>>>>> "F" == F Frank LeFever <flefever at ix.netcom.com> writes:
F> To begin with, if you are modeling based on a cartoon
F> (tripartite brain), you're "not playing with a full deck". ALL
F> parts of the human brain have evolved, and the evolution
F> includes intricate interrelationships among all parts. (no
F> disrespect to MacLean, who was one of my speakers at a NYNG
F> conference I organized many years ago).
In fact, our research is mainly inspired in Antonio Damasio
(e.g., "descartes error", 1994). His work suggests the existence of
two natures of stimulus processing in the brain --- at the level of
the amygdala (what we call "perceptual"), and at the level of the
neocortex (what we call "cognitive"). For instance, "primary emotion"
involve the perceptual layer, and the "secondary" ones make use of
both layers (using the prefrontal cortices to activate bodly
representations). But when defining the nature of each one of these
representations, one has to consider a third one, which seems to
reside at the level of the spinal cord (e.g., reflex reaction). Of
course that in the brain, everything is connected with (almost)
everything else, and the MacLean model is an extremelly rough
If one takes into account from the start this complexity, it
is difficult to advance in terms of modelling. The question remains:
what kinds of abstractions can one refer to when analysing the brain
in terms of large functional blocks?
F> You cannot describe "layers" without reference to
F> anatomy, inasmuch as "layers" itself seems to be an anatomical
F> concept (default case: metaphorical concept).
Sure, but AI researchers are seldom interested in anatomical
issues. My objectivo is to build models based on the brain, but
defined without resorting to anatomical explanations.
F> How do you "functionally" differentiate? Define your
F> terms. If you mean "reflex withdrawal", this can be done at a
F> spinal level. If you mean "increased heart rate, other
F> autonomic activity", you are involving higher levels.
But can those two kinds of response be defined in terms of
theirs proprieties? By "functional description" I mean the way they
function, the way they respond to stimulus, and possibly the way they
adapt. For instance, take the classification of knowledge in
non-declarative and declarative. In this classification there is no
mention to brain zones where this knowledge is hold, but rather a
definition of the proprieties of each of them. Reflex withdrawal is an
automatic immediate (reactive) response, while "emotions" seem to
involve visceral response (as Damasio terms it) based on an assessment
of the stimulus. If it weren't possible to functionally diferentiate
(am I using a computer-science-oriented definition of the word
"functional"?), then why not say that the spinal cord and the amygdala
are all the same? (Or why not divide the spinal cord into an arbitrary
large number of blocks, just because of their physical position?)
F> First, define "emotions". After that, we can talk.
Thats a tricky one...
My view is that the word "emotion", just like "intelligence",
"consciousness", and so on, cannot be precisely defined, since they
originate from common-sense knowledge, and from out linguistic
past. However one can build models which capture in some respects what
is commonly understood by "emotions" (or "inteligence"). Unfortunately
one cannot do better than this. However I guess it is more relevant to
work on these models rather than trying to define these
terms. Wouldn't you agree?
*** Rodrigo Martins de Matos Ventura, alias <Yoda>
*** yoda at isr.ist.utl.pt, http://www.isr.ist.utl.pt/~yoda
*** Teaching Assistant and MSc. Student at ISR:
*** Instituto de Sistemas e Robotica, Polo de Lisboa
*** Instituto Superior Tecnico, Lisboa, Portugal
*** PGP Public Key available on my homepage
*** Key fingerprint = 0C 0A 25 58 46 CF 14 99 CF 9C AF 9E 10 02 BB 2A