The explanatory gap, naive realism and levels of explanation
Michael Schmitz
interesting thread and commenting on my paper (so all the advertising I
did finally paid off).
Peter Shelton has moved the discussion into a very interesting direction
by rejecting attempts to find a new kind of explanation to close the
so-called explanatory gap and recommending instead that:
"The way to bridge the gap is to look at it
differently, which is where a philosophical perspective can contribute
to what is essentially a scientific enterprise."
With reference to Heidegger's notion of in-der-Welt-sein, he suggests
that it may be helpful here to be a direct perceptual realist. I
agree, but I also concur with Anders Weinstein that later Peter spoils
his point by saying things that seem inconsistent with this realism,
specifically when he states that:
"We experience the world,not necessarily
as it is "in itself",but as presented by the faculties and dependent on
the range of wavelengths transduced by eyes and ears..."
This reintroduces the physicalistic outlook that (I happily note) there
seemed to be an emerging consensus we want to get rid of: 'Out there' we
only find the wavelengths and other entities that physics proper deals
with. Ultimately it's all only atoms (or nowadays quarks, spin and charm
and what have you) in the void. Colors, sounds, people, chairs and all
the other familiar entities of our daily lives don't exist at all. There
are only the (illusory) appearances of such (non-existent) things, these
mere manifestations of our subjectivity.
If on the other hand we really stick to naive realism, we can overcome
the idea of the explanatory gap (or the hard problem) in the way that
Peter Shelton recommended, by changing our perspective on it rather than
by proposing a new type of explanation. For if we unambiguously accept
the reality of say, colors, we can see immediately that there is nothing
special about the alleged special gap between consciousness and
neurophysiology. There is also a 'gap' between colors and wavelengths,
or a 'hard problem' of why there is any connection between these two at
all; there are corresponding gaps and problems concerning heat, sounds
and in the end concerning all the more specific entities whose
conceptions involves notions relating to these properties.
The point is, however, once again spoilt, if, like Anders goes on to
suggest, we try to understand this situation in terms of a distinction
between type and token identity. This is another residue of physicalism.
Here I agree with Alfredo Pereira that "we are not talking about
identity in any sense of the word". Even though the token identity
theory has been proposed by a currently rather famous philosopher
(Davidson), it seems to be based on a quite simple error, a
misunderstanding of the relation between types and tokens. For it is
simply out of the question that one token should be an instance of
several types. A token of the type "brain-state" is a single brain state
and a token of the type "state of consciousness" a single, specific
states of consciousness. Tokens are tokens *of their types*, not 'bare
particulars' somehow neutral with regard to any sort of typing. (This
mistake is, however, very common.)
(Those interested in the issue of externalism also raised by Anders
Weinstein may want to look at the end of my posting in Psyche-B in the
"necessity vs. sufficiency thread.)
With regard to my earlier charge of dualism against Alfredo Pereira, I
must say that I may have misunderstood him. I thought he wanted to say
that the relevant neurophysiological entity would be a triggering cause
in the sense that it would trigger a spatiotemporally distinct conscious
event. But probably he only meant that while there may be a type of
neurophysiological state underlying conscious states which is not
spatiotemporally distinct from the latter - that for me is the crucial
point - it is, at this point, not a suitable task for empirical science
to search for such a state, that is, for the NCC. If that's the case, I
take back my accusation of dualism. With regard to the substantive (or
to some extent verbal) issue involved, it still seems to me that such a
state is what is naturally meant by the term "NCC" , and I'm also
prepared to bet that at least some scientists are - perhaps only quietly
and secretly - thinking about what the NCC in my sense might be. (For
more on this let me once again refer you to my posting in Psyche-B and
the postings by David Chalmers and John McCrone.)
Alfredo also criticizes my paper:
"Michael Schmitz' paper "Consciousness and Identification" (website
address
shown in his posting) is based on assumptions implying that multiscaling
phenomena are completely subjective (Frege's "Sinn") or completely
objective
(Frege's "reference"). As multiscaling in the mind/brain is NOT
subjective,
he concludes that it must be completely objective, and charges it as
dualism"
I'm not sure what Alfredo is driving at here, especially since in the
paper the point I make concerning Frege's notion of sense is a criticism
of the idea that Fregean senses or modes of access, that is ways of
identifying an object, could be merely subjectively different, such that
the difference would be unconnected to any differences on the objective
side of objects known. That is the case according to many materialists
who try to defend the identity theory by saying that we have a
1st-person (consciousness) and a third person (physiological) access to
what is still ontologically the same entity. But this, I argue among
other things, turns the identical object allegedly known in different
ways, into an unknowable third. This argument, it seems to me, goes into
a direction quite opposite to where Alfredo says it is going, by
stressing that subjective and objective notion are inextricably linked.
But perhaps this again is a misunderstanding on my side which Alfredo
can clear up when he starts using his computer again.
Michael Schmitz
mi...@zedat.fu-berlin.de
My paper "Consciousness and Identification" can be found under:
http://lgxserver.uniba.it/lei/MINDE/repbibl/t0000003.htm
William Seager
"...it is simply out of the question that one token
should be an instance of several types."
go by. Token are particulars which fall under
types. Thus my dog, rex, is a particular thing
which falls under many types, among them,
*dog*, *bull terrier*, *friendly*, *weighs
75 pounds*. Contrary to the above quote,
it is simply out of the questions that one
token should be an instance of only *one*
type.
We should be clear about this.
The follow up remark that "tokens are tokens
*of their types* not 'bare particulars' ..." is irrelevant
to the question of whether a token can fall under
only one type -- which claim is simply incoherent.
That tokens are tokens of their types is tautological,
and the issue of bare particulars is a red herring
here.
_______________________________
William Seager, Philosophy
University of Toronto at Scarborough
sea...@scar.utoronto.ca
www.scar.utoronto.ca/~seager
Anders N Weinstein
Michael Schmitz <PSYC...@LISTSERV.UH.EDU> wrote:
>The point is, however, once again spoilt, if, like Anders goes on to
>suggest, we try to understand this situation in terms of a distinction
>between type and token identity. This is another residue of physicalism.
Actually, I am agnostic about the sensibility of claiming even a token
identity between intentional psychological events and physical events
in the brain. (Check out Haugeland's paper "Weak Supervenience"
reprinted in his anthology _Having Thought_ for some reasons for
skepticism here.) I simply concede the possibility of token event
identities to those who want it, since token identity without type
identity does not secure any interesting reduction of psychological
predicates to physicalistic ones.
>Here I agree with Alfredo Pereira that "we are not talking about
>identity in any sense of the word". Even though the token identity
>theory has been proposed by a currently rather famous philosopher
>(Davidson), it seems to be based on a quite simple error, a
>misunderstanding of the relation between types and tokens. For it is
>simply out of the question that one token should be an instance of
>several types. A token of the type "brain-state" is a single brain state
>and a token of the type "state of consciousness" a single, specific
>states of consciousness. Tokens are tokens *of their types*, not 'bare
>particulars' somehow neutral with regard to any sort of typing. (This
>mistake is, however, very common.)
You need to be clear on what sort of thing you are identifying.
Davidson's position concerns token *events*, construed as he does
as particulars which fall under multiple descriptions.
On this view, for example, perhaps the event e1 of a certain neuron n1
changing its rate of firing from 50 to 100 hz is the very same event as
the event e2 of John Smith's deciding to accept the job in Boston (e1 = e2).
Then any property of e1 is a property of e2 by Leibniz' law, for example
e1 is a deciding.
It seems to me such a token identity does nothing to illuminate
intentional explanation of action and so is not worth fighting for. But
the key point is that even if we concede such an identity to those who
want it, there is no reason to believe we can give an explanation of
the intentional properties of the event, i.e. not just what makes it a
*deciding* but alsow what determines its particular
content (viz., to accept some contextually determined job in Boston),
without adverting to relationships to a institutions and other stuff
outside the skull.
Alfredo Pereira Jr
>With regard to my earlier charge of dualism against Alfredo Pereira, I
>must say that I may have misunderstood him. I thought he wanted to say
>that the relevant neurophysiological entity would be a triggering cause
>in the sense that it would trigger a spatiotemporally distinct conscious
>event. But probably he only meant that while there may be a type of
>neurophysiological state underlying conscious states which is not
>spatiotemporally distinct from the latter - that for me is the crucial
>point - it is, at this point, not a suitable task for empirical science
>to search for such a state, that is, for the NCC. If that's the case, I
>take back my accusation of dualism.
> With regard to the substantive (or
>to some extent verbal) issue involved, it still seems to me that such a
>state is what is naturally meant by the term "NCC" , and I'm also
>prepared to bet that at least some scientists are - perhaps only quietly
>and secretly - thinking about what the NCC in my sense might be.
We are almost in agreement now. Unfortunately things are more complicated
than your paragraph describes. In fact, I believe there are "TWO levels of
NCCs": one that current neuroscience is able to discover ( a kind of
neuronal mechanism or process that directly triggers consciousness) and
another one that is not spatiotemporally distinct from the conscious process
but requires a level of explanation that current neuroscience is not
prepared to handle. We are in need of epistemological discussion and serious
interdisciplinary work on this subject.
>Alfredo also criticizes my paper:
>"Michael Schmitz' paper "Consciousness and Identification" .....
>I'm not sure what Alfredo is driving at here, especially since in the
>paper the point I make concerning Frege's notion of sense is a criticism
>of the idea that Fregean senses or modes of access, that is ways of
>identifying an object, could be merely subjectively different, such that
>the difference would be unconnected to any differences on the objective
>side of objects known.
>That is the case according to many materialists
>who try to defend the identity theory by saying that we have a
>1st-person (consciousness) and a third person (physiological) access to
>what is still ontologically the same entity.
Now I am confused about your proposal.
I can see two alternatives:
a) you understand that the Fregean notion of "sense" is not merely
subjective, and then criticizes materialist/identity philosophers who use it
for being inconsistent;
b) you understand that the Fregean notion is merely subjective, and then
criticizes materialist/identity philosophers because such a notion wouldn't
give enough support for the distinction between different levels of access
(first- and third-person levels).
>But this, I argue among
>other things, turns the identical object allegedly known in different
>ways, into an unknowable third. This argument, it seems to me, goes into
>a direction quite opposite to where Alfredo says it is going, by
>stressing that subjective and objective notion are inextricably linked.
>But perhaps this again is a misunderstanding on my side which Alfredo
>can clear up when he starts using his computer again.
Discussing the subject "levels of explanation", I said that in your paper
such levels were thought to be merely subjective or merely objective.
Possibly my mistake was that I understood that you were assuming the Fregean
framework as exaustive.
If you are moving in the direction of a Subject-Object Interactionist view,
then we are more in agreement than I realized before. However, I have many
difficulties with the terms used in contemporary philosophy of mind. I can't
understand how Heideggerian "being-in-the -world" can be identified with
Gibsonian "direct perceptual realism". J.J. Gibson is the anti-Kantian man,
while Heidegger comes from Husserlian phenomenology that comes from Kantian
phenomenalism. Interactionism is basically Piagetian epistemology, that
comes from a reinterpretation of Kant, where subjective and objective forms
co-evolve dialectically.
On the issue Internalism/Externalism I think both have a part of truth.
Internalism is correct in the sense that from neuroscientific evidence any
content of consciousness must be represented by the brain; Externalism is
correct in the sense that the contents of consciousness (with few
exceptions) are about the world external to the brain and from scientific
evidence the generation of such contents requires interaction with the
environment and reafference.
Best Regards,
Alfredo Pereira Jr.