Martine
Nida-Rümelin
Université
Fribourg (CH)
Konstanz,
20.4.2002
1. Sketch of the
Argument
general version
(P1) The
nature of every physical property can in principle be captured by some physical
concept.
(P2) The
nature of phenomenal properties can only be captured by pure phenomenal
concepts.
(P3) Pure
phenomenal concepts and physical concepts are distinct.
Therefore:
( C)
Phenomenal properties are non-physical properties.
particular version
(P1) The
nature of every physical property can
in principle be captured by some physical concept.
(P2' ) The
nature of phenomenal blueness (PBL)
can only be captured by the pure phenomenal concept of blueness (PPB).
(P3') The pure phenomenal concept of blueness is
not a physical concept.
Therefore:
(C')
Phenomenal blueness is not a physical property.
2. Translation of P1
in a two-dimensional framework
PH(P): P is a
physical property
PH(C): C is a
physical concept
RF(P): the
function from possible worlds to extensions that appropriately represents the
property P
SI(C): the
secondary intension of the concept C
PI(C): the
primary intension of the concept C
(P1*) "P (PH(P) ® $ C (PH(C) Ù RF(P)=SI(C) Ù SI(C)=PI(C)))
For every
physical property P there is a physical concept C that fulfills the following
conditions:
(a) the
primary and secondary intension of the concept C coincide
(b) the
secondary intension of the concept C is identical with the function that
represents the property P.
This
translation presupposes the following definition:
Definition (grasping
properties by concepts):
The nature of
the property P can be captured by the concept C iff
(a) the
concept C has identical primary and secondary intension and
(b) the
function that represents P is identical with the secondary intension of C.
3. Translation of P2'
in a two-dimensional framework
A more
explicit formulation of P2':
(P2') (a) The
nature of phenomenal blueness can be captured
by the pure phenomenal concept of
blueness.
(b) The nature of phenomenal blueness
cannot be captured by any physical concept.
For the
argument only (b) is needed.
Translation of
(P2'b):
PBL:
Phenomenal blueness
(P2'b*) Ø$ C (PH(C) Ù RF(PBL) =
SI(C) Ù PI(C) = SI(C))
There is no
physical concept that fulfills the following conditions:
(a) C has
identical primary and secondary intensions
(b) The
secondary intension of C is identical with the function that represents
phenomenal blueness.
The conjunction
of (P1*) and (P2'b*) implies
(C*) Ø PH(PBL)
Phenomenal
blueness is not a physical property.
Remark: Using
"PBL" or "phenomenal blueness" as a name of a property
presupposes the ontological assumption that there is such a property (that
there is a property corresponding to our pure phenomenal concept).
3. An Argument for P2'b* (for the claim that phenomenal blueness cannot be grasped by any physical concept)
(P2'b*) is
implied by the conjunction of the following claims:
(C1) the premiss of cognitive independence between physical concepts and pure phenomenal concepts
Ø$ C (PH(C) Ù PI(C) =
PI(PPB))
There is no
physical concept with a primary intension that is identical with the primary
intension of the pure phenomenal concept of blueness.
(C2) a conceptual claim concerning pure phenomenal concepts
PI(PPB) =
SI(PPB)
The pure
phenomenal concept of blueness has the same primary and secondary intension.
(C3) a weak ontological premiss (exclusion of eliminativism)
RF(PBL) =
SI(PPB)
There is a
property, namely phenomenal blueness (PBL), that corresponds to the concept of
phenomenal blueness.
4. Proof of the
second premiss (P2'b*) using (C1), (C2) and (C3)
Indirect
proof:
Assumption:
(P2'b*) is false: there is some physical concept C* with identical primary and secondary intension such that its
secondary intension is identical with the function that represents phenomenal
blueness. C* fulfills the following condition:
(1) PH(C*) Ù PI (C*) = SI
(C*) Ù SI(C*) = RF(PBL)
(2) PI
(C*) = SI(C*) = SI(PPB) = PI(PPB)
(1) (1) and (C3)
(C2)
but PI (C*) =
PI (PPB) contradicts (C1).
APPENDIX
A1. The Generalised Kaplan Paradigm
simplified
FA (
<w1,w2>) = ... (extension)
FA is the two-dimensional function that corresponds to the
concept A
w1: the world taken as actual
w2: the world taken as counterfactual
FA ( <w1,w2>) the
extension of the concept A in counterfactual circumstances w2 if w1 is the
actual world.
Definition 1: The primary
intension that corresponds to a concept A is a one-place function defined
as follows:
For all worlds w: PIA (w) = FA(<w,w>)
For every concept A and every possible world w, PIA(w)
is the individual (class) that would be the referent (the extension) of the
concept A in the actual world if w were the actual world.
wactual : the actual world
Definition 2 : The secondary
intension corresponding to a
concept is a one-place function defined as follows:
For all worlds w : SIA (w) = FA(<wactual,w>)
The secondary intension describes how the
referent of A depends on the characteristics of the considered counterfactual
world given that the actual world is as it is.
Definition 3: A concept is
a rigid
designator iff for every possible world w the following is true: the function that would be the secondary
intension of the concept if w were the actual world is constant.
Definition 4: A concept is actuality-independent
(absolute) iff the referent
does not depend on the world taken as actual, that is iff:
for all w1, w1', w2 : FA ( <w1,w2>) = FA (
<w1',w2>)
intuitive formulation: the referent of the concept in
counterfactual situations does not depend on eventually unknown features of the
actual world.
- candidates for actuality-independent concepts:
poison/
democratic state/ the president of the United States in 2001 in its non-rigid
reading (concepts
corresponding to definite descriptions in their non-rigid reading)
Definition 5: A concept is actuality-dependent
iff it is not actuality-independent.
actuality-dependent concepts: water, gold, tigers,
concepts corresponding to definite descriptions in their rigid interpretation.
If a concept is actuality-independent, then its primary and its secondary intension
coincide.
Definition 6: A concept is
an super-rigid designator iff the
corresponding two-dimensional function is a constant. In other words: iff the
concept is actuality-independent (absolute) and rigid.
If A is a property concept then there are two
alternative ways to represent the corresponding two-dimensional function:
(A1) the values of the function are classes of
individuals (or of tuples of individuals)
(A2) the values of the function are properties
A.2. The pure phenomenal concept of
phenomenal blueness
(a) Independently of any eventually unknown features of our world the
PPB always picks out the same quality B. Choosing alternative A2 the primary
intension is constant. The value is always the quality B.
(b) When considering counterfactual circumstances the PPB always picks
out the quality B, independently of the counterfactual circumstances considered
and independently of eventually unknown features of the actual world. Choosing
alternative A2, the corresponding
two-dimensional function is constant.
FPPB(<w1,w2>) = the quality B
In other words (and more generally):
Thesis : Pure phenomenal concepts are super-rigid designators. Any such concept
picks out a particular qualitative kind (the same in all counterfactual worlds
and independently of features of the actual world).
FPPB((<w1,w2>) = the class of pairs <s,t> such
that s has an experience of the quality B at t in w2.