J.

Kim, Philosophy of Mind

Chap. 4 Mind as the Brain: The Psychoneural Identity Theory
In this chapter, Kim considers the psychoneural (or psychophysical, mind-
brain) identity theory, which holds that the mind is identical with the brain
and that for a creature to have mentality is for it to have a brain with
appropriate strucure and capacities.
Mind-Brain Correlations
We modern people, unlike some ancient Greeks, think that the organ
responsible for thoughts and feelings is the brain. What makes us think so is
the fact that there are pervasive and systematic correlations between mental
phenomena and neural states of the brain, which is known from considerable
empirical evidence. In view of the evidence, it is plausible that everything
that occurs in mental life has a state of the brain (or the central nervous
system) as its proximate physical basis. This idea can be formulated as
follows:
(Mind-Brain Correlation Thesis) For each type M of mental event that
occurs to an organism o, there exists a brain state of the kind B (Ms
neural correlate or substrate) such that M occurs to o at time t if
and only if B occurs to o at t.
According to this thesis, each type of mental event that can occur to an
organism has a neural correlate that is both necessary and sucient for its
occurrence. There are two points to be noted about the brain-mind correla-
tions in question. First, they are law-like: The fact that a particular type
of mental event occurs when some particular type of brain state obtains is
a matter of lawful regularity, not accidental, or coincidental, co-occurrences.
Second, even the smallest change in our mental life cannot occur unless there
are some specic (perhaps still unknown) changes in our brain state. To put
theses points shortly (though somewhat roughly), mentality supervenes on
brain states.
1
Making Sense of Mind-Brain Correlations
The question then arises about how we can explain or interpret the observed
systematic correlation between mental and neural properties or events. The
following are the examples of theories that have been proposed as the answer
to this question:
Causal interactionism holds that mental and physical events are di-
rectly causally related, just as the fall of the ambient temperature and
the freeze of the water in a pond are.
The theory of preestablished harmony holds that mental and physical
events are collateral eects of a common cause and that the cause is a
Gods action. On this view, the mind-body relation is analogous to the
relation among the clocks that were synchronized by the shopkeeper
before the shop opened in the morning.
Occasionalism, like the theory of preestablished harmony, holds that
mental and physical events are collateral eects of a common cause,
but it holds that what is responsible for mind-brain correlations is a
continued intervention of God, rather than a single action of God in
the past: when a particular type of mental event occurs, that serves
as an occasion for God to intervene and cause a particular type of
physical event, and vice versa. This way of explanation is similar to the
explanation that invokes continuous interventions of a little leprechaun
in order to account for the correlation among clocks.
The double aspect theory holds that mind and body are two aspects of
one underlying reality. On this view, the mind-body relation is analo-
gous to the relation between the temperature and pressure of a gas.
Emergentism holds that the correlations between mind and body are
brute fact that we must simply accept. According to this view, when
biological process attain a certain level of organizational complexity,
a wholly new type of phenomenon, namely, consciousness and ratio-
nality, emerges, and why and how these phenomena emerge is not
explainable in terms of the lower-level physical-biological facts.
The psychoneural identity theory advocates the identication of mental
states with physical processes in the brain. According to this view,
there are no mental events over and above, or in addition to, the neural
processes in the brain, just as there are no bolts of lightning over and
above atmospheric electrical discharges.
2
The Argument from Simplicity
One principal argument for the mind-brain identity theory is the simplicity
argument. The principle that urges simplicity as an important virtue of
theories and hypotheses is known as Ockhams razor, among whose standard
formulations are:
(I) Entities must not be multiplied beyond necessity.
(II) What can be done with fewer assumptions should not be done with
more.
Principle (I) urges us to adopt the simplest ontology possible, one that posits
no unnecessary entitiesthat is, entities that have no work to do. Principle
(II) can be taken as urging simplicity and economy in theory construction:
It recommends that we should choose the theory that gives the simplest,
most parsimonious descriptions and explanations of the phenomena in its
domainthat is, the theory that does its work with the fewest independent
hypotheses and assumptions. In view of Ockhams razor, there seem to be
the following three lines of consideration one might pursue in attempting to
argue in favor of the mind-brain identity theory on the ground of simplicity:
(1) Psychoneural identication reduces the number of putative entities
and thereby enhances ontological simplicity. If pain as a mental kind
is identied with its neural correlate, we simplify our ontology on type
level, and hence on token level as well.
(2) Psychoneural identication enhances conceptual or linguistic simplic-
ity as well. If all mental states are systematically identied with their
neural correlates, there is a sense in which mentalistic language is in
principle replaceable by a physical language. On the identity theory,
physical language is complete and universal in the sense that there are
no excess facts beyond physical facts that can only be described in
some nonphysical language.
(3) Psychoneural identication saves the need for adopting an intolera-
bly complex and bloated theory that includes among its basic laws
a huge and motley crowd of psychoneural correlation laws, each of
which is highly complex. If we move from psychoneural correlations to
psychoneural identities, we get a simple and elegant picture in which
our basic physics ultimately constitutes a complete and comprehensive
explanatory framework adequate for all aspects of the natural world.
3
However, there is room for doubt as to the claim that psychoneural iden-
tication simplies our total theory of the world: the identity theory might
seem to merely replace psychoneural correlations with an equal number of
psychoneural identities, which are not deducible from more basic physical-
biological laws any more than the correlations are. If so, the total empirical
content of a theory with psychoneural identities is at least equal to that of
a theory with the psychoneural correlations they replace, and version (II) of
the simplicity principle will seem to argue against psychoneural idendities.
Explanatory Arguments for Psychoneural Identity
Another principal argument for the identity theory is the explanatory argu-
ment. It claims that psychoneural identities help explain certain facts and
phenomena that would otherwise remain unexplained, or that they give the
best explanations of certain facts, better than the explanations aorded by
rivals. One version takes the explanandum to be psychoneural correlations,
claiming that psychoneural identities give the best explanation of psychoneu-
ral correlations. Another version claims that the identities explain certain
other facts about mental phenomena that would otherwise go unexplained.
Explanatory Argument I
The rst version claims that the psychoneural identity theory oers the best
explanation of both specic psychoneural correlations (e.g., pain occurs
i Cfs occurs) and the general fact of psychoneural correlations (i.e., the
mind-brain correlation thesis). However, even as for specic psychoneural
correlations, there are serious reasons to be skeptical about their purported
explanations oered by the identity theory being explanations at all:
(1) Given psychoneural identities, questions about psychoneural corre-
lations cancel themselves as explananda, since their presuppositions
about the very correlations are false. Even if we stick with i sen-
tences as explananda, it is doubtful that their purported explanations
oered by the identity theory could be informative.
(2) Psychoneural identication transforms the supposed explananda into
mere tautologies (like pain occurs i pain ocuurs) that seem neither
in need of explanation nor capable of receiving one.
In short, psychoneural identities disqualify psychoneural correlations as ex-
plananda, rather than explaining them. At best, the identities only enable
us to rewrite contentless tautologies by putting equals to equals.
4
(Kim mentions a more general problem of using the principle of inference
to the best explanation to support what is essentially philosophical claim:
since claims like pain = Cfs, unlike scientic hypotheses, have no predictive
implications of their own and hence no possibility of further tests, applying
the principle to such claims seems at best a misapplication of it.)
Explanatory Argument II
The second version claims that psychoneural identities enable us to explain
certain facts about mentality that would otherwise remain unexplained (e.g.,
the fact that pain causes a feeling of distress) by deriving them from neu-
rophysiological facts (e.g., the fact that Cfs causes neural state N) on the
basis of the psychoneural identities (e.g., pain = Cfs and distress = neu-
ral state N). However, this argument turns out to be problematic in that
the identities in question, again, seem to do no explanatory work and hence
are not qualied to benet from the principle of inference to the best ex-
planation: the identities enable us only to redescribe, in the familiar folk
language, a fact that has already been explained by laws of neurophysiology.
An Argument from Mental Causation
Yet another principal argument for the psychoneural identity theory is the
causal argument. Its currently inuential version goes as follows:
(i) Mental phenomena have eects in the physical world.
(ii) The physical world is causally closed. That is, if any physical event is
caused, it has a sucient physical cause (and a wholly physical causal
explanation).
(iii) Therefore, mental phenomena are physical phenomena.
Strictly speaking, the only proposition we are entitled to derive from (i)
and (ii) is that only those mental phenomena that cause physical events
are physical events; but given the transitivity of causation, every mental
event in any causal chain of mental events the last one of which causes a
physical event qualies as a physical event. Then all mental events should
be covered, unless there were a mental causal chain consisting exclusively of
mental events. (Even if there were one, the main physicalist point is made.)
Of course, the best candidates of physical events with which mental events
are identied are neural correlates of the mental, each of which is a necessary
and sucient condition for the occurrence of a mental event.
5
If we insist on mental events as a separate causes of physical events, we
will face the problem of causal overdetermination, two independent causes
bringing about a single eect. Epiphenomenalists might try to solve this
problem by claiming that a single physical cause brings about both a physical
event and a mental event; but then mental events would be rendered causally
inert, premiss (i) being rejected. At this point, the causal argument gives us
a choice between psychoneural identity and epiphenomenalism.
Against Psychoneural Identity Theory
The Epistemological Argument
Epistemological Objection 1. The objector claims that if a mental kind
is identical with its neural correlate, any person who knows something about
the former knows something about the latter as well, and that this conse-
quent is clearly false. However, the conditional asserted by the objector ap-
pears false, as is clear from the case of water and H
2
O. The objector might
insist that any person who knows something about water knows something
about H
2
O as well; but if we can concede that, there will be nothing wrong
about saying that e.g., medieval peasants knew something about Cfs.
Epistemological Objection 2. The objector claims that to make sense
of the supposed empirical character of psychoneural identities, we must ac-
knowledge the existence of phenomenal, qualitative characters of experience
distinct from neural properties. The reason is that if D
1
= D
2
is an em-
pirical truth, the two names or descriptions, D
1
and D
2
, must have
independent criteria of application, and it is by recognizing such properties
that we identify conscious experiences like pain.
To meet this objection, identity theorists must somehow show that sub-
jects do not identify mental states by noticing their qualitative properties.
They might try to analyzing away phenomenal properties, but it is hard to
conceive an analysis of them that is satisfactory from the rst-person point
of view. A more recent strategy hat has become popular with latter-day
type physicalists is the phenomenal concept strategy: the main idea is to
concede conceptual dierences between the mental and the neural but deny
that these dierences point to ontological dierences, that is, dierences in
the properties to which these concepts apply. On the basis of this idea, it
is claimed that a single property, presumably a physical-neural property,
is picked out by both a phenomenal and a neural concept. Whether this
strategy is of philosophical substance remains to be seen.
6
Epistemological Objection 3. The objector claims that we have private
direct accessthat is, rst-person access not based on inference from evince
or observationto (at least some of) our mental states, whereas we do not
have such access to any of our brain states; and that (at least some of)
mental states, therefore, cannot be identied with brain states.
To meet this objections, identity theorists might deny the claim that we
do not have direct private access to our brain states by saying that certain
brain states like Cfs can be known in two dierent modes of presentation
or under two dierent sorts of concepts, mental and physical; and that the
knowledge of these brain states is private under the mental mode or con-
cept, whereas it is public under the physical mode or concept. This reply
will likely stand or fall together with the latter-day type physicalists re-
ply to epistemological objection 2, which invokes phenomenal concepts; and
vindicating it will require us to answer and elucidate many questions about
the source and nature of the concepts or modes of presentation that give
rise to a very special type of knowledge, that is, knowledge by direct private
access.
The Modal Argument
Another main argument against the mind-brain identity theory goes as fol-
lows: (i) If mind-brain identities are true, they are necessarily true, since
expressions that refer to them are rigid designators (i.e., expressions that des-
ignate the same objects in all possible worlds); (ii) Mental events and their
neural correlates are existentially independent, that is, the former could have
been exist without the latter, and vice versa (since there are the possibilities
of disembodied souls and zombies); (iii) if X and Y are existentially inde-
pendent, X = Y is contingently true, if true; (iv) Therefore, mind-brain
identities are false.
Many identity theorists are likely to dispute premiss (ii). They will ar-
gue that although the existence of disembodied souls and zombies are con-
ceivable, conceivability does not entail real possibility, which is shown by a
posteriori necessary identities like water = H
2
O and heat = molecular
kinetic energy; and that psychoneural identities are necessary a posteriori
truths just like these scientic identities. Issues about conceivability and pos-
sibility are highly complex and contentious, and there has been no consensus
resolution as yet.
7
The Multiple Realization Argument
The most inuential argument against the identity theory is the multiple re-
alization argument. It derives the conclusion that no mental state type can be
identied with a physical state type from the premiss that any mental state
is multiply realizable: for any mental state M, there must be indenitely
many physical states that can realize it in all sorts of M-capable organisms
and systems. This argument bring about an early decline of psychoneural
identity theory, and it has had a critical impact on the way philosophy of
mind has developed. What made it distinctive was that it brought with it
a fresh and original conception of the mental, namely functionalism. This
conception oered an attractive alternative approach to the nature of mind,
and it is still the reigning orthodoxy on the nature of mentality and the
status of psychology.
Reductive and Nonreductive Physicalism
The psychoneural identity theory is a form of reductive physicalism, which
reductively identies (types of) mental states with (types of) neural states of
the brain. As reductionisms fortunes declines chiey on account of the inu-
ence of the multiple realization argument, nonreductive physicalism gained
strength and inuence. It comprises the following four claims: (1) Substance
Physicalism (See chapter 1); (2) Irreducibility of the Mental ; (3) Mind-Body
Supervenience or Realization (The realization relation is stronger than su-
pervenience); (4) Mental Causal Ecacy.
We can think of property dualism as the conjunction of (1), (2), and (4).
Besides its acceptance of substance physicalism, what makes nonreductive
physicalism a serious physicalism is its commitment to mind-body superve-
nience/realization. The irreducibility thesis can be taken to be an armation
of the autonomy of psychology and cognitive science as sciences in their own
right, not constrained by more basic sciences. In accepting the causal e-
cacy of the mental, the nonreductive physicalist not only acknowledges what
seems so familiar and obvious to common sense, but at the same time, it
declares psychology and cognitive science to be genuine sciences capable of
generating law-based causal explanations and predictions.
Nonreductive physicalism has reigned as the dominant and virtually un-
challenged position on the mind-body problem for the past several decades,
but there have recently been signicant objection and criticisms of its nonre-
ductive aspect. (Some of the diculties nonreductive physicalism faces in
regard to mental causation will be seen in chapter 7.)
8