[ Follow Ups ]
Posted by John Brady on October 28, 1999 at 17:51:50:
The Evolution of Pattern-Pools
Below I present the hypothesis that gene-pools have evolved a mechanism for modifying the distribution of trait occurrences that would otherwise occur due only to the evolution of individual organisms, giving a simple and compelling explanation for altruism. Also below are ideas for testing this hypothesis according to scientific method. Also below is speculation on the evolution of different methods of reproduction during the earliest origins of life, explaining death, mutation rates, throw-back rates.
From preliminary research I suspect these are new ideas, and could be big. But my main trade is computing and maths, so I have written this exposition for your forum to consider. I would appreciate it if someone could tick off each assertion below as validated/sub-judice/wrong/unclear. If this long and uneducated post adds nothing new to your discipline, then please forgive my interruption.
Explanation of 'Pattern-Pool Evolution':
I'll start with a heroism example. For a tribe of baboons living in a particular environmental niche, optimally the tribe needs n1 heroic males who will risk their lives to defend the tribe, and n2 coward males who will live to breed another day. Optimal n1 and optimal n2 don't vary from year to year just because the previous generation of males had a lot of bad luck with cheetahs last year - the tribe will be best off if n1 heroes are born, even if last year most of the heroes died without progeny. Ordinary evolution (which I will distinguish as 'organism evolution') predicts that a trait such as heroism is more likely to occur if one or more parents has it, less likely if only grandparents had it, and even less likely if it hasn't occurred for a while (although Dawkins points out that 'introns' explain why throwbacks can occur even if a trait is missing for many generations). Thus if only 'organism evolution' is at work, a tribe that has a bad year with cheetahs will have a sub-optimum number of heroes - bad news for the tribe! Wouldn't it be handy if that tribe ('pattern-pool') tended to produce the same optimal number of heroes regardless of how unlucky they were with cheetahs the previous year? In fact, it would be an adaptive advantage for that tribe compared to other tribes: hence 'pattern-pool evolution'.
What I'm suggesting is that pattern-pools:
* have evolved some mechanism for *modifying the distribution of trait
occurrences that would otherwise occur due only to 'organism evolution'* (which
I will call 'distribution modifications'),
* that these 'distribution modifications' arise randomly but tend to re-occur
as long as that pattern-pool lasts,
* that adaptive 'distribution modifications' will persist, and irrelevant or
mal-adaptive distribution modifications will be bred out.
When you consider the complexity / variety / order / chaos of genes and
ecologies, and the aeons of time available, I think that the evolved existence
of and continuing evolving refinement of such 'distribution modifications' are
no more remarkable than any other wonder of evolution.
These 'distribution modifications' would only be adaptive if they modify 'altruistic' traits. An 'altruistic trait' is here defined as any trait that is adaptive for the pattern-pool but mal-adaptive for the individual organism. Such altruistic adaptive traits include death, heroism, mutations, throw-backs and ageing (explanations follow, really). (There is no advantage in 'distribution modifications' of non-altruistic traits that can be more easily managed by 'organism evolution'.)
Although I talk about a mechanism for 'distribution modifications', my area
isn't biology, and the mechanism is a black box for me. But some thoughts:
* The mechanism for heroism and other altruistic traits would have evolved with
the first social organisms (jelly-fish?).
* The mechanisms for guaranteeing death would have evolved as an essential part
of sexual reproduction (see 'Kingdom Evolution' below). The fine tuning of just
*when* to kill a organism would have evolved along with the species.
* The mechanisms for mutations and throw-backs probably evolved even before
sexual reproduction.
My first guess is that these mechanisms are likely to be independent of genes per
se. Hence I've used the term 'pattern-pool', but if it doesn't cause
controversy then the term 'gene-pool evolution' is preferable. (The mechanism
for heroism 'distribution modifications' could be very hard to observe, just as
it's very hard to understand a computer neural network by examining its
component parts.)
I'm not talking about 'species evolution'. E.g. the optimal ratio of heroes to cowards may vary depending on whether a particular species of baboon lives in a drier or wetter niche. Nor is the unit of selection a tribe - obviously neighbouring tribes interbreed. The pattern-pool is a continuum spread over a species range, but this doesn't prevent evolution of adaptive features of the pattern-pool. (Incidentally, 'species evolution' could be considered a special case of 'pattern-pool evolution'.)
'Altruistic' Behaviours:
So far I've discussed 'pattern-pool' evolution using the altruistic trait of heroism as an example. Below are some other examples.
1: "It take all kinds to make the world go around". For each environmental niche, there is an optimal ratio of cowards vs. heroes, innovators vs. plodders, greedies vs. sharers. If a war kills most of the heroes and innovators, this 'pattern-pool evolution' predicts they would reappear in the next generation with the same frequency, but that many repeated disasters would eventually lower the percentage of heroes in a population.
2: Mutation Research & Development Budget: A corporation that spends too little or too much on R&D will go bust. Similarly, for each species in an environmental niche, there is an optimal percentage of energy that a pattern-pool should invest in 'Genetic Experimentation' vs. keeping its organisms alive and occupying their territorial slice of the niche. ('Genetic Experimentation' is increased by: dying younger, having more progeny, giving better care of progeny, having more mutations / throwbacks / other gene-jumbles.) Thus in a rainforest a plant would invest more energy into reproduction and even haphazard mutations / throw-backs / gene cross-overs, whereas in a colder climate a plant puts more emphasis on not losing its share of the canopy. A plant that produces mutant offspring (which could conceivably be *the* successful evolutionary branch for that gene-pool, but which are most likely un-viable) is behaving adaptively for its pattern-pool but mal-adaptively for itself. (Incidentally, a pattern-pool that doesn't ensure adequate diversity is vulnerable to environmental change; how are sequoias faring?)
3: "Death is adaptive". An essential parameter that a pattern-pool
uses to regulate its R&D budget is death. Pattern-pools have an adaptive
advantage if old organisms die at the optimal age (to make way for the fittest
new genetic experiments). Here are some criteria for the optimal time to kill
an organism:
* Give all individual organisms the same opportunity for reproduction: if they
have lived through sustained lean times then let them live longer; if they have
been able to grow fat (and thus have had ample *opportunity* for reproduction)
then kill them earlier to give the next generation a fair chance using the
niche.
* For monogamous animals, rapidly kill the remaining spouse on the death of a
mate, unless it's good at food gathering / child care.
4: Menopause: I'm not sure how easy it is to explain menopause (which is unique to humans?) with just 'organism evolution'. Is menopause adaptive for the organism because the orphans of elderly mums make it so much harder for mum's other relatives to survive? If you allow pattern-pool evolution, menopause is easy to explain.
5: Anti-wrinkle creams. A pattern-pool whose organisms are able to determine how many breeding years there are left remaining in an prospective mate has an adaptive advantage, although organisms who could breed but miss out on a potentially attractive mate because they are 'showing their age' have an individual disadvantage. This has particular application for humans, given our unique menopause, but should be true for all animals in which both parents rear the young.
6: 'Kingdom Evolution: Sexual Organisms Must Die. Back in the primordial soup, sexual reproduction evolved as a way of better-directed genetic experimentation, especially for more complex organisms. However, sexual reproduction is useless if immortal parents stay around to hog the niche. I speculate that back in the primordial soup there may have been several phyla (dead-end kingdoms, actually) that tried sex without mortality; these phyla might have extended to occupy most of the planet, before being over-run by a sexual phyla whose parents were considerate enough to die. I speculate that there may have been many dead-end kingdoms the failed because some of their organisms reverted to immortality, and that eventually a kingdom evolved (ours) that had so many fool-proof time-bombs embedded in each organism that none could live forever. This special case of 'pattern-pool evolution' could be called 'kingdom evolution'. Of course, modern technology may be able to disable many or all of those time-bombs, but perhaps shares in corporations seeking the elixir of youth are priced too high. 'Kingdom Evolution' could also explain mutation rates, throwbacks, and other gene-jumbling aspects of reproduction, unless you prefer to say "But that's just the way things are."
Testing the hypothesis of 'Pattern-pool evolution':
Here is one possible experiment:
1. Develop the equivalent of a Myers-Briggs personality trait questionnaire for
baboons. This isn't quite so stupid as it sounds. Psychologists have come some
way in the science of identifying a minimal set of attributes (i.e.
"traits" in their jargon) that maximise predictions of what a person
will do. Psychologists also have effective methods of figuring out which questions
have useful correlations in indicating one or more traits. Rather than teach a
baboon how to answer multiple-choice questions, develop questions like
"Does the baboon move towards or away from a photo of a cheetah?"
2. Test the baboons.
3. Remove all the heroes baboons before they breed, or even teach. (In a
seperate experiment, remove all the generous sharing baboons.)
4. Test subsequent generations. Does the distribution of heroes / cowards /
greedies / sharers in the next and subsequent generations match organism-only
evolution, or is the distribution significantly skewed towards altruism?
Here is another (somewhat inconclusive) prediction about the nature of altruistic social traits: they will tend to be either/or characteristics with a double humped distribution curve (like human testes size), rather than continuum characteristics with a single bell curve (such as skin colour). This would allow the pattern-pool to easily switch them on or off as required.
Currently gerontologists have all sorts of reasons why we age, why we die when we die, but 'pattern-pool evolution' offers just one (and claims the gerontologists' reasons as mere mechanisms :-) Perhaps there is a parallel between all the ancients' epicycles and the Newton's law of gravitational attraction? I hope so, because I can't think of many testable predictions in this area. (Although I predicted that thin rats would live longer before I heard on TV that they indeed do :-) But here's a prediction: all animals in which both parents rear the young will have a way of determining if a potential mate is likely to die during the period of rearing. Species who mate for life will be able to guess the life expectancy of a potential mate (as evidenced by mating choices).
I predict that pattern-pool evolution provides a single explanation for the rates of mutations / throw-backs / other gene-jumbles that are typically mal-adaptive for the organism.
That's it for one night: over to you folk.
Perhaps the most compelling reason for accepting the hypothesis of pattern-pool
evolution is to recognise the substantial advantage it would confer to a
gene-pool, and the likelihood that the complexity of reproduction offers ample
opportunity for evolving a mechanism for modifying the distribution of trait
occurrences that would otherwise occur due only to 'organism evolution'.
However, my speculations on 'kingdom evolution' must be as contentious as anything else on the origins of life.
Thanks for your attention.
P.S: An article on 'Altruism in Lions' (in the Sydney Morning Herald 21/1/96)
analysed the logic of altruism as though lions were capable of a syllogism.
Pulease! That is excessively academic-centric. Try teaching adult education to
realise that even intelligent adults may only learn logic through formal education.
Five year old children are often not clever enough to notice selfish behaviour
in others, and I doubt many lions have the IQ of a five year old human. My
point is that in animals (including even humans up until civilisation), actions
of cowardice or bravery are most likely determined by instinct modified by
inductive experience, and are certainly not determined by rational cogitation.
P.P.S: Many people have an emotional problem
when first considering socio-biology. The wide scope of human endeavour that it
pronounces on can overwhelm with nihilistic determinism. Given that this idea
of pattern-pool evolution makes deterministic predictions on different forms of
altruism, perhaps a cheery little reminder on the usefulness of free-will is in
order:
There is an easy solution to the apparent paradox between the concept of free
will and a deterministic universe (although I'm not sure if Particle Physicists
currently think the universe is deterministic). Determinism is feasibly useful
if you want to predict the path in molecules in a very small cup of tea for the
next few milli-seconds. But to predict exactly what I'll do with my life, I'd
need a parallel universe that is an exact copy of this one but which runs
faster... In other words, for all practical intents and purposes, I must use
the concept of free will. Although one might conceive of an intelligent being
that has absolutely no concept of free will, adding a 'free-will program
module' to the brain with confers a major adaptive advantage. So, child of
nature moulded by education, your life and your society is there for you to
shape to the best of your abilities - On yer bike!