Identity, Explanation, and Consciousness: A Reply to Kim

About This Presentation

Title:

Identity, Explanation, and Consciousness: A Reply to Kim

Description:

Identity, Explanation, and Consciousness: A Reply to Kim Brian McLaughlin Rutgers University Huxley on the Mystery of Consciousness How it is that anything so ... – PowerPoint PPT presentation

Number of Views:225

Avg rating:3.0/5.0

Slides: 115

Provided by: Philosophy6

Category:

more less

Transcript and Presenter's Notes

Title: Identity, Explanation, and Consciousness: A Reply to Kim

1
Identity, Explanation, and Consciousness A Reply
to Kim

Brian McLaughlin
Rutgers University

2
Huxley on the Mystery of Consciousness

How it is that anything so remarkable as a state
of consciousness comes about as a result of
irritating nervous tissue, is just as
unaccountable as the appearance of the Djin,
where Aladdin rubbed his lamp (Thomas Huxley
1866)

3
McGinn on the Mystery

The specific problem I want to discuss concerns
consciousness, the hard nut of the mind-body
problem. How is it possible for conscious states
to depend upon brain states? How can technicolour
phenomenology arise from soggy grey matter?...How
could the aggregation of millions of individually
insentient neurons generate subjective awareness?
We know that brains are the de facto causal
basis of consciousness, but we have, it seems, no
understanding whatever of how this can be so. It
strikes us as miraculous, eerie, even faintly
comic. Somehow, we feel, the water of the
physical brain is turned into the wine of
consciousness, but we draw a total blank on the
nature of this conversionThe mind-body problem
is the problem of understanding how the miracle
is wrought, thus removing the sense of deep
mystery. (McGinn 989, p.394-395)

4
Chalmers on the Mystery

There is no question that experience is closely
associated with physical processes in systems
such as brains. It seems that physical processes
give rise to experience, at least in the sense
that producing a physical system (such as a
brain) with the right physical properties
inevitably yields corresponding states of
experience. But how and why do physical
processes give rise to experience?...This is the
central mystery of consciousness (2000, p.248)

Suppose that we were able to confirm that the
following is true
Correlation Thesis. For any type of state of
phenomenal consciousness C there is a type of
neuro-scientific state N such that it is
nomologically necessary that a being is in C if
and only if the being is in N.

The correlation thesis itself would not offer a
solution to the problem of the place of
phenomenal consciousness in nature.
The correlation thesis is compatible with a
variety of would-be solutions to that problem,
including analytical functionalism, Cartesian
substance-dualism, emergent property dualism, and
neutral monism.

Indeed Huxley, McGinn, Chalmers and many other
philosophers maintain that the mystery of
phenomenal consciousness would remain How do the
neuro-scientific states in question give rise to
states of phenomenal consciousness?
How, for example, does N give rise to C, rather
than some other kind of state of phenomenal
consciousness distinct from C, or no state of
phenomenal consciousness at all?

Notice that a presupposition of this how-question
is that N gives rise to C.
If we believe that N gives rise to C, then we
will wonder how N gives rise to C.

It seems that it will either be a brute,
unexplainable fact that N gives rise to C or else
there will be some mechanism by which N gives
rise to C.
Either way, it is indeed hard to see how the
sense of mystery could be dispelled.

If N gives rise to C but not via any mechanism,
then it seems that the fact that N gives rise to
C will be a brute fact that we will simply have
to accept with what the emergentists of the early
twentieth century (Alexander and Broad, for
instance) called natural piety.
Their view was that among the fundamental laws of
nature are the laws correlating neuro-scientific
states with states of phenomenal consciousness.
In accepting this, we would not be dispelling or
removing the mystery of consciousness we would
simply be accepting it, with natural piety, as a
fact of life.

But if instead there is some physical mechanism
by which N gives rise to C, it is hard to see how
knowledge of it could possibly remove the sense
of mystery.
For the how-question would recur for the
relevant physical factor at work in the
mechanism if N gives rise to C by giving rise to
P which, in turn, gives rise to C, then how does
P give rise to C?

Were the correlation thesis true, however, one
epistemic possibility would be that N does not
give rise to Cthat the presupposition of the
how-question is false.
Rather than N somehow giving rise to C, it might
instead be the case that N is C.
There may be one type of state that is
conceptualized in two different ways two
conceptsthe concept of N and the concept of
Cthat answer to the same state type.

On this view, the explanation of why N is
nomologically correlated with C is that N C.

Consider, then
The Identity Thesis. Types of states of
phenomenal consciousness are identical with
neuro-scientific states.
The identity thesis entails the correlation
thesis.

It is a frequently made point that identities are
not themselves explainable.
There is no point to the question Why is A B?
There is no point, that is, unless the intent of
the question just is to ask why we should believe
that A B.
But that is a request for a justification, not a
request for an explanation.

Of course, we should not take the hypothesis that
N C to in fact explain why N is correlated with
C unless the identity hypothesis is epistemically
justified.

But the point to note is that the justification
for the identity hypothesis need not be
epistemically prior to the justification for the
correlation thesis.
The justification for the identity hypothesis
might be that it offers the best explanation of
the correlation.

Chris Hill and I have argued jointly and
independently that if the correlation thesis were
true, then the identity thesis would offer the
best explanation of why it is true.
We thus both offer a conditional defense for type
materialism or type physicalism for states of
phenomenal consciousness.
In their joint paper Conceptual Analysis,
Dualism, and the Explanatory Gap (Philosophical
Review 1999), Ned Block and Robert Stalnaker
argue that in the case of states of phenomenal
consciousness, type-type psycho-physical identity
claims could be defended by inferences to the
best explanation.

In his book, Physicalism or Something Near Enough
(Princeton 2005), Jaegwon Kim devotes a chapter
aptly entitled Explanatory Arguments for Type
Physicalism and Why They Dont Work to trying to
show that best explanation defenses of type
physicalism dont work.
In what remains, Ill respond to his objections.

20
Kims Main Claims

(1) Despite a surface similarity, the position
that Hill and I take is in fact incompatible with
the position that Block and Stalnaker take so we
cant all be right.
(2) There is reason to be skeptical that there is
any rule of inference to the best explanation.
(3) In any case, identities are not explanatory
so explanatory arguments are doomed to failure.
(4) Hill and I are best viewed as appealing to a
simplicity argument in favor of type identities,
while Block and Stalnaker are best viewed as
appealing to a causal argument for
psycho-physical identities.
(5) Neither simplicity arguments nor causal
arguments will establish psycho-physical
identities.
(6) The would-be defense of type physicalism is a
hopelessly flawed attempt to close the
explanatory gap.

Kim has one positive comment on our would-be
explanatory arguments.
He applauds the fact that we are trying to offer
a positive case for type physicalism, rather than
simply sitting back and saying I think conscious
states are type identical with physical states,
refute me if you can!.
He cites Brian Loar in this regard John Perrys
antecedent physicalism also comes to mind.

It should be noted, however, that there is an
important role for the antecedent physicalism
strategy.
For the leading arguments against type
physicalism attempt to show that it is a priori
false that types of states of phenomenal
consciousness are types of physical states.
Moreover, as will become clear in due course, Kim
fails to appreciate how work in antecedent
physicalism will be incorporated into an
explanatory argument for type physicalism.

23
Kim on Identity and Correlation

Kim states that he agrees with Jack Smarts
(1959) claim that A is correlated with B is
incompatible with A B.
As Kim puts it, identities and correlations
exclude each other.
He claims, moreover, that Stalnaker and Block
correctly take identities to exclude
correlations, while Hills and I mistakenly think
that identities explain correlations.
Hill and I take psychophysical identities to be
warranted by the fact that they offer the best
explanations of correlations.
Block and Stalnaker, however, know better because
they recognize that identity is incompatible with
correlation.
So despite surface similarities, the
Block-Stalnaker position is in fact incompatible
with the Hill-McLaughlin position.

24
In Response

The first point to note is that neither Hill nor
I make essential use of the word correlation.
The thesis that I call the correlation thesis
does not use the word correlate or any of its
cognates.
Correlation Thesis. For any type of state of
phenomenal consciousness C, there is some type of
neuro-scientific state N such that it is
nomologically necessary that a being is in C if
and only if the being is in N.

Moreover, as Kim himself acknowledges, the type
identity thesis implies the correlation thesis.
So by Kims own lights, the identity thesis and
the correlation thesis are compatible.
Thats what matters.
Its irrelevant whether the fact that A B
entails that A is not correlated with B.

I cannot, however, resist noting that Smart and
Kim are mistaken in thinking that correlation is
incompatible with identity.

Smarts paper was published in 1959, and so a few
years prior to H.P. Grices 1963 paper, The
Causal Theory of Perception, a paper in which
Grice distinguishes conversational implication
from semantic implication.
Smart and Kim have confused a conversational
implication with a semantic one.

A is correlated with B may well have the
conversational Doubt-or-Denial condition that A
B.
Subtleties aside, in an ordinary conversational
context, when making a statement, we are under a
prima facie obligation not to make a weaker
statement than we know to be true.
Statements of correlation are weaker than
statements of identity.
Thus perhaps ordinarily we shouldnt state that A
is correlated with B unless we doubted that A B,
or took ourselves to be in a position to deny
that A B.

But this D-or-D condition is a conversational
implication, and so can be cancelled.
And it is cancelled in the kind of explanatory
context in question.

In such an explanatory context, we are concerned
to do more than provide information we are
concerned to show that a certain claim is made
true by another.
That A B makes it the case that A is correlated
with B.

I think that Kim is also mistaken in claiming
that Block and Stalnaker take the position that
identities exclude correlations.

Kim offers the following quotation from Block and
Stalnaker in support of his reading
If we believe that heat is correlated with but
not identical to molecular kinetic energy, we
should regard as legitimate the question of why
the correlation exists and what its mechanism is.
But once we realize that heat is molecular
kinetic energy, questions like this will be seen
as wrong-headed.

Notice that they write, if we believe that heat
is correlated with but not identical to molecular
kinetic energy, which suggests they think
identity is compatible with correlation.

Moreover, Block and Stalnaker elsewhere speak of
mere correlation
if we were to accept mere correlations instead
of identities
identities allow a transfer of explanatory and
causal force not allowed by mere correlations

Correlation is compatible with identity.
Mere correlation is correlation without identity.

Kim has failed to identify any incompatibility
between the Hill-McLaughlin position and the
Block-Stalnaker position.
In fact, the positions seem to me to be
essentially the same.

Block and Stalnaker are of course correct in
calling wrong-headed the two-part question, Why
is there a correlation between heat and molecular
kinetic energy and what is the responsible
mechanism?
The second half of the question (What is the
mechanism?) is wrong-headed in that its
presupposition is false.

There is no mechanism that explains why heat is
correlated with molecular kinetic energy.
Heat is molecular kinetic energy.

39
Kim on Scientific Explanation

Kim claims
In science there seem to be only two principal
ways of explaining correlations(1) invoking a
single lower-level process(2) showing the
correlated phenomena to be collateral effects of
a common cause.

I (we) claim there is a third way by pointing
out that the correlates are identical.
Because A B can provide a correct,
informative answer to the question Why is it
that A is present when and only when B is?

Now I think that Kim would acknowledge that
identity claims can be informative answers to
such why questions.
But he would deny that this suffices for A B
to be explanatory in any scientific sense.

He maintains that scientific explanations provide
information about causal history or about an
underlying mechanism.
And of course identity claims dont do either.

43
In Response

But not all scientific explanations of
correlations provide information about causal
history or about an underlying mechanism.

When Maxells calculations showed that
electromagnetic waves have the same speed in a
vacuum as the known speed of light, he famously
make the bold conjecture that light waves
electromagnetic waves.

Maxwell also pointed out that electromagnetic
waves are refracted when going from one kind of
material to another in a manner that depends on
the refractive indices, K, of the material.
When it was established experimentally that light
refracted according to the principles that apply
to electromagnetic radiation, this was taken to
confirm Maxwells bold conjecture.

The hypothesis that light waves are
electro-magnetic waves was invoked to explain why
(1) electromagnetic waves have the same speed in
a vacuum as light waves, and why (2) the various
refractive indices in materials are exactly the
same for light waves and electro-magnetic waves.

This explanation, one of the greatest
achievements of classical physics, is an
explanation by appeal to identity.
Moreover, the identity claim was arrived at by
something like inference to the best explanation.

48
The Principle of Inference to the Best Explanation

As Kim notes, Hill, Block, and Stalnaker all
speak of the principle of inference to the best
explanation.
As he also notes, I dont.
I claim only that psychophysical identities offer
the best explanation of psychophysical
correlationsbest on grounds of overall coherence
and theoretical simplicity.

The reason that I didnt speak of the principle
of inference to the best explanation is that no
one knows what, exactly, that principle is.

While Block and Stalnaker say they are appealing
to the principle of inference to the best
explanation, they nowhere state the principle.

Hill (1991) offers the following formulation of
the principle If a theory provides a good
explanation of a set of facts, and the
explanation is better than any explanation
provided by a competing theory, then one has good
and sufficient reason to believe the theory is
true.

But echoing Bas van Fraseen, Kim asks what if the
explanation is only marginally better than its
competitors?
We might also ask, again echoing van Fraseen,
what if the theory is only the best of a bad lot?

Indeed Hills version of the principle is false.
At best, the fact that a theory provides a better
explanation of a set of facts than any of its
competitors is a reason to believe the theory is
true but it is by no means a sufficient reason
to believe that the theory is true.
The reason can be defeated, rebutted, by other
reasons.

In Harmans seminal paper, The Inference to the
Best Explanation, he spoke of a hypothesis
offering a sufficiently better explanation than
its competitors.
Such comparative judgments, he pointed out, will
be based on considerations such as which
hypothesis is simpler, which is more plausible,
which explains more, which is less ad hoc, and so
on.

These considerations, and considerations of
mutual support through explanatory coherence are
the kinds of considerations that I has in mind
when I spoke of considerations of overall
coherence and theoretical simplicity.

Kim correctly notes Inference to the best
explanation has been advertised as inductive
inference.
There is, he notes, supposed to be a rule that is
rationally compelling, objective, and ampilative.

Kim warns that we should be cautious, if not
suspicious, about any inference rule that
essentially rests on considerations of
explanation.
In defense of this warning, he simply references
van Fraseens (1989) well-known critical
discussion of the rule of inference to the best
explanation.

Van Fraseen remarks that the rule of inference to
the best explanation is a rule, as yet unknown,
which will make us revise our probabilities so as
to give bonus marks to the hypotheses that
explain observed phenomena.
And he argues there cannot exist any such
probabilistic rule, on pain of incoherence.

He imagines that the new rule of IBE would be a
recipe for adjusting our personal probabilities
in response to evidence, in a way that takes
into account the explanatory success of
hypotheses.

The idea, then, is that the rule will work so
that explanatory hypotheses will receive bonus
marks when we update our beliefs in response to
evidence.
But that, of course, as he shows, is incompatible
with what would result from our updating our
beliefs by Bayess rule of conditionalization,
and would leave us open to a Dutch Book.

61
In Response

One issue is whether there is a rule of inference
to the best explanation that is compatible with
updating ones beliefs by conditionalization.
Another issue is whether the idea that we can be
warranted in believing something on the grounds
that it offers the best explanation of something
is compatible with updating ones beliefs by
conditionalization.
Im mainly concerned with the latter.

There indeed cannot be a rule of inference to the
best explanation of the sort van Fraseen imagines
that is compatible with the results of updating
ones beliefs by conditionalization.

However, as van Fraseen himself acknowledged,
introducing such a rule is not the only way to
try to implement a notion of inference to the
best explanation within a Bayesian framework.

Indeed the natural way to try to handle
explanatory potential in a Bayesian framework is
by taking it into account in forming priors, that
is, in distributing ones initial degrees of
belief.

There would thus be two components to a Bayesian
implementation of a notion of inference to the
best explanation
(a) the comparative explanatory potential of the
hypothesis H with its competitors, and
(b) the likelihood of H on the evidence (the
explanandum) relative to the likelihoods of the
competitors on the evidence (the explanandum).
Explanatory potential would thus be taken into
account in the prior probabilities of H and its
competitors.

It will have to suffice to note that I am
optimistic that a notion of inference to the best
explanation can be incorporated into a Bayesian
framework.
Moreover, I think that correlations are always
evidence for a posteriori identities.
In some cases, the likelihood of identity on the
correlation will be vanishingly small but in
others, it will be strong.
It all depends on ones priors.

67
Hills IBE Argument (1991)

First Premise. If a theory provides a good
explanation of a set of facts, and the
explanation is better than any explanation
provided by a competing theory, then one has a
good and sufficient reason to believe that they
theory is true.
Second Premise. Type materialism provides a good
explanation of the psychophysical correlations
that are claimed to exist by the psychophysical
correlation thesis.
Third Premise. Moreover, the explanation that it
provides is superior to the explanation provided
by all competing theories.
Conclusion. Provided that the psychophysical
correlation thesis is true, we have good and
sufficient reason to believe that type
materialism is true.

68
Kim On Hills Arg.

Kim says There are other serious questions to
think about concerning this application of the
principle of inference to the best explanation.
Given that type physicalism is to be compared
with its rivals in point of explanatory power,
why are these theories tested only in respect of
how well they explain psychophysical
correlations? Inductive reasoning, whose aim is
to reach a conclusion that we should believe as
true, or be prepared to use as a guide to action,
must respect total evidence, and in the present
case this means the data, or evidence, to be
explained must be all the data relevant to the
issue at hand. It may not be exactly clear what
these data are in the present case, but surely it
is arbitrary only to consider the fact of
psychophysical correlations

69
Quote Continued

Dualists will say that among the relevant data
are such things as the presumptive privacy of the
first-person access to ones own mental states,
the persistence conditions of person, the
multiply realizability of mental properties, the
possibility of qualia inversions, the possibility
of zombies, and the like. These are all
contested and disputed issues however, like it
or not, dealing with them is the mind-body
problem. What we should not do, it seems to me,
is to pick the fact of psychoneural correlations
alone as the explanandum, and argue that since
type physicalism offers the best explanation of
this datum, it must be the overall winnerand,
moreover, it must be true! It is possible that
type physicalism gives the best overall account
of all these facts, but that must be shown, and
that would go far beyond the kind of explanatory
argument under consideration.

Second Premise. Type materialism provides a good
explanation of the psychophysical correlations
that are claimed to exist by the psychophysical
correlation thesis.
An adequate defense of the second premise
requires an answer to all of the would-be
arguments that type materialism is a priori
false the possibility of zombies, inverted
qualia, the subjectivity of conscious properties
and the objectivity of neuroscientific ones, and
so on.

The issue of multiply realizability would have to
be addressed as well in a defense of the Second
Premise.
And a defense would have to answer Kripkes
challenge that, for an a posteriori identity
claim to be justified, we must be able to explain
away its appearance of contingency.
For without an answer to this challenge, the
psychophysical identity claims will seem ad hoc.

Third Premise. Moreover, the explanation that it
provides is superior to the explanation provided
by all competing theories.
An adequate defense of this premise would require
consideration of much more than psychophysical
correlations.
An adequate defense would require that one
compare type physicalism with its competitors on
holistic grounds.

Recall Harmans claim that judgments of
comparative goodness of explanations, will be
based on considerations such as which hypothesis
is simpler, which is more plausible, which
explains more, which is less ad hoc, and so
forth.
The relevant notions of simplicity, plausibility,
and explanatory scope are holistic.

A defense of the third premise will no doubt
appeal, in part, to the theoretical unifying
power of psycho-physical identities, the fact
that it has fewer free parameters than its
competitors.

In summary, then, all of the mind-body issues
that Kim cites would have to be addressed in an
adequate defense of the premises of Hills best
explanation argument.
Hills attempt to offer this argument is in no
way whatsoever an attempt to avoid or ignore
those issues.

76
How are We Arguing?

While Block and Stalnaker speak of inference to
the best explanation, Kim claims that they seem
actually to be offering a familiar causal
argument for physicalism.
He is led to say this, I think, because their
examples of explanations achieved by
psychophysical identities indicate how
physicalism would vindicate our belief in the
causal efficacy of consciousness.

Kim maintains that Hill and I really seem to be
offering a simplicity argument for physicalism.
He is led to say this, I think, because of the
connection between simplicity and theoretical
unification.

But I dont offer a simplicity argument for type
physicalism.
Nor does Hill.
Hill in fact spends several pages of Sensation
explaining why simplicity arguments for
physicalism wont succeed.
And Block and Stalnaker dont offer a causal
argument for physicalism.
We all offer best explanation arguments.

In any case, Kim thinks that the causal argument
is the strongest of the three argumentsthe best
explanation argument, the simplicity argument,
and the causal argumentfor physicalism, but that
it is nevertheless inconclusive.

The causal argument is indeed a strong
argument.
It is deeply intuitive that consciousness has
causal efficacy.

I agree with Kim, though, that the causal
argument is not absolutely conclusive.
While I think epiphenomenalism is absurd, I dont
think it is incoherent.
Nor is the idea of fundamental downward causal
overdetermination of physical events by mental
events though it too is absurd.

In any case, neither simplicity arguments nor
causal arguments are stronger than best
explanation arguments.
The reason is that best explanation arguments
take into account considerations of simplicity
and causal considerations.
Best explanation arguments for physicalism take
full advantage of such considerations.

The central premise of best explanation arguments
for physicalism is that physicalism offers the
best explanation of the correlationsbest on
grounds of overall coherence and global
theoretical simplicity.

If one explanation of psychophysical correlations
coheres with our common sense view that conscious
states are causally efficacious and another
doesnt, that counts in favor of the explanation
that coheres with those viewsthat it so coheres
is a reason in its favor (a defeasible one, of
course).
Thus, considerations of causal efficacy come into
play in answering the question of whether
psycho-physical identities offer the best
explanations of psychophysical correlations and
so do considerations of simplicity relative to
our total theory.

Contra Kim, neither causal arguments nor
simplicity arguments are superior to a best
explanation argument.

86
Kim Identities Arent Explanatory

Kims main objection to explanatory arguments for
physicalism is that identities dont explain
anything.

If identity claims dont explain, then, of
course, we cannot infer them as the best
explanation of anything.
So Hill, Block, Stalnaker, and I are all
committed to an incoherent position.

Kim claims that identity claims function as
rewrite rules that merely transmit
explanations.

Talk of rewrite rules suggests definitions,
analyticity, and a priority.
But the relevant identities are a posteriori.

Kim maintains, however, that even such a
posteriori identity claims as Water is H20,
Light is electromagnetic radiation, and Heat
is molecular kinetic energy function as mere
rewrite rules in inferential contexts they
generate no explanatory connections between the
explanandum and the phenomena invoked in the
explanans they seem to to have explanatory
efficacy of their own.

Immediately after claiming that a posteriori
identities function merely as rewrite rules and
so have no explanatory efficacy of their own,
however, Kim says
Here I am speaking only of necessary identities
with rigid designators, not contingent identities
likeJerry is the fastest talker among
philosophers, which seem eminently explanatory.
(I dont think Kim means to distinguish necessary
from contingent identity, but only necessary
statements of identity from contingent ones.)

92
Some Questions for Kim

Jerry is the actual fastest talker among
philosophers is a necessary statement of
identity.
The locution actually, like in fact,
rigidifies an expression.
Could Jerry is the faster talker among
philosophers be explanatory, while Jerry is the
actual faster talker among philosophers not be?
Could Water is the watery stuff be explanatory,
while Water is the actual watery stuff not be?

While Jerry is the fastest talker among
philosophers and Jerry is the actual fastest
talker among philosophers are not
intersubstitutable is modal contexts, it is a
priori that the one is true when and only when
the other is.
If, as Kim holds, facts are the relata of the
explanation relation, is the fact that makes
Jerry is the fastest talker among philosophers
true different from the fact that makes Jerry is
the actual faster talker among philosophers true?

In any case, while a posteriori, the identity
claims in question are indeed metaphysically
necessary.
Thus, if premises P1Pn together with an identity
claim imply E, then P1Pn alone imply E.

But that does not make the identity claim
explanatorily idle.
It can figure essentially in the explanation of
E.
Explanations in physical science often include
mathematical truths.
Despite the fact that mathematical truths are
necessary truths, remove them from the deduction,
and one loses the explanation.

96
The Epistemic Dimension of Explanation

While it is controversial what the epistemic
conditions for explanation are, it should, I
think, be uncontroversial that there are such
conditions.
A posteriori identity claims can figure in
explanations in such a way that were they
removed, the epistemic conditions would fail to
be met.
They thus have an essential role in explanations.

Ive already mentioned the explanatory role of
Light is electromagnetism in Maxwells theory.
That water H20, and that salt NaCl (sodium
chloride) figure essentially in explanations of
why salt dissolves in water.

Consider, moreover, Kims third example of an a
posteriori identity that functions as a mere
rewrite rule, namely, Heat is molecular kinetic
energy.

The explanation of Boyles Law provided by
statistical mechanics makes essential use of this
identity.
Boyles Law states PV kT (pressure x volume
(a constant) k x temperature.

100

The explanation of this law in statistical
mechanics requires the hypotheses that heat is
molecular energy, and that pressure is force per
unit area.
The claim that heat is molecular energy plays an
essential role in the explanation.

101

Why, then, does Kim hold that identities function
only as rewrite rules that transmit explanations,
and so have no explanatory efficacy of their own?

102
Kims Metaphysical Picture

I think that Kims view is motivated by a certain
metaphysical picture he has of explanation.
The picture is this explanation is a relation,
and facts are the relata of the relation, where
facts are obtaining states of affairs.
The relevant notion of fact here is a thick or
coarse-grained one.

103

Kim tells us that Water freezes at 32 degrees
Fahrenheit says no more than H20 freezes at 32
degrees Fahrenheit.
Similarly, he says that if pain CFS, then Pain
occurs iff CFS occurs would say no more than
CFS occurs iff CFS occurs.
By says no more than he seems to mean
expresses the same fact as.

104

Notice that on this view, when Maxwell said that
light waves are electromagnetic waves, he was
saying no more than that light waves are light
waves.
Hardly a bold conjecture!

105
Hyperintensionality

I do not think explanation is a relation that one
fact bears to another in Kims coarse-grained
sense of fact.
Like understanding and knowledge, explanation has
a coarse-grained factual dimension.
But like knows why and understands why, the
linguistic context explains why is
hyperintensional.

106

Facts, in Kims sense, would explain other facts
only under descriptions or conceptualizations,
and so only under modes of presentation.
The modes of presentation are essential to a
deductions satisfying the epistemic conditions
required for being an explanation.

107

Only by taking into account the epistemic
dimension of explanation can we capture the idea
that explanations provide understanding, and give
reasons for belief.
An identity statement can be essential to the
epistemic dimension of explanation, can be
required for a deduction to meet the epistemic
conditions for being an explanation.
That is explanatory role enough.

108
The Explanatory Gap

Speaking of the recent resurgence of type
physicalism, Kim says
The central idea is that the explanatory gap
between consciousness and the brain could be
closed if we could identity states of
consciousness with states of the brain.
And he says in another place, If we can identify
pain with CFS there can be no gap between the
two its only when we allow them to be diverse,
the issue of an explanatory gap can arise. These
questions, and the gap, will remain to haunt us
if we refuse to acknowledge the identity of pain
with CFS.

109

I can see why Kim thinks this.
Given his view of his view of explanation as a
relation among (coarse-grained) facts, if states
of consciousness are identical with types of
physical states, then physical theory leaves no
fact unexplained (not even facts about what it is
like to have certain experiencesabout the
phenomenal characters of experiences).

110

While I am a type-physicalist about states of
phenomenal consciousness, I dont believe that
psycho-physical type identities will close the
explanatory gap.
The identity statements in question wont yield
the sort of explanation that proponents of the
gap are denying are possible.

111
Reduction via Functional Analysis

The kind of explanation in question is reductive
explanation via functional analysis.
Such explanations come in two varieties since a
functional analysis can be meaning giving or only
reference fixing.
A meaning giving functional analysis of a concept
would be an a priori definition of the concept in
physical and/or topic neutral terms.
A reference fixing functional analysis would
state an a priori physical and/or topic neutral
reference-fixing condition for the concept.

112

As proponents of the gap in effect argue,
phenomenal concepts lack either meaning giving or
reference fixing analyses, and so facts about the
phenomenal characters of experiences cannot be
reductively explained via functional analysis.
They are right this explanatory gap cannot be
closed.

113
Conceptual Dualism Without Property Dualism

But the fact that this explanatory gap cannot be
closed does not entail property dualism.
Physicalists can explain why the gap cannot be
closed in exactly the way that dualists explain
it phenomenal concepts lack either meaning
giving or reference fixing functional analyses.

114

Dualists go on to make an additional claim,
namely that conscious properties are distinct
from physical properties.
But property dualism isnt required to explain
why the gap cannot be closed.
An explanation in terms of the differences in the
conceptual roles of phenomenal concepts and
physical concepts will do.

Write a Comment

User Comments (0)