Tom asks:

> I may be out on a limb here, but I'll agree in that no single individual
> member of any species has a "distributed brain" [I think that's what you
were
> driving at], but what about "hives" -- ants, bees, etc., where the "hive"
as
> a whole moves with a sense of purpose?  [also known as "mod mentality"]

That's actually an excellent question/comment, and it requires a few hours'
answer to respond to properly. Nonetheless, let me provide a very brief answer
here.

One person has already written and said that "encephalization" doesn't appear
in Merriam-Webster's dictionary, but the word is common enough among
biologists. The process of encephalization implies evolutionary advance, thus you
won't see it in any single implementation but rather only when you view it over
time as a design evolves.

Similarly, let me also use two more words that you won't find in M-W:
"sematectonic" and "stigmergy," but which are just as commonly used in biology. And
in that regard, let me also refer you to a PhD engineering dissertation that
uses both words that I just stumbled onto now:

http://dochost.rz.hu-berlin.de/dissertationen/brueckner-sven-2000-06-21/HTML/b
rueckner-ch1.html

"Sematectonic" literally means "signal to build," just as "semaphore" means
"signal flag." Similarly, "stigmergy" means "marked work". The myrmecologist
(ant biologist) E.O. Wilson originally defined the word "sematectonic" to
supplant "stigmergy," believing it to be much more representative of what actually
occurs in a colony of social insects. Someone, without my knowledge, may have
since defined a hierarchy to these two words, giving them slightly different
meanings because the German fellow who wrote the dissertation I've referenced
above is using the ideas hierarchically, sematectonism being a subset of
stigmergy.

In the blood clotting cascade, in the immune system response, and in the
actions of ants in a hive, there is no central point of command and control. The
individual autonomous agents are responding to signals and triggers provided by
other individuals of their own or different castes. In this situation, the
resultant actions are, as Tom suggests, very much a case of distributed
computing. Because of the nature of this form of computation, the subject has been
much studied in engineering curricula for applications in things such as
autoconfiguring networks, load balancing in partitionable computing, traffic flow,
etc. The general rubric given to these kinds of investigations is called "complex
adaptive systems," and I'm sure that that's part of where HP is getting its
general name, "Adaptive Enterprise Computing," although there seems to be only
marginally little of the nature of autonomous agents interacting in HP's
design.

But the process doesn't have to be massive to be effective. The V-shaped
flight of geese is a form of identical autonomous agents which interact in a
highly specific manner so as to the benefit the group as a whole. The lead bird has
to work a little harder than the rest, each of whom get to swim in his wake,
and as a result the lead bird must change every so often. Nonetheless, this
simple example represents the benefit of colonial activity. In the adaptive
state space in which these animals would fly, this design represents at least a
local optimum, if not a global optimum.

All coloniality is like that. It would not exist -- nor persist -- if it did
not present distinct advantages over independent agents living independent
lives. In the 1900's, the anthropologist Alfred Kroeber defined the term
"superorganism" to represent the distinct break in architecture that a colony of
individuals represents over the actions of a mass of independent individuals
themselves. However that definition also obscures the fact that among biological
individuals, the colony superorganism is nothing more than a recapitulation of
the evolution of the original colonial structure of independent cells that
formed the very first agglomerated, task-partitioned multicellular organisms. It's
a repeated design, recapitulated at one higher level of organization.

But once task partitioning occurs between the components of the
phenotypically distinct caste members of a colonial structure -- and if the colony of
autonomous agents grows large enough -- the inevitable push is towards
encephalization. We don't see that encephalization in the ants, undoubtedly because the
colonies are not large enough nor are they complex enough in their phenotypic
partitioning, but if they were, I'm sure that they too would recapitulate the
same patterns seen 570 mya, at the beginning of the Phanerozoic, during that
time that's now described as the "invention of the phyla."

Wirt Atmar

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *