Here is something that I have been thinking about, for the chapter I am revising at the moment:
In my thesis, I focus on these two formulations of the conceivability argument:
The Two-Dimensional Argument against physicalism
(1): P&~Q is conceivable
(
CP): If S is (ideally) conceivable, then S is primarily possible
(2): If P&~Q is conceivable, P&~Q is primarily possible
(3): If P&~Q is primarily possible, either P&~Q is secondarily possible or
Russellian (type-F) monism is true
(4): If P&~Q is secondarily possible,
Physicalism is false
-----------
(5) Either
Physicalism is false or
Russellian (type-F) monism is true
The Two-Dimensional Argument against Type-identities
(1) The identity-statement ‘Pain=C-fibre firing’ is a
posteriori(2) If an identity statement A=B is a
posteriori, then the terms A and B have different primary
intensions (
DPIM)
(3) Therefore , ‘pain’ and ‘C-fibre firing’ have different primary
intensions(4) The primary and secondary
intension of ‘pain’ coincide
(5) The primary and secondary
intension of ‘C-fibre firing’ coincide
(6) Therefore, ‘pain’ and ‘C-fibre firing’ have different secondary
intensions(7) If ‘pain’ and ‘C-fibre firing’ have different secondary
intensions, then ‘pain=C-fibre firing’ is false.
---------------
(8) ‘Pain=C-fibre firing’ is false.
My critique focuses on premises
CP (conceivability-possibility link) and
DPIM (distinct primary
intension model), as a good type-B materialist would have it.
In particular, I defend the following view concerning concepts:
(
Non-reductive Ascriptivism) For any concept C, there is an application conditional like this: (
AC): ‘If x is F, then x falls under C’, where for most concepts C from level n, feature F is described using same-level concepts.
If this view is correct, then it follows that sentences such as 'If
microphysical truths are so and so, then water is H20' will not be a
priori true, since we will not be able to deduce truths involving macroscopic concepts such as WATER from lower-level truths merely a
priori. Then, we can use such conditional as a counterexample to
CP, since it will plausibly be an a
posteriori but 1-necessary conditional (ignoring some complications involving 'totalities' and '
indexicalities'. See Chalmers & Jackson "Conceptual Analysis and Reductive Explanation", Philosophical Review 2001, for further discussion).
But on my view there is an asymmetry concerning
CP and
DPIM: notice that the view I call non-reductive
ascriptivism does not entail that
DPIM is false (at least, when we focus on standard sentences not involving phenomenal concepts). For standard cases of a
posteriori necessities still seem to fit
DPIM. Consider the following examples: 'Water = H20'; 'Heat= molecular motion'. For these cases of a
posteriori necessities, the different terms in each identity seem to be associated with different primary
intensions.
Then, it seems that a strategy against conceivability arguments that relies upon non-reductive
ascriptivism exclusively, still has to face the challenge posed by the two-dimensional argument against
identity theses above. True, it is widely accepted that
physicalism is not committed to an identity thesis. Still, such argument can prove to be lethal for many positive views about consciousness that want to identify phenomenal properties with some sort of complex physical-functional properties.
On my view, such theorists should appeal to the so-called
phenomenal concept strategy, or
exceptionalist strategy, according to which
DPIM is correct in general but not when it comes to phenomenal concepts.
This sort of strategy (non-reductive
ascriptivism plus phenomenal concept strategy) is what I call a
mixed strategy against conceivability arguments. Notice that, according to this view,
CP is incorrect in general, whereas
DPIM is incorrect only when it comes to phenomenal concepts. In my thesis, I try to defend something like that.
Any thoughts are more than welcome!
