Re: complexity

20 Nov 1995 17:53:47 GMT

cc3265@CNSVAX.ALBANY.EDU wrote:
: In article <>, Troy Kelley <> writes:
: >I think you guys are thinking about this all wrong.
: >
: >First off, some books quote dolphine brains as being larger than ours.
: >
: >"The human brain weighs 1,400 grams, and the dolphin's 1,700. The
: >dolphin's cerebral cortex is larger than ours. It has twice the number
: >of convolutions, and 10 to 40 percent more nerve cells" (sorry I don't
: >have a page number)

: As I understand it (and I heard this at a dissertation defense, so, I'm
: sorry, I don't have a reference), dolphins do have large brains relative
: to body size, but their brains have expanded more in the temporal lobes,
: while most of ours is in the frontal. This makes sense, given that dolphins
: rely so much on auditory stimuli.

The initial neocortical processing of sonar data occurs there. I've seen
some tonotopic maps for bats, and the region associated with the
frequency of their squeak is massively expanded. In whales, it would be
even worse since the environment is noisy and non-linear. Target motion
analysis is used both in biological and artificial systems to maintain
tracks passively. This is done in the medial temporal cortex in mammals
for vision, and I have little doubt that much of the expansion of the
temporal cortex in whales is involved in similar processing for their
sonar. The following is relevant (and Copyright 1995 Erwin):

The key step in automatic tracking using bearing-only data in man-made
systems is the rapid association of the individual sound source with a
specific `bearing-only' track. For an isolated source in a quiet
environment, this is trivial, but the following problems exist
under realistic conditions:
\item the ocean environment is noisy (biological noise (e.g., snapping
shrimp, whale farts), distant shipping, icebergs, seismic testing, oil
rig drilling, wind, rain, and wave noise) and heterogeneous (water
density and salinity gradients, bottom type, multipath effects),

\item sound sources are quiet and may use tactics to confound their
emissions with other sources, and

\item self-noise (machinery noise, cavitation, other propellor noise,
and turbulence) can be significant in hiding useful information.
Crossing tracks are particularly hard to follow and can overwhelm the
data processing resources of a sonar system by generating many
possible alternative correlations. In practice, these systems use
experienced sonarmen to rapidly classify the crossing sound sources as
they separate and become distinguishable. In a high-traffic
environment, this classification workload becomes a constraint on the
quality of the tracks maintained.
---end insert---

I suspect the enormous expansion of the human frontal cortex is associated
with three things:
1. Planning,
2. Working memory,
3. Reafference associated with cued search (as opposed to general
attentiveness, which may be involved with the locus coeruleus).
If those are important to your definition of 'being human,' you probably
need to specifically associate human intelligence with frontal cortex

Harry Erwin
Home Page: (try a couple of times)
PhD student in comp neurosci: "Glitches happen" & "Meaning is emotional"