Re: Perceptual symbol systems

From: dan michaels (feedbackdroids_at_yahoo.com)
Date: 08/13/04 

Date: 13 Aug 2004 09:28:45 -0700

Traveler <traveler@nospam.com> wrote in message news:<mctkh0d1hhr6hqpbh6ioccvk408fu3r5gm@4ax.com>...

And it's true that the visual cortex is
> >> genetically designed to recognize edges, lines directions, etc... But
> >> all this stuff is just generic visual sensory capabilities. The eye
> >> and the visual cortex have nothing to do with the actual things that
> >> you look at. That is to say, they could not care less where you set
> >> you gaze: animals, trees, people, stars, rivers, rocks or what have
> >> you.
> >
> >
> >Well, in fact, that is the point. There is a reason we have those
> >particular 30+ cortical centers that each perform more or less
> >specific types of processing on the incoming visual information.
> >Natural selection selected for those particular processing types
> >because the information they provided to the organism enhanced
> >survival. The information selected for was common and repeatedly found
> >in the environment the organisms evolved within. Trees have vertical
> >edges, stones are blobs, other animals [predators/etc] move, all have
> >colors, objects close in have larger parallax differences than those
> >further away, on and on. The visual system organization didn't just
> >happen, it happened for a reason. The reason they're coded into the
> >genes is because they help the organism survive in the environment is
> >evolved into. If they didn't intimately help survival in the real
> >world, then the organisms would have gone extinct long ago.
>
> Certainly. But this applies only to lower animals. Humans, OTOH, don't
> have anything close to the degree of specialization that animals do.
> As I wrote earlier, some animals can recognize specific objects that
> they have never seen before. Our visual system will work just as well
> on Mars as in a virtual world. We don't just recognize vertical edges
> but all sorts of edges having various degrees of orientation. This is
> the mark of general visual system.
>

I think we both have similar ideas, but our interpretations are
slightly different. The perceptual systems of mammals are much more
general than those of lower animals which are much more hard-wired,
but still I think the processing centers were selected for during
evolution based upon fairly specific information they provide the
organism about the external environment. [there's probably a better
way to say this].

Now, higher animals do have to "learn" how to piece together the info
supplied by the 30+ visual centers into a coherent image. That's
fairly clear. This is one of the baby's main jobs in life. It's always
been my idea that the 30+ visual areas act essentially like
pre-filters, and their outputs then converge on more or less
tabula-rasa learning areas, in which the various area-outputs are
combined into a coherent "image" ... to use the word everyone around
here despises - where "binding" takes place. So, you have both
genetically-determined pre-filtering areas, along with more labile
learning centers which use the outputs of the pre-filter areas.

Coupla nites ago, I read Oliver Sack's essay called "To See and Not
See", regards a guy who had polio as a kid and lost his sight, and 45
years later was operated on and had cataracts removed, and his sight
partially restored. Unfortunately there was significant retinal
damage, esp in the macular region, but what he could see was rather
interesting. He could see colors, blobs, edges and lines, motion, and
even read letters on a wall chart with 20/100 acuity. Over time, his
vision got a little better but not much. What was most interesting to
me is that he could see bits and pieces of visual objects [like the
letters, and also parts of his cat's head like ears and eyes/etc], but
he could not coalesce the pieces together into a coherent image. IE,
he couldn't put the letters together to form readable words and he
couldn't see the cat's entire head as a whole, only the separate
parts.

In some cases, he could learn to perceive essentially an entire abject
as a whole, after he was able to touch and completely examine the
object by feel. This is interesting in light of how Piaget describes
babies as needing to make eye-hand see-touch associations in order to
build "internal cognitive structures" [a term you probably despise]
during their first year of life. It's also interesting in terms of
modularity of brain structure, regards other agnosias where people can
recognize objects, but not name them, etc/etc. The guy with
not-quite-restored sight was able to perceive a lot of the bits and
pieces presented to his higher centers by the visual pre-filters, but
was unable to integrate the signals properly.

> >Many of these centers exist in other mammals, but apparently humans
> >have more types. From an evolutionary perspective, these are all
> >basically outgrowths and reuse of processing centers which exist in
> >vertebrates lower than mammals, such as amphibians and reptiles; but
> >as mammals evolved, the processing became less specific and more
> >general, which allowed new species to appear. Supposedly objects that
> >don't move are invisible to frogs - frog survival apparently isn't
> >enhanced by being able to recognize a tree or a rock, although this
> >idea does seem strange. Apparently frogs don't move fast enuf that
> >they need to recognize a tree prior to running into it. Whatever.
> >Animals that move fast need to be able to "predict future" - as
> >Dennett says - so evolution gave them the necessary visual centers -
> >ability to perceive binocular parallax, etc.
>
> You will be surprised to learn that we are not that different than
> frogs. Humans cannot see anything either if there is no motion in our
> visual field. Our eyes are continually moving in tiny jumps called
> saccades. It was found that the moment the eye is immobilized (via
> medication and head restraint), we lose our ability to see stationary
> objects. We become frogs, so to speak. A frog's eye does not seem to
> have human-like saccadic capability. A frog sits there immobile and
> practically blind waiting for something interesting (prey) to move.
>
> An interesting observation in this context is that some spiders do
> have eye saccades even though their eyes cannot turn. They make up for
> it by having five eyes around their heads.
>
> >It's for all of these reasons that I say our perceptual systems give
> >us a "close-enough" internal representation of the outside world. We
> >have to learn the names for objects, etc, and we have to learn as
> >babies to distinquish object from background, and other
> >characteristics of the physical world, but this would not be very
> >effective if we didn't posses those 30+ visual centers provided by
> >evolutionary processes.
>
> My whole point is not there is no correlation between the environment
> and brain connectivity (of course there is), but that, a learning
> organism does not try to represent the environment in its brain. This
> is true, even if, after post-mortem dissection or through whatever
> other means, we do find a correlation between the environment and the
> brain of the organism. You have to remember that the organism does not
> see the environment separate from the "representation.". It can only
> see the "representation" which is not a representation because the
> subject (what is being represented) is not there. You must look at the
> problem from the point of view of the intelligence itself, not from
> your point of view. Representation is always from the observer's point
> of view.
>
> Louis Savain
>
> Artificial Intelligence From the Bible:
> http://users.adelphia.net/~lilavois/Seven/bible.html
>
> Falsifiable Predictions:
> http://users.adelphia.net/~lilavois/Seven/predictions.html

Relevant Pages

  • Re: Genetic or Epigenetic: The Causal Basis of Evolutionary Change
    ... The *reason* why most new mutations are neutral or deleterious is ... simply because the *environment* that the organism has evolved within ...
    (talk.origins)
  • Wholes & Parts
    ... According to the famous Scientist Richard Dawkins the concept of God has ... It is true that science must start by gradually collecting facts and, ... No organism exists in isolation. ... development and fate of the organism depends on the environment (and the ...
    (soc.religion.islam)
  • Re: Bet Hedging, Risk Aversion, Sex, and the Unit of Selection
    ... > behave as if they were risk averse maximizers of fitness (i.e. ... > On Population Growth in a Randomly Varying Environment ... > environment and other offspring adapted to different environments. ... An organism only HAS a few offspring. ...
    (sci.bio.evolution)
  • Livestock: Methicillin Resistant Staphylococcus Aureus
    ... To ask the Secretary of State for Environment, ... MRSA in farm animals; and if he will make a statement. ... The organism has been isolated from ... The European Food ...
    (uk.environment.conservation)
  • Re: On Evolution
    ... Epigenetic coding, determined by the environment, ... accounts for a much larger part of the complexity of an organism. ... which suggested that epigenetics "accounts for a much larger part of ... than only random mutations. ...
    (sci.bio.evolution)