(urth) 5HC : Skinner, Turing (fwd)

[Iorwerth Thomas]

>From: Maru Dubshinki <marudubshinki at gmail.com>

>!!
>Why is it every philosphical idea I think is amazingly clever turns
>out to have been stolen from me by some long dead philosopher?!
>"Hume's fork" ?  Dammit!
>Well, I hope I at least get credit for the adaptationist addition.
>Time to dispose of your arguments:
>Randomness in nature is bad for obvious reasons.  If any time I meet a
>dreaded lion or hippo (let's pretend I am a human pre-cursor living in
>Africa) I have a random chance of either a) fleeing in abject terror,
>wetting my pants, and the next time I go to that place I am
>extra-cautious or
>b) walking up to said dangerous, unfriendly creature and poking him in
>the snout. Or some other random action which doesn't involve
>killing/fleeing the creature.  It's equally obvious, I hope, that the
>set of actions which will preserve my life is shrinkingly smaller than
>the set of all possible actions.
>  So, random actions would be selected against, and actions with sense,
>which are non-random (ie, presented with an identical situation, the
>actions will not be utterly chaotic when iterated. If a creature is
>hungry and good food is near by, we should see it the majority of
>times engaging in the highly ordered, non-random sequence of going to
>it, checking it out, checking out teh surroudnings, eating it etc.)
>and fairly deterministic would selected for.

Oops.  My bad.  I misconstrued you.  I thought you were arguing that 
environment _alone_ rather than 'environment plus internal state' were 
determining factors for behaviour.  An action might be 'random' in the first 
sense but it wouldn't be in the second (though it might be free but not 
random, if you reject Hume's eating implement).  Sorry about that (esp. as 
the first view is so silly that I'd hate to ascribe it to anyone).

>Your analogy with WWII nuclear strategies is not really relevant.  It
>was an attempt to manipulate an opponents game-theoretic optimum
>strategy. The one acting 'crazy'/'random' was anything but.

I hope my remark makes sense now.

>And your second thing?
>Evolution is *very* fine-meshed.  The slightest advantage, compounded
>like interest over thousands, hundreds of thousands, millions, tens of
>millions etc will win out.
>If you wish to argue that 'non-enviromental' influence exist, please
>provide an example, any example.

(Caveat: I'm no biologist, and I'm willing to be told that I'm wrong on 
this) I imagine that there's a significant difference between the harshness 
of the selection of hairless tropical apes during a time of relative calm 
(when genetic drift will predominate) and the onset of an ice age.  Also, if 
a circumstance comes up rarely enough that, say, only one or two individuals 
are ever affected, it might not really impinge on an spiecies' history.  It 
might prove hard to predict the rection (though I'd have to be something 
_very_ odd not to invoke a flight or fight behaviour).

>For my purposes, 'enviroment' is anything which affects an organism's
>chances of living or reproducing.  This keeps everything nice and neat
>and evolutionary :)
>

Ok, I think I have a criterion for a 'non-environmental' [1] influence.  
Something which has not, up until this point, had a significant influence on 
the organism's chance of living or reproducing.  Good enough for ya? :)

Iorwerth


[1] Ok, it's not really.  But it's not environmental in the sense of the 
word you're using, if you see what I mean.