The Prime Goal -Reply

[Hans Blom, 960305b]

(Shannon Williams (960228.13:20))

Organisms control, above all, for the "need" to transmit their
genes to the next generation. Everything is subordinate to this
prime goal.

I disagree. If you Test this goal, you will realize that it fails
the Test. If all goals were subordinate to the need to procreate,
we would never have trouble breeding animals in captivity; Neutered
animals would be horribly depressed (or something) because of their
lack of control; And my mother would be a grandmother.

------------------- ASCII.ASC follows --------------------
I'm not ready yet to offer a full presentation of why I make such
statements, which may appear meaningless to others. And it makes a
difference, as well, whether one considers _one_ human (as in
psychology) or the aggregate of _all_ humans (as in sociology or
biology, especially evolution theory). But let me present some of
my thoughts.

Control, as we talk about it, is a rather high-level concept that,
somehow, originates from the interaction of the more basic laws of
physics. At the level of the atom or molecule we don't see it yet,
at the level of the organism we do. We seem to need some complex
arrangement of units before control arises.

The notion of "loop gain", for instance, is such a high-level con-
cept. It implies that a number of "things" are series-connected so
that they form a "circular causality" loop, where the properties
of _all_ those "things" contribute to the properties of the loop,
so that e.g. _one_ zero gain element may make the whole loop inef-
fective as a "control" system, however great the quality of all
other elements. Basically, the control loops of "things" are made
up out of simple molecules or aggregates of molecules.

It is that link between the basic laws of physics and the "emerg-
ent" phenomenon of "control" that interests me, and especially the
question: how come the kind of control that we so easily recognize
in living organisms (and in some of the artifacts of some living
organisms) came to be.

Let me link this to my Brownian motion experiments, in which a
particle, whose size varies as a function of the concentration of
some chemical, tends to remain in the vicinity of the maximum
concentration of that chemical. Whether one considers that as a
control process or not, it is often a rather capricious process:
although the particle's position may remain more or less constant
in the neighborhood of the maximum concentration over extended
periods of time, it may yet sooner or later disappear into the far
reaches of infinity, where the chemical's concentration is so low,
and its gradient is so flat, that the particle cannot find its way
"home" anymore.

As an aside: in itself, that is not "bad" if the particle were a
living organism such as a bacterium. This "getting lost" may allow
it to find another food source (high concentration) where it might
again come to rest, multiply and establish an additional colony,
and thus serve the "higher goal" -- if one wishes to use such
loaded words -- of gene multiplication.

Back to the original argument. The particle's positional homeo-
stasis (whether we call it control or not) is not robust. But
consider now a large ensemble of those particles. Although some
proportion of those particles may "escape" from the chemical's
maximum concentration, this can show up to be a slow process,
where the population's "half life" is rather long -- depending
upon the chemical's concentration and its gradient and on the
function that relates concentration and particle size.

Now view the particle as a living organism which, in addition to
the "movements" allowed to it by the size-change mechanism, can
multiply if sufficient food (a high enough concentration of the
chemical) is available. This extra phenomenon may now (more than)
compensate some individuals getting lost. In other words: the
homeostasis of the _population_ is robust whereas one individual
could not ensure its homeostasis over the long range; wheras one
individual could eventually "lose control", the population cannot
(until the food source is exhausted, that is). In the ensemble of
cells that make up our body we find something similar: the whole
is much more robust than each individual. And in the ensemble of
individuals that make up humanity, again. Note that the behavior
of _one_ individual is hardly important.

In the above example, the laws that governed the behavior of all
individuals are exactly the same, but the influences of the envi-
ronment provide a random factor, resulting in different individual
behaviors (if, for the moment, we call the path of a particle its
behavior). The situation does not change appreciably if the laws
that govern the behaviors are a "family" of laws, i.e. have some
probability distribution. In the example above: the size changes
of the particles need not be identical; the population homeostasis
will remain equally robust if some particles (but not too many)
cannot change their size at all. And the size at which a bacterium
divides needs not be identical for all members of the population;
the homeostasis will remain equally robust if some bacteria (but
not too many) cannot divide at all. The same is true for both
factors combined.

Of course the behavior of a population is not 100% robust either.
The process is, after all, determined by chance at the most basic
level, and it is theoretically possible that _all_ bacteria have
escaped at a certain moment. But this is unlikely in the extreme.
In practice, a population will disappear only after a major cata-
strophe such as exhaustion of the food source. In my size-change
example this results in a flat concentration gradient and a random
Brownian motion of all particles into the far reaches of infinity
-- maybe the "best" survival mechanism at that time, especially if
bacteria can encapsulate themselves so that they can survive long
periods of famine.

The proposed mechanisms are still simple enough to build a simu-
lation that can easily verify these results. Moreover, one can
make the simulation even more life-like if one allows the descen-
dants of a bacterium to be slightly different from its parent in
the parameters that describe its behavior (size vs. concentration,
size where division occurs). One can then also study the type of
evolutionary optimization ("adaptation") that evolution theory is
concerned with, noticing how the robustness of the population
"control" process is bound to increase as an automatic result.
Those who are familiar with genetic algorithms as optimization
methods will be acquainted with this phenomenon.

So what is the link with the highest level of the human hierarchy
of needs? First: that highest need must have originated from a
combination of the basic laws of physics, similar to -- but in a
much more complex fashion than -- the above example. And the best
current theory of how this came about is evolution theory, much as
Hawkins describes it in "The Selfish Gene", although the picture
that he portrays may be too simplistic when one takes the concept
of "genes" (as basic units) too literally.

Second: the behavior of each individual hardly matters. You and I
may have no children, but there are enough children _in the popu-
lation_ to guarantee the survival of the species. However, given
the extensive optimalizations that are possible in the genetic
"learning process" of very complex systems, you may "accidentally"
happen to find yourself in a position where you, too, somehow con-
tribute to the survival of the species. Evolution theory, as it is
often taught, places far too much emphasis on procreation as the
sole mechanism to ensure the survival into next generation. Pro-
creation is required, of course, to _obtain_ a new generation in
the first place, but mechanisms to ensure the _survival_ of the
next generation into their maturity must exist as well, and they
may be more subtle and more difficult to recognize. Grandparents
who do not reproduce anymore may contribute to the survival of
their children and grandchildren by contributing some of their
(material or mental) resources. Indeed, a complete stranger will
usually not be uninterested in -- if not genuinely concerned with
-- the welfare of his fellows, especially the young, in however
unobtruse or unconscious a manner.

Consider: if _all_ humans decided -- or found themselves in the
circumstances -- not to have children, humanity would stop to
exist. This is extremely unlikely, given the optimizations that
the evolutionary process has enforced and stored into our genes.
Although it is true that more than 99% of all species that have
ever existed are not there anymore, the continued existence of a
species is a remarkably robust process that may continue for a
great many millennea, as long as no major environmental cata-
strophes (including significant evolutionary "advances") occur.

You are right, however, that the basic forces of the evolutionary
process are so mild that they allow great latitude to each indivi-
dual, especially in these modern times where the new phenomenon of
voluntary birth control exists for the first time and in our West-
ern societies where the role of children in the support of society
is minimal compared to the villages of the poorer countries of the
world. But such a latitude exists at all levels and for all goals:
there are a great many different ways in which most goals can be
realized (how many ways are there to open a door? to go from A to
B? to ensure your nutrition? to contribute to the survival of our
species?). And that, again, has to do with the notion of robust-
ness itself: allowing a goal to be reached or maintained regard-
less of the great complexity and unpredictability of the environ-
ment is a lot more reliable if we have multiple action patterns
available. The more methods one has to realize a certain goal and
the more flexibly one can apply them, the more robust one's over-
all "control system". So it seems that the latitude that we indi-
viduals have is not a bug but a feature...

Third: the highest level goals viewed from the perspective _of the
population_ may be hard to recognize from the perspective _of the
individual_ because the mechanisms to achieve those goals allow so
much latitude, are so robust, have so low gain, and are therefore
so difficult to discover and to explore. They may resemble equili-
brium processes such as in my Brownian motion simulation much more
than high gain control processes that are so easy to recognize.

Greetings,

Hans

ยทยทยท

================================================================
Eindhoven University of Technology Eindhoven, the Netherlands
Dept. of Electrical Engineering Medical Engineering Group
email: j.a.blom@ele.tue.nl

Great man achieves harmony by maintaining differences; small man
achieves harmony by maintaining the commonality. Confucius