns its employment of an important theoretical innovation that is these days known
as a radical skeptical hypothesis.
A radical skeptical hypothesis i s a possibility of error that has two key features. The first is that it is a
scenario that is completely indistinguishable from one’s everyday life. The second is that it is a
scenario in which most of what one believes is false. These two features make radical sceptical
hypotheses very different from the kinds of possibilities of error that one usually encounters.
Consider the mundane error-possibility, noted in the last module, that one’s car, which one believes
to be parked outside, has in fact been stolen. There is nothing indistinguishable about this scenario,
as one only needs to go outside one’s house to verify whether it is true. Moreover, although it calls
some of one’s beliefs into question—most specifically one’s belief about the current location of one’s
car—it doesn’t call one’s beliefs as a whole into question at all. Radical sceptical hypotheses are thus
very different to everyday error-possibilities.
Descartes presents two very vivid radical skeptical hypotheses in his Meditations. The first is the
possibility that one is experiencing a particularly realistic dream. Dreams can be very credible (at least
while they are happening anyway), even to the point that one can be convinced that one is wide
awake. It thus seems at least possible that what one is experiencing right now could be just a
dream—i.e., that it is one of those hyper-real dreams that seem so real that one doesn’t realise that
one is dreaming. If one were in such a dream, then it would be completely indistinguishable from
normal waking life. For example, there would be no point pinching oneself to make sure that one is
awake, since if one is in the dream then the pinching would simply form part of the dreamlike state
itself.
If one were in the grip of such a dream, then although one would seem to be experiencing the world
around one (the ‘external world’) in the usual way, in fact one’s sensory experiences would not be
hooked-up to the external world at all. Instead, they would be the product of one’s imagination.
Given that one’s apparent sensory experiences in one’s dreams are the result of one’s imagination
rather than the external world, one can see how they might end up generating mostly false beliefs
about the external world. It could be, for example, that none of what you take yourself to have
experienced in the past really happened, and that you only believed that it happened because of what
you experienced in the dream. In that case, all your beliefs about these events could be false. (As the
Chinese philosopher Zhuang Zhou (c. 369 – c. 286 BC) put it, many years before Descartes, how is
one to know that one is a human being who is dreaming that one is a butterfly as opposed to being a
butterfly who is dreaming that one is a human being?)
We now have an error-possibility that is indistinguishable from everyday life but where most of one’s
beliefs are false (since one’s beliefs about the external world comprise a large component of your
beliefs as a whole). We thus have a radical skeptical hypothesis. Note too what this radical skeptical
17
hypothesis is calling into question, which is the truth of one’s beliefs about the external world. This is
why Cartesian skepticism is sometimes called external world skepticism.
As just noted, we are all familiar with dreams, and how they might deceive us. For Descartes,
however, the dreaming radical skeptical hypothesis is meant to just soften us up for an even more
extreme, and supernatural, error-possibility. This is the possibility that there is a malevolent demon
who has the power, and inclination, to systematically deceive us, albeit in a way that is completely
hidden from us. We know that our senses are fallible, so why couldn’t it be possible that there is a
creature who is able to ensure that our sensory experience is always misleading, but without this ever
being detectable, even in principle? Perhaps, for example, they can generate within us the sensory
experience of walking in a forest when we are in fact floating on water, or can make us believe that
we are high up on a mountain when in fact we are deep in a dark cave. While incredible, the point is
that this scenario does not seem at all impossible. But with this degree of deception in play there
wouldn’t be any way in which we could distinguish these misleading experiences from the genuine
article. Moreover, most of our beliefs formed on the basis of these deceptive experiences would tend
to be false—e.g., one would believe that one is on top of a mountain when one is in fact inside a
cave.
Thus far, the ‘evil demon’ error-possibility is much like the dream scenario, in that both are calling
the general veracity of our sensory experiences into question, and hence prompting us to doubt the
truth of our beliefs about the external world. The two skeptical scenarios diverge in terms of the
scope of the beliefs that they call into question, however. In particular, the evil demon skeptical
hypothesis seems able to not just make your sensory experiences unreliable, but also (just about) any
belief-forming process that you care to imagine.
For example, one might naturally suppose that one’s confidence in simple arithmetical truths, such as
that ‘4+4=8’, is immune to the dreaming skeptical hypothesis. After all, couldn’t one do these sums
even in one’s dreams, and thereby end up with a true mathematical belief even if all of one’s sensory
beliefs were false? Crucially, however, the evil demon could surely interfere with your mental
processes just as she interferes with your sensory experiences. In this way, she could make it feel like
you are undertaking elementary arithmetical calculations with obviously true conclusions when in fact
you are fundamentally in error. Under the influence of such an evil demon, couldn’t one form the
belief that ‘4+4=2’ with just as much conviction as one would form the belief that ‘4+4=8’ in
ordinary circumstances? If so, then nearly all of one’s beliefs could be undetectably false, where this
means not just one’s beliefs about the external world formed on the basis of apparent sensory
experiences, but also one’s beliefs formed in non-sensory ways, even those that are the result of
apparently impeccable rational processes like basic arithmetic.
The evil demon error-possibility is thus an even more radical skeptical hypothesis than the dreaming
hypothesis, in that it doesn’t just call the truth of one’s beliefs in an external world into question, but
also the truth of one’s beliefs that aren’t about such an external world. Indeed, the only belief whose
truth is not called into question by the evil demon radical skeptical hypothesis seems to be the cogito,
which shows just how extensive the skeptical scope of this error-possibility is.
18
Descartes’ 17th century audience clearly had no problem taking the idea of such a mischievous demon
seriously. In our more secular and scientific time, however, I imagine that many would find this
intellectual device perhaps a bit too fantastical. But this really doesn’t matter because we can easily
come up with error-possibilities that have exactly the same kinds of skepticism-inducing properties
without having to resort to such supernatural entities. Indeed, as we will see, all we need to do is go
to the cinema.
So, for example, here is a contemporary variation on the evil demon radical skeptical hypothesis that
you might be familiar with (in some form or other) from Hollywood movies such as The Matrix. Let’s
imagine that there are evil scientists who go about ‘harvesting’ brains from bodies and keeping those
brains alive in vats of nutrients where they are ‘fed’ experiences by supercomputers. Crucially, the
experiences that these ‘brains-in-a-vat’, or BIVs for short, are given are indistinguishable from one’s
everyday experiences. In addition, the BIVs have no memories of being abducted and put in the vats.
This means that the BIVs are completely unaware that they are now living their lives entirely within
the fake virtual environment created by the supercomputers. So, for example, the BIVs have
experiences as if they are walking around in their normal environment, interacting with friends and
colleagues, driving their cars to the shops, and so on. And yet none of this is actually happening,
since all of these experiences are in fact manufactured by the machines plugged into the vat.
Accordingly, most of what our BIVs believe is false, and hence they don’t know very much, if
anything. They are being systematically deceived, but not by a supernatural being as Descartes
envisaged, but rather by the evil scientists in charge of the technology that they are embedded in.
Notice how similar the BIV radical skeptical hypothesis is to Descartes’s evil demon scenario. Since
the supercomputers are generating fake sensory experiences, so they are blocking-off our sensory
access to the external world. But it isn’t just our sensory experiences that they are fabricating, since
they can fabricate any of our experiences at will. Accordingly, just like the evil demon, they can make
me believe that I have just conducted an impeccable chain of reasoning, such as an elementary piece
of arithmetic, even when I’ve done nothing of the sort. We thus get a skeptical scenario not only
suitable for the kind of external world skepticism that is distinctively Cartesian, but also one that can
call the truth of even our beliefs that aren’t concerned with the external world into question.
Now all this talk of BIVs is all very science fiction, of course. But note that no-one is saying that this
scenario is true, or even likely (just as Descartes wasn’t claiming that there really is an evil demon that
is deceiving you). Right now, the skeptic is merely noting that this hypothesis is possible—i.e., it is
something that could happen, no matter how unlikely it might be. And that is certainly true. After all,
the scenario as described does not depict anything impossible occurring, and so we ought to concede
that it is entirely possible.
19
Moreover, even if there were something about this scenario that might make us think that it is
impossible, then we could just come up with a different radical skeptical scenario. Since the BIV
scenario is found in the movies, let’s pick another radical skeptical hypothesis that we find in a
Hollywood film. This time consider the scenario as depicted in the film Inception. This depicts
fully-coherent dream-like states that people can inhabit in which everything seems completely
normal, when in fact nothing that one is apparently experiencing is actually happening. As we noted
with regard to Descartes’s skeptical dream hypothesis above, it does seem at least possible that all of
your experiences right now could be the product of a dream-like state. If that were so, however, then
most of what you believe would be false. You are not in fact sitting at a coffee shop drinking a coffee
with your friends right now, you didn’t visit your parents yesterday, and you won’t actually be driving
to the coast for the day with your pals tomorrow either—everything is rather just a part of the
dream-like state that you are presently inhabiting.
Or, since we are taking our cue from the movies, how about the radical skeptical scenario involved in
the film The Truman Show? This concerns someone who thinks he is leading a normal life, but who is
in fact the lead character in a TV show. Accordingly, everyone around him is an actor, and the
environment in which he interacts with others is nothing more than a TV set. As before, our hero’s
experiences on the TV set are indistinguishable from experiences one might have in everyday life,
where one is genuinely living a normal life rather than having the whole thing staged for one in a TV
studio. Moreover, a great deal of what he believes is false. He doesn’t live in a normal town, his
friends are not his friends but actors pretending to be his friends, the shops he goes to are not real
shops but sound stages, and so on. And since a lot of what he believes is false, it also follows that he
doesn’t know a lot of what he takes himself to know either.
(Notice that this skeptical scenario is a little different from the dream-like hypothesis depicted in
Inception, where one’s experiences aren’t hooked-up to the external world at all. Instead, although the
protagonist in The Truman Show is genuinely interacting with an external world, he fundamentally
misunderstands the nature of this interaction. In both cases the upshot is the same, in that our
deceived subject knows much less about the external world than she imagines).
The point is that once we understand how radical skeptical hypotheses are set-up, then it is fairly
straightforward to come up with new ones. There is thus very little to be gained by objecting to the
skeptic’s claim that such scenarios are at least possible. In any case, on the face of it, to concede this
is not to concede very much anyway. After all, although it is clear that the BIV, as with other victims
of radical skeptical hypotheses, has lots of false beliefs, and hence doesn’t know very much, why
should that have any relevance for us? Remember that the skeptic hasn’t claimed that we are BIVs
(or the victim of any other radical skeptical hypothesis for that matter). Indeed, she hasn’t even
claimed that such scenarios are likely, so for all that’s been said so far we can treat them as
completely far-fetched error-possibilities (which is what they seem to be). As a result, why should the
fact that the BIV doesn’t know very much have any bearing on what we know?
The skeptic now makes an important move, which is to point out that it is impossible to rule-out a radical
skeptical hypothesis. That is, it seems that one cannot know that one is not the victim of a radical
20
skeptical hypothesis. One cannot know, for instance, that one is not a BIV. For how would one go
about knowing that one is not a BIV, given that the experiences had by the BIV are indistinguishable
from one’s everyday experiences? For example, it would clearly be pointless to note that you can see
and feel your body right now, and hence conclude that you can’t be a bodiless BIV, since of course
the BIV also has the same experiences you do as if they have a body (albeit misleading experiences in
their case). Equally, there would be no point in grabbing objects in one’s environment—picking up a
cup on one’s desk, for example⎯and concluding that one can’t be a BIV since BIVs don’t physically
engage with their environment. This is because, as before, BIVs will have experiences just like you do
as if they are engaging with objects in their environment, even though in fact they are doing nothing
of the kind.
Interestingly, it would also be similarly pointless to respond to the BIV skeptical scenario by arguing
that one has scientific grounds for thinking that there are no BIVs at present. In particular, one
might be tempted to argue that since the current state of technology is not advanced enough for
there to be BIVs, hence we have a rational basis for ruling out this skeptical scenario. A moment’s
reflection reveals that considerations like this have no bearing at all on the issue in hand. After all,
one’s envatted counterpart could be being ‘fed’ similar information regarding the current state of
technological development and so will draw exactly the same conclusion, even though in their case
that conclusion would be false. Accordingly, that one believes that our current science couldn’t
support BIVs is neither here nor there when it comes to evaluating this skeptical scenario.
The upshot is that since radical skeptical hypotheses like the BIV scenario are indistinguishable from
everyday experiences, it follows that we cannot rule them out. And that in turn means that we are
unable to know that we are not the victims of such scenarios. Put another way, for all we know, we
could be BIVs (or in the grip of a hyper-real dream, or being deceived by an evil demon, and so on).
So the skeptic has now made two moves. The first has been to describe what a radical skeptical
hypothesis is, and to argue that such hypotheses depict possible scenarios. The second has been to
argue that we cannot rule-out these radical skeptical hypotheses, and hence that do not know that
they are false.
Has the skeptic done enough to motivate their skeptical conclusion that we don’t know much of
what we take ourselves to know (such as regarding the external world)? If one held that knowledge
demands that one is infallible or completely certain, then this would be a plausible conclusion to
draw. After all, the skeptic has alerted us to a class of error-possibilities that we cannot rule-out,
error-possibilities that call into question the truth of much of what we believe. Clearly, then, we
cannot credibly regard our everyday knowledge as being infallible, since we have just described a
skeptical scenario in which we would believe exactly what we do now but believe falsely (e.g., if one
actually were a BIV). Moreover, since we know that we cannot rule out these radical skeptical
error-possibilities, then how can we be absolutely certain that what we ordinarily believe is true?
Relatedly, if one is engaged in the epistemological enterprise of seeking a foundation for knowledge
that is indubitable, and hence absolutely certain, as Descartes was, then one is not going to hold that
one’s everyday beliefs could ever fulfil that role.
21
But as we saw in the last module, however, it doesn’t seem at all credible that the bar for knowledge
should be set so high as to demand infallibility or absolute certainty (or, for that matter,
indubitability). On the contrary, our everyday conception of knowledge seems perfectly happy with
the idea that knowledge can be fallible and not absolutely certain (and thus to a degree dubitable)
while being bona fide k nowledge nonetheless. What matters is that we have good epistemic reasons for
believing what we do, and we can have such reasons even in the absence of infallibility and absolutely
certainty. If that’s right, then insisting that knowledge is lacking would be akin to illicitly changing the
subject. The skeptic would in effect be arguing that we lack knowledge relative to a specific, and
highly demanding, account of ‘knowledge’ that is radically distinct from how we in fact use the
notion of knowledge in everyday discourse. But why should we care about that? To use an analogy
that a contemporary philosopher has offered, it would be like coming across someone who claims
that there are no doctors in New York City on the grounds that by ‘doctor’ they mean someone who
can cure any disease within 24 hours. No-one would be impressed by ‘doctor’ skepticism of this kind.
Why should radical scepticism about knowledge be any different?
The crux of the matter is that the mere fact that there are these radical skeptical hypotheses involving
massive error that we cannot rule-out doesn’t itself show that there is anything epistemically amiss
with our everyday knowledge. At most, it reminds us that such knowledge is fallible, and hence that
we shouldn’t be absolutely certain of what we believe. But we knew that already. As we noted, the
mere possibility of error does not suffice to undermine knowledge provided that this possibility is
remote. And for all the radical skeptic has said otherwise, these radical skeptical hypotheses are
depicting remote possibilities of error. For remember that the skeptic has not claimed that the radical
skeptical hypotheses they propose are likely, or even that they are true. So what is to stop us from
treating them as the far-fetched error-possibilities that they appear to be? If that’s right, then there
seems nothing amiss with supposing that we know most of what we take ourselves to know. It is just
that there are some error-possibilities—those picked out by radical skeptical hypotheses—that we
cannot know to be false.
Unfortunately, as we will now see, the radical skeptic is only just getting warmed up. In particular,
there is a way of supplementing the radical skeptical hypotheses with a further very credible principle
in order to serve the skeptic’s ends. In particular, this allows the radical skeptic to motivate her
skeptical conclusion without having to appeal to the idea that knowledge demands infallibility or
absolute certainty.
Here is the state of play. The radical skeptic has argued that we cannot know that we are not victims
of radical skeptical hypotheses, like the BIV scenario. But on the face of it that seems entirely
compatible with the idea that we know much of what we take ourselves to know—that I know where
I live, what I ate for breakfast today, what jersey I’m currently wearing, and so on (which are all
22
things, by the way, that the BIV doesn’t know). So how can the radical skeptic use the claim that we
can’t rule-out radical skeptical hypotheses to motivate her skeptical doubt?
This is where the radical skeptic introduces a principle that at first blush can seem entirely innocuous,
but which, as we will see, greatly strengthens their hand. Consider the following reasoning. Suppose
you know that the name of the capital of France begins with the letter ‘P’. Then aren’t you also in a
position to know that Madrid is not the capital of France (since ‘Madrid’ doesn’t begin with the letter
‘P’)? Or take another example. Suppose you know that the single most populous country on the
planet is China. Then aren’t you also in a position to know that India is not the most populous
country on the planet (given that there can be only one country that is the single most populous
country on the planet)?
The reasoning at issue here looks harmless, and that’s precisely because the inferences in play are so
obvious, given what you know. What such inferences illustrate is that if you know one proposition
(e.g., that the capital of France begins with the letter ‘P’, or that China is the single most populous
country on the planet), and you know that this proposition entails a second proposition (e.g., that
Madrid is not the capital of France, or that India is not the single most populous country on the
planet), then you know that second proposition too. What could be more innocuous than that?
The principle at issue here is the idea that knowledge is preserved, or ‘closed’, under known
entailment. This is why it is often called the closure principle for short. It is hard to see how such a
principle could possibly fail. How could it be that one knows one proposition, knows that this entails
a second proposition, and yet fail to know that second proposition? After all, precisely what it means
for one proposition to entail another is that if the entailing proposition is true, then the entailed
proposition must be true also. So if it is true that China is the single most populous country on the
planet, then it must also be true that India is not the single most populous country on the planet. But
if you know that something is true, and know that if this is true then something else must be true,
then how could you fail to know that this second claim is true as well? For example, how could one
know that China is the single most populous country on the planet, and know that if China is the
single most populous country on the planet then India is not the single most populous country on
the planet, and yet fail to know that India is not the single most populous country on the planet?
(Indeed, imagine someone who claimed to know both that China is the single most populous country
on the planet and that this entails that India is not the single most populous country on the planet,
but who nonetheless maintained that she didn’t know that India is not the single most populous
country on the planet. Does this even make sense? Wouldn’t we think that she must be confused in
some way, such as that she doesn’t really understand what she is claiming?)
The closure principle thus looks very compelling. The problem, however, is that the radical skeptic
can employ this seemingly harmless principle to motivate their radical skeptical doubt. For notice
that just about any everyday claim that you take yourself to know is going to be inconsistent with
some radical skeptical hypothesis or other. For example, right now I take myself to know that I am
wearing a shirt. I can see that I am wearing a shirt, I can feel it on my skin, I remember putting it on
23
this morning, other people can tell me that I am wearing a shirt, and so on. There is thus an
abundance of evidence that I am wearing a shirt. But if it is true that I am wearing a shirt, then it is
also true that I am not a BIV. After all, BIVs don’t have bodies, and so they can’t wear shirts. Thus it
can’t both be true that I am wearing a shirt and that I am a BIV. Put another way, that I am wearing
a shirt entails that I am not a BIV.
The difficulties start, however, once we begin to plug this entailment into a closure-style inference.
For suppose I do know that I am wearing a shirt. We’ve just noted that I also know that wearing a
shirt entails that I am not a BIV. But we have also agreed with the skeptic that I cannot rule out the
BIV skeptical hypothesis, and hence do not know that I’m not a BIV. Accordingly, it seems that I
know that I’m wearing a shirt, know that if I’m wearing a shirt then I’m not a BIV, but I don’t know
that I’m not a BIV. And that sounds entirely wrong. How can I know that I’m wearing a shirt, but
not whether I’m a shirtless BIV? Given the plausibility of the closure principle, it thus seems that if I
did know that I’m wearing a shirt, then I must be able to know that I’m not a BIV. Conversely, if it
really is impossible to know that I’m not a BIV, then it must also be impossible to know that I’m
wearing a shirt. With the closure principle in play, then, the radical skeptic can thus use our inability
to rule-out radical skeptical hypotheses to undermine our everyday knowledge, such as one’s
seemingly mundane knowledge that one is wearing a shirt.
And notice that what goes here for knowing that one is wearing a shirt applies to one’s knowledge of
a myriad range of everyday claims. For example, knowing that one is driving one’s car or that one is
playing the violin are both inconsistent with being a BIV, and hence one could run exactly the same
reasoning to show that such ‘knowledge’ is completely illusory. Moreover, as we noted above, the
BIV radical skeptical hypothesis is but one skeptical scenario, as we can easily cook up other radical
skeptical hypotheses. There is thus nothing stopping the radical skeptic from using the closure
principle to motivate skeptical doubt of a wide range of everyday claims that we think we know.
You should now be starting to see the shape of the puzzle. Prior to considering the closure principle,
it seemed that one could grant that one is unable to know the denials of radical skeptical hypotheses
without this having any implications for skepticism. This was because one’s knowledge of everyday
claims seemed to be unaffected by one’s inability to rule out such fantastical scenarios. But once we
realize that our everyday beliefs are in conflict with radical skeptical hypotheses, such that the truth
of the former entails the falsity of the latter, then this situation ceases to be stable. If we really do
know the everyday claims that we take ourselves to know (e.g., that one is wearing a shirt), then,
incredibly, it seems we must be able to know the denials of radical skeptical hypotheses after all (e.g.,
that one is not a BIV). Alternatively, insofar as the skeptic is right that we are unable to know the
denials of radical skeptical hypotheses, then it seems we must also be unable to know these everyday
claims as well. That is, if I really don’t know that I’m not a BIV, then I can’t know that I’m wearing a
shirt, and any other everyday claim that is inconsistent with a radical skeptical hypothesis. With the
closure principle in play, therefore, the skeptic seems able to motivate a radical skeptical doubt.
In fact, once we reflect on the matter, this skeptical conclusion ought not to be that surprising. We
have noted that knowledge is true belief that is grounded in good epistemic reasons. My putative
24
knowledge that I am wearing a shirt is, so I thought, grounded in such good epistemic reasons as that
I can see the shirt, I can remember putting it on, I can feel it on my torso, and so on. But if I am a
BIV, then I would seem to have all the same epistemic reasons and yet my belief that I am wearing a
shirt would be false. Indeed, this is precisely the reason why I can’t know that I’m not a BIV, in that I
have no good epistemic reasons for thinking that this is the case (on account of the fact that the
BIV’s experiences are indistinguishable from normal experiences). So how then can I have good
epistemic reasons for believing that I am presently wearing a shirt, and yet have no good epistemic
reasons for believing that I am not a (shirtless) BIV? The skeptical point is clear: insofar as you really
have no good epistemic reasons for believing that you are not a (shirtless) BIV, then you don’t really
have any good epistemic reasons for believing that you are wearing a shirt either. What you have are
merely apparently good epistemic reasons for believing everyday claims such as that you are wearing a
shirt, but since they are not reasons that distinguish between the everyday scenario (wearing a shirt)
and the incompatible skeptical scenario (being a shirtless BIV), they cannot be genuinely good
epistemic reasons at all. In short, while you might think that you have good epistemic reasons for
believing many of the things that you believe, in fact you don’t have any good reasons for believing
these things at all. And that’s why you don’t know many of the everyday claims that you take yourself
to know.
This highlights an important point about radical skepticism of this kind. The skeptical claim is not
merely that one does not know many of the things that one takes oneself to know. Rather, it is the
stronger thesis that one does not have any good epistemic reason for believing many of the things
that one takes oneself to know. The former claim is compatible with one having some good epistemic
reasons for believing what one does, but where these reasons are insufficient for knowledge. The
radical skeptical contention in play is, however, much more dramatic: that one has no good epistemic
reason at all for believing even such apparently mundane everyday claims as that one is presently
wearing a shirt. Accordingly, if your beliefs in this regard happen to be true, then that is just a matter
of luck from an epistemic point of view, just like it’s purely a matter of luck that the gullible person
who believes anything they are told might end up with some true beliefs.
The upshot is that it seems that if we want to have good epistemic reasons for believing the everyday
things that we take ourselves to know, then we need to be able to rule-out radical skeptical
hypotheses. And since we cannot do the latter, it follows that we don’t have good epistemic reasons
for much of what we believe, and hence we can’t know what we believe either. By employing the
closure principle the radical skeptic appears to have been able to turn the apparently harmless claim
that we can’t rule out radical skeptical scenarios into something with important skeptical
consequences. Even though the skeptic has offered us no reasons for thinking that the radical
skeptical hypotheses are true, or even likely, she is still able to use these scenarios to extract radical
skeptical conclusions. Note too that there is no appeal to infallibility (or complete certainty) going on
here. If the radical skeptic is right that we don’t have any epistemic reasons at all in support of our
everyday beliefs, then it follows that we lack even fallible (or somewhat uncertain) knowledge of
these everyday claims.
25
We are now in a position to put all the moving parts of this formulation of radical skepticism
together. In essence, the radical skeptic maintains that the following three claims are inconsistent,
which means that they cannot all be true (i.e., at least one of them must be false):
We have seen that (1) is eminently plausible. This is because radical skeptical hypotheses are precisely
characterized in such a way that they are indistinguishable from normal life. Life in the vat for the
BIV is exactly like a normal life out of the vat, even though most of the BIV’s beliefs are false. (2) is
also very plausible. Surely, as the closure principle demands, if one knows one proposition, and
knows that it entails a second proposition, then one must know the second proposition? But, as we
have seen, with (1) and (2) in play it seems that (3) becomes unsustainable. For if we did know these
everyday claims (such as that one is wearing a shirt), then given that we know that they are
inconsistent with radical skeptical hypotheses (e.g., that one is a BIV), it would seem to follow, via
the closure principle, that we must be able to know the denials of radical skeptical hypotheses. But
given that we cannot know the denials of radical skeptical hypotheses, it follows that we don’t know
the everyday claims either. And yet we do ordinarily suppose that we have lots of everyday
knowledge, in line with (3). Indeed, as we noted above, it seems that our inability to know the denials
of radical skeptical hypotheses means that we have no good epistemic reasons at all for our everyday
beliefs. Thus, we can’t consistently endorse all three of (1), (2), and (3).
I have deliberately expressed the skeptical challenge as an inconstant set of claims in order to
highlight two distinct ways of characterizing the problem of radical skepticism. We can think of
radical skepticism as either being a position, or as merely posing a paradox. So, for example, radical
skepticism understood as a position would respond to the inconsistent set of claims just listed and
conclude that since (1) and (2) are clearly true, hence it should be (3) that we reject. In particular,
radical skepticism as a position would conclude that from (1) and (2) it follows that we do not have
much of the everyday knowledge that we take ourselves to have, and hence that (3) is false.
Interestingly, however, the radical skeptic doesn’t need to go this far in order to intellectually disquiet
us. It is enough, after all, to notice that we seem to be independently committed to each of (1), (2)
and (3), and yet they can’t all be true. That is, all three of these claims appear to be rooted in our
ordinary way of thinking about knowledge, which is why we find each of these claims so compelling.
But there must be something amiss with our ordinary conception of knowledge if it leads to
inconsistent claims in this fashion.
This way of characterizing the problem of radical skepticism conceives it as a paradox rather than a
position. On this rendering, the radical skeptic doesn’t tell us which of the three claims that we
should give up, but only that we are committed to all three of them even though they can’t all be
26
true. There is a rich history in philosophy of posing paradoxes of this kind. That is, in thinking hard
about our most fundamental everyday concepts⎯time, freedom, causation, and so on—and showing
that they seem to generate inconsistent claims.
There are advantages to proposing the radical skeptical challenge in terms of a paradox rather than a
position. If one opts for the latter strategy, after all, then it is open to us to query the coherence of
this stance. We might ask the radical skeptic who denies (3), for example, how they are meant to live
their lives. How plausible is it that one can go about their daily existence without supposing that one
knows a great deal? As we noted at the end of module 1, it is hard to even make sense of the idea
that one can live one’s life completely deprived of knowledge without one’s existence becoming
absurd as a result. At the very least, the radical skeptic seems to owe us an explanation of how living
the radically skeptical life is even feasible.
If one proposes radical skepticism as a paradox, however, then one is under no such obligation to
explain oneself. After all, the paradox is simply composed of claims that we all seem compelled to
accept. The tension is thus arising exclusively out of our own commitments. The onus is therefore
on ourselves to work out how we are going to resolve this inconsistency in our thinking about
knowledge, preferably in such a way that avoids the skeptical option of rejecting (3).
In any case, notice that this skeptical challenge, whether formulated as a paradox or as a position,
satisfies the desiderata that we previously laid down for skeptical doubt. In particular, there is no
appeal to infallibility or complete certainty in this argument, but only to our regular, fallible,
conception of knowledge. Relatedly, it is not the mere possibility of error that is generating the
skeptical doubt, but rather the combination of this possibility of error and the closure principle. It is
the latter that is ensuring that the radical skeptical hypotheses, while apparently remote
error-possibilities, are nonetheless relevant to determining whether we have knowledge of everyday
claims. This is because of how we are aware that those everyday claims are inconsistent with radical
skeptical hypotheses. It is also worth reminding ourselves that the skeptic is not claiming that these
radical skeptical hypotheses are true (e.g., they are not claiming that we are BIVs), or even that we
have any rational basis for thinking that such hypotheses are likely. Given the closure principle, just
so long as these scenarios are possible and we are unable to rule them out (i.e., know them to be
false), then their inconsistency with our everyday beliefs will be enough to motivate the radical
skeptical challenge.
We thus seem to have a genuine radical skeptical argument on our hands. It is an argument that
purports to call our widespread knowledge of the world around us into question. In particular, the
argument isn’t directed at the truth of our beliefs, but rather at the epistemic grounds that we have
for those beliefs, even if true. Since, as we noted above, there is more to knowledge than mere true
belief, in that knowledge additionally demands epistemic grounds for one’s true belief, it follows that
so long as the skeptic can undermine these grounds for belief then she can thereby undermine our
knowledge. Indeed, if this radical skeptical argument works, then we not only lack knowledge of
much of what we believe, but we don’t even have any good epistemic reasons for believing what we
do.
27
In the next module we will consider some responses to this formulation of radical skepticism.
To learn more about the topics covered in this module, see chapter 2 of the book that
accompanies this MOOC, A Very Short Introduction to Skepticism, by Duncan Pritchard
(Oxford: Oxford University Press, 2019).
28
For an accessible introduction to the problem of radical skepticism, see Pritchard (2016, ch. 6). For
an in-depth account of the structure of skeptical arguments, see Pritchard (2015, part one). If you
want to read more about philosophical paradoxes, then a great place to start would be Sainsbury
(2009). For an entertaining collection of philosophical discussions of the film The Matrix, which is
probably the closest Hollywood has got to presenting a BIV-style radical skeptical scenario, see Grau
(2005).
Grau, C. (2005). Philosophers Explore the Matrix, Oxford: Oxford University Press.
Pritchard, D. H. (2016). Epistemology, L ondon: Palgrave Macmillan.
Pritchard, D. H. (2015). Epistemic Angst: Radical Skepticism and the Groundlessness of Our Believing,
Princeton, NJ: Princeton University Press.
Sainsbury, M. (2009). Paradoxes, (3rd Edn.), Cambridge: Cambridge University Press.
29
Module 3
Defending Knowledge
In this module we are going to be considering some responses to the skeptical problem that we
formulated in module 2. Let’s start with a very natural response to philosophical puzzles like this,
which is to insist on our commonsense principles and work back from there. In short, if radical
skepticism conflicts with commonsense—and surely it does—then isn’t that a sound basis for
dismissing radical skepticism?
That’s certainly a very reasonable place to start in our dealings with radical skepticism, since what
could be more reasonable than commonsense? Indeed, the idea that commonsense should be
theoretically privileged in this way also has a decent philosophical pedigree, in that it has been
endorsed in various forms by prominent philosophers over the years, including Thomas Reid
(1710-96) and G. E. Moore (1873-1958). One of the difficulties facing this approach to radical
skepticism, however, is that the skeptical puzzle that we are engaging with doesn’t seem to rest on
anything but commonsense. That is, the three claims that we have noted as being inconsistent above
all appear to arise out of our ordinary ways of thinking about knowledge. For example, as we noted
in the last module, none of these claims obviously presuppose contentious ideas about knowledge
such as that it is infallible or demands complete certainty. If that’s right, then it seems that it is
commonsense itself that is generating the skeptical puzzle, by virtue of the fact that our own
commonsense conception of knowledge leads to contradiction.
Now one might counter this by noting that although the inconsistent claims (1) to (3) that make up
our puzzle are all individually compelling, they nonetheless do appeal to notions that we don’t
ordinarily find in our everyday practices. In particular, we do not normally even consider radical
skeptical hypotheses in our day-to-day life. With that in mind, could we use an appeal to
commonsense to coherently argue that the esoteric nature of radical skeptical scenarios means that
we shouldn’t have to rule them out in order to have widespread everyday knowledge?
There are a number of difficulties with this suggestion. To begin with, it is far from clear that radical
skeptical hypotheses really are that esoteric. After all, as we noted previously, such scenarios are a
mainstay of Hollywood movies, including some blockbusters, so how plausible is it that they are
somehow at odds with commonsense thinking? Aren’t they in fact rather familiar possibilities of
error?
30
But even if such scenarios really were as arcane as this proposal suggests, that wouldn’t by itself
demonstrate that radical skeptical scenarios are contrary to commonsense ways of thinking. In order
to see this point we need to remind ourselves that in everyday life we are often constrained in lots of
incidental ways. For example, when evaluating whether a belief amounts to knowledge we simply
might not have the time to consider whether all the relevant possibilities of error have been excluded,
or may even lack the imagination to come up with the full list anyway. Moreover, our everyday
practices of evaluating knowledge will inevitably involve some short cuts and rules-of-thumb, as we
navigate ourselves around the practical hurdles that face us in day-to-day life.
Crucially, however, when we do philosophy it is usually important to set aside these purely practical
constraints, since they are normally irrelevant. This is especially so in the field of philosophy that
concerns us, which is epistemology (i.e., the area of philosophy devoted to questions about truth,
knowledge and related notions). Suppose, for example, that it turned out that when we consider, as
epistemologists, the basis for a particular knowledge claim we discover that in fact it is very shaky
due to the fact that there is a relevant possibility of error that most people in everyday life are simply
unaware of and hence don’t consider. (As it happens, this is not a merely theoretical possibility, as
cognitive scientists have demonstrated that there are a whole range of cognitive biases that influence
our reasoning, often negatively, of which we are often completely unaware). Would we conclude
from the fact that in everyday life we don’t consider this possibility of error that therefore we shouldn’t
consider it, and hence that the knowledge claim in question is entirely secure? Surely not. Instead we
would argue that while in fact we don’t consider this possibility of error, we ought to.
The point is that even if we don’t in fact consider radical skeptical hypotheses in our day-to-day lives,
that doesn’t by itself suffice to demonstrate that we ought not to. In any case, even if we granted that
there was something problematic about our consideration of radical skeptical hypotheses from a
commonsense perspective, it isn’t obvious how this would help us resolve the problem of radical
skepticism anyway. After all, the closure principle is surely rooted in our commonsense thinking
about knowledge. But if that’s right, then it seems to follow that we can only have knowledge of
everyday claims where we are also able to know whatever is entailed by these everyday claims. For if
it turns out that we can’t know the latter, then, via closure, it will follow that we can’t know the
former either. So, for example, if I can’t know that I am not presently shirtless, then I can’t know
that I’m presently wearing a shirt, since of course wearing a shirt entails that one is not shirtless.
Knowing that one is presently wearing a shirt thus demands, given closure, that one can know that
one is not presently shirtless.
We have just stated this implication of the closure principle without mentioning radical skeptical
hypotheses, so nothing we’ve expressed so far ought to be problematic from the perspective of
commonsense. But whether we mention radical skeptical hypotheses or not, it ought to be clear that
this implication of the closure principle will just as much apply to these scenarios. After all, one way
that I could be shirtless is to be a BIV who lacks a body to put a shirt on. So if in order to know that
I am presently wearing a shirt I need to be able to rule-out scenarios where I am shirtless, then that
means I need to be able to rule out the BIV radical skeptical scenario. It follows that regardless of
whether one explicitly mentions radical skeptical hypotheses, that one needs to rule them out in
31
order to have everyday knowledge does seem to be a consequence of our commonsense ways of
thinking about knowledge.
Simply appealing to commonsense thus don’t seem to offer much that is tangible when it comes to
responding to the skeptical puzzle. A more subtle way of employing commonsense in this regard was
famously offered by G. E. Moore, writing almost a century ago. His idea was that when philosophy
and commonsense conflict, then we are entitled to push back against philosophy by appealing to
commonsense. How would this idea play out with regard to the radical skeptical puzzle in hand?
Here is Moore’s proposal, in outline. From a commonsense perspective, (3) is clearly sacrosanct, as
giving up on that would be a disaster. The closure principle, and hence (2), also looks very secure. So
given that we need to reject at least one of the claims (1) to (3), how about we embrace (3) and (2)
and hence conclude that (1) must be false? Skepticism as a position, recall, advocates accepting that
we can’t know the denials of radical skeptical hypotheses (1), and hence concluding, via the closure
principle (2), that we don’t know much of what we ordinarily take ourselves to know (i.e., they
conclude that (3) is false). But Moore’s suggestion is to use this reasoning in reverse. That is, we
should embrace the idea that we do know the everyday things that we take ourselves to know (3),
and hence, via the closure principle (2), we should conclude that we do know the denials of radical
skeptical hypotheses after all (i.e., we should reject (1)).
In proposing this Moore grants two crucial points to the radical skeptic. The first is that the skeptic’s
reasoning is just as good as our reasoning. That is, he isn’t arguing that the radical skeptic’s way of
denying (3) is wrong. He is just saying that we can equally argue in the opposite direction and claim
that (1) is false. In effect, the thought is that we have a kind of dialectical stand-off here between
philosophy and commonsense, but that in the presence of such an impasse we are entitled to go with
commonsense and thus reject the philosophical (i.e., skeptical) conclusion that we lack everyday
knowledge.
The second concession that Moore makes to the radical skeptic is to grant that he can’t give an
explanation of why (1) is false. Moore’s claim is just that commonsense dictates that it must be false.
After all, all parties to this dispute agree that at least one of the three claims that make up the radical
skeptical puzzle must be false. And notice that denying any one of these three claims⎯including, for
that matter, denying (3)⎯generates mystery, given how plausible each of them is when taken
independently. So Moore’s contention is that while the falsity of (1) is mysterious, since the falsity of
any of these three claims would be mysterious, this is not in itself a count against rejecting (1).
How plausible is this Moorean response to the skeptical puzzle? Remember that in the previous
module we made the distinction between radical skepticism as a paradox and radical skepticism as a
position. Construed in the latter way, radical skepticism involves actively claiming that we lack
everyday knowledge—i.e., it involves denying (3). Construed in the former way, as merely a paradox,
however, then the radical skeptic doesn’t argue for any specific claim at all, but merely highlights that
we are independently committed to three claims that are inconsistent. As we noted in the last
module, radical skepticism construed as a paradox has fewer dialectical burdens than radical
32
skepticism as a position, not least because it doesn’t have to explain how one might coherently
embrace a radial skeptical conclusion.
This distinction between radical skepticism as a paradox and as a position is very important when it
comes to evaluating the persuasiveness of Moore’s anti-skeptical stance. Insofar as we focus on
radical skepticism as a position, then Moore’s response can look quite plausible. Both Moore and the
skeptic are claiming something highly counterintuitive—that we can know the denials of radical
skeptical hypotheses, and that we know very little of what we take ourselves to know, respectively.
Moreover, there does seem to be a fairly even dialectical stand-off when it comes to choosing
between these two positions. The skeptic grants (1) and (2), and hence denies (3), while the Moorean
grants (3) and (2), and hence concludes that (1) should be rejected. If the options do come down to a
clash between philosophy and commonsense in this way, then why not opt for commonsense and
hence reject the skeptical position that involves denying (3)?
The problems for Moore’s approach start to become apparent, however, once we shift our attention
to radical skepticism construed as a paradox rather than a position. This is because we are not now
trading off the virtues and vices of one proposal in light of a competing radical skeptical proposal.
Instead, we are faced with a fundamental tension apparently arising out of our own conception of
knowledge and trying to work out how best to resolve this tension. With that way of thinking about
radical skepticism in mind, what possible philosophical comfort can we possibly gain by being told
that (1) must be false, such that we can know the denials of radical skeptical hypotheses, if we are not
also told how this could ever be the case? Remember, after all, that we have already seen that
intuitively we cannot know the denials of radical skeptical hypotheses, which is why we ended up
endorsing (1) in the first place.
This means that if we are to turn Moore’s appeal to commonsense into a plausible anti-skeptical
proposal—one that has application even to the radical skeptical paradox at any rate—then we need
to combine it with an account of how (1) could be false. In particular, how could it be possible for us
to know the denials of radical skeptical hypotheses? The challenge here is to explain how such
knowledge could be possible given that radical skeptical hypotheses are in their nature
indistinguishable from our everyday experiences. That’s a tall order. Indeed, even if we could offer
such an account, it would on the face of it involve a highly revisionary account of knowledge. The
worry would then be that such a revisionary account is even less plausible than the claims that
generated the radical skeptical puzzle in the first place. The concern would be that we remain more
convinced by the radical skeptic’s contention that there is something deeply amiss with our ordinary
ways of thinking about knowledge than with the revisionary anti-skeptical proposal on offer.
A different kind of response to the skeptical problem involves the idea that perhaps there is some
sort of context-shift in play in the skeptical reasoning. It can certainly feel as if something of this kind
is happening when one first engages with the problem. As we just noted in our discussion of
33
skepticism and commonsense, we don’t usually even consider radical skeptical hypotheses in our
day-to-day life. Could it therefore be the case that the radical skeptic is somehow illicitly raising the
standards for knowledge, and that this is what is generating the skeptical problem?
Here is one way of putting some flesh on the bones of this idea. Perhaps ‘knows’ is a
context-sensitive term. The idea would be that sometimes this term is associated with a very
undemanding epistemic standard, and hence is easy to satisfy, while at other times it is associated
with a very demanding epistemic standard, and hence is hard (if not impossible) to satisfy. This could
explain why our everyday usage of this term involves us ascribing lots of knowledge to each other, as
the thought would be that in day-to-day contexts we use ‘knows’ in an undemanding way. Perhaps
what happens when we engage with radical skepticism, however, is that we shift from using ‘knows’
in this undemanding way to employing it in its more austere rendering. It would thus be no surprise
that we find ourselves no longer widely attributing knowledge, since ‘knows’ now means something
far more restrictive than it did previously. Could the radical skeptical problem really turn on a shift of
context of this kind?
It is certainly true that some of the words that we use are inherently context-sensitive, in the sense
that one needs to know specific information about the situation in which they are used in order to
work out what is meant. Perhaps the clearest example of this are indexicals, like ‘I’, ‘here’, and ‘now’.
When I say that ‘I am hungry’, I mean that Duncan Pritchard is hungry, but when you make the very
same assertion you are making a claim about you and not about me. Understanding statements
involving the word ‘I’ thus requires one to know who is speaking. Similarly, in order to understand
what is meant by an assertion involving ‘here’ or ‘now’, it will be important to know when and where,
respectively, the assertion was made. That expressions involving indexicals have this feature explains
why no-one thinks that two people are disagreeing if one person says ‘I am hungry’ and the other
person says ‘I am not hungry’. If the same person made both assertions one after the other, then we
would be puzzled, since they would seem to be contradicting themselves. But if two people make
these assertions then there is no contradiction, since the ‘I’ in each case refers to a different person.
So there is clearly a precedent for there being terms in our language that are context-sensitive in
roughly the way that it is being suggested that ‘knows’ might be. Indexicals are perhaps not the best
examples to focus upon in this regard, however, as there is no appeal to standards in play here. In
terms of ‘I’, for example, all that matters is who is speaking. But other context-sensitive terms do
involve an appeal to standards. Take a term like ‘tall’, for example. I’m a couple of inches over six
feet in height, and hence in most contexts I would count as tall. But in the context of picking a
basketball team, however, I probably wouldn’t count as tall at all. There is no contradiction here—it
is not that I am simultaneously both tall and not tall. Rather, our assessments of tallness are made
relative to a particular standard. At a couple of inches over six feet in height, I am tall relative to the
standard of average person height, but that’s not tall relative to the standard of average basketball player
height. What goes here for ‘tall’ also applies to a range of other context-sensitive terms, like ‘big’,
‘heavy’, ‘wide’, ‘empty’, ‘flat’, and so on.
34
Could what is happening here with a term like ‘tall’ be what is occurring when we engage with radical
skepticism? Just as it is ordinarily true to say that I am ‘tall’, since it is relative to an undemanding
standard for tallness, so it is ordinarily true to say that we ‘know’ lots of things, since that is relative
to an undemanding standard for knowledge. But just as it is not true to say that I am ‘tall’ relative to
the demanding standards for tallness employed by (say) a basketball coach, so it is not true to say that
I ‘know’ a lot of things relative to the more demanding standards for knowledge employed by the
radical skeptic.
One advantage of this contextualist response to radical skepticism is that there is a sense in which both
we’re right and the skeptic is right. After all, our ordinary practices of ascribing knowledge to each
other come out as correct, as they are implicitly picking out an undemanding standard for ‘knows’.
But the radical skeptic is also correct to contend, relative to the demanding standard for ‘knows’ that
she employs, that we lack a lot of the knowledge that we take ourselves to have. This looked like we
were contradicting each other, but insofar as ‘knows’ is a context-sensitive term then there no more a
contraction in play here than you are contradicting me when you say ‘I am not hungry’ just after I
have declared ‘I am hungry’.
Relatedly, the contextualist also has a way of claiming that, properly understood, the three claims that
make up the radical skeptical paradox above are not really in conflict. Take (3), the claim that we
have lots of everyday knowledge. According to contextualism, this is true relative to the everyday
standards for ‘knows’, but false relative to the high standards for ‘knows’ employed by the radical
skeptic. The contextualist can also explain why (1), the claim that we are unable to know the denials
of radical skeptical hypotheses, is so compelling. After all, since (1) explicitly invokes radical skeptical
hypotheses, it already brings in the problem of radical skepticism, and hence appeals to the
demanding standards for ‘knows’ that the radical skeptic employs. If considering the problem of
radical skepticism means raising the standards for ‘knows’, then evaluating whether we know the
denials of radical skeptical hypotheses should be sufficient to take us into the more demanding
skeptical context.
Conceding that (1) is true does not create any conflict with (3) and the closure principle, represented
by (2), however, so long as we keep the relevant context fixed. In any high-standards context where
(1) is at issue, then (3) is no longer true, and so there is no tension with the closure principle. That is,
once we start thinking about radical skeptical hypotheses, as we would need to in order to evaluate
(1), then we are in a skeptical context where ‘knows’ picks out a demanding standard. But in that
context (3) is no longer true, as ‘knows’ there will also pick out a demanding standard as well. Put
another way, we don’t ‘know’ everyday claims relative to demanding skeptical standards.
And in any context where (3) comes out as true—where we do ‘know’ everyday claims, relative to the
relevant undemanding standards—then the problem of radical skepticism is by definition not under
consideration, and hence neither are radical skeptical hypotheses. Thus the question of whether (1) is
true simply doesn’t arise, and so there is no tension with the closure principle (and thus (2)). As soon
as the question of whether (1) is true does arise, however, then (1) becomes true. But then (3)
35
becomes false, as the standards for ‘knows’ have just risen to mean that we no longer ‘know’
everyday claims either.
Either way, then, there is no single context where the closure principle generates a tension between
(1) and (3). Thus although it looked like claims (1)-(3) were jointly inconsistent, according to the
contextualist this is not the case. What we have failed to do is realize that ‘knows’ is a
context-sensitive term in the way that the contextualist has set out.
One might think that contextualism offers an elegant way of dealing with the problem of radical
skepticism. But it faces some fairly serious objections. One kind of concern is that it is hard to
understand why we were ever taken in by the problem of radical skepticism if contextualism is the
right way to respond to it. No-one was ever confused about how what is meant by expressions
involving indexicals could depend on context. Similarly, no-one has been led astray by the fact that
words like ‘tall’ can be used in very different ways relative to different contexts of use. Put another
way, there is no parallel skeptical puzzle involving ‘I’ or ‘tall’. No-one gets puzzled, for example, that
I count as tall relative to the standards for tallness employed in everyday contexts, but I don’t count
as tall relative to the standards employed by basketball coaches.
But if we are not confused by our usage of other context-sensitive terms, then why is ‘knows’ so
different? That is, why didn’t we recognize immediately that there was a context-shift going on in the
radical skeptical argument, just as we do when it comes to our usage of other context-sensitive
terms? Is it really plausible that a puzzle that has perplexed philosophers for such a long term simply
turns on a simple feature of language that we would ordinarily identify in a flash? Remember too in
this regard that ‘knows’ is a pretty basic term, one that we are all familiar with—it’s not as if it’s an
arcane word, or that it has a very specialized meaning.
There is another reason for thinking that contextualism doesn’t get to the bottom of the radical
skeptical problem. For the contextualist response to radical skepticism to gain a purchase on this
difficulty, then it is crucial that we satisfy some epistemic standard for knowledge, albeit only a weak
one. After all, the contextualist line is precisely that in everyday contexts our knowledge ascriptions
are correct because the standards in play are so undemanding. This is how it can come out as true
that I ‘know’ something so mundane as that I am currently wearing a shirt.
The problem, however, is that the radical skeptical problem that we posed earlier seems to exclude
even this possibility. Remember that the radical skeptical puzzle was posed at the level of knowledge.
In particular, the claim was that even if our beliefs are true, they do not amount to knowledge
because we have no epistemic reason at all for thinking that our beliefs are true. As we noted, in
normal circumstances I might think that I have all sorts of epistemic reasons for thinking, say, that I
am wearing a shirt, but once the closure principle and radical skeptical hypotheses like the BIV
scenario are in play, then these epistemic reasons seem to be shown to be illusory. After all, I know
that if I were a BIV, then I wouldn’t be wearing a shirt, but I don’t know that I’m not a BIV. So how
can I have any good epistemic reasons for thinking that I am wearing a shirt right now? Aren’t my
epistemic reasons for believing that I’m wearing a shirt right now the very same reasons that the BIV
36
would offer for thinking that she is wearing a shirt? But if that’s right, then in what sense do I have
any good epistemic reasons at all?
The point is that insofar as radical skepticism generates this result, then contextualism is simply
irrelevant to the problem in hand. This is because it would follow that we don’t satisfy any standard
for knowledge, not even the weakest standard imaginable. Accordingly, appealing to standards for
‘knows’ gains us no purchase on the skeptical problem whatsoever, as the difficulty is still with us. In
short, we still lack knowledge, even relative to the low everyday standards described by the
contextualist. And if we lack knowledge even relative to these undemanding epistemic standards,
then we lack it simpliciter, just as the radical skeptic alleges.
Inverting Skepticism
A more radical approach to the problem of radical skepticism is sketched by Ludwig Wittgenstein
(1889-1951) in his final notebooks, published posthumously as On Certainty (1969). As we noted
above, Wittgenstein’s contemporary Moore thought that our everyday commonsense certainties had
a special role to play in our response to radical skepticism. So did Wittgenstein, but his conception of
the special role they are meant to play is radically different. Whereas Moore thought that these
commonsense certainties provided us with a sound epistemic basis to push back against radical
skeptical doubt, Wittgenstein instead claimed that they were essentially arational (i.e., neither
rationally nor irrationally held). Let’s try to unpack this thought, and why Wittgenstein thought that it
provided us with anything constructive to say about radical skepticism. After all, on the face of it, to
say that our most basic convictions are arationally held sounds an awful lot like Wittgenstein was
agreeing with the radical skeptic.
In order to understand what Wittgenstein was proposing, we first need to step back a little and make
a few observations about rational evaluations. We have previously noted that knowledge is true belief
that is appropriately grounded in epistemic reasons. When we make rational evaluations—for
example, when we try to determine whether a belief has the right kind of rational support such that it
amounts to knowledge—we normally do so against a backdrop of accepted claims that are not
themselves in question. So, for example, suppose we want to work our whether someone’s belief that
the tree that they are looking at is an oak amounts to knowledge. To do so we will consider various
epistemically relevant factors. These might include why they believe what they do (e.g., is it because
of the way the tree looks or is it based on someone else’s testimony?), how responsible they were in
forming their belief (e.g., did they inspect the tree closely, or merely take a quick glance at it?), how
likely it would be that this person could be wrong about this subject matter (e.g., are there types of
tree in the vicinity that look like oaks but which aren’t oaks?), and so on. In undertaking a rational
evaluation in this way we are evaluating whether one belief (in this case about whether this tree is an
oak) amounts to knowledge relative to a background of claims that are already accepted as
knowledge (e.g., that the tree looks like a normal oak tree, that there aren’t non-oak trees in the
vicinity that look like oaks, that one can spot an oak just by looking at it without the need for special
checks, and so on). This is the sense in which rational evaluations are normally local, in that we are
37
not rationally evaluating all of our beliefs all at once, but only a sub-set of those beliefs, relative to a
wider set of beliefs that are not in doubt (and which are thus treated as knowledge).
In contrast, the radical skeptic’s rational evaluations are not local in this way, but global. After all, the
radical skeptic is trying to determine whether our beliefs as a whole enjoy appropriate epistemic
grounds, and so amount to knowledge. Indeed, that’s just the whole point of introducing radical
skeptical scenarios, which are precisely situations in which one’s beliefs are massively in error (but
which are nonetheless indistinguishable from normal life, which is why we are unable to rule them
out).
Normally we rule-out error-possibilities by appealing to things that we know that are not themselves
called into question by that error-possibility. For example, one might rule-out the error-possibility
that the tree before one is not an oak but rather an elm by pointing out that one knows the
difference between the appearances of these two trees, and hence can tell them apart. Here we are
appealing to our knowledge (in this case of what oaks and elms look like) in order to rule-out the
error-possibility that what we think is an oak is actually an elm.
We cannot coherently take this kind of line with radical skeptical scenarios, however. This is because
they are error-possibilities that call all (or at least most anyway) of one’s beliefs into question, and
hence call into question even one’s background knowledge. That’s why we cannot appeal to our
memories or experiences (or even our beliefs about the current technological state-of-play) in order
to rule-out the BIV skeptical hypothesis, as this very scenario raises doubts about their veracity.
That radical skepticism is in the business of undertaking global rational evaluations rather than the
local ones that we are ordinarily familiar with is not in itself a count against it. As we noted when we
discussed the relevance of commonsense when it comes to the problem of radical skepticism, the
radical skeptic is plausibly offering us a ‘purified’ version of our everyday practices. Normally we
rationally evaluate our beliefs in a piecemeal fashion, and it makes sense to do so, given the practical
constraints that we ordinarily operate under. We just don’t have the time, much less the inclination,
to question all our beliefs all at once. But when we do philosophy we are not constrained by such
limitations, and hence we can ask what the epistemic standing of our beliefs are as a whole. The
radical skeptic’s contention is that once we step back to do this we discover that our beliefs have no
sound epistemic basis at all, and hence do not amount to knowledge (or, at least, that we don’t have
any plausible story available to us as to why they amount to knowledge).
Interestingly, it is not just the radical skeptic who attempts to rationally evaluate our beliefs as a
whole. For notice that this is also the goal of the traditional anti-skeptic too. While the radical skeptic
rationally evaluates our beliefs as a whole in a negative fashion, and so finds them wanting, the
traditional anti-skeptic rationally evaluates our beliefs as a whole with a view to producing a positive
verdict (i.e., as showing that they amount to knowledge, contra the radical skeptic). Recall that
Descartes was trying to do this with his foundationalism, whereby we are able to show that our
beliefs as a whole are in good epistemic order since they are supported by indubitable foundations.
Moore was trying to do something in the same broad vein by appealing to our commonsense
38
certainties. In particular, he was trying to show that such certainties provide us with a rational basis
to push back against radical skepticism and thereby assure ourselves that our beliefs, as a whole, were
perfectly in order from a rational point of view just as they are. (Or, at least, that if there is anything
wrong with our beliefs in this regard, it is not something that the radical skeptic has exposed).
This is where Wittgenstein comes in. His basic idea was that both the radical skeptic and the
traditional anti-skeptic are employing a false conception of the nature of rational evaluation. In
particular, Wittgenstein argued that it is not an incidental feature of our everyday rational evaluations
that they are local. Instead, it is rather part of the very nature of rational evaluation that it is localised
in this way. In particular, he claimed that the very idea that one could undertake a global rational
evaluation—i.e., to rationally evaluate all of one’s beliefs all at once—is simply incoherent. If that’s
right, then both the radical skeptic and the traditional anti-skeptic are making the same fundamental
error.
The reason why Wittgenstein made this claim was that he thought that all rational evaluations
necessarily take place against a backdrop of a kind of primitive, but absolute, certainty. Since this
certainty is needed for rational evaluations to occur, it is sometimes called a ‘hinge’ certainty
(following a famous metaphor that Wittgenstein employs in this regard). This certainty is in effect an
underlying tacit conviction that one is not radically in error. Wittgenstein held that this is manifested
in the kind of brute certainty that we exhibit when it comes to those everyday claims that are, for us,
optimally certain. This might sound an awful lot like Moore, and indeed it is, in that Wittgenstein is
interested in the very same everyday certainties that Moore highlighted. But there is also a crucial
difference. While Moore thought that these everyday certainties had a special rational status,
Wittgenstein instead held that they are immune to rational evaluation. Not only can we make no
sense of them being rationally grounded, but they are not thereby irrational either. They are rather
the hinges that must be in place in order for rational evaluations to occur, and hence cannot
themselves be rationally evaluated. If this is right, then it follows that there cannot be universal
rational evaluations, since there will always need to be the hinge certainties in place for any rational
evaluation to occur, and which are themselves immune to rational evaluation.
Let’s consider an example that Wittgenstein discusses in this respect. Moore famously offered his
belief that he has hands as one of the commonsense claims that he was optimally certain of. Thus, if
any philosophical argument, like the radical skeptical puzzle that we have encountered, purports to
show that he doesn’t know he has hands, then Moore contends that it is more reasonable to reject
that reasoning than to accept that he doesn’t know that he has hands. In particular, as we saw above,
it would seem that what Moore would have to claim is that we must be able to know the denials of
radical skeptical hypotheses (even though he would grant that he cannot explain how). Wittgenstein
agrees that this claim that one has hands is optimally certain for us, at least in normal conditions, but
he doesn’t think that this bestows it with any special rational status as Moore does. Wittgenstein
argues that the fact that this is so certain for us in normal conditions means that we can neither make
sense of there being any coherent grounds for doubting it or for there being any coherent grounds in
favour of believing it.
39
We might think this is puzzling, since isn’t my belief that I have hands something that is grounded in
my experiences of my hands, my seeing them and feeling them and so forth? Wittgenstein contends
that this is a mistake. As he puts it:
“If a blind man were to ask me “Have you got two hands?” I should not make sure by looking. If I were to
have any doubt of it, then I don’t know why I should trust my eyes. For why shouldn’t I test my eyes
by looking to find out whether I see my two hands? What is to be tested by what?”
The point is that such a certainty is not something that is itself rationally evaluated, but rather part of
the unquestioned backdrop against which other claims are rationally evaluated. It is thus what makes
rational evaluations possible in the first place. (Compare one’s conviction that one has hands with
one’s conviction that one’s keys are in one’s pocket. If someone asks you if you have your keys, it
makes perfect sense to tap one’s sides to see if the tell-tale jingle of the keys in your pocket is there,
or even to take them out to show the other person. But it makes no sense at all if someone asks you
whether you have hands to produce them and say, ‘oh yes, here they are!’).
Wittgenstein’s point is that these everyday certainties manifest our general conviction that we are not
fundamentally mistaken in our beliefs. But that we are not fundamentally mistaken in our beliefs is
not something that we have any reason for holding. It is rather just part of what it is to be a believer,
and for that matter a doubter, in the first place.
If we have no rational basis for these hinge certainties, then it follows that they can’t be known, at
least if we continue to maintain that knowledge requires true belief grounded in appropriate
epistemic reasons. But one might wonder why this claim doesn’t just collapse into radical skepticism.
If our rational evaluations presuppose a hinge certainty that is lacking in rational support, then
doesn’t that mean that our beliefs are ultimately completely groundless (i.e., completely lacking in
epistemic reasons)? And doesn’t that in turn entail that we don’t know anything, just as the radical
skeptic contends?
Notice, however, that while it’s true on the Wittgensteinian picture that we don’t have knowledge of
the hinge certainties, there’s also a sense in which we don’t fail to know them either. That is, this is
not something that we are ignorant of, as if it is something that we could have known but failed to.
Wittgenstein’s claim is that there is no sense to the idea of a rational evaluation that didn’t
presuppose this backdrop of hinge certainty, and hence that one cannot rationally evaluate the hinge
itself as this needs to be in place for a rational evaluation to occur. The hinge certainty is thus in a
sense neither something that we could know or fail to know (i.e., be ignorant of)—it simply isn’t the
kind of thing that is in the market for knowledge.
This marks a subtle difference between Wittgenstein’s proposal and radical skepticism. They both
agree that the hinge certainty is unknown, but the explanation is different in each case. For
Wittgenstein, it is unknown because the hinge certainty is simply not the kind of thing that could be
known, and it is a philosophical mistake to think otherwise. For the radical skeptic, in contrast, the
hinge certainty is the kind of claim that could be known, it is just that we have failed to know it (i.e.,
40
we are ignorant of it). Although subtle, this difference is important. If we are simply ignorant of the
hinge certainty, then the idea that our everyday beliefs, buttressed by local rational evaluations, lack
supporting epistemic reasons looks very plausible. But if the hinge certainty is not even in the market
for knowledge, then why should the fact that it is unknown have any negative ramifications for the
epistemic pedigree of our everyday beliefs? This is why I have described this Wittgensteinian
approach as ‘inverting’ skepticism, in that it takes a key contention that the radical skeptic makes and
attempts to turn it on its head so that it can be employed against radical skepticism.
Even if one grants this point to Wittgenstein, however, it can still look mysterious how this
anti-skeptical line is to have any application to the particular formulation of the radical skeptical
puzzle that we are dealing with. On the face of it, the idea seems to be that our basic certainty that
we are not radically in error is encapsulated in our conviction that we are not the victims of radical
skeptical hypotheses. Sure, one might go through the motions of entertaining such scenarios, of
thinking that one might be a BIV for example, but the certainty that is manifest in our day-to-day
actions reveals that there is in fact no doubt in play here at all. So, just as the radical skeptic contends,
we don’t know the denials of radical skeptical hypotheses, in line with (1), though with the proviso that
we are not ignorant of such claims either. And yet, the thinking seems to go, the epistemic support
enjoyed by our everyday beliefs is perfectly in order as it is, even though this is the product of local
rational evaluations that take place against a backdrop of primitive arational certainty. The
Wittgensteinian thus seems to endorse (3) as well.
So does that mean that the Wittgensteinian line involves denying (2), and hence rejecting the closure
principle? It can certainly seem that this must be implied by this approach, since what other option is
there? The thought appears to be that the closure principle looks innocuous because we are implicitly
taking it as given that there can be universal rational evaluations. Ordinarily, the closure principle will
only take you from one locally evaluated piece of knowledge to another locally evaluated piece of
knowledge, as when one concludes from the fact that the tree is an oak that it is not an elm (given
that one knows that no tree is both an oak and an elm). What is distinctive about instances of the
closure principle that the radical skeptic employs, however, is that they take us from everyday, locally
evaluated, knowledge claims to the denials of radical skeptical hypotheses, where the latter are the
kind of thing which, if known, would involve a global rational evaluation. The Wittgensteinian line
on closure thus seems to be the thought that instances of the closure principle of this kind are far
from innocuous, in that they presuppose a conception of the nature of rational evaluation which
Wittgenstein maintains is completely false.
If that’s right, then the Wittgensteinian response to the problem of rational skepticism is to deny the
closure principle, and thus (2). But given the evident plausibility of the closure principle, this is going
to be a difficult approach to take, as it inevitably involves denying something that we find highly
intuitive.
My goal in this module has not been to argue for any particular response to the problem of radical
skepticism, or even to offer a comprehensive overview of all the responses available. Instead, I’ve
presented a representative sample of anti-skeptical proposals so that the reader can get a sense of
how one might go about approaching this difficulty. As we have seen, all of the anti-skeptical lines
we have looked have their own difficulties, with no account offering us a ‘slam-dunk’ on this score.
Does that mean that the situation is hopeless? Not at all. In fact, that there are a range of responses
available, each with their own merits and demerits, is entirely normal for a philosophical puzzle, and
hence is nothing to be surprised about. What it tells us is that the puzzle in hand is likely to be deep
and important, such that rather than being amenable to a straightforward solution it instead prompts
us to look more closely at the subject matter in question, and in the process refine our thinking about
it. In the case of the radical skeptical paradox, this could mean, for example, trying to make sense of
how our system of rational evaluation might necessarily incorporate arational hinge commitments (as
Wittgenstein suggested), or figuring out how ‘knowledge’ could be a context-sensitive notion in
disguise (as the contextualism contends). The point of considering these anti-skeptical lines is thus
not to convince you of a particular solution, much less to convince you that there is no solution
available, but rather to demonstrate that there are philosophically interesting responses that can be
offered to the problem of radical skepticism. With this in mind, it would be just as premature to
throw in the towel to radical skepticism as it would be to opt for one of these anti-skeptical
proposals and declare victory.
To learn more about the topics covered in this module, see chapter 3 of the book that
accompanies this MOOC, A Very Short Introduction to Skepticism, by Duncan Pritchard
(Oxford: Oxford University Press, 2019).
42
See Moore (1925; 1939) for his own main responses to radical skepticism, although notice that the
skeptical problem that he is dealing with is rather different to how we are formulating it here (e.g., he
never discusses BIVs or the closure principle). For an overview of recent work on contextualism and
skepticism, see Rysiew (2010). The notebooks that make up Wittgenstein’s On Certainty are published
as Wittgenstein (1969). As with Moore, note that Wittgenstein’s characterisation of the skeptical
problem is somewhat different to how we have characterised it. The quotation is from Wittgenstein
(1969, §125). The interpretation of Wittgenstein that I offer here is my own. See Pritchard (2015, part
two) for the details. (Note that I further claim in this work that my interpretation of Wittgenstein has
the resources to preserve the closure principle while still evading the radical skeptical puzzle). For a
survey of recent work on Wittgensteinian epistemology, and his associated response to radical
skepticism, see Pritchard (2017).
Moore, G. E. (1925). ‘A Defence of Common Sense’, Contemporary British Philosophy (2nd series), J. (ed.)
H. Muirhead, London: Allen & Unwin.
Moore, G. E. (1939). ‘Proof of an External World’, Proceedings of the British Academy 25, 273-300.
Pritchard, D. H. (2017). ‘Wittgenstein on Hinge Commitments and Radical Scepticism in On
Certainty’, Blackwell Companion to Wittgenstein, (eds.) H.-J. Glock & J. Hyman, 563-75, Oxford:
Blackwell.
Pritchard, D. H. (2015). Epistemic Angst: Radical Skepticism and the Groundlessness of Our Believing,
Princeton, NJ: Princeton University Press.
Rysiew, P. (2010). ‘Contextualism’, Routledge Companion to Epistemology, (eds.) S. Bernecker & D. H.
Pritchard, 523-35, London: Routledge.
Wittgenstein, L. (1969). On Certainty, (eds.) G. E. M. Anscombe & G. H. von Wright, (tr.) D. Paul &
G. E. M. Anscombe, Oxford: Blackwell.
43
Module 4
Skepticism as a Way of Life
We began this MOOC by noting a distinction between a healthy, moderate skepticism and a radical
skepticism that had all kinds of challenging implications. The former is the antidote to gullibility, in
that it prompts us to question what we are told and not merely to take it at face value. Such moderate
skepticism is thus to be encouraged. Where skepticism becomes problematic is when it slips into the
more radical variety. Whereas a moderate skepticism scrutinizes particular claims, a radical skepticism
is skeptical about the truth of our beliefs en masse. This has troubling ramifications, as we saw in
module 1. If all our beliefs are open to question, then why be committed to the truth of anything?
Moreover, once we determine that we have no grip on the truth, then why should it matter any more
what is true? Radical skepticism thus invites relativism about truth, such that the truth is no longer
something objective, but just whatever someone claims the truth to be.
But what are the motivations for radical skepticism? Those who express a radical skepticism⎯or its
close cousin, relativism⎯in public life rarely offer any theoretical basis for it. Think, for example, of
those who question the authority of science, who offer conspiracy theories, who happily assert that
contemporary politics is in a ‘post-truth’ phase, and so on. These claims, which presuppose radical
skepticism/relativism, do not bring with them any justification for this presupposition.
We saw in module 1 that some of the rationales that might be offered for radical skepticism are in
fact highly dubious. For example, the undeniable fact that we are fallible creatures who are
sometimes in error does not in itself offer a justification for radical skepticism. That we are
sometimes mistaken does not mean that we are always mistaken. Similarly, that we hardly ever, if
ever, have reasons to be completely certain in what we believe is not a rationale for radical doubt
either. The point is that human knowledge doesn’t demand infallibility or complete certainty, and
hence that our beliefs lack these features doesn’t show that they fail to amount to knowledge.
That knowledge doesn’t entail infallibility or complete certainty was also why radical skepticism
doesn’t follow directly from the fact that we cannot rule-out radical skeptical hypotheses. Recall that
these are scenarios, like the BIV scenario, where one is radically deceived about the world around
one, but undetectably so. It seems that by definition we cannot come to know that we are not in such
a scenario. But this by itself doesn’t have any radical skeptical implications. It just reminds us that
there is always the possibility of error, and that’s just to reiterate the point that we are fallible⎯and
hence to some degree quite rightly uncertain⎯creatures.
44
Interestingly, however, although appeals to infallibility and complete certainty are not motivations for
radical skepticism, they do offer rationales for moderate skepticism. If we are sometimes mistaken,
and are often not in a position to be completely confident of what we believe, then of course we
should be careful about what beliefs we endorse.
Indeed, this is precisely why we find a moderate skepticism at work in the scientific method, because
scientists recognize that our means of finding the truth is imperfect. This is why even well-founded
scientific claims are nonetheless treated as provisional, in the sense of being open to being
reevaluated if new evidence comes to light. Far from science representing a dogmatic stance about
the world around us, the scientific method in fact incorporates such a moderate skepticism.
Indeed, the emergence of what we now think of as the scientific method during the period of history
known as the scientific revolution (roughly from the middle of the 16th Century to the 18th Century)
is widely regarded as being in large part due to the rediscovery of ancient skeptical texts during the
Renaissance. By questioning the received wisdom of the past, and in particular the authority of the
Church which was often appealed to as underpinning this received wisdom, scientists were able to
discover new and important truths, such as that the earth sun did not orbit the earth as previously
supposed, and much else besides. Interestingly, however, while radical forms of skepticism did
emerge in this period, the ancient forms of skepticism that our new scientists were responding to at
this time are more concerned with localized forms of doubt. In examining this period of history we
are thus witnessing how a moderate skepticism can be harnessed to enhance our grip on the truth, and
in the process spur the kind of scientific progress that we now take for granted.
That said, we also noted in module 2 that there is a powerful argument that can be offered in support
of radical skepticism. This employs radial skeptical hypotheses, but doesn’t rest only on appealing to
these scenarios (since it also makes use of the closure principle). We also saw in module 3 that while
there are philosophical responses to this argument, they all face problems of their own.
Significantly, however, this radical skeptical argument is never presented by those in public life who
exhibit such radical skeptical tendencies, and this is no accident. Part of the reason for this, of course,
is that those who espouse radical skepticism in public life are probably unfamiliar with the
philosophical ideas behind it. But even if they were aware of the radical skeptical paradox, they would
still be unwise to try to use it to motivate their particular skeptical stance. This is because far from
supporting their skepticism, it would instead serve to undermine it by exposing just how radical the
scope of their doubt is.
Consider someone who advances a general skepticism about science, who thinks that man-made
climate change is a hoax, that vaccines are part of a global conspiracy to harm our children, and so
on. As we’ve previously noted, once one’s skepticism becomes this broad, then it is hard to see how
it could be contained—if one is skeptical about science, then why not be skeptical about everything?
Accordingly, one might think that an argument that purports to show that knowledge is impossible
would be a boon to such a person. The problem, however, is that it is in fact important to our
45
science skeptic that she doesn’t generalize her doubt in this way. After all, she wants to maintain that
she is in a privileged epistemic position relative to everyone else; that she can see through the
conspiracy that everyone else is falling for. But that means that she needs to credit herself with
knowledge that others are lacking, and hence a truly radical skepticism is not her friend. Our science
skeptic is thus trying to pull-off a kind of intellectual high-wire act, whereby she advances a dramatic
form of skepticism which is nonetheless not quite so dramatic that it thereby also undermines the
knowledge that she needs to advance her case in the first place.
Such an intellectual high-wire act is not sustainable, as we’ve previously noted. Once one’s doubt
becomes extensive enough, then it also undermines one’s own grounds for doubt and hence
becomes self-destructive. This doesn’t mean that there is no problem of radical skepticism, only that
this problem is not best understood as a position—i.e., as a view that someone coherently
advances—as opposed to being a puzzle or paradox. In any case, this is why there would be nothing
to be gained by our science skeptic (or anyone else in public life who advances a sufficiently general
form of skepticism) to appeal to the radical skeptical paradox in support of their stance. To do so
would expose just how radical the skepticism is play is, and thereby undermine it.
This point highlights how radical skepticism when advanced in public life (as opposed to in a
philosophical discussion) is in a certain sense fake. Our science skeptic is happy to advance doubts
that call most of what we believe into question, while nonetheless trying to maintain that her grounds
for such doubts are entirely secure and beyond doubt. But that’s simply not credible. Moreover, this
kind of intellectual ‘double-dealing’ is also manifest in how our science skeptic lives her life. Such a
person is usually quite content to receive first-world healthcare, to travel on planes with the
confidence that they won’t fall out of the sky, and so on. But if she is really a skeptic about science,
then how can that be the case?
The crucial point is that once we make explicit what radical skepticism involves, then it is hard, if not
impossible, to even make sense of someone adopting this as a stance. How should we make sense of
someone who claims not to believe anything? How do we make sense of what they do, what they
care about⎯including the radically skeptical claims that they make⎯and so forth if we take them
seriously as being genuinely skeptical about everything? This is why those who espouse radical
skeptical ideas in public life do not cloak them in these terms (even though they implicitly trade on
skeptical themes), since to do so would be contrary to their purposes in advancing their skepticism in
the first place. They want us to believe them, after all, even if they don’t want us to believe much else.
Interestingly, similar points apply when it comes to those who put forward relativistic ideas about
truth in public life, as this also involves a healthy dose of fakery. Those who try to convince us that
objective truth doesn’t matter in political life, for example, because there is no such thing as objective
truth, are not in fact giving up on an objective conception of truth at all. Do you think that their lack
of concern for the objective truth extends to matters of practical concern to them? For example, do
you think they would be so sanguine about what is objectively true if it were the tax authorities
declaring that they were not due the tax refund to which they are entitled? Or if they were falsely
accused of a serious crime? This is the sense in which their commitment to relativism is fake, since if
46
they really were relativists about truth then they wouldn’t regard anything as objectively true. Like
those who advance radical skeptical ideas in public life, these relativists are trying to pull off an
intellectual high-wire act, whereby they are relativists about objective truth when it suits them but not
when it doesn’t. Such a position is no more sustainable in this case as it was with our science skeptic.
These points about the incoherence of radical skeptical and relativistic ideas in public life highlights
something important. For often these ideas are presented as being challenges to authority, as thus as
liberating narratives that one can employ to undermine existing power-structures. In fact, the reality
is almost always the opposite, in that such ideas in face serve to support existing power structures
and keep them in place—authority has nothing to fear from them. For example, if science is totally
debunked, then why should we listen to scientists who demand action on man-made climate change?
Or if there is no such thing as objective truth, then it also cannot be an objective matter that there is
injustice in society. How then could one motivate the political basis for confronting such injustice?
In any case, the argument for radical skepticism that we looked at in module 2 was powerful precisely
because it was not essentially tied to a radical skeptical position. Rather, it was primarily put forward as
a paradox⎯i.e., as exposing deep tensions within our own concept of knowledge. As we saw,
paradoxes can be coherently posed without this requiring anyone to advocate for a particular way of
resolving that paradox. Radical skepticism as a paradox is thus distinct from radical skepticism as a
position—i.e., as a lived stance that one takes. That the latter often leads to incoherence thus doesn’t
entail that there is anything essentially amiss with the former.
Those who espouse radical skeptical and relativistic ideas in public life are therefore unable to
legitimately appeal to the radical skeptical paradox in support of their claims. Moreover, as we also
saw in module 3, just as a case can be made in support of this paradox, there are also philosophical
responses that can be offered too. It would therefore be premature to conclude that there is an
insolvable skeptical problem afflicting our knowledge.
In what follows we are going to be putting radical skepticism to one side and considering instead
how one might make a positive case for a moderate skepticism. To this end we will be looking at an
idea that we get from the ancients, and in particular from the work of possibly the greatest
philosopher who ever lived, Aristotle (384-322 BC). This idea concerns the role of the virtues, and the
intellectual virtues in particular, in the ‘good life’ of human flourishing; what the ancient Greeks called
eudaimonia. As we will see, understanding the role that the intellectual virtues play in the good life will
enable us to see how embracing a moderate skepticism could be necessary for living such a life (in
contrast to embracing radical skepticism, which would be inimical to it). Relatedly, it will also help us
to resolve a possible tension between adopting a healthy moderately skeptical attitude while at the
same time living a life of genuine conviction.
Philosophy is often accused of being focussed on entirely abstract matters, unconnected from the
pressing issues that confront us in everyday life. This certainly wasn’t true of Aristotle, however, who
was concerned to offer us practical advice about how to lead good lives. When we think of ethics in
modern life we tend to equate it with morality—i.e., with what is specifically morally good (and bad).
But for Aristotle, as for many ancient philosophers, ethics had a much broader meaning as being
concerned with the more general question of how one ought to live; what constitutes a good life.
Morality may be part of such a life, but it is but one part, as there are other important elements to a
good life. A life that is lacking in achievements, in significant personal relationships, or in aesthetic
experience, would be impoverished as a result, but none of these things are essentially concerned
with morality. There thus might be more to living a good life than simply living a moral life. It is the
broader ethical question of how to live a good life that Aristotle attempted to answer.
He approached this question by appealing to the role of the virtues in the good life. These are
distinctive kinds of character traits—i.e., dispositions that one has to behave in certain ways—that
are particularly admirable. It includes such traits as courage, generosity, and kindness. The virtues
have a number of interesting properties, at least according to Aristotle. For example, virtues lie in
opposition to vices, which are character traits that are not admirable at all (indeed, they are often
contemptible). Indeed, a virtue lies between two vices, a vice of excess and a vice of deficiency. Take
courage, for instance. The lack of courage—cowardice—is a vice, but one can also take excessive
risks and thereby exhibit the very different vice of being rash or foolhardy. Being courageous is thus
having the good judgement to be disposed to act between these two extremes, what Aristotle called
the golden mean.
We are not born with virtues, but must acquire them through practice, especially by emulating those
around us who have these admirable traits. One has to learn to be courageous, as opposed to being
cowardly or rash. And one needs to cultivate this virtue once acquired, or it is apt to be lost. It is
properties like this that set virtues apart from mere skills. We can be born with certain skills (e.g., our
hardwired skills to perceive our environment via our senses), and there can be skills that are learnt in
such a way that they are unlikely to be lost regardless of whether one cultivates them thereafter (e.g.,
riding a bicycle). Another important difference between skills and virtues is that the latter bring with
them distinctive kinds of motivations. Someone who has the virtue of generosity, for example, will
be genuinely motivated to help others. Merely acting as if one is generous—in order to impress one’s
peers, for example—is not the same as being virtuous. Skills, in contrast, often don’t demand
motivations in this way. Indeed, being able to act in ways that give others the (false) impression that
one is generous might well be a very skilful thing to do.
But the big difference between mere skills and virtues is that the latter has a special kind of value, and
this is because of the vital role that they play in a good life. In essence, Aristotle held that the good
life is the life of virtue. Notice that he isn’t suggesting that such a life will be good in the sense that it
will always involve pleasure rather than pain, or that it will be a life without conflict and suffering. In
fact, Aristotle took it for granted that all lives were likely to involve pain, conflict and suffering. The
point is rather that the way to confront such ills is by being armed with the virtues. If one is kind,
generous, courageous and so on, then one can flourish as a human being even despite facing such
48
difficulties. Moreover, facing these difficulties in such a virtuous manner is in any case far better than
being insulated from them in an empty life of meaningless pleasure. Pleasure without virtue never
turns out well, as I think we all know.
I noted that ethics for Aristotle is not just about morality, and this is reflected in the virtues
themselves. Some of them, like kindness and generosity, do seem to be particularly moral. But others
are less so, such as courage. In fact, Aristotle held that being virtuous didn’t simply involve knowing
what the right thing to do is from a moral point of view, but also consists in having the specifically
intellectual virtues too. After all, how can one know the right thing to do if one doesn’t have knowledge
in the first place? And yet the intellectual virtues are vital to ensuring that one is a knower.
The intellectual virtues include such traits as conscientiousness and open-mindedness. It also includes
intellectual variants of more general kinds of virtue, as when someone is specifically intellectually
courageous (think, for instance, of a scientist who develops a radical new idea) or when someone
exhibits intellectual humility (as opposed to intellectual arrogance). So, for example, to be intellectually
conscientious is to form one’s beliefs by attending to all the relevant evidence and being able to
appropriately weigh up this evidence in reaching a decision. It is just such a trait that we would want
in a judge who was hearing our trial, for example, or in a doctor who was making a decision about
whether we needed an operation. This intellectual virtue lies between two corresponding intellectual
vices. On the one hand, there is the vice of deficiency of not attending to the evidence at all in
reaching one’s decision, but rather, for example, opting for whatever outcome suits your interests.
On the other hand, there is also the vice of excess of being so attentive to all the evidence, regardless
of its relevance to the issue in hand, that one is simply swamped and so unable to reach a decision.
The golden mean involves having the wisdom to navigate between these two extremes.
Having this picture in play of how there is an intellectual element to the good life helps us to clarify
how a radical skepticism would be inimical to such a life. It also helps us to see just what might be
admirable about a localised skepticism. Let’s take these points in turn. We have noted previously that
there is a deep incoherency about trying to actively embrace radical skepticism. How can we even
make sense of someone who claims not to know anything and to doubt everything? We have also
noted that radical skepticism poses an existential challenge because our inability to know anything
would make our lives absurd. The Aristotelian picture gives us the means to develop this last point.
The idea that without knowledge our lives would be absurd is a purely negative claim, in that it
merely tells us that some essential component to a meaningful life would be lacking. In particular, it
doesn’t tell us how having knowledge might make our lives meaningful (perhaps they would still be
absurd, but for different reasons). This is where the Aristotelian account of the virtues is helpful, in
that it offers a positive story about what is involved in a meaningful life, one that makes essential use
of the intellectual virtues, and thus the knowledge that these virtues generate. In particular, given that
manifesting the virtues requires knowledge, how could one lead a virtuous life if one doubts
everything? For example, how could someone have the good judgement to be courageous and
generous, and thereby know how best to act in this regard, while living such a life of radical doubt?
The Aristotelian picture thus tells us why, if radical skepticism were true, then one’s life would be
49
lacking in a fundamental respect, in that the life of flourishing, eudaimonia, would simply be
unavailable.
Interestingly, however, while a life of radical skepticism is in conflict with the virtues, an attitude of
moderate skepticism seems to be licensed by the virtues, particularly the intellectual virtues. Indeed,
the intellectual virtues tend to characteristically involve a moderate skepticism. Isn’t that part of what
is involved in being conscientious when one weighs up the evidence, rather than leaping to conclusions?
Or in being open-minded in one’s opinions, as opposed to being dogmatic? Moderate skepticism is thus
often what is dictated by the golden mean, where radical skepticism represents a vice of excess (i.e.,
of excessive doubt), with dogmatism representing the opposing vice of deficiency (i.e., of insufficient
doubt). Accordingly, on an Aristotelian conception of the virtues, and of the intellectual virtues in
particular, we can account for the importance of moderate skepticism by showing how it is
conducive to a life of flourishing. Relatedly, we can also explain just what is problematic about
attempting to live a life of radical skepticism, as this would be inimical to living such a flourishing
life. (To reiterate our point from earlier, remember that in saying this we are not thereby claiming
that the problem of radical skepticism qua paradox is thereby resolved. Radical skepticism as a
position and radical skepticism as a paradox are very different beasts).
Earlier I described how the intellectual virtues could be an essential part of a life of flourishing, a
good life. I’ve also described how a moderate skepticism could be viewed as a manifestation of the
intellectual virtues, unlike a radical skepticism. There is still an important puzzle remaining about
how any kind of skepticism could be compatible with a life of flourishing. The crux of the matter is
that we also treat conviction as being essential to a good life too. We want people to have, as we might
put it, the strength of their convictions, to stick by their principles and not simply abandon their
most cherished beliefs at the first sign of dissent. Arguably at least, someone who lacked conviction
would be lacking something crucial to the good life of human flourishing. Indeed, isn’t such
conviction part of what it is to manifest the virtue of being intellectually courageous? But how can
we square the importance of conviction to the good life with the apparently opposing role that we
have described for moderate skepticism?
In order to see why this conflict is in fact illusory, we first need to unpack what is involved in
conviction in this regard, at least insofar as it is a plausible part of a life of flourishing. To begin with,
notice that there is nothing particularly laudable about people who stick to their convictions
regardless of the rational support they have for those beliefs, or whether they even have any rational
support for them. Merely holding onto one’s beliefs no matter what looks like mere dogmatism, and
it hard to see why that would be an important part of a life of flourishing (much less that it has
anything to do with the manifestation of intellectual virtue).
Relatedly, in saying that we want people to be willing to stick to their convictions in the face of
dissent, I take it that what we mean is that we want people to not simply give up their convictions
50
because of pressure from other people. For example, if someone who holds anti-racist views finds
themselves surrounded by people expressing racist opinions, we would think it disappointing if they
‘went with the flow’ and duly parroted the racist views of those around them. The reason why this
would be disappointing is that it would display a willingness to change one’s mind even though one has
been given no reason for doing so, but merely because it is convenient (in this case socially convenient). In
contrast, sometimes one is given excellent reasons to change one’s mind, and when this happens one
ought to do so. Being willing to change one’s mind when one has been given good reasons for doing
so is not a sign that one lacks conviction, but rather demonstrates that one is rational. This is
certainly something that the intellectually virtuous person would do.
Just as we would we would find it disappointing if someone abandoned their beliefs merely due to
social pressure, so we would find it disheartening if someone were to change their mind when
offered only very weak reasons for doing so. If one has formed one’s beliefs appropriately, and hence
has good reasons in support of them, then one ought not to change one’s mind at the first sign of
any argument against one’s views, no matter how flimsy. Indeed, if one has really thought through
why one is so convinced about this subject matter, then one should be able to rebut weak
counterevidence were it to be offered—for example, if someone offers a manifestly implausible
justification for their opposing views. It follows that there is no inherent reason why someone who
has conviction in her beliefs should be unwilling to listen to counterevidence to her views.
So insofar as conviction is laudable, and thus the kind of thing that might be part of the good life of
human flourishing, then it ought to be grounded in reasons and it should be open to
counterevidence. Call this reasonable conviction. We can thus set aside types of unreasonable conviction
that are merely dogmatic, where this means either not grounded in reasons or closed-off from
counterevidence (or both). Reasonable conviction seems part of an intellectually virtuous life. Indeed,
it seems to be something that is entirely compatible with one embracing the kind of moderate
skepticism that we noted above. That requires that one is suitably sensitive to reasons, but as we just
described it that’s precisely what a reasonable conviction involves.
Even with the scope of conviction so restricted, there is still a puzzle here. If one has really thought
through the issues and arrived at a certain conclusion, then doesn’t that mean that the matter is
settled? If so, then why should one consider counterevidence? (And if it isn’t settled, then doesn’t
that mean that conviction isn’t warranted?) Indeed, if one takes oneself to have impeccable rational
support for a particular claim, then isn’t that in itself a good reason to regard any evidence against
that claim as being misleading? Conversely, if one does take such counterevidence seriously, then
doesn’t that mean downgrading one’s own conviction accordingly? After all, if responding to such
counterevidence seriously doesn’t entail downgrading one’s conviction, then doesn’t that mean that
one simply isn’t taking it seriously, but is rather (perhaps secretly) adopting a dogmatic stance?
We can bring this point out by considering cases of disagreement. Imagine that you have a certain
political view, a position which you have thought through rationally over a long time. Suppose you
are now confronted by a group of people who, in turns out, have also thought through these matters
but have come to the exact opposite conclusion. Since they have also thoroughly contemplated these
51
issues, they are able to marshal counterarguments of their own, counterarguments that aren’t
obviously flawed. How should you respond? In particular, what does having reasonable conviction in
your views demand of you?
One might think that having reasonable conviction here means being unwilling to even consider
these counterarguments. After all, you have thought through these issues and come to a conclusion.
The matter is thus settled for you. But as we noted above, that sort of response does seem dogmatic.
After all, can’t reasonable people come to very different opinions, especially regarding such a
contentious subject matter as politics? If that’s so, isn’t that a reason to show a little humility and not
be so sure of one’s own opinions? But that seems to suggest that the alternative to sticking to one’s
guns in the face of disagreement is to downgrade one’s confidence in one’s beliefs. But how does
that allow any intellectual space for reasonable conviction?
We can bring this issue into sharper relief by characterising it as an apparent conflict between
intellectual virtue and conviction, both of which, as we have noted, seem to be important to the good
life of human flourishing. This is because intellectual humility itself seems to be an important
intellectual virtue, one that is closely related to the moderate skepticism that we saw above was
plausibly part of the virtuous life. Isn’t the lack of intellectual humility intellectual arrogance, or
dogmatism? And wouldn’t someone who embraced the moderate skepticism that we set out be
inclined to be intellectually humble in their opinions rather than intellectually arrogant? In particular,
when faced with disagreements with apparently reasonable people, wouldn’t they be willing not just
to rationally engage with these people, but also to downgrade their confidence in their own opinions,
at least temporarily? Indeed, one might think that intellectual humility just is a matter of having a
downgraded assessment of one’s intellectual abilities and achievements, such that rather than
regarding oneself as knowing it all, one instead treats oneself as being a highly fallible agent with an
imperfect grasp of the truth, and hence willing to learn from others around one. But on this
conception of intellectual humility, what room could there possibly be for conviction in a virtuous
life?
This apparent tension between intellectual humility as an intellectual virtue and conviction can be
resisted, but to do so we need to understand exactly what intellectual humility as an intellectual virtue
looks like. The mistake made in the foregoing is to treat such a virtue as being entirely inwards
focussed. What I mean by this is that the account of intellectual humility in play is all about how one
assesses oneself, from an intellectual point of view.
The way we just expressed this point was that to be intellectually humble involves one effectively
having an inaccurate (at least potentially anyway) conception of one’s intellectual abilities and
achievements, such that one regards them as less impressive than they might in fact be. That might
seem initially appealing—don’t humble people in general have such a downgraded conception of
their abilities and achievements? On reflection, however, it ought to be clear that this can’t be the
right way to think about humility, intellectual or otherwise. If it were, then it would mean having an
accurate conception of one’s abilities and achievements would be a vice. Applied to humility in
general that’s puzzling, but applied to intellectual humility in particular it’s straightforwardly bizarre.
52
Remember that the intellectual virtues are devoted to excellence in intellectual matters. If so, how
could they demand inaccuracy in one’s beliefs?
Taking this point on board, one might be tempted to argue instead that the point is not that one
should have a downgraded assessment of one’s intellectual abilities and achievements, but that one
should instead embrace one’s all too human intellectual failings, one’s fallibility and so forth. This is
sometimes called ‘owning one’s intellectual limitations’. This avoids the problem of an intellectual
virtue requiring inaccuracy, but it faces problems of its own, again resulting from the
inwards-directed nature of the proposal. For imagine someone who is clearly intellectually superior to
those around them—much smarter, much less prone to mistakes, and so on. Moreover, she isn’t just
intellectually superior, but she also knows full well that she is. If intellectual humility is just a matter
of owning one’s intellectual limitations, and one’s intellectual limitations are so much less than
everyone else’s, then what would be wrong about such a person acting in intellectually superior ways
to those around them? For example, what would be wrong with such a person discounting other
peoples’ viewpoints and the reasons they can offer in support of those viewpoints? It seems that on
this conception of intellectual humility this kind of behaviour would be entirely compatible with one
being intellectually humble.
The crux of the matter is that we need to think of intellectual humility in outwards- facing (or
other-directed) terms rather than in an inwards-facing (or self-directed) way. What I mean by this is
that rather than such a virtue being focussed on how one regards one’s own intellectual abilities and
achievements it should instead be geared towards how one treats other people. In particular, what
makes someone intellectually humble is the respectful way in which they intellectually engage with
other people: whether they are willing to listen to alternative viewpoints, to explain the reasons they
have for their own opinions, to rationally debate subject matters, and so on. The point is that while
having an accurate conception of one’s own intellectual abilities and achievements, and thus one’s
intellectual limitations, might well be something that an intellectually virtuous person would have, it
is not this that grounds one’s intellectual humility but rather whether one displays such intellectual
respect for others.
Once we understand that this is the right way to think about intellectual humility, then the apparent
conflict between acting as this intellectual virtue demands and reasonable conviction disappears.
Suppose one has thought through the issues behind one’s views in great detail, and is confronted
with someone who has opposing views, but who has clearly not put as much thought into their own
position. Being intellectually humble doesn’t mean that one cannot retain one’s confidence in one’s
opinions even when faced with this disagreement. What it does mean, however, is that one should be
intellectually respectful of the other person and their opinions. One should be willing to discuss their
position, explain one’s own stance, and so on.
One might be suspicious of such a proposal for the following reason. If one isn’t really persuaded by
this opposing view, then isn’t one’s engagement with their position a kind of play-acting, as if one is
pretending to take their position seriously when in fact one isn’t? But notice that if one were
play-acting in this way, then this precisely wouldn’t be a manifestation of one’s intellectual humility.
53
Remember that intellectual virtues, like virtues more generally, are grounded in virtuous motivations.
Recall that we noted above that merely acting as if one is generous—because one enjoys receiving
the admiration of others, say—is not enough to manifest the virtue of generosity. It is in addition
important that one’s acting in a generous way is rooted in the right kinds of motivations, such as a
concern for others. It follows that it is not enough to manifest the virtue of intellectual humility to
act as if one is intellectually humble. Rather, to be intellectually humble involves behaving in ways
that genuinely reflect one’s intellectual concern for others, and there can be no play-acting involved
in that.
That intellectual humility is compatible with reasonable conviction is crucial to the modern world we
live in. We need the intellectual virtues if we are to be flourishing human beings, and as we saw
embracing a moderate skepticism is part of the intellectually virtuous life. But we also need
reasonable conviction. Not all points of view are rationally on a par, and that means that the
intellectually virtuous should be willing to defend their opinions when necessary, and not simply
capitulate in the face of such disagreement. Especially when it comes to our public life⎯to the
defence of our political institutions, of the authority of science, and so forth⎯we need the virtue of
intellectual courage. What is also vital, however, is that we manifest that intellectual courage in ways
that are respectful of others, and that requires the virtue of intellectual humility.
If it were to turn out that intellectual virtue demanded such capitulations, then that would be a
problem. Indeed, once we start capitulating in this way, then one can easily see how a moderate
skepticism might collapse into the kind of radical skepticism that we saw above had such pernicious
social consequences. If we are willing to downgrade our confidence in our views in this way, even
though the evidence doesn’t warrant this, then why should we be confident of anything we believe?
The path to a radical skepticism, one that invites relativism about truth—such that we no longer care
what is really true—is thus opened up.
But the choice between intellectual conviction and intellectual virtue, and the virtues of a moderate
skepticism in particular, is a false one. We can have both. And that means that we can incorporate
such moderate skepticism into our lives in an intellectually virtuous way without this meaning that we
have to lose the strength of our convictions in the process.
To learn more about the topics covered in this module, see chapter 4 of the book that
accompanies this MOOC, A Very Short Introduction to Skepticism, by Duncan Pritchard
(Oxford: Oxford University Press, 2019).
54
55
The main work where Aristotle develops his ethical views, and thereby discusses the virtues
(including the intellectual virtues), is the Nichomacean Ethics. For an excellent contemporary translation
of this work, see Aristotle (1999). For a key contemporary defence of an Aristotelian account of the
intellectual virtues, see Zagzebski (1995). For a recent overview of work on the epistemology of
disagreement, see Frances (2018). For a helpful recent overview of the literature on intellectual
humility, see Snow (2018). For a defence of a version of the kind of ‘outward-facing’ account of
intellectual humility offered here, see Roberts & Woods (2007). For further discussion of how this
account of intellectual humility can be compatible with reasonable conviction, such that we don’t
have to automatically lower our confidence in our beliefs in the face of a disagreement with an
epistemic peer, see Pritchard (2018).
Aristotle (1999). Nicomachean Ethics, (2nd Edn.), (ed.) T. Irwin, Indianapolis, IN: Hackett.
Frances, B. (2018). ‘Disagreement’, Stanford Encyclopedia of Philosophy, (ed.) E. Zalta,
https://plato.stanford.edu/entries/disagreement/.
Pritchard, D. H. (2018). ‘Intellectual Humility and the Epistemology of Disagreement’, Synthese
[Online First, DOI: https://doi.org/10.1007/s11229-018-02024-5].
Roberts, R. C., & Wood, W. J. (2007). Intellectual Virtues: An Essay in Regulative Epistemology, Oxford:
Oxford University Press.
Snow, N. (2018). ‘Intellectual Humility’, Routledge Handbook of Virtue Epistemology, (ed.) H. Battaly, ch.
15, London: Routledge.
Zagzebski, L. (1996). Virtues of the Mind: An Inquiry into the Nature of Virtue and the Ethical Foundations of
Knowledge, Cambridge: Cambridge University Press.