THE IDENTITY THEORY

The
Identity
Theory
is
the
hypothesis
that
-not
necessarily
but
as
a
matter
of
fact-every
experience
is
identical
with
some
physical
state.
Specifically
with
some
neurochemical
state.
I
contend
that
we
who
accept
the
material-
istic
working
hypothesis
that
physical
phenomena
have
none
but
purely
physical
explanations
must
accept
the
identity
theory.
This
is
to
say
more
than
do
most
friends
of
the
theory,
who
say
only
that
we
are
free
to
accept
it,
and
should
for
the
sake
of
some
sort
of
economy
or
elegance.
I
do
not
need
to
make
a
case
for
the
identity
theory
on
grounds
of
economy
since
I
believe
it
can
and
should
rest
on
a
stronger
foundation.
My
argument
is
this:
The
definitive
characteristic
of
any
(sort
of)
experience
as
such
is
its
causal
role,
its
syndrome
of
most
typical
causes
and
effects.
But
we
materialists
believe
that
these
causal
roles
which
belong
by
analytic
necessity
to
experiences
be-
long
in
fact
to
certain
physical
states.
Since
those
physical
states
possess
the
definitive
characteristics
of
experience,
they
must
be
the
experiences.
My
argument
parallels
an
argument
which
we
will
find
un-
controversial.
Consider
cylindrical
combination
locks
for
bicycle
chains.
The
definitive
characteristic
of
their
state
of
being
un-
locked
is
the
causal
role
of
that
state,
the
syndrome
of
its
most
typical
causes
and
effects:
namely,
that
setting
the
combination
typically
causes
the
lock
to
be
unlocked
and
that
being
unlocked
1
Experiences
herein
are
to
be
taken
in
general
as
universals,
not
as
ab-
stract
particulars.
2
States
also
are
to
be
taken
in
general
as
universals.
I
shall
not
dis-
tinguish
between
processes,
events,
phenomena,
and
states
in
a
strict
sense.
3
I
am
therefore
invulnerable
to
Brandt's
objection
that
the
identity
theory
is
not
clearly
more
economical
than
a
certain
kind
of
dualism.
Doubts
about
the
typically
causes
the
lock
to
open
when
gently
pulled.
That
is
all
we
need
know
in
order
to
ascribe
to
the
lock
the
state
of
being
or
of
not
being
unlocked.
But
we
may
learn
that,
as
a
matter
of
fact,
the
lock
contains
a
row
of
slotted
discs;
setting
the
combination
typically
causes
the
slots
to
be
aligned;
and
alignment
of
the
slots
typically
causes
the
lock
to
open
when
gently
pulled.
So
alignment
of
slots
occupies
precisely
the
causal
role
that
we
ascribed
to
being
unlocked
by
analytic
necessity,
as
the
definitive
characteristic
of
being
unlocked
(for
these
locks).
Therefore
alignment
of
slots
is
identical
with
being
unlocked
(for
these
locks).
They
are
one
and
the
same
state.

We
must
understand
that
the
identity
theory
asserts
that
certain
physical
states
are
experiences,
introspectible
processes
or
activities,
not
that
they
are
the
supposed
intentional
objects
that
experiences
are
experiences
of.
If
these
objects
of
experience
really
exist
separate
from
experiences
of
them,
or
even
as
abstract
parts
thereof,
they
may
well
also
be
something
physical.
Perhaps
they
are
also
neural,
or
perhaps
they
are
abstract
constituents
of
veridically
perceived
surroundings,
or
perhaps
they
are
something
else,
or
nothing
at
all;
but
that
is
another
story.
So
I
am
not
claiming
that
an
experience
of
seeing
red,
say,
is
itself
somehow
a
red
neural
state.
Shaffer
has
argued
that
the
identity
theory
is
impossible
be-
cause
(abstract
particular)
experiences
are,
by
analytic
necessity,
unlocated,
whereas
the
(abstract
particular)
neural
events
that
they
supposedly
are
have
a
location
in
part
of
the
subject's
ner-
vous
system.4
But
I
see
no
reason
to
believe
that
the
principle
that
experiences
are
unlocated
enjoys
any
analytic,
or
other,
neces-
sity.
Rather
it
is
a
metaphysical
prejudice
which
has
no
claim
to
be
respected.
Or
if
there
is,
after
all,
a
way
in
which
it
is
analytic
that
experiences
are
unlocated,
that
way
is
irrelevant:
perhaps
in
our
presystematic
thought
we
regard
only
concreta
as
located
in
a
primary
sense,
and
abstracta
as
located
in
a
merely
derivative
sense
by
their
inherence
in
located
conereta.
But
this
possible
source
of
analytic
unlocatedness
for
experiences
does
not
meet
the
needs
of
Shaffer
's
argument.
For
neural
events
are
abstracta
too.
Whatever
unlocatedness
accrues
to
experiences
not
because
they
are
mental
but
because
they
are
abstract
must
accrue
as
much
to
neural
events.
So
it
does
not
discriminate
between
the
two.

The
identity
theory
says
that
experience-aseriptions
have
the
same
reference
as
certain
neural-state-ascriptions:
both
alike
refer
to
the
neural
states
which
are
experiences.
It
does
not
say
that
these
ascriptions
have
the
same
sense.
They
do
not;
experience-
ascriptions
refer
to
a
state
by
specifying
the
causal
role
that
be-
longs
to
it
accidentally,
in
virtue
of
causal
laws,
whereas
neural-
state-aseriptions
refer
to
a
state
by
describing
it
in
detail.
There-
fore
the
identity
theory
does
not
imply
that
whatever
is
true
of
experiences
as
such
is
likewise
true
of
neural
states
as
such,
nor
conversely.
For
a
truth
about
things
of
any
kind
as
such
is
about
things
of
that
kind
not
by
themselves,
but
together
with
the
sense
of
expressions
by
which
they
are
referred
to
as
things
of
that
kind.5
So
it
is
pointless
to
exhibit
various
discrepancies
between
what
is
true
of
experiences
as
such
and
what
is
true
of
neural
states
as
such.
We
can
explain
those
discrepancies
without
denying
psycho-
physical
identity
and
without
admitting
that
it
is
somehow
identity
of
a
defective
sort.
We
must
not
identify
an
experience
itself
with
the
attribute
that
is
predicated
of
somebody
by
saying
that
he
is
having
that
experience.
The
former
is
whatever
state
it
is
that
occupies
a
certain
definitive
causal
role;
the
latter
is
the
attribute
of
being
in
whatever
state
it
is
that
occupies
that
causal
role.
By
this
distinc-
tion
we
can
answer
the
objection
that,
since
experience-ascriptions
and
neural-state-descriptions
are
admittedly
never
synonymous
and
since
attributes
are
identical
just
in
case
they
are
predicated
by
synonymous
expressions,
therefore
experiences
and
neural
states
cannot
be
identical
attributes.
The
objection
does
establish
a
non-
identity,
but
not
between
experiences
and
neural
states.
(It
is
un-
fair
to
blame
the
identity
theory
for
needing
the
protection
of
so
suspiciously
subtle
a
distinction,
for
a
parallel
distinction
is
needed
elsewhere.
Blue
is,
for
instance,
the
color
of
my
socks,
but
blue
is
not
the
attribute
predicated
of
things
by
saying
they
are
the
color
of
my
socks,
since;
.
.
is
blue'
and
'.
.
.
is
the
color
of
my
socks'
are
not
synonymous.)

The
first
of
my
two
premises
for
establishing
the
identity
theory
is
the
principle
that
the
definitive
characteristic
of
any
experience
as
such
is
its
causal
role.

By
analytic
necessity
these
conditions
are
true
of
the
experience
and
jointly
distinctive
of
it.
My
first
premise
is
an
elaboration
and
generalization
of
Smart's
theory
that
avowals
of
experience
are,
in
effect,
of
the
form
'What
is
going
on
in
me
is
like
what
is
going
on
in
me
when
followed
by
specification
of
typical
stimuli
for,
or
responses
to,
the
ex-
perience.7
I
wish
to
add
explicitly
that
.
may
be
an
elaborate
logical
compound
of
clauses
if
necessary;
that
.
.
.
must
specify
typical
causes
or
effects
of
the
experience,
not
mere
accompani-
ments;
that
these
typical
causes
and
effects
may
include
other
ex-
periences;
and
that
the
formula
does
not
apply
only
to
first-person
reports
of
experience.
This
is
not
a
materialist
principle,
nor
does
it
ascribe
ma-
terialism
to
whoever
speaks
of
experiences.
Rather
it
is
an
ac-
count
of
the
parlance
common
to
all
who
believe
that
experiences
are
something
or
other
real
and
that
experiences
are
efficacious
outside
their
own
realm.
It
is
neutral
between
theories-or
a
lack
of
any
theory-about
what
sort
of
real
and
efficacious
things
experiences
are:
neural
states
or
the
like,
pulsations
of
ectoplasm
or
the
like,
or
just
experiences
and
nothing
else.
It
is
not
neutral,
however,
between
all
current
theories
of
mind
and
body.
Epi-
phenomenalist
and
parallelist
dualism
are
ruled
out
as
contra-
dictory
because
they
deny
the
efficacy
of
experience.
Behaviorism
as
a
thoroughgoing
dispositional
analysis
of
all
mental
states,
in-
eluding
experiences,8
is
likewise
ruled
out
as
denying
the
reality
inad
a
fortiori
the
efficacy
of
experiences.
For
a
pure
disposition
is
a
fictitious
entity.
The
expressions
that
ostensibly
denote
dispo-
sitions
are
best
construed
as
syncategorematic
parts
of
statements
of
the
lawlike
regularities
in
which
(as
we
say)
the
dispositions
are
manifest.
Yet
the
principle
that
experiences
are
defined
by
their
causal
6
It
would
do
no
harm
to
allow
the
set
of
conditions
to
be
infinite,
so
long
as
it
is
recursive.
But
I
doubt
the
need
for
this
relaxation.

Smart
's
concession
that
his
formula
does
lnot
really
translate
avowals
is
unnecessary.
It
results
from
a
bad
example:
'I
have
a
pain'
is
not
trans-
latable
as
'What
is
going
on
in
me
is
like
what
goes
on
when
a
pin
is
stuck
into
me',
because
the
concept
of
pain
might
be
introduced
without
mention
of
pins.
Indeed;
but
the
objection
is
no
good
against
the
tranLslation
'What
is
going
on
in
me
is
like
what
goes
on
when
(i.e.
when
and
because)
my
skiin
is
damaged
'.
8
Any
theory
of
mind
and
body
is
compatible
with
a
dispositional
analysis
of
mental
states
other
than
experiences
or
with
so-callel
"methodological
behaviorism.

The roles
is
itself
behaviorist
in
origin,
in
that
it
inherits
the
be-
haviorist
discovery
that
the
(ostensibly)
causal
connections
be-
tween
an
experience
and
its
typical
occasions
and
manifestations
somehow
contain
a
component
of
analytic
necessity.
But
my
principle
improves
on
the
original
behavioristic
embodiment
of
that
discovery
in
several
ways:
First,
it
allows
experiences
to
be
something
real
and
so
to
be
the
effects
of
their
occasions
and
the
causes
of
their
manifestations,
as
common
opinion
supposes
them
to
be.
Second,
it
allows
us
to
include
other
experiences
among
the
typical
causes
and
effects
by
which
an
experience
is
defined.
It
is
crucial
that
we
should
be
able
to
do
so
in
order
that
we
may
do
justice,
in
defining
experiences
by
their
causal
roles,
to
the
introspective
accessibility
which
is
such
an
important
feature
of
any
experience.
For
the
introspective
accessibility
of
an
experi-
ence
is
its
propensity
reliably
to
cause
other
(future
or
simultane-
ous)
experiences
directed
intentionally
upon
it,
wherein
we
are
aware
of
it.
The
requisite
freedom
to
interdefine
experienes
is
not
available
in
general
under
behaviorism;
interdefinition
of
experi-
ences
is
permissible
only
if
it
can
in
principle
be
eliminated,
which
is
so
only
if
it
happens
to
be
possible
to
arrange
experiences
in
a
hierarchy
of
definitional
priority.
We,
on
the
other
hand,
may
allow
interdefinition
with
no
such
constraint.
We
may
expect
to
get
mutually
interdefined
families
of
experiences,
but
they
will
do
us
no
harm.
There
will
be
no
reason
to
identify
anything
with
one
experience
in
such
a
family
without
regard
to
the
others
-but
why
should
there
be
?
Whatever
occupies
the
definitive
causal
role
of
an
experience
in
such
a
family
does
so
by
virtue
of
its
own
membership
in
a
causal
isomorph
of
the
family
of
ex-
periences,
that
is,
in
a
system
of
states
having
the
same
pat-
tern
of
causal
connections
with
one
another
and
the
same
causal
connections
with
states
outside
the
family,
viz.,
stimuli
and
be-
havior.
The
isomorphism
guarantees
that
if
the
family
is
identi-
fied
throughout
with
its
isomorph
then
the
experiences
in
the
family
will
have
their
definitive
causal
roles.
So,
ipso
facto,
the
isomorphism
requires
us
to
accept
the
identity
of
all
the
experi-
ences
of
the
family
with
their
counterparts
in
the
causal
isomorph
of
the
family.
Putnam
discusses
an
analogous
case
for
maehines:
a
family
of
("
logi-
cal"
or
"functional")
states
defined
by
their
causal
roles
and
mutually
interdefined,
and
a
causally
isomorphic
system
of
("structural")
states
otherwise
defined.
He
does
not
equate
the
correlated
logical
and
structural
states.
"Minds
and
Machines,"
in
Dimensions
of
Mind.

Third.
we
are
not
obliged
to
define
an
experience
by
the
causes
and
effects
of
exactly
all
and
only
its
occurrences.
We
can
be
content
rather
merely
to
identify
the
experience
as
that
state
which
is
typically
caused
in
thus-and-such
ways
and
typically
causes
thus-and-such
effects,
saying
nothing
about
its
causes
and
effects
in
a
(small)
residue
of
exceptional
cases.
A
definition
by
causes
and
effects
in
typical
cases
suffices
to
determine
what
the
experience
is,
and
the
fact
that
the
experience
has
some
character-
istics
or
other
besides
its
definitive
causal
role
confers
a
sense
upon
aseriptions
of
it
in
some
exceptional
cases
for
which
its
definitive
typical
causes
and
effects
are
absent
(and
likewise
upon
denials
of
it
in
some
cases
for
which
they
are
present).
Be-
haviorism
does
not
acknowledge
the
fact
that
the
experience
is
something
apart
from
its
definitive
occasions
and
manifestations,
and
so
must
require
that
the
experience
be
defined
by
a
strictly
necessary
and
sufficient
condition
in
terms
of
them.
Otherwise
the
behaviorist
has
merely
a
partial
explication
of
the
experience
by
criteria,
which
can
never
give
more
than
a
presumption
that
the
experience
is
present
or
absent,
no
matter
how
much
we
know
about
the
subject's
behavior
and
any
lawlike
regularities
that
may
govern
it.
Relaxation
of
the
requirement
for
a
strictly
necessary
and
sufficient
condition
is
welcome.
As
anybody
who
has
tried
to
implement
behaviorism
knows,
it
is
usually
easy
to
find
con-
ditions
which
are
almost
necessary
and
sufficient
for
an
experience.
All
the
work-and
all
the
complexity
which
renders
it
incredible
that
the
conditions
found
should
be
known
implicitly
by
every
speaker-comes
in
trying
to
cover
a
few
exceptional
cases.
In
fact,
it
is
just
impossible
to
cover
some
atypical
cases
of
experi-
ences
behavioristically:
the
case
of
a
perfect
actor
pretending
to
have
an
experience
he
does
not
really
have;
and
the
case
of
a
total
paralytic
who
cannot
manifest
any
experience
he
does
have
(both
cases
under
the
stipulation
that
the
pretense
or
paralysis
will
last
for
the
rest
of
the
subject's
life
no
matter
what
happens,
in
virtue
of
regularities
just
as
lawlike
as
those
by
which
the
behaviorist
seeks
to
define
experiences).
It
is
possible,
and
probably
good
analytic
strategy,
to
re-
construe
any
supposed
pure
dispositional
state
rather
as
a
state
defined
by
its
causal
role.
The
advantages
in
general
are
those
we
have
seen
in
this
case:
the
state
becomes
recognized
as
real
and
efficacious;
unrestricted
mutual
interdefinition
of
the
state
and
others
of
its
sort
becomes
permissible
and
it
becomes
intelligible
that
the
state
may
sometimes
occur
despite
prevention
of
its
de-
finitive
manifestations.
I
do
not
offer
to
prove
my
principle
that
the
definitive
char-
acteristics
of
experiences
as
such
are
their
causal
roles.
It
would
be
verified
by
exhibition
of
many
suitable
analytic
statements
saying
that
various
experiences
typically
have
thus-and-such
causes
and
effects.
Many
of
these
statements
have
been
collected
by
behaviorists;
I
inherit
these
although
I
explain
their
status
some-
what
differently.
Behaviorism
is
widely
accepted.
I
am
content
to
rest
my
case
on
the
argument
that
my
principle
can
accommo-
date
what
is
true
in
behaviorism
and
can
escape
attendant
diffi-
culties.

My
second
premise
is
the
plausible
hypothesis
that
there
is
some
unified
body
of
scientific
theories,
of
the
sort
we
now
accept,
which
together
provide
a
true
and
exhaustive
account
of
all
physi-
cal
phenomena.

D.L.