Brain, Behaviour and Extensionalism

Joe Legris jalegris at xympatico.ca
Sat Apr 10 12:11:57 EST 2004


Wolf Kirchmeir wrote:
> JXStern wrote:
> 
>> I suppose you still don't care to try to address this theory I have
>> about my computer, but it trumps anything you care to say until you
>> adjust your claims appropriately. 
> 
> 
> The computer "contains a program" in a metaphoric sense only. That sense 
> is useful at the level ay which we interact with a computer (when we 
> "run" or "exit" a program, when we "input data" via the keyboard, etc.) 
> But this language does not describe what actually happens in a computer.
> 
> What a computer actually contains is circuitry, and, when it operates, 
> electricity flowing through that circuitry. Some of the circuit elements 
> are so constituted that their behaviour changes as electricity flows 
> through them. The result is that the pattern of the circuitry changes 
> peridocally (about 400,000,000 times a second on my machine), and in 
> turn the flow of electricity through the circuitry changes, too - it 
> takes different paths. Some of these paths end or begin at "peripherals" 
> - the CRT screen, the printer, the keyboard, the mouse, etc. What I type 
> on the keyboard changes the patterns of the circuits and eventually (a 
> few milliseconds later), the image on the screen changes. One could say 
> that my typing is the "stimulus" and the image on the screen is the 
> "response".
> 
> Where is the program, exactly? In which circuits is it contained? Etc. 
> The proper answer is "that's a misleading question". The program 
> _exists_ only while the computer functions, when electricity flows 
> through its circuits, and millions of transistors etc change their 
> states from one tick of the clock to the next. Shut off the power, and 
> there's no program. ***
> 
> One could identify the program with a particular pattern of some part(s) 
> of the cirucit, but this pattern comes into being only when the power 
> flows through those ciruits (I'm talking about RAM, which must be 
> "refreshed" at regular intervals.) You could also identify the program 
> with its execution (ie, there is no prgram until it's executed), but 
> again, where is the program? It exists only while it "runs", and it will 
> run "in" different parts of the computer at different times. (Some parts 
> of the computer, eg the graphics card, run the same program at all 
> times, however, but even they do so with different data, so that each 
> run is different.)
> 
> IMO, the analogy with brain functions is obvious, and quite strong. DFor 
> example, "I" exist only while this brain of mine functions. "I" 
> disappear when this brain functions in different ways (when I sleep, or 
> when I'm anesthetised, etc.) It's not much of a strecth to say that "I" 
> is a program that "runs" much of but not all of the time.  And just as 
> the computer's behaviour gives the impression that there is something 
> "in there", so my functioning gives the impression that there's 
> something "in there." One could talk about chains of causation, I 
> suppose, but they all have the same general pattern: Input --> something 
> happens inside --> Output, which doesn't get us very far (apart from 
> figuring out how certain input configurations result in certain output 
> configurations, which can be very useful knowledge).
> 
> The question of how and why "I" feel that "I" exist and "have 
> experiences" is not answered by positing souls, minds, etc. That's 
> merely using alternative terminology. The best general answer to "I"'s 
> existence is that it's an "emergent phenomenon", and a variety of 
> mathematical concepts help one understand, sort of, what that phrase may 
> mean. But there is at present no specific answer. That seems to bother a 
> lot of people. But the AI people shouldn't be worried. They should 
> concentrate on building systems such that Input --> cicuitry --> Output 
>  more or less matches the desire behaviour of the system. That's enough 
> of a challenge to take up all oner's time, IMO.
> 
> ** Footnote on memory: People talk about RAM and disks as if they were a 
> sort of memory. Let's agree that this is useful metaphor. What then is 
> actually "stored" in RAM or on disk? Nothing, actually. What exists in 
> RAM or on disk is a pattern of magnetic fields. We say the data ie 
> "written" to storage - that is, patterns of magnetic domains are 
> impressed on the disk. We say the the data is "retrieved" - but what 
> actually happpens is that the flow of elctricity through the read head 
> is modified, and, once again, circuitry patterns change, and electricity 
> flows along different paths inside the computer. While this is 
> occurring, the machine may be said to "remember" what was "stored" on 
> disk. But the memory only _exists_ while the electricity flows. No 
> electricity, no memory. IOW, there is no "memory", there is only 
> "remembering". And if the storage media cannot be read (which is already 
> a problem, BTW, for data stored on old disk formats), then there can be 
> no "remembering".
> 
> In human beings, too, there is no "memory", there is only "remembering", 
> that is, the functioning of brain circuits under the influence of 
> "stored data" (which appears to be differential strengths of synapses, 
> as far is known to date). Interfere with the brain circuitry that 
> functions when we "remember something", and we no longer remember. A 
> blow on the head, drugs, intense emotion, etc, can all change the 
> brain's circuitry, and so change, extinguish, an even create "memories".
> 
> In both computer and humans, "remembering" may result in observable 
> behaviour. In both computer and humans, most "remembering" happens 
> internally only: it's one of the operations of the program.
> 
> Having said all this, I do know that analogies are not proof, but they 
> are aids to insight. Behaviourially, they "satisfy" -- that is, they 
> "reinforce themselves" (label quotes).
> 
> That's my story, and I'm sticking to it. Or it's sticking to me. Hard to 
> tell, really. :-)
> 

To say that a computer contains a program and that memory stores 
information is no more metaphorical that saying a cup contains water and 
  a cellar stores potatoes.

Where in the cup is the water? It is just a collection of molecules 
confined by gravity and intermolecular forces. Some of have vaporized 
and are no longer in the cup at all. In fact, if you wait long enough, 
and the humidity is high, every single molecule will eventually be 
replaced, but it's still a cup of water.

What is really stored in the cellar? People have placed fresh potatoes 
there and will come back later to retrieve them, but no one checks to 
see that those retrieved are exactly the same potatoes as those stored 
because it doesn't matter.

It is the functional relationship that counts. Information processing 
and storage is as purely functional as you can get and that is why it is 
so useful.

--
Joe Legris




More information about the Neur-sci mailing list