From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!psych.toronto.edu!michael Mon May 25 14:05:24 EDT 1992
Article 5653 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!psych.toronto.edu!michael
>From: michael@psych.toronto.edu (Michael Gemar)
Subject: Re: Comments on Searle - What could causal powers be?
Organization: Department of Psychology, University of Toronto
References: <1992May11.202715.47273@spss.com> <1992May12.003415.7383@psych.toronto.edu> <1992May12.194417.31787@spss.com>
Message-ID: <1992May14.162236.21843@psych.toronto.edu>
Date: Thu, 14 May 1992 16:22:36 GMT

In article <1992May12.194417.31787@spss.com> markrose@spss.com (Mark Rosenfelder) writes:
>In article <1992May12.003415.7383@psych.toronto.edu> michael@psych.toronto.edu 
>(Michael Gemar) writes (quoting me):
>>>OK, let's say that the Bolivian economy, looked at in a certain way, can be
>>>seen to implement an algorithm that passes the Turing Test.  But surely
>>>there's nothing that *makes* it do that. 
>>
>>Sure there is, namely, the principles of economics, along with the initial
>>conditions and any "input" into the system.
>>
>>> All the actions and states which
>>>go to make up the Bolivian economy have a different explanation. 
>>
>>So do the activities of our brains.  They are (making the safe assumption
>>that you are a materialist) 
>
>Not a safe assumption, actually, but for the purposes of discussing AI I am.
>
>>purely explicable on the physical, or
>>biochemical, or neurological level.  No need for any of this spooky
>>cognitive stuff at all...
>
>Presumably cognitive events have neurochemical explanations.  There's nothing
>spooky here if you're travelling from the top down.  Similarly the economic
>transactions of Bolivia can be explained in terms of human psychology,
>the physics of the materials involved, etc.  

OK so far.

>The problem comes in going the other way.  Examining the neurochemistry of
>the brain, is there any way to tell that there's a cognitive level built on
>top of it which needs explaining?  Examining the economy of Bolivia, is
>there any way to tell whether there's a meaningful pattern built on top of
>it (e.g. an algorithm it happens to be implementing)? 
>
>These aren't rhetorical questions; I'd really like to know, if you have
>an answer.  On the other hand I hope you'll agree that there is in fact a
>cognitive level in the brain, but it is at best an open question whether
>there is an algorithmic level in the Bolivian economy.

I agree that there is a cognitive level, but only because I know through
introspection that *I* have one.  There is nothing (again, assuming materialism)
about my or anyone else's behaviour that is not explicable in terms of 
physical causes.  Note that I am not being an elimativist - I think that this
view is perfectly compatible with the view that cognitive processes are real
and play a causal role (in the same way that rivers are decribable purely
in terms of physical properties, and yet are also real geographical entities).
No, I *don't* think that there is necessarily anything in the neurochemistry
of any *particular* brain which begs for a cognitive explanation.  Rather, it
is the fact that brains with non-identical physical makeups nevertheless
exhibit similar behaviour that demands a level of explanation beyond the 
physical - at least, if you want to make generalizations regarding the
causes of behaviour.

  As far as Bolivia goes, I think that you are confusing levels when you equate
the cognitive level in people to the algorithmic in economies (or presumably 
computers).  As I understand it, for functionalism the algorithmic level
lies *below* the cognitive - it is what *causes* cognitive phenomena. 
To compare across entities at these different levels is to misapply the
terms.  As far as the Bolivian economy demanding a *cognitive* explanation
(which I take to be the real question of interest), the answer seems to me 
to be the same as that given for different brains above - if commonalities
in behaviour (however defined) across entities with different *physical*
makeups can be explained by postulating similarities in *abstract* properties
(such as cognitive functions), then it's appropriate to do so.  In this
way, the Bolivian economy should be treated in the same way that any
other entity is -such as a Martian, or a primate, or even another human.
If cognitive processes help explain its behaviour, then they are reasonable
to postulate.

  Of course, it may be possible to describe the behaviour of *any*
collection of matter in such a way as to meet the above criteria for
applying cognitive terms (given that what counts as "behaviour" is not
objectively in the world, and given the infinite properties that one can
observe changes in).  This is a problem for functionalism.  Just as a 
reminder, the Bolivian economy example is intended to be a reductio...

>>Your problem about ceasing to be an implementation only applies to the
>>case in which the Bolivian economy *fortuitously* implements a 
>>Turing-Test-passing program, and not to the case where it is
>>arranged and manipulated *intentionally* (say by Donald Trump :-) to
>>implement such a program.  Sure, it's unlikely that the Bolivian
>>economy *currently* has mental states.  What I want to know is if
>>it could *in principle*.  You seemed to suggest that the answer is
>>"no", but it is unclear to me why.
>
>Well, it's unclear to me how you and Donald propose to execute programs
>on this device of yours, Bolivia.  If you're thinking that the economy is
>so close to a functioning algorithm now that all it needs is a few nudges
>from a rich investor, I just don't buy it-- if it's operating like an
>algorithm it must be fortuitous and it's going to dissipate in time.

I agree.  I don't think that the Bolivian economy is currently 
running an *easy-to-detect* algorithm.  (However, recall that  
behaviour can be defined over whatever properties *you* choose, and
there are a great many possible properties, even when restricted to
the economics of South American countries.) 

>(Is David Chalmers here?  Won't we have counterfactual troubles here?

Only presumably if we are relying on fortutious arrangements - which we
need not.

>If some Bolivian decided to buy a poncho instead of a sack of potatoes,
>would your algorithm disappear?)

Well, your cognitive processes wouldn't necessarily disappear if a neuron
misfired.  

>If you're thinking the Donald is going to pay every citizen of Bolivia
>to follow economic rules that end up implementing an algorithm-- OK, fine,
>you've got your computer, but I don't see that you necessarily have the
>economy of Bolivia anymore.  You have a bunch of Bolivians playing a game,
>not unlike Searle's Chinese Gym.


The Donald doesn't have to pay *every* citizen - merely arrange the
initial conditions and rules such that the economy runs *itself*
in the appropriate algorithm.  (Sure this is an economic idealization,
but we can allow for tweaking on the fly by Trump, and, as noted
above, presumably your brain processes are 100% accurate either.).

In any case, I don't want to quibble over what counts as the *real*
economy of Bolivia, and what counts as merely a computer being acted
out by Bolivian citizens.  To be honest, I don't see how a functionalist
can assert that there *is* a distinction... In any event, other entities
(schools of fish, interacting galaxies) can be used which are less
contentious as far as their existence, and still run through the 
example. 

- michael


