From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!tdatirv!sarima Thu Jan 16 17:20:14 EST 1992
Article 2695 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!tdatirv!sarima
>From: sarima@tdatirv.UUCP (Stanley Friesen)
Newsgroups: comp.ai.philosophy
Subject: Re: Waking up is hard to do, but somebody's got to do it
Message-ID: <370@tdatirv.UUCP>
Date: 13 Jan 92 23:54:09 GMT
References: <60551@netnews.upenn.edu> <334@tdatirv.UUCP> <60759@netnews.upenn.edu> <350@tdatirv.UUCP> <61668@netnews.upenn.edu> <363@tdatirv.UUCP> <61968@netnews.upenn.edu>
Reply-To: sarima@tdatirv.UUCP (Stanley Friesen)
Organization: Teradata Corp., Irvine
Lines: 214

In article <61968@netnews.upenn.edu> weemba@libra.wistar.upenn.edu (Matthew P Wiener) writes:
|In article <363@tdatirv.UUCP>, sarima@tdatirv (Stanley Friesen) writes:
|
|>That is my point, sort of.  It evolved, so the added benefit of sleeping
|>*must* exceed the danger.
|
|My point is that the dangers of sleep to species are not as great as the
|danger to individuals.

Hmm, I am not disputing this.  I do not think it relevant.  To become a species
feature it must first appear in individuals of the species and survive the
selection process long enough to become established.

Species are *not* uniform, unvarying wholes.  They are aggregates of
individuals.  Any disadvantageous feature that has less deletrious alternatives
available in the gene pool will tend to become rare to non-extistant. Always.

The only way this can be blocked is the lack of alternatives that do not have
other, more serious, disadvantages.

|>			    I think my model explains the benefit quite well.
|
|But your model--preventative maintanance--has been explored and found
|wanting for decades.

Hmm, this certainly seems unlikely to me.  Are you sure all approaches to
testing for PM have been tried?  My own body certainly *seems* to recover
from exertion and stress faster if I sleep than if I do not.  Even injuries
seem to heal faster while asleep than while awake.  [Not to mention emotional
stress].

Now, I am quite well aware how easy it is for the huma mind to fool itself,
but these effects seem so strong to me that *some* explanation for them
is necessary.

|>Especially since sleep does *not* seem to be correlated with anything
|>like 'intelligence'.
|
|Actually, there is a correlation.  High EQ (encephelization quotient) species
|seem to need more REM sleep than low EQ species.  Whether this correlation is
|due to an underlying relationship or not is unknown.  If sleep is needed for
|thermoregulation, than larger bodies will lead independently to both higher
|intelligence and simpler thermoregulation.

But REM sleep is only a small part of sleep.  It is certainly not the sole
purpose for sleep.  This just implies that different functional designs must
divide the PM resources up in different ways, with more cognitive creatures
spending relatively more resources on cognitive PM.  (Certainly the association
between REM sleep and mental stress and memory performance supports such a
possibility].

|>		       Even reptiles enter torpor of some sort at
|>various times.  It is not entirely clear that this is really any
|>different than mammalian sleep.
|
|Who knows?  Reptiles don't have REM sleep.  Very few mammal species (dolphins,
|spiny anteaters, ????) don't.

Dolphins *don't* have REM sleep?  That really plays havoc with sleep as
a quantum result of brain activity.  Dolphins are the closest thing amoung
non-human animals to a conscious being.

Even then, REM sleep is, in my model, the data processing phase of the PM,
rather like running fsck on a UNIX system.  Reptiles do have a much simpler
brain than any mammal, so the lack of a need for DP PM is understandable.

|>|Desire to survive is better than just instinct to survive.
|
|>What does this mean?  What is a desire except a particular class of instinct?
|>Or do you restrict the word 'instinct' to refer to insect style pre-wired
|>behavior?  If so, there is precious little instinct amoung amniotes.
|>[In short, I suspect the 'survival instinct' of even reptiles takes the form
|>of *feelings*, like fear, hunger and so forth].
|
|I refer to an active self-awareness to get food--tastes good!--rather than 
|stop the stomach from growling--less filling!

But, I think that it is very questionable to assume that animals *don't*
eat because it tastes good (in part).  Certainly my sister's pet dog seems
to have some very definate ideas about taste, and what she (the dog) likes.

This whole thing is a continuum.  Animals have positive and negative
reinforcement subsystems built into them.  These combine with learning
to provide conditioned responses to various stimuli.  Attraction towards
positive renforcement is a major part of most land vertebrates' behavior.
And just about any higher vertebrate builds a layer of generalization
on top of the basic conditioned response, allowing it to act on the
*anticipation* of pain or pleasure.  My sister's dog certainly shows quite
a bit of capacity for anticipation.

How does self-awareness differ from a special, advanced version of,
anticipation based on abstract understanding?
[Hmm, that didn't quite come out right - I am trying to suggest that
the distinction is a fuzzy, imprecise one, rather than a simple Aristotelean
binary one].

|>|And warmth has nothing to do with it?
|
|>Not in the tropics.  Humans live in buildings even in the warmest climates.
|
|You'll also note that predators like lions generally sleep more than their
|typical prey like zebras.  They eat their fill and laze around for days at
|a time.

Yep, because meat is a much more concentrated source of energy, so less
is needed.  Thus the zebra *must* spend most of its time eating, while the
lion simply does not have to do anywhere near as much.  Given a relatively
safe place to sleep (the middle of a pride with guards posted), one might
as well sleep when there is nothing else to do.  I certainly tend to.

|>There is a problem here. Almost all Cretaceous mammals were *nocturnal*,
|>sleeping, if at all, during the *day*, when the dinosaurs *were* active.
|
|This is known?  Considering how varied dolphin sleep is from other mammals,
|or even human from shrew, I find it quite difficult to buy the above based
|on extrapolation from today's patterns.  Is there some other determination?

Yes this is known.  It is based on comparative anatomy and taxonomic
relationships.  Nocturnal animals have identifiable anatomical features
that show up in fossils (such as large eyes, which are preserved as large
eye sockets).  Also it is possible to derive an evolutionary tree of the
vaious groups of mammals, and they generally trace back to nocturnal ancestral
forms.

|Are you implying that mammals did not need preventative maintence 1e8 years BP?

No.  But sleep does not fossilize, so we do *not* know if they slept in any
direct way.  I tend to assume they did, since all living mammals do, but that
is only a reasonable conclusion, not a fact.  Of course, I do find it hard to
imagine what a nocturnal animal would be doing during the day if not sleeping!

|>Also, given that crocodiles doze, and birds apparently sleep, it is quite
|>likely that dinosaurs themselves slept from time to time.
|
|Right.  And they probably had some intelligence to boot.

Very little, at least in the case of crocs and the plant-eating dinosaurs
most closely related to the ancestry of birds.  (Of course the carnivorous
dinosaurs that were even more closely related to birds *were* rather
intelligent - the most intelligent forms almost reaching avian levels of
intelligence, but certainly not mammliam levels).

|>					     It is almost as if any level of
|>processing above that of reflex requires some form of sleep.
|
|Yes!  Take out the "almost as if" and you've got what I've been arguing
|for all along.

This seems to be almost to trivialize your hypothesis.  This is so widespread
that your 'consciousness' would exist in fish, and perhaps in some insects.
This seems to be rather an over-extension of the term.  [At least I seem
to remember you claiming that sleep is related to consciousness].

I might accept that a shrew is conscious, but a dogfish?

|>But what did they die *of*?  What was the *proximal* cause?  If it was
|>any sort of physical collapse (like heart failure), this is perfectly
|>consistant with the PM hypothesis.
|
|The mice went on eating gorges, lost weight, developed skin lesions on
|their paws and tails, and eventually their thermoregulation collapsed.
|No proximal cause was identified.  No other problems were found, from
|internal organs to the immune system.
|
|Note that PM might be relevant to the paw and tail ulcers, but that
|would be strange--why were there no other ulcers?  None under the fur,
|and none internally.  Just the exposed naked skin.

Perhaps the naked skin was more susceptible to damage?  It is certainly
more exposed to radiation.  In fact the syndrome you describe sounds
suspiciously like radiation sickness.

Even the eating can be related to PM, as an attemtp to compensate for the
lack of sleep by increasing the energy budget enough to support waking PM.

|It would be interesting to do this experiment with their exposed skin
|protected.

Yes it would.  It might also be interesting to do a genetic assay of the
lesioned tissues to see if they might be suffering from high mutation
levels.  [Contrary to popular belief, living tissue *does* have ways
of repairing genetic damage, whatever its source.  So a failure of the
DNA repair mechanism might cause most of the observed symptoms].

|>				     In fact this level of toxicity in
|>unrested tissues would *easily* override the much smaller risk of death
|>while sleeping.
|
|What toxicity?

Death due to tissue degeneration.  I was internally modelling the death
as being due to the accumulation of metabolic wastes.  Perhaps a weak
assumption.  But still viable, many metabolic toxins are mutagens.

|>What reason is there to think the death in this experiment was due to
|>*mental* causes?
|
|None.  The belief in that link is from the known cognitive distortions in
|sleep-deprived humans.  A mere PM need would have to do a lot of explaining.
|
|The mice's death is consistent with a cognitive need for sleep: they no
|longer know to curl up their tails and tuck in their paws for heat
|conservation, so they no longer know how to keep warm, and so they die.

Hmm, sounds interesting.  It does seem to be somewhat related to the memory
effects of REM deprivation in humans.

I have modelled REM sleep as a form of background data processing that
clears up emotional backlog and commits important medium term memories
into long term memories (migration of data to more stable media).
-- 
---------------
uunet!tdatirv!sarima				(Stanley Friesen)



