Philosophy of Mind and The Mind Body Problem
Philosophy of Mind and The Mind Body Problem
An obvious starting place, perhaps, is to roughly define it as the stuff that goes on in
our heads things like mental arithmetic calculations, daydreaming, planning
revenge, or thinking about yesterdays roast dinner. But this doesnt seem quite right.
After all, it seems like there are lots of other phenomena that we would probably want
to class as mental that dont seem to fit this description. A feeling of pain when I
knock my shin, for instance, or my growing feeling of hunger, or a ticklish feeling
between my shoulder blades that I cant reach.
One thing that does seem to be in common to all of these things is that there is
something it is like to experience them, so perhaps this is the way to pick out the
mental things. This is more promising, but we still might not think that this is going
to be enough as our full answer theres nothing its really like to experience the
belief that Paris is the capital of France, but surely beliefs are exactly the kind of thing
we think of as mental.
There is obviously much more to be said about what we should count as the mark of
the mental, and brainstorming this question might be a nice place to start students
thinking about the philosophy of mind. In what follows, however, I wont further
pursue this question any further. I trust that we have a rough-and-ready
understanding of what minds and mental phenomena are that will do for our purposes.
Moving away from this question, the focus of these notes will be one of the central
problems in the philosophy of mind. This is the problem called the mind-body
problem.
Questions in the philosophy of mind generally fall into two categories. The first is a
category of epistemological questions that ask about what we can know when it comes
to the mind. These include questions about how we can know about the mental states
of others, how we can know about our own mental states, and the nature of these
forms of knowledge. The other category consists in metaphysical questions that ask
about the fundamental structure of the mind. The problem that has dominated the
philosophy of mind over the last three centuries or so, and which will provide the
basis of these notes, is a metaphysical one: the mind-body problem.
1
On the face of it, the mind-body problem seems to be a fairly simple one. It amounts
to the challenge to say what the relation is between our minds, or mental states, and
our bodies, or brains. Why should we be moved to consider this problem at all?
The real force of the problem comes from two sets of intuitions, both of which seem
pretty compelling. The first of these is the intuition that the mind and the body are
two very different kinds of things. The body is an extended object in the physical
world it has a size, a weight, a shape, a spatio-temporal location, I can feel it, taste
it, smell and see it. And as a physical object in the world, it is just as well placed as
other worldly physical objects to causally interact with the things around it. My body
gets onto trains, sits on chairs, picks up coffee cups, and shakes hands with other
bodies. Moreover, all of this is done in a way that would seem to be perfectly
respectable to a physicist. Just like other objects in the world, our bodies conform to
the natural and physical laws the laws of gravity, causation, motion and so on. In
short, our bodies just seem to be physical things like any other. And the same, I take
it, can be said of our brains.
The mind, by contrast, does not seem to be like this at all. Our mental states do not
seem to be the kinds of things we could locate in the physical world or encounter
through the senses to talk in this way just seems to get the sort of thing were
dealing with wrong. If we start asking whether our mental states are subject to laws
of gravity, or could get onto trains, or weighs more than a bag of sugar then it just
looks like weve made a mistake. As pointed out above, its not east to say just what
properties we can truly ascribe to our mental states, but it will probably be something
like consciousness, intentionality, subjectivity, or the fact that theres something its
like to be in that state all features that seem to have no place in our purely physical
description of our bodies.
At first sight, then, it looks like our bodies and our minds are very different kinds of
entities. It is, on the other hand, plain to anyone who has ever had both a mind and a
body that they are more deeply and inextricably bound up together than this intuition
of distinctness might at first lead us to believe. My body isnt just a vehicle for my
mental life. Every thought I think, every action I perform, or sensation I experience,
rather, involves a complex and indissoluble engagement of both mind and body; as
the founder of contemporary philosophy of mind, Ren Descartes, famously put it: I
am not only lodged in my body as a pilot in a vessel, but that I am besides so
intimately conjoined, and as it were intermixed with it, that my mind and body
compose a certain unity. If we take the intuitions of the last two paragraphs too
seriously, however, then we are in danger of slipping into just this kind of distinct
picture of the mind-body relation.
What, then, is the right way to understand the metaphysical relation between my mind
and my body that will accommodate both the intuition of distinctness and the thought
that our minds and bodies are deeply interconnected? This is the mind-body problem.
In these notes I will outline four ways you might try to respond to the mind-body
problem.
2
The first, substance dualism, says that the mind and the body really are distinct kinds
of thing. The body is made up of a physical substance, whereas the mind is
something more like a ghostly soul, and as such is made out of a kind of ghostly
mental substance which is non-physical. In the first part of these notes I will outline
substance dualism, and set out some arguments both in support of it and against it.
A third option is a kind of compromise between the first two. Supporters of non-
reductive materialism agree with reductive materialists that there is only physical
substance in the world. They disagree with the reductive materialist, however, in the
claim that the physical stuff is all there is to mental phenomena. Even if our mental
states and events depend on, or are made up out of, our physical states, facts about our
mental lives cant be reduced to facts about our physical make-up. In part three I will
summarise the claims characteristic of non-reductive physicalism, then critically
present two such views anomalous monism and functionalism.
Teaching tips
If youre planning to teach a course on theories of the mind, its crucial to get
the students to see the force of the mind-body problem, and be persuaded by
both sets of intuitions so that they will understand why there is a puzzle here
in the first place.
That the mind (or the soul) is distinct from the is an idea that is commonly
found across different religions and societies drawing on these kinds of
exampels could be one way to urge the first intuition of the mind-body
problem.
With regards to the second intuition, try to press the thought that were not so
alienated from our bodies as to think that they are just receptacles for our
minds surely our minds and our bodies causally interact in all sorts of ways.
Use lots of examples of mind-body causation, and of bodily sensations which
seem to be both mental and bodily.
3
Discussion questions
What do you think are the central defining marks of the mental? Can you
think of any counterexamples to each others suggestions?
In what ways do you think that the mind and the body or brain are different
sorts of thing? In what way do you think that they are intimately related?
Before turning to what various philosophers have said in response to the
mind-body problem, what do you intuitively think is the relation between
your mind and your body, or your brain?
Further resources
4
1. Substance Dualism
Duality of substances
What differentiates the two? Well, a substance is a kind of thing that has, or
instantiates, properties. We could identify the substance of water, for instance, by
pointing to the kinds of properties it has (of being composed of hydrogen and oxygen,
for instance, or of being liquid, or drinkable, or coming out of the taps and so on).
These kinds of properties give us a way to hook onto the differences between
different sorts of substances. If, for instance, we wanted to contrast the water
substance-kind with another say, the substance of gold we could easily do so by
appealing to their different properties. The substance of water has the properties
listed above, while gold substance has properties like being yellowish in colour, being
a chemical element with the symbol au, and so on.
According to the substance dualist, we can identify and contrast physical substance
and mental substance using exactly the same kind of method. The central features of
physical (or material) are things like having spatial extension in the world, being
decomposable, and being divisible (well come back to this property in a moment
when we come to the argument from indivisibility). Others listed by Descartes
include things like spatial location, being an object that can be sensed, passivity,
being law-governed, measurability, and publicity.
Mental substance, by contrast, is a substance that does not have any of these
properties. Its not spatially extended, divisible, decomposable, or have any of the
other physical properties mentioned in the last paragraph. This is because its not
made up of the physical stuff we can see in the world around us. Rather, its a kind of
ghostly or ethereal substance of the kind that things like souls, or minds are made of.
Its properties, for Descartes, are things like being the seat of consciousness and
thought, being private, being active, being indivisible, lacking spatial location or
extension, and being indestructable. (Of course, in terms of this taxonomy, the
example given above was a bit misleading, since both water and gold will count as
physical substances).
Human persons
Us humans are made up of a combination of both of these substances. Lets take our
bodies first. Our bodies, for the substance dualist, are built out of purely physical
substance. We can easily see that this is the case by noticing that it has all the kinds
of properties that Descartes uses as a marker of the physical (its spatially extended,
decomposable etc.) As a physical object, it has no problem in navigating its way
around the physical world. It can pick things up, bump into things, adjust the position
of things, and so on. When it engages in this kind of interaction in the physical world,
according to Descartes, we should think of it as moving around just like a machine
5
would. Its propelled into its various behaviours and exerts its causal influence on
things in the world according to purely physical-mechanistic laws.
This, however, is not all there is to being a human. Indeed, this kind of purely
physical description of our engagement with the world seems to miss out the most
important bit, which is that we also have a kind of rich awareness of whats going on;
we are conscious of the world, and of ourselves in the world, and can think about
whats happening in all sorts of ways. That is, we also have minds. These minds,
weve already seen, are the things that account for our consciousness and thoughts.
This is because according to Descartes, having consciousness and thoughts are the
kinds of properties that, for the substance dualist, are distinctive of mental substance.
Mind-body problem
Like the other theories of mind to be discussed in these notes, substance dualism can
be seen as a response to the mind-body problem. Just as a quick recap, the mind-body
problem was the challenge to explain how our minds relate to our bodies. The force
of this question, we saw, came from two sets of intuitions that seemed to stand in
tension with each other. On the one hand, the mind and the body seem to be two very
different kinds of thing, but on the other hand, they seem to be very intimately related.
Substance dualism responds to the mind-body problem by taking the first set of
intuitions very seriously. The reason why it seems to us as if the mind and the body
are very different sorts of entity is because they are! The first set of intuitions, then,
have been accounted for on this view. There is, however, still a task left for the
substance dualist to do. Anyone who holds this view still has to account for the
second set of intuitions about how closely the mind and the body are intermingled.
As we will see in two of the arguments against substance dualism, this is really the
views weak spot. It turns out to be very difficult to say how the mind and body
interact once you say that they are made up of different substances.
Descartes himself was very aware of this task facing his theory, and unlike some of
the other substance dualists we will see (occasionalists, parallelists and
epiphenomenalists), give a direct answer to it. Unfortunately, his answer was not
altogether convincing. His solution was to say that our minds can causally interact
with our physical bodies in the same way in which we might pull a lever on a machine
to set it in motion. This is not as metaphorical as you might think for Descartes, this
lever really did exist, and was located in the pineal gland, a small endocrine gland
located in the brain just above the cerebellum.
The resultant fusing of mind and body when the lever is pulled, for Descartes, is so
intimate that it gives rise to certain properties that can be had by neither the mind
nor the body alone. Sensations, for instance, such as pains, itches, tickles and
tingles, can be ascribed only where the mind and the body are so joined; they
are, strictly speaking, neither purely mental nor purely physical properties, but
properties of a thing made up of the fusion of mind and body.
6
Any substance dualist account like this, on which the mind is taken to be causally
interactive with the body, is called interactionism.
Historical interest
It should be kept in mind that substance dualism is very much out of favour in
contemporary philosophy of mind (with a very few exceptions - see further reading).
In large part, its decline in popularity came from a science-led 19th century
movement to bring our understanding of the mind under a world-view that appeals
only to entities we would recognise as part of the natural, or physical, world. As a
part of this effort to naturalise the mind, philosophers started looking for a way to
answer the mind-body problem that avoided the substance dualists mysterious
ghosts in the machine (as the philosopher Gilbert Ryle called it). Some of the
results of this search will be outlined in the later sections on materialist or physicalist
accounts of the mind.
Because of this, substance dualism is now normally taught as part of the historical
overview of the mind-body problem, and as a useful alternative with which to contrast
the substance monist physicalist theories that dominate contemporary philosophy of
mind.
Property dualism
A final quick cautionary note before we move on to the arguments for and against this
view: substance dualism, which says that there are two substances in the world,
should not be confused with property, or aspect, dualism, which says that there is
only one kind of substance in the world, but that the substance in question has both
mental and physical properties.
Teaching tips
Its likely that most students will intuitively be quite sympathetic to dualism,
so this is a good place to start in a course on the philosophy of mind.
This also means, however, that the arguments against it might be a bit of a
hard-sell. Its important, however, that the students are persuaded of the
inadequacy of substance dualism if they are going to stay on board with the
search for the right physicalist theory of mind for the rest of the course, so its
worth spending a bit of time on why it cant be right.
As mentioned in the last section, one of the reasons substance dualism is
intuitively attractive might be because it is a picture of the mind thats pretty
pervasive across various difference cultures and times, and fits in with any
view on which we have non-bodily souls or spirits. As suggested in the last
section, starting with these kinds of examples can work as a good way into the
subject.
Discussion questions
Descartes thought that only human animals had minds as well as physical-
mechanistic bodies, and that other animals only had the physical body. How
7
plausible do you think this is? Do you think animals have mental states like
beliefs, intentions, regrets or hopes?
What do you think of the idea that our minds and our bodies are completely
distinct in the way suggested by the substance dualist? Is your body just a
kind of physical vehicle for your mind? Is your mind something separate that
controls that vehicle like a lever on a machine?
What do you think of the idea that our minds are made of something non-
physical? Are you much persuaded by the idea that its made of a kind of
non-physical substance? What would the substance be like?
If you are attracted by the idea that our minds are really separate from our
bodies, can you articulate the reasons why you like the idea? Is there
something special about the mind that this theory would be able to explain?
Further resources
Essential reading for this topic is Descartes sixth meditation from Meditations
on First Philosophy (lots of different editions available, a good one is from the
Cambridge Texts in the History of Philosophy series edited by John
Cottingham.) Although this is a very rich text, it is also eminently
approachable and by no means an unenjoyable read.
For introductory accounts of substance dualism, try Chapter 2 of Jaegwon
Kims Philosophy of Mind, Cambridge MA: Westview Press (2006), Chapter
1 of Stephen Priests Theories of Mind, London: Penguin (1991), and Chapter
2 of Philosophy of Mind (a beginners guide) by Edward Feser, Oxford:
Oneworld Publications (2005) to get you started, but most philosophy of mind
study guides will have a chapter on substance dualism.
There are very few modern day substance dualists. One exception to this is
John Foster, whose defence of Cartesian dualism can be found in The
Immaterial Self: A defence of the Cartesian dualist conception of the mind,
New York: Routledge (1996).
8
1.1 Argument for substance dualism: the indivisibility argument
Leibnizs law
The main philosophical tool that will be needed for the argument is the
indiscernibility of identicals principle or, more familiarly, Leibnizs law (named
after the 18th century German philosopher and mathematician Wilhelm Gottfried
Leibniz, to whom we owe this principle). Leibnizs law is the principle that two
things can be identical only if they share all the same properties. This, really, is
just common sense. If two things are identical, then surely they must share all
the same properties; they are one and the same thing! We might, of course, have
different names for that thing, which probably explains why the question of
identity would have come up in the first place. We might, for instance, ask
whether Clark Kent and Superman are identical. Given that they really are
identical, however, that man must surely have the same properties whichever
name we use for him.
With Leibnizs law in hand, we are ready to give the basic shape of the
indivisibility argument as follows:
I have given the arguments schematic form here to show that the same template
could be used to generate various different arguments for substance dualism
all you need to do is find a property that the body has and that the mind lacks, or
vice versa. In the case of the indivisibility argument, property x stands for the
property of being divisible.
9
(6) Therefore, by Leibnizs law, the mind and the body are not identical
Lets go through each of these steps in turn. What does it mean to say that the
body is divisible? Gruesomely enough, this means exactly what it says: I could, if
I so chose, cut my body into halves or thirds or quarters, or I could separate my
left leg from the rest of my body, or, more generally, I could carve myself up any
way I like.
Less dramatically, the idea here is that the body is a material object, and it is a
feature of material objects that we can divide them into smaller pieces without
compromising the material nature of those body-bits. Indeed, for Descartes, this
kind of divisibility is one of the defining features of things made of material
substance (see section 1 on substance dualism). Cut my material body into two,
and what youll get is just two (somewhat smaller) material objects in the world.
This, really, is all there is to the first premise in the argument.
The second premise, that says that the mind is not divisible, is no trickier than
the first. It just says that the mind is not the kind of thing that we could chop up
into smaller pieces. After all, what would it mean to say that the mind is
divisible? Can I cut my thoughts in two? Or my headache into thirds? It just
doesnt make sense to talk like this when thinking about the mental domain, in
the way that it is perfectly intelligible when thinking about the body.
Descartes describes this indivisibility of the mind in terms of the mind being a
simple substance, by which he merely means that it is not a substance that can be
divided up into smaller proper parts. It is an indivisible whole in itself, which
either exists in its entirety, or doesnt exist at all.
Descartes, then, takes the mind and the body to differ in the kinds of properties
they have at the very least, for all that has been said here, they differ in terms of
their divisibility properties.
It is at this point in the argument that we need to make use of Leibnizs law, the
principle introduced above that if any two things are identical, then they must
share all of the same properties. After all, all that we have shown so far is that
the body has a property that the mind lacks namely, that of divisibility. In
order to move from here to the conclusion that the body is distinct from the
mind, we must think that a difference in properties is enough to establish a
difference in entity. Luckily, this is precisely the move made permissible by
Leibnizs law: given that the body and the mind do not share all of the same
properties, according to that law they cant be the same thing.
10
Teaching tips
As remarked above, this is one of the simplest arguments in support of
substance dualism, so a good one to ease the students into the topic.
Because of its simplicity, its also a good argument to get students to
develop their more general philosophical skill in handling arguments and
for showing them what a syllogistic argument looks like.
You could also use it to introduce them to the notions of validity (if all the
premises are true then the conclusion must be true) and soundness (valid
+ the premises are true).
You could also get them to think about how an argument of this kind of
syllogistic form could be undermined i.e. either by denying its validity,
or by denying its soundness by rejecting one of the premises.
The only slightly tricky part of this argument is Leibnizs law. One way to
help them get a grip on it could be to try to get them to come up with
counterexamples to it, which will hopefully get them to think carefully
about what the principle amounts to.
Discussion questions
Once we have Leibnizs law in hand, it looks like there might be other
arguments we could generate using the same argumentative template.
What kinds of properties might we appeal to? Consciousness? Object-
directness/intentionality? You could have a look back at the properties
that Descartes takes to distinguish mental and physical substance for
inspiration.
It seems like premise 1 is pretty robust, but what about premise 2? Is it
right to say that we cant make sense of dividing up the mind? What
about people with multiple personality disorders? Or what about sleep
episodes dont they cut up our episodes of awareness in some sense?
Are these kinds of divisibility still different enough from the divisibility of
the material body for the argument from divisibility to still go through?
Further resources
11
There is a wealth of material available on disassociative and multiple
personality disorders, but here are a few popular articles and medical
resources that might be helpful as a starting point:
o http://psychology.about.com/od/psychotherapy/f/psychvspsych.
htm
o http://www.guardian.co.uk/lifeandstyle/2011/sep/30/kim-
noble-woman-with-100-personalities
o http://www.psychologytoday.com/conditions/dissociative-
identity-disorder-multiple-personality-disorder
o http://www.mind.org.uk/mental_health_a-
z/8039_dissociative_disorders
o http://www.nami.org/Content/ContentGroups/Helpline1/Dissoci
ative_Identity_Disorder_(formerly_Multiple_Personality_Disorder).
htm
Stanford Encyclopedias online resource is often a good starting place with a
new topic it provides detailed topic overviews that are serious-minded and
somewhat advanced, but that are nevertheless fairly accessible and clearly
written. They are also a great resource for further reading suggestions, as each
topic is accompanied by an extensive topic bibliography. The entry on
Leibnizs law is here: http://plato.stanford.edu/entries/identity-
indiscernible/
12
1.2 Argument for substance dualism: the conceivability argument
The most well-known argument for dualism the so-called conceivability argument
begins with the thought that if I set myself to the task it is fairly easy to imagine
waking up one morning and finding to my great surprise that my body has
disappeared. I can, in my imagined scene, seem to see the room around me and the
crumpled bedclothes beneath me. Everything seems to be in place, but when I look
down, the familiar assembly of arms, knees and toes Im used to seeing when I look
downwards just isnt there. This might, of course, all seem a bit fanciful. What is
important for the argument from conceivability, however, is not that this scene is in
any way plausible or likely to happen, but only that it is imaginable.
This might, at first, not seem terribly impressive. Cant we imagine all sorts of
things? Why should we be interested in what things we can imagine? Well, it turns
out that there are some things that we dont seem to be able to imagine. Try putting
your mind, for instance, to the attempt to imagine a second scene in which you wake
up one morning to find that 2+2 no longer equals 4, or again a third in which you
awake to find that the inside corners of all of the triangles in the world no longer add
up to 180 degrees. No matter how hard you try, imagining these things just seems to
be impossible. Unlike the first scenario, these last two cases dont seem to provide a
coherent enough narrative to be imaginable as a scene; what would it be to imagine
2+2 equaling something other than 4, or to imagine a triangle that didnt have its
internal vertices add up to 180 degrees? We can, of course, imagine ourselves
making mistakes about these matters, but that is a very different thing to imagining a
world in which they are actually true.
The first step in the argument, then, is to establish that unlike these other cases, we
can imagine a mind existing without a body, and so importantly imagine a mind
existing without a brain. How do we get from here to an argument for substance
dualism?
Conceivability-entails-possibility principle
More precisely put, the next move in the argument employs the principle that the
conceivability of a given state of affairs entails its metaphysical possibility, where
metaphysical possibility can just be thought of as possible ways in which the world
might have been if the natural laws of the world had turned out differently. It might
be helpful to contrast this metaphysical possibility with the more limited physical
possibility, which marks the ways things might have turned out given the natural laws
that really do govern our world. It is, for instance, metaphysically but not physically
13
possible that a human being could jump over the channel, or that gravity could only
be operative on alternate days. (It should be noted that this conceivability-entails-
possibility principle is far from uncontroversially accepted in the literature see
further resources.)
Here, then, is state of the argument so far: we have seen that a bodiless (and so
brainless) mind is conceivable, and have found that this conceivability reveals such a
state of affairs to be metaphysically possible. So far, however, this only shows that
this is a way that the world might have been. But we are trying to argue for substance
dualism, which is a position that says that as things really are the mind is not identical
to the brain. How does the metaphysical possibility of a mind existing without a brain
show us that the mind is actually a different thing to the brain?
Perhaps an example will help to illustrate why this is. Dhaulagiri is the Nepali name
for the mountain that we know more familiarly as Mt Everest. Here, then, we have
an identity-statement that we probably have a better handle on than the mind-brain
identity, and which might help to get us going on this tricky part of the argument.
The identity-statement in question is the claim that Dhaulagiri is identical to Mt
Everest. The two are just one and the same thing. When you climb one of the them,
then, you also thereby climb the other, when you photograph one you also photograph
the other. You couldnt do otherwise given that they are the same thing.
The same thing, then, applies to the supposed identity of the mind and the brain. To
say, along with the substance dualists opponent, that they are identical is to say that
they are one and the same thing. But if they really are one and the same thing, then it
makes no sense to say that they could be separated in other metaphysically possible
14
scenarios how could something be separated from itself, even if the natural laws of
the world were different? The suggestion just makes no sense. Their identity, if it
really is an identity, must therefore hold in all metaphysically possible worlds. But
we have already said that the conceivability of a disembodied mind shows that it is
metaphysically possible that a mind could exist independently of a body. The mind
and the brain, therefore, cant really be identical. Substance dualism must be correct.
Zombies
With these zombies in hand, we can construct the argument in exactly the same form
as the original conceivability argument:
The transitions between these moves in the argument parallel the transitions in the
original: the move from (4) to (5) appeals to the conceivability-entails-possibility
principle, and the move from (5) to (6) works by recognising that if its possible for a
seeming identity to come apart, then it cant really be a true identity. If the mind
really is just one and the same thing as the brain, that is to say, then zombies shouldnt
be possible. This is because saying that zombies are metaphysically possible at the
15
same time as saying that the mind is identical to the brain is just the same thing as
saying that it is metaphysically possible for one thing (the mind/brain) to both be
present and not to be present at the same time, but this is clearly impossible, even in
merely metaphysically possible worlds!
Ive given the zombie argument here using the term body throughout, but this is
really just a placeholder for whichever sort of physical state or property a particular
version of physicalism identifies with the mind. As we will see in later sections, for
example, one version of physicalism thinks that mental states are really nothing more
than behavioural dispositions. A zombie argument against this so-called behaviourist
version of substance physicalism, then, can be constructed by replacing mention of a
body with behavioural dispositions, and likewise for other versions of reductive
physicalism.
Teaching tips
There are several tricky steps to this argument. The necessity of true identities,
for instance, is a difficult idea to get a grip on so worth spending some time
on. Using lots of examples will help.
The conceivability-implies-possibility principle is also a little slippery. In
particular, students might struggle to grasp that there is anything that is not
conceivable. Having a few inconceivable examples up your sleeve might help
here (e.g. a square circle, parallel lines that meet, something that is both red all
over and green all over, an object that both exists and doesnt exist at the same
time, etc. anything that violates a logical truth will do.)
You could come back to the zombie argument when you get to reductive
physicalism, and try to get students to construct their own zombie arguments
against specific forms of reductive physicalism.
Discussion questions
Can you think of some other examples of scenarios that are metaphysically
possible but not physically possible?
Can you think of your own example of a true identity? Can you explain to a
partner why it is metaphysically necessary (i.e. cant be a non-identity in any
metaphysically possible world?)?
Do you find the conceivability argument and/or the zombie argument
persuasive? If so, are you convinced of the truth of substance dualism? If not,
can you explain which bit of the argument(s) you think dont work?
Further resources
16
Chalmers The Conscious Mind, Oxford: OUP (1996), or David Stoljars The
Conceivability Argument and Two Conceptions of the Physical, in
Philosophical Perspectives, Volume 15, James E. Tomberlin (ed.), Oxford:
Blackwell, 393413.
The Stanford Encyclopedia entry on zombies and zombie arguments can be
found here: http://plato.stanford.edu/entries/zombies/#3
There is also a chapter on zombies in Tim Cranes introductory book,
Elements of Mind, Oxford: OUP, 2001.
For a rather detailed exploration of the conceivability-entails-possibility
principle, see David Chalmers Does Conceivability Entail Possibility? in T.
Gendler & J. Hawthorne, eds) Conceivability and Possibility (Oxford
University Press, 2002), pp.145-200. The second half is a little dense, but the
opening sections are more accessible and provide a helpful and in-depth
discussion of the principle. An online version of the paper is available on
Chalmers personal website: http://consc.net/papers/conceivability.html
The Gendler and Hawthorne volume just mentioned is a collection papers on
the topic of whether conceivability entails possibility but perhaps worth
mentioning that some of these are a little technical. The book has a clear and
detailed introduction, which can be found here:
http://fds.oup.com/www.oup.co.uk/pdf/0-19-825090-8.pdf
17
1.3 Argument against substance dualism: the interaction problem
The first problem that I will set out for the substance dualist is the interaction
problem, a problem often taken to be fatal for substance dualism. It was, as far as we
know, first put forward by Princess Elizabeth of Bohemia in a letter to Descartes.
(See further resources).
The problem emerges from the idea that the mental and the physical realm are
causally interactive. This is just to say that things that happen in the mind sometimes
cause physical things to happen, and in return that things that happen in the physical
body or physical world sometimes cause mental events to take place.
That this kind of interaction of the mind and the body is just a common sense idea can
be brought out with a pair of examples illustrating this causal interplay in both
directions. In the first direction, the suggestion is that the mind can cause physical
events to happen in the body. This just seems obvious: to give a recent example, a
moment ago my desire to appease my sweet-tooth (a mental state) caused me to walk
to the kitchen and get some chocolate (a physical event involving complex bodily
movements). Likewise, it seems undeniable that causation can also go in the opposite
direction. A brutal toe-stubbing (physical) can naturally be said to cause feelings of
pain and self-pity (mental).
The interaction problem amounts to the challenge to the substance dualist to explain
how such causal interaction could take place between substances as radically different
on the substance dualists account as the mind and the body. How is it possible?
It is deeply mysterious, if the mind really is made up of a non-spatially-extended
substance, how it could exert causal influence over a spatially located object in the
world like ones body. How can a non-located entity causally affect a located one?
Or, equally perplexing, how could an ordinary physical object like a body serve to
cause changes in a non-spatial ghostly substance like the mind? Where would these
causal exchanges take place? It certainly doesnt look anything like our normal
picture of causation on which, for example, I kick a ball and it moves off with
predictable speed and direction. And yet, the thought that there is such causal
interaction between the mind and the body is very difficult to resist.
The substance dualist is left with an explanatory gap which has yet to be filled in
satisfactorily; even Descartes had little to say in response to it. Notice that it is really
this failure on the part of the substance dualist rather than the argument itself that
gives the interaction problem its force. After all, we rarely take the fact that we dont
yet know how to explain something as decisive reason to reject the phenomena that
gave rise to the explanatory question in the first place scientific disciplines would
hardly get very far if we did! It is, then, the fact, that the explanatory gap in this case
is so apparently difficult to fill in that really gives the interaction problem its teeth.
Of course, even if the substance dualist cannot find a way to explain how it is that the
mind and the body could causally interact, there is a second kind of response
18
available. The substance dualist might simply deny the premise that the mind and the
body really do interact in the ways that we suppose. There have been a number of
such substance dualist responses (none of which, it has to be said, have proved to be
very convincing.) I will outline the two best known such responses here, both of
which attempt to solve the interaction problem by denying the intuition of mind-body
interaction.
Occasionalism
Parallelism
A second response that uses the same kind of strategy to answer the interaction
problem is called parallelism. Like the occasionalist, the parallelist holds that the
mental realm and the physical realm are, in fact, causally insulated from one another.
Unlike the occasionalist, however, the parallelist does not think that God intervenes
on each occasion to ensure that the two realms appear to be in line with each other.
According to the parallelist, rather, both realms proceed in their course in absolute
isolation from each other, but nevertheless always end up matching up so that to some
it may seem as if they interact.
There is, of course, something a bit suspicious about this supposed coincidence,
which the strongest form of this view attempts to explain by appeal to a higher power.
Leibnizs pre-established harmony version of parallelism, for instance, holds that the
mental and the physical realms appear to be causally linked only because God has
already arranged that the two will progress in perfect coordination. A helpful way to
think of this, perhaps, is that it is a bit like God pressing play on a silent film and an
independently recorded soundtrack at precisely the same time.
Teaching tips
As noted above, the force of the interaction problem really comes from the
substance dualists failure to come up with a satisfying response to the
explanatory challenge. One way to get students to see the power of this
problem is to get them brainstorming on the part of the substance dualist, each
time trying to show that their suggested answer wont work. This way they
will get to feel the frustration of the substance dualist for themselves, and
hopefully be moved by the force of the challenge.
19
Discussion questions
How convincing do you find the interaction problem? Arent there other
distinct kinds of things that can causally interact despite their difference in
kind? (E.g. what about gasses interacting with solids or liquids? Response
these are nevertheless all part of a spatially extended world, and it is the non-
spatial nature of the mind that generates the real mystery in the case of the
interaction problem.)
This problem could be a platform to start a discussion about what we mean by
causation. It seems like, if this is going to be a problem, we have to think of
causation as the kind of thing that happens when objects push other objects
around the world. But are there other ways of understanding what causation is
that doesnt depend on anything to do with the nature of the objects involved?
E.g. what about a counterfactual story of causation, where A causes B if and
only if B wouldnt have happened if A hadnt happened? Or a probabilistic
story of causation, on which As cause Bs if and only if As raise the probability
of Bs occurring? Do we even really have an intuitive grip on what causation is
beyond constant conjunction (i.e. one thing always seeming to happen after
another thing)?
Even if, like the occasionalist or the parallelist, you deny the causal
interactiveness between the mind and the body, it still looks like you would
have to explain why we have such a profoundly held conviction that the mind
and the body do causally impact on each other. (Such an explanation is called
an error theory, and is required whenever an argument suggests that our
everyday intuitions are wildly off the mark). What could the occasionalists or
the parallelists error theory be? Do you find them convincing?
Further resources
20
1.4 Argument against substance dualism: the argument from the
causal closure of the physical
A second argument often made against substance dualism the argument from the
causal closure of the physical is a close relation of the argument from the last
section insofar as they both pivot on the intuition of causal contact between the
mental and physical domains. Unlike the interaction problem, however, which was
motivated by the intuition of mind-body causation going in both directions, the
argument from the causal closure of the physical is only interested in a single
direction of causality: namely, the apparent causal influence that our mental lives
have over our bodies and the things that we do with them. We have the intuition that
the mind can make an actual difference to what happens in the physical world.
The intuitive compulsion to think that our mental states and events have this kind of
causal influence over our bodies is, we saw in the last section, so strong as to be
almost irresistible. If asked to explain why I just reached my arm towards my coffee
cup, my answer is likely to mention something to do with my state of mind just before
the action occurred my desire for a sip of coffee, for instance, or a fatigued feeling
paired with my belief that coffee-drinking is just the thing to shake me out of it. In
fact, put this way it just looks like an obvious everyday truth that things that happen in
my mental life are the kinds of things that often genuinely influence physiological
changes that take place in my body (a movement of the relevant limb, etc.) The
argument from the causal closure of the physical uses this intuitively compelling
causal claim together with a pair of equally plausible-sounding additional premises to
generate an apparent contradiction in the substance dualists account.
The premises
(1) is just a statement of the intuition about mental causality discussed both in relation
to the interaction problem, and earlier in this section. Enough, I hope, has been said
to motivate this premise. The only thing that I have added here is that our mental
states cause physical events often, or systematically. All I mean by this is that there is
a kind of regularity between the kinds of mental states I have and the kinds of
physical effects they cause (coffee-desire in the right kinds of contexts, for instance,
systematically cause arm-reachings towards coffee cups). When we get to premise
21
(3) it will, I hope, become clear why I have had to add this in in order to get a tension
going between the premises.
Premise (2) is a statement of the thesis of the causal closure of the physical that gives
this argument its name. There are various ways in which we might have chosen to
formulate this thesis. A strong formulation, for instance, would have it that every
physical event has a unique physical cause. The way I have formulated it in the above
argument is a little weaker than this; it says only that for every physical event there is
at least one sufficient physical event that has caused it. This, really, is just the
thought that for any given physical event (the event of my fingertips reaching the
coffee cup, for example) there will be a fully specifiable physical explanation that al.
Reaching for a coffee cup, for instance, can be traced back to neuron activity in the
cerebellum and motor cortex. These neurons fire to the neuromuscular junction
causing contraction in certain muscles and relaxation in the opposing muscles in the
arm and hand.
We could, according to (2), describe all of the physical happenings in the world
(including our bodily movements) in this way, using purely physical language. If that
is right, then for every physical event we are sure to find a physical cause to think
otherwise, the thought is, would be something like believing in telekenesis.
There are also somewhat more theory-laden arguments than this that we could have
appealed to to motivate the second premise. The arguments I have in mind rely on the
fundamental scientific principle of energy conservation which, roughly speaking, says
that the total amount of energy within a closed physical system cant ever go up or
down, even if it can move around the system. If this is right (and it is almost
universally accepted) then any physical event must be wholly describable as an
exchange of energy occurring entirely within the physical system in question. This is
probably more than would be needed to motivate (2) at this level so I wont pursue
this line of thought here, but see further resources below.
Premise (3) says that physical events are not systematically overdetermined, or, to put
it another way, that they do not typically have more than one sufficient cause. Why
should we accept this premise? Surely we can think of examples in which a physical
event does seem to be overdetermined in this way? To take a stock example, imagine
the scene of an execution by firing squad in which each shooter fires a bullet at their
common target at exactly the same time. Suppose that all of the bullets enter the body
at precisely the same instant, and moreover, that each shot would have been fatal by
itself. The targets death, in this case, is overdetermined it is the result of multiple
independently sufficient causes. It doesnt seem problematic in this case to say that
no single shooter caused the victims death, even if each would have done if the
others had not been there. Why, then, cant we think of the causes of our bodily
physical events in the same way? That is, why cant we just say that they are
systematically caused by both the physical causes mentioned in premise 2 and the
mental causes mentioned in premise 1 in just the same way that the targets death was
caused by all of the shooters at once?
There is, as it turns out, a wealth of reasons to reject postulating this kind of
systematic overdetermination. One of these is that if each of these causes really were
sufficient for its effect, then it would be a deeply puzzling coincidence that the two
22
causes nonetheless always happen in tandem. If my desire for coffee is really
sufficient to cause my reaching for the cup, then why doesnt it sometimes do so by
itself? Why, come to that, arent the purely physical causes sometimes efficacious in
the absence of any mental cause?
With all these reasons for accepting it, then, (3)s rejection of overdetermination looks
itself to be suspiciously overdetermined!
So long as we accept premises (1) to (3) it looks like the conclusion must follow that
the mental causes of physical events must be identical to those events physical
causes. This seems to be the only way to reconcile all three premises, each of which
as we have seen seem to be in good standing on their own; if we are in agreement
with premises (2) and (3) that all physical events have at least and at most one cause
which is itself physical, then the only way to hold on to the powerful intuition that
mental events regularly cause physical events is to say that those mental events are
just the same thing as those physical events. The neuronal activity to which I can
trace back the physical causal history of my arm-reaching, for instance, must just be
one and the same thing as the mental event of desiring coffee that featured in my
commonsense explanation of that action. To say this much, however, is to deny
substance dualism.
A response that has sometimes been given to this argument on behalf of the substance
dualist comes from a view called epiphenomenalism, which uses the strategy of
simply denying that mental states and events ever cause any physical effects.
According to the epiphenomalist, while it is correct to say that physical things can
cause mental happenings, those mental events cannot in turn assert causal influence
over anything physical; they are causally inefficacious epiphenomena that are
caused but are not themselves causes of anything. Epiphenomenalists thus neutralize
the argument from the causal closure of the physical by denying premise (1).
This view is, in a way, more palatable than occasionalism and parallelism (from the
last section) in that it denies only one half of the mind-body causality intuition where
occasionalists and parallelists deny it in both directions. Having said that, however,
this restraint on the part of the epiphenomenalist also makes it less satisfying than the
other two responses in another respect. By denying that mind-body causation is
possible in either direction, at least occasionalists and parallelists have in hand a
23
response to both the argument from the causal closure of the physical and the
interaction problem. Epiphenomenalists, by contrast, are still no better off than their
fellow interactionist substance dualists in the face of the causal interaction problem.
In fact, epiphenomenalists seem to face a revived version of that problem; if it is right
that mental occurrences can have physical causes then it is even more mysterious why
they cant in turn causally influence the physical. Moreover, unlike occasionalism
and parallelism, epiphenomenalism itself does not provide us with any explanation of
why it seems so strongly to us as if there are sometimes mental causes of physical
effects.
The commonalities and differences between the arguments make the interaction
problem and the argument from the causal closure of the physical naturally
complementing arguments to teach alongside each other; while they both exploit the
same family of intuitions about the causal interactions between the mental and the
physical realms, they are arguments of very different kinds. The interaction problem
is a kind of brute argument that works by simply confronting the substance dualist
with a seemingly unanswerable explanatory challenge. While the force of the
argument is immediately felt, then, it can nonetheless feel like a less-than-satisfying
victory over the substance dualist. By contrast, although the argument from the causal
closure of the physical relies on a number of premises that require a bit of working
out, the syllogistic form of the argument can make its conclusion feel more
gratifyingly earned than that of the interaction problem. Its conclusion is also a bit
sturdier whoever accepts the premises of this argument is seemingly obliged to
conclude that substance dualism is mistaken.
Teaching tips
This is another nice argument for demonstrating the notions of soundness and
validity, and showing what you would need to do if you wanted to reject the
conclusion of the argument i.e. either argue that its invalid, or argue that its
unsound by denying one of the premises.
This argument has a few steps that rely on things other than pure intuition, so
will be a harder sell to students than the more intuition based arguments (like
the interaction problem). For the moves in the argument that are a bit more
theoretical (e.g. the problems with systematic overdetermination) use lots of
examples to show the students the attraction of these kinds of theoretical
virtues.
Discussion questions
24
Another way of pushing back against this argument is to refuse to accept
premise (3)s rejection of systematic overdetermination. Although I gave a
number of arguments above, none were decisive they were based, rather on
considerations of theoretical virtues (e.g. Ockhams razor), or on inferences to
the best explanation. How attractive a response would it be to say that
physical events are systematically overdetermined?
How might a parallelist or an occasionalist respond to the argument from the
causal closure of the physical?
If epiphenomenalism was correct, would there be a problem in even stating it
as a theory? (I.e., wouldnt the fact of our talking about mental processes
count as a physical effect of mental causes?).
By saying that mental states and events are causally inert, the
epiphenomenalist denies both that they cause physical events but also that they
cause other mental events. How plausible is this claim? Can we make sense
of a view on which one thought cant cause another?
Further resources
25
1.5 Argument against substance dualism: the problem of other
minds
So far the problems weve looked at for substance dualism are metaphysical;
they work by revealing that if substance dualism turned out to be correct then
certain problems would arise in accounting for the way we take the world to be,
or for our understanding of the structure of reality. In this section well look at
an epistemological problem for substance dualism, or problems to do with what
we can know if the substance dualist is right. After all, even if the mind-body
problem is itself a metaphysical question, if a particular response to it has
unacceptable epistemological implications, then thats still going to count as a
point against that view.
More specifically, well look at the epistemological problem of other minds for
substance dualism.
A problem family
First, a quick cautionary note. Although it is often called the problem of other
minds, there are in fact a handful of related problems which are far from unique
to substance dualism. Even if mental states are not distinct from physical states,
for instance, there is still a conceptual question about how we could possibly
develop mental concepts that can be applied both first- and third-personally
given that we seem to learn about them and how to apply them from our own
case. Likewise, a non-dualist will still have to explain how it is that we come to
know about others mental states, given that they do not seem to be the kinds of
things that we can directly perceive with our senses. Moreover, even a non-
dualist has to give us an account of how we should deal with the possibility of
actors who are only pretending to have the mental states that they present
themselves as having. Dont these pretenders give us reason to be suspicious
about the claim that we can ever know what mental state someone else is in?
I mention these problems to put them to one side in this section I will consider
a problem of other minds that is specific to the substance dualist; roughly, the
problem that if the mind and the body are really distinct then given that I only
ever encounter bodies in the world I have no reason to think that there are any
minds in existence other than my own.
In this case, the reductio would run as follows. If substance dualism is correct,
then the mind is composed of a non-spatially extended, non-physical substance
(so, non-physically detectable substance), which could conceivably exist
26
independently of a body. This means that if substance dualism is correct, then
there is no way of knowing about the existence of mental states other than ones
own after all, if the mind and the body could conceivably be separated then
nothing about our interactions with others bodies is going to count as a reliable
guide to the existence or non-existence of minds or of specific mental states.
From this point in the argument, there are actually not one but two potential
absurdities to be drawn. The first is that if substance dualism is correct, then we
have no way of knowing that there are minds other than ones own. The second is
that if substance dualism is correct, then we dont have any way of knowing which
mental states anyone else is undergoing.
But surely these are both obviously false! Surely we do know that we are not
alone in this world, or that other minded people exist; to deny this is to accept
that for all I know the position of solipsism, which says that I alone exist in the
universe, is correct. And likewise, surely we do sometimes know what kind of
mental state someone else is in. When I see you writhing around on the floor
clutching a gushing wound in your side, I dont have to ask before concluding
that you are in pain.
Teaching tips
Discussion questions
27
in? Would this also work in response to the second way of putting the
problem, i.e. that we have no way of knowing whether someone is
minded?
Why do we call this an epistemological problem rather than a
metaphysical one? (Note that all of the premises and the conclusion are
formulated in terms of what we can know, and not what is the case. For
instance, one of the results is not that solipsism is correct, but that for all I
know solipsism is correct, etc.)
What is so wrong with solipsism? Or the idea that we dont really know
which mental states others are undergoing? After all, the only thing going
for those premises are really our intuitions about those claims, and surely
mere intuitions are the first thing to be given up in the face of
philosophical reasoning?
o Is this line of thinking encouraged by the conceivability of
scenarios like Descartes evil demon, or the matrix? Dont they
show that I really might be the only thing that exists?
On the other hand, isnt the thought that we know were in a world with
other minded people in which we sometimes know what theyre thinking
and feeling pretty fundamental to our world view? Surely, then, the
burden of proof lies with the substance dualist to show us that we have
good reason to question it?
Further resources
The problem discussed here is far from the most common out of the
family of other-minds problems mentioned above, so be wary of
indiscriminate searching for material on the problem of other minds
discussed here. Stanford encyclopedia has a good entry on this topic, that
should help with disambiguation between the various related problems
http://plato.stanford.edu/entries/other-minds/
28
2. Reductive materialism
Broadly speaking, materialism is the view that everything that exists is fundamentally
material in nature. The label materialism is often used interchangeably with the
more modern-sounding physicalism, and I will do the same here. The difference
between them is that physicalism refines the central claim of materialism (that
everything is fundamentally made of matter) to the claim that everything is
fundamentally physical. This difference, however, will not matter for our purposes.
You can be a materialist about lots of different things. You might, for instance be a
materialist about biology if you think that all biological entities in the world are
ultimately material in nature, or that as long as you fixed all of the physical facts
about the world, the biological facts would be fixed too.
Materialists in the philosophy of mind are not very surprisingly! materialists about
the mind. Their claim is that all mental phenomena are fundamentally material or
physical in nature. From now on Ill use materialism and physicalism just to mean
materialism or physicalism about the mind.
Supervenience
This definition, however, is not going to be very helpful until we get clearer on what
it means to say that all mental phenomena are fundamentally material. There are a few
different things it could mean, some weaker and some stronger. One of the stronger
things it could mean is that all mental phenomena can be reduced to physical ones
Ill come to that in a moment. However, even those that go in for this full-blooded, or
reductive, materialism also accept a related weaker claim. This weaker claim says
that mental phenomena essentially depend on, or supervene on, physical ones. Before
explaining what a reduction of the mental to the physical would be, then, Ill first
explain what it means for the mental to supervene on the physical. (Well see later
that it will be useful to already have the notion of supervenience in hand when we
come to non-reductive physicalist theories).
An example would probably be helpful here. Ill use the example of the property of
softness. Suppose that I have an indescribably soft marshmallow in my hand. Its
clear that in some sense the softness of the marshmallow depends on its physical
make-up it is, that is to say, surely facts about the way its physically constituted in
virtue of which it is so soft. With the philosophical tool of supervenience in hand,
however, we can say something more precise than this. Namely, we can say that the
29
marshmallow couldnt differ in how soft it was unless it also differed physically from
the one I have in my hand now. In other words: as long as any two marshmallows are
exactly the same in every one of their physical features, then will have to also be
exactly the same in their level of softness. As noted above, however, this dependence
only goes one way. Theres no reason why we couldnt swap sugar for artificial
sweetener in one of the marshmallows, but still get precisely the same level of
softness as in the original sugar-marshmallow. What this shows is that softness is a
property that supervenes on physical properties, but those physical properties (e.g. the
crystalline structure of the marshmallows make up) dont in turn supervene on the
softness ones.
Materialists of all stripes are committed to the claim that mental properties supervene
on physical properties. There can be no difference in what mental properties I have
(like being in pain, or remembering the beach at Hunstanton, etc.) without a
difference in my physical properties (the physical composition of my brain, for
instance). An intriguing upshot of this, and one which might provide the basis for a
class exercise, is that if I had a twin who was exactly physically alike to me in every
respect, then if physicalism is correct she would also be exactly like me in all of her
mental properties.
Reductive materialists, like all materialists, accept the supervenience of the mental on
the physical. But they also go in for something stronger than supervenience. They
claim that mental properties can be reduced to physical properties. What does it mean
for one thing to be reducible to another? Roughly speaking, to reduce one domain to
another is to say that the first domain is really nothing over and above the second.
However, there are still at least two things we might mean when we say this.
The first thing we might mean is called explanatory reduction. This holds that
everything in the first domain can be explained exhaustively in terms of the second.
This kind of thing is also sometimes call analytic reduction, in the sense that I can
give a complete analysis of all the phenomena in the first domain in terms of the
second.
Lets have an example. Take vision. I might think that everything about vision can
be explained or analysed using purely physical terms, such as those relating to the
wavelengths of visible light, for instance, or to the anatomy of the optic system, and
so on. To say that this kind of explanation would be exhaustive is too say that vision
is explanatorily reducible to the physical domain. It doesnt deny that there is such a
thing as vision, its just that an explanation or analysis of that phenomenon given in
purely physical terms wouldnt leave anything out we wouldnt be missing anything
if we were to reduce all talk of vision to talk of light waves etc.
The second kind of reduction we might mean is called ontological reduction. When
we say that one kind of thing is ontologically reducible to a second, what we are
saying is that the second kind of thing is the only kind of thing that actually exists
when we really get down to the bottom of things. The first kind of thing, if it is
ontologically reducible to a second, is actually just composed of the (truly existing)
second kind.
30
I might, for example, think that tables are ontologically reducible to their atomic
structures. Although it looks like there are such things as tables in the world, the only
things that really exist are molecules, and the thing that I thought was a table is really
just a table-shaped molecular arrangement. We can still talk about tables, since doing
so seems to be helpful for our conversational purposes sometimes it would be rather
a nuisance if our philosophical metaphysical theories forced us to only ever talk about
things in terms of their atomic structure! Its just that strictly speaking, what were
really talking about when we have a conversation about tables are collections of
atoms arranged in a table-shaped way. Tables themselves dont, strictly speaking,
exist.
There are actually a few more ways of understanding reduction than these two, but
these are the only ones that will be needed for our purposes. (See further resources).
Notice that saying that something is supervenient on another is saying something very
weak. All thats needed to make a claim like this true is a systematic dependence of
one set of properties on another. To claim something is reducible to another, by
contrast, is to make a much stronger claim.
The stronger claim actually brings with it the weaker claim for free. Take the last
example I gave. If the table is ontologically reducible to its molecular components,
then its table-properties must also supervene on its molecular properties. This is
because if tables are really nothing over and above the molecules that make them up,
then there can be no difference to whether or not we see something as a table without
there also being a difference in how the molecules are arranged. Likewise, so long as
there are two exactly similar arrangements of molecules, those molecule-groupings
will stand and fall together with respect to whether or not they count as being
arranged table-wise or not.
Reduction claims, then, always entail supervenience claims. Given their weakness,
however, its not really surprising that supervenience claims dont likewise entail
reduction claims. Suppose, for instance, that I think that dancing-properties
supervene on physical ones there can be no difference as to whether and how
someone is performing a dance without a corresponding difference to the physical
properties of the dancers bodily movements. I might assent to this, but still resist the
stronger claim that dancing-properties are reducible to physical ones. I might still
think that a complete physical analysis of whats happening will still leave something
out, either explanatorily or ontologically speaking.
Reductive materialist accounts of the mind come in both flavours of reduction they
sometimes say that the mind is explanatorily reducible to the physical, and sometimes
that the reduction is ontological. In what follows I will present two different
reductive accounts: logical behaviourism and identity theories. These are both
reductive accounts of the strongest kind they are ontologically reductive accounts.
31
Other than whether they are explanatorily or ontologically reductive, the main
difference between different reductive physicalist accounts of the mind lies in the
choice what kind of physical property the mind is taken to be reducible to. For logical
behaviourists, for instance, mental properties are reducible to physical behavioural
properties, whereas for identity theorists they are reducible to physical properties of
the brain.
Teaching tips
A fun way to test the students intuitions about whether or not they are
materialists is to use the question proposed by Saul Kripke: If God came
along and fixed all of the physical facts about the world, would he have more
work to do in order for there to be mental properties in the world? Those who
dont think theres any more work to do are naturally inclined reductive
materialists.
Its good to have a few examples of properties that might be taken to
supervene on the physical to get the students to really grasp this relation. A
couple of suggestions for supervenient properties might be biological,
chemical, cultural, aesthetic, tactile, communicative, cartographical,
representational (almost anything you can think of that isnt purely physical!).
You could try to get the students to think of a few themselves.
To get a better hold of what the mind-body supervenience claim amounts to,
you could use clones/teletransporter/doppelganger cases etc. to cook up
scenarios where there is someone exactly physically identical to you in every
way. If the mental supervenes on the physical then this entails that your clone
will also be mentally identical to you.
Discussion questions
Can you think of any other properties not already mentioned that supervene on
physical properties?
Can you think of examples where one thing is explanatorily reducible to
another? What about where something is ontologically reducible?
What would a reductive physicalist account of dancing say? (Where its
explanatorily reductive? And what about where its ontologically reductive?)
Further reading
32
o Oxford Reference Online
http://www.oxfordreference.com/view/10.1093/acref/9780199541430.
001.0001/acref-9780199541430
o Routledge Encyclopedia (requires login)
http://www.rep.routledge.com/LOGIN?sessionid=cabfbb1fa320dc49d
30b91cf84d4517a&authstatuscode=400
o Internet Encyclopedia of Philosophy
http://www.iep.utm.edu
Heres a link to a website that lists some different kinds of reduction in
philosophy, that go beyond the two discussed here:
http://www.liv.ac.uk/~pcknox/teaching/phil/sred.htm
For a detailed chapter on ontological and explanatory reduction, see Chapter
15 of Tim Cranes Elements of Mind, Oxford: OUP, 2001. The final chapter
of Jaegwon Kims Philosophy of Mind, Cambridge MA: Westview Press
(2006) also covers a number of issues connected to questions about reduction
in philosophy of mind.
Almost all philosophy of mind introductory textbooks will have a section on
reductive physicalist accounts. See, for instance, the section on reductive
physicalism in William Jaworskis Philosophy of Mind, Oxford: OUP (2011).
33
2.1 Logical Behaviourism
The first reductive account I will discuss is called logical behaviourism (also
sometimes known as philosophical behaviourism, to distinguish it from
methodological behaviourism in psychology). According to logical behaviourism
the physical properties to which mental properties are properly reducible to are
behavioural ones. On this view, to ascribe someone a mental state is to ascribe to
them a set of behavioural dispositions. To say that someone is in pain, for
instance, is just to say that they are currently disposed to display certain
characteristic pain behaviours writhing around on the floor, say, or clutching
the injured body part and saying ouch. To say that someone thinks its going to
be sunny today is to say that that are disposed to wear a t-shirt and pick up their
sunglasses on their way out of the door. To say that someone believes the train
leaves at 11am is to say that they are disposed to plan their morning accordingly.
More generally speaking, we can say that to be in a particular mental state is to
be disposed to display a given range of associated behaviours in response to a
given range of stimuli.
Saying that these dispositions are associated with these states of mind is not, of
course, to say anything particularly revelatory. The crucial point about logical
behaviourism is that it says something much stronger than this its not just that
if you are in pain then you will be disposed to writhe etc.; according to the
behaviourist your pain is your disposition to exhibit those behaviours. Using the
locution introduced in the last section, we can say that your pain-state is
ontologically reducible to your behavioural dispositions. Metaphysically speaking,
it is really nothing over and above them. Put this way, behaviourism looks like a
pretty striking position, and it might not come as much of a surprise that it is not
a particularly favoured one in contemporary philosophy of mind. Saying that, it
enjoyed a brief burst of popularity around the middle of the last century, so it
must have something going for it. What considerations, then, could there be in
its favour?
Verificationism
34
A historical explanation for the emergence of behaviourism as a theory of mind is
that it provides a natural companion to this theory of meaning that was so
popular at the time. After all, if the meaning of a sentence is given by its mode of
verification, then what should we make of sentences attributing mental states to
others? What mode of verification can we apply to those sentences? Not
introspection that is only available first personally, so cant be the way to verify
our mental state ascriptions to other people. Well, it seems like we normally rely
on behaviour to tell us about others states of mind. According to verificationism,
to take this seriously as the relevant mode of verification is to say that the
meaning of mental-state attributions is given by statements about behaviour, or
behavioural dispositions.
Explanatory power
There is another argument that might still be taken to work in its favour (though
admittedly an argument so lightweight that it has little hope of standing up to the
battery of arguments we are about to see against the view). This is a
consideration of explanatory power: the behaviourist, we might think, can
explain things that are left mysterious on some other theories of mind namely,
the connection between the mind and behaviour.
It is, I think, pretty intuitive that there is a deep and systematic relation between
mental states and their associated behavioural profile, or their characteristic
dispositional inputs and outputs. There seems to be something significant and
non-arbitrary, for instance, about the link between the state of fear and
threatening kinds of stimuli on the one hand, and
trembling/sweating/fleeing/etc. behaviours on the other. The behaviourist can
easily accommodate this intuition. The reason that we think that there are
important connections between mental states and their behavioural
dispositional profiles is because those states are nothing over and above those
profiles. At the very least, then, we can say that the behaviourist wields some
attractive explanatory leverage.
Qualia
The main problem with behaviourism is not difficult to spot: it misses out what it
feels like to be in pain. To say that the mind is reducible to behavioural
dispositions leaves out what we surely think is one of the most central and
ineliminable features of the mental realm namely, that having a mind, or being
in certain mental states, comes with a certain experiential quality; in the famous
phrase from Thomas Nagel, there is something it is like to be minded (see further
35
resources). Such experiential properties of the mind the what-it-is-like-ness
of a given mental state are sometimes called qualia (a single one of these
properties is called a quale).
The first objection to behaviourism, then, is that it cannot account for the
existence of qualia. It cant explain why being in a state of pain feels like
something. You might say that just as above there was an argument for
behaviourism on the grounds that it has some kind of explanatory power, this is
an argument against behaviourism on the grounds that it lacks explanatory
power with respect to the existence of qualia.
In an effort to avoid revealing her birth defect, she perfects a fear act in which
she accurately mimics these fear-behaviours whenever she recognises that a
fear-stimulus is present. The act, let us suppose, is faultless. It is true to say of
her, then, that she has a full ordinary set of fear dispositions, which are activated
in typical fear situations. Surely, however, we would not want to say that she
ever really undergoes the state of fear. The whole point of the story is that she is
only pretending. To say that someone has the relevant dispositions, then, is not
sufficient for saying that they have the mental state in question.
36
Mental causation
Teaching tips
Discussion questions
Do you think theres anything right about the behaviourist account of the
mind? How would you defend the theory if you had to stick up for it on
the behaviourists behalf?
Are there some mental phenomena that are more conducive to a
behaviourist analysis than others?
37
Taking a mental state other than pain, could you construct a thought
experiment that shows that having the relevant behavioural dispositions
are neither necessary nor sufficient for having the mental state?
Which of the arguments against behaviourism do you find most
persuasive and why?
Further reading
38
2.2 Identity theories
While logical behaviourism reached its peak of popularity in the first half of the
last century, identity theory (sometimes also known as central state materialism)
was introduced by a small group of Australian philosophers at the university of
Adelaide in the late 40s and 50s, and expanded to international popularity
among the philosophical community during the 50s and 60s.
In a slogan the identity theory is the view that mental states are brain states.
Type/token distinction
Before going any further, it will first be important to get clear on a certain
distinction between what we call types and what we call tokens of a given thing.
Consider the following inscription:
glee
How many letters are there here? Well, it depends on how youre counting them.
One correct answer would be that there are four letters here a g, an l, an e
and another e. If this is the answer youre after, then it looks like the question
you were asking is how many actual letter items there are here. In terms of the
type/token distinction, we would say that youre counting the letter tokens.
This, however, is not the only way you could answer the question. You could also
say that there are only three letters here g, l and e. If this is your answer,
then you are probably interested in how many different kinds of letters there are.
We would say that you are counting letter types.
The same kind of distinction can be made when it comes to talking about mental
states. Take, for instance, the mental state of thinking that the moon is made of
cheese. How many mental states are there here? Well, again, it depends on what
youre counting. If youre counting mental state types, then there is just one: the
state of thinking that the moon is made of cheese. On the other hand, you might
be asking something quite different you might be interested in how many
actual events there are, or have been in the history of the universe, in which the
thought that the moon is made of cheese was actually had. To get an answer to
this question would require some pretty extreme research methods! In principle,
however, I hope the idea is clear. We might be counting thought-types (of which
there is only one), or we might be counting thought-tokens (which may well run
into the billions).
39
Likewise for brain states. Consider a brain state (lets call it B for short) that
involves a specified pattern of neurons firing in the right parietal lobe.
Supposing that B sometimes does actually take place in normal human beings,
how many brain states should we say there are? Again, your answer to this
question is going to depend on what youre really asking. If youre asking how
many types of B-states there are, then there is just one. If youre asking how
many tokens of B-states there are in the world at any given time, then there may
well be many.
Using this distinction, we are now in a position to say that the original identity
theory of mind, as developed by J.J. Smart, Place and others in Australia in the
1940s and 50s, was a type identity theory. Take any mental state type you please
thinking of ones grandmother, experiencing tingling in ones fingertips,
listening to an earworm of Beethovens fifth, there is literally an infinite number
to choose from. According to the type identity theorist, that mental state type is
identical to a particular brain state type. Every time anyone thinks of their
grandma, they are activating exactly the same kind of brain state.
The strength of this claim needs underlining. Type identity theorists are not just
saying that whenever anyone thinks of their grandma, this mental state causes
exactly the same kind of brain state to fire up every time, or that the activation of
that sort of brain state will systematically correspond to grandma type thoughts.
The claim is that the grandma-thought-type is the brain-state-type. This is why
they are called identity theories; on their view the mental state type is identical
to the brain state type.
Scientific hypothesis
The type identity theory is not supposed to fall out of our intuitions about what
the mind is in the same way that, for example, we might think that substance
dualism does. When it was proposed, it was not done so in the spirit of common
sense. It was, rather, put forward as a substantive scientific hypothesis, or as an
empirical discovery about the world. To see what I mean, think about the first
time that anyone proposed that water is H2O, or that heat is molecular energy.
These are type identity claims, just like the type identity theorists claim about
mental states and brain states. The first, for instance, says that a certain type of
natural stuff in the world water is type-identical to a certain kind of chemical
composition hydrogen and oxygen molecules, bonded together in a specific
way. This suggestion must have come as something of a surprise to most people
when scientists first proposed it, but it didnt really matter that the claim was not
obvious from our intuitions about water. It wasnt supposed to be intuitive. It
was a scientific discovery.
According to the type identity theorist about the mind we should think about the
mind-brain identity in the same kind of way. Given the seeming untenability of
substance dualism (that emerges from the problems with that view that weve
already seen), the type identity theorist thinks that we should be open to the
40
scientific discovery about the mind and the brain: just like water turned out to be
identical to H20, our mental states turn out to be identical to our brain states.
One important upshot of this status of the type identity theory as a scientific
hypothesis is that it means that a certain line of attack against the theory cant
even get off the ground. The line of attack I have in mind says that the mind cant
be identical with the brain, because otherwise we would have noticed this fact.
Or, put another way, if mental states are really identical with brain states, how
come we can easily pick out different mental states, and yet be completely
ignorant as to which brain states they are? I hope it is clear that these kinds of
objection are non-starters. Given that the mind-brain identity is a scientific
discovery just like the water-H20 identity it is no surprise that a bit of
scientific investigation is required before we can work out which mental state
type is identical to which brain state type.
Even if this isnt a very good kind of objection to type identity theory, however,
there are a number of much better ones available. Here are two.
Leibnizs law
We can use Leibnizs law to generate problems for type identity theories; all we
have to do is come up with properties that are not shared between the mind and
the brain. There are two ways of doing this: we can think of properties that the
mind has but the brain lacks, or we can look for properties of the brain that are
not had by the mind. At least on the face of it, there seem to be plenty examples
of both (but see discussion questions for possible responses). This, really, is very
similar to the arguments seen earlier from Descartes for substance dualism.
As an example of the first kind of property, take a pain that Im feeling in my left
foot. That pain has the property of being located in my foot. Clearly, however,
this isnt a property shared with my foot-pain brain state which, presumably, is
being instantiated in my brain. We can also go back to those qualia again (the
experiential properties of the mind). There is something it is like to have a foot-
pain, but as far as I know, theres nothing in particular its like to be in brain state
B.
Going in the other direction, there are also properties that the brain seems to
have, but that the mind lacks. Brain states, for instance, have certain frequencies,
but would we really want to say that this is also true of my mental foot-pain?
41
The type identity theorist may well have ways of defending herself against these
kinds of apparent problems, and they are not normally taken to be decisive
against identity theories. At the very least, however, there is a challenge here
that calls for a response.
Multiple realizability
By contrast, the problem of multiple realizability is often taken to be fatal for the
type identity theory (first put forward by Hillary Putnam, see further resources).
One way of putting this objection is to say that its really rather chauvinistic of us
humans to assume that the mental state of, say, pain is type-identical with brain
state B, where brain state B is the kind of brain state that we humans are in
whenever we are in pain. Surely, after all, there are lots of animals sensitive to
pain whose brains dont look anything like ours. Would we really want to use
this as a reason to deny that they feel pain? Take, for instance, the squid. Squids
brains are no doubt very different to ours made fom different stuff, differently
structured, etc. They couldnt, then, ever be in brain state B; they just dont have
the right kinds of brains for it. But if type identity theory is correct, then brain
state B just is the mental state of pain. If squids can never be in brain state B,
then the type identity theorist will have to say that they can never be in pain.
Poor squid! Given that we dont really want to say this, then the fact that it comes
out as a result of the type identity theory means that we must have gone wrong
somewhere along the way.
This objection packs an even more forceful punch once we leave the boundaries
of the actual world and look at how things could have been in other possible
worlds (which, remember, is a way that we have of testing metaphysical
possibilities). Imagine that theres a world whose inhabitants are really similar
to us. So similar, in fact, that you wouldnt be able to tell them and us apart from
the outside they look like us, they have full and tiring lives like us, they fall in
love, have favourite breakfast places and so on. Open up one of their skulls,
however, and the difference is easy to spot. Where our brains are normally
located, these creatures have a sophisticated arrangement of silicone chips. Like
the squid, this means that these creatures could never be in brain state B (indeed,
they could never be in any brain state they have no brains!). But if brain state B
is identical to the mental state of pain, then we will have to say that they can
never be in pain either. But surely its even less plausible to say this in the case
of these creatures, who look and act just like us, than it was with the squid.
42
A solution?: Token identity theory
There is a way of refining the identity theory of mind to make it compatible with
this multiple realizability of mental states that has been taken by some
philosophers intent on salvaging something from the theory. Rather than
looking at types of mental state, the token identity theorist is interested in mental
state tokens. Take a specific mental state that you might be currently undergoing
say, a feeling of strong elation. When you undergo that very feeling, you are in
a token mental state that can never be repeated. This is so even if you feel the
same kind of elation every morning when you wake up in that case, it would be
true to say that you have recurring mental states of the same type every morning
(namely, the elation-type). No two instances of that type, however, can count as
the same token; we would just say that they are both tokens of the same type.
Likewise for brain states. The token identity theorist unlike the type identity
theorist has no special use for types of brain state. The token identity theorist,
rather, is interested in actual neural events going on on a given occasion. They
are, that is to say, only interested in brain state tokens.
The revised theory says that any given mental state token (e.g. my actual feeling
of elation at 19:32 on 23rd June 2013) will be identical with a token brain state
(e.g. firing of synapses in position. in the brain at frequency at 19:32 on 23 rd
June 2013.) Mental states are still brain states on this version of the theory, but
the identities can only ever be worked out on a case-by-case basis. There is no
expectation that all tokens of a given type of mental state will be realized in a
brain state of a common type. The advantages of identity theory, then, (namely,
its strongly scientific and anti-dualist spirit) are preserved, but the multiple
realizability objection is avoided.
Teaching tips
You could use some of our contemporary ways of talking to introduce this
kind of theory: Shes really brainy, shes got a good head on her
shoulders, wheres your head at?, I wish I had her brains, I cant get my
head around it etc.
Be sure to use lots of examples to explain the type/token distinction, and
maybe try to get them to come up with a few of their own. Its easy to
confuse the type/token distinction with (the different)
universal/particular distinction, which is the distinction between a
general property like redness or beauty and specific instantiations of it.
Like with all of the reductive theories, it is important to be clear about just
how strong the claim is. In the case of identity theories, the claim isnt
just that mental states (types or tokens) are systematically correlated
with brain states (types/tokens). The mental states ARE the brain states.
It is easy to grant these theories unwarranted plausibility by unwittingly
slipping back to the weaker claim.
43
If students do find it completely implausible, however, then this shouldnt
be used as a point against it; just remind them that the identity theorist
can appeal to fact that this is a scientific hypothesis, not an intuitive claim.
Discussion questions
Further reading
A good place to start on this topic is Tim Cranes Elements of Mind, Oxford:
OUP, 2001, which has a chapter both on identity theory and on the problem of
multiple realizability.
Almost all other introductory books on the philosophy of mind will
likewise have a chapter on identity theory see, for instance, Chapter 3 of
Jaegwon Kims Philosophy of Mind, Cambridge MA: Westview Press (2006),
or Chapter 6 of John Heils Philosophy of Mind, London: Routledge (2004).
Hillary Putnam presents the multiple realizability objection in
Psychological Predicates in W.H. Capitan and D.D. Merrill (eds.), Art, Mind,
and Religion, Pittsburgh: University of Pittsburgh Press, 3748. A
(questionably scanned) copy of it available here:
http://www.phil.uu.nl/~joel/3027/3027PutnamPsychPredicates.pdf). Theres
also a good entry on it in the Stanford encyclopedia here:
http://plato.stanford.edu/entries/multiple-realizability/.
Another useful place to look at the multiple realizabilty objection is Jaegwon
Kims Multiple realizability and the metaphysics of reduction in Philosophy
and Phenomenological Research , Vol. 52, No. 1 (Mar., 1992), pp. 1-26
For a classic formulation of type identity theory, see J.J.C. Smarts Sensations
and Brain Processes in Philosophical Review, 68, pp.141-156. Pdf version
here:
44
http://philosophyfaculty.ucsd.edu/faculty/rarneson/Courses/SMARTJACKphil
1.pdf
45
3. Non-reductive materialism
With these notions already in hand its not difficult to spell out what non-
reductive materialism amounts to. Non-reductive materialists agree with
reductive materialists about the supervenience claim, but deny the stronger
reducibility claim that reductive materialists accept. For the non-reductive
materialist, mental states are not reducible to any kind of physical state, they
merely supervene on them. And just to remind ourselves: to say that one thing
supervenes on another is to say that there can be no change to what properties
there are of the first kind without there also being a difference to what
properties of the second kind the object has.
Non-reductive accounts of the mind think that mental properties are like this too.
Suppose that neuroscientists were able to provide a fully comprehensive report
about the physical properties of my brain. According to the non-reductive
physicalist, reading this report to me wouldnt amount to telling me about my
mental properties. Even though what mental properties I have might depend on
the kinds of physical facts about my brain that are listed in that report, they
arent reducible to those facts.
46
By admitting that there really is something special about the mental realm that
cant be reduced to the physical, non-reductive accounts of the mind come across
as far more palatable than hard-nosed reductive accounts. By nevertheless
insisting that the mental still supervenes on the physical, however, they avoid
the difficulties of substance dualism. Non-reductionism, then, gives us an
attractive way to hold on to the thought that mental properties are instantiated
by, or are causally dependent on, physical states in some way while stopping
short of claiming that the two are one and the same thing. These days, most
philosophers would probably count themselves as non-reductive physcialists.
The simplest way to have a non-reductive account of the mind is just to state that
mental properties supervene on physical ones (or, more precisely, to state that
they merely so supervene, given that identity also counts as supervenience and
supervenience accounts are designed to stop short of identity claims).
To say this much, however, is not really to say very much at all. There are, after
all, lots of different ways that two sets of properties could be related that would
count as one of them supervening on the other. The first set might just be
causally dependent on the first, or it might be realized or instantiated by it, or
constituted by it, or even identical with it (though, to be fair to the supervenience
theorist, I did just say that identity was explicitly ruled out). It also isnt to say
anything about how strong the supervenience is supposed to be. Is it just that as
a matter of fact, the first set of properties supervenes on the second (what we
might call de facto supervenience), or is the supervenience a matter of logical
necessity (de jure supervenience)? To give an account that simply refuses to
further specify the relation between the mental and the physical, then, feels a
little unsatisfying. We havent explained very much at all.
Teaching tips
Discussion questions
Can you think of any other properties that you think supervene on the
physical domain, but are non-reducible to it?
47
Further reading
48
3.1 Functionalism
This view is non-reductive because its supporters dont think that mental states
are reducible to physical states. Talk of mental states on this view, rather,
amounts to talk of the inhabitants of certain functional or causal roles, which
themselves might be manifested in any number of ways. A full inventory of the
physical facts about the world, then, wont be enough to tell you all the mental
facts too; the latter cannot be reduced to the former (either ontologically or
explanatorily).
Functionally definability
Some things in the world offer themselves up to what well call a functional
characterisation. Lets introduce a couple of examples to demonstrate what I
mean. Take the example of necklaces. What is it to be a necklace? Well,
necklaces come in all sorts of shapes and sizes, and can be made out of lots of
different materials. You might have a necklace in the form of a delicate gold
pendant, or it might be a chunky string of glass beads, or a strand of peals, or
even a daisy chain made of flowers. The point here is that what matters when it
comes to settling the question of whether or not something counts as a necklace
is not what its made of. What matters is whether it does the job that necklaces do.
Does it hang around your neck and look aesthetically pleasing? If it does, then
youve got yourself a necklace no matter what its made of. A necklace, it seems,
is a functionally definable kind of object; we can define what it is in terms of the
role it plays rather than in terms of what its made of.
Not all objects are functionally definable in this way. Take, for instance, the
question of what it is to be a pearl. Is it alright in this case to say that we dont
really mind what its made of? That we will count it as a pearl so long as it
performs the pearl-role, or does the job that pearls do? Certainly not. To be a
pearl is to be composed of calcium carbonate crystals that have been hardened
over time in the soft tissue of a shelled mollusc, and nothing else will do. There
are, then, some kinds of objects that are simply unsuitable for functional
characterisation.
49
calcium carbonate crystals, for the identity theorist mental states just are brain
states (where we might be talking either type or token, depending on what kind
of identity theorist you are.)
An example
Take the example of pain. What makes something a mental state of pain?
According to the functionalist, its not the fact that the state is realised in a
particular brain state. For the functionalist, rather, what makes something the
mental state of pain is that the state plays a certain kind of role. We might, for
instance, say that the role of a pain state is to respond to tissue damage with both
certain kinds of further mental states (anxiety, self-pity, thoughts of retaliation,
etc.) and with certain kinds of outwards bodily behaviours, like saying ow,
crying, clutching the injured body part, and so on. As long as this kind of causal
profile is instantiated, then I will count as being in pain, regardless of how that
state is physically realised be it brain tissue, silicone chips, squid-brain, or
anything else.
Notice that as long as the functionalist thinks that these states are physically
realised somehow, then they will have to ascribe to some kind of token identity
theory. Each occurrence of a pain state identified by the functional role just
described will as a matter of fact be identical to a particular token physical
state. The difference with straight token-identity theorists, however, is that
functionalists dont think that this identity plays any role in what makes the state
a pain state. All of that work is done by the states functional role. It just also
turns out that the state so characterised will have to be physically realised
somehow.
The first difference is like the difference between the necklace and the pearl. For
the behaviourist, our mental states are like pearls. On their account, my mental
state of pain just is my writhing on the floor, holding my foot, etc. This kind of
50
behaviour, or my disposition to engage in it, is what my state of pain is, just as a
pearl just is a calcium carbonate deposit.
Functionalism is not like this, because functionalism doesnt say anything about
what my mental state is made up of. My state of pain, for the functionalist, is
more like the necklace what counts as my state of pain is just whatever it is that
performs the right kind of function, just as what counts as a necklace is just
whether or not it plays the right role.
The second difference is in the kinds of inputs and outputs that are permitted
into the analysis. For a behaviourist, these inputs and outputs will all be
outwardly perceivable stimuli and behaviours the observability of these
behavioural profiles, remember, was one of the advantages of behaviourism.
Such restrictions, by contrast, dont hold for the functionalist, who can also list
other inner mental states and events as part of a states functional role.
Physicalist or dualist?
As far as I know, there arent actually any substance dualist functionalists out
there. The point here is only that strictly speaking, functionalism doesnt entail
physicalism, so if you wanted you could be both a dualist and a functionalist. The
reason that as a matter of fact all functionalists have been physicalists, then,
must come from the independent reasons to favour physicalism over substance
dualism that weve already come across.
If, however, you are a physicalist and a functionalist, then you will have earned
yourself an extra argument against the substance dualist, or, at least, you will
have a way of responding to the conceivability argument that might not be
available to other physicalists. That argument, remember, was based on the
conceivability of having a mind while lacking a body. Given that conceivability is
supposed to be a good guide to metaphysical possibility, this was supposed to
show us that a separation of mind and body is metaphysically possible. But
identities are metaphysical necessities. So the mind is distinct from the body.
Functionalism gives us a way to respond to the first step of this argument
namely, to the claim that we can imagine undergoing a mental state in the
absence of a body. If the mental state is, as the functionalist would have it,
whatever performs the relevant function, then the functionalist can just respond
51
that this isnt really a case of imagining a non-physically instantiated mind.
Rather, there must be some other kind of physical stuff around doing the job.
Advantages of functionalism
Functionalism says that mental state is identical to whatever performs the right
kind of functional role. Its flexibility with regards to what this thing is what
makes the view so appealing in the face of the multiple realizability intuition
there are lots of different things that it could be, so we dont have to
chauvinistically restrict these mental states to creatures with brains just like
ours.
This virtue of the account, however, also ends up being its vice: it is this very
liberalism that is the source of the most common kinds of complaints against
functionalism. The view, this kind of objection goes, is too liberal, and by going
in for it we will end up ascribing mental states where we really shouldnt. Here
are two examples of this kind of problem for functionalism.
This argument, which was most famously given by Ned Block in the late 1970s, is
based on the thought that we could arrange systems to perform exactly the same
functional role as a given mental state, but where most of us would strongly
resist the claim that any mental state is being had.
52
into this system that is connected, again via a two way radio signal, to a human
body. That body communicates stimuli into the system, and performs output
behaviours on the basis of instructions coming out of the system. When a
stimulus of, for instance, a ticklish-feeling is introduced into the network,
communication between individuals in the system progresses in a way that
exactly mirrors the pattern of neural signals that really do fire up in my brain
when I actually get tickled. After the requisite activity has passed through the
system, suppose that a signal returns to the body from the network instructing it
that it should perform behaviours of giggling, flushing, seeking to avoid the
source of the tickling, and so on.
Together, the Chinese populants in this story perform the functional role
associated with the mental state of feeling ticklish a ticklish stimulus is fed into
the system, and the right kinds of behavioural outputs result. (We might
additionally suppose that it will also result in fresh patterns of signaling activity
within the network to parallel other mental states caused by the ticklish feeling).
But on the functionalists account, remember, a mental state is identifiable as
whatever it is that performs the appropriate functional role. In this case, this is
the population of China! Most think that this is a pretty unpalatable result, which
means that the functionalist theory must have gone wrong somewhere.
One of the reasons that we might resist the functionalists reading of this
scenario might be because we think that there is something important being
missed out: namely, that there is something it is like to feel ticklish (those pesky
qualia again!). Not only does this thought experiment expose the apparent
weakness of functionalism that it is overly liberal in what it will allow in as a
mental state, then, it also makes salient the fact that this view seems to have little
to say about how we should account for the conscious experiential features of
mental states. (See the Shoemaker text in further resources for more on this line
of objection).
The thought experiment begins with an imagined scenario that would seem to
support the strong AI thesis. Imagine that there is a computer which is
programmed in such a way that it can input strings of Chinese of characters, and
output sentences of English. If, as functionalism has it, the state of
understanding Chinese amounts to nothing more than the instantiation of the
relevant understanding Chinese functional role, then this computer will count
as understanding Chinese.
53
You might already be loath to accept to this functionalist result. Surely, you
might protest, its a computer, and computers dont understand anything at all.
At this point, the argument is a bit like the Chinese nation argument; it works by
simply showing that the functionalist view of the mind will allow some systems
to have mental states where our intuitions say there arent any.
Teaching tips
54
Discussion questions
What objects can you think of that are functionally definable? And that
arent functionally definable?
Can you come up with some examples of functional roles for different
mental states? (Remember to include both physical behaviours and other
mental states in their outputs).
Can you come up with your own example showing the problematic
liberalism of functionalism? I.e. can you come up with an example where
there a functional role normally associated with a given mental state is
realized, but where we would not want to say that a mental state is being
had?
What could the functionalist say in response to the Chinese nation and to
the Chinese room arguments? Are there ways to make palatable the
thought that there really are mental states here?
Further reading
55
3.2 Anomalous Monism
Just to be really clear, lets go through some terminology thats already cropped
up. First of all, why does Davidson call this position anomalous monism?
The account is monistic, because it takes it that there is only one kind of stuff in
the world. This can be contrasted with substance dualism, which is dualistic
because it thinks that there are two kinds of substances in the world the
physical and the mental. More specifically, the anomalous monist thinks that the
only kind of substance there is in the world is the physical stuff. Because of this
it is a physicalist monistic account.
That accounts for the monism part, but what about the anomalous bit? Well,
the word anomalous is derived from the Greek word nmos, meaning law, or
custom. If something is nomological, it is lawlike. If it is anomalous then it is
not lawlike. Just think of the word anomaly, which we use to describe
something that doesnt fit in with a pattern or lawlike regularity. Davidsons
physicalist monism is anomalous, because it takes it that mental states, events
and processes do not correspond to physical events in a lawlike way.
Finally, I said that this is a non-reductive account of the mind. Why is that? Well,
Davidson thinks that all mental occurrences supervene on things that happen in
the physical world. He has to say this given that as a physicalist monist he
thinks that physical stuff is the only kind of stuff that exists. He doesnt,
however, think that the mental events will be explained by a complete physical
account of the world. In the absence of lawlike connections that could tell us
56
which kinds of mental events correspond to which physical ones, there is no way
of telling which mental states and events there are from the physical facts. While
the mental supervenes on the physical, then, it is not reducible to it. And this is
just this combination of commitments that we used to characterise non-reductive
physicalist accounts.
An inconsistent triad
Davidson argues for his position using whats known as an inconsistent triad
(also sometimes known as a trilemma). An inconsistent triad is a list of three
claims, all of which are either well supported or intuitively compelling. The
claims, however, are in tension with each other it looks like they cant all be
true at the same time. This is why we call it an inconsistent triad: it is a puzzle
composed of three inconsistent claims, where no single one of them looks like an
obvious candidate for elimination.
Claim (1) is one weve already come across in earlier arguments. It says that
there are at least some mental events that sometimes causally interact with the
physical domain. Weve already spent some time urging this intuition elsewhere,
but if you want a quick reminder about why we might want to endorse this claim
have a look back to the interaction problem for substance dualism.
Claim (2) says that wherever one thing causes another thing, this must happen in
a strictly lawlike way. Although this claim has sometimes been questioned, this
seems at least on the face of it like a pretty plausible claim. Ill use an example to
try to show why this is.
Suppose that I throw a pebble into my pond in North London, and just at the
moment that it hits the water, a storm erupts in Leicester. It would, I take it, be a
pretty tall task to persuade anyone that my pebble-throwing caused the storm.
57
Consider now, however, a development of the case where it turns out that a
storm reliably and systematically erupts every time I throw a pebble into my
pond. Suppose, that is to say, that we can establish the presence of a lawlike, or
nomological connection between my pebble-throwings and the storm-eruptings.
Once we add this into the case then it stops looking quite so ludicrous to say my
throwing the pebbles is what is causing the storms to erupt, even if we dont
have any idea how the causal connection between them works. The principle
that causality is always lawlike, then, seems to be one that is supported by our
intuitions about when we can say that one thing causes another.
Claim (3) is the thesis that there are no strict laws on the basis of which mental
events can either be predicted or explained, or on the basis of which mental
events can themselves be used to predict or explain other events. Another way
of putting claim (3) is to say that there are no such things as strict psychophysical
laws, or laws connecting the mental realm to the physical. Although Davidson
first introduces this claim as an assumption about which, presumably, he thinks
we would mostly be in agreement, he does later give us two further arguments
for the claim. (Actually, there are more than two arguments for it, but the two
main ones will do for our purposes).
The first is the consideration of the holism of the mental. Something is holistic if
treating or explaining things about its parts can only be carried out by reference
to the whole. (Think of holistic medicine, for instance, that works by focusing on
a patients whole wellbeing, rather than on individual ailments). The holistic
argument for the anomaly claim is that we can only use mental events to predict
and explain physical events when we look at our mental lives holistically. My
mental event of thinking its raining, for instance, will only explain why I pick up
my umbrella in the context of a whole range of further mental states (my desire
to leave the house, my disliking getting wet, my sense of urgency that prevents
me from waiting until its stopped raining, etc.). Because of this, we cant say that
the mental state of thinking that its raining is systematically, or nomologically,
connected to any given physical state. It only leads to any physical events in the
world in virtue of its place in a holistic network of mental states.
The second argument comes from the thought that for any given set of physical
facts, there are lots of different ways we might interpret someones mental
states. When we observe someones brain-states and behaviour, that is to say,
that doesnt tell us straight-off whats going on with them mentally speaking. We
might call this the indeterminacy of translation between the physical and the
mental even once we have in place a complete list of a persons physical
58
properties at a given time, it might be indeterminate how we should translate
that talk of physical properties into mental talk there might be different mental
ascriptions that fit the physical facts equally well. If this is right, however, then
there cant be any strict laws connecting the physical and the mental, since if
such laws did exist, then they would give us a straightforward way to settle these
kinds of translation questions.
Inconsistency
Lets say that we agree with all three of Davidsons claims. Then it looks like we
have a problem. (1) and (2) together entail that mental events must be
nomologically connected to the physical, but this is precisely whats denied by
(3). What to do?
Instead of getting rid of any of these claims, Davidson finds an ingenious way to
make them compatible. What we have to do, he says, is just to recognise an
important difference between causal connections, which apply to things no
matter how we describe them, and lawlike connections, which only apply to
things under certain ways of describing them.
Lets have an example. Superman and Clark Kent are the same person. Which
name we use for that man, however, really depends on how we are thinking of
him, or what kind of description were thinking of him under. When we call him
Superman, for instance, we are thinking of him under a description like the
superhero with the blue and red costume. This, however, is clearly not how we
are thinking of him when we call him Clark Kent. When we use that name, we
are thinking of him as fitting a description like the geeky Daily Planet reporter.
What does it mean to say that causal connections dont depend on how an object
is described? Well, it means that it doesnt really matter which description were
using when we talk about things that causally happen to the man weve been
talking about. If Lois Lane causes Clark Kent to blush, then this is no different to
saying that shes causing Superman to blush; theyre one and the same man, so
one of them cant blush without the other doing so too. Likewise, if Superman
causes the villain to flee the city, then so does Clark Kent (even if most people
dont know it). They are just one and the same person, so one of them cant cause
villain-fleeings without the other doing the same. More generally, the moral here
is that as long as we have a single object in play, it really doesnt matter how we
describe it when giving an account of a causal event all that matters is that that
thing (which we call by both names) caused something to happen.
This, however, is not the same if we are thinking not about causal relations, but
about lawlike connections that serve to predict and explain things. Supermans
handsome face, for instance, might serve to explain Lois Lanes weak knees, or
might serve to predict that she will become flushed at the sight of him. For these
kinds of connections, it does seem to matter that weve called him Superman
instead of Clark Kent. After all, Clark Kents face is, as a matter of fact, the same
59
thing as Supermans face, but Clark Kents face cant be used to predict or explain
the same things as Supermans face did. Lois Lane doesnt become weak-kneed or
flushed when she thinks of the face as Clark Kents. When it comes to lawlike
connections that are used to explain and predict things, then, it does make a
difference how we describe them.
Anomalous Monism
How does all of this help Davidson to solve the inconsistent triad? The answer
comes in two parts.
The second part of the answer is that Davidson thinks that it does matter
whether were thinking of these events in their mental or in their physical guise
when it comes to the nomological connections that are used to predict or explain
things. This is how he manages to also accommodate claim (3), that mental
events cant explain or predict, or be explained or predicted by, physical events.
This is because explanations are sensitive to the way things are described, and
we have already seen that there are reasons to think that when events are
described mentally, they cant feature in explanations of, or be explained by,
physically-described events. This is the anomalous bit of anomalous monism.
Token identities
When I say that Davidson is a physicalist, notice that he cant be anything like a
type identity physicalist. This is because if certain kinds of mental states were
type-identical to certain kinds of physical states, then this fact would serve to
provide us with the very kinds of psychophysical nomological connections that
hes denying in claim (3). Instead, we can think of Davidson as a more refined
and developed kind of token identity theorist. Given his physicalist monism, he
will be happy to allow that every mental state is token-identical with a physical
state, but given his commitment to the anomaly of the mental, he must resist that
there are any type-identities.
The main objection that has been raised against anomalous monism is the
complaint that if you are an anomalous monist, then you cant avoid slipping into
epiphenomalism. Epiphenomalism, remember, is the view that mental states,
events and processes are causally inefficacious.
60
In a way, this is a bit like the problem that we came across in the causal closure of
the physical objection to substance dualism, which said that despite our strong
intuitions to the contrary, mental properties cant really cause anything in the
physical realm, because physical events are always already fully causally
accounted for within the physical domain. The objection here is pretty much the
same. Given that physical events are fully causally explained by other physical
events, what work is there left to do by the mental stuff? They must be causally
redundant, so epiphenomenalism must be correct. However, most people think
of epiphenomenalism as a result to be avoided at all costs, so would not be happy
to accept it as an upshot of their view. Anomalous monism is in trouble, then, if
really results in epiphenomenalism.
Davidson does have some things to say in response to this objection which I will
not cover here, but see further resources.
Teaching tips
This position is among the trickier ones among the ones covered in these
note, but almost everything needed to explain it will have already been
covered in the foregoing topics, so this probably one to teach towards the
end of a course. (E.g. token-identities, non-reductive physicalism etc.)
The main new thing that will need to be introduced here is the thought
that you can describe things in different ways, and that those descriptions
dont matter to causal questions, but do matter to explanatory or
predictive questions. Lots of examples will help here.
This position offers another opportunity to introduce students to a new
form of argument the inconsistent triad. Have a few simple toy
examples of inconsistent triads ready to demonstrate how they work as
an argument-type.
Discussion questions
61
Further reading
62
4. Eliminative Materialism
This claim is far more radical than any of the physicalist theories of mind that
weve seen so far. Even the reductive physicalists still thought that there are
such things as mental states, or that talking about mental states is still useful for
our purposes. For them the point was that when we talk about them, we are
really be talking about physical phenomena, or things that can be fully analysed
in physical terms. For eliminativists this is still too concessive. They think its
really only a matter of time and scientific advancement before we realise that the
things that we commonly think of as mental states in fact have no place in a
scientifically more complete and accurate worldview.
Folk psychology
I should point out that this explanation for how we explain and predict
behaviour is not the only one available. It amounts to whats called the theory-
theory a theory that says that we use a theory to make sense of peoples
behaviour. The main competitor to the theory-theory is called simulation theory,
which (very roughly put) holds that we predict and explain others actions by
effectively putting ourselves in their place and asking ourselves what we would
do. Given that the eliminativist assumes that we ascribe to theory-theory, one
way to refute the view is to reject that assumption. I have included some
references for the theory-theory vs. simulation theory debate in the further
resources, but this way of arguing against eliminativism involves introducing a
fairly substantial new topic area, so this might not be the most profitable line to
pursue. Other arguments against eliminativism are noted below.
63
To recap, then, the eliminativists target for elimination are the mental states,
events and processes that we postulate in our folk psychological explanations of
each others behaviour.
The eliminativist challenges us to say how we can we so sure that were not in
the same position when it comes to mental states. Once our scientific theories
have sufficiently matured, whats to say that we wont end up discarding serious
talk of mental states in just the same way that we have discarded serious talk of
witches? Just as it turned out that there were really no such things as witches,
the eliminativist predicts that its going to turn out that there were really never
any such things as mental states.
Stagnation
One of the tests for a good theory in the sciences is whether or not it gives rise to
fertile research. A research area is fertile when it leads to a good amount of
further active research, whereas a research area is stagnant when no further
64
progress is being made in it. In a way, this is the objection that the folk
psychological theory of mind is a bit of a dead-end; we dont seem to be getting
anything new out of it. By contrast, neuroscientific theories are flourishing
were learning more and more all the time about how our brain activities
correspond to our behaviours. The eliminativist recommendation is that we
should drop the stagnant theory in favour of more fertile ones that ultimately
promise us a much more complete understanding of the phenomena.
Another general theoretical problem with the theory of folk psychology is that it
doesnt seem to be much supported by other adjacent theories. In general, a
theory gets much of its support by featuring in a kind of mutually supportive
network of theories. The theory, for instance, that dinosaurs became extinct
about 66 million years ago because of an asteroid hitting the earth is supported
by a number of other nearby theories: natural selection theories, for example, or
theories about extinction patterns, and geological theories about asteroid
impacts on the earths surface and so on. Our dinosaur-extinction theory gets
support from consistency with the predictions of neighbouring theories.
Folk psychology doesnt seem to be like this. What other theories do we have
that would support the postulation of mental states? Indeed, the prognosis for
folk psychology seems to be even worse than this. For didnt our discussion of
the causal closure of the physical show that our postulation of mental states is
not only unsupported by neighbouring theories, but in fact even causes problems
for them?
Explanatory power
65
akratic actions (where we act contrary to our beliefs about the best way to act)
and the kinds of unconscious beliefs and desires that we find in Freudian
theories of psychoanalysis. These things, the eliminativist thinks, just cant be
explained using our everyday psychological theories, and this explanatory
weakness of folk psychology should give us reason to be suspicious of that
theory.
Eliminativism is a radical and controversial position, and I think its fair to say
that most philosophers would be extremely reluctant to take it overly seriously.
The problem, however, is in pinning down exactly whats wrong with it. After all,
the argument against it has got to be something stronger than mere repugnance
towards the view if you feel some resistance to the suggestion that we should
stop talking seriously about mental states, just think about how people in the
16th century would have reacted if you had said the same thing to them about
witches.
Given that the arguments for eliminativism were based on the theoretical
weakness of folk psychology, we might try to argue against eliminativism by
defending the credentials of that theory. This has been the tack taken by many
opponents of eliminativism. Other arguments against eliminativism involve
arguing that the view is internally inconsistent in some way that it is somehow
unintelligible, or self-refuting. I will not cover these arguments here, since they
move beyond the considerations of theoretical virtue that I have been focusing
on (but see further resources.)
Theoretical success
You might not think it from the eliminativists onslaught, but folk psychology is
actually a very successful theory. If theory-theory is correct, then we use it all the
time, in every one of our interactions with each other, to predict what others will
say or do. Even in the simple matter of having a conversation with you, for
instance, I will normally theorise that you are a willing conversational
participant, and so will plan my conversation on the expectation that you are not
about the bolt without warning. Indeed, it is plausibly the remarkable reliability
and accuracy of these theories that allows us to navigate the interpersonal world
so skillfully. When you think about it, its pretty amazing that we manage to
coordinate our behaviour that springs from inner mental activity into seamlessly
executed interpersonal activities. This kind of thing happens both on a small
scale like when we form a plan to meet tomorrow at noon, or when we move a
table together to the big, like sorting ourselves into communities, and
arranging large-scale projects together.
66
brain states? Probably not. Could we explain that sort of thing using folk
psychology? Yes, very easily we do so all the time. Even if there are some
things that folk psychology does not explain then, there are many more things
that it does explain.
Teaching tips
Discussion questions
All of these arguments are based on a theory that were supposed to apply
to others to explain and predict their behaviour. But dont we have a
more direct way of knowing that mental states really do exist? That is,
cant we just introspect them?
o An eliminativist could respond that were so immersed in our folk
psychological theories that we unwittingly apply those theories to
ourselves too when we take ourselves to be introspecting whats
really there. (Dont we post-hoc rationalise our actions all the
time, and in doing so postulate things like beliefs and desires for
ourselves?) The possibility that this is happening makes our
introspective evidence pretty unreliable.
For the general theoretical problems with folk psychology, what help do
you think appeal to the soft sciences like psychology, or Freudian analysis
might give us?
Are there more problems for the eliminativist when it comes to states
with qualia (experiential properties) like pain, then propositional states
like beliefs and desires? After all, we can feel pains, so surely we must be
right that they exist.
o An eliminativist could respond by saying that although its right
that we sometimes have these experiences, were wrong to try to
categorise them as pains. This is because the theoretical
postulate of a pain is a confused notion that doesnt really track
anything real in the world, or have a properly worked out
explanatory role. People on morphine, for instance, often report
having pains, but where they dont have any unpleasantness
67
associated with the feeling. Given these kinds of cases, its not
clear that weve hooked on to anything thats really out there in the
world with our notion of a pain.
Can you think of any other kinds of mental or behavioural phenomena
that cant be explained by our everyday psychological explanations?
What could the eliminativist say in response to the theoretical success of
folk psychology? Didnt people in the 16th century also think that the
witch-theory was highly successful?
o Perhaps there are problems with assessing a theoretical
framework from inside it. I.e., if we are implicitly assuming the
truth of folk psychology then it will be difficult to really assess its
success. There is always the risk of confirmation bias, which is
the risk that one will tend to process the data in favour of a
working hypothesis by ignoring anomalies, making salient the
cases that fit the hypothesis, etc.
Further reading
68