On Tue, 3 Jun 2003 13:54:00 +0000 (UTC), dan{at}oricomtech.com (dan
michaels) wrote:

>Over on the robotics-AI forums, we have a discussion going about
>possible brain mechanisms. One of the things we have been hashing out
>lately is the role temporal phenomena play in operation of brains and
>organisms. We realize that most AI models all but ignore this, to
>concentrate on static and/or spatial models.
>
>So, I have been thinking about temporal activities of the brain,
>vis-a-vis how they might have developed during early evolution. It is
>just a conjecture, but let's say early on, movement activities were
>the first introduction of the organisms to temporal phenomena, whereas
>sensory activities were more spatial and on-off like, rather than
>graded and temporal. IOW, this is saying that early organisms had very
>primitive sensory apparatus that simply signaled encounters, like
>bumps, with the outside world, but that to escape from these bumps,
>the organisms had to carry out some successive string of muscle
>activities, or whatever they used for propulsion.
>
>IOW, I am trying to say that the sensory side was more tuned to
>spatial input, while the movement side was more tuned to temporal.
>This would mean that the organisms needed to develop a brain which
>could create the temporal sequences needed for the movements to be
>carried out. IOW, the part of the brain that first evolved the most
>sophisticated temporal apparatus was the movement related parts. I
>assume this makes some sense ?????????
>
>For instance, an arthropod needed to evolve very complex timing
>mechanisms to control its 6 or 8 legs during walking, whereas its
>sensory apparatus probably did not need to deal with such complex
>timing issues.
>
>The next step is that, if we bootstrap off of this idea as we go up
>the evoltionary chain, then it would stand to figure that the movement
>end of all brains would be the place where the temporal aspects were
>of paramount importance. Does this make sense ????
>
>Now the question. I know that most or all sensory organs of the
>various kinds are modified hair cells, but what are muscle fibers? How
>did they come about? Actin and myosin - much different from sensory
>hair cells. These are made to create temporal movements, under control
>of a central timing mechanism, as compared to hair cells which respond
>to movements.
>
>Which might have evolved first? Ability to move or ability to sense?
>Or are these things of the same basic origin? Like cilia in protozoa
>that might either move or respond to movement? Maybe cilia don't
>sense, but only move.
>
>This is all kind of jumbled, but I am trying to think of how the
>temporal aspects of the brain evolved, and am thinking it must have
>been more connected to the movement side of the organism, and not to
>the sensory side. Related to this might be one of the reasons why AI
>has pretty much drawn a blank so far, since most concentration has
>been on the sensory side [ie, vision] or high-level function [ie,
>symbolics] - both of which are normally modelled without worrying
>about temporal issues. There seems to have been little concentration
>regarding integration of this with the effector side where temporal
>comes in.
>
>Anyone got any idea what I am talking about here? [if not, never mind
>:)].
>
>
>- dan michaels
>==========================

I think you are off on a side track taking you nowhere.  

Clearly, temporal sequencing of motor activity is necessary to produce
any patterned activity.  This in seen in all the rhythmic activities
of locomotion: swimming, walking, flying; in respiration: breathing
and ventilation of gills; in digestion: GI motility, peristalsis. But
temporal analysis of sensory activity is just as necessary to detect
motion, to localize stimulus source, to differentiate novelty from
constancy.  You are asking "chicken vs egg" questions about which came
first.  Nervous systems have the ability to do both spatial and
temporal processing.  For example, spatial and temporal summation are
both fundamental aspects of the way synaptic integration occurs.  Many
systems, sensory and motor, are separated into phasic vs tonic,
clearly a temporal processing distinction.  At the same time, many
systems organized as parallel processors with a population of similar
cells located on some surface show lateral inhibition, clearly a
spatial processing distinction.

Another "chick vs egg" question is which came first, motility or
receptiveness to the outside? Both of these long predate any notion of
a nervous system.  Both are fundamental cellular processes dating back
to prokaryotic times.  Single-celled eukaryotes show motility using
both ciliary (tubulin-based) mechanisms and "muscle" (actin-based
mechanisms).  Also single-celled eukaryotes show responsiveness to the
environment using membrane based proteins that initiate signaling
systems mediated by ion channel gating and membrane potentials.  Think
of a paramecium bumping into something.  It backs up, then moves
forward at an angle to detour around the object. It seems very
goal-oriented but is the simple working out of a simple
receptor-signalling-effector chain.

If you are interested in nervous systems, most thinking is that early
multicellular eukaryotes, presumably something like todays jellyfish
(OK, not necessarily jellyfish, but Cnidaria nonetheless) used
chemical signals in the water to help control and coordinate the two
major problems of life: eating and reproducing.  Nervous coordination
was a way of integrating and coordinating the activity of many cells
each of which had independent capabilities both to detect events in
the environment (sensory) and to respond to them (motor).  Throw into
the mix the notion of secretion as another major way that cells
respond to stimuli.  You now get simple signaling systems where one
cell detects an environmental signal and responds by secreting a
chemical.  A different cell detects the first cell's secretion and
responds by doing something that is appropriate to deal with the
orignal environmental situation.  Extend the chain to include more and
more cells.  Now throw into the mix the notion that the
stimulus/response system within a single cell can involve intermediate
signals which can be either chemicals or membrane potentials to
connect the physically separated receptor and effector. Finally make
your chain of signalling cells very elongated in shape so that the
messages travel most of the way intracellularly through electrical
potentials and only very short distances between cells by secretions.
Voila! A nervous system!

The major point is that all the aspects of nervous function, sensory
reception, synaptic transmitter reception, transmitter secretion,
electrogenesis of both graded and all-or-none signals, and different
mechanisms of motility are all basic elements in the toolkit of all
cells. Nervous systems are johny-come-lately's taking advantage of
these cell phenomena to do the coordination job.
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 6/4/03 6:11:57 AM