on this page

Or send us an email

Application form

Pathways programs

Letters to my students

How-to-do-it guide

Essay archive

Ask a philosopher

Pathways e-journal

Features page

Downloads page

Pathways portal

Pathways to Philosophy

Geoffrey Klempner CV
G Klempner

International Society for Philosophers
ISFP site

PHILOSOPHY PATHWAYS electronic journal


P H I L O S O P H Y   P A T H W A Y S                   ISSN 2043-0728

Issue No. 189
21st November 2014


Edited by Eric DeJardin

I. 'A Quasi-Realist Approach to Morality' by Kamala Vadlamani

II. 'What the Ability Hypothesis Teaches -- The Conceptual basis of
Moralizing' by Matthew Sims

III. 'Doubting Descartes: Skeptical Scope and the Dream and Deceiver
Arguments in the First Meditation' by Eric DeJardin



This edition of 'Philosophy Pathways' is dedicated to the work of
philosophy students who are pursuing a BA with the University of
London's International Programme (UoLIP)
ba-diphe-certhe-philosophy. For those of you who are unaware of
UoLIP, it's a distance learning program that was founded by the
University of London in 1858. It provides students all over the world
with the opportunity to pursue University of London degrees without
having to leave their homes, their families or their jobs. In
addition, the cost of the program is extremely low -- the total cost
of the BA is actually less than what the average full-time student in
the U.S. pays for a semester's worth of coursework. However, the
flexibility and cost effectiveness of the UoLIP philosophy BA is
achieved in a unique way: registered UoLIP philosophy students only
receive (1) syllabi for the courses they're studying, which include
reading lists and a very brief overview of course topics, (2) exam
questions and examiners reports from the previous seven years, (3)
access to an online library, (4) the opportunity to be tested, once a
year, at approved examination centers around the world, and (5) the
opportunity to have their yearly exams graded in London by the same
examiners who grade the work of the brick and mortar students at the
University of London's eighteen constituent colleges. Note, UoLIP
students do not sit for lectures or receive any feedback whatsoever
from University of London faculty throughout the year while they work
through their course materials. Indeed, the only feedback students
receive from UoLIP is in the form of one exam grade per course at the
end of each academic year!

Although the UoLIP courses are designed to be studied entirely on
one's own, many UoLIP students receive tutorial assistance from
outside the University of London. Indeed, the three students whose
work is featured in this issue are all former students of Dr.
Geoffrey Klempner, the founder of the 'Philosophy Pathways' journal.
And at the beginning of the 2014-2015 academic year, the graduate
students at Birkbeck College, which is the lead college of the UoLIP
philosophy BA program, set up an arrangement with UoLIP to provide
tutorial support to interested UoLIP students at a very reasonable
cost. The availability of this sort of tutorial support from
experienced philosophers, along with the presence of various
community promoting websites (such as the University of London
International Programme Philosophy BA Facebook site:
http://www.facebook.com/groups/393151104153897/), help fill in the
gaps in the educational experience that are a consequence of UoLIP's
commitment to keeping costs low and their programs of study highly

As you might imagine, studying with UoLIP is not for everyone. There
is no hand-holding or spoon feeding whatsoever. However, I believe
that a self-motivated, hard-working student can achieve amazing
results, especially with the aid of experienced tutors. And this
conclusion seems to be borne out by the data: In a comment on a post
on distance learning on Brian Leiter's 'Leiter Reports: A Philosophy
Blog', Dr. Samuel Guttenplan, the former Director of the University
of London International Programme in Philosophy, wrote that 'Quite a
few students completing the [UoLIP philosophy] BA got First Class
Honours and went on to postgraduate work at e.g. Cornell, Oxford,
London and several other universities'
admission-in-north-america.html). Now that's very encouraging!

With that brief introduction to UoLIP out of the way, I shall now
introduce the essays that are included in this UoLIP dedicated issue.

The first essay, 'A Quasi-Realist Approach to Morality', is by Kamala
Vadlamani. In it, Kamala explicates Blackburn's moral quasi-realism,
which is an attempt to retain the moral language of the moral
realist, with its concomitant logic of moral discourse, while
remaining uncommitted to the sort of metaphysical suppositions that
moral realists often find themselves encumbered with. Kamala begins
by clarifying the quasi-realist position, and concludes 'that
quasi-realism falls into the trap of subjectivism -- the
mind-dependent account of moral truths -- and thus doesn't entirely
escape the charge of relativism'. She concludes the essay by
considering 'the kind of normative ethics... compatible with quasi

The Second essay, 'What the Ability Hypothesis Teaches: The
Conceptual Basis of Moralizing', is by Matthew Sims. In it, Matthew
posits and defends an application of the Ability Hypothesis response
to Jackson's Knowledge Argument to moral reasoning. Matthew argues
that what he calls the '3Rs' of the Ability Hypothesis --
'remembering, recognizing and representing' -- presuppose the
possession of concepts that are necessary for empathizing. And, since
empathizing is necessary for moral reasoning, he argues, the 3Rs too
are thus necessary for moral reasoning. Hence, Matthew defends the
conclusion that moral reasoning necessarily involves a facility with
specific conceptual resources.

The Third essay, 'Doubting Descartes: Skeptical Scope and the Dream
and Deceiver Arguments in the 'First Meditation'' is mine. Since I'm
a first year UoLIP student, I decided to write an essay on that
quintessential first year philosophy student topic, Cartesian Doubt.
I argue that the famous Dream Argument, which Descartes advances yet
ultimately rejects in his attempt to establish a highly specific
variant of universal doubt, is much stronger than Descartes would
lead us to believe, at least in his 'First Meditation.' More
specifically, I argue that his Dream Argument has the conceptual
resources to refute each of the three objections he adduced for
ultimately rejecting it, and that it is capable of establishing the
circumscribed version of universal doubt he sought in his 'First

I would like to thank Dr. Klempner for extending the opportunity of
editing this issue to me. It's been very challenging, but very
rewarding as well. And I'd like to thank Kamala and Matthew for their
excellent submissions, their many helpful suggestions and, not least,
for their encouragement, motivation and support. This project is very
much a collaborative one, and has only come to fruition because of our
mutual commitment to it.

Inquiries about tutorial assistance for UoLIP students are currently
handled by Karl Egerton, who can be reached at
kegert01@mail.bbk.ac.uk with the subject heading 'UoLIP tutor'.

(c) Eric DeJardin 2014

Email: dejardin@cox.net

About the editor:



The rejection of ethnocentrism, with its attendant sexist and racist
beliefs, in western liberal democracies has led to an erosion of
belief in moral objectivism -- the view that there are universal
moral principles. Multiculturalism has replaced ethnocentrism and
engendered the rise of moral relativism -- the belief in a theory
that there are no universally valid moral principles. So, all moral
principles are valid in the context of respective cultures and
societies, according to relativism. But moral relativism itself poses
problems, because it does not do justice, nor does it accurately
address conflict or disagreement. There is something fundamentally
infuriating about confronting, for example, the way women are
oppressed by Islamic radicals and being told that the differences in
attitudes are merely cultural. Some of us may want to say no; it is
not merely a case of cultural differences, but something objectively
wrong about oppressing women. In this paper, I explore Blackburn's
quasi-realist theory of ethics to see how he defends against the
challenge from relativism, and what kind of normative ethics are
compatible with quasi-realism.

Origins of quasi-realism

David Hume in his Treatise showed how morality is 'more properly felt
than judged of' and the drive to action comes from passions and
sentiments rather than reason. Hume argued forcefully that ethics is
primarily a practical subject and that moral distinctions are not
based on reason, but instead derived from moral sense. So how should
we interpret this? Mackie interprets this to mean that we can call
something virtuous if and because it produces a particular kind of
pleasure, and evil, if it produces a particular kind of pain. The
virtuousness or evilness is not in the objects themselves but in the
sentiments they evoke in us.[1] Hume's position on ethics attracted
few philosophers at that time but in the twentieth century objections
to realist theories in ethics, brought the focus back to Humean

In the twentieth century G.E Moore proposed what came to be known as
the 'Open Question Argument' against Moral Naturalism -- the view
that there are objective morals facts and properties and that these
are natural facts and properties.[2] Moore's argument tried to show
that any serious ethical theory needed to explain the distinction
between questions like 'What is goodness? And questions like 'What
things are good? '[3]. Judgments of Goodness according to Moore,
offer us an opportunity to pass normative judgments, where we can
condemn or endorse. The Open Question takes scientific and empirical
facts to be settled, but leaves open to interpretation, judgments on
morality. Moore, himself argued for a type of ethical intuitionism --
the view that moral judgments have a unique identity that can only be
evaluated by the process of intuition.[4] Confronted with the Open
Question, philosophers began to look to a new approach to ethical
matters, leading to the emotivism of A.J Ayer and Charles Stevenson.
Emotivism concentrated on the practical aspects of expressing
emotion, but their work too fell short of providing a satisfactory
theory of metaethics because in many cases, ethical discourse can be
understood to be unemotional. R.M Hare's prescriptivism provided a
better formulation for issuing ethical prescriptions, and provided a
broad canvas for identifying what exactly is being expressed. Hare's
work focused on moving from language oriented discourse to a
practical action theory. His moral judgments are therefore more
prescriptive rather than descriptive. However it is important to
remember that their prescriptive nature is not because of some
natural (descriptivism) or non-natural (Moorean intuitionism)
properties but because they guide action. Ayer's Emotivism and Hare's
Prescriptivism led the way towards a broadly Expressivist theory in
ethics, by making the assumption that 'the meaning of a sentence is
to be understood on the basis of the effects it is used to achieve or
the speech act it is used to perform.'[5]

The Expressivist approach conceives normative propositions to be
directives, and an obvious advantage of this approach, is that it
takes into consideration the motives of moral commitments. For
example, take the proposition 'X was a bad man'. Expressivism denies
that the speaker is describing her own mind (and thus differentiating
itself from subjectivism) but is instead voicing a position that must
be held. As we shall see by the end of this essay this balancing act
is very hard to pull off.

Quasi-realism flows from expressivism and has been described as an
antirealist, non-cognitive theory in metaethics, which claims that
ethical sentences do not express propositions, but instead express
emotional attitudes as though they are real properties. It thereby
earns the right to speak of morals in realist terms, while firmly
remaining in the anti-realist camp. Blackburn himself, recently
(November 14-15 2014) at a Conference on Moral Sentimentalism held at
Holy Cross College, USA, has expressed being uncomfortable with being
described as a Non-Cognitivist.

One way to approach quasi-realism is to view it as a means to rescue
expressivism from error theory associated with J.L Mackie. Mackie
states that 'the assertion that there are objective values or
intrinsically prescriptive entities or features of some kind, which
ordinary moral judgments presuppose is, I hold it not meaningless but
false.'[6] According to the above statement then, moral judgments
refer to truth-apt representations of fact, and since there are no
truth apt-representations of fact, moral assertions are false. Moral
discourse is thus referencing entities that are not there. We are
therefore in perpetual error when we assume that there are moral
entities, thus, the name 'error theory.' Mackie's error theory has
been described as a descriptivist antirealist position because it
maintains that there are no moral facts, and that moral judgments
describe the world. This combination of moral antirealism and
descriptivism is what makes Blackburn think Mackie's position is

Blackburn's response to Mackie's Error Theory is to reject that there
is any error. He denies that any 'first-order ethical practice
embodies any mistake',[8] one of the reasons being, that Mackie
himself never showed what a practical system of ethics would look
like if it was free of error.[9]

Is Quasi-realism fictionalism?

In a paper titled 'Quasi-realism is Fictionalism',[10] David Lewis
argued that the quasi-realism of Blackburn is a kind of moral
fictionalism. Lewis grants that quasi-realism succeeds 'on its own
terms', so that the quasi-realist is entitled to 'echo' everything
the realist says. He draws attention to the similarity quasi-realism
shares with fictionalism by noting that the quasi-realist echoes
everything the realist says but his assertions are quasi-assertions,
because 'they are preceded, explicitly or implicitly, by a disowning
preface. That preface, is to be found in the endorsement of
projectivism that precedes and motivates his advocacy of
quasi-realism... It is something the quasi-realist says the realist
will not echo.'[11] Lewis argues further, that since the
quasi-realist wants to say everything the realist wants to, he must
either be a realist or is pretending that realism is true. But since
he is not a realist, and is instead making believe realism is true;
he is a fictionalist. Does Lewis have a point? Is quasi-realism
really another term for Moral fictionalism? Not really; Lewis'
argument is predicated on the point that the quasi-realist wants to
say everything the realist wants to say but either she is a realist
or just pretending to be one. But this is exactly why Blackburn
rejected Mackie's error theory. Blackburn's objection was that there
was something fishy about promoting an ethical theory, with fingers
crossed behind your back.

The sense in which the quasi-realist wants to 'say everything the
realist does' is not the same as the sense in which the realist makes
assertions, nor does she want to pretend that she does. The
quasi-realist does not want to make any assertions at all when she
utters first-order moral pronouncements, which the realist absolutely
wants to do. The quasi-realist certainly says things that sound like
what a realist would say, but she wants people to understand them
differently. For the quasi-realist, her first-order moral
pronouncements are just expressions of attitudes. Taking this
argument further, we can say that expressing an attitude requires no
commitment to a realist style belief in moral propositions. In a
rebuttal to Lewis,[12] Blackburn respectfully argues that while Lewis
thinks that quasi-realism gains luster from its association with
fictionalism, he disagrees with Lewis on this. Blackburn views
fictionalism with deep suspicion, especially its application to
evaluative thoughts and philosophical discourse. Blackburn recognizes
that the term quasi-realism could prove misleading and evoke ideas of
an 'as if' philosophy. Meaning, that we can talk of morals and ethics
as if they exist, when in truth there are none. How then does one talk
of ethics in realist terms and ascribe it to an anti-realist?

In making a case against moral fictionalism, Blackburn suggests that
fictionalism should not just be a doctrine where we talk of something
as being true, which it is not, but something more ambitious -- a
doctrine, where falsity is integral to practice and where we must
take refuge in make-believe once this falsity is exposed.[13] The
difference with the quasi-realist, Blackburn argues, is that falsity
is not integral to practice, and therefore there is no need for
make-believe. When we talk of fictions, he further argues, we must
know the contrast with fact. Using the argument from Mackie,
Blackburn asserts Mackie himself never showed clearly what a
non-error based moral theory would look like. On the contrary, Mackie
makes several straightforward assertions on morality and what a good
life should look like in same old (supposedly tainted) vocabulary.
Mackie's thesis concludes that according to error theory, there is no
such thing as first-order moralizing but adopts the Humean view that
one can accept the social function of morality, and choose the moral
views and moral positions to adopt. Blackburn's point against Mackie
is simple: 'why should we choose to fall into error'[14] he asks.
Quasi-realism can therefore be interpreted as an attempt to save
expressivism from error theory by showing that 'ordinary moral
thought is not infected root and branch with philosophical myth.'[15]

Blackburn's defense against charges of moral fictionalism thus
focuses on proving how an error-based moral theory is not sustainable
in the way quasi-realism is, and how therefore fictionalism is not a
good interpretation of his anti-realist position.[16]

Meeting the threat from relativism

The threat from relativism is real if we are to accept Blackburn's
own description that 'an ethic is the propositional reflection of the
dispositions and attitudes, policies and stances of people'.[17] It
follows from this description that different dispositions and
attitudes will emerge from different ethics -- each having its own
truth apt claim. This in turn makes it eminently hospitable to the
slippery ethical slope called relativism, with its threat of
legitimizing different ethics, all in conflict with each other. But
Blackburn argues that there is no threat to quasi-relativism from
relativism because there is no problem of moral truth. Moral
opinions, he argues, do not represent actual moral facts but assess
actions, choices and attitudes.[18] They reflect which attitude or
action to adopt in situations. For example, suppose I express the
attitude that human beings should not be beheaded, and then I meet a
member of the terrorist group ISIS, that holds the view that
beheading is permissible. What quasi-realism allows me to say is that
'this view is wrong'. The relativist may respond that it is merely
your opinion, conditioned by your environment. The relativist can
also question the quasi-realist, being an antirealist, how one can
state that beheading people is a moral wrong, if as she believes,
there are no objective moral facts? The response, according to
Blackburn, is that being an antirealist does not mean that there are
'no facts of an ethical or normative kind'.[19] Quasi-realism does
not provide an explanation for ethical facts, but explains why the
reasons that propel philosophers to realism need not do so. The ISIS
group can believe beheading is correct without it being so.
Quasi-realists can state beheading is wrong whether anyone believes
it is so. So what exactly does this mean? By stating that beheading
is wrong, the quasi-realist regards any potential challenge to the
belief that beheading (or beating blind puppies or abusing women and
so on) is wrong, as being indefensible. This is what Blackburn means
when he says that quasi-realism regards a moral claim as being true,
independent of anyone's opinion about. The very definition of
expressivism used to mean that moral sentences can be explained
without appealing to their truth-conditions, but Blackburn seems to
have embraced that school of thought (still firmly in the
expressivist camp) that does not shy away from speaking about moral
truths or even moral facts. Blackburn endorses a minimalist or
deflationary theory of truth. As he says, 'To say that an ethical
view is true is just to reaffirm it, so it is if we add the weighty
words 'really', 'true', 'fact', and so on. To say that is objectively
true is to affirm that its truth does not vary with what we happen to
think about it, and once more this is an internal, first-order
ethical position.'[20]

Quasi-realism is thus comfortable with the idea that we can have
objective views, but if we were to change our minds about those
views, the underlying ethical facts would not change.

What kind of normative ethics are compatible with quasi-realism?

Blackburn argues that modern moral ethical theories are handicapped
by their Aristotelian and Kantian trappings in how they deal with
modern ethical conflicts.[21] Those trappings, he says lead one group
to assert, for example, that murderers will not flourish and another
to assert that 'they have transgressed against some rational
constraint on practical reasoning.'[22] Both have failed to capture
the picture accurately because murderers do get away with murder and
there really is no proof of a Kantian theorem of practical reasoning
that they trespass against. So these schools of ethical thought offer
no practical solutions to our everyday moral conflicts. Instead,
Blackburn argues, what is wrong in coming out and saying murder is
wrong; or oppressing women is wrong; or beating puppies is wrong? Is
that not enough? Why we should feel compelled to say anything more,
he asks. The right way to deal with murderers is to make sure, as a
society, that they be punished. It is up to individual societies to
punish transgressions.

On charges of being subjective about ethical issues, Blackburn argues
that it is a mistake to see it from the view point of a single BIG
Question. Instead, he argues, they are many little questions. So if
someone accuses me of being subjective when I condemn the Taliban's
oppression of woman, it is possible she is biased or unable to see a
different point of view, but she has no overarching single charge
that she can make stick. Individual judgments vary according
situations, and I could be mistaken in some situations, where if I
have a disagreement with another person about whether factory farming
is ethical or not, one of us is objectively right. The point Blackburn
is trying to make is that charges will vary because there really is no
single all-encompassing morality. Quasi-realism frees us in a way to
follow a practical ethic without committing to any moral absolutes.


At the very outset, I set myself two tasks in this essay. One, to
judge whether Blackburn's quasi-realism provides an adequate defense
against moral relativism, and second, what kind of normative ethics
are compatible with it?

To the first, Blackburn tries to walk a thin line between rejecting
the kind of relativism that asserts that had our conative impulses
been different, we would have embraced different ethical positions;
and had our conative impulses been different, we would have embraced
different ethical positions, and these positions would have been
right for us to embrace.[23] So, is what Blackburn suggesting that:
'I cannot be wrong about morality, but others can?' If that is the
case, then it resembles what the Australian philosopher David Stove
called the 'Ishmael Effect' a reference to Melville's Moby Dick,
whose narrator Ishmael, ends the book by stating 'I alone escaped to
tell the tale' -- an assertion that is hard to accept, given the tale
he tells about being rammed by the whale and being plunged into the
depths of the ocean.[24] Blackburn uses this reference, ironically,
to point out the flaws in the relativist position. The relativist
cannot claim that all human beliefs are subjective except, the belief
that all human beliefs are subjective. She cannot be an exception to a
fate that she claims befalls everyone.[25] Ironically, Blackburn
himself falls into the same trap that he claims the relativist is
vulnerable to. His position smells too much like subjectivism -- the
mind-dependent approach to moral discourse; to which charge, his only
defense is that that no other assessment is possible, because one
cannot evaluate and pass judgment without using those very same parts
of the brain that enable one to pass judgment.[26]

To the second question: about what kind of normative ethics are
compatible with quasi-realism? The answer is: whatever you want it to
be. It is true that this position can be deeply unsettling, especially
because we tend to grasp at 'Big Truths' that ultimately disappoint,
and in this way, we seem to be defined by our epistemic insecurities.
Blackburn, in his defense, seems to offer a way out. His is not a
'Big' overarching morality but instead, suggests that we address each
moral problem like individual problems; according to the beliefs and
attitudes, we hold individually. In this way, Simon Blackburn's
quasi-realism succeeds in threading the needle to allow us to, in one
breath, reject moral absolutes while retaining moral convictions.


Blackburn, S. (1993). 'Errors and Phenomenology of Value' in Essays
in Quasi-realism. In S. Blackburn, Essays in quasi-realism (pp.
149-165). New York: Oxford University Press.

Blackburn, S. (1993). Essays in Quasi-realism. New York: Oxford
University Press.

Blackburn, S. (1998). Ruling Passions: a theory of practical
reasoning. Oxford: Clarendon press.

Blackburn, S. (2005). 'Quasi-realism no fictionalism' in (Ed) Mark
Kalderon 'Fictionalism in metaphysics'. Oxford: Oxford University

Blackburn, S. (2005). Truth: a guide. Oxford University Press.

Blackburn, S. (2006). 'Antirealism, expressivism and quasi-realism'
in The Oxford Book of Ethical Theory(Ed) by David Copp. Oxford
University Press.

Blackburn, S. (Nov 10, 2010). 'Is Objective Moral Justification
possible on a Quasi-realist Foundation?'. Inquiry:An
Interdisciplinary Journal of Philosophy, 42:2, 213-227, DOI:

Hurka, T. M. (2010, March 25). Retrieved from Stanford Encyclopedia
of Philosophy: http://plato.stanford.edu/entries/moore-moral/

Kauppinen, A. (Spring 2014). Moral Sentimentalism. Retrieved from
Stanford Encyclopedia of Philosophy:

Kim, S. (n.d.). Retrieved from Internet Encylopedia of Philosophy:

Lewis, D. (2005). Quasi-realism is fictionalism' in Kalderon, Mark
(Ed) , Fictionalism in Metaphysics. Oxford University Press.

Mackie, J.L (1977). Ethics: inventing right and wrong . London:
Penguin Books.

Mackie, J.L (1980). Hume's Moral Theory. London: Routledge and Kegan

Moore, A. (July 2002). 'Quasi-realism and Relativism'. Philosophy and
Phenomenological Research, Vol LXV, No1.


1. Mackie, 'Hume's Moral Theory'

2. http://plato.stanford.edu/entries/naturalism-moral/

3. Moore G.E , 1903 'Principia Ethica'

4. http://plato.stanford.edu/entries/moore-moral/

5. Kauppinen, Antti 'Moral Sentimentalism' Stanford Encylopedia of
Philosophy Spring 2014 page 21
moral-sentimentalism (Kauppinen, Spring 2014)

6. Mackie J.L (1977) ' Ethics: inventing right and wrong' London.
Penguin Books page 40

7. Moral realism in 'The Internet Encyclopedia of Philosophy'

8. Blackburn, Simon (1999) 'Is Objective Moral Justification possible
on a Quasi-realist Foundation' Inquiry: An Interdisciplinary Journal
of Philosophy 42:2

9. Blackburn, Simon (1999) 'Is Objective Moral Justification possible
on a Quasi-realist Foundation' Inquiry: An Interdisciplinary Journal
of Philosophy 42:2

10. Lewis, David ' Quasi-realism is fictionalism' in Kalderon, Mark
(Ed) , Fictionalism in Metaphysics

11. Lewis, David 'Quasi-realism is fictionalism' in Kalderon, Mark (
Ed) , Fictionalism in Metaphysics, page 315

12. Blackburn, Simon, ' Quasi-realism no Fictionalism' ' in Kalderon,
Mark (Ed) , Fictionalism in Metaphysics

13. Ibid

14. Blackburn, Simon 'Errors and Phenomenology of Value' in Essays in
Quasi-realism page 150

15. Blackburn, Simon 'Antirealism, expressivism and quasi-realism'
page 154 in The Oxford Book of Ethical Theory

16. Kalderon, Mark ' Introduction' in Kalderon, Mark (Ed) ,
Fictionalism in Metaphysics

17. Blackburn, Simon quoted in (Moore, July 2002)' Quasi-realism and
Relativism' Philosophy and Phenomenological Research Vol. LXV, No 1,
July 2002

18. Blackburn, Simon (1999) 'Is Objective Moral Justification
possible on a Quasi-realist Foundation' Inquiry: An Interdisciplinary
Journal of Philosophy 42:2 page 214

19. Ibid

20. Blackburn, Simon ' Ruling Passions' page 296

21. Blackburn, Simon (1999) 'Is Objective Moral Justification
possible on a Quasi-realist Foundation' Inquiry: An Interdisciplinary
Journal of Philosophy 42:2 page 222

22. Ibid

23. Moore A.W. 'Quasi-realism and Relativism' Philosophy and
Phenomenological Research Vol LXV , No1 , July 2002

24. Quoted in Blackburn, Simon 'Truth' page 47

25. Blackburn, Simon ' Truth' page 47

26. Blackburn, Simon 'Antirealism, expressivism and quasi-realism'
page 154 in The Oxford Book of Ethical Theory

(c) Kamala Vadlamani 2014

Email: kamala.vadlamani@gmail.com



Jackson in his Knowledge argument (1982) concluded that one who would
have a certain phenomenal experience for the first time would learn
something new despite having previously known all physical knowledge
pertaining to that kind of experience. He claimed that experience
makes non-physical information available, and thus physicalism, the
metaphysical position that all information is physical information,
is incorrect. One attempt at defending physicalism, put forth by both
Nemirow (1980) and Lewis (1983) was known as The Ability Hypothesis,
AH. AH, rather than concluding that one gains propositional knowledge
or no new knowledge at all, holds that the kind of knowledge gained is
only that which is involved in gaining certain abilities; those of
remembering, recognizing and representing the phenomenal character in
question; let's for brevity refer to these three as the 3 Rs.
Furthermore, Lewis argued that the 3 Rs are necessary and sufficient
conditions for knowing what some phenomenal experience is like.
Michael Tye (2000), however in his paper 'Knowing What it is like:
The Ability Hypothesis and The Knowledge Argument' successfully
argues how the AH fails to do what it set out to do. The
fined-grainedness of experience outstrips our ability to recognize a
difference between two shades of red that are located next to one
another on the colour spectrum. The average human just cannot
recognize, remember or imagine a difference between red #4 and red
#5. Despite lacking these abilities, Tye points out that we can
successfully experience both shades. Thus, acquisition of the 3 Rs is
not necessary for knowing what it is like to have a particular
phenomenal experience and the AH fails.

Why aren't acquisition of the 3 Rs necessary? Consider Baby Harry. He
is one month old and there is a good chance that he cannot see
anything further than 12 inches away from him. He does have the
ability to see colour but can barely hold his eyes fixed long enough
to track a moving rattle in front of his face. Considering Harry's
situation, there is a good chance that he lacks all of the concepts
that you and I have. Despite life without concepts, he is still able
to have phenomenal experiences. Thus having phenomenal experiences
does not require concepts. With this in mind, how in the case of
Jackson's Knowledge argument, could Mary's gaining the abilities to
remember, recognize and represent be claimed necessary and sufficient
for the kind of knowledge that she gains from experience? It seems
that if Harry can experience what he does without concepts, Mary's
gaining certain abilities which presuppose certain concepts for their
success is something additional to anything she learns having a
certain phenomenal experience.

If this is correct, it should come as no surprise as to why the AH
was not successful. Any hypothesis made in response to a question
which requires by the very nature of its subject matter a
non-conceptual 'response' is bound for failure; any meaningful answer
in any language will fail, given there is no language in the absence
of concepts. This being said, although 'knowing what it's like'
phenomenally may outstrip those conditions formulated by the AH,
there seems to be something intuitively correct about those same
conditions when viewed as integral to our moralizing; 'knowing what
it's like humanly'. In this essay, I will firstly argue that gaining
the 3 Rs is necessary to empathize. Lastly, I argue that because
empathy is something that is necessary for moralizing, moral values
and attitudes are dependent upon the concepts underlying the 3 Rs as
given by the AH. Thus the insights of the AH illustrate that to
moralize is to make use of various conceptual abilities.


The ability to recognize some object, x, successfully presupposes an
ability to identify that x which one has previously been attentive
to. Such an identification demands of an agent that she be acquainted
with various criteria to which she can apply the concepts of
'similarity' and 'difference'. It is only through the application of
these two concepts that one can say (or think) of something x that
given similarity in criteria, C, it the same type or token as
previously experienced. Whether it is the same x now at time t that
was experienced at t1 or it is another token x of the same type, the
concept of similarity (and difference) are necessary for recognition.
Similarly, the ability to remember can generally be defined in terms
of one's successes in bringing to mind past experiences or facts.
Such successes are evidenced in the often practical natures of memory
application. The success of my remembering that the water in one of
the tea cups setting before me is boiling hot, is evidenced in my not
severely burning myself. Without possession of the concepts of
'similarity' and 'difference' with respect to C, the boundary between
imagination and memory becomes blurred.

1. One objection to this line of argument is that a cat may certainly
recognize its owner, but on the given account this would entail that
animals posses concepts and there is something wrong about this. To
this I might argue that non-human animals and very young children are
limited to having feelings of familiarity with respect to certain
stimuli. This is to be distinguished from recognition; a process that
surely seems to require one's having classificatory concepts and is
thus itself required for language competence. If words only felt
familiar and were not recognized, meaning could never get off the
ground; a sentence might be a guessing game with no rules nor
consequences other than sounds of air being produced.[1]

Because the abilities to recognize, remember and represent presuppose
the possession of the basic concepts of similarity and difference with
respect to some particular criterial set, and that the successful
application of these concepts require identification or
re-identification of some object, the 3 Rs must therefore somehow
play an essential role in the identification and re-identification of
objects. This, may seem painfully obvious to some. If it's not
obvious, consider the following; Is it possible for you to identity
the very house you live in and yet simultaneously lack the ability to
recognize it? Lacking the ability to recognize x is just to say that
one lacks the ability to identify x. The same follows with
remembering x. Classifying a certain cat as a Russian Blue requires
that one apply a certain relational concept of likeness to the object
in question and to some criterial model. The path from one to the
other necessarily involves remembering.

Representing x however, seems to be a bit more difficult to prove
necessary with respect to identification. If we can identify an
icosagon drawn on a page but we cannot represent the same figure in
our minds, then representing that figure cannot be necessary for the
ability to identify it. One must ask with respect to the
fine-grainedness of identification cases, like that of a icosagon,
can one really ever identify that polygon as such in the first place?
Isn't it more likely that what we are doing when we think we are
identifying such a many-sided figure is counting the sides
independently rather than recognizing the figure as a whole? If this
the case then our identification of it is not as a twenty-sided
figure per se, but as figure that contains 5 four lined segments of
equal length; this of course depending upon our ability to
discriminate line composing line segments. Now given the limits of
the length which it is possible to hold an image in mind and the fact
that we cannot compare one mental image with another which serves as
an imaginary stable model, it comes as no surprise that we cannot
represent a twenty-sided figure or discriminate it from one that is
nineteen-sided; we are not in the position to count the sides as we
can when it is drawn upon a page. If the average person hasn't the
ability to discriminate between a nineteen-sided figure and a
twenty-sided figure, even when drawn on a page, without counting the
sides of the two figures, why should we expect our ability to
represent such a figure in mind to be any less dependent upon

I would like to emphasize a very important the idea which comes to
the fore when thinking about the limits of our ability to represent;
that having these limits forces us to decide which features of
objects are important enough for us to say that some object is the
same type as another. Such features, it would seem, are chosen with
respect to our purposes. Two dogs are of the same breed and could
visually be identical but there is only one that I am interested in
identifying after I've been informed that only one of them is rabid.
This interest is telling of my value for my own health which is the
purpose of my choosing to discriminate between the two otherwise
identical dogs.

One purpose which seems to be of extreme importance with regard to
the ability to identify objects according to some relevant feature is
that of defending human value; and this bring us to the notion of
empathy. Empathy, can be defined in many ways; an ability to
harmonize emotionally, the ability to 'put oneself in another's
shoes', the ability to imagine what it's like for someone else. It is
this last simple definition which I would like to concentrate on for a

Just how is the ability to identify x related to the ability to
empathize with x? Maybe this is the wrong question to pursue; the
better one being 'how is the ability to identify some object, o, as x
related to the ability to empathize with o?' Looking back at the
simple definition we have chosen to consider, one answer could be
that they are related through an ability to imagine. Now Prinz
rejects this kind of definition of empathy, saying of one very much
like it that,

     [First] the appeal to imagination seems overly
     intellectual. Imagination sounds like a kind of mental act
     that requires effort on the part of the imaginer.
     (Prinz, p. 212)
     It seems inflated to call a contagion ['catching the
     emotion another person feels'] an imaginative act.
     (Prinz, p. 212)

However, this is exactly what we have seen above not to be the case.
Unlike phenomenally experiencing redness, representing something does
require effort and presupposes acquaintance with various concepts --
such as that of likeness. In terms of considering empathy being a
contagion, I can see no reason to disagree that this could be the
case. However, its being such is consistent with empathizing
requiring the abilities to identify and to imagine. How can this be
so? Well, the emotion of fear could also be said to be a contagion
and yet it seems that to 'catch' this emotion one must have the
ability to recognize that someone is feeling fearful and recognize
that that someone is like us enough to matter. The autumn wind in
downtown Chicago sometimes screams but it hardly manages to elicit

Now it may be uninteresting to ask the question, 'can one empathize
with something one cannot identify?' However, to ask the question,
'can one empathize with something one cannot identify with?' is more
interesting. We do not empathize with rocks, mountains or trees. It
certainly could be debated whether or not we empathize with lower
animals such as a snakes, dogs, or cats. However, to completely deny
that when a dog cries out, one is not moved and cannot understand the
dog's pain in such a way as to remind one of one's own pain, would be
to ignore something of great importance; that is that empathy seems
to be a gradient phenomenon. We all do not share the same levels of
empathy for the same objects nor do all classes of object inspire
empathy to the same degree. One thing that may be certain, however,
is that most normally psychologically functioning agents are most
likely to feel empathy towards those things which most resemble

Perhaps this is too fast and not careful enough. If empathy requires
the ability to represent to oneself how it feels for some other
thing, then such a requirement presupposes firstly, that the other
thing can enter into causal relations in a similar fashion as do I.
Secondly, it presupposes that those causal relations end in states
that are similar to those which I have. Thirdly it presupposes that
their behavior is as a good a guide to their feelings as my own are.
If by our natures as metaphysically subjective beings we can only
know what our own subjective states are phenomenally like, then these
presuppositions are just three variations of a way to make the claim
that empathizing requires some minimal degree of identification on
the part of the empathizer with those empathized with. The higher
this degree is, seemingly the less there is to infer or imagine in
terms of what someone who is like me must be feeling.

This kind of view I assume will be met with an objection similar to
that of Prinz's; that empathizing does not require that one go
through some criterial list and tic as many boxes as one sees fit in
order to react to some state of affairs afterward. It is something
that happens automatically without thinking about likeness or
similarity or our own cases for that matter. I would have to agree
that it does 'seem' automatic. However, perception's not 'seeming'
like it's a process does not entail that it's not a process.
Furthermore, there is no reason for anyone to hold the position that
identifying with someone, which can be considered a type of
conceptual process, and the feeling of empathy must occur
simultaneously. If we recognize ourselves by means of being related
causally to the world in a way particular to the very spatial
location that each of us individually occupy at any moment in time,
then the concept of likeness to my own case will be that which
characterizes others' cases by my inferring a similarity in causal
interaction. Once my own case is established, it is something which I
have the ability to remember, recognize and represent generally
without the need for computation or inference. It is what I refer to
when I think and utter 'I'. Perhaps this comes close to what Frege
had in mind when claiming that I-thoughts are 'primitive' and what
Kripke labels as 'autonomous designation'. If this is true, then
identifying something to be causally like me, may only seem automatic
given that one part of the relation in question is always available to


It was noted above that the criteria with which we judge something as
being similar to (or different from) are sensitive and telling of our
purposes and values. If the criterion for determining that something
is human is bound to the degree to which it is similar to us and
assuming that psychologically normal agents do have a sense of
self-esteem, then such a criterion of humanity is telling of our
value for a class of things to which we invariably belong. This
results in the notion whereby those whom we can identify with the
most -- those who we recognize as human by our own standards and
those who we can thereby infer via our own case 'what it is like' for
them -- are those whom we most ultimately value. Our moralizing, in
particular having of reactive attitudes we do, could be considered
evidence of this. If a tree falls and kills a man in a hurricane, we
neither hold it nor the hurricane responsible for murder. If a dog
bites its owner, although the owner might very well be angry as a
result, if she is a rational being she will not morally judge the
dog. We do, however, hold those who we feel are like us causally, and
hence who are like us motivationally, accountable.

Now of course there is a danger looming in the background; that of
the moral egoist. A position like this would reason something like,
'it is because you are similar to me, that I can empathize with your
situation -- but it is really my own situation that is making me
uneasy and I find it morally wrong that I feel this way regardless of
how you may feel.' This position is considered by many an undesirable
one, however, is not a position that follows simply from one holding
that empathy is essential to the having of moral reactive attitudes.
Just because I am not in a metaphysical position to ever be able to
know your pain the way you know it or experience redness exactly as
you experience it, does not entail that my ability to identify with
you via my own case will land me in the position of a moral egoist
any more than my being limited to experience the world given my
particular spacial extension at any given time must land me in a
position of being a solipsist.

Strawson in his seminal essay, 'Freedom and Resentment' put it very
nicely. Crudely paraphrased, it is only those who we feel to be
morally accountable agents that we have reactive attitudes towards.
Our moral reactive attitudes are essential to us as humans. Others,
who for some reason or another lack the ability to have reactive
attitudes are as a result 'objectified' and held unaccountable. One
way to interpret this is that our humanity is a common identity
amongst seemingly causally efficacious beings and it is this which
grounds our ability to empathize and thus moralize; holding others
like us responsible. It justifies our moral reactive attitudes even
if we turn out not to be as causally efficacious as we consider
ourselves to be. So if we hold our humanity as a reason for having
the moral reactive attitudes we do, then given our metaphysical
circumstances as humans, the inability to empathize without some
sense of being able to identify with someone via our own subjective
cases mustn't lead to a moral egoist view; only a view that is
unmistakably human.

Can moral judgement occur without the ability to empathize? Sure, why
not? It is possible to train a group of people to utter 'stealing is
wrong' and sometimes this kind of passive judgement -- when
reinforced by punitive acts -- is all that is necessary to dissuade
when there is a lack of empathy to be had. Can moral reactive
attitudes occur without empathy? Surely not! If what has been put
forth has been correct, moral reactive attitudes depend on a sense of
identification between beings who can from their own case imagine what
it is like for someone else. It is this knowing what it's like for
someone else which requires the 3 Rs and thus they are essential the
kind of moralizing that we are personally invested in. It is due to
our being personally invested that our moral judgements based upon
our reactive attitudes are self-motivating; we are our own enforcers
of those acts which uphold that which we value.

But what about morally condemning acts that do not involve directly
harming people? Dumping toxic waste into the ocean is considered
wrong but one doesn't usually have feelings of empathy towards the
ocean. 'Wrong' in this case could just be a judgement resulting from
a social convention. Again it would be hasty to conclude that because
these kinds of moral judgements do occur that all moral judgements are
of this kind; for this example can be explained otherwise. Firstly,
this objection assumes that empathy is only something felt towards a
moral victim. In judging such an act wrong, it is our empathy which
is extended towards those who commit such acts which justify our
holding them as causally and morally accountable. It is their
motivations that are disapproved of given that they are human enough
to be held accountable. Secondly, the ocean is full of animals and
for many with an extended sense of empathy, being an animal is being
similar enough to us to imagine what it's like for them; that they
could like us feel pain and thus to be a reason for concern. Hence,
the abilities necessary for empathy are essential to a moralizing
that counts motivations as those things that are the objects of our
approbation or disapprobation.

It has thus been argued that empathy is necessary for a kind of
moralizing that involves reactive attitudes and that empathy itself
necessarily involves three conceptual abilities. Thus, the kind of
moralizing which is based upon our moral reactive attitudes must
necessarily require the concepts inherent in the 3 Rs. If this is the
case, then perhaps the the insights of the AH can be directed away
from anti-physicalism and redirected toward an argument against the
objectivity of moral values. Such an argument is beyond the scope of
this paper, however, I mention it only as a possibility that becomes
available as the result of accepting this argument as plausible.


I believe it was Moore that once said something to the effect that
many of the problems of philosophy are the result of asking the wrong
questions. The Knowledge argument poses such a question. By demanding
that a description be given of a subjective phenomenal experience --
something that by its nature falls short of analysis by means of
objective concepts -- it is asking for something that is
metaphysically impossible. This however doesn't prove that
physicalism is wrong any more than my not being able to occupy the
same space as you proves that physicalism is incorrect. Such a
question is the wrong one to ask and in agreeing to answer it, the AH
has not a chance but to fail. However, I have argued that the 3 Rs
when understood as basic to the concepts that ground empathy, get us
closer to understanding how moral reactive attitudes depend upon
concepts. Take away these concepts and one not only takes away the
accountability necessary for having moral attitudes but one does away
with any chance of being human.


Frege. G., 1918, 'Der Gedanke. Eine Logische Untersuchung', in
Beitrage zur Philosophie des deutschen Idealismus, I (1918-1919):
58-77. Translated as 'Thoughts', by P. Geach and R. Stoothoff, in
McGuinness (ed.) 1984.

Jackson, F. 1982 'Epiphenomenal Qualia,' Philosophical Quarterly, 32,

Kripke, S., 2008, 'Frege's Theory of Sense and Reference: Some
Exegetical Notes', Theoria, Blackwell Publishing Ltd.

Lewis, D. 1990 'What Experience Teaches Us,' in Mind and Cognition: A
Reader, ed. by W. Lycan, (Oxford: Blackwells).

Nemirow, L. 1980 'Review of Nagel's Mortal Questions,' Philosophical
Review, 89, 473-477.
Prinz, J., 2011, 'Is Empathy Necessary for Morality', in Amy Coplan
and Peter Goldie, Empathy: Philosophical and Psychological
Perspectives, Oxford University Press.

Strawson, P. F., 1962. 'Freedom and Resentment,' Proceedings of the
British Academy, 48: 1-25. Reprinted in Fischer and Ravizza, 1993.

Tye, M., 2000, 'Knowing What it is Like: The Ability Hypothesis and
the Knowledge Argument,' in M. Tye, Consciousness, Color, and
Content, Cambridge, MA: MIT Press.


1. I would like to thank E.F. Dejardin for bringing this objection to
my attention.

(c) Matthew Sims 2014

Email: sims303@hotmail.com




Let's call an argument the conclusion of which calls into doubt our
ability to ascertain either the veracity or justification of some
particular belief or set of beliefs a 'Skeptical Argument.'[1] And
let's say that two or more skeptical arguments have the same
'skeptical scope' if they call into doubt our ability to ascertain
the veracity or justification of the same set of beliefs. In his
'First Meditation', Descartes employs two main Skeptical Arguments to
establish the conclusion that he has grounds to doubt everything he
formerly believed. Let's call this conclusion 'Descartes' Doubt'. He
uses both arguments, which we shall call the 'Dream Argument' and the
'Deceiver Argument,' because he believes that they differ in skeptical
scope. More precisely, he argues that only the Deceiver Argument can
establish Descartes' Doubt. For he believes that the skeptical scope
of the Dream Argument is circumscribed by what we shall call his
'Three Responses' to it. We shall refer to them as the 'General
Response,' the 'Basic Response' and the 'Mathematical Response.'

I shall argue that Descartes is wrong about the skeptical scope of
the Dream Argument. For I think that it has the resources to refute
each of the Three Responses. That is, I shall argue that the Dream
Argument has the same skeptical scope that Descartes concludes the
Deceiver Argument has. Hence, I conclude that by Descartes'
standards, it too can establish Descartes' Doubt.

My argument concerning the skeptical scope of the Dream Argument
assumes the following claims. First, I shall assume that the Deceiver
Argument successfully establishes Descartes' Doubt. Second, I shall
assume that it does so because the ground of doubt on which it rests
is immune to the Three Responses. Hence, it follows that I shall also
assume that if the Dream Argument can withstand the Three Responses,
then it too can establish Descartes' Doubt. Therefore, all I'm trying
to establish in this essay is that the Dream Argument can withstand
the Three Reposes. For given the three plausible assumptions laid out
above, if I can establish that conclusion, I'll have shown that the
Dream Argument has the same skeptical scope as the Deceiver Argument.

This is not to say that each of these assumptions is uncontroversial;
I grant that they could each be questioned. Rather, I've presented
them here only to delimit the scope of my argument from the outset to
the issue of whether the Dream Argument can refute the Three Responses.

This essay shall proceed as follows. In Section One, I shall explain
some preliminary notions that are essential to understanding the
argument I shall subsequently develop. Then, in Section Two, I shall
explicate the Dream Argument. Finally, in Section Three I shall
explain the Three Responses, and show that the Dream Argument can be
used to refute each of them.


The 'I' of Descartes' 'First Meditation' is an empiricist (Frankfurt
1970, pg. 32). That is, he believes that everything we know is
ultimately grounded in sense perception. We shall refer to him as
'The Empiricist'.

We must clarify, however, the sort of empiricism to which The
Empiricist is committed. Importantly, it does not appear to be the
sophisticated empiricism of the trained philosopher, which admits the
possibility of a priori knowledge while merely denying that it
pertains to the world we experience. Rather, it seems to be what
Frankfurt calls the 'naive empiricism' of the 'philosophical novice'
who 'ascribes to the senses... many things that are not properly to
be found in them' (Frankfurt 1970, pp. 61-62). Frankfurt defends this
conception of The Empiricist by referencing a passage from the
Conversation with Burman, in which Descartes identifies The
Empiricist with 'a man who is first beginning to philosophize' and
who is thus 'limited to the senses' (Frankfurt 1970, pg. 62). As
such, Descartes continues, he is incapable of considering anything
'in the abstract and separated from matter and particular instances'
(Frankfurt 1970, pg. 62). Hence, The Empiricist of the 'First
Meditation' would be unable to entertain the sort of abstract
reflections on a priori propositions that are readily available to
the sophisticated empiricist. In this essay, I shall adopt this
Frankfurtian conception of The Empiricist.

While the empiricism of The Empiricist may be naive, he is not. For
as the 'First Meditation' opens we learn that The Empiricist has
decided to question his empiricism, since he desires certainty, yet
has discovered that some of his beliefs are false. Those isolated
instances of false belief, he fears, may by symptomatic of epistemic
rot at the base of his doxastic structure. He thus concludes that if
certainty is his aim, he can best achieve it by first rejecting all
of his dubious beliefs. And since, qua empiricist, he judges his
sensory beliefs in general to be both the most reliable and the most
fundamental of all his beliefs, he decides to begin his investigation
by questioning them.

But how is he to determine which beliefs are dubious? Since his aim
is certainty, The Empiricist adopts a stringent standard: If a ground
for doubt that he believes is merely possible can be adduced for a
particular belief, then it's a sufficient ground, and that belief
must be jettisoned as if it were false. Let's call a ground for doubt
that satisfies this inordinately low standard of dubiousness a
'Cartesian Ground'.

The Empiricist concludes, by the end of the 'First Meditation', that
he has found Cartesian Grounds for doubting all of his former
beliefs. That is, he has at that point established Descartes' Doubt.
As we said above, Descartes' Doubt is ultimately established by what
we have called the Deceiver Argument. However, Descartes in fact
formulates the Deceiver Argument as one horn of a dilemma. Either the
faculties by which we come to hold beliefs or acquire knowledge, which
we shall call our 'cognitive faculties', were created by an omnipotent
god, or they were not. If they were created by omnipotent god, then he
may have designed them such that we're continuously deceived, both
when we experience the world and when we reason about it. Thus, if an
omnipotent god exists, The Empiricist has Cartesian Grounds to doubt
everything he believes. But if we're not the product of an omnipotent
god, then our cause, being less than omnipotent, is likely to have
produced us with unreliable cognitive faculties. For they would then
lack, as it were, the guarantee of reliability that omnipotence alone
can provide. Thus, if an omnipotent god does not exist, The Empiricist
has Cartesian Grounds to doubt everything he believes. Hence, since he
cannot confirm that whatever caused his existence has provided him
with reliable cognitive faculties, The Empiricist has Cartesian
Grounds for doubting any belief or claim to knowledge that they

I shall argue that we need not consider the atheistic horn of
Descartes' dilemma. For as long as one concedes the possibility that
an omnipotent god exists, Descartes's Doubt is generated. And I
suspect that even most atheists would grant that god's existence is
minimally possible, even if they deny that certain well specified
conceptions of an omnipotent god could possibly obtain. Therefore, we
shall suppose that Descartes' Doubt is adequately generated by the
theistic horn of the dilemma alone, i.e. by the Deceiver Argument.

We've stipulated above that the Deceiver Argument provides The
Empiricist with Cartesian Grounds for doubting everything that he
believes. But the Deceiver Argument is only adduced after The
Empiricist posits and then rejects what we have called the Dream
Argument. For The Empiricist argues that what we have called his
Three Responses to the Dream Argument permit him to maintain his
empiricism even if he grants the argument's conclusion.

We've now completed our account of the fundamental elements of
Descartes' 'First Meditation' that are relevant to the issue we're
investigating in this essay. In the next section, I shall develop the
version of the Dream Argument with which we shall subsequently be


Descartes' Dream Argument presupposes that we have genuine
experiences while dreaming. Let's call these 'Dreaming Experiences',
and contrast them with our ordinary non-dream experiences, which we
shall call 'Waking Experiences'. (I shall ignore, for the sake of
simplicity, non-ordinary Waking Experiences such as hallucinations,
illusions, etc.) It begins with the premise that our Dreaming
Experiences often fail to correspond with the Waking Experiences we
would be having if we were instead awake at the time of a dream's
occurrence. For example, I may dream that I'm standing near the ocean
while I'm in fact lying in bed in the desert. Let's call this the
'Correspondence Failure Premise'. The argument further supposes that
it's minimally possible for any imaginable Dreaming Experience to
match the degree of vivacity that's attained by our most vivid Waking
Experiences. Let's call this the 'Vivacity Premise'. Finally, the
argument contains the implicit premise that whatever I can experience
while awake is a possible element of Dreaming Experience as well.
Let's call this the 'Experiential Premise'.

From the conjunction of the Vivacity and Experiential premises, it
follows that I can never be certain that what I'm experiencing at any
given time is a Waking Experience as opposed to a Dreaming Experience.
Why? Because any possible Waking Experience is also a possible
Dreaming Experience, and both Dreaming Experiences and Waking
Experiences can attain the same degree of vivacity. That is, they can
in principle be qualitatively indistinguishable qua experiences. If we
couple this sub-conclusion with the Correspondence Failure Premise, we
can infer the ultimate conclusion of the Dream Argument, viz. that I
have Cartesian Grounds for doubting any specific experience I might
have. For if I can never be certain that I'm not dreaming, then given
the frequent failure of correspondence between Dreaming and Waking
Experiences, I always have Cartesian Grounds for doubting the
veridicality of the specific content of any particular experience.

Indeed, even If my Dreaming Experience did correspond perfectly with
what my Waking Experience at any moment, I still could not conclude
that my Dreaming Experience is veridical. First, I would be unable to
adduce any reasons to support the conclusion that my Dreaming and
Waking Experiences coincide. For to have a Dreaming Experience just
is to be cut off, as it were, from Waking Experience. And this point
is not weakened by the fact that our Dreaming Experience can
sometimes be affected by phenomena we would also experience were we
awake. Indeed, we often notice ex post facto that although, say, a
particular sound in the vicinity of our sleeping body may have
altered our Dreaming Experience, we don't necessarily experience it
as we would have were we instead awake. The sound of a book that's
been knocked to the floor by my cat can be transformed by the alchemy
of Dreaming Experience into the sound of a gunshot by a masked

But second, and more importantly, the Dream Argument implicitly
distinguishes 'sensations' from 'perceptions'. Let's say that
perception necessarily involves a causal relation between a subject
and an object in which the object somehow causes a sensation in the
perceiving subject. And let's say that a sensation of an object in an
act of perception is best characterized as a mental representation, of
some sort, of the object perceived. Now what the Dream Argument
presupposes is that sensation can occur in the absence of perception;
that is, it supposes that perception is a sufficient but not necessary
condition of sensation. For ordinary Dreaming Experiences just are
sensations sans perceptions. Hence, it follows that even if my
Dreaming and Waking Experiences did correspond perfectly, I still
would not be justified in concluding that my Dreaming Experiences
were veridical. For as sensations alone, the content of Dreaming
Experiences would lack any causal relation (at least at the time of
their occurrence) with the objects they seem to represent.

Finally, one might object that the Dream Argument is self-refuting,
for the Correspondence Failure premise presupposes that Dreaming
Experiences can be distinguished from Waking Experiences, while the
argument's conclusion asserts that the two cannot be distinguished.
But this objection can be met by requiring only that one both possess
the concept of a dream, and concede that dreams are possible. For then
the Dream Argument can be run without presupposing that one has
previously distinguished Waking Experiences from Dreaming Experiences.

Now that we've formulated the Dream Argument, we can examine the
Three Responses that The Empiricist raises to it, and show how the
Dream Argument can refute each of them.


The Empiricist grants the conclusion of the Dream Argument, but
proceeds to develop his Three Responses in an attempt to defend his
empiricism from it. The first of them that he considers is the
General Response. The Empiricist reasons that the general types of
objects that Dreaming Experiences comprise must exist to serve as the
source of the specific content of our dreams. So, even if the specific
tree or hand or star that I'm experiencing now doesn't exist, trees
and hands and stars must exist. And it's my previous sensations of
those types of objects that provided me with the imaginative
resources to dream about them now. But I'd argue that we can conceive
of the possibility of waking from a dream to discover that none of the
general objects it comprised accurately represented a type of extant
object. I may dream tonight of a world that comprises nothing but a
consistent taxonomy of cubist objects, none of which exist in the
actual world of Waking Experience. Hence, the Dream Argument provides
us with Cartesian Grounds for doubting the existence of the general
types of objects we encounter in dreams. Therefore, the Dream
Argument can withstand the General Response.

The Empiricist tacitly concedes the weakness of the General Response
(Descartes 1985, pg. 13). But not so with the Basic Response, which
maintains that not merely all actual general objects, but all
possible general objects, must be composed of fundamental elements,
such as extension, shape and size. Hence, even if our dream contents
comprise nothing but non-existent general objects, they must
minimally comprise the basic elements out of which all possible
general objects are composed. And the ultimate source of our
knowledge of these basic elements is sense experience. Thus, in this
attenuated sense, our Dreaming Experience accurately represents at
least the basic constituents of the real world of Waking Experience.

We may employ Descartes' supposition that god may exist to refute the
Basic Response with the resources of the Dream Argument alone. For god
is posited to exist both immaterially and eternally; that is, he
exists in a mode that's devoid of matter and that's outside, or
rather not in, space-time. Further, it's plausible to suppose that
Descartes' god, which resembles a disembodied mind, has something
analogous to the experiences we commonly attribute to the sorts of
minds with which we're familiar. Finally, it's possible that if such
a god exists, he has the power to create beings that, like himself,
are both immaterial and possess the capacity for experience. Let's
call such beings 'spirits'.

Now is it impossible that I'm in fact a spirit in an immaterial world
who's merely dreaming that I'm an embodied being in a world that
comprises general objects composed of basic elements? If we can
coherently define the notion of 'dreaming' so it applies both to
beings like us and to spirits, then it's not impossible that I'm such
a being having such a dream. And it certainly seems prima facie
plausible that we can so define the notion of dreaming. For example,
nothing is obviously incoherent about the notion of dreaming on a
substance dualistic account of the human person, some versions of
which would seem to require that immaterial minds dream. Further, I
now believe that I am (at least) a material being in a material
world. But it seems possible that I could dream tonight that I'm a
spiritual being in an immaterial world. That is, the notion that I
could dream that I'm a spirit in an immaterial world is not obviously
inconceivable. Yet if this is correct, then our dreams even now need
not involve any of the elements that the Basic Response is concerned
with. And hence our dreams can in principle differ so radically from
reality that not even the basic elements that they comprise are
common to both Dreaming and Waking experiences. Let's call this
attempted refutation of the Basic Response the Immaterialism Argument.

I shall consider two objections to the Immaterialism Argument. The
first is that it supposes that there is a coherent notion of dreaming
that applies equally well to both embodied persons and to spirits.
Hence, to support the notion that it's possible that I'm a dreaming
spirit, I must provide that definition, which I've thus far failed to

But I don't think this is so, at least given the standards that The
Empiricist has set for himself and his project. For recall that a
ground for doubt qualifies as a Cartesian Ground for The Empiricist
if it's merely possible that it obtains. For example, it may be the
case that the notion of an omnipotent god is in fact incoherent, and
hence that it's impossible that such a god exists. But as far as I
know, this has never been conclusively demonstrated. Let's grant,
arguendo, that this is correct: no one has successfully shown that
the notion of an omnipotent god is incoherent. Further, though, let's
grant that the notion of an omnipotent god does in fact represent an
impossibility. Given these conditions, would The Empiricist thus be
debarred from appealing to the possibility that an omnipotent god
exists in making his skeptical case? It certainly doesn't seem to me
as if he would be. For The Empiricist has no way of knowing that an
omnipotent god cannot exist; for all he knows, it's minimally
possible. And this is sufficient to provide him with Cartesian
Grounds for any doubt that the existence of an omnipotent god could
legitimately raise.

This example clarifies the notion of a Cartesian Ground of doubt by
precisifying the conception of 'possibility' it rests on. For it
shows that we must distinguish epistemic possibility from
metaphysical possibility when judging whether a ground for doubt is
to qualify as a Cartesian Ground. Let's (loosely) say that a claim is
epistemically possible if it's possible 'for all we know.' And let's
(loosely) say that a claim is metaphysically possible if it's a
coherent claim, i.e. if it's not self-contradictory. I'd argue that a
Cartesian Ground for doubt need only be epistemically possible to
constitute a sufficient ground for doubt. For a posited ground of
doubt for some belief can be both metaphysically impossible and
epistemically possible; that is, we can believe that a ground for
doubt is possible even if it's in fact impossible. And in such a case
we have no choice, given The Empiricist's methodological constraints,
but to judge such a ground for doubt an adequate Cartesian Ground.

Hence, a ground for doubt can be both metaphysically impossible and a
Cartesian Ground, provided one does not know (in the strong sense)
that it's metaphysically impossible. And this brings us back to the
notion of a dreaming spirit. As long as the notion of a dreaming
spirit is not known to be a metaphysical impossibility, it can
constitute a Cartesian Ground for doubt. Therefore, we are not
required to show that it's a coherent concept -- that is, that it's
metaphysically possible -- if we are to make use of it in an argument
against the Basic Response.

The second objection to the Immaterialism Argument that I shall
consider concerns the Vivacity Premise of the Dream Argument. Recall
that a Dreaming Experience must be capable of attaining the degree of
vivacity achieved by our most vivid Waking Experiences if it's to
provide us with Cartesian Grounds for doubting the veridicality of
any given experience. But it's highly implausible that a material
being could have a dream about existing immaterially that's as vivid
as his Waking Experience. For what sorts of experiences would such a
dream comprise? A material being would, it seems, lack the
imaginative resources that the Immaterialism Argument requires of
him. Hence, the Immaterialism Argument cannot refute the Basic

I grant that it's highly implausible to suppose that a material being
would possess the imaginative resources to produce Dreaming
Experiences about an immaterial world that are as vivid as his Waking
Experiences. I further grant that were you to ask me what a Dreaming
Experience about an immaterial world could possibly be like, I would
have no answer. Indeed, if I did have such a dream, I may not be able
to identify it as comprising Dreaming Experiences that are about, or
represent, an immaterial world. But none of these concessions
demonstrate that it's impossible that I should have such Dreaming
Experiences, or that they should be as vivid as my Waking
Experiences. Hence, for all I know, I could have such a dream
tonight, and awake tomorrow exclaiming, 'So that's what it's like!';
but then it follows that it's epistemically possible that I should
have such Dreaming Experiences. And this is all that's needed to get
the Immaterialism Argument off the ground.

I thus conclude that the Dream Argument can refute the Basic Response
via the Immaterialism Argument. For I may be a spirit dreaming that a
material world exists, when none actually does; hence there need not
be any general objects, and thus no basic elements to compose them.
Further, such 'spirit dreams' no more need to be a product of god's
active and willful deception than do our own dreams. Hence, we can
use the supposition that god exists in concert with the Dream
Argument to refute the Basic Response in a way that clearly
distinguishes it from the Deceiver Argument.

Finally, I shall argue that the Dream Argument can handle the
Mathematical Response as well. For The Empiricist believes that all
knowledge, and thus all mathematical truths, are ultimately grounded
in sense experience. But we have shown above that the Dream Argument
can render the existence both of the general objects of experience
and of their basic constituents dubious on Cartesian Grounds. Hence,
the supposition that either one exists must be rejected as if it were
false, in accord with the requirements of Descartes' method. And this
leaves The Empiricist, insofar as he's employing Descartes' method,
with no reliable information about the nature of the external world.
But then it follows that it leaves him, qua (naive) empiricist, with
an utterly groundless mathematics as well. Hence, the Dream Argument
provides The Empiricist with Cartesian Grounds for doubting all
mathematical truths.

One might object that mathematical truths are knowable a priori, and
hence are not dependent on any knowledge whatsoever of the nature of
the world of experience. Therefore, the Dream Argument's ability to
withstand the Basic Response by providing Cartesian Grounds for
doubting the existence of the material world can do nothing to rebut
the Mathematical Response on those grounds alone. The Deceiver
Argument, however, by rendering the reliability of our cognitive
faculties dubious, does have the resources to refute the Mathematical
Response. And if the Dream Argument leaves the Mathematical Response
untouched, it follows (given our stipulation that the Deceiver
Argument can refute the Three Responses) that its skeptical scope is
narrower than that of the Deceiver Argument.

But I'd argue that this objection fundamentally misconceives the
conception of The Empiricist that Descartes presupposes. For recall
from Section One that he is a naive empiricist, and hence does not
consider the truth of any claims, including mathematical claims, to
be completely divorced from or independent of the data of sense
perception. But then it follows that if the Dream Argument can
provide Cartesian Grounds for doubting the existence of the basic
elements that mathematical claims are 'about', it can provide
Cartesian Grounds for doubting the truth of mathematical claims

Indeed, I'd argue that even if The Empiricist were a sophisticated
empiricist, the Dream Argument could provide him with Cartesian
Grounds for doubting a priori truths as well. For I can certainly
conceive of dreaming about a man who is certain that two plus two is
five, and I can even conceive of dreaming that he has reasons for
believing this obvious falsehood to be true. Is it impossible that I
should dream that he explains those reasons to me, and persuades me
that he's right? I can certainly be taken in by fallacious reasoning
while I'm awake, and thus be led to accept as true conclusions that
require me to reject beliefs I had formerly thought to be certain.
Isn't it then conceivable that I should be similarly misled while
dreaming? I have no reason to conclude that it's impossible; but then
it follows that I have Cartesian Grounds for concluding that even
beliefs that I now judge to be knowable a priori are dubious.

I thus conclude that Descartes' Dream Argument has the resources to
refute each of the Three Responses that The Empiricist of the 'First
Meditation' raises against it.


I have argued that, contrary to Descartes' arguments in the 'First
Meditation', the Dream Argument can supply The Empiricist with
Cartesian Grounds for doubting the Three Responses that he adduces to
reject it. Hence, I conclude that Descartes underestimated the power
and skeptical scope of the Dream Argument in his 'First Meditation.'


Descartes, R. The Philosophical Writings of Descartes, vol.2.
translated by John Cottingham, Robert

Stoothoff and Dugald Murdoch, Cambridge: Cambridge University Press,

Frankfurt, H. G. 1970. Demons, Dreamers, and Madmen. Indianapolis:


1. I would like to thank Joanne Lovesey, Kamala Vadlamani, Matthew
Simms and Richard Chappell for their helpful comments and
suggestions, both on an earlier, shorter version of this essay, and
on the current version.

(c) Eric DeJardin 2014

Email: dejardin@cox.net

 Philosophy Pathways is the electronic newsletter for the
 Pathways to Philosophy distance learning program

 To subscribe or cancel your subscription please email your
 request to philosophypathways@fastmail.net

 The views expressed in this newsletter do not necessarily
 reflect those of the Editors.

Pathways to Philosophy

Original Newsletter
Home Page
Pathways Home Page