[Wittrs] Further Thoughts on Dennett, Searle and the Conundrum of Dualism

  • From: "SWM" <SWMirsky@xxxxxxx>
  • To: wittrsamr@xxxxxxxxxxxxx
  • Date: Fri, 26 Mar 2010 20:28:16 -0000

When Gordon said:

"I want to know if Dennett believes the CRA makes sense only if one considers 
mental phenomena as evidence of NON-physical substances of properties. Does he?

I replied:

"My view is yes, based on his text though, again, he does not lay out a formal 
argument as I have done. But I think the text of his informal argument pretty 
much makes the same point."


I was probably responding too breezily so let me add this:

I think Dennett shares the view I hold that, for the CR to be evidence that 
leads to the conclusion of the CRA, that computers can't produce consciousness, 
one has to think that consciousness is irreducible (which means irreducible to 
anything else that is part of the physical universe). Dennett does not use 
"irreducible" (at least I don't think he does) but asks, rather, why is it that 
the CR seems to suggest to us that processes of the sort found in the CR can't 
produce consciousness rather than that the CR just may not be up to the task?

He answers his own question by reminding us that what's going on in the CR not 
only doesn't look like real understanding (it's just rote symbol manipulation) 
but it isn't performing any of the many other things our brains do when we have 
an instance of understanding in ourselves. This boils down to looking at what 
it means when we have an instance of understanding. Dennett puts it this way: 
for the CR to understand it would need to be able to make many, many 
connections at many levels. It would need knowledge of lots of things including 
the world, the relations of symbols to what is symbolized, itself, purposes, 
others' behavior and so forth. Every instance of understanding we have reflects 
just this vast array, this network of associations.

But Searle's CR has none of that. What it has is a processor taking symbol X, 
matching it to symbol Y and outputting symbol Y.

Dennett's point is that this is not how our own understanding works so why 
should we expect the CR to have understanding, absent the complexity of 
information processing he describes? So the idea that "more of the same" can't 
do it because it's just "more of the same" looks convincing but Dennett asks 
why should it be?

It seems to convince until you look more closely at what we mean by 
understanding and ask why couldn't a machine, running more processes of the 
type described by Dennett, succeed where the CR fails while not being 
qualitatively different in terms of its constituents from the CR -- and then it 
no longer seems quite so self-evident.

More importantly perhaps, look at the phrase "more of the same". Implicit in 
that idea is that all the understanding is packed into each and every instance 
of the CR processing an inputted symbol. That implies that the understanding 
"property" is attached to the process itself, an aspect of it, one of those 
mental properties dualists tell us about which are ontologically irreducible. 
It's just that some physical processes have them and some don't, sort of like 
saying some processes make a lot of noise while others are silent, some involve 
flashing lights and some don't, etc.

Now, if we look closely, we see we can break every process down to its 
constituents, too. So there's really nothing like an atomic process just as 
there aren't atomic facts or atomic propositions. Even a process is a system of 
sorts and every property of any process is ascribable to other phenomena that 
underlie and constitute it. In the same sense, I think Dennett is telling us 
that, not only can there be no process property per se (they are ALL system 
properties) but, if that is the case, then we have to think of the world and 
understanding in system terms. And, in that case, if everything is a system, 
and understanding is similarly a system property, then there is no reason why a 
property of a larger system must already be found in its constituent 
processes/systems. Some things require more complexity in organization than 
other things.

Is this a claim that takes "mental phenomena as evidence of NON-physical 
substances of properties" as you put it (I assume you meant "or" rather than 
that second "of")?

Because of my points elsewhere about "physical/non-physical" I'm not sure how 
to answer this. But note that the issue of the CR is whether computational 
processes running on computers can suffice to produce consciousness so the 
question is whether one physical platform can do what another physical platform 
(brains) is known to be able to do. We know Searle sees mental events as just 
the other side of the coin of physical events, not as separate phenomena in 
themselves. Yet, his CRA, in order to reach a conclusion that is generalizable 
to all potential R's from the example of the CR, must do something Searle isn't 
willing to do when he talks about brains and that is to say that computational 
platforms can't do what brain platforms can because, when viewed in isolation, 
their processes evidence no understanding.

So Searle, who insists that physical platforms are the sole causes of minds in 
the world through brains, at the least, is in the strange position of saying 
that another physical platform, computers can't do what the brain does because 
its parts aren't themselves capable of understanding anything. But if you look 
at the individual parts of the brain, they aren't either.

So on Dennett's view, as I understand and interpret it, (and on my view though 
that's less relevant here since you have said you want to address Dennett's 
position) Searle falls into dualism with the CRA even though he avoids it with 
his explicit statements about brains. Because he doesn't explicitly acknowledge 
any of this, Searle can be taken as both affirming and denying that 
consciousness is non-physical, something I could agree with if he meant 
"physical" differently in his affirmation and denial. (I would say that mental 
phenomena are physically derived but not physical entities.) But Searle can't 
do this with regard to the CR if the CRA is to reach the conclusion he asserts 
for it.

SWM

=========================================
Need Something? Check here: http://ludwig.squarespace.com/wittrslinks/

Other related posts: